From ea75deef1a738d25502cfbb2caa564270b271525 Mon Sep 17 00:00:00 2001 From: Richard Fitzgerald Date: Mon, 15 Aug 2022 13:31:38 +0100 Subject: [PATCH 001/158] ASoC: cs42l42: Only report button state if there was a button interrupt Only report a button state change if the interrupt status shows that there was a button event. Previously the code would always drop into the button reporting at the end of interrupt handling if the jack was present. If neither of the button report interrupts were pending it would report all buttons released. This could then lead to a button being reported as released while it is still pressed. Fixes: c5b8ee0879bc ("ASoC: cs42l42: Report jack and button detection") Signed-off-by: Richard Fitzgerald Link: https://lore.kernel.org/r/20220815123138.3810249-1-rf@opensource.cirrus.com Signed-off-by: Mark Brown --- sound/soc/codecs/cs42l42.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/sound/soc/codecs/cs42l42.c b/sound/soc/codecs/cs42l42.c index d545a593a251..daafd4251ce6 100644 --- a/sound/soc/codecs/cs42l42.c +++ b/sound/soc/codecs/cs42l42.c @@ -1617,7 +1617,6 @@ static irqreturn_t cs42l42_irq_thread(int irq, void *data) unsigned int current_plug_status; unsigned int current_button_status; unsigned int i; - int report = 0; mutex_lock(&cs42l42->irq_lock); if (cs42l42->suspended) { @@ -1711,13 +1710,15 @@ static irqreturn_t cs42l42_irq_thread(int irq, void *data) if (current_button_status & CS42L42_M_DETECT_TF_MASK) { dev_dbg(cs42l42->dev, "Button released\n"); - report = 0; + snd_soc_jack_report(cs42l42->jack, 0, + SND_JACK_BTN_0 | SND_JACK_BTN_1 | + SND_JACK_BTN_2 | SND_JACK_BTN_3); } else if (current_button_status & CS42L42_M_DETECT_FT_MASK) { - report = cs42l42_handle_button_press(cs42l42); - + snd_soc_jack_report(cs42l42->jack, + cs42l42_handle_button_press(cs42l42), + SND_JACK_BTN_0 | SND_JACK_BTN_1 | + SND_JACK_BTN_2 | SND_JACK_BTN_3); } - snd_soc_jack_report(cs42l42->jack, report, SND_JACK_BTN_0 | SND_JACK_BTN_1 | - SND_JACK_BTN_2 | SND_JACK_BTN_3); } } From dcdfa3471f9c28ee716c687d85701353e2e86fde Mon Sep 17 00:00:00 2001 From: Pieterjan Camerlynck Date: Sat, 13 Aug 2022 10:33:52 +0200 Subject: [PATCH 002/158] ASoC: fsl_sai: fix incorrect mclk number in error message In commit c3ecef21c3f26 ("ASoC: fsl_sai: add sai master mode support") the loop was changed to start iterating from 1 instead of 0. The error message however was not updated, reporting the wrong clock to the user. Signed-off-by: Pieterjan Camerlynck Acked-by: Shengjiu Wang Link: https://lore.kernel.org/r/20220813083353.8959-1-pieterjan.camerlynck@gmail.com Signed-off-by: Mark Brown --- sound/soc/fsl/fsl_sai.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/fsl/fsl_sai.c b/sound/soc/fsl/fsl_sai.c index 7523bb944b21..d430eece1d6b 100644 --- a/sound/soc/fsl/fsl_sai.c +++ b/sound/soc/fsl/fsl_sai.c @@ -1306,7 +1306,7 @@ static int fsl_sai_probe(struct platform_device *pdev) sai->mclk_clk[i] = devm_clk_get(dev, tmp); if (IS_ERR(sai->mclk_clk[i])) { dev_err(dev, "failed to get mclk%d clock: %ld\n", - i + 1, PTR_ERR(sai->mclk_clk[i])); + i, PTR_ERR(sai->mclk_clk[i])); sai->mclk_clk[i] = NULL; } } From c6e14bb9f50df7126ca64405ae807d8bc7b39f9a Mon Sep 17 00:00:00 2001 From: Srinivas Kandagatla Date: Tue, 16 Aug 2022 17:52:29 +0100 Subject: [PATCH 003/158] ASoC: qcom: sm8250: add missing module owner Add missing module owner to able to build and load this driver as module. Fixes: aa2e2785545a ("ASoC: qcom: sm8250: add sound card qrb5165-rb5 support") Signed-off-by: Srinivas Kandagatla Link: https://lore.kernel.org/r/20220816165229.7971-1-srinivas.kandagatla@linaro.org Signed-off-by: Mark Brown --- sound/soc/qcom/sm8250.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/qcom/sm8250.c b/sound/soc/qcom/sm8250.c index ce4a5713386a..98a2fde9e004 100644 --- a/sound/soc/qcom/sm8250.c +++ b/sound/soc/qcom/sm8250.c @@ -270,6 +270,7 @@ static int sm8250_platform_probe(struct platform_device *pdev) if (!card) return -ENOMEM; + card->owner = THIS_MODULE; /* Allocate the private data */ data = devm_kzalloc(dev, sizeof(*data), GFP_KERNEL); if (!data) From ecdb10df7e0d83bfd12fb8f71e28ea4753e3716a Mon Sep 17 00:00:00 2001 From: Yang Yingliang Date: Thu, 18 Aug 2022 16:17:51 +0800 Subject: [PATCH 004/158] ASoC: SOF: ipc4-topology: fix wrong use of sizeof in sof_ipc4_widget_setup_comp_src() It should be size of the struct sof_ipc4_src, not data pointer pass to sof_update_ipc_object(). Fixes: b85f4fc40d56 ("ASoC: SOF: add ipc4 SRC module support") Signed-off-by: Yang Yingliang Acked-by: Peter Ujfalusi Link: https://lore.kernel.org/r/20220818081751.2407066-1-yangyingliang@huawei.com Signed-off-by: Mark Brown --- sound/soc/sof/ipc4-topology.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/sof/ipc4-topology.c b/sound/soc/sof/ipc4-topology.c index af072b484a60..c6abfaf5d532 100644 --- a/sound/soc/sof/ipc4-topology.c +++ b/sound/soc/sof/ipc4-topology.c @@ -771,7 +771,7 @@ static int sof_ipc4_widget_setup_comp_src(struct snd_sof_widget *swidget) goto err; ret = sof_update_ipc_object(scomp, src, SOF_SRC_TOKENS, swidget->tuples, - swidget->num_tuples, sizeof(src), 1); + swidget->num_tuples, sizeof(*src), 1); if (ret) { dev_err(scomp->dev, "Parsing SRC tokens failed\n"); goto err; From b66905e04dc714825aa6cffb950e281b46bbeafe Mon Sep 17 00:00:00 2001 From: Jack Wang Date: Thu, 18 Aug 2022 12:53:54 +0200 Subject: [PATCH 005/158] RDMA/rtrs-clt: Use the right sg_cnt after ib_dma_map_sg When iommu is enabled, we hit warnings like this: WARNING: at rtrs/rtrs.c:178 rtrs_iu_post_rdma_write_imm+0x9b/0x110 rtrs warn on one sge entry length is 0, which is unexpected. The problem is ib_dma_map_sg augments the SGL into a 'dma mapped SGL'. This process may change the number of entries and the lengths of each entry. Code that touches dma_address is iterating over the 'dma mapped SGL' and must use dma_nents which returned from ib_dma_map_sg(). So pass the count return from ib_dma_map_sg. Fixes: 6a98d71daea1 ("RDMA/rtrs: client: main functionality") Link: https://lore.kernel.org/r/20220818105355.110344-3-haris.iqbal@ionos.com Signed-off-by: Jack Wang Reviewed-by: Aleksei Marov Signed-off-by: Leon Romanovsky --- drivers/infiniband/ulp/rtrs/rtrs-clt.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/ulp/rtrs/rtrs-clt.c b/drivers/infiniband/ulp/rtrs/rtrs-clt.c index baecde41d126..449904dac0a9 100644 --- a/drivers/infiniband/ulp/rtrs/rtrs-clt.c +++ b/drivers/infiniband/ulp/rtrs/rtrs-clt.c @@ -1004,7 +1004,8 @@ rtrs_clt_get_copy_req(struct rtrs_clt_path *alive_path, static int rtrs_post_rdma_write_sg(struct rtrs_clt_con *con, struct rtrs_clt_io_req *req, struct rtrs_rbuf *rbuf, bool fr_en, - u32 size, u32 imm, struct ib_send_wr *wr, + u32 count, u32 size, u32 imm, + struct ib_send_wr *wr, struct ib_send_wr *tail) { struct rtrs_clt_path *clt_path = to_clt_path(con->c.path); @@ -1024,12 +1025,12 @@ static int rtrs_post_rdma_write_sg(struct rtrs_clt_con *con, num_sge = 2; ptail = tail; } else { - for_each_sg(req->sglist, sg, req->sg_cnt, i) { + for_each_sg(req->sglist, sg, count, i) { sge[i].addr = sg_dma_address(sg); sge[i].length = sg_dma_len(sg); sge[i].lkey = clt_path->s.dev->ib_pd->local_dma_lkey; } - num_sge = 1 + req->sg_cnt; + num_sge = 1 + count; } sge[i].addr = req->iu->dma_addr; sge[i].length = size; @@ -1142,7 +1143,7 @@ static int rtrs_clt_write_req(struct rtrs_clt_io_req *req) */ rtrs_clt_update_all_stats(req, WRITE); - ret = rtrs_post_rdma_write_sg(req->con, req, rbuf, fr_en, + ret = rtrs_post_rdma_write_sg(req->con, req, rbuf, fr_en, count, req->usr_len + sizeof(*msg), imm, wr, &inv_wr); if (ret) { From 56c310de0b4b3aca1c4fdd9c1093fc48372a7335 Mon Sep 17 00:00:00 2001 From: Jack Wang Date: Thu, 18 Aug 2022 12:53:55 +0200 Subject: [PATCH 006/158] RDMA/rtrs-srv: Pass the correct number of entries for dma mapped SGL ib_dma_map_sg() augments the SGL into a 'dma mapped SGL'. This process may change the number of entries and the lengths of each entry. Code that touches dma_address is iterating over the 'dma mapped SGL' and must use dma_nents which returned from ib_dma_map_sg(). We should use the return count from ib_dma_map_sg for futher usage. Fixes: 9cb837480424e ("RDMA/rtrs: server: main functionality") Link: https://lore.kernel.org/r/20220818105355.110344-4-haris.iqbal@ionos.com Signed-off-by: Jack Wang Reviewed-by: Aleksei Marov Signed-off-by: Leon Romanovsky --- drivers/infiniband/ulp/rtrs/rtrs-srv.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/infiniband/ulp/rtrs/rtrs-srv.c b/drivers/infiniband/ulp/rtrs/rtrs-srv.c index 34c03bde5064..4894e7329d88 100644 --- a/drivers/infiniband/ulp/rtrs/rtrs-srv.c +++ b/drivers/infiniband/ulp/rtrs/rtrs-srv.c @@ -595,7 +595,7 @@ static int map_cont_bufs(struct rtrs_srv_path *srv_path) struct sg_table *sgt = &srv_mr->sgt; struct scatterlist *s; struct ib_mr *mr; - int nr, chunks; + int nr, nr_sgt, chunks; chunks = chunks_per_mr * mri; if (!always_invalidate) @@ -610,19 +610,19 @@ static int map_cont_bufs(struct rtrs_srv_path *srv_path) sg_set_page(s, srv->chunks[chunks + i], max_chunk_size, 0); - nr = ib_dma_map_sg(srv_path->s.dev->ib_dev, sgt->sgl, + nr_sgt = ib_dma_map_sg(srv_path->s.dev->ib_dev, sgt->sgl, sgt->nents, DMA_BIDIRECTIONAL); - if (nr < sgt->nents) { - err = nr < 0 ? nr : -EINVAL; + if (!nr_sgt) { + err = -EINVAL; goto free_sg; } mr = ib_alloc_mr(srv_path->s.dev->ib_pd, IB_MR_TYPE_MEM_REG, - sgt->nents); + nr_sgt); if (IS_ERR(mr)) { err = PTR_ERR(mr); goto unmap_sg; } - nr = ib_map_mr_sg(mr, sgt->sgl, sgt->nents, + nr = ib_map_mr_sg(mr, sgt->sgl, nr_sgt, NULL, max_chunk_size); if (nr < 0 || nr < sgt->nents) { err = nr < 0 ? nr : -EINVAL; @@ -641,7 +641,7 @@ static int map_cont_bufs(struct rtrs_srv_path *srv_path) } } /* Eventually dma addr for each chunk can be cached */ - for_each_sg(sgt->sgl, s, sgt->orig_nents, i) + for_each_sg(sgt->sgl, s, nr_sgt, i) srv_path->dma_addr[chunks + i] = sg_dma_address(s); ib_update_fast_reg_key(mr, ib_inc_rkey(mr->rkey)); From 221ab1f0bf46236cf1a3fef5298ff5894acfb0c5 Mon Sep 17 00:00:00 2001 From: Jiaxin Yu Date: Sat, 20 Aug 2022 15:19:25 +0800 Subject: [PATCH 007/158] ASoC: mediatek: mt8186: fix DMIC record noise When the first DMIC recording is power down, mtkaif_dmic will be reset. This will cause configuration error in the second DMIC recording. So do not reset mtkaif_dmic except in "MTKAIF_DMIC Switch" kcontrol. Signed-off-by: Jiaxin Yu Link: https://lore.kernel.org/r/20220820071925.13557-1-jiaxin.yu@mediatek.com Signed-off-by: Mark Brown --- sound/soc/mediatek/mt8186/mt8186-dai-adda.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/sound/soc/mediatek/mt8186/mt8186-dai-adda.c b/sound/soc/mediatek/mt8186/mt8186-dai-adda.c index 266704556f37..094402470dc2 100644 --- a/sound/soc/mediatek/mt8186/mt8186-dai-adda.c +++ b/sound/soc/mediatek/mt8186/mt8186-dai-adda.c @@ -271,9 +271,6 @@ static int mtk_adda_ul_event(struct snd_soc_dapm_widget *w, /* should delayed 1/fs(smallest is 8k) = 125us before afe off */ usleep_range(125, 135); mt8186_afe_gpio_request(afe->dev, false, MT8186_DAI_ADDA, 1); - - /* reset dmic */ - afe_priv->mtkaif_dmic = 0; break; default: break; From cf5071876baf995f8f98e86ef06f85a58feda63c Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 23 Aug 2022 10:09:56 +0200 Subject: [PATCH 008/158] ASoC: nau8821: Implement hw constraint for rates nau8821 driver restricts the sample rate with over sampling rate, but currently it barely bails out at hw_params with -EINVAL error (with a kernel message); this doesn't help for user-space to recognize which rate can be actually used. This patch introduces the proper hw constraint for adjusting the available range of the sample rate depending on the OSR setup, as well as some code cleanup, for improving the communication with user-space. Now applications can know the valid rate beforehand and reduces the rate appropriately without errors. Signed-off-by: Takashi Iwai Link: https://lore.kernel.org/r/20220823081000.2965-2-tiwai@suse.de Signed-off-by: Mark Brown --- sound/soc/codecs/nau8821.c | 66 +++++++++++++++++++++----------------- 1 file changed, 36 insertions(+), 30 deletions(-) diff --git a/sound/soc/codecs/nau8821.c b/sound/soc/codecs/nau8821.c index 2d21339932e6..4a72b94e8410 100644 --- a/sound/soc/codecs/nau8821.c +++ b/sound/soc/codecs/nau8821.c @@ -670,28 +670,40 @@ static const struct snd_soc_dapm_route nau8821_dapm_routes[] = { {"HPOR", NULL, "Class G"}, }; -static int nau8821_clock_check(struct nau8821 *nau8821, - int stream, int rate, int osr) +static const struct nau8821_osr_attr * +nau8821_get_osr(struct nau8821 *nau8821, int stream) { - int osrate = 0; + unsigned int osr; if (stream == SNDRV_PCM_STREAM_PLAYBACK) { + regmap_read(nau8821->regmap, NAU8821_R2C_DAC_CTRL1, &osr); + osr &= NAU8821_DAC_OVERSAMPLE_MASK; if (osr >= ARRAY_SIZE(osr_dac_sel)) - return -EINVAL; - osrate = osr_dac_sel[osr].osr; + return NULL; + return &osr_dac_sel[osr]; } else { + regmap_read(nau8821->regmap, NAU8821_R2B_ADC_RATE, &osr); + osr &= NAU8821_ADC_SYNC_DOWN_MASK; if (osr >= ARRAY_SIZE(osr_adc_sel)) - return -EINVAL; - osrate = osr_adc_sel[osr].osr; + return NULL; + return &osr_adc_sel[osr]; } +} - if (!osrate || rate * osrate > CLK_DA_AD_MAX) { - dev_err(nau8821->dev, - "exceed the maximum frequency of CLK_ADC or CLK_DAC"); +static int nau8821_dai_startup(struct snd_pcm_substream *substream, + struct snd_soc_dai *dai) +{ + struct snd_soc_component *component = dai->component; + struct nau8821 *nau8821 = snd_soc_component_get_drvdata(component); + const struct nau8821_osr_attr *osr; + + osr = nau8821_get_osr(nau8821, substream->stream); + if (!osr || !osr->osr) return -EINVAL; - } - return 0; + return snd_pcm_hw_constraint_minmax(substream->runtime, + SNDRV_PCM_HW_PARAM_RATE, + 0, CLK_DA_AD_MAX / osr->osr); } static int nau8821_hw_params(struct snd_pcm_substream *substream, @@ -699,7 +711,8 @@ static int nau8821_hw_params(struct snd_pcm_substream *substream, { struct snd_soc_component *component = dai->component; struct nau8821 *nau8821 = snd_soc_component_get_drvdata(component); - unsigned int val_len = 0, osr, ctrl_val, bclk_fs, clk_div; + unsigned int val_len = 0, ctrl_val, bclk_fs, clk_div; + const struct nau8821_osr_attr *osr; nau8821->fs = params_rate(params); /* CLK_DAC or CLK_ADC = OSR * FS @@ -708,27 +721,19 @@ static int nau8821_hw_params(struct snd_pcm_substream *substream, * values must be selected such that the maximum frequency is less * than 6.144 MHz. */ - if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) { - regmap_read(nau8821->regmap, NAU8821_R2C_DAC_CTRL1, &osr); - osr &= NAU8821_DAC_OVERSAMPLE_MASK; - if (nau8821_clock_check(nau8821, substream->stream, - nau8821->fs, osr)) { - return -EINVAL; - } + osr = nau8821_get_osr(nau8821, substream->stream); + if (!osr || !osr->osr) + return -EINVAL; + if (nau8821->fs * osr->osr > CLK_DA_AD_MAX) + return -EINVAL; + if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) regmap_update_bits(nau8821->regmap, NAU8821_R03_CLK_DIVIDER, NAU8821_CLK_DAC_SRC_MASK, - osr_dac_sel[osr].clk_src << NAU8821_CLK_DAC_SRC_SFT); - } else { - regmap_read(nau8821->regmap, NAU8821_R2B_ADC_RATE, &osr); - osr &= NAU8821_ADC_SYNC_DOWN_MASK; - if (nau8821_clock_check(nau8821, substream->stream, - nau8821->fs, osr)) { - return -EINVAL; - } + osr->clk_src << NAU8821_CLK_DAC_SRC_SFT); + else regmap_update_bits(nau8821->regmap, NAU8821_R03_CLK_DIVIDER, NAU8821_CLK_ADC_SRC_MASK, - osr_adc_sel[osr].clk_src << NAU8821_CLK_ADC_SRC_SFT); - } + osr->clk_src << NAU8821_CLK_ADC_SRC_SFT); /* make BCLK and LRC divde configuration if the codec as master. */ regmap_read(nau8821->regmap, NAU8821_R1D_I2S_PCM_CTRL2, &ctrl_val); @@ -843,6 +848,7 @@ static int nau8821_digital_mute(struct snd_soc_dai *dai, int mute, } static const struct snd_soc_dai_ops nau8821_dai_ops = { + .startup = nau8821_dai_startup, .hw_params = nau8821_hw_params, .set_fmt = nau8821_set_dai_fmt, .mute_stream = nau8821_digital_mute, From 5628560e90395d3812800a8e44a01c32ffa429ec Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 23 Aug 2022 10:09:57 +0200 Subject: [PATCH 009/158] ASoC: nau8824: Fix semaphore unbalance at error paths The semaphore of nau8824 wasn't properly unlocked at some error handling code paths, hence this may result in the unbalance (and potential lock-up). Fix them to handle the semaphore up properly. Signed-off-by: Takashi Iwai Link: https://lore.kernel.org/r/20220823081000.2965-3-tiwai@suse.de Signed-off-by: Mark Brown --- sound/soc/codecs/nau8824.c | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-) diff --git a/sound/soc/codecs/nau8824.c b/sound/soc/codecs/nau8824.c index ad54d70f7d8e..10bdfebe92d5 100644 --- a/sound/soc/codecs/nau8824.c +++ b/sound/soc/codecs/nau8824.c @@ -1043,6 +1043,7 @@ static int nau8824_hw_params(struct snd_pcm_substream *substream, struct snd_soc_component *component = dai->component; struct nau8824 *nau8824 = snd_soc_component_get_drvdata(component); unsigned int val_len = 0, osr, ctrl_val, bclk_fs, bclk_div; + int err = -EINVAL; nau8824_sema_acquire(nau8824, HZ); @@ -1059,7 +1060,7 @@ static int nau8824_hw_params(struct snd_pcm_substream *substream, osr &= NAU8824_DAC_OVERSAMPLE_MASK; if (nau8824_clock_check(nau8824, substream->stream, nau8824->fs, osr)) - return -EINVAL; + goto error; regmap_update_bits(nau8824->regmap, NAU8824_REG_CLK_DIVIDER, NAU8824_CLK_DAC_SRC_MASK, osr_dac_sel[osr].clk_src << NAU8824_CLK_DAC_SRC_SFT); @@ -1069,7 +1070,7 @@ static int nau8824_hw_params(struct snd_pcm_substream *substream, osr &= NAU8824_ADC_SYNC_DOWN_MASK; if (nau8824_clock_check(nau8824, substream->stream, nau8824->fs, osr)) - return -EINVAL; + goto error; regmap_update_bits(nau8824->regmap, NAU8824_REG_CLK_DIVIDER, NAU8824_CLK_ADC_SRC_MASK, osr_adc_sel[osr].clk_src << NAU8824_CLK_ADC_SRC_SFT); @@ -1090,7 +1091,7 @@ static int nau8824_hw_params(struct snd_pcm_substream *substream, else if (bclk_fs <= 256) bclk_div = 0; else - return -EINVAL; + goto error; regmap_update_bits(nau8824->regmap, NAU8824_REG_PORT0_I2S_PCM_CTRL_2, NAU8824_I2S_LRC_DIV_MASK | NAU8824_I2S_BLK_DIV_MASK, @@ -1111,15 +1112,17 @@ static int nau8824_hw_params(struct snd_pcm_substream *substream, val_len |= NAU8824_I2S_DL_32; break; default: - return -EINVAL; + goto error; } regmap_update_bits(nau8824->regmap, NAU8824_REG_PORT0_I2S_PCM_CTRL_1, NAU8824_I2S_DL_MASK, val_len); + err = 0; + error: nau8824_sema_release(nau8824); - return 0; + return err; } static int nau8824_set_fmt(struct snd_soc_dai *dai, unsigned int fmt) @@ -1128,8 +1131,6 @@ static int nau8824_set_fmt(struct snd_soc_dai *dai, unsigned int fmt) struct nau8824 *nau8824 = snd_soc_component_get_drvdata(component); unsigned int ctrl1_val = 0, ctrl2_val = 0; - nau8824_sema_acquire(nau8824, HZ); - switch (fmt & SND_SOC_DAIFMT_MASTER_MASK) { case SND_SOC_DAIFMT_CBM_CFM: ctrl2_val |= NAU8824_I2S_MS_MASTER; @@ -1171,6 +1172,8 @@ static int nau8824_set_fmt(struct snd_soc_dai *dai, unsigned int fmt) return -EINVAL; } + nau8824_sema_acquire(nau8824, HZ); + regmap_update_bits(nau8824->regmap, NAU8824_REG_PORT0_I2S_PCM_CTRL_1, NAU8824_I2S_DF_MASK | NAU8824_I2S_BP_MASK | NAU8824_I2S_PCMB_EN, ctrl1_val); From 92283c86260d8712b55f97eada13b3c8b2f469b2 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 23 Aug 2022 10:09:58 +0200 Subject: [PATCH 010/158] ASoC: nau8824: Implement hw constraint for rates nau8824 driver restricts the sample rate with over sampling rate, but currently it barely bails out at hw_params with -EINVAL error (with a kernel message); this doesn't help for user-space to recognize which rate can be actually used. This patch introduces the proper hw constraint for adjusting the available range of the sample rate depending on the OSR setup, as well as some code cleanup, for improving the communication with user-space. Now applications can know the valid rate beforehand and reduces the rate appropriately without errors. Signed-off-by: Takashi Iwai Link: https://lore.kernel.org/r/20220823081000.2965-4-tiwai@suse.de Signed-off-by: Mark Brown --- sound/soc/codecs/nau8824.c | 67 +++++++++++++++++++++----------------- 1 file changed, 38 insertions(+), 29 deletions(-) diff --git a/sound/soc/codecs/nau8824.c b/sound/soc/codecs/nau8824.c index 10bdfebe92d5..15596452ca37 100644 --- a/sound/soc/codecs/nau8824.c +++ b/sound/soc/codecs/nau8824.c @@ -1014,27 +1014,42 @@ static irqreturn_t nau8824_interrupt(int irq, void *data) return IRQ_HANDLED; } -static int nau8824_clock_check(struct nau8824 *nau8824, - int stream, int rate, int osr) +static const struct nau8824_osr_attr * +nau8824_get_osr(struct nau8824 *nau8824, int stream) { - int osrate; + unsigned int osr; if (stream == SNDRV_PCM_STREAM_PLAYBACK) { + regmap_read(nau8824->regmap, + NAU8824_REG_DAC_FILTER_CTRL_1, &osr); + osr &= NAU8824_DAC_OVERSAMPLE_MASK; if (osr >= ARRAY_SIZE(osr_dac_sel)) - return -EINVAL; - osrate = osr_dac_sel[osr].osr; + return NULL; + return &osr_dac_sel[osr]; } else { + regmap_read(nau8824->regmap, + NAU8824_REG_ADC_FILTER_CTRL, &osr); + osr &= NAU8824_ADC_SYNC_DOWN_MASK; if (osr >= ARRAY_SIZE(osr_adc_sel)) - return -EINVAL; - osrate = osr_adc_sel[osr].osr; + return NULL; + return &osr_adc_sel[osr]; } +} - if (!osrate || rate * osr > CLK_DA_AD_MAX) { - dev_err(nau8824->dev, "exceed the maximum frequency of CLK_ADC or CLK_DAC\n"); +static int nau8824_dai_startup(struct snd_pcm_substream *substream, + struct snd_soc_dai *dai) +{ + struct snd_soc_component *component = dai->component; + struct nau8824 *nau8824 = snd_soc_component_get_drvdata(component); + const struct nau8824_osr_attr *osr; + + osr = nau8824_get_osr(nau8824, substream->stream); + if (!osr || !osr->osr) return -EINVAL; - } - return 0; + return snd_pcm_hw_constraint_minmax(substream->runtime, + SNDRV_PCM_HW_PARAM_RATE, + 0, CLK_DA_AD_MAX / osr->osr); } static int nau8824_hw_params(struct snd_pcm_substream *substream, @@ -1042,7 +1057,8 @@ static int nau8824_hw_params(struct snd_pcm_substream *substream, { struct snd_soc_component *component = dai->component; struct nau8824 *nau8824 = snd_soc_component_get_drvdata(component); - unsigned int val_len = 0, osr, ctrl_val, bclk_fs, bclk_div; + unsigned int val_len = 0, ctrl_val, bclk_fs, bclk_div; + const struct nau8824_osr_attr *osr; int err = -EINVAL; nau8824_sema_acquire(nau8824, HZ); @@ -1054,27 +1070,19 @@ static int nau8824_hw_params(struct snd_pcm_substream *substream, * than 6.144 MHz. */ nau8824->fs = params_rate(params); - if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) { - regmap_read(nau8824->regmap, - NAU8824_REG_DAC_FILTER_CTRL_1, &osr); - osr &= NAU8824_DAC_OVERSAMPLE_MASK; - if (nau8824_clock_check(nau8824, substream->stream, - nau8824->fs, osr)) - goto error; + osr = nau8824_get_osr(nau8824, substream->stream); + if (!osr || !osr->osr) + goto error; + if (nau8824->fs * osr->osr > CLK_DA_AD_MAX) + goto error; + if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) regmap_update_bits(nau8824->regmap, NAU8824_REG_CLK_DIVIDER, NAU8824_CLK_DAC_SRC_MASK, - osr_dac_sel[osr].clk_src << NAU8824_CLK_DAC_SRC_SFT); - } else { - regmap_read(nau8824->regmap, - NAU8824_REG_ADC_FILTER_CTRL, &osr); - osr &= NAU8824_ADC_SYNC_DOWN_MASK; - if (nau8824_clock_check(nau8824, substream->stream, - nau8824->fs, osr)) - goto error; + osr->clk_src << NAU8824_CLK_DAC_SRC_SFT); + else regmap_update_bits(nau8824->regmap, NAU8824_REG_CLK_DIVIDER, NAU8824_CLK_ADC_SRC_MASK, - osr_adc_sel[osr].clk_src << NAU8824_CLK_ADC_SRC_SFT); - } + osr->clk_src << NAU8824_CLK_ADC_SRC_SFT); /* make BCLK and LRC divde configuration if the codec as master. */ regmap_read(nau8824->regmap, @@ -1550,6 +1558,7 @@ static const struct snd_soc_component_driver nau8824_component_driver = { }; static const struct snd_soc_dai_ops nau8824_dai_ops = { + .startup = nau8824_dai_startup, .hw_params = nau8824_hw_params, .set_fmt = nau8824_set_fmt, .set_tdm_slot = nau8824_set_tdm_slot, From bed41de0f679c516de45cfeb2c40c412bc5e0c0b Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 23 Aug 2022 10:09:59 +0200 Subject: [PATCH 011/158] ASoC: nau8825: Implement hw constraint for rates nau8825 driver restricts the sample rate with over sampling rate, but currently it barely bails out at hw_params with -EINVAL error (with a kernel message); this doesn't help for user-space to recognize which rate can be actually used. This patch introduces the proper hw constraint for adjusting the available range of the sample rate depending on the OSR setup, as well as some code cleanup, for improving the communication with user-space. Now applications can know the valid rate beforehand and reduces the rate appropriately without errors. Signed-off-by: Takashi Iwai Link: https://lore.kernel.org/r/20220823081000.2965-5-tiwai@suse.de Signed-off-by: Mark Brown --- sound/soc/codecs/nau8825.c | 83 +++++++++++++++++++++----------------- 1 file changed, 45 insertions(+), 38 deletions(-) diff --git a/sound/soc/codecs/nau8825.c b/sound/soc/codecs/nau8825.c index 54ef7b0fa878..8213273f501e 100644 --- a/sound/soc/codecs/nau8825.c +++ b/sound/soc/codecs/nau8825.c @@ -1247,27 +1247,42 @@ static const struct snd_soc_dapm_route nau8825_dapm_routes[] = { {"HPOR", NULL, "Class G"}, }; -static int nau8825_clock_check(struct nau8825 *nau8825, - int stream, int rate, int osr) +static const struct nau8825_osr_attr * +nau8825_get_osr(struct nau8825 *nau8825, int stream) { - int osrate; + unsigned int osr; if (stream == SNDRV_PCM_STREAM_PLAYBACK) { + regmap_read(nau8825->regmap, + NAU8825_REG_DAC_CTRL1, &osr); + osr &= NAU8825_DAC_OVERSAMPLE_MASK; if (osr >= ARRAY_SIZE(osr_dac_sel)) - return -EINVAL; - osrate = osr_dac_sel[osr].osr; + return NULL; + return &osr_dac_sel[osr]; } else { + regmap_read(nau8825->regmap, + NAU8825_REG_ADC_RATE, &osr); + osr &= NAU8825_ADC_SYNC_DOWN_MASK; if (osr >= ARRAY_SIZE(osr_adc_sel)) - return -EINVAL; - osrate = osr_adc_sel[osr].osr; + return NULL; + return &osr_adc_sel[osr]; } +} - if (!osrate || rate * osr > CLK_DA_AD_MAX) { - dev_err(nau8825->dev, "exceed the maximum frequency of CLK_ADC or CLK_DAC\n"); +static int nau8825_dai_startup(struct snd_pcm_substream *substream, + struct snd_soc_dai *dai) +{ + struct snd_soc_component *component = dai->component; + struct nau8825 *nau8825 = snd_soc_component_get_drvdata(component); + const struct nau8825_osr_attr *osr; + + osr = nau8825_get_osr(nau8825, substream->stream); + if (!osr || !osr->osr) return -EINVAL; - } - return 0; + return snd_pcm_hw_constraint_minmax(substream->runtime, + SNDRV_PCM_HW_PARAM_RATE, + 0, CLK_DA_AD_MAX / osr->osr); } static int nau8825_hw_params(struct snd_pcm_substream *substream, @@ -1276,7 +1291,9 @@ static int nau8825_hw_params(struct snd_pcm_substream *substream, { struct snd_soc_component *component = dai->component; struct nau8825 *nau8825 = snd_soc_component_get_drvdata(component); - unsigned int val_len = 0, osr, ctrl_val, bclk_fs, bclk_div; + unsigned int val_len = 0, ctrl_val, bclk_fs, bclk_div; + const struct nau8825_osr_attr *osr; + int err = -EINVAL; nau8825_sema_acquire(nau8825, 3 * HZ); @@ -1286,29 +1303,19 @@ static int nau8825_hw_params(struct snd_pcm_substream *substream, * values must be selected such that the maximum frequency is less * than 6.144 MHz. */ - if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) { - regmap_read(nau8825->regmap, NAU8825_REG_DAC_CTRL1, &osr); - osr &= NAU8825_DAC_OVERSAMPLE_MASK; - if (nau8825_clock_check(nau8825, substream->stream, - params_rate(params), osr)) { - nau8825_sema_release(nau8825); - return -EINVAL; - } + osr = nau8825_get_osr(nau8825, substream->stream); + if (!osr || !osr->osr) + goto error; + if (params_rate(params) * osr->osr > CLK_DA_AD_MAX) + goto error; + if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) regmap_update_bits(nau8825->regmap, NAU8825_REG_CLK_DIVIDER, NAU8825_CLK_DAC_SRC_MASK, - osr_dac_sel[osr].clk_src << NAU8825_CLK_DAC_SRC_SFT); - } else { - regmap_read(nau8825->regmap, NAU8825_REG_ADC_RATE, &osr); - osr &= NAU8825_ADC_SYNC_DOWN_MASK; - if (nau8825_clock_check(nau8825, substream->stream, - params_rate(params), osr)) { - nau8825_sema_release(nau8825); - return -EINVAL; - } + osr->clk_src << NAU8825_CLK_DAC_SRC_SFT); + else regmap_update_bits(nau8825->regmap, NAU8825_REG_CLK_DIVIDER, NAU8825_CLK_ADC_SRC_MASK, - osr_adc_sel[osr].clk_src << NAU8825_CLK_ADC_SRC_SFT); - } + osr->clk_src << NAU8825_CLK_ADC_SRC_SFT); /* make BCLK and LRC divde configuration if the codec as master. */ regmap_read(nau8825->regmap, NAU8825_REG_I2S_PCM_CTRL2, &ctrl_val); @@ -1321,10 +1328,8 @@ static int nau8825_hw_params(struct snd_pcm_substream *substream, bclk_div = 1; else if (bclk_fs <= 128) bclk_div = 0; - else { - nau8825_sema_release(nau8825); - return -EINVAL; - } + else + goto error; regmap_update_bits(nau8825->regmap, NAU8825_REG_I2S_PCM_CTRL2, NAU8825_I2S_LRC_DIV_MASK | NAU8825_I2S_BLK_DIV_MASK, ((bclk_div + 1) << NAU8825_I2S_LRC_DIV_SFT) | bclk_div); @@ -1344,17 +1349,18 @@ static int nau8825_hw_params(struct snd_pcm_substream *substream, val_len |= NAU8825_I2S_DL_32; break; default: - nau8825_sema_release(nau8825); - return -EINVAL; + goto error; } regmap_update_bits(nau8825->regmap, NAU8825_REG_I2S_PCM_CTRL1, NAU8825_I2S_DL_MASK, val_len); + err = 0; + error: /* Release the semaphore. */ nau8825_sema_release(nau8825); - return 0; + return err; } static int nau8825_set_dai_fmt(struct snd_soc_dai *codec_dai, unsigned int fmt) @@ -1420,6 +1426,7 @@ static int nau8825_set_dai_fmt(struct snd_soc_dai *codec_dai, unsigned int fmt) } static const struct snd_soc_dai_ops nau8825_dai_ops = { + .startup = nau8825_dai_startup, .hw_params = nau8825_hw_params, .set_fmt = nau8825_set_dai_fmt, }; From be919239fbcab19290bfd6802c7ad1dc946c515b Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 23 Aug 2022 10:10:00 +0200 Subject: [PATCH 012/158] ASoC: nau8540: Implement hw constraint for rates nau8540 driver restricts the sample rate with over sampling rate, but currently it barely bails out at hw_params with -EINVAL error (with a kernel message); this doesn't help for user-space to recognize which rate can be actually used. This patch introduces the proper hw constraint for adjusting the available range of the sample rate depending on the OSR setup, as well as some code cleanup, for improving the communication with user-space. Now applications can know the valid rate beforehand and reduces the rate appropriately without errors. Signed-off-by: Takashi Iwai Link: https://lore.kernel.org/r/20220823081000.2965-6-tiwai@suse.de Signed-off-by: Mark Brown --- sound/soc/codecs/nau8540.c | 42 +++++++++++++++++++++++++++----------- 1 file changed, 30 insertions(+), 12 deletions(-) diff --git a/sound/soc/codecs/nau8540.c b/sound/soc/codecs/nau8540.c index 58f70a02f18a..0626d5694c22 100644 --- a/sound/soc/codecs/nau8540.c +++ b/sound/soc/codecs/nau8540.c @@ -357,17 +357,32 @@ static const struct snd_soc_dapm_route nau8540_dapm_routes[] = { {"AIFTX", NULL, "Digital CH4 Mux"}, }; -static int nau8540_clock_check(struct nau8540 *nau8540, int rate, int osr) +static const struct nau8540_osr_attr * +nau8540_get_osr(struct nau8540 *nau8540) { + unsigned int osr; + + regmap_read(nau8540->regmap, NAU8540_REG_ADC_SAMPLE_RATE, &osr); + osr &= NAU8540_ADC_OSR_MASK; if (osr >= ARRAY_SIZE(osr_adc_sel)) + return NULL; + return &osr_adc_sel[osr]; +} + +static int nau8540_dai_startup(struct snd_pcm_substream *substream, + struct snd_soc_dai *dai) +{ + struct snd_soc_component *component = dai->component; + struct nau8540 *nau8540 = snd_soc_component_get_drvdata(component); + const struct nau8540_osr_attr *osr; + + osr = nau8540_get_osr(nau8540); + if (!osr || !osr->osr) return -EINVAL; - if (rate * osr > CLK_ADC_MAX) { - dev_err(nau8540->dev, "exceed the maximum frequency of CLK_ADC\n"); - return -EINVAL; - } - - return 0; + return snd_pcm_hw_constraint_minmax(substream->runtime, + SNDRV_PCM_HW_PARAM_RATE, + 0, CLK_ADC_MAX / osr->osr); } static int nau8540_hw_params(struct snd_pcm_substream *substream, @@ -375,7 +390,8 @@ static int nau8540_hw_params(struct snd_pcm_substream *substream, { struct snd_soc_component *component = dai->component; struct nau8540 *nau8540 = snd_soc_component_get_drvdata(component); - unsigned int val_len = 0, osr; + unsigned int val_len = 0; + const struct nau8540_osr_attr *osr; /* CLK_ADC = OSR * FS * ADC clock frequency is defined as Over Sampling Rate (OSR) @@ -383,13 +399,14 @@ static int nau8540_hw_params(struct snd_pcm_substream *substream, * values must be selected such that the maximum frequency is less * than 6.144 MHz. */ - regmap_read(nau8540->regmap, NAU8540_REG_ADC_SAMPLE_RATE, &osr); - osr &= NAU8540_ADC_OSR_MASK; - if (nau8540_clock_check(nau8540, params_rate(params), osr)) + osr = nau8540_get_osr(nau8540); + if (!osr || !osr->osr) + return -EINVAL; + if (params_rate(params) * osr->osr > CLK_ADC_MAX) return -EINVAL; regmap_update_bits(nau8540->regmap, NAU8540_REG_CLOCK_SRC, NAU8540_CLK_ADC_SRC_MASK, - osr_adc_sel[osr].clk_src << NAU8540_CLK_ADC_SRC_SFT); + osr->clk_src << NAU8540_CLK_ADC_SRC_SFT); switch (params_width(params)) { case 16: @@ -515,6 +532,7 @@ static int nau8540_set_tdm_slot(struct snd_soc_dai *dai, static const struct snd_soc_dai_ops nau8540_dai_ops = { + .startup = nau8540_dai_startup, .hw_params = nau8540_hw_params, .set_fmt = nau8540_set_fmt, .set_tdm_slot = nau8540_set_tdm_slot, From 1faa6f8274e2b08a38c0cca74113dfb26c6ad7b7 Mon Sep 17 00:00:00 2001 From: Shengjiu Wang Date: Tue, 23 Aug 2022 17:35:08 +0800 Subject: [PATCH 013/158] ASoC: fsl_mqs: Fix supported clock DAI format The MQS works as codec DAI, not cpu DAI. It is clock consumer, not clock privider. Fixes: 3b14c15a333b ("ASoC: fsl: Update to use set_fmt_new callback") Signed-off-by: Shengjiu Wang Link: https://lore.kernel.org/r/1661247308-2650-1-git-send-email-shengjiu.wang@nxp.com Signed-off-by: Mark Brown --- sound/soc/fsl/fsl_mqs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/fsl/fsl_mqs.c b/sound/soc/fsl/fsl_mqs.c index c1e2f671191b..4922e6795b73 100644 --- a/sound/soc/fsl/fsl_mqs.c +++ b/sound/soc/fsl/fsl_mqs.c @@ -122,7 +122,7 @@ static int fsl_mqs_set_dai_fmt(struct snd_soc_dai *dai, unsigned int fmt) } switch (fmt & SND_SOC_DAIFMT_CLOCK_PROVIDER_MASK) { - case SND_SOC_DAIFMT_BP_FP: + case SND_SOC_DAIFMT_CBC_CFC: break; default: return -EINVAL; From 3942499fba11de048c3ac1390b808e9e6ae88de5 Mon Sep 17 00:00:00 2001 From: Peter Ujfalusi Date: Tue, 23 Aug 2022 15:15:53 +0300 Subject: [PATCH 014/158] ASoC: SOF: Kconfig: Make IPC_FLOOD_TEST depend on SND_SOC_SOF Make sure that the IPC_FLOOD client can not be built in when SND_SOC_SOF is built as module. Fixes: 6e9548cdb30e5 ("ASoC: SOF: Convert the generic IPC flood test into SOF client") Reported-by: kernel test robot Signed-off-by: Peter Ujfalusi Reviewed-by: Pierre-Louis Bossart Link: https://lore.kernel.org/r/20220823121554.4255-2-peter.ujfalusi@linux.intel.com Signed-off-by: Mark Brown --- sound/soc/sof/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/sof/Kconfig b/sound/soc/sof/Kconfig index e90f173d067c..cfb244e4e142 100644 --- a/sound/soc/sof/Kconfig +++ b/sound/soc/sof/Kconfig @@ -196,6 +196,7 @@ config SND_SOC_SOF_DEBUG_ENABLE_FIRMWARE_TRACE config SND_SOC_SOF_DEBUG_IPC_FLOOD_TEST tristate "SOF enable IPC flood test" + depends on SND_SOC_SOF select SND_SOC_SOF_CLIENT help This option enables a separate client device for IPC flood test From 2cf520ffbcbd55c0f2b4276065444d7526d9d197 Mon Sep 17 00:00:00 2001 From: Peter Ujfalusi Date: Tue, 23 Aug 2022 15:15:54 +0300 Subject: [PATCH 015/158] ASoC: SOF: Kconfig: Make IPC_MESSAGE_INJECTOR depend on SND_SOC_SOF Make sure that the IPC_MESSAGE_INJECTOR client can not be built in when SND_SOC_SOF is built as module. Fixes: cac0b0887e530 ("ASoC: SOF: Convert the generic IPC message injector into SOF client") Reported-by: kernel test robot Signed-off-by: Peter Ujfalusi Reviewed-by: Pierre-Louis Bossart Link: https://lore.kernel.org/r/20220823121554.4255-3-peter.ujfalusi@linux.intel.com Signed-off-by: Mark Brown --- sound/soc/sof/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/sof/Kconfig b/sound/soc/sof/Kconfig index cfb244e4e142..37f7df5fde17 100644 --- a/sound/soc/sof/Kconfig +++ b/sound/soc/sof/Kconfig @@ -215,6 +215,7 @@ config SND_SOC_SOF_DEBUG_IPC_FLOOD_TEST_NUM config SND_SOC_SOF_DEBUG_IPC_MSG_INJECTOR tristate "SOF enable IPC message injector" + depends on SND_SOC_SOF select SND_SOC_SOF_CLIENT help This option enables the IPC message injector which can be used to send From 5c5c2baad2b55cc0a4b190266889959642298f79 Mon Sep 17 00:00:00 2001 From: Nathan Chancellor Date: Tue, 9 Aug 2022 18:08:09 -0700 Subject: [PATCH 016/158] ASoC: mchp-spdiftx: Fix clang -Wbitfield-constant-conversion A recent change in clang strengthened its -Wbitfield-constant-conversion to warn when 1 is assigned to a 1-bit signed integer bitfield, as it can only be 0 or -1, not 1: sound/soc/atmel/mchp-spdiftx.c:505:20: error: implicit truncation from 'int' to bit-field changes value from 1 to -1 [-Werror,-Wbitfield-constant-conversion] dev->gclk_enabled = 1; ^ ~ 1 error generated. The actual value of the field is never checked, just that it is not zero, so there is not a real bug here. However, it is simple enough to silence the warning by making the bitfield unsigned, which matches the mchp-spdifrx driver. Fixes: 06ca24e98e6b ("ASoC: mchp-spdiftx: add driver for S/PDIF TX Controller") Link: https://github.com/ClangBuiltLinux/linux/issues/1686 Link: https://github.com/llvm/llvm-project/commit/82afc9b169a67e8b8a1862fb9c41a2cd974d6691 Signed-off-by: Nathan Chancellor Reviewed-by: Nick Desaulniers Link: https://lore.kernel.org/r/20220810010809.2024482-1-nathan@kernel.org Signed-off-by: Mark Brown --- sound/soc/atmel/mchp-spdiftx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/atmel/mchp-spdiftx.c b/sound/soc/atmel/mchp-spdiftx.c index 4850a177803d..ab2d7a791f39 100644 --- a/sound/soc/atmel/mchp-spdiftx.c +++ b/sound/soc/atmel/mchp-spdiftx.c @@ -196,7 +196,7 @@ struct mchp_spdiftx_dev { struct clk *pclk; struct clk *gclk; unsigned int fmt; - int gclk_enabled:1; + unsigned int gclk_enabled:1; }; static inline int mchp_spdiftx_is_running(struct mchp_spdiftx_dev *dev) From 4ee6fc271b59e805301371ea3862f558a23d9c7b Mon Sep 17 00:00:00 2001 From: Bard Liao Date: Mon, 22 Aug 2022 21:02:11 +0200 Subject: [PATCH 017/158] ASoC: SOF: ipc4-topology: fix alh_group_ida max value group_id is from 0 ~ ALH_MULTI_GTW_COUNT - 1, not 0 ~ ALH_MULTI_GTW_COUNT. Fixes: a150345aa7584 ("ASoC: SOF: ipc4-topology: add SoundWire/ALH aggregation support") Reported-by: kernel test robot Reported-by: Dan Carpenter Reviewed-by: Rander Wang Signed-off-by: Bard Liao Signed-off-by: Pierre-Louis Bossart Link: https://lore.kernel.org/r/20220822190211.170537-1-pierre-louis.bossart@linux.intel.com Signed-off-by: Mark Brown --- sound/soc/sof/ipc4-topology.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/sof/ipc4-topology.c b/sound/soc/sof/ipc4-topology.c index c6abfaf5d532..64929dc9af39 100644 --- a/sound/soc/sof/ipc4-topology.c +++ b/sound/soc/sof/ipc4-topology.c @@ -1251,7 +1251,7 @@ sof_ipc4_prepare_copier_module(struct snd_sof_widget *swidget, if (blob->alh_cfg.count > 1) { int group_id; - group_id = ida_alloc_max(&alh_group_ida, ALH_MULTI_GTW_COUNT, + group_id = ida_alloc_max(&alh_group_ida, ALH_MULTI_GTW_COUNT - 1, GFP_KERNEL); if (group_id < 0) From ea532c29972df96fda20393d9bf057e898f5e965 Mon Sep 17 00:00:00 2001 From: Shengjiu Wang Date: Thu, 25 Aug 2022 20:27:39 +0800 Subject: [PATCH 018/158] ASoC: fsl_aud2htx: register platform component before registering cpu dai There is no defer probe when adding platform component to snd_soc_pcm_runtime(rtd), the code is in snd_soc_add_pcm_runtime() snd_soc_register_card() -> snd_soc_bind_card() -> snd_soc_add_pcm_runtime() -> adding cpu dai -> adding codec dai -> adding platform component. So if the platform component is not ready at that time, then the sound card still registered successfully, but platform component is empty, the sound card can't be used. As there is defer probe checking for cpu dai component, then register platform component before cpu dai to avoid such issue. And the behavior of imx_pcm_dma_init() is same as common devm_snd_dmaengine_pcm_register(), so use devm_snd_dmaengine_pcm_register() instead Signed-off-by: Shengjiu Wang Link: https://lore.kernel.org/r/1661430460-5234-1-git-send-email-shengjiu.wang@nxp.com Signed-off-by: Mark Brown --- sound/soc/fsl/fsl_aud2htx.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/sound/soc/fsl/fsl_aud2htx.c b/sound/soc/fsl/fsl_aud2htx.c index 873295f59ad7..bc1b9c6df95c 100644 --- a/sound/soc/fsl/fsl_aud2htx.c +++ b/sound/soc/fsl/fsl_aud2htx.c @@ -234,6 +234,16 @@ static int fsl_aud2htx_probe(struct platform_device *pdev) regcache_cache_only(aud2htx->regmap, true); + /* + * Register platform component before registering cpu dai for there + * is not defer probe for platform component in snd_soc_add_pcm_runtime(). + */ + ret = devm_snd_dmaengine_pcm_register(&pdev->dev, NULL, 0); + if (ret) { + dev_err(&pdev->dev, "failed to pcm register\n"); + return ret; + } + ret = devm_snd_soc_register_component(&pdev->dev, &fsl_aud2htx_component, &fsl_aud2htx_dai, 1); @@ -242,10 +252,6 @@ static int fsl_aud2htx_probe(struct platform_device *pdev) return ret; } - ret = imx_pcm_dma_init(pdev); - if (ret) - dev_err(&pdev->dev, "failed to init imx pcm dma: %d\n", ret); - return ret; } From b1cd3fd42db7593a2d24c06f1c53b8c886592080 Mon Sep 17 00:00:00 2001 From: Shengjiu Wang Date: Thu, 25 Aug 2022 20:27:40 +0800 Subject: [PATCH 019/158] ASoC: fsl_aud2htx: Add error handler for pm_runtime_enable Call pm_runtime_disable() when error happens in probe() Signed-off-by: Shengjiu Wang Link: https://lore.kernel.org/r/1661430460-5234-2-git-send-email-shengjiu.wang@nxp.com Signed-off-by: Mark Brown --- sound/soc/fsl/fsl_aud2htx.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/sound/soc/fsl/fsl_aud2htx.c b/sound/soc/fsl/fsl_aud2htx.c index bc1b9c6df95c..1e421d9a03fb 100644 --- a/sound/soc/fsl/fsl_aud2htx.c +++ b/sound/soc/fsl/fsl_aud2htx.c @@ -241,6 +241,7 @@ static int fsl_aud2htx_probe(struct platform_device *pdev) ret = devm_snd_dmaengine_pcm_register(&pdev->dev, NULL, 0); if (ret) { dev_err(&pdev->dev, "failed to pcm register\n"); + pm_runtime_disable(&pdev->dev); return ret; } @@ -249,6 +250,7 @@ static int fsl_aud2htx_probe(struct platform_device *pdev) &fsl_aud2htx_dai, 1); if (ret) { dev_err(&pdev->dev, "failed to register ASoC DAI\n"); + pm_runtime_disable(&pdev->dev); return ret; } From ead54ced6321099978d30d62dc49c282a6e70574 Mon Sep 17 00:00:00 2001 From: Shiraz Saleem Date: Wed, 24 Aug 2022 10:43:59 -0500 Subject: [PATCH 020/158] RDMA/irdma: Fix drain SQ hang with no completion SW generated completions for outstanding WRs posted on SQ after QP is in error target the wrong CQ. This causes the ib_drain_sq to hang with no completion. Fix this to generate completions on the right CQ. [ 863.969340] INFO: task kworker/u52:2:671 blocked for more than 122 seconds. [ 863.979224] Not tainted 5.14.0-130.el9.x86_64 #1 [ 863.986588] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 863.996997] task:kworker/u52:2 state:D stack: 0 pid: 671 ppid: 2 flags:0x00004000 [ 864.007272] Workqueue: xprtiod xprt_autoclose [sunrpc] [ 864.014056] Call Trace: [ 864.017575] __schedule+0x206/0x580 [ 864.022296] schedule+0x43/0xa0 [ 864.026736] schedule_timeout+0x115/0x150 [ 864.032185] __wait_for_common+0x93/0x1d0 [ 864.037717] ? usleep_range_state+0x90/0x90 [ 864.043368] __ib_drain_sq+0xf6/0x170 [ib_core] [ 864.049371] ? __rdma_block_iter_next+0x80/0x80 [ib_core] [ 864.056240] ib_drain_sq+0x66/0x70 [ib_core] [ 864.062003] rpcrdma_xprt_disconnect+0x82/0x3b0 [rpcrdma] [ 864.069365] ? xprt_prepare_transmit+0x5d/0xc0 [sunrpc] [ 864.076386] xprt_rdma_close+0xe/0x30 [rpcrdma] [ 864.082593] xprt_autoclose+0x52/0x100 [sunrpc] [ 864.088718] process_one_work+0x1e8/0x3c0 [ 864.094170] worker_thread+0x50/0x3b0 [ 864.099109] ? rescuer_thread+0x370/0x370 [ 864.104473] kthread+0x149/0x170 [ 864.109022] ? set_kthread_struct+0x40/0x40 [ 864.114713] ret_from_fork+0x22/0x30 Fixes: 81091d7696ae ("RDMA/irdma: Add SW mechanism to generate completions on error") Link: https://lore.kernel.org/r/20220824154358.117-1-shiraz.saleem@intel.com Reported-by: Kamal Heib Signed-off-by: Shiraz Saleem Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/irdma/utils.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/irdma/utils.c b/drivers/infiniband/hw/irdma/utils.c index fdf4cc88cb91..c8b92355ddea 100644 --- a/drivers/infiniband/hw/irdma/utils.c +++ b/drivers/infiniband/hw/irdma/utils.c @@ -2598,7 +2598,7 @@ void irdma_generate_flush_completions(struct irdma_qp *iwqp) spin_unlock_irqrestore(&iwqp->lock, flags2); spin_unlock_irqrestore(&iwqp->iwscq->lock, flags1); if (compl_generated) - irdma_comp_handler(iwqp->iwrcq); + irdma_comp_handler(iwqp->iwscq); } else { spin_unlock_irqrestore(&iwqp->iwscq->lock, flags1); mod_delayed_work(iwqp->iwdev->cleanup_wq, &iwqp->dwork_flush, From 27cfde795a96aef1e859a5480489944b95421e46 Mon Sep 17 00:00:00 2001 From: Michael Guralnik Date: Tue, 23 Aug 2022 13:51:50 +0300 Subject: [PATCH 021/158] RDMA/cma: Fix arguments order in net device validation Fix the order of source and destination addresses when resolving the route between server and client to validate use of correct net device. The reverse order we had so far didn't actually validate the net device as the server would try to resolve the route to itself, thus always getting the server's net device. The issue was discovered when running cm applications on a single host between 2 interfaces with same subnet and source based routing rules. When resolving the reverse route the source based route rules were ignored. Fixes: f887f2ac87c2 ("IB/cma: Validate routing of incoming requests") Link: https://lore.kernel.org/r/1c1ec2277a131d277ebcceec987fd338d35b775f.1661251872.git.leonro@nvidia.com Signed-off-by: Michael Guralnik Signed-off-by: Leon Romanovsky --- drivers/infiniband/core/cma.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 46d06678dfbe..be317f2665a9 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -1841,8 +1841,8 @@ cma_ib_id_from_event(struct ib_cm_id *cm_id, } if (!validate_net_dev(*net_dev, - (struct sockaddr *)&req->listen_addr_storage, - (struct sockaddr *)&req->src_addr_storage)) { + (struct sockaddr *)&req->src_addr_storage, + (struct sockaddr *)&req->listen_addr_storage)) { id_priv = ERR_PTR(-EHOSTUNREACH); goto err; } From 6292b4ba607de5bcf4ab0e57892a2f8068e6b997 Mon Sep 17 00:00:00 2001 From: Luca Ceresoli Date: Wed, 24 Aug 2022 10:31:02 +0200 Subject: [PATCH 022/158] docs: i2c: i2c-topology: fix typo "intension" should have probably been "intention", however "intent" seems even better. Reported-by: Bagas Sanjaya Signed-off-by: Luca Ceresoli Acked-by: Peter Rosin Signed-off-by: Wolfram Sang --- Documentation/i2c/i2c-topology.rst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/i2c/i2c-topology.rst b/Documentation/i2c/i2c-topology.rst index 7cb53819778e..c9ed3b4d6085 100644 --- a/Documentation/i2c/i2c-topology.rst +++ b/Documentation/i2c/i2c-topology.rst @@ -103,7 +103,7 @@ ML2. It is not safe to build arbitrary topologies with two (or more) I.e. the select-transfer-deselect transaction targeting e.g. device address 0x42 behind mux-one may be interleaved with a similar operation targeting device address 0x42 behind mux-two. The - intension with such a topology would in this hypothetical example + intent with such a topology would in this hypothetical example be that mux-one and mux-two should not be selected simultaneously, but mux-locked muxes do not guarantee that in all topologies. From 12c035a1a840e07a351c277d7a468269fa603000 Mon Sep 17 00:00:00 2001 From: Luca Ceresoli Date: Wed, 24 Aug 2022 10:31:03 +0200 Subject: [PATCH 023/158] docs: i2c: i2c-topology: fix incorrect heading "Etc" here was never meant to be a heading, it became one while converting to ReST. It would be easy to just convert it to plain text, but rather remove it and add an introductory text before the list that conveys the same meaning but with a better reading flow. Fixes: ccf988b66d69 ("docs: i2c: convert to ReST and add to driver-api bookset") Signed-off-by: Luca Ceresoli Acked-by: Peter Rosin Signed-off-by: Wolfram Sang --- Documentation/i2c/i2c-topology.rst | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/Documentation/i2c/i2c-topology.rst b/Documentation/i2c/i2c-topology.rst index c9ed3b4d6085..326a1198551e 100644 --- a/Documentation/i2c/i2c-topology.rst +++ b/Documentation/i2c/i2c-topology.rst @@ -5,6 +5,8 @@ I2C muxes and complex topologies There are a couple of reasons for building more complex I2C topologies than a straight-forward I2C bus with one adapter and one or more devices. +Some example use cases are: + 1. A mux may be needed on the bus to prevent address collisions. 2. The bus may be accessible from some external bus master, and arbitration @@ -14,9 +16,6 @@ than a straight-forward I2C bus with one adapter and one or more devices. from the I2C bus, at least most of the time, and sits behind a gate that has to be operated before the device can be accessed. -Etc -=== - These constructs are represented as I2C adapter trees by Linux, where each adapter has a parent adapter (except the root adapter) and zero or more child adapters. The root adapter is the actual adapter that issues From 92d5d6434d285510a3563ebace24306e240615e1 Mon Sep 17 00:00:00 2001 From: Luca Ceresoli Date: Wed, 24 Aug 2022 10:31:04 +0200 Subject: [PATCH 024/158] docs: i2c: i2c-topology: reorder sections more logically The sequence of sections is a bit confusing here: * we list the mux locking scheme for existing drivers before introducing what mux locking schemes are * we list the caveats for each locking scheme (which are tricky) before the example of the simple use case Restructure it entirely with the following logic: * Intro ("I2C muxes and complex topologies") * Locking - mux-locked - example - caveats - parent-locked - example - caveats * Complex examples * Mux type of existing device drivers While there, also apply some other improvements: * convert the caveat list from a table (with only one column carrying content) to a bullet list. * add a small introductory text to bridge the gap from listing the use cases to telling about the hardware components to handle them and then the device drivers that implement those. * make empty lines usage more uniform Signed-off-by: Luca Ceresoli Acked-by: Peter Rosin Signed-off-by: Wolfram Sang --- Documentation/i2c/i2c-topology.rst | 211 ++++++++++++++++------------- 1 file changed, 114 insertions(+), 97 deletions(-) diff --git a/Documentation/i2c/i2c-topology.rst b/Documentation/i2c/i2c-topology.rst index 326a1198551e..48fce0f7491b 100644 --- a/Documentation/i2c/i2c-topology.rst +++ b/Documentation/i2c/i2c-topology.rst @@ -16,7 +16,10 @@ Some example use cases are: from the I2C bus, at least most of the time, and sits behind a gate that has to be operated before the device can be accessed. -These constructs are represented as I2C adapter trees by Linux, where +Several types of hardware components such as I2C muxes, I2C gates and I2C +arbitrators allow to handle such needs. + +These components are represented as I2C adapter trees by Linux, where each adapter has a parent adapter (except the root adapter) and zero or more child adapters. The root adapter is the actual adapter that issues I2C transfers, and all adapters with a parent are part of an "i2c-mux" @@ -34,46 +37,7 @@ Locking ======= There are two variants of locking available to I2C muxes, they can be -mux-locked or parent-locked muxes. As is evident from below, it can be -useful to know if a mux is mux-locked or if it is parent-locked. The -following list was correct at the time of writing: - -In drivers/i2c/muxes/: - -====================== ============================================= -i2c-arb-gpio-challenge Parent-locked -i2c-mux-gpio Normally parent-locked, mux-locked iff - all involved gpio pins are controlled by the - same I2C root adapter that they mux. -i2c-mux-gpmux Normally parent-locked, mux-locked iff - specified in device-tree. -i2c-mux-ltc4306 Mux-locked -i2c-mux-mlxcpld Parent-locked -i2c-mux-pca9541 Parent-locked -i2c-mux-pca954x Parent-locked -i2c-mux-pinctrl Normally parent-locked, mux-locked iff - all involved pinctrl devices are controlled - by the same I2C root adapter that they mux. -i2c-mux-reg Parent-locked -====================== ============================================= - -In drivers/iio/: - -====================== ============================================= -gyro/mpu3050 Mux-locked -imu/inv_mpu6050/ Mux-locked -====================== ============================================= - -In drivers/media/: - -======================= ============================================= -dvb-frontends/lgdt3306a Mux-locked -dvb-frontends/m88ds3103 Parent-locked -dvb-frontends/rtl2830 Parent-locked -dvb-frontends/rtl2832 Mux-locked -dvb-frontends/si2168 Mux-locked -usb/cx231xx/ Parent-locked -======================= ============================================= +mux-locked or parent-locked muxes. Mux-locked muxes @@ -88,40 +52,8 @@ full transaction, unrelated I2C transfers may interleave the different stages of the transaction. This has the benefit that the mux driver may be easier and cleaner to implement, but it has some caveats. -==== ===================================================================== -ML1. If you build a topology with a mux-locked mux being the parent - of a parent-locked mux, this might break the expectation from the - parent-locked mux that the root adapter is locked during the - transaction. - -ML2. It is not safe to build arbitrary topologies with two (or more) - mux-locked muxes that are not siblings, when there are address - collisions between the devices on the child adapters of these - non-sibling muxes. - - I.e. the select-transfer-deselect transaction targeting e.g. device - address 0x42 behind mux-one may be interleaved with a similar - operation targeting device address 0x42 behind mux-two. The - intent with such a topology would in this hypothetical example - be that mux-one and mux-two should not be selected simultaneously, - but mux-locked muxes do not guarantee that in all topologies. - -ML3. A mux-locked mux cannot be used by a driver for auto-closing - gates/muxes, i.e. something that closes automatically after a given - number (one, in most cases) of I2C transfers. Unrelated I2C transfers - may creep in and close prematurely. - -ML4. If any non-I2C operation in the mux driver changes the I2C mux state, - the driver has to lock the root adapter during that operation. - Otherwise garbage may appear on the bus as seen from devices - behind the mux, when an unrelated I2C transfer is in flight during - the non-I2C mux-changing operation. -==== ===================================================================== - - Mux-locked Example ------------------- - +~~~~~~~~~~~~~~~~~~ :: @@ -152,6 +84,43 @@ This means that accesses to D2 are lockout out for the full duration of the entire operation. But accesses to D3 are possibly interleaved at any point. +Mux-locked caveats +~~~~~~~~~~~~~~~~~~ + +When using a mux-locked mux, be aware of the following restrictions: + +[ML1] + If you build a topology with a mux-locked mux being the parent + of a parent-locked mux, this might break the expectation from the + parent-locked mux that the root adapter is locked during the + transaction. + +[ML2] + It is not safe to build arbitrary topologies with two (or more) + mux-locked muxes that are not siblings, when there are address + collisions between the devices on the child adapters of these + non-sibling muxes. + + I.e. the select-transfer-deselect transaction targeting e.g. device + address 0x42 behind mux-one may be interleaved with a similar + operation targeting device address 0x42 behind mux-two. The + intent with such a topology would in this hypothetical example + be that mux-one and mux-two should not be selected simultaneously, + but mux-locked muxes do not guarantee that in all topologies. + +[ML3] + A mux-locked mux cannot be used by a driver for auto-closing + gates/muxes, i.e. something that closes automatically after a given + number (one, in most cases) of I2C transfers. Unrelated I2C transfers + may creep in and close prematurely. + +[ML4] + If any non-I2C operation in the mux driver changes the I2C mux state, + the driver has to lock the root adapter during that operation. + Otherwise garbage may appear on the bus as seen from devices + behind the mux, when an unrelated I2C transfer is in flight during + the non-I2C mux-changing operation. + Parent-locked muxes ------------------- @@ -160,28 +129,10 @@ Parent-locked muxes lock the parent adapter during the full select- transfer-deselect transaction. The implication is that the mux driver has to ensure that any and all I2C transfers through that parent adapter during the transaction are unlocked I2C transfers (using e.g. -__i2c_transfer), or a deadlock will follow. There are a couple of -caveats. - -==== ==================================================================== -PL1. If you build a topology with a parent-locked mux being the child - of another mux, this might break a possible assumption from the - child mux that the root adapter is unused between its select op - and the actual transfer (e.g. if the child mux is auto-closing - and the parent mux issues I2C transfers as part of its select). - This is especially the case if the parent mux is mux-locked, but - it may also happen if the parent mux is parent-locked. - -PL2. If select/deselect calls out to other subsystems such as gpio, - pinctrl, regmap or iio, it is essential that any I2C transfers - caused by these subsystems are unlocked. This can be convoluted to - accomplish, maybe even impossible if an acceptably clean solution - is sought. -==== ==================================================================== - +__i2c_transfer), or a deadlock will follow. Parent-locked Example ---------------------- +~~~~~~~~~~~~~~~~~~~~~ :: @@ -211,10 +162,30 @@ When there is an access to D1, this happens: 9. M1 unlocks its parent adapter. 10. M1 unlocks muxes on its parent. - This means that accesses to both D2 and D3 are locked out for the full duration of the entire operation. +Parent-locked Caveats +~~~~~~~~~~~~~~~~~~~~~ + +When using a parent-locked mux, be aware of the following restrictions: + +[PL1] + If you build a topology with a parent-locked mux being the child + of another mux, this might break a possible assumption from the + child mux that the root adapter is unused between its select op + and the actual transfer (e.g. if the child mux is auto-closing + and the parent mux issues I2C transfers as part of its select). + This is especially the case if the parent mux is mux-locked, but + it may also happen if the parent mux is parent-locked. + +[PL2] + If select/deselect calls out to other subsystems such as gpio, + pinctrl, regmap or iio, it is essential that any I2C transfers + caused by these subsystems are unlocked. This can be convoluted to + accomplish, maybe even impossible if an acceptably clean solution + is sought. + Complex Examples ================ @@ -260,8 +231,10 @@ This is a good topology:: When device D1 is accessed, accesses to D2 are locked out for the full duration of the operation (muxes on the top child adapter of M1 are locked). But accesses to D3 and D4 are possibly interleaved at -any point. Accesses to D3 locks out D1 and D2, but accesses to D4 -are still possibly interleaved. +any point. + +Accesses to D3 locks out D1 and D2, but accesses to D4 are still possibly +interleaved. Mux-locked mux as parent of parent-locked mux @@ -393,3 +366,47 @@ This is a good topology:: When D1 or D2 are accessed, accesses to D3 and D4 are locked out while accesses to D5 may interleave. When D3 or D4 are accessed, accesses to all other devices are locked out. + + +Mux type of existing device drivers +=================================== + +Whether a device is mux-locked or parent-locked depends on its +implementation. The following list was correct at the time of writing: + +In drivers/i2c/muxes/: + +====================== ============================================= +i2c-arb-gpio-challenge Parent-locked +i2c-mux-gpio Normally parent-locked, mux-locked iff + all involved gpio pins are controlled by the + same I2C root adapter that they mux. +i2c-mux-gpmux Normally parent-locked, mux-locked iff + specified in device-tree. +i2c-mux-ltc4306 Mux-locked +i2c-mux-mlxcpld Parent-locked +i2c-mux-pca9541 Parent-locked +i2c-mux-pca954x Parent-locked +i2c-mux-pinctrl Normally parent-locked, mux-locked iff + all involved pinctrl devices are controlled + by the same I2C root adapter that they mux. +i2c-mux-reg Parent-locked +====================== ============================================= + +In drivers/iio/: + +====================== ============================================= +gyro/mpu3050 Mux-locked +imu/inv_mpu6050/ Mux-locked +====================== ============================================= + +In drivers/media/: + +======================= ============================================= +dvb-frontends/lgdt3306a Mux-locked +dvb-frontends/m88ds3103 Parent-locked +dvb-frontends/rtl2830 Parent-locked +dvb-frontends/rtl2832 Mux-locked +dvb-frontends/si2168 Mux-locked +usb/cx231xx/ Parent-locked +======================= ============================================= From 35508d2424097f9b6a1a17aac94f702767035616 Mon Sep 17 00:00:00 2001 From: Yang Ling Date: Tue, 23 Aug 2022 19:17:25 +0800 Subject: [PATCH 025/158] MIPS: loongson32: ls1c: Fix hang during startup The RTCCTRL reg of LS1C is obselete. Writing this reg will cause system hang. Fixes: 60219c563c9b6 ("MIPS: Add RTC support for Loongson1C board") Signed-off-by: Yang Ling Tested-by: Keguang Zhang Acked-by: Keguang Zhang Signed-off-by: Thomas Bogendoerfer --- arch/mips/loongson32/ls1c/board.c | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/mips/loongson32/ls1c/board.c b/arch/mips/loongson32/ls1c/board.c index e9de6da0ce51..9dcfe9de55b0 100644 --- a/arch/mips/loongson32/ls1c/board.c +++ b/arch/mips/loongson32/ls1c/board.c @@ -15,7 +15,6 @@ static struct platform_device *ls1c_platform_devices[] __initdata = { static int __init ls1c_platform_init(void) { ls1x_serial_set_uartclk(&ls1x_uart_pdev); - ls1x_rtc_set_extclk(&ls1x_rtc_pdev); return platform_add_devices(ls1c_platform_devices, ARRAY_SIZE(ls1c_platform_devices)); From 53b774718f784d614c8bfed20a3b32fb753604f3 Mon Sep 17 00:00:00 2001 From: Bruce Duncan Date: Thu, 25 Aug 2022 15:01:41 +0200 Subject: [PATCH 026/158] docs: i2c: piix4: Fix typos, add markup, drop link [JD: Update the subject One more typo fixed Drop the link to lm-sensors' README, it's irrelevant] Signed-off-by: Bruce Duncan Reviewed-by: Jean Delvare Signed-off-by: Wolfram Sang --- Documentation/i2c/busses/i2c-piix4.rst | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-) diff --git a/Documentation/i2c/busses/i2c-piix4.rst b/Documentation/i2c/busses/i2c-piix4.rst index cc9000259223..07fe6f6f4b18 100644 --- a/Documentation/i2c/busses/i2c-piix4.rst +++ b/Documentation/i2c/busses/i2c-piix4.rst @@ -64,7 +64,7 @@ correct address for this module, you could get in big trouble (read: crashes, data corruption, etc.). Try this only as a last resort (try BIOS updates first, for example), and backup first! An even more dangerous option is 'force_addr='. This will not only enable the PIIX4 like -'force' foes, but it will also set a new base I/O port address. The SMBus +'force' does, but it will also set a new base I/O port address. The SMBus parts of the PIIX4 needs a range of 8 of these addresses to function correctly. If these addresses are already reserved by some other device, you will get into big trouble! DON'T USE THIS IF YOU ARE NOT VERY SURE @@ -86,15 +86,15 @@ If you own Force CPCI735 motherboard or other OSB4 based systems you may need to change the SMBus Interrupt Select register so the SMBus controller uses the SMI mode. -1) Use lspci command and locate the PCI device with the SMBus controller: +1) Use ``lspci`` command and locate the PCI device with the SMBus controller: 00:0f.0 ISA bridge: ServerWorks OSB4 South Bridge (rev 4f) The line may vary for different chipsets. Please consult the driver source - for all possible PCI ids (and lspci -n to match them). Lets assume the + for all possible PCI ids (and ``lspci -n`` to match them). Let's assume the device is located at 00:0f.0. 2) Now you just need to change the value in 0xD2 register. Get it first with - command: lspci -xxx -s 00:0f.0 + command: ``lspci -xxx -s 00:0f.0`` If the value is 0x3 then you need to change it to 0x1: - setpci -s 00:0f.0 d2.b=1 + ``setpci -s 00:0f.0 d2.b=1`` Please note that you don't need to do that in all cases, just when the SMBus is not working properly. @@ -109,6 +109,3 @@ which can easily get corrupted due to a state machine bug. These are mostly Thinkpad laptops, but desktop systems may also be affected. We have no list of all affected systems, so the only safe solution was to prevent access to the SMBus on all IBM systems (detected using DMI data.) - -For additional information, read: -http://www.lm-sensors.org/browser/lm-sensors/trunk/README From 55af9d498556f0860eb89ffa7677e8d73f6f643f Mon Sep 17 00:00:00 2001 From: Chengchang Tang Date: Mon, 29 Aug 2022 18:50:18 +0800 Subject: [PATCH 027/158] RDMA/hns: Fix supported page size The supported page size for hns is (4K, 128M), not (4K, 2G). Fixes: cfc85f3e4b7f ("RDMA/hns: Add profile support for hip08 driver") Link: https://lore.kernel.org/r/20220829105021.1427804-2-liangwenpeng@huawei.com Signed-off-by: Chengchang Tang Signed-off-by: Wenpeng Liang Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/hns/hns_roce_hw_v2.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h index f96debac30fe..6cf07355348d 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h @@ -83,7 +83,7 @@ #define HNS_ROCE_V2_QPC_TIMER_ENTRY_SZ PAGE_SIZE #define HNS_ROCE_V2_CQC_TIMER_ENTRY_SZ PAGE_SIZE -#define HNS_ROCE_V2_PAGE_SIZE_SUPPORTED 0xFFFFF000 +#define HNS_ROCE_V2_PAGE_SIZE_SUPPORTED 0xFFFF000 #define HNS_ROCE_V2_MAX_INNER_MTPT_NUM 2 #define HNS_ROCE_INVALID_LKEY 0x0 #define HNS_ROCE_INVALID_SGE_LENGTH 0x80000000 From 0c8b5d6268d92d141bfd64d21c870d295a84dee1 Mon Sep 17 00:00:00 2001 From: Wenpeng Liang Date: Mon, 29 Aug 2022 18:50:19 +0800 Subject: [PATCH 028/158] RDMA/hns: Fix wrong fixed value of qp->rq.wqe_shift The value of qp->rq.wqe_shift of HIP08 is always determined by the number of sge. So delete the wrong branch. Fixes: cfc85f3e4b7f ("RDMA/hns: Add profile support for hip08 driver") Fixes: 926a01dc000d ("RDMA/hns: Add QP operations support for hip08 SoC") Link: https://lore.kernel.org/r/20220829105021.1427804-3-liangwenpeng@huawei.com Signed-off-by: Wenpeng Liang Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/hns/hns_roce_qp.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/drivers/infiniband/hw/hns/hns_roce_qp.c b/drivers/infiniband/hw/hns/hns_roce_qp.c index 48d3616a6d71..7bee7f6c5e70 100644 --- a/drivers/infiniband/hw/hns/hns_roce_qp.c +++ b/drivers/infiniband/hw/hns/hns_roce_qp.c @@ -462,11 +462,8 @@ static int set_rq_size(struct hns_roce_dev *hr_dev, struct ib_qp_cap *cap, hr_qp->rq.max_gs = roundup_pow_of_two(max(1U, cap->max_recv_sge) + hr_qp->rq.rsv_sge); - if (hr_dev->caps.max_rq_sg <= HNS_ROCE_SGE_IN_WQE) - hr_qp->rq.wqe_shift = ilog2(hr_dev->caps.max_rq_desc_sz); - else - hr_qp->rq.wqe_shift = ilog2(hr_dev->caps.max_rq_desc_sz * - hr_qp->rq.max_gs); + hr_qp->rq.wqe_shift = ilog2(hr_dev->caps.max_rq_desc_sz * + hr_qp->rq.max_gs); hr_qp->rq.wqe_cnt = cnt; if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_RQ_INLINE && From 45baad7dd98f4d83f67c86c28769d3184390e324 Mon Sep 17 00:00:00 2001 From: Yixing Liu Date: Mon, 29 Aug 2022 18:50:20 +0800 Subject: [PATCH 029/158] RDMA/hns: Remove the num_qpc_timer variable The bt number of qpc_timer of HIP09 increases compared with that of HIP08. Therefore, qpc_timer_bt_num and num_qpc_timer do not match. As a result, the driver may fail to allocate qpc_timer. So the driver needs to uniquely uses qpc_timer_bt_num to represent the bt number of qpc_timer. Fixes: 0e40dc2f70cd ("RDMA/hns: Add timer allocation support for hip08") Link: https://lore.kernel.org/r/20220829105021.1427804-4-liangwenpeng@huawei.com Signed-off-by: Yixing Liu Signed-off-by: Wenpeng Liang Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/hns/hns_roce_device.h | 1 - drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 3 +-- drivers/infiniband/hw/hns/hns_roce_hw_v2.h | 2 +- drivers/infiniband/hw/hns/hns_roce_main.c | 2 +- 4 files changed, 3 insertions(+), 5 deletions(-) diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h index f848eedc6a23..d24996526c4d 100644 --- a/drivers/infiniband/hw/hns/hns_roce_device.h +++ b/drivers/infiniband/hw/hns/hns_roce_device.h @@ -730,7 +730,6 @@ struct hns_roce_caps { u32 num_qps; u32 num_pi_qps; u32 reserved_qps; - int num_qpc_timer; u32 num_srqs; u32 max_wqes; u32 max_srq_wrs; diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c index cbdafaac678a..c780646bd60a 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c @@ -1977,7 +1977,7 @@ static void set_default_caps(struct hns_roce_dev *hr_dev) caps->num_mtpts = HNS_ROCE_V2_MAX_MTPT_NUM; caps->num_pds = HNS_ROCE_V2_MAX_PD_NUM; - caps->num_qpc_timer = HNS_ROCE_V2_MAX_QPC_TIMER_NUM; + caps->qpc_timer_bt_num = HNS_ROCE_V2_MAX_QPC_TIMER_BT_NUM; caps->cqc_timer_bt_num = HNS_ROCE_V2_MAX_CQC_TIMER_BT_NUM; caps->max_qp_init_rdma = HNS_ROCE_V2_MAX_QP_INIT_RDMA; @@ -2273,7 +2273,6 @@ static int hns_roce_query_pf_caps(struct hns_roce_dev *hr_dev) caps->max_rq_sg = le16_to_cpu(resp_a->max_rq_sg); caps->max_rq_sg = roundup_pow_of_two(caps->max_rq_sg); caps->max_extend_sg = le32_to_cpu(resp_a->max_extend_sg); - caps->num_qpc_timer = le16_to_cpu(resp_a->num_qpc_timer); caps->max_srq_sges = le16_to_cpu(resp_a->max_srq_sges); caps->max_srq_sges = roundup_pow_of_two(caps->max_srq_sges); caps->num_aeq_vectors = resp_a->num_aeq_vectors; diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h index 6cf07355348d..64797109bab6 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h @@ -36,11 +36,11 @@ #include #define HNS_ROCE_V2_MAX_QP_NUM 0x1000 -#define HNS_ROCE_V2_MAX_QPC_TIMER_NUM 0x200 #define HNS_ROCE_V2_MAX_WQE_NUM 0x8000 #define HNS_ROCE_V2_MAX_SRQ_WR 0x8000 #define HNS_ROCE_V2_MAX_SRQ_SGE 64 #define HNS_ROCE_V2_MAX_CQ_NUM 0x100000 +#define HNS_ROCE_V2_MAX_QPC_TIMER_BT_NUM 0x100 #define HNS_ROCE_V2_MAX_CQC_TIMER_BT_NUM 0x100 #define HNS_ROCE_V2_MAX_SRQ_NUM 0x100000 #define HNS_ROCE_V2_MAX_CQE_NUM 0x400000 diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c index c8af4ebd7cbd..4ccb217b2841 100644 --- a/drivers/infiniband/hw/hns/hns_roce_main.c +++ b/drivers/infiniband/hw/hns/hns_roce_main.c @@ -725,7 +725,7 @@ static int hns_roce_init_hem(struct hns_roce_dev *hr_dev) ret = hns_roce_init_hem_table(hr_dev, &hr_dev->qpc_timer_table, HEM_TYPE_QPC_TIMER, hr_dev->caps.qpc_timer_entry_sz, - hr_dev->caps.num_qpc_timer, 1); + hr_dev->caps.qpc_timer_bt_num, 1); if (ret) { dev_err(dev, "Failed to init QPC timer memory, aborting.\n"); From 4e2a2ed96adcb6151a15ea40db933d97f2ae7aaa Mon Sep 17 00:00:00 2001 From: Lad Prabhakar Date: Mon, 29 Aug 2022 22:47:30 +0100 Subject: [PATCH 030/158] dt-bindings: i2c: renesas,riic: Fix 'unevaluatedProperties' warnings With 'unevaluatedProperties' support implemented, there's a number of warnings when running dtbs_check: arch/arm64/boot/dts/renesas/r9a07g043u11-smarc.dtb: i2c@10058000: Unevaluated properties are not allowed ('resets' was unexpected) From schema: Documentation/devicetree/bindings/i2c/renesas,riic.yaml The main problem is that bindings schema marks resets as a required property for RZ/G2L (and alike) SoC's but resets property is not part of schema. So to fix this just add a resets property with maxItems set to 1. Signed-off-by: Lad Prabhakar Reviewed-by: Geert Uytterhoeven Acked-by: Krzysztof Kozlowski Signed-off-by: Wolfram Sang --- Documentation/devicetree/bindings/i2c/renesas,riic.yaml | 3 +++ 1 file changed, 3 insertions(+) diff --git a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml index 2f315489aaae..d3c0d5c427ac 100644 --- a/Documentation/devicetree/bindings/i2c/renesas,riic.yaml +++ b/Documentation/devicetree/bindings/i2c/renesas,riic.yaml @@ -60,6 +60,9 @@ properties: power-domains: maxItems: 1 + resets: + maxItems: 1 + required: - compatible - reg From 3f5df3ac646e21a79a421ae4037c4ef0632bcaa9 Mon Sep 17 00:00:00 2001 From: Ian Rogers Date: Tue, 30 Aug 2022 09:48:40 -0700 Subject: [PATCH 031/158] perf metric: Return early if no CPU PMU table exists Previous behavior is to segfault if there is no CPU PMU table and a metric is sought. To reproduce compile with NO_JEVENTS=1 then request a metric, for example, "perf stat -M IPC true". Committer testing: Before: $ make -k NO_JEVENTS=1 BUILD_BPF_SKEL=1 O=/tmp/build/perf-urgent -C tools/perf install-bin $ perf stat -M IPC true Segmentation fault (core dumped) $ After: $ perf stat -M IPC true Usage: perf stat [] [] -M, --metrics monitor specified metrics or metric groups (separated by ,) $ Fixes: 00facc760903be66 ("perf jevents: Switch build to use jevents.py") Signed-off-by: Ian Rogers Tested-by: Arnaldo Carvalho de Melo Cc: Alexander Shishkin Cc: Andi Kleen Cc: Caleb Biggers Cc: Florian Fischer Cc: Ian Rogers Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: John Garry Cc: Kan Liang Cc: Kshipra Bopardikar Cc: Mark Rutland Cc: Miaoqian Lin Cc: Namhyung Kim Cc: Perry Taylor Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Richter Cc: Xing Zhengjun Link: https://lore.kernel.org/r/20220830164846.401143-3-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/metricgroup.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/tools/perf/util/metricgroup.c b/tools/perf/util/metricgroup.c index 464475fd6b9a..c93bcaf6d55d 100644 --- a/tools/perf/util/metricgroup.c +++ b/tools/perf/util/metricgroup.c @@ -1655,6 +1655,9 @@ int metricgroup__parse_groups(const struct option *opt, struct evlist *perf_evlist = *(struct evlist **)opt->value; const struct pmu_events_table *table = pmu_events_table__find(); + if (!table) + return -EINVAL; + return parse_groups(perf_evlist, str, metric_no_group, metric_no_merge, NULL, metric_events, table); } From 873aefb376bbc0ed1dd2381ea1d6ec88106fdbd4 Mon Sep 17 00:00:00 2001 From: Alex Williamson Date: Mon, 29 Aug 2022 21:05:40 -0600 Subject: [PATCH 032/158] vfio/type1: Unpin zero pages There's currently a reference count leak on the zero page. We increment the reference via pin_user_pages_remote(), but the page is later handled as an invalid/reserved page, therefore it's not accounted against the user and not unpinned by our put_pfn(). Introducing special zero page handling in put_pfn() would resolve the leak, but without accounting of the zero page, a single user could still create enough mappings to generate a reference count overflow. The zero page is always resident, so for our purposes there's no reason to keep it pinned. Therefore, add a loop to walk pages returned from pin_user_pages_remote() and unpin any zero pages. Cc: stable@vger.kernel.org Reported-by: Luboslav Pivarc Reviewed-by: David Hildenbrand Link: https://lore.kernel.org/r/166182871735.3518559.8884121293045337358.stgit@omen Signed-off-by: Alex Williamson --- drivers/vfio/vfio_iommu_type1.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/drivers/vfio/vfio_iommu_type1.c b/drivers/vfio/vfio_iommu_type1.c index db516c90a977..8706482665d1 100644 --- a/drivers/vfio/vfio_iommu_type1.c +++ b/drivers/vfio/vfio_iommu_type1.c @@ -558,6 +558,18 @@ static int vaddr_get_pfns(struct mm_struct *mm, unsigned long vaddr, ret = pin_user_pages_remote(mm, vaddr, npages, flags | FOLL_LONGTERM, pages, NULL, NULL); if (ret > 0) { + int i; + + /* + * The zero page is always resident, we don't need to pin it + * and it falls into our invalid/reserved test so we don't + * unpin in put_pfn(). Unpin all zero pages in the batch here. + */ + for (i = 0 ; i < ret; i++) { + if (unlikely(is_zero_pfn(page_to_pfn(pages[i])))) + unpin_user_page(pages[i]); + } + *pfn = page_to_pfn(pages[0]); goto done; } From 17e4732d1d8a859fbb56e5f050e05d3142b88f96 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Thu, 25 Aug 2022 19:04:17 +0100 Subject: [PATCH 033/158] dt-bindings: riscv: sifive-l2: add a PolarFire SoC compatible The l2 cache on PolarFire SoC is cross between that of the fu540 and the fu740. It has the extra interrupt from the fu740 but the lower number of cache-sets. Add a specific compatible to avoid the likes of: mpfs-polarberry.dtb: cache-controller@2010000: interrupts: [[1], [3], [4], [2]] is too long Fixes: 34fc9cc3aebe ("riscv: dts: microchip: correct L2 cache interrupts") Reviewed-by: Rob Herring Signed-off-by: Conor Dooley --- .../bindings/riscv/sifive-l2-cache.yaml | 77 ++++++++++++------- 1 file changed, 48 insertions(+), 29 deletions(-) diff --git a/Documentation/devicetree/bindings/riscv/sifive-l2-cache.yaml b/Documentation/devicetree/bindings/riscv/sifive-l2-cache.yaml index 69cdab18d629..ca3b9be58058 100644 --- a/Documentation/devicetree/bindings/riscv/sifive-l2-cache.yaml +++ b/Documentation/devicetree/bindings/riscv/sifive-l2-cache.yaml @@ -17,9 +17,6 @@ description: acts as directory-based coherency manager. All the properties in ePAPR/DeviceTree specification applies for this platform. -allOf: - - $ref: /schemas/cache-controller.yaml# - select: properties: compatible: @@ -33,11 +30,16 @@ select: properties: compatible: - items: - - enum: - - sifive,fu540-c000-ccache - - sifive,fu740-c000-ccache - - const: cache + oneOf: + - items: + - enum: + - sifive,fu540-c000-ccache + - sifive,fu740-c000-ccache + - const: cache + - items: + - const: microchip,mpfs-ccache + - const: sifive,fu540-c000-ccache + - const: cache cache-block-size: const: 64 @@ -72,29 +74,46 @@ properties: The reference to the reserved-memory for the L2 Loosely Integrated Memory region. The reserved memory node should be defined as per the bindings in reserved-memory.txt. -if: - properties: - compatible: - contains: - const: sifive,fu540-c000-ccache +allOf: + - $ref: /schemas/cache-controller.yaml# -then: - properties: - interrupts: - description: | - Must contain entries for DirError, DataError and DataFail signals. - maxItems: 3 - cache-sets: - const: 1024 + - if: + properties: + compatible: + contains: + enum: + - sifive,fu740-c000-ccache + - microchip,mpfs-ccache -else: - properties: - interrupts: - description: | - Must contain entries for DirError, DataError, DataFail, DirFail signals. - minItems: 4 - cache-sets: - const: 2048 + then: + properties: + interrupts: + description: | + Must contain entries for DirError, DataError, DataFail, DirFail signals. + minItems: 4 + + else: + properties: + interrupts: + description: | + Must contain entries for DirError, DataError and DataFail signals. + maxItems: 3 + + - if: + properties: + compatible: + contains: + const: sifive,fu740-c000-ccache + + then: + properties: + cache-sets: + const: 2048 + + else: + properties: + cache-sets: + const: 1024 additionalProperties: false From 0dec364ffeb6149aae572ded1e34d4b444c23be6 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Thu, 25 Aug 2022 19:04:18 +0100 Subject: [PATCH 034/158] riscv: dts: microchip: use an mpfs specific l2 compatible PolarFire SoC does not have the same l2 cache controller as the fu540, featuring an extra interrupt. Appease the devicetree checker overlords by adding a PolarFire SoC specific compatible to fix the below sort of warnings: mpfs-polarberry.dtb: cache-controller@2010000: interrupts: [[1], [3], [4], [2]] is too long Fixes: 0fa6107eca41 ("RISC-V: Initial DTS for Microchip ICICLE board") Fixes: 34fc9cc3aebe ("riscv: dts: microchip: correct L2 cache interrupts") Reviewed-by: Heinrich Schuchardt Signed-off-by: Conor Dooley --- arch/riscv/boot/dts/microchip/mpfs.dtsi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/riscv/boot/dts/microchip/mpfs.dtsi b/arch/riscv/boot/dts/microchip/mpfs.dtsi index 74493344ea41..6d9d455fa160 100644 --- a/arch/riscv/boot/dts/microchip/mpfs.dtsi +++ b/arch/riscv/boot/dts/microchip/mpfs.dtsi @@ -185,7 +185,7 @@ ranges; cctrllr: cache-controller@2010000 { - compatible = "sifive,fu540-c000-ccache", "cache"; + compatible = "microchip,mpfs-ccache", "sifive,fu540-c000-ccache", "cache"; reg = <0x0 0x2010000 0x0 0x1000>; cache-block-size = <64>; cache-level = <2>; From 35503ce12a2c3d5d9a94e3cd85a06739b0120f79 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Wed, 31 Aug 2022 14:40:41 +0200 Subject: [PATCH 035/158] perf script: Skip dummy event attr check Hongtao Yu reported problem when displaying uregs in perf script for system wide perf.data: # perf script -F uregs | head -10 Samples for 'dummy:HG' event do not have UREGS attribute set. Cannot print 'uregs' field. The problem is the extra dummy event added for system wide, which does not have proper sample_type setup. Skipping attr check completely for dummy event as suggested by Namhyung, because it does not have any samples anyway. Reported-by: Hongtao Yu Suggested-by: Namhyung Kim Signed-off-by: Jiri Olsa Acked-by: Ian Rogers Cc: Alexander Shishkin Cc: Ingo Molnar Cc: Mark Rutland Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20220831124041.219925-1-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-script.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index 13580a9c50b8..304d234d8e84 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -566,6 +566,8 @@ static struct evsel *find_first_output_type(struct evlist *evlist, struct evsel *evsel; evlist__for_each_entry(evlist, evsel) { + if (evsel__is_dummy_event(evsel)) + continue; if (output_type(evsel->core.attr.type) == (int)type) return evsel; } From c0a50cd389c3ed54831e240023dd12bafa56b3a6 Mon Sep 17 00:00:00 2001 From: Hannes Reinecke Date: Wed, 24 Aug 2022 08:00:33 +0200 Subject: [PATCH 036/158] scsi: lpfc: Return DID_TRANSPORT_DISRUPTED instead of DID_REQUEUE When the driver hits an internal error condition returning DID_REQUEUE the I/O will be retried on the same ITL nexus. This will inhibit multipathing, resulting in endless retries even if the error could have been resolved by using a different ITL nexus. Return DID_TRANSPORT_DISRUPTED to allow for multipath to engage and route I/O to another ITL nexus. Link: https://lore.kernel.org/r/20220824060033.138661-1-hare@suse.de Reviewed-by: James Smart Signed-off-by: Hannes Reinecke Signed-off-by: Martin K. Petersen --- drivers/scsi/lpfc/lpfc_scsi.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c index 084c0f9fdc3a..938a5e435943 100644 --- a/drivers/scsi/lpfc/lpfc_scsi.c +++ b/drivers/scsi/lpfc/lpfc_scsi.c @@ -4272,7 +4272,7 @@ lpfc_fcp_io_cmd_wqe_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pwqeIn, lpfc_cmd->result == IOERR_ABORT_REQUESTED || lpfc_cmd->result == IOERR_RPI_SUSPENDED || lpfc_cmd->result == IOERR_SLER_CMD_RCV_FAILURE) { - cmd->result = DID_REQUEUE << 16; + cmd->result = DID_TRANSPORT_DISRUPTED << 16; break; } if ((lpfc_cmd->result == IOERR_RX_DMA_FAILED || @@ -4562,7 +4562,7 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pIocbIn, lpfc_cmd->result == IOERR_NO_RESOURCES || lpfc_cmd->result == IOERR_ABORT_REQUESTED || lpfc_cmd->result == IOERR_SLER_CMD_RCV_FAILURE) { - cmd->result = DID_REQUEUE << 16; + cmd->result = DID_TRANSPORT_DISRUPTED << 16; break; } if ((lpfc_cmd->result == IOERR_RX_DMA_FAILED || From da6d507f5ff328f346b3c50e19e19993027b8ffd Mon Sep 17 00:00:00 2001 From: Yang Yingliang Date: Tue, 23 Aug 2022 12:42:37 +0800 Subject: [PATCH 037/158] scsi: lpfc: Add missing destroy_workqueue() in error path Add the missing destroy_workqueue() before return from lpfc_sli4_driver_resource_setup() in the error path. Link: https://lore.kernel.org/r/20220823044237.285643-1-yangyingliang@huawei.com Fixes: 3cee98db2610 ("scsi: lpfc: Fix crash on driver unload in wq free") Reviewed-by: James Smart Signed-off-by: Yang Yingliang Signed-off-by: Martin K. Petersen --- drivers/scsi/lpfc/lpfc_init.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c index c69c5a0979ec..55a1ad6eed03 100644 --- a/drivers/scsi/lpfc/lpfc_init.c +++ b/drivers/scsi/lpfc/lpfc_init.c @@ -8053,7 +8053,7 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba) /* Allocate device driver memory */ rc = lpfc_mem_alloc(phba, SGL_ALIGN_SZ); if (rc) - return -ENOMEM; + goto out_destroy_workqueue; /* IF Type 2 ports get initialized now. */ if (bf_get(lpfc_sli_intf_if_type, &phba->sli4_hba.sli_intf) >= @@ -8481,6 +8481,9 @@ out_free_bsmbx: lpfc_destroy_bootstrap_mbox(phba); out_free_mem: lpfc_mem_free(phba); +out_destroy_workqueue: + destroy_workqueue(phba->wq); + phba->wq = NULL; return rc; } From 2b36209ca8182c5e0221dd416106d434de433d34 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Sun, 21 Aug 2022 15:04:59 -0700 Subject: [PATCH 038/158] scsi: core: Revert "Call blk_mq_free_tag_set() earlier" Revert the patch series "Call blk_mq_free_tag_set() earlier" because it introduces a deadlock if the scsi_remove_host() caller holds a reference on a device, target or host. Link: https://lore.kernel.org/r/20220821220502.13685-2-bvanassche@acm.org Fixes: f323896fe6fa ("scsi: core: Call blk_mq_free_tag_set() earlier") Reported-by: syzbot+bafeb834708b1bb750bc@syzkaller.appspotmail.com Tested-by: Kenneth R. Crudup Signed-off-by: Bart Van Assche Signed-off-by: Martin K. Petersen --- drivers/scsi/hosts.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c index 0738238ed6cc..20c1f5420ba6 100644 --- a/drivers/scsi/hosts.c +++ b/drivers/scsi/hosts.c @@ -197,8 +197,6 @@ void scsi_remove_host(struct Scsi_Host *shost) * the dependent SCSI targets and devices are gone before returning. */ wait_event(shost->targets_wq, atomic_read(&shost->target_count) == 0); - - scsi_mq_destroy_tags(shost); } EXPORT_SYMBOL(scsi_remove_host); @@ -309,8 +307,8 @@ int scsi_add_host_with_dma(struct Scsi_Host *shost, struct device *dev, return error; /* - * Any resources associated with the SCSI host in this function except - * the tag set will be freed by scsi_host_dev_release(). + * Any host allocation in this function will be freed in + * scsi_host_dev_release(). */ out_del_dev: device_del(&shost->shost_dev); @@ -326,7 +324,6 @@ int scsi_add_host_with_dma(struct Scsi_Host *shost, struct device *dev, pm_runtime_disable(&shost->shost_gendev); pm_runtime_set_suspended(&shost->shost_gendev); pm_runtime_put_noidle(&shost->shost_gendev); - scsi_mq_destroy_tags(shost); fail: return error; } @@ -360,6 +357,9 @@ static void scsi_host_dev_release(struct device *dev) kfree(dev_name(&shost->shost_dev)); } + if (shost->tag_set.tags) + scsi_mq_destroy_tags(shost); + kfree(shost->shost_data); ida_free(&host_index_ida, shost->host_no); From 70e8d057bef5367709f4eda15b8d9a40d1b45e90 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Sun, 21 Aug 2022 15:05:00 -0700 Subject: [PATCH 039/158] scsi: core: Revert "Simplify LLD module reference counting" Revert the patch series "Call blk_mq_free_tag_set() earlier" because it introduces a deadlock if the scsi_remove_host() caller holds a reference on a device, target or host. Link: https://lore.kernel.org/r/20220821220502.13685-3-bvanassche@acm.org Fixes: 1a9283782df2 ("scsi: core: Simplify LLD module reference counting") Reported-by: syzbot+bafeb834708b1bb750bc@syzkaller.appspotmail.com Tested-by: Kenneth R. Crudup Signed-off-by: Bart Van Assche Signed-off-by: Martin K. Petersen --- drivers/scsi/scsi.c | 9 +++------ drivers/scsi/scsi_sysfs.c | 9 +++++++++ 2 files changed, 12 insertions(+), 6 deletions(-) diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c index 086ec5b5862d..c59eac7a32f2 100644 --- a/drivers/scsi/scsi.c +++ b/drivers/scsi/scsi.c @@ -586,13 +586,10 @@ EXPORT_SYMBOL(scsi_device_get); */ void scsi_device_put(struct scsi_device *sdev) { - /* - * Decreasing the module reference count before the device reference - * count is safe since scsi_remove_host() only returns after all - * devices have been removed. - */ - module_put(sdev->host->hostt->module); + struct module *mod = sdev->host->hostt->module; + put_device(&sdev->sdev_gendev); + module_put(mod); } EXPORT_SYMBOL(scsi_device_put); diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c index 9dad2fd5297f..282b32781e8c 100644 --- a/drivers/scsi/scsi_sysfs.c +++ b/drivers/scsi/scsi_sysfs.c @@ -452,6 +452,9 @@ static void scsi_device_dev_release_usercontext(struct work_struct *work) struct scsi_vpd *vpd_pg0 = NULL, *vpd_pg89 = NULL; struct scsi_vpd *vpd_pgb0 = NULL, *vpd_pgb1 = NULL, *vpd_pgb2 = NULL; unsigned long flags; + struct module *mod; + + mod = sdev->host->hostt->module; scsi_dh_release_device(sdev); @@ -518,11 +521,17 @@ static void scsi_device_dev_release_usercontext(struct work_struct *work) if (parent) put_device(parent); + module_put(mod); } static void scsi_device_dev_release(struct device *dev) { struct scsi_device *sdp = to_scsi_device(dev); + + /* Set module pointer as NULL in case of module unloading */ + if (!try_module_get(sdp->host->hostt->module)) + sdp->host->hostt->module = NULL; + execute_in_process_context(scsi_device_dev_release_usercontext, &sdp->ew); } From d94b2d00f7bfa0c555e7b8724b1faa037d57d150 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Sun, 21 Aug 2022 15:05:01 -0700 Subject: [PATCH 040/158] scsi: core: Revert "Make sure that hosts outlive targets" Revert the patch series "Call blk_mq_free_tag_set() earlier" because it introduces a deadlock if the scsi_remove_host() caller holds a reference on a device, target or host. Link: https://lore.kernel.org/r/20220821220502.13685-4-bvanassche@acm.org Fixes: 16728aaba62e ("scsi: core: Make sure that hosts outlive targets") Reported-by: syzbot+bafeb834708b1bb750bc@syzkaller.appspotmail.com Tested-by: Kenneth R. Crudup Signed-off-by: Bart Van Assche Signed-off-by: Martin K. Petersen --- drivers/scsi/hosts.c | 8 -------- drivers/scsi/scsi_scan.c | 7 ------- include/scsi/scsi_host.h | 3 --- 3 files changed, 18 deletions(-) diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c index 20c1f5420ba6..26bf3b153595 100644 --- a/drivers/scsi/hosts.c +++ b/drivers/scsi/hosts.c @@ -190,13 +190,6 @@ void scsi_remove_host(struct Scsi_Host *shost) transport_unregister_device(&shost->shost_gendev); device_unregister(&shost->shost_dev); device_del(&shost->shost_gendev); - - /* - * After scsi_remove_host() has returned the scsi LLD module can be - * unloaded and/or the host resources can be released. Hence wait until - * the dependent SCSI targets and devices are gone before returning. - */ - wait_event(shost->targets_wq, atomic_read(&shost->target_count) == 0); } EXPORT_SYMBOL(scsi_remove_host); @@ -406,7 +399,6 @@ struct Scsi_Host *scsi_host_alloc(struct scsi_host_template *sht, int privsize) INIT_LIST_HEAD(&shost->starved_list); init_waitqueue_head(&shost->host_wait); mutex_init(&shost->scan_mutex); - init_waitqueue_head(&shost->targets_wq); index = ida_alloc(&host_index_ida, GFP_KERNEL); if (index < 0) { diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c index ac6059702d13..4c1efd6a3b0c 100644 --- a/drivers/scsi/scsi_scan.c +++ b/drivers/scsi/scsi_scan.c @@ -406,14 +406,9 @@ static void scsi_target_destroy(struct scsi_target *starget) static void scsi_target_dev_release(struct device *dev) { struct device *parent = dev->parent; - struct Scsi_Host *shost = dev_to_shost(parent); struct scsi_target *starget = to_scsi_target(dev); kfree(starget); - - if (atomic_dec_return(&shost->target_count) == 0) - wake_up(&shost->targets_wq); - put_device(parent); } @@ -528,8 +523,6 @@ static struct scsi_target *scsi_alloc_target(struct device *parent, starget->max_target_blocked = SCSI_DEFAULT_TARGET_BLOCKED; init_waitqueue_head(&starget->sdev_wq); - atomic_inc(&shost->target_count); - retry: spin_lock_irqsave(shost->host_lock, flags); diff --git a/include/scsi/scsi_host.h b/include/scsi/scsi_host.h index aa7b7496c93a..b6e41ee3d566 100644 --- a/include/scsi/scsi_host.h +++ b/include/scsi/scsi_host.h @@ -690,9 +690,6 @@ struct Scsi_Host { /* ldm bits */ struct device shost_gendev, shost_dev; - atomic_t target_count; - wait_queue_head_t targets_wq; - /* * Points to the transport data (if any) which is allocated * separately From f782201ebc2b5f6c7a22e586e8eb81fe1d9e4801 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Sun, 21 Aug 2022 15:05:02 -0700 Subject: [PATCH 041/158] scsi: core: Revert "Make sure that targets outlive devices" Revert the patch series "Call blk_mq_free_tag_set() earlier" because it introduces a deadlock if the scsi_remove_host() caller holds a reference on a device, target or host. Link: https://lore.kernel.org/r/20220821220502.13685-5-bvanassche@acm.org Fixes: fe442604199e ("scsi: core: Make sure that targets outlive devices") Reported-by: syzbot+bafeb834708b1bb750bc@syzkaller.appspotmail.com Tested-by: Kenneth R. Crudup Signed-off-by: Bart Van Assche Signed-off-by: Martin K. Petersen --- drivers/scsi/scsi_scan.c | 2 -- drivers/scsi/scsi_sysfs.c | 20 +++----------------- include/scsi/scsi_device.h | 2 -- 3 files changed, 3 insertions(+), 21 deletions(-) diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c index 4c1efd6a3b0c..91ac901a6682 100644 --- a/drivers/scsi/scsi_scan.c +++ b/drivers/scsi/scsi_scan.c @@ -521,8 +521,6 @@ static struct scsi_target *scsi_alloc_target(struct device *parent, starget->state = STARGET_CREATED; starget->scsi_level = SCSI_2; starget->max_target_blocked = SCSI_DEFAULT_TARGET_BLOCKED; - init_waitqueue_head(&starget->sdev_wq); - retry: spin_lock_irqsave(shost->host_lock, flags); diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c index 282b32781e8c..aa70d9282161 100644 --- a/drivers/scsi/scsi_sysfs.c +++ b/drivers/scsi/scsi_sysfs.c @@ -443,9 +443,7 @@ static void scsi_device_cls_release(struct device *class_dev) static void scsi_device_dev_release_usercontext(struct work_struct *work) { - struct scsi_device *sdev = container_of(work, struct scsi_device, - ew.work); - struct scsi_target *starget = sdev->sdev_target; + struct scsi_device *sdev; struct device *parent; struct list_head *this, *tmp; struct scsi_vpd *vpd_pg80 = NULL, *vpd_pg83 = NULL; @@ -454,6 +452,8 @@ static void scsi_device_dev_release_usercontext(struct work_struct *work) unsigned long flags; struct module *mod; + sdev = container_of(work, struct scsi_device, ew.work); + mod = sdev->host->hostt->module; scsi_dh_release_device(sdev); @@ -516,9 +516,6 @@ static void scsi_device_dev_release_usercontext(struct work_struct *work) kfree(sdev->inquiry); kfree(sdev); - if (starget && atomic_dec_return(&starget->sdev_count) == 0) - wake_up(&starget->sdev_wq); - if (parent) put_device(parent); module_put(mod); @@ -1538,14 +1535,6 @@ static void __scsi_remove_target(struct scsi_target *starget) goto restart; } spin_unlock_irqrestore(shost->host_lock, flags); - - /* - * After scsi_remove_target() returns its caller can remove resources - * associated with @starget, e.g. an rport or session. Wait until all - * devices associated with @starget have been removed to prevent that - * a SCSI error handling callback function triggers a use-after-free. - */ - wait_event(starget->sdev_wq, atomic_read(&starget->sdev_count) == 0); } /** @@ -1656,9 +1645,6 @@ void scsi_sysfs_device_initialize(struct scsi_device *sdev) list_add_tail(&sdev->same_target_siblings, &starget->devices); list_add_tail(&sdev->siblings, &shost->__devices); spin_unlock_irqrestore(shost->host_lock, flags); - - atomic_inc(&starget->sdev_count); - /* * device can now only be removed via __scsi_remove_device() so hold * the target. Target will be held in CREATED state until something diff --git a/include/scsi/scsi_device.h b/include/scsi/scsi_device.h index 3113471ca375..2493bd65351a 100644 --- a/include/scsi/scsi_device.h +++ b/include/scsi/scsi_device.h @@ -309,8 +309,6 @@ struct scsi_target { struct list_head devices; struct device dev; struct kref reap_ref; /* last put renders target invisible */ - atomic_t sdev_count; - wait_queue_head_t sdev_wq; unsigned int channel; unsigned int id; /* target id ... replace * scsi_device.id eventually */ From 8fe4ce5836e932f5766317cb651c1ff2a4cd0506 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Thu, 25 Aug 2022 17:26:34 -0700 Subject: [PATCH 042/158] scsi: core: Fix a use-after-free There are two .exit_cmd_priv implementations. Both implementations use resources associated with the SCSI host. Make sure that these resources are still available when .exit_cmd_priv is called by waiting inside scsi_remove_host() until the tag set has been freed. This commit fixes the following use-after-free: ================================================================== BUG: KASAN: use-after-free in srp_exit_cmd_priv+0x27/0xd0 [ib_srp] Read of size 8 at addr ffff888100337000 by task multipathd/16727 Call Trace: dump_stack_lvl+0x34/0x44 print_report.cold+0x5e/0x5db kasan_report+0xab/0x120 srp_exit_cmd_priv+0x27/0xd0 [ib_srp] scsi_mq_exit_request+0x4d/0x70 blk_mq_free_rqs+0x143/0x410 __blk_mq_free_map_and_rqs+0x6e/0x100 blk_mq_free_tag_set+0x2b/0x160 scsi_host_dev_release+0xf3/0x1a0 device_release+0x54/0xe0 kobject_put+0xa5/0x120 device_release+0x54/0xe0 kobject_put+0xa5/0x120 scsi_device_dev_release_usercontext+0x4c1/0x4e0 execute_in_process_context+0x23/0x90 device_release+0x54/0xe0 kobject_put+0xa5/0x120 scsi_disk_release+0x3f/0x50 device_release+0x54/0xe0 kobject_put+0xa5/0x120 disk_release+0x17f/0x1b0 device_release+0x54/0xe0 kobject_put+0xa5/0x120 dm_put_table_device+0xa3/0x160 [dm_mod] dm_put_device+0xd0/0x140 [dm_mod] free_priority_group+0xd8/0x110 [dm_multipath] free_multipath+0x94/0xe0 [dm_multipath] dm_table_destroy+0xa2/0x1e0 [dm_mod] __dm_destroy+0x196/0x350 [dm_mod] dev_remove+0x10c/0x160 [dm_mod] ctl_ioctl+0x2c2/0x590 [dm_mod] dm_ctl_ioctl+0x5/0x10 [dm_mod] __x64_sys_ioctl+0xb4/0xf0 dm_ctl_ioctl+0x5/0x10 [dm_mod] __x64_sys_ioctl+0xb4/0xf0 do_syscall_64+0x3b/0x90 entry_SYSCALL_64_after_hwframe+0x46/0xb0 Link: https://lore.kernel.org/r/20220826002635.919423-1-bvanassche@acm.org Fixes: 65ca846a5314 ("scsi: core: Introduce {init,exit}_cmd_priv()") Cc: Ming Lei Cc: Christoph Hellwig Cc: Mike Christie Cc: Hannes Reinecke Cc: John Garry Cc: Li Zhijian Reported-by: Li Zhijian Tested-by: Li Zhijian Signed-off-by: Bart Van Assche Signed-off-by: Martin K. Petersen --- drivers/scsi/hosts.c | 16 +++++++++++++--- drivers/scsi/scsi_lib.c | 6 +++++- drivers/scsi/scsi_priv.h | 2 +- drivers/scsi/scsi_scan.c | 1 + drivers/scsi/scsi_sysfs.c | 1 + include/scsi/scsi_host.h | 2 ++ 6 files changed, 23 insertions(+), 5 deletions(-) diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c index 26bf3b153595..9857dba09c95 100644 --- a/drivers/scsi/hosts.c +++ b/drivers/scsi/hosts.c @@ -182,6 +182,15 @@ void scsi_remove_host(struct Scsi_Host *shost) mutex_unlock(&shost->scan_mutex); scsi_proc_host_rm(shost); + /* + * New SCSI devices cannot be attached anymore because of the SCSI host + * state so drop the tag set refcnt. Wait until the tag set refcnt drops + * to zero because .exit_cmd_priv implementations may need the host + * pointer. + */ + kref_put(&shost->tagset_refcnt, scsi_mq_free_tags); + wait_for_completion(&shost->tagset_freed); + spin_lock_irqsave(shost->host_lock, flags); if (scsi_host_set_state(shost, SHOST_DEL)) BUG_ON(scsi_host_set_state(shost, SHOST_DEL_RECOVERY)); @@ -245,6 +254,9 @@ int scsi_add_host_with_dma(struct Scsi_Host *shost, struct device *dev, if (error) goto fail; + kref_init(&shost->tagset_refcnt); + init_completion(&shost->tagset_freed); + /* * Increase usage count temporarily here so that calling * scsi_autopm_put_host() will trigger runtime idle if there is @@ -317,6 +329,7 @@ int scsi_add_host_with_dma(struct Scsi_Host *shost, struct device *dev, pm_runtime_disable(&shost->shost_gendev); pm_runtime_set_suspended(&shost->shost_gendev); pm_runtime_put_noidle(&shost->shost_gendev); + kref_put(&shost->tagset_refcnt, scsi_mq_free_tags); fail: return error; } @@ -350,9 +363,6 @@ static void scsi_host_dev_release(struct device *dev) kfree(dev_name(&shost->shost_dev)); } - if (shost->tag_set.tags) - scsi_mq_destroy_tags(shost); - kfree(shost->shost_data); ida_free(&host_index_ida, shost->host_no); diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index ef08029a0079..96e7e3eaca29 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -1983,9 +1983,13 @@ int scsi_mq_setup_tags(struct Scsi_Host *shost) return blk_mq_alloc_tag_set(tag_set); } -void scsi_mq_destroy_tags(struct Scsi_Host *shost) +void scsi_mq_free_tags(struct kref *kref) { + struct Scsi_Host *shost = container_of(kref, typeof(*shost), + tagset_refcnt); + blk_mq_free_tag_set(&shost->tag_set); + complete(&shost->tagset_freed); } /** diff --git a/drivers/scsi/scsi_priv.h b/drivers/scsi/scsi_priv.h index 429663bd78ec..f385b3f04d6e 100644 --- a/drivers/scsi/scsi_priv.h +++ b/drivers/scsi/scsi_priv.h @@ -94,7 +94,7 @@ extern void scsi_run_host_queues(struct Scsi_Host *shost); extern void scsi_requeue_run_queue(struct work_struct *work); extern void scsi_start_queue(struct scsi_device *sdev); extern int scsi_mq_setup_tags(struct Scsi_Host *shost); -extern void scsi_mq_destroy_tags(struct Scsi_Host *shost); +extern void scsi_mq_free_tags(struct kref *kref); extern void scsi_exit_queue(void); extern void scsi_evt_thread(struct work_struct *work); diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c index 91ac901a6682..5d27f5196de6 100644 --- a/drivers/scsi/scsi_scan.c +++ b/drivers/scsi/scsi_scan.c @@ -340,6 +340,7 @@ static struct scsi_device *scsi_alloc_sdev(struct scsi_target *starget, kfree(sdev); goto out; } + kref_get(&sdev->host->tagset_refcnt); sdev->request_queue = q; q->queuedata = sdev; __scsi_init_queue(sdev->host, q); diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c index aa70d9282161..5d61f58399dc 100644 --- a/drivers/scsi/scsi_sysfs.c +++ b/drivers/scsi/scsi_sysfs.c @@ -1476,6 +1476,7 @@ void __scsi_remove_device(struct scsi_device *sdev) mutex_unlock(&sdev->state_mutex); blk_mq_destroy_queue(sdev->request_queue); + kref_put(&sdev->host->tagset_refcnt, scsi_mq_free_tags); cancel_work_sync(&sdev->requeue_work); if (sdev->host->hostt->slave_destroy) diff --git a/include/scsi/scsi_host.h b/include/scsi/scsi_host.h index b6e41ee3d566..9b0a028bf053 100644 --- a/include/scsi/scsi_host.h +++ b/include/scsi/scsi_host.h @@ -557,6 +557,8 @@ struct Scsi_Host { struct scsi_host_template *hostt; struct scsi_transport_template *transportt; + struct kref tagset_refcnt; + struct completion tagset_freed; /* Area to keep a shared tag map */ struct blk_mq_tag_set tag_set; From 12f35199a2c0551187edbf8eb01379f0598659fa Mon Sep 17 00:00:00 2001 From: "yangx.jy@fujitsu.com" Date: Wed, 31 Aug 2022 08:16:29 +0000 Subject: [PATCH 043/158] RDMA/srp: Set scmnd->result only when scmnd is not NULL This change fixes the following kernel NULL pointer dereference which is reproduced by blktests srp/007 occasionally. BUG: kernel NULL pointer dereference, address: 0000000000000170 PGD 0 P4D 0 Oops: 0002 [#1] PREEMPT SMP NOPTI CPU: 0 PID: 9 Comm: kworker/0:1H Kdump: loaded Not tainted 6.0.0-rc1+ #37 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.15.0-29-g6a62e0cb0dfe-prebuilt.qemu.org 04/01/2014 Workqueue: 0x0 (kblockd) RIP: 0010:srp_recv_done+0x176/0x500 [ib_srp] Code: 00 4d 85 ff 0f 84 52 02 00 00 48 c7 82 80 02 00 00 00 00 00 00 4c 89 df 4c 89 14 24 e8 53 d3 4a f6 4c 8b 14 24 41 0f b6 42 13 <41> 89 87 70 01 00 00 41 0f b6 52 12 f6 c2 02 74 44 41 8b 42 1c b9 RSP: 0018:ffffaef7c0003e28 EFLAGS: 00000282 RAX: 0000000000000000 RBX: ffff9bc9486dea60 RCX: 0000000000000000 RDX: 0000000000000102 RSI: ffffffffb76bbd0e RDI: 00000000ffffffff RBP: ffff9bc980099a00 R08: 0000000000000001 R09: 0000000000000001 R10: ffff9bca53ef0000 R11: ffff9bc980099a10 R12: ffff9bc956e14000 R13: ffff9bc9836b9cb0 R14: ffff9bc9557b4480 R15: 0000000000000000 FS: 0000000000000000(0000) GS:ffff9bc97ec00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000170 CR3: 0000000007e04000 CR4: 00000000000006f0 Call Trace: __ib_process_cq+0xb7/0x280 [ib_core] ib_poll_handler+0x2b/0x130 [ib_core] irq_poll_softirq+0x93/0x150 __do_softirq+0xee/0x4b8 irq_exit_rcu+0xf7/0x130 sysvec_apic_timer_interrupt+0x8e/0xc0 Fixes: ad215aaea4f9 ("RDMA/srp: Make struct scsi_cmnd and struct srp_request adjacent") Link: https://lore.kernel.org/r/20220831081626.18712-1-yangx.jy@fujitsu.com Signed-off-by: Xiao Yang Acked-by: Bart Van Assche Signed-off-by: Leon Romanovsky --- drivers/infiniband/ulp/srp/ib_srp.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 7720ea270ed8..d7f69e593a63 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -1961,7 +1961,8 @@ static void srp_process_rsp(struct srp_rdma_ch *ch, struct srp_rsp *rsp) if (scmnd) { req = scsi_cmd_priv(scmnd); scmnd = srp_claim_req(ch, req, NULL, scmnd); - } else { + } + if (!scmnd) { shost_printk(KERN_ERR, target->scsi_host, "Null scmnd for RSP w/tag %#016llx received on ch %td / QP %#x\n", rsp->tag, ch - target->ch, ch->qp->qp_num); From 7e1afce5866e02b45bf88c27dd7de1b9dfade1cc Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Wed, 31 Aug 2022 14:59:00 +0200 Subject: [PATCH 044/158] ALSA: usb-audio: Inform the delayed registration more properly The info message that was added in the commit a4aad5636c72 ("ALSA: usb-audio: Inform devices that need delayed registration") is actually useful to know the need for the delayed registration. However, it turned out that this doesn't catch the all cases; namely, this warned only when a PCM stream is attached onto the existing PCM instance, but it doesn't count for a newly created PCM instance. This made confusion as if there were no further delayed registration. This patch moves the check to the code path for either adding a stream or creating a PCM instance. Also, make it simpler by checking the card->registered flag instead of querying each snd_device state. Fixes: a4aad5636c72 ("ALSA: usb-audio: Inform devices that need delayed registration") Link: https://bugzilla.kernel.org/show_bug.cgi?id=216082 Link: https://lore.kernel.org/r/20220831125901.4660-1-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/usb/stream.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/sound/usb/stream.c b/sound/usb/stream.c index ceb93d798182..40b7821c6c99 100644 --- a/sound/usb/stream.c +++ b/sound/usb/stream.c @@ -495,6 +495,10 @@ static int __snd_usb_add_audio_stream(struct snd_usb_audio *chip, return 0; } } + + if (chip->card->registered) + chip->need_delayed_register = true; + /* look for an empty stream */ list_for_each_entry(as, &chip->pcm_list, list) { if (as->fmt_type != fp->fmt_type) @@ -502,9 +506,6 @@ static int __snd_usb_add_audio_stream(struct snd_usb_audio *chip, subs = &as->substream[stream]; if (subs->ep_num) continue; - if (snd_device_get_state(chip->card, as->pcm) != - SNDRV_DEV_BUILD) - chip->need_delayed_register = true; err = snd_pcm_new_stream(as->pcm, stream, 1); if (err < 0) return err; From 2027f114686e0f3f1f39971964dfc618637c88c2 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Wed, 31 Aug 2022 14:59:01 +0200 Subject: [PATCH 045/158] ALSA: usb-audio: Register card again for iface over delayed_register option When the delayed registration is specified via either delayed_register option or the quirk, we delay the invocation of snd_card_register() until the given interface. But if a wrong value has been set there and there are more interfaces over the given interface number, snd_card_register() call would be missing for those interfaces. This patch catches up those missing calls by fixing the comparison of the interface number. Now the call is skipped only if the processed interface is less than the given interface, instead of the exact match. Fixes: b70038ef4fea ("ALSA: usb-audio: Add delayed_register option") Link: https://bugzilla.kernel.org/show_bug.cgi?id=216082 Link: https://lore.kernel.org/r/20220831125901.4660-2-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/usb/card.c | 2 +- sound/usb/quirks.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/usb/card.c b/sound/usb/card.c index d356743de2ff..706d249a9ad6 100644 --- a/sound/usb/card.c +++ b/sound/usb/card.c @@ -699,7 +699,7 @@ static bool check_delayed_register_option(struct snd_usb_audio *chip, int iface) if (delayed_register[i] && sscanf(delayed_register[i], "%x:%x", &id, &inum) == 2 && id == chip->usb_id) - return inum != iface; + return iface < inum; } return false; diff --git a/sound/usb/quirks.c b/sound/usb/quirks.c index 9bfead5efc4c..5b4d8f5eade2 100644 --- a/sound/usb/quirks.c +++ b/sound/usb/quirks.c @@ -1764,7 +1764,7 @@ bool snd_usb_registration_quirk(struct snd_usb_audio *chip, int iface) for (q = registration_quirks; q->usb_id; q++) if (chip->usb_id == q->usb_id) - return iface != q->interface; + return iface < q->interface; /* Register as normal */ return false; From ff878b408a03bef5d610b7e2302702e16a53636e Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 1 Sep 2022 14:41:36 +0200 Subject: [PATCH 046/158] ALSA: usb-audio: Split endpoint setups for hw_params and prepare One of the former changes for the endpoint management was the more consistent setup of endpoints at hw_params. snd_usb_endpoint_configure() is a single function that does the full setup, and it's called from both PCM hw_params and prepare callbacks. Although the EP setup at the prepare phase is usually skipped (by checking need_setup flag), it may be still effective in some cases like suspend/resume that requires the interface setup again. As it's a full and single setup, the invocation of snd_usb_endpoint_configure() includes not only the USB interface setup but also the buffer release and allocation. OTOH, doing the buffer release and re-allocation at PCM prepare phase is rather superfluous, and better to be done only in the hw_params phase. For those optimizations, this patch splits the endpoint setup to two phases: snd_usb_endpoint_set_params() and snd_usb_endpoint_prepare(), to be called from hw_params and from prepare, respectively. Note that this patch changes the driver operation slightly, effectively moving the USB interface setup again to PCM prepare stage instead of hw_params stage, while the buffer allocation and such initializations are still done at hw_params stage. And, the change of the USB interface setup timing (moving to prepare) gave an interesting "fix", too: it was reported that the recent kernels caused silent output at the beginning on playbacks on some devices on Android, and this change casually fixed the regression. It seems that those devices are picky about the sample rate change (or the interface change?), and don't follow the too immediate rate changes. Meanwhile, Android operates the PCM in the following order: - open, then hw_params with the possibly highest sample rate - close without prepare - re-open, hw_params with the normal sample rate - prepare, and start streaming This procedure ended up the hw_params twice with different rates, and because the recent kernel did set up the sample rate twice one and after, it screwed up the device. OTOH, the earlier kernels didn't set up the USB interface at hw_params, hence this problem didn't appear. Now, with this patch, the USB interface setup is again back to the prepare phase, and it works around the problem automagically. Although we should address the sample rate problem in a more solid way in future, let's keep things working as before for now. Fixes: bf6313a0ff76 ("ALSA: usb-audio: Refactor endpoint management") Cc: Reported-by: chihhao chen Link: https://lore.kernel.org/r/87e6d6ae69d68dc588ac9acc8c0f24d6188375c3.camel@mediatek.com Link: https://lore.kernel.org/r/20220901124136.4984-1-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/usb/endpoint.c | 23 +++++++++-------------- sound/usb/endpoint.h | 6 ++++-- sound/usb/pcm.c | 14 ++++++++++---- 3 files changed, 23 insertions(+), 20 deletions(-) diff --git a/sound/usb/endpoint.c b/sound/usb/endpoint.c index 0d7b73bf7945..a42f2ce19455 100644 --- a/sound/usb/endpoint.c +++ b/sound/usb/endpoint.c @@ -758,7 +758,8 @@ bool snd_usb_endpoint_compatible(struct snd_usb_audio *chip, * The endpoint needs to be closed via snd_usb_endpoint_close() later. * * Note that this function doesn't configure the endpoint. The substream - * needs to set it up later via snd_usb_endpoint_configure(). + * needs to set it up later via snd_usb_endpoint_set_params() and + * snd_usb_endpoint_prepare(). */ struct snd_usb_endpoint * snd_usb_endpoint_open(struct snd_usb_audio *chip, @@ -1290,12 +1291,13 @@ out_of_memory: /* * snd_usb_endpoint_set_params: configure an snd_usb_endpoint * + * It's called either from hw_params callback. * Determine the number of URBs to be used on this endpoint. * An endpoint must be configured before it can be started. * An endpoint that is already running can not be reconfigured. */ -static int snd_usb_endpoint_set_params(struct snd_usb_audio *chip, - struct snd_usb_endpoint *ep) +int snd_usb_endpoint_set_params(struct snd_usb_audio *chip, + struct snd_usb_endpoint *ep) { const struct audioformat *fmt = ep->cur_audiofmt; int err; @@ -1378,18 +1380,18 @@ static int init_sample_rate(struct snd_usb_audio *chip, } /* - * snd_usb_endpoint_configure: Configure the endpoint + * snd_usb_endpoint_prepare: Prepare the endpoint * * This function sets up the EP to be fully usable state. - * It's called either from hw_params or prepare callback. + * It's called either from prepare callback. * The function checks need_setup flag, and performs nothing unless needed, * so it's safe to call this multiple times. * * This returns zero if unchanged, 1 if the configuration has changed, * or a negative error code. */ -int snd_usb_endpoint_configure(struct snd_usb_audio *chip, - struct snd_usb_endpoint *ep) +int snd_usb_endpoint_prepare(struct snd_usb_audio *chip, + struct snd_usb_endpoint *ep) { bool iface_first; int err = 0; @@ -1410,9 +1412,6 @@ int snd_usb_endpoint_configure(struct snd_usb_audio *chip, if (err < 0) goto unlock; } - err = snd_usb_endpoint_set_params(chip, ep); - if (err < 0) - goto unlock; goto done; } @@ -1440,10 +1439,6 @@ int snd_usb_endpoint_configure(struct snd_usb_audio *chip, if (err < 0) goto unlock; - err = snd_usb_endpoint_set_params(chip, ep); - if (err < 0) - goto unlock; - err = snd_usb_select_mode_quirk(chip, ep->cur_audiofmt); if (err < 0) goto unlock; diff --git a/sound/usb/endpoint.h b/sound/usb/endpoint.h index 6a9af04cf175..e67ea28faa54 100644 --- a/sound/usb/endpoint.h +++ b/sound/usb/endpoint.h @@ -17,8 +17,10 @@ snd_usb_endpoint_open(struct snd_usb_audio *chip, bool is_sync_ep); void snd_usb_endpoint_close(struct snd_usb_audio *chip, struct snd_usb_endpoint *ep); -int snd_usb_endpoint_configure(struct snd_usb_audio *chip, - struct snd_usb_endpoint *ep); +int snd_usb_endpoint_set_params(struct snd_usb_audio *chip, + struct snd_usb_endpoint *ep); +int snd_usb_endpoint_prepare(struct snd_usb_audio *chip, + struct snd_usb_endpoint *ep); int snd_usb_endpoint_get_clock_rate(struct snd_usb_audio *chip, int clock); bool snd_usb_endpoint_compatible(struct snd_usb_audio *chip, diff --git a/sound/usb/pcm.c b/sound/usb/pcm.c index d45d1d7e6664..b604f7e95e82 100644 --- a/sound/usb/pcm.c +++ b/sound/usb/pcm.c @@ -443,17 +443,17 @@ static int configure_endpoints(struct snd_usb_audio *chip, if (stop_endpoints(subs, false)) sync_pending_stops(subs); if (subs->sync_endpoint) { - err = snd_usb_endpoint_configure(chip, subs->sync_endpoint); + err = snd_usb_endpoint_prepare(chip, subs->sync_endpoint); if (err < 0) return err; } - err = snd_usb_endpoint_configure(chip, subs->data_endpoint); + err = snd_usb_endpoint_prepare(chip, subs->data_endpoint); if (err < 0) return err; snd_usb_set_format_quirk(subs, subs->cur_audiofmt); } else { if (subs->sync_endpoint) { - err = snd_usb_endpoint_configure(chip, subs->sync_endpoint); + err = snd_usb_endpoint_prepare(chip, subs->sync_endpoint); if (err < 0) return err; } @@ -551,7 +551,13 @@ static int snd_usb_hw_params(struct snd_pcm_substream *substream, subs->cur_audiofmt = fmt; mutex_unlock(&chip->mutex); - ret = configure_endpoints(chip, subs); + if (subs->sync_endpoint) { + ret = snd_usb_endpoint_set_params(chip, subs->sync_endpoint); + if (ret < 0) + goto unlock; + } + + ret = snd_usb_endpoint_set_params(chip, subs->data_endpoint); unlock: if (ret < 0) From 6b66ca0bac1b9cee7608d7c4dc59b699458b4cb8 Mon Sep 17 00:00:00 2001 From: Yicong Yang Date: Thu, 25 Aug 2022 17:20:07 +0800 Subject: [PATCH 047/158] arch_topology: Make cluster topology span at least SMT CPUs Currently cpu_clustergroup_mask() will return CPU mask if cluster span more or the same CPUs as cpu_coregroup_mask(). This will result topology borken on non-Cluster SMT machines when building with CONFIG_SCHED_CLUSTER=y. Test with: qemu-system-aarch64 -enable-kvm -machine virt \ -net none \ -cpu host \ -bios ./QEMU_EFI.fd \ -m 2G \ -smp 48,sockets=2,cores=12,threads=2 \ -kernel $Image \ -initrd $Rootfs \ -nographic \ -append "rdinit=init console=ttyAMA0 sched_verbose loglevel=8" We'll get below error: [ 3.084568] BUG: arch topology borken [ 3.084570] the SMT domain not a subset of the CLS domain Since cluster is a level higher than SMT, fix this by making cluster spans at least SMT CPUs. Fixes: bfcc4397435d ("arch_topology: Limit span of cpu_clustergroup_mask()") Cc: Sudeep Holla Cc: Vincent Guittot Cc: Ionela Voinescu Reviewed-by: Sudeep Holla Reviewed-by: Ionela Voinescu Signed-off-by: Yicong Yang Link: https://lore.kernel.org/r/20220825092007.8129-1-yangyicong@huawei.com Signed-off-by: Greg Kroah-Hartman --- drivers/base/arch_topology.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/base/arch_topology.c b/drivers/base/arch_topology.c index eaa1b8d2d39d..59880b5d9fe2 100644 --- a/drivers/base/arch_topology.c +++ b/drivers/base/arch_topology.c @@ -724,7 +724,7 @@ const struct cpumask *cpu_clustergroup_mask(int cpu) */ if (cpumask_subset(cpu_coregroup_mask(cpu), &cpu_topology[cpu].cluster_sibling)) - return get_cpu_mask(cpu); + return cpu_smt_mask(cpu); return &cpu_topology[cpu].cluster_sibling; } From aded3cad909581c60335037112c4f86bbfe90f17 Mon Sep 17 00:00:00 2001 From: Sander Vanheule Date: Sun, 21 Aug 2022 17:01:47 +0200 Subject: [PATCH 048/158] kunit: fix assert_type for comparison macros When replacing KUNIT_BINARY_*_MSG_ASSERTION() macros with KUNIT_BINARY_INT_ASSERTION(), the assert_type parameter was not always correctly transferred. Specifically, the following errors were introduced: - KUNIT_EXPECT_LE_MSG() uses KUNIT_ASSERTION - KUNIT_ASSERT_LT_MSG() uses KUNIT_EXPECTATION - KUNIT_ASSERT_GT_MSG() uses KUNIT_EXPECTATION A failing KUNIT_EXPECT_LE_MSG() test thus prevents further tests from running, while failing KUNIT_ASSERT_{LT,GT}_MSG() tests do not prevent further tests from running. This is contrary to the documentation, which states that failing KUNIT_EXPECT_* macros allow further tests to run, while failing KUNIT_ASSERT_* macros should prevent this. Revert the KUNIT_{ASSERTION,EXPECTATION} switches to fix the behaviour for the affected macros. Fixes: 40f39777ce4f ("kunit: decrease macro layering for integer asserts") Signed-off-by: Sander Vanheule Reviewed-by: Daniel Latypov Reviewed-by: Brendan Higgins Signed-off-by: Shuah Khan --- include/kunit/test.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/include/kunit/test.h b/include/kunit/test.h index c958855681cc..840a2c375065 100644 --- a/include/kunit/test.h +++ b/include/kunit/test.h @@ -826,7 +826,7 @@ do { \ #define KUNIT_EXPECT_LE_MSG(test, left, right, fmt, ...) \ KUNIT_BINARY_INT_ASSERTION(test, \ - KUNIT_ASSERTION, \ + KUNIT_EXPECTATION, \ left, <=, right, \ fmt, \ ##__VA_ARGS__) @@ -1116,7 +1116,7 @@ do { \ #define KUNIT_ASSERT_LT_MSG(test, left, right, fmt, ...) \ KUNIT_BINARY_INT_ASSERTION(test, \ - KUNIT_EXPECTATION, \ + KUNIT_ASSERTION, \ left, <, right, \ fmt, \ ##__VA_ARGS__) @@ -1157,7 +1157,7 @@ do { \ #define KUNIT_ASSERT_GT_MSG(test, left, right, fmt, ...) \ KUNIT_BINARY_INT_ASSERTION(test, \ - KUNIT_EXPECTATION, \ + KUNIT_ASSERTION, \ left, >, right, \ fmt, \ ##__VA_ARGS__) From 793f55b2971e3a95d77ad08e9da2a3dc6c946cd7 Mon Sep 17 00:00:00 2001 From: Nico Pache Date: Wed, 10 Aug 2022 19:40:56 -0400 Subject: [PATCH 049/158] kunit: fix Kconfig for build-in tests USB4 and Nitro Enclaves Both the USB4 and Nitro Enclaves KUNIT tests are now able to be compiled if KUNIT is compiled as a module. This leads to issues if KUNIT is being packaged separately from the core kernel and when KUNIT is run baremetal without the required driver compiled into the kernel. Fixes: 635dcd16844b ("thunderbolt: test: Use kunit_test_suite() macro") Fixes: fe5be808fa6c ("nitro_enclaves: test: Use kunit_test_suite() macro") Signed-off-by: Nico Pache Reviewed-by: David Gow Reviewed-by: Andra Paraschiv Acked-by: Brendan Higgins Signed-off-by: Shuah Khan --- drivers/thunderbolt/Kconfig | 3 +-- drivers/virt/nitro_enclaves/Kconfig | 2 +- 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/thunderbolt/Kconfig b/drivers/thunderbolt/Kconfig index e76a6c173637..f12d0a3ee3e2 100644 --- a/drivers/thunderbolt/Kconfig +++ b/drivers/thunderbolt/Kconfig @@ -29,8 +29,7 @@ config USB4_DEBUGFS_WRITE config USB4_KUNIT_TEST bool "KUnit tests" if !KUNIT_ALL_TESTS - depends on (USB4=m || KUNIT=y) - depends on KUNIT + depends on USB4 && KUNIT=y default KUNIT_ALL_TESTS config USB4_DMA_TEST diff --git a/drivers/virt/nitro_enclaves/Kconfig b/drivers/virt/nitro_enclaves/Kconfig index ce91add81401..dc4d25c26256 100644 --- a/drivers/virt/nitro_enclaves/Kconfig +++ b/drivers/virt/nitro_enclaves/Kconfig @@ -17,7 +17,7 @@ config NITRO_ENCLAVES config NITRO_ENCLAVES_MISC_DEV_TEST bool "Tests for the misc device functionality of the Nitro Enclaves" if !KUNIT_ALL_TESTS - depends on NITRO_ENCLAVES && KUNIT + depends on NITRO_ENCLAVES && KUNIT=y default KUNIT_ALL_TESTS help Enable KUnit tests for the misc device functionality of the Nitro From 3e48940abee88b8dbbeeaf8a07e7b2b6be1271b3 Mon Sep 17 00:00:00 2001 From: Pattara Teerapong Date: Thu, 1 Sep 2022 14:40:36 +0000 Subject: [PATCH 050/158] ALSA: aloop: Fix random zeros in capture data when using jiffies timer In loopback_jiffies_timer_pos_update(), we are getting jiffies twice. First time for playback, second time for capture. Jiffies can be updated between these two calls and if the capture jiffies is larger, extra zeros will be filled in the capture buffer. Change to get jiffies once and use it for both playback and capture. Signed-off-by: Pattara Teerapong Cc: Link: https://lore.kernel.org/r/20220901144036.4049060-1-pteerapong@chromium.org Signed-off-by: Takashi Iwai --- sound/drivers/aloop.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/sound/drivers/aloop.c b/sound/drivers/aloop.c index 9b4a7cdb103a..12f12a294df5 100644 --- a/sound/drivers/aloop.c +++ b/sound/drivers/aloop.c @@ -605,17 +605,18 @@ static unsigned int loopback_jiffies_timer_pos_update cable->streams[SNDRV_PCM_STREAM_PLAYBACK]; struct loopback_pcm *dpcm_capt = cable->streams[SNDRV_PCM_STREAM_CAPTURE]; - unsigned long delta_play = 0, delta_capt = 0; + unsigned long delta_play = 0, delta_capt = 0, cur_jiffies; unsigned int running, count1, count2; + cur_jiffies = jiffies; running = cable->running ^ cable->pause; if (running & (1 << SNDRV_PCM_STREAM_PLAYBACK)) { - delta_play = jiffies - dpcm_play->last_jiffies; + delta_play = cur_jiffies - dpcm_play->last_jiffies; dpcm_play->last_jiffies += delta_play; } if (running & (1 << SNDRV_PCM_STREAM_CAPTURE)) { - delta_capt = jiffies - dpcm_capt->last_jiffies; + delta_capt = cur_jiffies - dpcm_capt->last_jiffies; dpcm_capt->last_jiffies += delta_capt; } From c7943bb324e503baeeba3df2bc5ca8a377111bfa Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Sat, 27 Aug 2022 00:34:51 +0300 Subject: [PATCH 051/158] drm/edid: Handle EDID 1.4 range descriptor h/vfreq offsets MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit EDID 1.4 introduced some extra flags in the range descriptor to support min/max h/vfreq >= 255. Consult them to correctly parse the vfreq limits. Note that some combinations of the flags are documented as "reserved" (as are some other values in the descriptor) but explicitly checking for those doesn't seem particularly worthwile since we end up with bogus results whether we decode them or not. v2: Increase the storage to u16 to make it work (Jani) Note the "reserved" values situation (Jani) v3: Document the EDID version number in the defines Drop some bogus (u8) casts Cc: stable@vger.kernel.org Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/6519 References: https://gitlab.freedesktop.org/drm/intel/-/issues/6484 Reviewed-by: Jani Nikula Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20220826213501.31490-2-ville.syrjala@linux.intel.com Reviewed-by: Manasi Navare --- drivers/gpu/drm/drm_debugfs.c | 4 ++-- drivers/gpu/drm/drm_edid.c | 24 ++++++++++++++++++------ include/drm/drm_connector.h | 4 ++-- include/drm/drm_edid.h | 5 +++++ 4 files changed, 27 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/drm_debugfs.c b/drivers/gpu/drm/drm_debugfs.c index 493922069c90..01ee3febb813 100644 --- a/drivers/gpu/drm/drm_debugfs.c +++ b/drivers/gpu/drm/drm_debugfs.c @@ -377,8 +377,8 @@ static int vrr_range_show(struct seq_file *m, void *data) if (connector->status != connector_status_connected) return -ENODEV; - seq_printf(m, "Min: %u\n", (u8)connector->display_info.monitor_range.min_vfreq); - seq_printf(m, "Max: %u\n", (u8)connector->display_info.monitor_range.max_vfreq); + seq_printf(m, "Min: %u\n", connector->display_info.monitor_range.min_vfreq); + seq_printf(m, "Max: %u\n", connector->display_info.monitor_range.max_vfreq); return 0; } diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c index bbc25e3b7220..eaa819381281 100644 --- a/drivers/gpu/drm/drm_edid.c +++ b/drivers/gpu/drm/drm_edid.c @@ -5971,12 +5971,14 @@ static void drm_parse_cea_ext(struct drm_connector *connector, } static -void get_monitor_range(const struct detailed_timing *timing, - void *info_monitor_range) +void get_monitor_range(const struct detailed_timing *timing, void *c) { - struct drm_monitor_range_info *monitor_range = info_monitor_range; + struct detailed_mode_closure *closure = c; + struct drm_display_info *info = &closure->connector->display_info; + struct drm_monitor_range_info *monitor_range = &info->monitor_range; const struct detailed_non_pixel *data = &timing->data.other_data; const struct detailed_data_monitor_range *range = &data->data.range; + const struct edid *edid = closure->drm_edid->edid; if (!is_display_descriptor(timing, EDID_DETAIL_MONITOR_RANGE)) return; @@ -5992,18 +5994,28 @@ void get_monitor_range(const struct detailed_timing *timing, monitor_range->min_vfreq = range->min_vfreq; monitor_range->max_vfreq = range->max_vfreq; + + if (edid->revision >= 4) { + if (data->pad2 & DRM_EDID_RANGE_OFFSET_MIN_VFREQ) + monitor_range->min_vfreq += 255; + if (data->pad2 & DRM_EDID_RANGE_OFFSET_MAX_VFREQ) + monitor_range->max_vfreq += 255; + } } static void drm_get_monitor_range(struct drm_connector *connector, const struct drm_edid *drm_edid) { - struct drm_display_info *info = &connector->display_info; + const struct drm_display_info *info = &connector->display_info; + struct detailed_mode_closure closure = { + .connector = connector, + .drm_edid = drm_edid, + }; if (!version_greater(drm_edid, 1, 1)) return; - drm_for_each_detailed_block(drm_edid, get_monitor_range, - &info->monitor_range); + drm_for_each_detailed_block(drm_edid, get_monitor_range, &closure); DRM_DEBUG_KMS("Supported Monitor Refresh rate range is %d Hz - %d Hz\n", info->monitor_range.min_vfreq, diff --git a/include/drm/drm_connector.h b/include/drm/drm_connector.h index a1705d6b3fba..7df7876b2ad5 100644 --- a/include/drm/drm_connector.h +++ b/include/drm/drm_connector.h @@ -319,8 +319,8 @@ enum drm_panel_orientation { * EDID's detailed monitor range */ struct drm_monitor_range_info { - u8 min_vfreq; - u8 max_vfreq; + u16 min_vfreq; + u16 max_vfreq; }; /** diff --git a/include/drm/drm_edid.h b/include/drm/drm_edid.h index 2181977ae683..1ed61e2b30a4 100644 --- a/include/drm/drm_edid.h +++ b/include/drm/drm_edid.h @@ -92,6 +92,11 @@ struct detailed_data_string { u8 str[13]; } __attribute__((packed)); +#define DRM_EDID_RANGE_OFFSET_MIN_VFREQ (1 << 0) /* 1.4 */ +#define DRM_EDID_RANGE_OFFSET_MAX_VFREQ (1 << 1) /* 1.4 */ +#define DRM_EDID_RANGE_OFFSET_MIN_HFREQ (1 << 2) /* 1.4 */ +#define DRM_EDID_RANGE_OFFSET_MAX_HFREQ (1 << 3) /* 1.4 */ + #define DRM_EDID_DEFAULT_GTF_SUPPORT_FLAG 0x00 #define DRM_EDID_RANGE_LIMITS_ONLY_FLAG 0x01 #define DRM_EDID_SECONDARY_GTF_SUPPORT_FLAG 0x02 From f0c86a2bae4fd12bfa8bad4d43fb59fb498cdd14 Mon Sep 17 00:00:00 2001 From: Zhengjun Xing Date: Fri, 26 Aug 2022 22:00:57 +0800 Subject: [PATCH 052/158] perf stat: Fix L2 Topdown metrics disappear for raw events MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In perf/Documentation/perf-stat.txt, for "--td-level" the default "0" means the max level that the current hardware support. So we need initialize the stat_config.topdown_level to TOPDOWN_MAX_LEVEL when “--td-level=0” or no “--td-level” option. Otherwise, for the hardware with a max level is 2, the 2nd level metrics disappear for raw events in this case. The issue cannot be observed for the perf stat default or "--topdown" options. This commit fixes the raw events issue and removes the duplicated code for the perf stat default. Before: # ./perf stat -e "cpu-clock,context-switches,cpu-migrations,page-faults,instructions,cycles,ref-cycles,branches,branch-misses,{slots,topdown-retiring,topdown-bad-spec,topdown-fe-bound,topdown-be-bound,topdown-heavy-ops,topdown-br-mispredict,topdown-fetch-lat,topdown-mem-bound}" sleep 1 Performance counter stats for 'sleep 1': 1.03 msec cpu-clock # 0.001 CPUs utilized 1 context-switches # 966.216 /sec 0 cpu-migrations # 0.000 /sec 60 page-faults # 57.973 K/sec 1,132,112 instructions # 1.41 insn per cycle 803,872 cycles # 0.777 GHz 1,909,120 ref-cycles # 1.845 G/sec 236,634 branches # 228.640 M/sec 6,367 branch-misses # 2.69% of all branches 4,823,232 slots # 4.660 G/sec 1,210,536 topdown-retiring # 25.1% Retiring 699,841 topdown-bad-spec # 14.5% Bad Speculation 1,777,975 topdown-fe-bound # 36.9% Frontend Bound 1,134,878 topdown-be-bound # 23.5% Backend Bound 189,146 topdown-heavy-ops # 182.756 M/sec 662,012 topdown-br-mispredict # 639.647 M/sec 1,097,048 topdown-fetch-lat # 1.060 G/sec 416,121 topdown-mem-bound # 402.063 M/sec 1.002423690 seconds time elapsed 0.002494000 seconds user 0.000000000 seconds sys After: # ./perf stat -e "cpu-clock,context-switches,cpu-migrations,page-faults,instructions,cycles,ref-cycles,branches,branch-misses,{slots,topdown-retiring,topdown-bad-spec,topdown-fe-bound,topdown-be-bound,topdown-heavy-ops,topdown-br-mispredict,topdown-fetch-lat,topdown-mem-bound}" sleep 1 Performance counter stats for 'sleep 1': 1.13 msec cpu-clock # 0.001 CPUs utilized 1 context-switches # 882.128 /sec 0 cpu-migrations # 0.000 /sec 61 page-faults # 53.810 K/sec 1,137,612 instructions # 1.29 insn per cycle 881,477 cycles # 0.778 GHz 2,093,496 ref-cycles # 1.847 G/sec 236,356 branches # 208.496 M/sec 7,090 branch-misses # 3.00% of all branches 5,288,862 slots # 4.665 G/sec 1,223,697 topdown-retiring # 23.1% Retiring 767,403 topdown-bad-spec # 14.5% Bad Speculation 2,053,322 topdown-fe-bound # 38.8% Frontend Bound 1,244,438 topdown-be-bound # 23.5% Backend Bound 186,665 topdown-heavy-ops # 3.5% Heavy Operations # 19.6% Light Operations 725,922 topdown-br-mispredict # 13.7% Branch Mispredict # 0.8% Machine Clears 1,327,400 topdown-fetch-lat # 25.1% Fetch Latency # 13.7% Fetch Bandwidth 497,775 topdown-mem-bound # 9.4% Memory Bound # 14.1% Core Bound 1.002701530 seconds time elapsed 0.002744000 seconds user 0.000000000 seconds sys Fixes: 63e39aa6ae103451 ("perf stat: Support L2 Topdown events") Reviewed-by: Kan Liang Signed-off-by: Xing Zhengjun Cc: Alexander Shishkin Cc: Andi Kleen Cc: Ian Rogers Cc: Ingo Molnar Cc: Jiri Olsa Cc: Kan Liang Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20220826140057.3289401-1-zhengjun.xing@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-stat.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index 54cd29d07ca8..0b4a62e4ff67 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -1932,6 +1932,9 @@ setup_metrics: free(str); } + if (!stat_config.topdown_level) + stat_config.topdown_level = TOPDOWN_MAX_LEVEL; + if (!evsel_list->core.nr_entries) { if (target__has_cpu(&target)) default_attrs0[0].config = PERF_COUNT_SW_CPU_CLOCK; @@ -1948,8 +1951,6 @@ setup_metrics: } if (evlist__add_default_attrs(evsel_list, default_attrs1) < 0) return -1; - - stat_config.topdown_level = TOPDOWN_MAX_LEVEL; /* Platform specific attrs */ if (evlist__add_default_attrs(evsel_list, default_null_attrs) < 0) return -1; From c749b275056d4d1023af125b320c91a24d6856b8 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Sat, 3 Sep 2022 08:05:08 +0200 Subject: [PATCH 053/158] Revert "arch_topology: Make cluster topology span at least SMT CPUs" This reverts commit 6b66ca0bac1b9cee7608d7c4dc59b699458b4cb8 as it breaks the build on some arches as reported by the kernel test robot. Link: https://lore.kernel.org/r/202209030824.SouwDV5M-lkp@intel.com Reported-by: kernel test robot Fixes: 6b66ca0bac1b ("arch_topology: Make cluster topology span at least SMT CPUs") Cc: Sudeep Holla Cc: Vincent Guittot Cc: Ionela Voinescu Cc: Yicong Yang Signed-off-by: Greg Kroah-Hartman --- drivers/base/arch_topology.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/base/arch_topology.c b/drivers/base/arch_topology.c index 59880b5d9fe2..eaa1b8d2d39d 100644 --- a/drivers/base/arch_topology.c +++ b/drivers/base/arch_topology.c @@ -724,7 +724,7 @@ const struct cpumask *cpu_clustergroup_mask(int cpu) */ if (cpumask_subset(cpu_coregroup_mask(cpu), &cpu_topology[cpu].cluster_sibling)) - return cpu_smt_mask(cpu); + return get_cpu_mask(cpu); return &cpu_topology[cpu].cluster_sibling; } From 748008e1da926a814cc0a054c81ca614408b1b0c Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Tue, 23 Aug 2022 18:38:19 +0800 Subject: [PATCH 054/158] block: don't add partitions if GD_SUPPRESS_PART_SCAN is set Commit b9684a71fca7 ("block, loop: support partitions without scanning") adds GD_SUPPRESS_PART_SCAN for replacing part function of GENHD_FL_NO_PART. But looks blk_add_partitions() is missed, since loop doesn't want to add partitions if GENHD_FL_NO_PART was set. And it causes regression on libblockdev (as called from udisks) which operates with the LO_FLAGS_PARTSCAN. Fixes the issue by not adding partitions if GD_SUPPRESS_PART_SCAN is set. Fixes: b9684a71fca7 ("block, loop: support partitions without scanning") Signed-off-by: Ming Lei Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/r/20220823103819.395776-1-ming.lei@redhat.com Signed-off-by: Jens Axboe --- block/partitions/core.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/block/partitions/core.c b/block/partitions/core.c index fc1d70384825..b8112f52d388 100644 --- a/block/partitions/core.c +++ b/block/partitions/core.c @@ -596,6 +596,9 @@ static int blk_add_partitions(struct gendisk *disk) if (disk->flags & GENHD_FL_NO_PART) return 0; + if (test_bit(GD_SUPPRESS_PART_SCAN, &disk->state)) + return 0; + state = check_partition(disk); if (!state) return 0; From 0d1b756acf60da5004c1e20ca4462f0c257bf6e1 Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Fri, 2 Sep 2022 23:59:18 +0200 Subject: [PATCH 055/158] RDMA/siw: Pass a pointer to virt_to_page() Functions that work on a pointer to virtual memory such as virt_to_pfn() and users of that function such as virt_to_page() are supposed to pass a pointer to virtual memory, ideally a (void *) or other pointer. However since many architectures implement virt_to_pfn() as a macro, this function becomes polymorphic and accepts both a (unsigned long) and a (void *). If we instead implement a proper virt_to_pfn(void *addr) function the following happens (occurred on arch/arm): drivers/infiniband/sw/siw/siw_qp_tx.c:32:23: warning: incompatible integer to pointer conversion passing 'dma_addr_t' (aka 'unsigned int') to parameter of type 'const void *' [-Wint-conversion] drivers/infiniband/sw/siw/siw_qp_tx.c:32:37: warning: passing argument 1 of 'virt_to_pfn' makes pointer from integer without a cast [-Wint-conversion] drivers/infiniband/sw/siw/siw_qp_tx.c:538:36: warning: incompatible integer to pointer conversion passing 'unsigned long long' to parameter of type 'const void *' [-Wint-conversion] Fix this with an explicit cast. In one case where the SIW SGE uses an unaligned u64 we need a double cast modifying the virtual address (va) to a platform-specific uintptr_t before casting to a (void *). Fixes: b9be6f18cf9e ("rdma/siw: transmit path") Cc: linux-rdma@vger.kernel.org Signed-off-by: Linus Walleij Link: https://lore.kernel.org/r/20220902215918.603761-1-linus.walleij@linaro.org Signed-off-by: Leon Romanovsky --- drivers/infiniband/sw/siw/siw_qp_tx.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/sw/siw/siw_qp_tx.c b/drivers/infiniband/sw/siw/siw_qp_tx.c index 1f4e60257700..7d47b521070b 100644 --- a/drivers/infiniband/sw/siw/siw_qp_tx.c +++ b/drivers/infiniband/sw/siw/siw_qp_tx.c @@ -29,7 +29,7 @@ static struct page *siw_get_pblpage(struct siw_mem *mem, u64 addr, int *idx) dma_addr_t paddr = siw_pbl_get_buffer(pbl, offset, NULL, idx); if (paddr) - return virt_to_page(paddr); + return virt_to_page((void *)paddr); return NULL; } @@ -533,13 +533,23 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx, struct socket *s) kunmap_local(kaddr); } } else { - u64 va = sge->laddr + sge_off; + /* + * Cast to an uintptr_t to preserve all 64 bits + * in sge->laddr. + */ + uintptr_t va = (uintptr_t)(sge->laddr + sge_off); - page_array[seg] = virt_to_page(va & PAGE_MASK); + /* + * virt_to_page() takes a (void *) pointer + * so cast to a (void *) meaning it will be 64 + * bits on a 64 bit platform and 32 bits on a + * 32 bit platform. + */ + page_array[seg] = virt_to_page((void *)(va & PAGE_MASK)); if (do_crc) crypto_shash_update( c_tx->mpa_crc_hd, - (void *)(uintptr_t)va, + (void *)va, plen); } From 414d38ba871092aeac4ed097ac4ced89486646f7 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Sun, 4 Sep 2022 09:27:50 +0200 Subject: [PATCH 056/158] ALSA: hda/sigmatel: Keep power up while beep is enabled It seems that the beep playback doesn't work well on IDT codec devices when the codec auto-pm is enabled. Keep the power on while the beep switch is enabled. Link: https://bugzilla.suse.com/show_bug.cgi?id=1200544 Link: https://lore.kernel.org/r/20220904072750.26164-1-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_sigmatel.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c index 61df4d33c48f..066bfccd2587 100644 --- a/sound/pci/hda/patch_sigmatel.c +++ b/sound/pci/hda/patch_sigmatel.c @@ -209,6 +209,7 @@ struct sigmatel_spec { /* beep widgets */ hda_nid_t anabeep_nid; + bool beep_power_on; /* SPDIF-out mux */ const char * const *spdif_labels; @@ -4443,6 +4444,26 @@ static int stac_suspend(struct hda_codec *codec) return 0; } + +static int stac_check_power_status(struct hda_codec *codec, hda_nid_t nid) +{ + struct sigmatel_spec *spec = codec->spec; + int ret = snd_hda_gen_check_power_status(codec, nid); + +#ifdef CONFIG_SND_HDA_INPUT_BEEP + if (nid == spec->gen.beep_nid && codec->beep) { + if (codec->beep->enabled != spec->beep_power_on) { + spec->beep_power_on = codec->beep->enabled; + if (spec->beep_power_on) + snd_hda_power_up_pm(codec); + else + snd_hda_power_down_pm(codec); + } + ret |= spec->beep_power_on; + } +#endif + return ret; +} #else #define stac_suspend NULL #endif /* CONFIG_PM */ @@ -4455,6 +4476,7 @@ static const struct hda_codec_ops stac_patch_ops = { .unsol_event = snd_hda_jack_unsol_event, #ifdef CONFIG_PM .suspend = stac_suspend, + .check_power_status = stac_check_power_status, #endif }; From 5666a274a6d54372d6b79b1f78682a9d827e679e Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 1 Sep 2022 18:37:34 +0200 Subject: [PATCH 057/158] driver core: fix driver_set_override() issue with empty strings Python likes to send an empty string for some sysfs files, including the driver_override field. When commit 23d99baf9d72 ("PCI: Use driver_set_override() instead of open-coding") moved the PCI core to use the driver core function instead of hand-rolling their own handler, this showed up as a regression from some userspace tools, like DPDK. Fix this up by actually looking at the length of the string first instead of trusting that userspace got it correct. Fixes: 23d99baf9d72 ("PCI: Use driver_set_override() instead of open-coding") Cc: Krzysztof Kozlowski Cc: Bjorn Helgaas Cc: "Rafael J. Wysocki" Cc: Andy Shevchenko Cc: stable Reported-by: Stephen Hemminger Tested-by: Huisong Li Reviewed-by: Stephen Hemminger Reviewed-by: Krzysztof Kozlowski Link: https://lore.kernel.org/r/20220901163734.3583106-1-gregkh@linuxfoundation.org Signed-off-by: Greg Kroah-Hartman --- drivers/base/driver.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/base/driver.c b/drivers/base/driver.c index 15a75afe6b84..676b6275d5b5 100644 --- a/drivers/base/driver.c +++ b/drivers/base/driver.c @@ -63,6 +63,12 @@ int driver_set_override(struct device *dev, const char **override, if (len >= (PAGE_SIZE - 1)) return -EINVAL; + /* + * Compute the real length of the string in case userspace sends us a + * bunch of \0 characters like python likes to do. + */ + len = strlen(s); + if (!len) { /* Empty string passed - clear override */ device_lock(dev); From dec9b2f1e0455a151a7293c367da22ab973f713e Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Fri, 2 Sep 2022 16:59:15 +0200 Subject: [PATCH 058/158] debugfs: add debugfs_lookup_and_remove() There is a very common pattern of using debugfs_remove(debufs_lookup(..)) which results in a dentry leak of the dentry that was looked up. Instead of having to open-code the correct pattern of calling dput() on the dentry, create debugfs_lookup_and_remove() to handle this pattern automatically and properly without any memory leaks. Cc: stable Reported-by: Kuyo Chang Tested-by: Kuyo Chang Link: https://lore.kernel.org/r/YxIaQ8cSinDR881k@kroah.com Signed-off-by: Greg Kroah-Hartman --- fs/debugfs/inode.c | 22 ++++++++++++++++++++++ include/linux/debugfs.h | 6 ++++++ 2 files changed, 28 insertions(+) diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index 3dcf0b8b4e93..232cfdf095ae 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -744,6 +744,28 @@ void debugfs_remove(struct dentry *dentry) } EXPORT_SYMBOL_GPL(debugfs_remove); +/** + * debugfs_lookup_and_remove - lookup a directory or file and recursively remove it + * @name: a pointer to a string containing the name of the item to look up. + * @parent: a pointer to the parent dentry of the item. + * + * This is the equlivant of doing something like + * debugfs_remove(debugfs_lookup(..)) but with the proper reference counting + * handled for the directory being looked up. + */ +void debugfs_lookup_and_remove(const char *name, struct dentry *parent) +{ + struct dentry *dentry; + + dentry = debugfs_lookup(name, parent); + if (!dentry) + return; + + debugfs_remove(dentry); + dput(dentry); +} +EXPORT_SYMBOL_GPL(debugfs_lookup_and_remove); + /** * debugfs_rename - rename a file/directory in the debugfs filesystem * @old_dir: a pointer to the parent dentry for the renamed object. This diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h index c869f1e73d75..f60674692d36 100644 --- a/include/linux/debugfs.h +++ b/include/linux/debugfs.h @@ -91,6 +91,8 @@ struct dentry *debugfs_create_automount(const char *name, void debugfs_remove(struct dentry *dentry); #define debugfs_remove_recursive debugfs_remove +void debugfs_lookup_and_remove(const char *name, struct dentry *parent); + const struct file_operations *debugfs_real_fops(const struct file *filp); int debugfs_file_get(struct dentry *dentry); @@ -225,6 +227,10 @@ static inline void debugfs_remove(struct dentry *dentry) static inline void debugfs_remove_recursive(struct dentry *dentry) { } +static inline void debugfs_lookup_and_remove(const char *name, + struct dentry *parent) +{ } + const struct file_operations *debugfs_real_fops(const struct file *filp); static inline int debugfs_file_get(struct dentry *dentry) From c2e406596571659451f4b95e37ddfd5a8ef1d0dc Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Fri, 2 Sep 2022 14:31:07 +0200 Subject: [PATCH 059/158] sched/debug: fix dentry leak in update_sched_domain_debugfs Kuyo reports that the pattern of using debugfs_remove(debugfs_lookup()) leaks a dentry and with a hotplug stress test, the machine eventually runs out of memory. Fix this up by using the newly created debugfs_lookup_and_remove() call instead which properly handles the dentry reference counting logic. Cc: Major Chen Cc: stable Cc: Ingo Molnar Cc: Peter Zijlstra Cc: Juri Lelli Cc: Vincent Guittot Cc: Dietmar Eggemann Cc: Steven Rostedt Cc: Ben Segall Cc: Mel Gorman Cc: Daniel Bristot de Oliveira Cc: Valentin Schneider Cc: Matthias Brugger Reported-by: Kuyo Chang Tested-by: Kuyo Chang Acked-by: Peter Zijlstra (Intel) Link: https://lore.kernel.org/r/20220902123107.109274-2-gregkh@linuxfoundation.org Signed-off-by: Greg Kroah-Hartman --- kernel/sched/debug.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c index bb3d63bdf4ae..667876da8382 100644 --- a/kernel/sched/debug.c +++ b/kernel/sched/debug.c @@ -416,7 +416,7 @@ void update_sched_domain_debugfs(void) char buf[32]; snprintf(buf, sizeof(buf), "cpu%d", cpu); - debugfs_remove(debugfs_lookup(buf, sd_dentry)); + debugfs_lookup_and_remove(buf, sd_dentry); d_cpu = debugfs_create_dir(buf, sd_dentry); i = 0; From 85eaeb5058f0f04dffb124c97c86b4f18db0b833 Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Wed, 24 Aug 2022 09:10:36 +0300 Subject: [PATCH 060/158] IB/core: Fix a nested dead lock as part of ODP flow Fix a nested dead lock as part of ODP flow by using mmput_async(). From the below call trace [1] can see that calling mmput() once we have the umem_odp->umem_mutex locked as required by ib_umem_odp_map_dma_and_lock() might trigger in the same task the exit_mmap()->__mmu_notifier_release()->mlx5_ib_invalidate_range() which may dead lock when trying to lock the same mutex. Moving to use mmput_async() will solve the problem as the above exit_mmap() flow will be called in other task and will be executed once the lock will be available. [1] [64843.077665] task:kworker/u133:2 state:D stack: 0 pid:80906 ppid: 2 flags:0x00004000 [64843.077672] Workqueue: mlx5_ib_page_fault mlx5_ib_eqe_pf_action [mlx5_ib] [64843.077719] Call Trace: [64843.077722] [64843.077724] __schedule+0x23d/0x590 [64843.077729] schedule+0x4e/0xb0 [64843.077735] schedule_preempt_disabled+0xe/0x10 [64843.077740] __mutex_lock.constprop.0+0x263/0x490 [64843.077747] __mutex_lock_slowpath+0x13/0x20 [64843.077752] mutex_lock+0x34/0x40 [64843.077758] mlx5_ib_invalidate_range+0x48/0x270 [mlx5_ib] [64843.077808] __mmu_notifier_release+0x1a4/0x200 [64843.077816] exit_mmap+0x1bc/0x200 [64843.077822] ? walk_page_range+0x9c/0x120 [64843.077828] ? __cond_resched+0x1a/0x50 [64843.077833] ? mutex_lock+0x13/0x40 [64843.077839] ? uprobe_clear_state+0xac/0x120 [64843.077860] mmput+0x5f/0x140 [64843.077867] ib_umem_odp_map_dma_and_lock+0x21b/0x580 [ib_core] [64843.077931] pagefault_real_mr+0x9a/0x140 [mlx5_ib] [64843.077962] pagefault_mr+0xb4/0x550 [mlx5_ib] [64843.077992] pagefault_single_data_segment.constprop.0+0x2ac/0x560 [mlx5_ib] [64843.078022] mlx5_ib_eqe_pf_action+0x528/0x780 [mlx5_ib] [64843.078051] process_one_work+0x22b/0x3d0 [64843.078059] worker_thread+0x53/0x410 [64843.078065] ? process_one_work+0x3d0/0x3d0 [64843.078073] kthread+0x12a/0x150 [64843.078079] ? set_kthread_struct+0x50/0x50 [64843.078085] ret_from_fork+0x22/0x30 [64843.078093] Fixes: 36f30e486dce ("IB/core: Improve ODP to use hmm_range_fault()") Reviewed-by: Maor Gottlieb Signed-off-by: Yishai Hadas Link: https://lore.kernel.org/r/74d93541ea533ef7daec6f126deb1072500aeb16.1661251841.git.leonro@nvidia.com Signed-off-by: Leon Romanovsky --- drivers/infiniband/core/umem_odp.c | 2 +- kernel/fork.c | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/core/umem_odp.c b/drivers/infiniband/core/umem_odp.c index 186ed8859920..d39e16c211e8 100644 --- a/drivers/infiniband/core/umem_odp.c +++ b/drivers/infiniband/core/umem_odp.c @@ -462,7 +462,7 @@ retry: mutex_unlock(&umem_odp->umem_mutex); out_put_mm: - mmput(owning_mm); + mmput_async(owning_mm); out_put_task: if (owning_process) put_task_struct(owning_process); diff --git a/kernel/fork.c b/kernel/fork.c index 90c85b17bf69..8a9e92068b15 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1225,6 +1225,7 @@ void mmput_async(struct mm_struct *mm) schedule_work(&mm->async_put_work); } } +EXPORT_SYMBOL_GPL(mmput_async); #endif /** From 9ca05b0f27de928be121cccf07735819dc9e1ed3 Mon Sep 17 00:00:00 2001 From: Maher Sanalla Date: Mon, 29 Aug 2022 12:02:27 +0300 Subject: [PATCH 061/158] RDMA/mlx5: Rely on RoCE fw cap instead of devlink when setting profile When the RDMA auxiliary driver probes, it sets its profile based on devlink driverinit value. The latter might not be in sync with FW yet (In case devlink reload is not performed), thus causing a mismatch between RDMA driver and FW. This results in the following FW syndrome when the RDMA driver tries to adjust RoCE state, which fails the probe: "0xC1F678 | modify_nic_vport_context: roce_en set on a vport that doesn't support roce" To prevent this, select the PF profile based on FW RoCE capability instead of relying on devlink driverinit value. To provide backward compatibility of the RoCE disable feature, on older FW's where roce_rw is not set (FW RoCE capability is read-only), keep the current behavior e.g., rely on devlink driverinit value. Fixes: fbfa97b4d79f ("net/mlx5: Disable roce at HCA level") Reviewed-by: Shay Drory Reviewed-by: Michael Guralnik Reviewed-by: Saeed Mahameed Signed-off-by: Maher Sanalla Link: https://lore.kernel.org/r/cb34ce9a1df4a24c135cb804db87f7d2418bd6cc.1661763459.git.leonro@nvidia.com Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/mlx5/main.c | 2 +- .../net/ethernet/mellanox/mlx5/core/main.c | 23 +++++++++++++++++-- include/linux/mlx5/driver.h | 19 +++++++-------- 3 files changed, 32 insertions(+), 12 deletions(-) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index fc94a1b25485..883d7c60143e 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -4336,7 +4336,7 @@ static int mlx5r_probe(struct auxiliary_device *adev, dev->mdev = mdev; dev->num_ports = num_ports; - if (ll == IB_LINK_LAYER_ETHERNET && !mlx5_is_roce_init_enabled(mdev)) + if (ll == IB_LINK_LAYER_ETHERNET && !mlx5_get_roce_state(mdev)) profile = &raw_eth_profile; else profile = &pf_profile; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c index bec8d6d0b5f6..4870a88cecb7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c @@ -494,6 +494,24 @@ static int max_uc_list_get_devlink_param(struct mlx5_core_dev *dev) return err; } +bool mlx5_is_roce_on(struct mlx5_core_dev *dev) +{ + struct devlink *devlink = priv_to_devlink(dev); + union devlink_param_value val; + int err; + + err = devlink_param_driverinit_value_get(devlink, + DEVLINK_PARAM_GENERIC_ID_ENABLE_ROCE, + &val); + + if (!err) + return val.vbool; + + mlx5_core_dbg(dev, "Failed to get param. err = %d\n", err); + return MLX5_CAP_GEN(dev, roce); +} +EXPORT_SYMBOL(mlx5_is_roce_on); + static int handle_hca_cap_2(struct mlx5_core_dev *dev, void *set_ctx) { void *set_hca_cap; @@ -597,7 +615,8 @@ static int handle_hca_cap(struct mlx5_core_dev *dev, void *set_ctx) MLX5_CAP_GEN_MAX(dev, num_total_dynamic_vf_msix)); if (MLX5_CAP_GEN(dev, roce_rw_supported)) - MLX5_SET(cmd_hca_cap, set_hca_cap, roce, mlx5_is_roce_init_enabled(dev)); + MLX5_SET(cmd_hca_cap, set_hca_cap, roce, + mlx5_is_roce_on(dev)); max_uc_list = max_uc_list_get_devlink_param(dev); if (max_uc_list > 0) @@ -623,7 +642,7 @@ static int handle_hca_cap(struct mlx5_core_dev *dev, void *set_ctx) */ static bool is_roce_fw_disabled(struct mlx5_core_dev *dev) { - return (MLX5_CAP_GEN(dev, roce_rw_supported) && !mlx5_is_roce_init_enabled(dev)) || + return (MLX5_CAP_GEN(dev, roce_rw_supported) && !mlx5_is_roce_on(dev)) || (!MLX5_CAP_GEN(dev, roce_rw_supported) && !MLX5_CAP_GEN(dev, roce)); } diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h index 96b16fbe1aa4..d6338fb449c8 100644 --- a/include/linux/mlx5/driver.h +++ b/include/linux/mlx5/driver.h @@ -1279,16 +1279,17 @@ enum { MLX5_TRIGGERED_CMD_COMP = (u64)1 << 32, }; -static inline bool mlx5_is_roce_init_enabled(struct mlx5_core_dev *dev) -{ - struct devlink *devlink = priv_to_devlink(dev); - union devlink_param_value val; - int err; +bool mlx5_is_roce_on(struct mlx5_core_dev *dev); - err = devlink_param_driverinit_value_get(devlink, - DEVLINK_PARAM_GENERIC_ID_ENABLE_ROCE, - &val); - return err ? MLX5_CAP_GEN(dev, roce) : val.vbool; +static inline bool mlx5_get_roce_state(struct mlx5_core_dev *dev) +{ + if (MLX5_CAP_GEN(dev, roce_rw_supported)) + return MLX5_CAP_GEN(dev, roce); + + /* If RoCE cap is read-only in FW, get RoCE state from devlink + * in order to support RoCE enable/disable feature + */ + return mlx5_is_roce_on(dev); } #endif /* MLX5_DRIVER_H */ From 74b30b3ad5cec95d2647e796d10137438a098bc1 Mon Sep 17 00:00:00 2001 From: Chris Mi Date: Mon, 29 Aug 2022 12:02:28 +0300 Subject: [PATCH 062/158] RDMA/mlx5: Set local port to one when accessing counters When accessing Ports Performance Counters Register (PPCNT), local port must be one if it is Function-Per-Port HCA that HCA_CAP.num_ports is 1. The offending patch can change the local port to other values when accessing PPCNT after enabling switchdev mode. The following syndrome will be printed: # cat /sys/class/infiniband/rdmap4s0f0/ports/2/counters/* # dmesg mlx5_core 0000:04:00.0: mlx5_cmd_check:756:(pid 12450): ACCESS_REG(0x805) op_mod(0x1) failed, status bad parameter(0x3), syndrome (0x1e5585) Fix it by setting local port to one for Function-Per-Port HCA. Fixes: 210b1f78076f ("IB/mlx5: When not in dual port RoCE mode, use provided port as native") Reviewed-by: Mark Bloch Signed-off-by: Chris Mi Link: https://lore.kernel.org/r/6c5086c295c76211169e58dbd610fb0402360bab.1661763459.git.leonro@nvidia.com Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/mlx5/mad.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/infiniband/hw/mlx5/mad.c b/drivers/infiniband/hw/mlx5/mad.c index 293ed709e5ed..b4dc52392275 100644 --- a/drivers/infiniband/hw/mlx5/mad.c +++ b/drivers/infiniband/hw/mlx5/mad.c @@ -166,6 +166,12 @@ static int process_pma_cmd(struct mlx5_ib_dev *dev, u32 port_num, mdev = dev->mdev; mdev_port_num = 1; } + if (MLX5_CAP_GEN(dev->mdev, num_ports) == 1) { + /* set local port to one for Function-Per-Port HCA. */ + mdev = dev->mdev; + mdev_port_num = 1; + } + /* Declaring support of extended counters */ if (in_mad->mad_hdr.attr_id == IB_PMA_CLASS_PORT_INFO) { struct ib_class_port_info cpi = {}; From 9b7d4be967f16f79a2283b2338709fcc750313ee Mon Sep 17 00:00:00 2001 From: Maor Gottlieb Date: Mon, 29 Aug 2022 12:02:29 +0300 Subject: [PATCH 063/158] RDMA/mlx5: Fix UMR cleanup on error flow of driver init The cited commit removed from the cleanup flow of umr the checks if the resources were created. This could lead to null-ptr-deref in case that we had failure in mlx5_ib_stage_ib_reg_init stage. Fix it by adding new state to the umr that can say if the resources were created or not and check it in the umr cleanup flow before destroying the resources. Fixes: 04876c12c19e ("RDMA/mlx5: Move init and cleanup of UMR to umr.c") Reviewed-by: Michael Guralnik Signed-off-by: Maor Gottlieb Link: https://lore.kernel.org/r/4cfa61386cf202e9ce330e8d228ce3b25a36326e.1661763459.git.leonro@nvidia.com Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/mlx5/mlx5_ib.h | 1 + drivers/infiniband/hw/mlx5/umr.c | 3 +++ 2 files changed, 4 insertions(+) diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index 2e2ad3918385..e66bf72f1f04 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -708,6 +708,7 @@ struct mlx5_ib_umr_context { }; enum { + MLX5_UMR_STATE_UNINIT, MLX5_UMR_STATE_ACTIVE, MLX5_UMR_STATE_RECOVER, MLX5_UMR_STATE_ERR, diff --git a/drivers/infiniband/hw/mlx5/umr.c b/drivers/infiniband/hw/mlx5/umr.c index e00b94d1b1ea..d5105b5c9979 100644 --- a/drivers/infiniband/hw/mlx5/umr.c +++ b/drivers/infiniband/hw/mlx5/umr.c @@ -177,6 +177,7 @@ int mlx5r_umr_resource_init(struct mlx5_ib_dev *dev) sema_init(&dev->umrc.sem, MAX_UMR_WR); mutex_init(&dev->umrc.lock); + dev->umrc.state = MLX5_UMR_STATE_ACTIVE; return 0; @@ -191,6 +192,8 @@ destroy_pd: void mlx5r_umr_resource_cleanup(struct mlx5_ib_dev *dev) { + if (dev->umrc.state == MLX5_UMR_STATE_UNINIT) + return; ib_destroy_qp(dev->umrc.qp); ib_free_cq(dev->umrc.cq); ib_dealloc_pd(dev->umrc.pd); From bc7a2c9b17773c89f2f9c09ac9f9233716d6cd08 Mon Sep 17 00:00:00 2001 From: Wenpeng Liang Date: Mon, 5 Sep 2022 10:38:15 +0800 Subject: [PATCH 064/158] MAINTAINERS: Update maintainers of HiSilicon RoCE Weihang has moved to work in other technical areas, and Haoyue will maintain this module instead of him. Signed-off-by: Wenpeng Liang Link: https://lore.kernel.org/r/20220905023815.1477684-1-liangwenpeng@huawei.com Signed-off-by: Leon Romanovsky --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 8a5012ba6ff9..6b26f4d5a1c9 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9205,8 +9205,8 @@ F: Documentation/ABI/testing/debugfs-hisi-zip F: drivers/crypto/hisilicon/zip/ HISILICON ROCE DRIVER +M: Haoyue Xu M: Wenpeng Liang -M: Weihang Li L: linux-rdma@vger.kernel.org S: Maintained F: Documentation/devicetree/bindings/infiniband/hisilicon-hns-roce.txt From 8423f0b6d513b259fdab9c9bf4aaa6188d054c2d Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 5 Sep 2022 08:07:14 +0200 Subject: [PATCH 065/158] ALSA: pcm: oss: Fix race at SNDCTL_DSP_SYNC There is a small race window at snd_pcm_oss_sync() that is called from OSS PCM SNDCTL_DSP_SYNC ioctl; namely the function calls snd_pcm_oss_make_ready() at first, then takes the params_lock mutex for the rest. When the stream is set up again by another thread between them, it leads to inconsistency, and may result in unexpected results such as NULL dereference of OSS buffer as a fuzzer spotted recently. The fix is simply to cover snd_pcm_oss_make_ready() call into the same params_lock mutex with snd_pcm_oss_make_ready_locked() variant. Reported-and-tested-by: butt3rflyh4ck Reviewed-by: Jaroslav Kysela Cc: Link: https://lore.kernel.org/r/CAFcO6XN7JDM4xSXGhtusQfS2mSBcx50VJKwQpCq=WeLt57aaZA@mail.gmail.com Link: https://lore.kernel.org/r/20220905060714.22549-1-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/core/oss/pcm_oss.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/sound/core/oss/pcm_oss.c b/sound/core/oss/pcm_oss.c index 90c3a367d7de..02df915eb3c6 100644 --- a/sound/core/oss/pcm_oss.c +++ b/sound/core/oss/pcm_oss.c @@ -1672,14 +1672,14 @@ static int snd_pcm_oss_sync(struct snd_pcm_oss_file *pcm_oss_file) runtime = substream->runtime; if (atomic_read(&substream->mmap_count)) goto __direct; - err = snd_pcm_oss_make_ready(substream); - if (err < 0) - return err; atomic_inc(&runtime->oss.rw_ref); if (mutex_lock_interruptible(&runtime->oss.params_lock)) { atomic_dec(&runtime->oss.rw_ref); return -ERESTARTSYS; } + err = snd_pcm_oss_make_ready_locked(substream); + if (err < 0) + goto unlock; format = snd_pcm_oss_format_from(runtime->oss.format); width = snd_pcm_format_physical_width(format); if (runtime->oss.buffer_used > 0) { From 6a02a61e81c231cc5c680c5dbf8665275147ac52 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Fri, 12 Aug 2022 14:03:17 -0700 Subject: [PATCH 066/158] nvmet: fix a use-after-free Fix the following use-after-free complaint triggered by blktests nvme/004: BUG: KASAN: user-memory-access in blk_mq_complete_request_remote+0xac/0x350 Read of size 4 at addr 0000607bd1835943 by task kworker/13:1/460 Workqueue: nvmet-wq nvme_loop_execute_work [nvme_loop] Call Trace: show_stack+0x52/0x58 dump_stack_lvl+0x49/0x5e print_report.cold+0x36/0x1e2 kasan_report+0xb9/0xf0 __asan_load4+0x6b/0x80 blk_mq_complete_request_remote+0xac/0x350 nvme_loop_queue_response+0x1df/0x275 [nvme_loop] __nvmet_req_complete+0x132/0x4f0 [nvmet] nvmet_req_complete+0x15/0x40 [nvmet] nvmet_execute_io_connect+0x18a/0x1f0 [nvmet] nvme_loop_execute_work+0x20/0x30 [nvme_loop] process_one_work+0x56e/0xa70 worker_thread+0x2d1/0x640 kthread+0x183/0x1c0 ret_from_fork+0x1f/0x30 Cc: stable@vger.kernel.org Fixes: a07b4970f464 ("nvmet: add a generic NVMe target") Signed-off-by: Bart Van Assche Signed-off-by: Christoph Hellwig --- drivers/nvme/target/core.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/nvme/target/core.c b/drivers/nvme/target/core.c index a1345790005f..7f4083cf953a 100644 --- a/drivers/nvme/target/core.c +++ b/drivers/nvme/target/core.c @@ -735,6 +735,8 @@ static void nvmet_set_error(struct nvmet_req *req, u16 status) static void __nvmet_req_complete(struct nvmet_req *req, u16 status) { + struct nvmet_ns *ns = req->ns; + if (!req->sq->sqhd_disabled) nvmet_update_sq_head(req); req->cqe->sq_id = cpu_to_le16(req->sq->qid); @@ -745,9 +747,9 @@ static void __nvmet_req_complete(struct nvmet_req *req, u16 status) trace_nvmet_req_complete(req); - if (req->ns) - nvmet_put_namespace(req->ns); req->ops->queue_response(req); + if (ns) + nvmet_put_namespace(ns); } void nvmet_req_complete(struct nvmet_req *req, u16 status) From 51bdc8bb82525cd70feb92279c8b7660ad7948dd Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 5 Sep 2022 15:06:30 +0200 Subject: [PATCH 067/158] ALSA: hda/sigmatel: Fix unused variable warning for beep power change The newly added stac_check_power_status() caused a compile warning when CONFIG_SND_HDA_INPUT_BEEP is disabled. Fix it. Fixes: 414d38ba8710 ("ALSA: hda/sigmatel: Keep power up while beep is enabled") Reported-by: kernel test robot Link: https://lore.kernel.org/r/20220905130630.2845-1-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_sigmatel.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c index 066bfccd2587..7f340f18599c 100644 --- a/sound/pci/hda/patch_sigmatel.c +++ b/sound/pci/hda/patch_sigmatel.c @@ -4447,7 +4447,9 @@ static int stac_suspend(struct hda_codec *codec) static int stac_check_power_status(struct hda_codec *codec, hda_nid_t nid) { +#ifdef CONFIG_SND_HDA_INPUT_BEEP struct sigmatel_spec *spec = codec->spec; +#endif int ret = snd_hda_gen_check_power_status(codec, nid); #ifdef CONFIG_SND_HDA_INPUT_BEEP From d5b81ced74afded85619ffbbe9c32ba9d82c9b1e Mon Sep 17 00:00:00 2001 From: Naohiro Aota Date: Wed, 31 Aug 2022 13:55:48 +0900 Subject: [PATCH 068/158] btrfs: zoned: fix API misuse of zone finish waiting The commit 2ce543f47843 ("btrfs: zoned: wait until zone is finished when allocation didn't progress") implemented a zone finish waiting mechanism to the write path of zoned mode. However, using wait_var_event()/wake_up_all() on fs_info->zone_finish_wait is wrong and wait_var_event() just hangs because no one ever wakes it up once it goes into sleep. Instead, we can simply use wait_on_bit_io() and clear_and_wake_up_bit() on fs_info->flags with a proper barrier installed. Fixes: 2ce543f47843 ("btrfs: zoned: wait until zone is finished when allocation didn't progress") CC: stable@vger.kernel.org # 5.16+ Signed-off-by: Naohiro Aota Signed-off-by: David Sterba --- fs/btrfs/ctree.h | 2 -- fs/btrfs/disk-io.c | 1 - fs/btrfs/inode.c | 7 +++---- fs/btrfs/zoned.c | 3 +-- 4 files changed, 4 insertions(+), 9 deletions(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 9ef162dbd4bc..df8c99c99df9 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1088,8 +1088,6 @@ struct btrfs_fs_info { spinlock_t zone_active_bgs_lock; struct list_head zone_active_bgs; - /* Waiters when BTRFS_FS_NEED_ZONE_FINISH is set */ - wait_queue_head_t zone_finish_wait; /* Updates are not protected by any lock */ struct btrfs_commit_stats commit_stats; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index b0345aa16c01..129711773ae5 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3070,7 +3070,6 @@ void btrfs_init_fs_info(struct btrfs_fs_info *fs_info) init_waitqueue_head(&fs_info->transaction_blocked_wait); init_waitqueue_head(&fs_info->async_submit_wait); init_waitqueue_head(&fs_info->delayed_iputs_wait); - init_waitqueue_head(&fs_info->zone_finish_wait); /* Usable values until the real ones are cached from the superblock */ fs_info->nodesize = 4096; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index bea0adf28735..a95bac0c3e20 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1644,10 +1644,9 @@ static noinline int run_delalloc_zoned(struct btrfs_inode *inode, done_offset = end; if (done_offset == start) { - struct btrfs_fs_info *info = inode->root->fs_info; - - wait_var_event(&info->zone_finish_wait, - !test_bit(BTRFS_FS_NEED_ZONE_FINISH, &info->flags)); + wait_on_bit_io(&inode->root->fs_info->flags, + BTRFS_FS_NEED_ZONE_FINISH, + TASK_UNINTERRUPTIBLE); continue; } diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c index b150b07ba1a7..fb25ba7fcb2a 100644 --- a/fs/btrfs/zoned.c +++ b/fs/btrfs/zoned.c @@ -2007,8 +2007,7 @@ static int do_zone_finish(struct btrfs_block_group *block_group, bool fully_writ /* For active_bg_list */ btrfs_put_block_group(block_group); - clear_bit(BTRFS_FS_NEED_ZONE_FINISH, &fs_info->flags); - wake_up_all(&fs_info->zone_finish_wait); + clear_and_wake_up_bit(BTRFS_FS_NEED_ZONE_FINISH, &fs_info->flags); return 0; } From cac5c44c48c9fb9cc31bea15ebd9ef0c6462314f Mon Sep 17 00:00:00 2001 From: Shin'ichiro Kawasaki Date: Fri, 26 Aug 2022 16:42:15 +0900 Subject: [PATCH 069/158] btrfs: zoned: set pseudo max append zone limit in zone emulation mode The commit 7d7672bc5d10 ("btrfs: convert count_max_extents() to use fs_info->max_extent_size") introduced a division by fs_info->max_extent_size. This max_extent_size is initialized with max zone append limit size of the device btrfs runs on. However, in zone emulation mode, the device is not zoned then its zone append limit is zero. This resulted in zero value of fs_info->max_extent_size and caused zero division error. Fix the error by setting non-zero pseudo value to max append zone limit in zone emulation mode. Set the pseudo value based on max_segments as suggested in the commit c2ae7b772ef4 ("btrfs: zoned: revive max_zone_append_bytes"). Fixes: 7d7672bc5d10 ("btrfs: convert count_max_extents() to use fs_info->max_extent_size") CC: stable@vger.kernel.org # 5.12+ Reviewed-by: Johannes Thumshirn Reviewed-by: Naohiro Aota Signed-off-by: Shin'ichiro Kawasaki Signed-off-by: David Sterba --- fs/btrfs/zoned.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c index fb25ba7fcb2a..f576e30ceb33 100644 --- a/fs/btrfs/zoned.c +++ b/fs/btrfs/zoned.c @@ -421,10 +421,19 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device, bool populate_cache) * since btrfs adds the pages one by one to a bio, and btrfs cannot * increase the metadata reservation even if it increases the number of * extents, it is safe to stick with the limit. + * + * With the zoned emulation, we can have non-zoned device on the zoned + * mode. In this case, we don't have a valid max zone append size. So, + * use max_segments * PAGE_SIZE as the pseudo max_zone_append_size. */ - zone_info->max_zone_append_size = - min_t(u64, (u64)bdev_max_zone_append_sectors(bdev) << SECTOR_SHIFT, - (u64)bdev_max_segments(bdev) << PAGE_SHIFT); + if (bdev_is_zoned(bdev)) { + zone_info->max_zone_append_size = min_t(u64, + (u64)bdev_max_zone_append_sectors(bdev) << SECTOR_SHIFT, + (u64)bdev_max_segments(bdev) << PAGE_SHIFT); + } else { + zone_info->max_zone_append_size = + (u64)bdev_max_segments(bdev) << PAGE_SHIFT; + } if (!IS_ALIGNED(nr_sectors, zone_sectors)) zone_info->nr_zones++; From 6ca64ac2763149fb66c0b4bf12f5e0977a88e51d Mon Sep 17 00:00:00 2001 From: Johannes Thumshirn Date: Mon, 5 Sep 2022 05:38:24 -0700 Subject: [PATCH 070/158] btrfs: zoned: fix mounting with conventional zones Since commit 6a921de58992 ("btrfs: zoned: introduce space_info->active_total_bytes"), we're only counting the bytes of a block group on an active zone as usable for metadata writes. But on a SMR drive, we don't have active zones and short circuit some of the logic. This leads to an error on mount, because we cannot reserve space for metadata writes. Fix this by also setting the BLOCK_GROUP_FLAG_ZONE_IS_ACTIVE bit in the block-group's runtime flag if the zone is a conventional zone. Fixes: 6a921de58992 ("btrfs: zoned: introduce space_info->active_total_bytes") Signed-off-by: Johannes Thumshirn Signed-off-by: David Sterba --- fs/btrfs/zoned.c | 81 ++++++++++++++++++++++++------------------------ 1 file changed, 40 insertions(+), 41 deletions(-) diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c index f576e30ceb33..62e7007a7e46 100644 --- a/fs/btrfs/zoned.c +++ b/fs/btrfs/zoned.c @@ -1187,7 +1187,7 @@ int btrfs_ensure_empty_zones(struct btrfs_device *device, u64 start, u64 size) * offset. */ static int calculate_alloc_pointer(struct btrfs_block_group *cache, - u64 *offset_ret) + u64 *offset_ret, bool new) { struct btrfs_fs_info *fs_info = cache->fs_info; struct btrfs_root *root; @@ -1197,6 +1197,21 @@ static int calculate_alloc_pointer(struct btrfs_block_group *cache, int ret; u64 length; + /* + * Avoid tree lookups for a new block group, there's no use for it. + * It must always be 0. + * + * Also, we have a lock chain of extent buffer lock -> chunk mutex. + * For new a block group, this function is called from + * btrfs_make_block_group() which is already taking the chunk mutex. + * Thus, we cannot call calculate_alloc_pointer() which takes extent + * buffer locks to avoid deadlock. + */ + if (new) { + *offset_ret = 0; + return 0; + } + path = btrfs_alloc_path(); if (!path) return -ENOMEM; @@ -1332,6 +1347,13 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new) else num_conventional++; + /* + * Consider a zone as active if we can allow any number of + * active zones. + */ + if (!device->zone_info->max_active_zones) + __set_bit(i, active); + if (!is_sequential) { alloc_offsets[i] = WP_CONVENTIONAL; continue; @@ -1398,45 +1420,23 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new) __set_bit(i, active); break; } - - /* - * Consider a zone as active if we can allow any number of - * active zones. - */ - if (!device->zone_info->max_active_zones) - __set_bit(i, active); } if (num_sequential > 0) cache->seq_zone = true; if (num_conventional > 0) { - /* - * Avoid calling calculate_alloc_pointer() for new BG. It - * is no use for new BG. It must be always 0. - * - * Also, we have a lock chain of extent buffer lock -> - * chunk mutex. For new BG, this function is called from - * btrfs_make_block_group() which is already taking the - * chunk mutex. Thus, we cannot call - * calculate_alloc_pointer() which takes extent buffer - * locks to avoid deadlock. - */ - /* Zone capacity is always zone size in emulation */ cache->zone_capacity = cache->length; - if (new) { - cache->alloc_offset = 0; - goto out; - } - ret = calculate_alloc_pointer(cache, &last_alloc); - if (ret || map->num_stripes == num_conventional) { - if (!ret) - cache->alloc_offset = last_alloc; - else - btrfs_err(fs_info, + ret = calculate_alloc_pointer(cache, &last_alloc, new); + if (ret) { + btrfs_err(fs_info, "zoned: failed to determine allocation offset of bg %llu", - cache->start); + cache->start); + goto out; + } else if (map->num_stripes == num_conventional) { + cache->alloc_offset = last_alloc; + cache->zone_is_active = 1; goto out; } } @@ -1504,13 +1504,6 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new) goto out; } - if (cache->zone_is_active) { - btrfs_get_block_group(cache); - spin_lock(&fs_info->zone_active_bgs_lock); - list_add_tail(&cache->active_bg_list, &fs_info->zone_active_bgs); - spin_unlock(&fs_info->zone_active_bgs_lock); - } - out: if (cache->alloc_offset > fs_info->zone_size) { btrfs_err(fs_info, @@ -1535,10 +1528,16 @@ out: ret = -EIO; } - if (!ret) + if (!ret) { cache->meta_write_pointer = cache->alloc_offset + cache->start; - - if (ret) { + if (cache->zone_is_active) { + btrfs_get_block_group(cache); + spin_lock(&fs_info->zone_active_bgs_lock); + list_add_tail(&cache->active_bg_list, + &fs_info->zone_active_bgs); + spin_unlock(&fs_info->zone_active_bgs_lock); + } + } else { kfree(cache->physical_map); cache->physical_map = NULL; } From 8cbb2b50ee2dcb082675237eaaa48fe8479f8aa5 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Thu, 25 Aug 2022 10:25:05 +0200 Subject: [PATCH 071/158] asm-generic: Conditionally enable do_softirq_own_stack() via Kconfig. Remove the CONFIG_PREEMPT_RT symbol from the ifdef around do_softirq_own_stack() and move it to Kconfig instead. Enable softirq stacks based on SOFTIRQ_ON_OWN_STACK which depends on HAVE_SOFTIRQ_ON_OWN_STACK and its default value is set to !PREEMPT_RT. This ensures that softirq stacks are not used on PREEMPT_RT and avoids a 'select' statement on an option which has a 'depends' statement. Link: https://lore.kernel.org/YvN5E%2FPrHfUhggr7@linutronix.de Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Arnd Bergmann --- arch/Kconfig | 3 +++ arch/arm/kernel/irq.c | 2 +- arch/parisc/kernel/irq.c | 2 +- arch/powerpc/kernel/irq.c | 4 ++-- arch/s390/include/asm/softirq_stack.h | 2 +- arch/sh/kernel/irq.c | 2 +- arch/sparc/kernel/irq_64.c | 2 +- arch/x86/include/asm/irq_stack.h | 2 +- arch/x86/kernel/irq_32.c | 2 +- include/asm-generic/softirq_stack.h | 2 +- 10 files changed, 13 insertions(+), 10 deletions(-) diff --git a/arch/Kconfig b/arch/Kconfig index 5dbf11a5ba4e..8b311e400ec1 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -923,6 +923,9 @@ config HAVE_SOFTIRQ_ON_OWN_STACK Architecture provides a function to run __do_softirq() on a separate stack. +config SOFTIRQ_ON_OWN_STACK + def_bool HAVE_SOFTIRQ_ON_OWN_STACK && !PREEMPT_RT + config ALTERNATE_USER_ADDRESS_SPACE bool help diff --git a/arch/arm/kernel/irq.c b/arch/arm/kernel/irq.c index 034cb48c9eeb..fe28fc1f759d 100644 --- a/arch/arm/kernel/irq.c +++ b/arch/arm/kernel/irq.c @@ -70,7 +70,7 @@ static void __init init_irq_stacks(void) } } -#ifndef CONFIG_PREEMPT_RT +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK static void ____do_softirq(void *arg) { __do_softirq(); diff --git a/arch/parisc/kernel/irq.c b/arch/parisc/kernel/irq.c index fbb882cb8dbb..b05055f3ba4b 100644 --- a/arch/parisc/kernel/irq.c +++ b/arch/parisc/kernel/irq.c @@ -480,7 +480,7 @@ static void execute_on_irq_stack(void *func, unsigned long param1) *irq_stack_in_use = 1; } -#ifndef CONFIG_PREEMPT_RT +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK void do_softirq_own_stack(void) { execute_on_irq_stack(__do_softirq, 0); diff --git a/arch/powerpc/kernel/irq.c b/arch/powerpc/kernel/irq.c index 0f17268c1f0b..9ede61a5a469 100644 --- a/arch/powerpc/kernel/irq.c +++ b/arch/powerpc/kernel/irq.c @@ -199,7 +199,7 @@ static inline void check_stack_overflow(unsigned long sp) } } -#ifndef CONFIG_PREEMPT_RT +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK static __always_inline void call_do_softirq(const void *sp) { /* Temporarily switch r1 to sp, call __do_softirq() then restore r1. */ @@ -335,7 +335,7 @@ void *mcheckirq_ctx[NR_CPUS] __read_mostly; void *softirq_ctx[NR_CPUS] __read_mostly; void *hardirq_ctx[NR_CPUS] __read_mostly; -#ifndef CONFIG_PREEMPT_RT +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK void do_softirq_own_stack(void) { call_do_softirq(softirq_ctx[smp_processor_id()]); diff --git a/arch/s390/include/asm/softirq_stack.h b/arch/s390/include/asm/softirq_stack.h index af68d6c1d584..1ac5115d3115 100644 --- a/arch/s390/include/asm/softirq_stack.h +++ b/arch/s390/include/asm/softirq_stack.h @@ -5,7 +5,7 @@ #include #include -#ifndef CONFIG_PREEMPT_RT +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK static inline void do_softirq_own_stack(void) { call_on_stack(0, S390_lowcore.async_stack, void, __do_softirq); diff --git a/arch/sh/kernel/irq.c b/arch/sh/kernel/irq.c index 909276738078..4e6835de54cf 100644 --- a/arch/sh/kernel/irq.c +++ b/arch/sh/kernel/irq.c @@ -149,7 +149,7 @@ void irq_ctx_exit(int cpu) hardirq_ctx[cpu] = NULL; } -#ifndef CONFIG_PREEMPT_RT +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK void do_softirq_own_stack(void) { struct thread_info *curctx; diff --git a/arch/sparc/kernel/irq_64.c b/arch/sparc/kernel/irq_64.c index 41fa1be980a3..72da2e10e255 100644 --- a/arch/sparc/kernel/irq_64.c +++ b/arch/sparc/kernel/irq_64.c @@ -855,7 +855,7 @@ void __irq_entry handler_irq(int pil, struct pt_regs *regs) set_irq_regs(old_regs); } -#ifndef CONFIG_PREEMPT_RT +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK void do_softirq_own_stack(void) { void *orig_sp, *sp = softirq_stack[smp_processor_id()]; diff --git a/arch/x86/include/asm/irq_stack.h b/arch/x86/include/asm/irq_stack.h index 63f818aedf77..147cb8fdda92 100644 --- a/arch/x86/include/asm/irq_stack.h +++ b/arch/x86/include/asm/irq_stack.h @@ -203,7 +203,7 @@ IRQ_CONSTRAINTS, regs, vector); \ } -#ifndef CONFIG_PREEMPT_RT +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK /* * Macro to invoke __do_softirq on the irq stack. This is only called from * task context when bottom halves are about to be reenabled and soft diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c index e5dd6da78713..01833ebf5e8e 100644 --- a/arch/x86/kernel/irq_32.c +++ b/arch/x86/kernel/irq_32.c @@ -132,7 +132,7 @@ int irq_init_percpu_irqstack(unsigned int cpu) return 0; } -#ifndef CONFIG_PREEMPT_RT +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK void do_softirq_own_stack(void) { struct irq_stack *irqstk; diff --git a/include/asm-generic/softirq_stack.h b/include/asm-generic/softirq_stack.h index d3e2d81656e0..2a67aed9ac52 100644 --- a/include/asm-generic/softirq_stack.h +++ b/include/asm-generic/softirq_stack.h @@ -2,7 +2,7 @@ #ifndef __ASM_GENERIC_SOFTIRQ_STACK_H #define __ASM_GENERIC_SOFTIRQ_STACK_H -#if defined(CONFIG_HAVE_SOFTIRQ_ON_OWN_STACK) && !defined(CONFIG_PREEMPT_RT) +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK void do_softirq_own_stack(void); #else static inline void do_softirq_own_stack(void) From 4fa07edbb7eacfb56b3aa64f590e9f38e7f1042c Mon Sep 17 00:00:00 2001 From: Jiapeng Chong Date: Mon, 5 Sep 2022 10:04:36 +0800 Subject: [PATCH 072/158] io_uring/notif: Remove the unused function io_notif_complete() The function io_notif_complete() is defined in the notif.c file, but not called elsewhere, so delete this unused function. io_uring/notif.c:24:20: warning: unused function 'io_notif_complete' [-Wunused-function]. Link: https://bugzilla.openanolis.cn/show_bug.cgi?id=2047 Reported-by: Abaci Robot Signed-off-by: Jiapeng Chong Reviewed-by: Pavel Begunkov Link: https://lore.kernel.org/r/20220905020436.51894-1-jiapeng.chong@linux.alibaba.com Signed-off-by: Jens Axboe --- io_uring/notif.c | 8 -------- 1 file changed, 8 deletions(-) diff --git a/io_uring/notif.c b/io_uring/notif.c index 38d77165edc3..e37c6569d82e 100644 --- a/io_uring/notif.c +++ b/io_uring/notif.c @@ -21,14 +21,6 @@ static void __io_notif_complete_tw(struct io_kiocb *notif, bool *locked) io_req_task_complete(notif, locked); } -static inline void io_notif_complete(struct io_kiocb *notif) - __must_hold(¬if->ctx->uring_lock) -{ - bool locked = true; - - __io_notif_complete_tw(notif, &locked); -} - static void io_uring_tx_zerocopy_callback(struct sk_buff *skb, struct ubuf_info *uarg, bool success) From 1f05f65bddd6958d25b133f886da49c1d4bff3fa Mon Sep 17 00:00:00 2001 From: Alexandru Gagniuc Date: Mon, 5 Sep 2022 09:28:04 -0500 Subject: [PATCH 073/158] hwmon: (tps23861) fix byte order in resistance register The tps23861 registers are little-endian, and regmap_read_bulk() does not do byte order conversion. On BE machines, the bytes were swapped, and the interpretation of the resistance value was incorrect. To make it work on both big and little-endian machines, use le16_to_cpu() to convert the resitance register to host byte order. Signed-off-by: Alexandru Gagniuc Fixes: fff7b8ab22554 ("hwmon: add Texas Instruments TPS23861 driver") Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20220905142806.110598-1-mr.nuke.me@gmail.com Signed-off-by: Guenter Roeck --- drivers/hwmon/tps23861.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/hwmon/tps23861.c b/drivers/hwmon/tps23861.c index 42762e87b014..f7c59ff7ae8e 100644 --- a/drivers/hwmon/tps23861.c +++ b/drivers/hwmon/tps23861.c @@ -493,18 +493,20 @@ static char *tps23861_port_poe_plus_status(struct tps23861_data *data, int port) static int tps23861_port_resistance(struct tps23861_data *data, int port) { - u16 regval; + unsigned int raw_val; + __le16 regval; regmap_bulk_read(data->regmap, PORT_1_RESISTANCE_LSB + PORT_N_RESISTANCE_LSB_OFFSET * (port - 1), ®val, 2); - switch (FIELD_GET(PORT_RESISTANCE_RSN_MASK, regval)) { + raw_val = le16_to_cpu(regval); + switch (FIELD_GET(PORT_RESISTANCE_RSN_MASK, raw_val)) { case PORT_RESISTANCE_RSN_OTHER: - return (FIELD_GET(PORT_RESISTANCE_MASK, regval) * RESISTANCE_LSB) / 10000; + return (FIELD_GET(PORT_RESISTANCE_MASK, raw_val) * RESISTANCE_LSB) / 10000; case PORT_RESISTANCE_RSN_LOW: - return (FIELD_GET(PORT_RESISTANCE_MASK, regval) * RESISTANCE_LSB_LOW) / 10000; + return (FIELD_GET(PORT_RESISTANCE_MASK, raw_val) * RESISTANCE_LSB_LOW) / 10000; case PORT_RESISTANCE_RSN_SHORT: case PORT_RESISTANCE_RSN_OPEN: default: From 160f3549a907a50e51a8518678ba2dcf2541abea Mon Sep 17 00:00:00 2001 From: Sagi Grimberg Date: Mon, 5 Sep 2022 13:54:17 +0300 Subject: [PATCH 074/158] nvme-tcp: fix UAF when detecting digest errors We should also bail from the io_work loop when we set rd_enabled to true, so we don't attempt to read data from the socket when the TCP stream is already out-of-sync or corrupted. Fixes: 3f2304f8c6d6 ("nvme-tcp: add NVMe over TCP host driver") Reported-by: Daniel Wagner Signed-off-by: Sagi Grimberg Reviewed-by: Daniel Wagner Signed-off-by: Christoph Hellwig --- drivers/nvme/host/tcp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c index 044da18c06f5..54b4e8a7fe42 100644 --- a/drivers/nvme/host/tcp.c +++ b/drivers/nvme/host/tcp.c @@ -1229,7 +1229,7 @@ static void nvme_tcp_io_work(struct work_struct *w) else if (unlikely(result < 0)) return; - if (!pending) + if (!pending || !queue->rd_enabled) return; } while (!time_after(jiffies, deadline)); /* quota is exhausted */ From 3770a42bb8ceb856877699257a43c0585a5d2996 Mon Sep 17 00:00:00 2001 From: Sagi Grimberg Date: Mon, 5 Sep 2022 18:07:06 +0300 Subject: [PATCH 075/158] nvme-tcp: fix regression that causes sporadic requests to time out When we queue requests, we strive to batch as much as possible and also signal the network stack that more data is about to be sent over a socket with MSG_SENDPAGE_NOTLAST. This flag looks at the pending requests queued as well as queue->more_requests that is derived from the block layer last-in-batch indication. We set more_request=true when we flush the request directly from .queue_rq submission context (in nvme_tcp_send_all), however this is wrongly assuming that no other requests may be queued during the execution of nvme_tcp_send_all. Due to this, a race condition may happen where: 1. request X is queued as !last-in-batch 2. request X submission context calls nvme_tcp_send_all directly 3. nvme_tcp_send_all is preempted and schedules to a different cpu 4. request Y is queued as last-in-batch 5. nvme_tcp_send_all context sends request X+Y, however signals for both MSG_SENDPAGE_NOTLAST because queue->more_requests=true. ==> none of the requests is pushed down to the wire as the network stack is waiting for more data, both requests timeout. To fix this, we eliminate queue->more_requests and only rely on the queue req_list and send_list to be not-empty. Fixes: 122e5b9f3d37 ("nvme-tcp: optimize network stack with setting msg flags according to batch size") Reported-by: Jonathan Nicklin Signed-off-by: Sagi Grimberg Tested-by: Jonathan Nicklin Signed-off-by: Christoph Hellwig --- drivers/nvme/host/tcp.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c index 54b4e8a7fe42..d5871fd6f769 100644 --- a/drivers/nvme/host/tcp.c +++ b/drivers/nvme/host/tcp.c @@ -121,7 +121,6 @@ struct nvme_tcp_queue { struct mutex send_mutex; struct llist_head req_list; struct list_head send_list; - bool more_requests; /* recv state */ void *pdu; @@ -320,7 +319,7 @@ static inline void nvme_tcp_send_all(struct nvme_tcp_queue *queue) static inline bool nvme_tcp_queue_more(struct nvme_tcp_queue *queue) { return !list_empty(&queue->send_list) || - !llist_empty(&queue->req_list) || queue->more_requests; + !llist_empty(&queue->req_list); } static inline void nvme_tcp_queue_request(struct nvme_tcp_request *req, @@ -339,9 +338,7 @@ static inline void nvme_tcp_queue_request(struct nvme_tcp_request *req, */ if (queue->io_cpu == raw_smp_processor_id() && sync && empty && mutex_trylock(&queue->send_mutex)) { - queue->more_requests = !last; nvme_tcp_send_all(queue); - queue->more_requests = false; mutex_unlock(&queue->send_mutex); } From 8d44e6044a0e885acdd01813768a0b27906d64fd Mon Sep 17 00:00:00 2001 From: Mohan Kumar Date: Mon, 5 Sep 2022 22:54:20 +0530 Subject: [PATCH 076/158] ALSA: hda/tegra: Align BDL entry to 4KB boundary AZA HW may send a burst read/write request crossing 4K memory boundary. The 4KB boundary is not guaranteed by Tegra HDA HW. Make SW change to include the flag AZX_DCAPS_4K_BDLE_BOUNDARY to align BDLE to 4K boundary. Signed-off-by: Mohan Kumar Link: https://lore.kernel.org/r/20220905172420.3801-1-mkumard@nvidia.com Signed-off-by: Takashi Iwai --- sound/pci/hda/hda_tegra.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/sound/pci/hda/hda_tegra.c b/sound/pci/hda/hda_tegra.c index 7debb2c76aa6..976a112c7d00 100644 --- a/sound/pci/hda/hda_tegra.c +++ b/sound/pci/hda/hda_tegra.c @@ -474,7 +474,8 @@ MODULE_DEVICE_TABLE(of, hda_tegra_match); static int hda_tegra_probe(struct platform_device *pdev) { const unsigned int driver_flags = AZX_DCAPS_CORBRP_SELF_CLEAR | - AZX_DCAPS_PM_RUNTIME; + AZX_DCAPS_PM_RUNTIME | + AZX_DCAPS_4K_BDLE_BOUNDARY; struct snd_card *card; struct azx *chip; struct hda_tegra *hda; From e53f47f6c1a56d2af728909f1cb894da6b43d9bf Mon Sep 17 00:00:00 2001 From: Dongxiang Ke Date: Tue, 6 Sep 2022 10:49:28 +0800 Subject: [PATCH 077/158] ALSA: usb-audio: Fix an out-of-bounds bug in __snd_usb_parse_audio_interface() There may be a bad USB audio device with a USB ID of (0x04fa, 0x4201) and the number of it's interfaces less than 4, an out-of-bounds read bug occurs when parsing the interface descriptor for this device. Fix this by checking the number of interfaces. Signed-off-by: Dongxiang Ke Link: https://lore.kernel.org/r/20220906024928.10951-1-kdx.glider@gmail.com Signed-off-by: Takashi Iwai --- sound/usb/stream.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/usb/stream.c b/sound/usb/stream.c index 40b7821c6c99..f10f4e6d3fb8 100644 --- a/sound/usb/stream.c +++ b/sound/usb/stream.c @@ -1106,7 +1106,7 @@ static int __snd_usb_parse_audio_interface(struct snd_usb_audio *chip, * Dallas DS4201 workaround: It presents 5 altsettings, but the last * one misses syncpipe, and does not produce any sound. */ - if (chip->usb_id == USB_ID(0x04fa, 0x4201)) + if (chip->usb_id == USB_ID(0x04fa, 0x4201) && num >= 4) num = 4; for (i = 0; i < num; i++) { From c0a454b9044fdc99486853aa424e5b3be2107078 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Mon, 5 Sep 2022 15:22:55 +0100 Subject: [PATCH 078/158] arm64/bti: Disable in kernel BTI when cross section thunks are broken GCC does not insert a `bti c` instruction at the beginning of a function when it believes that all callers reach the function through a direct branch[1]. Unfortunately the logic it uses to determine this is not sufficiently robust, for example not taking account of functions being placed in different sections which may be loaded separately, so we may still see thunks being generated to these functions. If that happens, the first instruction in the callee function will result in a Branch Target Exception due to the missing landing pad. While this has currently only been observed in the case of modules having their main code loaded sufficiently far from their init section to require thunks it could potentially happen for other cases so the safest thing is to disable BTI for the kernel when building with an affected toolchain. [1]: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106671 Reported-by: D Scott Phillips [Bits of the commit message are lifted from his report & workaround] Signed-off-by: Mark Brown Link: https://lore.kernel.org/r/20220905142255.591990-1-broonie@kernel.org Cc: # v5.10+ Signed-off-by: Will Deacon --- arch/arm64/Kconfig | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index 9fb9fff08c94..1ce7685ad5de 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -1887,6 +1887,8 @@ config ARM64_BTI_KERNEL depends on CC_HAS_BRANCH_PROT_PAC_RET_BTI # https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94697 depends on !CC_IS_GCC || GCC_VERSION >= 100100 + # https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106671 + depends on !CC_IS_GCC # https://github.com/llvm/llvm-project/commit/a88c722e687e6780dcd6a58718350dc76fcc4cc9 depends on !CC_IS_CLANG || CLANG_VERSION >= 120000 depends on (!FUNCTION_GRAPH_TRACER || DYNAMIC_FTRACE_WITH_REGS) From 37137ec26c2c03039d8064c00f6eae176841ee0d Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 6 Sep 2022 11:03:19 +0200 Subject: [PATCH 079/158] ALSA: hda: Once again fix regression of page allocations with IOMMU The last fix for trying to recover the regression on AMD platforms, unfortunately, leaded to yet another regression: it turned out that IOMMUs don't like the usage of raw page allocations. This is yet another attempt for addressing the log saga; at this time, we re-use the existing buffer allocation mechanism with SG-pages although we require only single pages. The SG buffer allocation itself was confirmed to work for stream buffers, so it's relatively easy to adapt for other places. The only problem is: although the HD-audio code is accessing the address directly via dmab->address field, SG-pages don't set up it. For the ease of adaption, we now set up the dmab->addr field from the address of the first page as default, so that it can run with the HD-audio driver code as-is without the excessive call of snd_sgbuf_get_addr() multiple times; that's the only change in the memalloc helper side. The rest is nothing but a flip of the dma_type field in the HD-audio side. Fixes: a8d302a0b770 ("ALSA: memalloc: Revive x86-specific WC page allocations again") Reported-by: Mikhail Gavrilov Tested-by: Mikhail Gavrilov Cc: Link: https://lore.kernel.org/r/CABXGCsO+kB2t5QyHY-rUe76npr1m0-5JOtt8g8SiHUo34ur7Ww@mail.gmail.com Link: https://bugzilla.kernel.org/show_bug.cgi?id=216112 Link: https://bugzilla.kernel.org/show_bug.cgi?id=216363 Link: https://lore.kernel.org/r/20220906090319.23358-1-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/core/memalloc.c | 9 +++++++-- sound/pci/hda/hda_intel.c | 2 +- 2 files changed, 8 insertions(+), 3 deletions(-) diff --git a/sound/core/memalloc.c b/sound/core/memalloc.c index b665ac66ccbe..cfcd8eff4139 100644 --- a/sound/core/memalloc.c +++ b/sound/core/memalloc.c @@ -543,10 +543,13 @@ static void *snd_dma_noncontig_alloc(struct snd_dma_buffer *dmab, size_t size) dmab->dev.need_sync = dma_need_sync(dmab->dev.dev, sg_dma_address(sgt->sgl)); p = dma_vmap_noncontiguous(dmab->dev.dev, size, sgt); - if (p) + if (p) { dmab->private_data = sgt; - else + /* store the first page address for convenience */ + dmab->addr = snd_sgbuf_get_addr(dmab, 0); + } else { dma_free_noncontiguous(dmab->dev.dev, size, sgt, dmab->dev.dir); + } return p; } @@ -780,6 +783,8 @@ static void *snd_dma_sg_fallback_alloc(struct snd_dma_buffer *dmab, size_t size) if (!p) goto error; dmab->private_data = sgbuf; + /* store the first page address for convenience */ + dmab->addr = snd_sgbuf_get_addr(dmab, 0); return p; error: diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c index a77165bd92a9..b20694fd69de 100644 --- a/sound/pci/hda/hda_intel.c +++ b/sound/pci/hda/hda_intel.c @@ -1817,7 +1817,7 @@ static int azx_create(struct snd_card *card, struct pci_dev *pci, /* use the non-cached pages in non-snoop mode */ if (!azx_snoop(chip)) - azx_bus(chip)->dma_type = SNDRV_DMA_TYPE_DEV_WC; + azx_bus(chip)->dma_type = SNDRV_DMA_TYPE_DEV_WC_SG; if (chip->driver_type == AZX_DRIVER_NVIDIA) { dev_dbg(chip->card->dev, "Enable delay in RIRB handling\n"); From d91c411c744b55e860fbafc9a499f4f22d64c762 Mon Sep 17 00:00:00 2001 From: ZhenGuo Yin Date: Tue, 6 Sep 2022 16:46:19 +0800 Subject: [PATCH 080/158] drm/ttm: update bulk move object of ghost BO MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit [Why] Ghost BO is released with non-empty bulk move object. There is a warning trace: WARNING: CPU: 19 PID: 1582 at ttm/ttm_bo.c:366 ttm_bo_release+0x2e1/0x2f0 [amdttm] Call Trace: amddma_resv_reserve_fences+0x10d/0x1f0 [amdkcl] amdttm_bo_put+0x28/0x30 [amdttm] amdttm_bo_move_accel_cleanup+0x126/0x200 [amdttm] amdgpu_bo_move+0x1a8/0x770 [amdgpu] ttm_bo_handle_move_mem+0xb0/0x140 [amdttm] amdttm_bo_validate+0xbf/0x100 [amdttm] [How] The resource of ghost BO should be moved to LRU directly, instead of using bulk move. The bulk move object of ghost BO should set to NULL before function ttm_bo_move_to_lru_tail_unlocked. v2: set bulk move to NULL manually if no resource associated with ghost BO Fixed: 5b951e487fd6bf5f ("drm/ttm: fix bulk move handling v2") Signed-off-by: ZhenGuo Yin Reviewed-by: Christian König Signed-off-by: Christian König Link: https://patchwork.freedesktop.org/patch/msgid/20220906084619.2545456-1-zhenguo.yin@amd.com --- drivers/gpu/drm/ttm/ttm_bo_util.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c index 1cbfb00c1d65..57a27847206f 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_util.c +++ b/drivers/gpu/drm/ttm/ttm_bo_util.c @@ -239,6 +239,9 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo, if (fbo->base.resource) { ttm_resource_set_bo(fbo->base.resource, &fbo->base); bo->resource = NULL; + ttm_bo_set_bulk_move(&fbo->base, NULL); + } else { + fbo->base.bulk_move = NULL; } dma_resv_init(&fbo->base.base._resv); From 72cd652b73dd77df6f26bd450e804ee29232669f Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Mon, 5 Sep 2022 19:49:28 +0530 Subject: [PATCH 081/158] perf affinity: Fix out of bound access to "sched_cpus" mask The affinity code in "affinity_set" function access array named "sched_cpus". The size for this array is allocated in affinity_setup function which is nothing but value from get_cpu_set_size. This is used to contain the cpumask value for each cpu. While setting bit for each cpu, it calls "set_bit" function which access index in sched_cpus array. If we provide a command-line option to -C which is more than the number of CPU's present in the system, the set_bit could access an array member which is out-of the array size. This is because currently, there is no boundary check for the CPU. This will result in seg fault: <<>> ./perf stat -C 12323431 ls Perf can support 2048 CPUs. Consider raising MAX_NR_CPUS Segmentation fault (core dumped) <<>> Fix this by adding boundary check for the array. After the fix from powerpc system: <<>> ./perf stat -C 12323431 ls 1>out Perf can support 2048 CPUs. Consider raising MAX_NR_CPUS Performance counter stats for 'CPU(s) 12323431': msec cpu-clock context-switches cpu-migrations page-faults cycles instructions branches branch-misses 0.001192373 seconds time elapsed <<>> Reported-by: Nageswara R Sastry Signed-off-by: Athira Jajeev Tested-by: Arnaldo Carvalho de Melo Tested-by: Nageswara R Sastry Cc: Jiri Olsa Cc: Kajol Jain Cc: Madhavan Srinivasan Cc: Michael Ellerman Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220905141929.7171-1-atrajeev@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/affinity.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/tools/perf/util/affinity.c b/tools/perf/util/affinity.c index 4d216c0dc425..4ee96b3c755b 100644 --- a/tools/perf/util/affinity.c +++ b/tools/perf/util/affinity.c @@ -49,8 +49,14 @@ void affinity__set(struct affinity *a, int cpu) { int cpu_set_size = get_cpu_set_size(); - if (cpu == -1) + /* + * Return: + * - if cpu is -1 + * - restrict out of bound access to sched_cpus + */ + if (cpu == -1 || ((cpu >= (cpu_set_size * 8)))) return; + a->changed = true; set_bit(cpu, a->sched_cpus); /* From cbd7bfc7fd99acdde58ec2b0bce990158fba1654 Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Mon, 5 Sep 2022 19:49:29 +0530 Subject: [PATCH 082/158] tools/perf: Fix out of bound access to cpu mask array The cpu mask init code in "record__mmap_cpu_mask_init" function access "bits" array part of "struct mmap_cpu_mask". The size of this array is the value from cpu__max_cpu().cpu. This array is used to contain the cpumask value for each cpu. While setting bit for each cpu, it calls "set_bit" function which access index in "bits" array. If we provide a command line option to -C which is greater than the number of CPU's present in the system, the set_bit could access an array member which is out-of the array size. This is because currently, there is no boundary check for the CPU. This will result in seg fault: <<>> ./perf record -C 12341234 ls Perf can support 2048 CPUs. Consider raising MAX_NR_CPUS Segmentation fault (core dumped) <<>> Debugging with gdb, points to function flow as below: <<>> set_bit record__mmap_cpu_mask_init record__init_thread_default_masks record__init_thread_masks cmd_record <<>> Fix this by adding boundary check for the array. After the patch: <<>> ./perf record -C 12341234 ls Perf can support 2048 CPUs. Consider raising MAX_NR_CPUS Failed to initialize parallel data streaming masks <<>> With this fix, if -C is given a non-exsiting CPU, perf record will fail with: <<>> ./perf record -C 50 ls Failed to initialize parallel data streaming masks <<>> Reported-by: Nageswara R Sastry Signed-off-by: Athira Jajeev Tested-by: Arnaldo Carvalho de Melo Tested-by: Nageswara R Sastry Cc: Jiri Olsa Cc: Kajol Jain Cc: Madhavan Srinivasan Cc: Michael Ellerman Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220905141929.7171-2-atrajeev@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-record.c | 26 ++++++++++++++++++++------ 1 file changed, 20 insertions(+), 6 deletions(-) diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 4713f0f3a6cf..09b68d76bbdc 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -3358,16 +3358,22 @@ static struct option __record_options[] = { struct option *record_options = __record_options; -static void record__mmap_cpu_mask_init(struct mmap_cpu_mask *mask, struct perf_cpu_map *cpus) +static int record__mmap_cpu_mask_init(struct mmap_cpu_mask *mask, struct perf_cpu_map *cpus) { struct perf_cpu cpu; int idx; if (cpu_map__is_dummy(cpus)) - return; + return 0; - perf_cpu_map__for_each_cpu(cpu, idx, cpus) + perf_cpu_map__for_each_cpu(cpu, idx, cpus) { + /* Return ENODEV is input cpu is greater than max cpu */ + if ((unsigned long)cpu.cpu > mask->nbits) + return -ENODEV; set_bit(cpu.cpu, mask->bits); + } + + return 0; } static int record__mmap_cpu_mask_init_spec(struct mmap_cpu_mask *mask, const char *mask_spec) @@ -3379,7 +3385,9 @@ static int record__mmap_cpu_mask_init_spec(struct mmap_cpu_mask *mask, const cha return -ENOMEM; bitmap_zero(mask->bits, mask->nbits); - record__mmap_cpu_mask_init(mask, cpus); + if (record__mmap_cpu_mask_init(mask, cpus)) + return -ENODEV; + perf_cpu_map__put(cpus); return 0; @@ -3461,7 +3469,12 @@ static int record__init_thread_masks_spec(struct record *rec, struct perf_cpu_ma pr_err("Failed to allocate CPUs mask\n"); return ret; } - record__mmap_cpu_mask_init(&cpus_mask, cpus); + + ret = record__mmap_cpu_mask_init(&cpus_mask, cpus); + if (ret) { + pr_err("Failed to init cpu mask\n"); + goto out_free_cpu_mask; + } ret = record__thread_mask_alloc(&full_mask, cpu__max_cpu().cpu); if (ret) { @@ -3702,7 +3715,8 @@ static int record__init_thread_default_masks(struct record *rec, struct perf_cpu if (ret) return ret; - record__mmap_cpu_mask_init(&rec->thread_masks->maps, cpus); + if (record__mmap_cpu_mask_init(&rec->thread_masks->maps, cpus)) + return -ENODEV; rec->nr_threads = 1; From 6ea9da51a5c00c9d9309c4c9aa21cbe63c799c56 Mon Sep 17 00:00:00 2001 From: Zixuan Tan Date: Fri, 26 Aug 2022 01:00:58 +0800 Subject: [PATCH 083/158] perf genelf: Switch deprecated openssl MD5_* functions to new EVP API Switch to the flavored EVP API like in test-libcrypto.c, and remove the bad gcc #pragma. Inspired-by: 5b245985a6de5ac1 ("tools build: Switch to new openssl API for test-libcrypto") Signed-off-by: Zixuan Tan Cc: Alexander Shishkin Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/CABwm_eTnARC1GwMD-JF176k8WXU1Z0+H190mvXn61yr369qt6g@mail.gmail.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/genelf.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/tools/perf/util/genelf.c b/tools/perf/util/genelf.c index 953338b9e887..ed28a0dbcb7f 100644 --- a/tools/perf/util/genelf.c +++ b/tools/perf/util/genelf.c @@ -30,10 +30,6 @@ #define BUILD_ID_URANDOM /* different uuid for each run */ -// FIXME, remove this and fix the deprecation warnings before its removed and -// We'll break for good here... -#pragma GCC diagnostic ignored "-Wdeprecated-declarations" - #ifdef HAVE_LIBCRYPTO_SUPPORT #define BUILD_ID_MD5 @@ -45,6 +41,7 @@ #endif #ifdef BUILD_ID_MD5 +#include #include #endif #endif @@ -142,15 +139,20 @@ gen_build_id(struct buildid_note *note, static void gen_build_id(struct buildid_note *note, unsigned long load_addr, const void *code, size_t csize) { - MD5_CTX context; + EVP_MD_CTX *mdctx; if (sizeof(note->build_id) < 16) errx(1, "build_id too small for MD5"); - MD5_Init(&context); - MD5_Update(&context, &load_addr, sizeof(load_addr)); - MD5_Update(&context, code, csize); - MD5_Final((unsigned char *)note->build_id, &context); + mdctx = EVP_MD_CTX_new(); + if (!mdctx) + errx(2, "failed to create EVP_MD_CTX"); + + EVP_DigestInit_ex(mdctx, EVP_md5(), NULL); + EVP_DigestUpdate(mdctx, &load_addr, sizeof(load_addr)); + EVP_DigestUpdate(mdctx, code, csize); + EVP_DigestFinal_ex(mdctx, (unsigned char *)note->build_id, NULL); + EVP_MD_CTX_free(mdctx); } #endif From 4efa8e314351cb3e7f229ddd3571069edf38f99a Mon Sep 17 00:00:00 2001 From: Shang XiaoJing Date: Tue, 6 Sep 2022 11:29:06 +0800 Subject: [PATCH 084/158] perf c2c: Prevent potential memory leak in c2c_he_zalloc() Free allocated resources when zalloc() fails for members in c2c_he, to prevent potential memory leak in c2c_he_zalloc(). Signed-off-by: Shang XiaoJing Reviewed-by: Leo Yan Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20220906032906.21395-4-shangxiaojing@huawei.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-c2c.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/tools/perf/builtin-c2c.c b/tools/perf/builtin-c2c.c index 653e13b5037e..438fc222e213 100644 --- a/tools/perf/builtin-c2c.c +++ b/tools/perf/builtin-c2c.c @@ -146,15 +146,15 @@ static void *c2c_he_zalloc(size_t size) c2c_he->cpuset = bitmap_zalloc(c2c.cpus_cnt); if (!c2c_he->cpuset) - return NULL; + goto out_free; c2c_he->nodeset = bitmap_zalloc(c2c.nodes_cnt); if (!c2c_he->nodeset) - return NULL; + goto out_free; c2c_he->node_stats = zalloc(c2c.nodes_cnt * sizeof(*c2c_he->node_stats)); if (!c2c_he->node_stats) - return NULL; + goto out_free; init_stats(&c2c_he->cstats.lcl_hitm); init_stats(&c2c_he->cstats.rmt_hitm); @@ -163,6 +163,12 @@ static void *c2c_he_zalloc(size_t size) init_stats(&c2c_he->cstats.load); return &c2c_he->he; + +out_free: + free(c2c_he->nodeset); + free(c2c_he->cpuset); + free(c2c_he); + return NULL; } static void c2c_he_free(void *he) From 5da431b71d4b9be3c8cf6786eff9e2d41a5f9f65 Mon Sep 17 00:00:00 2001 From: Qu Wenruo Date: Thu, 18 Aug 2022 15:06:44 +0800 Subject: [PATCH 085/158] btrfs: fix the max chunk size and stripe length calculation [BEHAVIOR CHANGE] Since commit f6fca3917b4d ("btrfs: store chunk size in space-info struct"), btrfs no longer can create larger data chunks than 1G: mkfs.btrfs -f -m raid1 -d raid0 $dev1 $dev2 $dev3 $dev4 mount $dev1 $mnt btrfs balance start --full $mnt btrfs balance start --full $mnt umount $mnt btrfs ins dump-tree -t chunk $dev1 | grep "DATA|RAID0" -C 2 Before that offending commit, what we got is a 4G data chunk: item 6 key (FIRST_CHUNK_TREE CHUNK_ITEM 9492758528) itemoff 15491 itemsize 176 length 4294967296 owner 2 stripe_len 65536 type DATA|RAID0 io_align 65536 io_width 65536 sector_size 4096 num_stripes 4 sub_stripes 1 Now what we got is only 1G data chunk: item 6 key (FIRST_CHUNK_TREE CHUNK_ITEM 6271533056) itemoff 15491 itemsize 176 length 1073741824 owner 2 stripe_len 65536 type DATA|RAID0 io_align 65536 io_width 65536 sector_size 4096 num_stripes 4 sub_stripes 1 This will increase the number of data chunks by the number of devices, not only increase system chunk usage, but also greatly increase mount time. Without a proper reason, we should not change the max chunk size. [CAUSE] Previously, we set max data chunk size to 10G, while max data stripe length to 1G. Commit f6fca3917b4d ("btrfs: store chunk size in space-info struct") completely ignored the 10G limit, but use 1G max stripe limit instead, causing above shrink in max data chunk size. [FIX] Fix the max data chunk size to 10G, and in decide_stripe_size_regular() we limit stripe_size to 1G manually. This should only affect data chunks, as for metadata chunks we always set the max stripe size the same as max chunk size (256M or 1G depending on fs size). Now the same script result the same old result: item 6 key (FIRST_CHUNK_TREE CHUNK_ITEM 9492758528) itemoff 15491 itemsize 176 length 4294967296 owner 2 stripe_len 65536 type DATA|RAID0 io_align 65536 io_width 65536 sector_size 4096 num_stripes 4 sub_stripes 1 Reported-by: Wang Yugui Fixes: f6fca3917b4d ("btrfs: store chunk size in space-info struct") Signed-off-by: Qu Wenruo Reviewed-by: David Sterba Signed-off-by: David Sterba --- fs/btrfs/space-info.c | 2 +- fs/btrfs/volumes.c | 3 +++ 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/space-info.c b/fs/btrfs/space-info.c index d0cbeb7ae81c..435559ba94fa 100644 --- a/fs/btrfs/space-info.c +++ b/fs/btrfs/space-info.c @@ -199,7 +199,7 @@ static u64 calc_chunk_size(const struct btrfs_fs_info *fs_info, u64 flags) ASSERT(flags & BTRFS_BLOCK_GROUP_TYPE_MASK); if (flags & BTRFS_BLOCK_GROUP_DATA) - return SZ_1G; + return BTRFS_MAX_DATA_CHUNK_SIZE; else if (flags & BTRFS_BLOCK_GROUP_SYSTEM) return SZ_32M; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 064ab2a79c80..f63ff91e2883 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -5267,6 +5267,9 @@ static int decide_stripe_size_regular(struct alloc_chunk_ctl *ctl, ctl->stripe_size); } + /* Stripe size should not go beyond 1G. */ + ctl->stripe_size = min_t(u64, ctl->stripe_size, SZ_1G); + /* Align to BTRFS_STRIPE_LEN */ ctl->stripe_size = round_down(ctl->stripe_size, BTRFS_STRIPE_LEN); ctl->chunk_size = ctl->stripe_size * data_stripes; From edca5a2c373db61efa959307c13ed9156b1c14d9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Mon, 29 Aug 2022 16:58:34 +0300 Subject: [PATCH 086/158] drm/i915/bios: Copy the whole MIPI sequence block MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Turns out the MIPI sequence block version number and new block size fields are considered part of the block header and are not included in the reported new block size field itself. Bump up the block size appropriately so that we'll copy over the last five bytes of the block as well. For this particular machine those last five bytes included parts of the GPIO op for the backlight on sequence, causing the backlight no longer to turn back on: Sequence 6 - MIPI_SEQ_BACKLIGHT_ON Delay: 20000 us - GPIO index 0, number 0, set 0 (0x00) + GPIO index 1, number 70, set 1 (0x01) Cc: stable@vger.kernel.org Fixes: e163cfb4c96d ("drm/i915/bios: Make copies of VBT data blocks") Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/6652 Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20220829135834.8585-1-ville.syrjala@linux.intel.com Reviewed-by: Jani Nikula (cherry picked from commit a06289f3f72431f3777af95ea1226b5b0abdc426) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/i915/display/intel_bios.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/gpu/drm/i915/display/intel_bios.c b/drivers/gpu/drm/i915/display/intel_bios.c index 198a2f4920cc..7d6eb9ad7a02 100644 --- a/drivers/gpu/drm/i915/display/intel_bios.c +++ b/drivers/gpu/drm/i915/display/intel_bios.c @@ -479,6 +479,13 @@ init_bdb_block(struct drm_i915_private *i915, block_size = get_blocksize(block); + /* + * Version number and new block size are considered + * part of the header for MIPI sequenece block v3+. + */ + if (section_id == BDB_MIPI_SEQUENCE && *(const u8 *)block >= 3) + block_size += 5; + entry = kzalloc(struct_size(entry, data, max(min_size, block_size) + 3), GFP_KERNEL); if (!entry) { From e1cab970574c001d83e59ca8388c474a57a1afb6 Mon Sep 17 00:00:00 2001 From: Rodrigo Vivi Date: Wed, 31 Aug 2022 17:45:38 -0400 Subject: [PATCH 087/158] drm/i915/slpc: Let's fix the PCODE min freq table setup for SLPC We need to inform PCODE of a desired ring frequencies so PCODE update the memory frequencies to us. rps->min_freq and rps->max_freq are the frequencies used in that request. However they were unset when SLPC was enabled and PCODE never updated the memory freq. v2 (as Suggested by Ashutosh): if SLPC is in use, let's pick the right frequencies from the get_ia_constants instead of the fake init of rps' min and max. v3: don't forget the max <= min return v4: Move all the freq conversion to intel_rps.c. And the max <= min check to where it belongs. v5: (Ashutosh) Fix old comment s/50 HZ/50 MHz and add a doc explaining the "raw format" Fixes: 7ba79a671568 ("drm/i915/guc/slpc: Gate Host RPS when SLPC is enabled") Cc: # v5.15+ Cc: Ashutosh Dixit Tested-by: Sushma Venkatesh Reddy Signed-off-by: Rodrigo Vivi Reviewed-by: Ashutosh Dixit Link: https://patchwork.freedesktop.org/patch/msgid/20220831214538.143950-1-rodrigo.vivi@intel.com (cherry picked from commit 018a7bdbb090b9155a6509a0d1a684db4afaa5b1) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/i915/gt/intel_llc.c | 19 ++++++----- drivers/gpu/drm/i915/gt/intel_rps.c | 50 +++++++++++++++++++++++++++++ drivers/gpu/drm/i915/gt/intel_rps.h | 2 ++ 3 files changed, 61 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_llc.c b/drivers/gpu/drm/i915/gt/intel_llc.c index 14fe65812e42..1d19c073ba2e 100644 --- a/drivers/gpu/drm/i915/gt/intel_llc.c +++ b/drivers/gpu/drm/i915/gt/intel_llc.c @@ -12,6 +12,7 @@ #include "intel_llc.h" #include "intel_mchbar_regs.h" #include "intel_pcode.h" +#include "intel_rps.h" struct ia_constants { unsigned int min_gpu_freq; @@ -55,9 +56,6 @@ static bool get_ia_constants(struct intel_llc *llc, if (!HAS_LLC(i915) || IS_DGFX(i915)) return false; - if (rps->max_freq <= rps->min_freq) - return false; - consts->max_ia_freq = cpu_max_MHz(); consts->min_ring_freq = @@ -65,13 +63,8 @@ static bool get_ia_constants(struct intel_llc *llc, /* convert DDR frequency from units of 266.6MHz to bandwidth */ consts->min_ring_freq = mult_frac(consts->min_ring_freq, 8, 3); - consts->min_gpu_freq = rps->min_freq; - consts->max_gpu_freq = rps->max_freq; - if (GRAPHICS_VER(i915) >= 9) { - /* Convert GT frequency to 50 HZ units */ - consts->min_gpu_freq /= GEN9_FREQ_SCALER; - consts->max_gpu_freq /= GEN9_FREQ_SCALER; - } + consts->min_gpu_freq = intel_rps_get_min_raw_freq(rps); + consts->max_gpu_freq = intel_rps_get_max_raw_freq(rps); return true; } @@ -130,6 +123,12 @@ static void gen6_update_ring_freq(struct intel_llc *llc) if (!get_ia_constants(llc, &consts)) return; + /* + * Although this is unlikely on any platform during initialization, + * let's ensure we don't get accidentally into infinite loop + */ + if (consts.max_gpu_freq <= consts.min_gpu_freq) + return; /* * For each potential GPU frequency, load a ring frequency we'd like * to use for memory access. We do this by specifying the IA frequency diff --git a/drivers/gpu/drm/i915/gt/intel_rps.c b/drivers/gpu/drm/i915/gt/intel_rps.c index fb3f57ee450b..7bb967034679 100644 --- a/drivers/gpu/drm/i915/gt/intel_rps.c +++ b/drivers/gpu/drm/i915/gt/intel_rps.c @@ -2126,6 +2126,31 @@ u32 intel_rps_get_max_frequency(struct intel_rps *rps) return intel_gpu_freq(rps, rps->max_freq_softlimit); } +/** + * intel_rps_get_max_raw_freq - returns the max frequency in some raw format. + * @rps: the intel_rps structure + * + * Returns the max frequency in a raw format. In newer platforms raw is in + * units of 50 MHz. + */ +u32 intel_rps_get_max_raw_freq(struct intel_rps *rps) +{ + struct intel_guc_slpc *slpc = rps_to_slpc(rps); + u32 freq; + + if (rps_uses_slpc(rps)) { + return DIV_ROUND_CLOSEST(slpc->rp0_freq, + GT_FREQUENCY_MULTIPLIER); + } else { + freq = rps->max_freq; + if (GRAPHICS_VER(rps_to_i915(rps)) >= 9) { + /* Convert GT frequency to 50 MHz units */ + freq /= GEN9_FREQ_SCALER; + } + return freq; + } +} + u32 intel_rps_get_rp0_frequency(struct intel_rps *rps) { struct intel_guc_slpc *slpc = rps_to_slpc(rps); @@ -2214,6 +2239,31 @@ u32 intel_rps_get_min_frequency(struct intel_rps *rps) return intel_gpu_freq(rps, rps->min_freq_softlimit); } +/** + * intel_rps_get_min_raw_freq - returns the min frequency in some raw format. + * @rps: the intel_rps structure + * + * Returns the min frequency in a raw format. In newer platforms raw is in + * units of 50 MHz. + */ +u32 intel_rps_get_min_raw_freq(struct intel_rps *rps) +{ + struct intel_guc_slpc *slpc = rps_to_slpc(rps); + u32 freq; + + if (rps_uses_slpc(rps)) { + return DIV_ROUND_CLOSEST(slpc->min_freq, + GT_FREQUENCY_MULTIPLIER); + } else { + freq = rps->min_freq; + if (GRAPHICS_VER(rps_to_i915(rps)) >= 9) { + /* Convert GT frequency to 50 MHz units */ + freq /= GEN9_FREQ_SCALER; + } + return freq; + } +} + static int set_min_freq(struct intel_rps *rps, u32 val) { int ret = 0; diff --git a/drivers/gpu/drm/i915/gt/intel_rps.h b/drivers/gpu/drm/i915/gt/intel_rps.h index 1e8d56491308..4509dfdc52e0 100644 --- a/drivers/gpu/drm/i915/gt/intel_rps.h +++ b/drivers/gpu/drm/i915/gt/intel_rps.h @@ -37,8 +37,10 @@ u32 intel_rps_get_cagf(struct intel_rps *rps, u32 rpstat1); u32 intel_rps_read_actual_frequency(struct intel_rps *rps); u32 intel_rps_get_requested_frequency(struct intel_rps *rps); u32 intel_rps_get_min_frequency(struct intel_rps *rps); +u32 intel_rps_get_min_raw_freq(struct intel_rps *rps); int intel_rps_set_min_frequency(struct intel_rps *rps, u32 val); u32 intel_rps_get_max_frequency(struct intel_rps *rps); +u32 intel_rps_get_max_raw_freq(struct intel_rps *rps); int intel_rps_set_max_frequency(struct intel_rps *rps, u32 val); u32 intel_rps_get_rp0_frequency(struct intel_rps *rps); u32 intel_rps_get_rp1_frequency(struct intel_rps *rps); From 672d6ca758651f0ec12cd0d59787067a5bde1c96 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Fri, 2 Sep 2022 10:03:18 +0300 Subject: [PATCH 088/158] drm/i915: Implement WaEdpLinkRateDataReload MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A lot of modern laptops use the Parade PS8461E MUX for eDP switching. The MUX can operate in jitter cleaning mode or redriver mode, the first one resulting in higher link quality. The jitter cleaning mode needs to know the link rate used and the MUX achieves this by snooping the LINK_BW_SET, LINK_RATE_SELECT and SUPPORTED_LINK_RATES DPCD accesses. When the MUX is powered down (seems this can happen whenever the display is turned off) it loses track of the snooped link rates so when we do the LINK_RATE_SELECT write it no longer knowns which link rate we're selecting, and thus it falls back to the lower quality redriver mode. This results in unstable high link rates (eg. usually 8.1Gbps link rate no longer works correctly). In order to avoid all that let's re-snoop SUPPORTED_LINK_RATES from the sink at the start of every link training. Unfortunately we don't have a way to detect the presence of the MUX. It looks like the set of laptops equipped with this MUX is fairly large and contains devices from multiple manufacturers. It may also still be growing with new models. So a quirk doesn't seem like a very easily maintainable option, thus we shall attempt to do this unconditionally on all machines that use LINK_RATE_SELECT. Hopefully this extra DPCD read doesn't cause issues for any unaffected machine. If that turns out to be the case we'll need to convert this into a quirk in the future. Cc: stable@vger.kernel.org Cc: Jason A. Donenfeld Cc: Ankit Nautiyal Cc: Jani Nikula Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/6205 Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20220902070319.15395-1-ville.syrjala@linux.intel.com Tested-by: Aaron Ma Tested-by: Jason A. Donenfeld Reviewed-by: Jani Nikula (cherry picked from commit 25899c590cb5ba9b9f284c6ca8e7e9086793d641) Signed-off-by: Rodrigo Vivi --- .../drm/i915/display/intel_dp_link_training.c | 22 +++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/drivers/gpu/drm/i915/display/intel_dp_link_training.c b/drivers/gpu/drm/i915/display/intel_dp_link_training.c index 9feaf1a589f3..d213d8ad1ea5 100644 --- a/drivers/gpu/drm/i915/display/intel_dp_link_training.c +++ b/drivers/gpu/drm/i915/display/intel_dp_link_training.c @@ -671,6 +671,28 @@ intel_dp_prepare_link_train(struct intel_dp *intel_dp, intel_dp_compute_rate(intel_dp, crtc_state->port_clock, &link_bw, &rate_select); + /* + * WaEdpLinkRateDataReload + * + * Parade PS8461E MUX (used on varius TGL+ laptops) needs + * to snoop the link rates reported by the sink when we + * use LINK_RATE_SET in order to operate in jitter cleaning + * mode (as opposed to redriver mode). Unfortunately it + * loses track of the snooped link rates when powered down, + * so we need to make it re-snoop often. Without this high + * link rates are not stable. + */ + if (!link_bw) { + struct intel_connector *connector = intel_dp->attached_connector; + __le16 sink_rates[DP_MAX_SUPPORTED_RATES]; + + drm_dbg_kms(&i915->drm, "[CONNECTOR:%d:%s] Reloading eDP link rates\n", + connector->base.base.id, connector->base.name); + + drm_dp_dpcd_read(&intel_dp->aux, DP_SUPPORTED_LINK_RATES, + sink_rates, sizeof(sink_rates)); + } + if (link_bw) drm_dbg_kms(&i915->drm, "[ENCODER:%d:%s] Using LINK_BW_SET value %02x\n", From 151e0e0fdb4d6365305598af3b9072d44e896ef0 Mon Sep 17 00:00:00 2001 From: Matthew Auld Date: Mon, 5 Sep 2022 11:53:29 +0100 Subject: [PATCH 089/158] drm/i915: consider HAS_FLAT_CCS() in needs_ccs_pages Just move the HAS_FLAT_CCS() check into needs_ccs_pages. This also then fixes i915_ttm_memcpy_allowed() which was incorrectly reporting true on DG1, even though it doesn't have small-BAR or flat-CCS. References: https://gitlab.freedesktop.org/drm/intel/-/issues/6605 Fixes: efeb3caf4341 ("drm/i915/ttm: disallow CPU fallback mode for ccs pages") Signed-off-by: Matthew Auld Cc: Nirmoy Das Reviewed-by: Nirmoy Das Link: https://patchwork.freedesktop.org/patch/msgid/20220905105329.41455-1-matthew.auld@intel.com (cherry picked from commit 873fef8833ea794526b7f4179088e565078fe0e8) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/i915/gem/i915_gem_object.c | 3 +++ drivers/gpu/drm/i915/gem/i915_gem_ttm.c | 2 +- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.c b/drivers/gpu/drm/i915/gem/i915_gem_object.c index 389e9f157ca5..85482a04d158 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.c @@ -723,6 +723,9 @@ bool i915_gem_object_needs_ccs_pages(struct drm_i915_gem_object *obj) bool lmem_placement = false; int i; + if (!HAS_FLAT_CCS(to_i915(obj->base.dev))) + return false; + for (i = 0; i < obj->mm.n_placements; i++) { /* Compression is not allowed for the objects with smem placement */ if (obj->mm.placements[i]->type == INTEL_MEMORY_SYSTEM) diff --git a/drivers/gpu/drm/i915/gem/i915_gem_ttm.c b/drivers/gpu/drm/i915/gem/i915_gem_ttm.c index f131dc065f47..6f3ab7ade41a 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_ttm.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_ttm.c @@ -297,7 +297,7 @@ static struct ttm_tt *i915_ttm_tt_create(struct ttm_buffer_object *bo, i915_tt->is_shmem = true; } - if (HAS_FLAT_CCS(i915) && i915_gem_object_needs_ccs_pages(obj)) + if (i915_gem_object_needs_ccs_pages(obj)) ccs_pages = DIV_ROUND_UP(DIV_ROUND_UP(bo->base.size, NUM_BYTES_PER_CCS_BYTE), PAGE_SIZE); From baf2c002402702accc30dcc8f19199f303638306 Mon Sep 17 00:00:00 2001 From: Daniel Bristot de Oliveira Date: Tue, 23 Aug 2022 17:20:29 +0200 Subject: [PATCH 090/158] rv/monitors: Make monitor's automata definition static Monitor's automata definition is only used locally, so make them static for all existing monitors. Link: https://lore.kernel.org/all/202208210332.gtHXje45-lkp@intel.com Link: https://lore.kernel.org/all/202208210358.6HH3OrVs-lkp@intel.com Link: https://lkml.kernel.org/r/a50e27c3738d6ef809f4201857229fed64799234.1661266564.git.bristot@kernel.org Fixes: ccc319dcb450 ("rv/monitor: Add the wwnr monitor") Fixes: 8812d21219b9 ("rv/monitor: Add the wip monitor skeleton created by dot2k") Reported-by: kernel test robot Signed-off-by: Daniel Bristot de Oliveira Signed-off-by: Steven Rostedt (Google) --- kernel/trace/rv/monitors/wip/wip.h | 2 +- kernel/trace/rv/monitors/wwnr/wwnr.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/trace/rv/monitors/wip/wip.h b/kernel/trace/rv/monitors/wip/wip.h index c1c47e2305ef..dacc37b62a2c 100644 --- a/kernel/trace/rv/monitors/wip/wip.h +++ b/kernel/trace/rv/monitors/wip/wip.h @@ -27,7 +27,7 @@ struct automaton_wip { bool final_states[state_max_wip]; }; -struct automaton_wip automaton_wip = { +static struct automaton_wip automaton_wip = { .state_names = { "preemptive", "non_preemptive" diff --git a/kernel/trace/rv/monitors/wwnr/wwnr.h b/kernel/trace/rv/monitors/wwnr/wwnr.h index d1afe55cdd4c..118e576b91b4 100644 --- a/kernel/trace/rv/monitors/wwnr/wwnr.h +++ b/kernel/trace/rv/monitors/wwnr/wwnr.h @@ -27,7 +27,7 @@ struct automaton_wwnr { bool final_states[state_max_wwnr]; }; -struct automaton_wwnr automaton_wwnr = { +static struct automaton_wwnr automaton_wwnr = { .state_names = { "not_running", "running" From fa4b9df00af4a749d4d715d7e11562121b7885ac Mon Sep 17 00:00:00 2001 From: Daniel Bristot de Oliveira Date: Tue, 23 Aug 2022 17:33:16 +0200 Subject: [PATCH 091/158] MAINTAINERS: Add Runtime Verification (RV) entry Add a Runtime Verification (RV) entry in the MAINTAINERS file with Steven Rostedt and myself as maintainers. Link: https://lkml.kernel.org/r/b24c13553b6947a8da16d884ca464e4233eb8fb7.1661268579.git.bristot@kernel.org Cc: Joe Perches Signed-off-by: Daniel Bristot de Oliveira Signed-off-by: Steven Rostedt (Google) --- MAINTAINERS | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index d30f26e07cd3..b89b70af0b07 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17735,6 +17735,17 @@ L: linux-rdma@vger.kernel.org S: Maintained F: drivers/infiniband/ulp/rtrs/ +RUNTIME VERIFICATION (RV) +M: Daniel Bristot de Oliveira +M: Steven Rostedt +L: linux-trace-devel@vger.kernel.org +S: Maintained +F: Documentation/trace/rv/ +F: include/linux/rv.h +F: include/rv/ +F: kernel/trace/rv/ +F: tools/verification/ + RXRPC SOCKETS (AF_RXRPC) M: David Howells M: Marc Dionne From f72c3a2d344b3933dedd42fdaf7d3e1aa4b8cce1 Mon Sep 17 00:00:00 2001 From: Lukas Bulwahn Date: Thu, 25 Aug 2022 13:59:27 +0200 Subject: [PATCH 092/158] MAINTAINERS: add scripts/tracing/ to TRACING The files in scripts/tracing/ belong to the TRACING subsystem. Add a corresponding file entry for TRACING. Link: https://lkml.kernel.org/r/20220825115927.20598-1-lukas.bulwahn@gmail.com Signed-off-by: Lukas Bulwahn Signed-off-by: Steven Rostedt (Google) --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index b89b70af0b07..93ffebc3e6c0 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -20612,6 +20612,7 @@ F: include/*/ftrace.h F: include/linux/trace*.h F: include/trace/ F: kernel/trace/ +F: scripts/tracing/ F: tools/testing/selftests/ftrace/ TRACING MMIO ACCESSES (MMIOTRACE) From 54be5509422ebee333709c92441aa7185ca182fa Mon Sep 17 00:00:00 2001 From: Alison Schofield Date: Mon, 29 Aug 2022 10:10:48 -0700 Subject: [PATCH 093/158] tracepoint: Allow trace events in modules with TAINT_TEST Commit 2852ca7fba9f ("panic: Taint kernel if tests are run") introduced a new taint type, TAINT_TEST, to signal that an in-kernel test module has been loaded. TAINT_TEST taint type defaults into a 'bad_taint' list for kernel tracing and blocks the creation of trace events. This causes a problem for CXL testing where loading the cxl_test module makes all CXL modules out-of-tree, blocking any trace events. Trace events are in development for CXL at the moment and this issue was found in test with v6.0-rc1. Link: https://lkml.kernel.org/r/20220829171048.263065-1-alison.schofield@intel.com Fixes: 2852ca7fba9f7 ("panic: Taint kernel if tests are run") Reported-by: Ira Weiny Suggested-by: Dan Williams Tested-by: Ira Weiny Reviewed-by: David Gow Signed-off-by: Alison Schofield Signed-off-by: Steven Rostedt (Google) --- kernel/tracepoint.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/kernel/tracepoint.c b/kernel/tracepoint.c index 64ea283f2f86..ef42c1a11920 100644 --- a/kernel/tracepoint.c +++ b/kernel/tracepoint.c @@ -571,7 +571,8 @@ static void for_each_tracepoint_range( bool trace_module_has_bad_taint(struct module *mod) { return mod->taints & ~((1 << TAINT_OOT_MODULE) | (1 << TAINT_CRAP) | - (1 << TAINT_UNSIGNED_MODULE)); + (1 << TAINT_UNSIGNED_MODULE) | + (1 << TAINT_TEST)); } static BLOCKING_NOTIFIER_HEAD(tracepoint_notify_list); @@ -647,7 +648,7 @@ static int tracepoint_module_coming(struct module *mod) /* * We skip modules that taint the kernel, especially those with different * module headers (for forced load), to make sure we don't cause a crash. - * Staging, out-of-tree, and unsigned GPL modules are fine. + * Staging, out-of-tree, unsigned GPL, and test modules are fine. */ if (trace_module_has_bad_taint(mod)) return 0; From 54c3931957f6a6194d5972eccc36d052964b2abe Mon Sep 17 00:00:00 2001 From: Yipeng Zou Date: Thu, 1 Sep 2022 18:45:14 +0800 Subject: [PATCH 094/158] tracing: hold caller_addr to hardirq_{enable,disable}_ip Currently, The arguments passing to lockdep_hardirqs_{on,off} was fixed in CALLER_ADDR0. The function trace_hardirqs_on_caller should have been intended to use caller_addr to represent the address that caller wants to be traced. For example, lockdep log in riscv showing the last {enabled,disabled} at __trace_hardirqs_{on,off} all the time(if called by): [ 57.853175] hardirqs last enabled at (2519): __trace_hardirqs_on+0xc/0x14 [ 57.853848] hardirqs last disabled at (2520): __trace_hardirqs_off+0xc/0x14 After use trace_hardirqs_xx_caller, we can get more effective information: [ 53.781428] hardirqs last enabled at (2595): restore_all+0xe/0x66 [ 53.782185] hardirqs last disabled at (2596): ret_from_exception+0xa/0x10 Link: https://lkml.kernel.org/r/20220901104515.135162-2-zouyipeng@huawei.com Cc: stable@vger.kernel.org Fixes: c3bc8fd637a96 ("tracing: Centralize preemptirq tracepoints and unify their usage") Signed-off-by: Yipeng Zou Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace_preemptirq.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/trace/trace_preemptirq.c b/kernel/trace/trace_preemptirq.c index 95b58bd757ce..1e130da1b742 100644 --- a/kernel/trace/trace_preemptirq.c +++ b/kernel/trace/trace_preemptirq.c @@ -95,14 +95,14 @@ __visible void trace_hardirqs_on_caller(unsigned long caller_addr) } lockdep_hardirqs_on_prepare(); - lockdep_hardirqs_on(CALLER_ADDR0); + lockdep_hardirqs_on(caller_addr); } EXPORT_SYMBOL(trace_hardirqs_on_caller); NOKPROBE_SYMBOL(trace_hardirqs_on_caller); __visible void trace_hardirqs_off_caller(unsigned long caller_addr) { - lockdep_hardirqs_off(CALLER_ADDR0); + lockdep_hardirqs_off(caller_addr); if (!this_cpu_read(tracing_irq_cpu)) { this_cpu_write(tracing_irq_cpu, 1); From cecf8e128ec69149fe53c9a7bafa505a4bee25d9 Mon Sep 17 00:00:00 2001 From: "Masami Hiramatsu (Google)" Date: Sun, 4 Sep 2022 13:12:29 +0900 Subject: [PATCH 095/158] tracing: Fix to check event_mutex is held while accessing trigger list Since the check_user_trigger() is called outside of RCU read lock, this list_for_each_entry_rcu() caused a suspicious RCU usage warning. # echo hist:keys=pid > events/sched/sched_stat_runtime/trigger # cat events/sched/sched_stat_runtime/trigger [ 43.167032] [ 43.167418] ============================= [ 43.167992] WARNING: suspicious RCU usage [ 43.168567] 5.19.0-rc5-00029-g19ebe4651abf #59 Not tainted [ 43.169283] ----------------------------- [ 43.169863] kernel/trace/trace_events_trigger.c:145 RCU-list traversed in non-reader section!! ... However, this file->triggers list is safe when it is accessed under event_mutex is held. To fix this warning, adds a lockdep_is_held check to the list_for_each_entry_rcu(). Link: https://lkml.kernel.org/r/166226474977.223837.1992182913048377113.stgit@devnote2 Cc: stable@vger.kernel.org Fixes: 7491e2c44278 ("tracing: Add a probe that attaches to trace events") Signed-off-by: Masami Hiramatsu (Google) Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace_events_trigger.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/kernel/trace/trace_events_trigger.c b/kernel/trace/trace_events_trigger.c index cb866c3141af..918730d74932 100644 --- a/kernel/trace/trace_events_trigger.c +++ b/kernel/trace/trace_events_trigger.c @@ -142,7 +142,8 @@ static bool check_user_trigger(struct trace_event_file *file) { struct event_trigger_data *data; - list_for_each_entry_rcu(data, &file->triggers, list) { + list_for_each_entry_rcu(data, &file->triggers, list, + lockdep_is_held(&event_mutex)) { if (data->flags & EVENT_TRIGGER_FL_PROBE) continue; return true; From 93d71986a60c37395e0a1d2f7fe170ef1e0a8ba4 Mon Sep 17 00:00:00 2001 From: Xiu Jianfeng Date: Tue, 6 Sep 2022 22:12:10 +0800 Subject: [PATCH 096/158] rv/reactor: add __init/__exit annotations to module init/exit funcs Add missing __init/__exit annotations to module init/exit funcs. Link: https://lkml.kernel.org/r/20220906141210.132607-1-xiujianfeng@huawei.com Fixes: 135b881ea885 ("rv/reactor: Add the printk reactor") Fixes: e88043c0ac16 ("rv/reactor: Add the panic reactor") Signed-off-by: Xiu Jianfeng Acked-by: Daniel Bristot de Oliveira Signed-off-by: Steven Rostedt (Google) --- kernel/trace/rv/reactor_panic.c | 4 ++-- kernel/trace/rv/reactor_printk.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/kernel/trace/rv/reactor_panic.c b/kernel/trace/rv/reactor_panic.c index b698d05dd069..d65f6c25a87c 100644 --- a/kernel/trace/rv/reactor_panic.c +++ b/kernel/trace/rv/reactor_panic.c @@ -24,13 +24,13 @@ static struct rv_reactor rv_panic = { .react = rv_panic_reaction }; -static int register_react_panic(void) +static int __init register_react_panic(void) { rv_register_reactor(&rv_panic); return 0; } -static void unregister_react_panic(void) +static void __exit unregister_react_panic(void) { rv_unregister_reactor(&rv_panic); } diff --git a/kernel/trace/rv/reactor_printk.c b/kernel/trace/rv/reactor_printk.c index 31899f953af4..4b6b7106a477 100644 --- a/kernel/trace/rv/reactor_printk.c +++ b/kernel/trace/rv/reactor_printk.c @@ -23,13 +23,13 @@ static struct rv_reactor rv_printk = { .react = rv_printk_reaction }; -static int register_react_printk(void) +static int __init register_react_printk(void) { rv_register_reactor(&rv_printk); return 0; } -static void unregister_react_printk(void) +static void __exit unregister_react_printk(void) { rv_unregister_reactor(&rv_printk); } From 991df3dd5144f2e6b1c38b8d20ed3d4d21e20b34 Mon Sep 17 00:00:00 2001 From: Sreekanth Reddy Date: Tue, 6 Sep 2022 19:19:08 +0530 Subject: [PATCH 097/158] scsi: mpt3sas: Fix use-after-free warning Fix the following use-after-free warning which is observed during controller reset: refcount_t: underflow; use-after-free. WARNING: CPU: 23 PID: 5399 at lib/refcount.c:28 refcount_warn_saturate+0xa6/0xf0 Link: https://lore.kernel.org/r/20220906134908.1039-2-sreekanth.reddy@broadcom.com Signed-off-by: Sreekanth Reddy Signed-off-by: Martin K. Petersen --- drivers/scsi/mpt3sas/mpt3sas_scsih.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/mpt3sas/mpt3sas_scsih.c b/drivers/scsi/mpt3sas/mpt3sas_scsih.c index def37a7e5980..bd6a5f1bd532 100644 --- a/drivers/scsi/mpt3sas/mpt3sas_scsih.c +++ b/drivers/scsi/mpt3sas/mpt3sas_scsih.c @@ -3670,6 +3670,7 @@ static struct fw_event_work *dequeue_next_fw_event(struct MPT3SAS_ADAPTER *ioc) fw_event = list_first_entry(&ioc->fw_event_list, struct fw_event_work, list); list_del_init(&fw_event->list); + fw_event_work_put(fw_event); } spin_unlock_irqrestore(&ioc->fw_event_lock, flags); @@ -3751,7 +3752,6 @@ _scsih_fw_event_cleanup_queue(struct MPT3SAS_ADAPTER *ioc) if (cancel_work_sync(&fw_event->work)) fw_event_work_put(fw_event); - fw_event_work_put(fw_event); } ioc->fw_events_cleanup = 0; } From d29f59051d3a07b81281b2df2b8c9dfe4716067f Mon Sep 17 00:00:00 2001 From: Tasos Sahanidis Date: Wed, 7 Sep 2022 04:18:00 +0300 Subject: [PATCH 098/158] ALSA: emu10k1: Fix out of bounds access in snd_emu10k1_pcm_channel_alloc() The voice allocator sometimes begins allocating from near the end of the array and then wraps around, however snd_emu10k1_pcm_channel_alloc() accesses the newly allocated voices as if it never wrapped around. This results in out of bounds access if the first voice has a high enough index so that first_voice + requested_voice_count > NUM_G (64). The more voices are requested, the more likely it is for this to occur. This was initially discovered using PipeWire, however it can be reproduced by calling aplay multiple times with 16 channels: aplay -r 48000 -D plughw:CARD=Live,DEV=3 -c 16 /dev/zero UBSAN: array-index-out-of-bounds in sound/pci/emu10k1/emupcm.c:127:40 index 65 is out of range for type 'snd_emu10k1_voice [64]' CPU: 1 PID: 31977 Comm: aplay Tainted: G W IOE 6.0.0-rc2-emu10k1+ #7 Hardware name: ASUSTEK COMPUTER INC P5W DH Deluxe/P5W DH Deluxe, BIOS 3002 07/22/2010 Call Trace: dump_stack_lvl+0x49/0x63 dump_stack+0x10/0x16 ubsan_epilogue+0x9/0x3f __ubsan_handle_out_of_bounds.cold+0x44/0x49 snd_emu10k1_playback_hw_params+0x3bc/0x420 [snd_emu10k1] snd_pcm_hw_params+0x29f/0x600 [snd_pcm] snd_pcm_common_ioctl+0x188/0x1410 [snd_pcm] ? exit_to_user_mode_prepare+0x35/0x170 ? do_syscall_64+0x69/0x90 ? syscall_exit_to_user_mode+0x26/0x50 ? do_syscall_64+0x69/0x90 ? exit_to_user_mode_prepare+0x35/0x170 snd_pcm_ioctl+0x27/0x40 [snd_pcm] __x64_sys_ioctl+0x95/0xd0 do_syscall_64+0x5c/0x90 ? do_syscall_64+0x69/0x90 ? do_syscall_64+0x69/0x90 entry_SYSCALL_64_after_hwframe+0x63/0xcd Signed-off-by: Tasos Sahanidis Cc: Link: https://lore.kernel.org/r/3707dcab-320a-62ff-63c0-73fc201ef756@tasossah.com Signed-off-by: Takashi Iwai --- sound/pci/emu10k1/emupcm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/pci/emu10k1/emupcm.c b/sound/pci/emu10k1/emupcm.c index b2701a4452d8..48af77ae8020 100644 --- a/sound/pci/emu10k1/emupcm.c +++ b/sound/pci/emu10k1/emupcm.c @@ -124,7 +124,7 @@ static int snd_emu10k1_pcm_channel_alloc(struct snd_emu10k1_pcm * epcm, int voic epcm->voices[0]->epcm = epcm; if (voices > 1) { for (i = 1; i < voices; i++) { - epcm->voices[i] = &epcm->emu->voices[epcm->voices[0]->number + i]; + epcm->voices[i] = &epcm->emu->voices[(epcm->voices[0]->number + i) % NUM_G]; epcm->voices[i]->epcm = epcm; } } From b7e97872a65e1d57b4451769610554c131f37a0a Mon Sep 17 00:00:00 2001 From: Dennis Maisenbacher Date: Tue, 6 Sep 2022 09:39:28 +0200 Subject: [PATCH 099/158] nvmet: fix mar and mor off-by-one errors Maximum Active Resources (MAR) and Maximum Open Resources (MOR) are 0's based vales where a value of 0xffffffff indicates that there is no limit. Decrement the values that are returned by bdev_max_open_zones and bdev_max_active_zones as the block layer helpers are not 0's based. A 0 returned by the block layer helpers indicates no limit, thus convert it to 0xffffffff (U32_MAX). Fixes: aaf2e048af27 ("nvmet: add ZBD over ZNS backend support") Suggested-by: Niklas Cassel Signed-off-by: Dennis Maisenbacher Reviewed-by: Chaitanya Kulkarni Signed-off-by: Christoph Hellwig --- drivers/nvme/target/zns.c | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-) diff --git a/drivers/nvme/target/zns.c b/drivers/nvme/target/zns.c index c7ef69f29fe4..835bfda86fcf 100644 --- a/drivers/nvme/target/zns.c +++ b/drivers/nvme/target/zns.c @@ -100,6 +100,7 @@ void nvmet_execute_identify_cns_cs_ns(struct nvmet_req *req) struct nvme_id_ns_zns *id_zns; u64 zsze; u16 status; + u32 mar, mor; if (le32_to_cpu(req->cmd->identify.nsid) == NVME_NSID_ALL) { req->error_loc = offsetof(struct nvme_identify, nsid); @@ -130,8 +131,20 @@ void nvmet_execute_identify_cns_cs_ns(struct nvmet_req *req) zsze = (bdev_zone_sectors(req->ns->bdev) << 9) >> req->ns->blksize_shift; id_zns->lbafe[0].zsze = cpu_to_le64(zsze); - id_zns->mor = cpu_to_le32(bdev_max_open_zones(req->ns->bdev)); - id_zns->mar = cpu_to_le32(bdev_max_active_zones(req->ns->bdev)); + + mor = bdev_max_open_zones(req->ns->bdev); + if (!mor) + mor = U32_MAX; + else + mor--; + id_zns->mor = cpu_to_le32(mor); + + mar = bdev_max_active_zones(req->ns->bdev); + if (!mar) + mar = U32_MAX; + else + mar--; + id_zns->mar = cpu_to_le32(mar); done: status = nvmet_copy_to_sgl(req, 0, id_zns, sizeof(*id_zns)); From 371a982cd2b01487295cd87abec82357e268457b Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Thu, 1 Sep 2022 08:30:39 -0700 Subject: [PATCH 100/158] nvme: requeue aen after firmware activation The driver prevents async event work while handling a processing paused event, but someone needs to restart it after the controller returns to a live state. Link: https://bugzilla.kernel.org/show_bug.cgi?id=216400 Signed-off-by: Keith Busch Signed-off-by: Christoph Hellwig --- drivers/nvme/host/core.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index 2429b11eb9a8..70ebf27ad10e 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -4702,6 +4702,8 @@ static void nvme_fw_act_work(struct work_struct *work) nvme_start_queues(ctrl); /* read FW slot information to clear the AER */ nvme_get_fw_slot_info(ctrl); + + queue_work(nvme_wq, &ctrl->async_event_work); } static u32 nvme_aer_type(u32 result) @@ -4714,9 +4716,10 @@ static u32 nvme_aer_subtype(u32 result) return (result & 0xff00) >> 8; } -static void nvme_handle_aen_notice(struct nvme_ctrl *ctrl, u32 result) +static bool nvme_handle_aen_notice(struct nvme_ctrl *ctrl, u32 result) { u32 aer_notice_type = nvme_aer_subtype(result); + bool requeue = true; trace_nvme_async_event(ctrl, aer_notice_type); @@ -4733,6 +4736,7 @@ static void nvme_handle_aen_notice(struct nvme_ctrl *ctrl, u32 result) */ if (nvme_change_ctrl_state(ctrl, NVME_CTRL_RESETTING)) { nvme_auth_stop(ctrl); + requeue = false; queue_work(nvme_wq, &ctrl->fw_act_work); } break; @@ -4749,6 +4753,7 @@ static void nvme_handle_aen_notice(struct nvme_ctrl *ctrl, u32 result) default: dev_warn(ctrl->device, "async event result %08x\n", result); } + return requeue; } static void nvme_handle_aer_persistent_error(struct nvme_ctrl *ctrl) @@ -4764,13 +4769,14 @@ void nvme_complete_async_event(struct nvme_ctrl *ctrl, __le16 status, u32 result = le32_to_cpu(res->u32); u32 aer_type = nvme_aer_type(result); u32 aer_subtype = nvme_aer_subtype(result); + bool requeue = true; if (le16_to_cpu(status) >> 1 != NVME_SC_SUCCESS) return; switch (aer_type) { case NVME_AER_NOTICE: - nvme_handle_aen_notice(ctrl, result); + requeue = nvme_handle_aen_notice(ctrl, result); break; case NVME_AER_ERROR: /* @@ -4791,7 +4797,9 @@ void nvme_complete_async_event(struct nvme_ctrl *ctrl, __le16 status, default: break; } - queue_work(nvme_wq, &ctrl->async_event_work); + + if (requeue) + queue_work(nvme_wq, &ctrl->async_event_work); } EXPORT_SYMBOL_GPL(nvme_complete_async_event); From 12faad5e5cf2372af2d51f348b697b5edf838daf Mon Sep 17 00:00:00 2001 From: Sindhu-Devale Date: Tue, 6 Sep 2022 17:32:40 -0500 Subject: [PATCH 101/158] RDMA/irdma: Report the correct max cqes from query device Report the correct max cqes available to an application taking into account a reserved entry to detect overflow. Fixes: b48c24c2d710 ("RDMA/irdma: Implement device supported verb APIs") Signed-off-by: Sindhu-Devale Signed-off-by: Shiraz Saleem Link: https://lore.kernel.org/r/20220906223244.1119-2-shiraz.saleem@intel.com Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/irdma/verbs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/irdma/verbs.c b/drivers/infiniband/hw/irdma/verbs.c index 9b07b8af2997..f272a323d065 100644 --- a/drivers/infiniband/hw/irdma/verbs.c +++ b/drivers/infiniband/hw/irdma/verbs.c @@ -39,7 +39,7 @@ static int irdma_query_device(struct ib_device *ibdev, props->max_send_sge = hw_attrs->uk_attrs.max_hw_wq_frags; props->max_recv_sge = hw_attrs->uk_attrs.max_hw_wq_frags; props->max_cq = rf->max_cq - rf->used_cqs; - props->max_cqe = rf->max_cqe; + props->max_cqe = rf->max_cqe - 1; props->max_mr = rf->max_mr - rf->used_mrs; props->max_mw = props->max_mr; props->max_pd = rf->max_pd - rf->used_pds; From 6b227bd32db778eddc6f3b22cc72a28dda0f2272 Mon Sep 17 00:00:00 2001 From: Sindhu-Devale Date: Tue, 6 Sep 2022 17:32:41 -0500 Subject: [PATCH 102/158] RDMA/irdma: Return error on MR deregister CQP failure The MR deregister CQP can fail if an MW is bound to it. Return an appropriate error for this case. Fixes: b48c24c2d710 ("RDMA/irdma: Implement device supported verb APIs") Signed-off-by: Sindhu-Devale Signed-off-by: Shiraz Saleem Link: https://lore.kernel.org/r/20220906223244.1119-3-shiraz.saleem@intel.com Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/irdma/utils.c | 13 ++++++++----- drivers/infiniband/hw/irdma/verbs.c | 6 +++++- 2 files changed, 13 insertions(+), 6 deletions(-) diff --git a/drivers/infiniband/hw/irdma/utils.c b/drivers/infiniband/hw/irdma/utils.c index c8b92355ddea..075defaabee5 100644 --- a/drivers/infiniband/hw/irdma/utils.c +++ b/drivers/infiniband/hw/irdma/utils.c @@ -590,11 +590,14 @@ static int irdma_wait_event(struct irdma_pci_f *rf, cqp_error = cqp_request->compl_info.error; if (cqp_error) { err_code = -EIO; - if (cqp_request->compl_info.maj_err_code == 0xFFFF && - cqp_request->compl_info.min_err_code == 0x8029) { - if (!rf->reset) { - rf->reset = true; - rf->gen_ops.request_reset(rf); + if (cqp_request->compl_info.maj_err_code == 0xFFFF) { + if (cqp_request->compl_info.min_err_code == 0x8002) + err_code = -EBUSY; + else if (cqp_request->compl_info.min_err_code == 0x8029) { + if (!rf->reset) { + rf->reset = true; + rf->gen_ops.request_reset(rf); + } } } } diff --git a/drivers/infiniband/hw/irdma/verbs.c b/drivers/infiniband/hw/irdma/verbs.c index f272a323d065..892467c91262 100644 --- a/drivers/infiniband/hw/irdma/verbs.c +++ b/drivers/infiniband/hw/irdma/verbs.c @@ -3009,6 +3009,7 @@ static int irdma_dereg_mr(struct ib_mr *ib_mr, struct ib_udata *udata) struct irdma_pble_alloc *palloc = &iwpbl->pble_alloc; struct irdma_cqp_request *cqp_request; struct cqp_cmds_info *cqp_info; + int status; if (iwmr->type != IRDMA_MEMREG_TYPE_MEM) { if (iwmr->region) { @@ -3039,8 +3040,11 @@ static int irdma_dereg_mr(struct ib_mr *ib_mr, struct ib_udata *udata) cqp_info->post_sq = 1; cqp_info->in.u.dealloc_stag.dev = &iwdev->rf->sc_dev; cqp_info->in.u.dealloc_stag.scratch = (uintptr_t)cqp_request; - irdma_handle_cqp_op(iwdev->rf, cqp_request); + status = irdma_handle_cqp_op(iwdev->rf, cqp_request); irdma_put_cqp_request(&iwdev->rf->cqp, cqp_request); + if (status) + return status; + irdma_free_stag(iwdev, iwmr->stag); done: if (iwpbl->pbl_allocated) From dcb23bbb1de7e009875fdfac2b8a9808a9319cc6 Mon Sep 17 00:00:00 2001 From: Sindhu-Devale Date: Tue, 6 Sep 2022 17:32:42 -0500 Subject: [PATCH 103/158] RDMA/irdma: Return correct WC error for bind operation failure When a QP and a MR on a local host are in different PDs, the HW generates an asynchronous event (AE). The same AE is generated when a QP and a MW are in different PDs during a bind operation. Return the more appropriate IBV_WC_MW_BIND_ERR for the latter case by checking the OP type from the CQE in error. Fixes: 551c46edc769 ("RDMA/irdma: Add user/kernel shared libraries") Signed-off-by: Sindhu-Devale Signed-off-by: Shiraz Saleem Link: https://lore.kernel.org/r/20220906223244.1119-4-shiraz.saleem@intel.com Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/irdma/uk.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/irdma/uk.c b/drivers/infiniband/hw/irdma/uk.c index daeab5daed5b..d003ad864ee4 100644 --- a/drivers/infiniband/hw/irdma/uk.c +++ b/drivers/infiniband/hw/irdma/uk.c @@ -1005,6 +1005,7 @@ int irdma_uk_cq_poll_cmpl(struct irdma_cq_uk *cq, int ret_code; bool move_cq_head = true; u8 polarity; + u8 op_type; bool ext_valid; __le64 *ext_cqe; @@ -1187,7 +1188,6 @@ int irdma_uk_cq_poll_cmpl(struct irdma_cq_uk *cq, do { __le64 *sw_wqe; u64 wqe_qword; - u8 op_type; u32 tail; tail = qp->sq_ring.tail; @@ -1204,6 +1204,8 @@ int irdma_uk_cq_poll_cmpl(struct irdma_cq_uk *cq, break; } } while (1); + if (op_type == IRDMA_OP_TYPE_BIND_MW && info->minor_err == FLUSH_PROT_ERR) + info->minor_err = FLUSH_MW_BIND_ERR; qp->sq_flush_seen = true; if (!IRDMA_RING_MORE_WORK(qp->sq_ring)) qp->sq_flush_complete = true; From 2c8844431d065ae15a6b442f5769b60aeaaa07af Mon Sep 17 00:00:00 2001 From: Sindhu-Devale Date: Tue, 6 Sep 2022 17:32:43 -0500 Subject: [PATCH 104/158] RDMA/irdma: Use s/g array in post send only when its valid Send with invalidate verb call can pass in an uninitialized s/g array with 0 sge's which is filled into irdma WQE and causes a HW asynchronous event. Fix this by using the s/g array in irdma post send only when its valid. Fixes: 551c46e ("RDMA/irdma: Add user/kernel shared libraries") Signed-off-by: Sindhu-Devale Signed-off-by: Shiraz Saleem Link: https://lore.kernel.org/r/20220906223244.1119-5-shiraz.saleem@intel.com Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/irdma/uk.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/irdma/uk.c b/drivers/infiniband/hw/irdma/uk.c index d003ad864ee4..a6e5d350a94c 100644 --- a/drivers/infiniband/hw/irdma/uk.c +++ b/drivers/infiniband/hw/irdma/uk.c @@ -497,7 +497,8 @@ int irdma_uk_send(struct irdma_qp_uk *qp, struct irdma_post_sq_info *info, FIELD_PREP(IRDMAQPSQ_IMMDATA, info->imm_data)); i = 0; } else { - qp->wqe_ops.iw_set_fragment(wqe, 0, op_info->sg_list, + qp->wqe_ops.iw_set_fragment(wqe, 0, + frag_cnt ? op_info->sg_list : NULL, qp->swqe_polarity); i = 1; } From a261786fdc0a5bed2e5f994dcc0ffeeeb0d662c7 Mon Sep 17 00:00:00 2001 From: Sindhu-Devale Date: Tue, 6 Sep 2022 17:32:44 -0500 Subject: [PATCH 105/158] RDMA/irdma: Report RNR NAK generation in device caps Report RNR NAK generation when device capabilities are queried Fixes: b48c24c2d710 ("RDMA/irdma: Implement device supported verb APIs") Signed-off-by: Sindhu-Devale Signed-off-by: Shiraz Saleem Link: https://lore.kernel.org/r/20220906223244.1119-6-shiraz.saleem@intel.com Signed-off-by: Leon Romanovsky --- drivers/infiniband/hw/irdma/verbs.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/irdma/verbs.c b/drivers/infiniband/hw/irdma/verbs.c index 892467c91262..9b207f5084eb 100644 --- a/drivers/infiniband/hw/irdma/verbs.c +++ b/drivers/infiniband/hw/irdma/verbs.c @@ -46,8 +46,11 @@ static int irdma_query_device(struct ib_device *ibdev, props->max_sge_rd = hw_attrs->uk_attrs.max_hw_read_sges; props->max_qp_rd_atom = hw_attrs->max_hw_ird; props->max_qp_init_rd_atom = hw_attrs->max_hw_ord; - if (rdma_protocol_roce(ibdev, 1)) + if (rdma_protocol_roce(ibdev, 1)) { + props->device_cap_flags |= IB_DEVICE_RC_RNR_NAK_GEN; props->max_pkeys = IRDMA_PKEY_TBL_SZ; + } + props->max_ah = rf->max_ah; props->max_mcast_grp = rf->max_mcg; props->max_mcast_qp_attach = IRDMA_MAX_MGS_PER_CTX; From 81c12e922b97b94f64e84aa9bfe5a9d1fca2dc25 Mon Sep 17 00:00:00 2001 From: Yu Zhao Date: Wed, 31 Aug 2022 00:38:18 -0600 Subject: [PATCH 106/158] Revert "swiotlb: panic if nslabs is too small" This reverts commit 0bf28fc40d89b1a3e00d1b79473bad4e9ca20ad1. Reasons: 1. new panic()s shouldn't be added [1]. 2. It does no "cleanup" but breaks MIPS [2]. v2: properly solved the conflict [3] with commit 20347fca71a38 ("swiotlb: split up the global swiotlb lock") Reported-by: kernel test robot Reported-by: Dan Carpenter [1] https://lore.kernel.org/r/CAHk-=wit-DmhMfQErY29JSPjFgebx_Ld+pnerc4J2Ag990WwAA@mail.gmail.com/ [2] https://lore.kernel.org/r/20220820012031.1285979-1-yuzhao@google.com/ [3] https://lore.kernel.org/r/202208310701.LKr1WDCh-lkp@intel.com/ Fixes: 0bf28fc40d89b ("swiotlb: panic if nslabs is too small") Signed-off-by: Yu Zhao Signed-off-by: Christoph Hellwig --- kernel/dma/swiotlb.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c index c5a9190b218f..dd8863987e0c 100644 --- a/kernel/dma/swiotlb.c +++ b/kernel/dma/swiotlb.c @@ -326,9 +326,6 @@ void __init swiotlb_init_remap(bool addressing_limit, unsigned int flags, swiotlb_adjust_nareas(num_possible_cpus()); nslabs = default_nslabs; - if (nslabs < IO_TLB_MIN_SLABS) - panic("%s: nslabs = %lu too small\n", __func__, nslabs); - /* * By default allocate the bounce buffer memory from low memory, but * allow to pick a location everywhere for hypervisors with guest @@ -341,8 +338,7 @@ retry: else tlb = memblock_alloc_low(bytes, PAGE_SIZE); if (!tlb) { - pr_warn("%s: Failed to allocate %zu bytes tlb structure\n", - __func__, bytes); + pr_warn("%s: failed to allocate tlb structure\n", __func__); return; } From 2995b8002cefc7b0b00b8f9a0ce36601a8c390c0 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Mon, 15 Aug 2022 20:28:40 +0100 Subject: [PATCH 107/158] dma-debug: improve search for partial syncs When bucket_find_contains() tries to find the original entry for a partial sync, it manages to constrain its search in a way that is both too restrictive and not restrictive enough. A driver which only uses single mappings rather than scatterlists might not set max_seg_size, but could still technically perform a partial sync at an offset of more than 64KB into a sufficiently large mapping, so we could stop searching too early before reaching a legitimate entry. Conversely, if no valid entry is present and max_range is large enough, we can pointlessly search buckets that we've already searched, or that represent an impossible wrapping around the bottom of the address space. At worst, the (legitimate) case of max_seg_size == UINT_MAX can make the loop infinite. Replace the fragile and frankly hard-to-follow "range" logic with a simple counted loop for the number of possible hash buckets below the given address. Reported-by: Yunfei Wang Signed-off-by: Robin Murphy Signed-off-by: Christoph Hellwig --- kernel/dma/debug.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/kernel/dma/debug.c b/kernel/dma/debug.c index 2caafd13f8aa..18c93c2276ca 100644 --- a/kernel/dma/debug.c +++ b/kernel/dma/debug.c @@ -350,11 +350,10 @@ static struct dma_debug_entry *bucket_find_contain(struct hash_bucket **bucket, unsigned long *flags) { - unsigned int max_range = dma_get_max_seg_size(ref->dev); struct dma_debug_entry *entry, index = *ref; - unsigned int range = 0; + int limit = min(HASH_SIZE, (index.dev_addr >> HASH_FN_SHIFT) + 1); - while (range <= max_range) { + for (int i = 0; i < limit; i++) { entry = __hash_bucket_find(*bucket, ref, containing_match); if (entry) @@ -364,7 +363,6 @@ static struct dma_debug_entry *bucket_find_contain(struct hash_bucket **bucket, * Nothing found, go back a hash bucket */ put_hash_bucket(*bucket, *flags); - range += (1 << HASH_FN_SHIFT); index.dev_addr -= (1 << HASH_FN_SHIFT); *bucket = get_hash_bucket(&index, flags); } From 3f0461613ebcdc8c4073e235053d06d5aa58750f Mon Sep 17 00:00:00 2001 From: Chao Gao Date: Fri, 19 Aug 2022 16:45:37 +0800 Subject: [PATCH 108/158] swiotlb: avoid potential left shift overflow The second operand passed to slot_addr() is declared as int or unsigned int in all call sites. The left-shift to get the offset of a slot can overflow if swiotlb size is larger than 4G. Convert the macro to an inline function and declare the second argument as phys_addr_t to avoid the potential overflow. Fixes: 26a7e094783d ("swiotlb: refactor swiotlb_tbl_map_single") Signed-off-by: Chao Gao Reviewed-by: Dongli Zhang Signed-off-by: Christoph Hellwig --- kernel/dma/swiotlb.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c index dd8863987e0c..1ce8977d911c 100644 --- a/kernel/dma/swiotlb.c +++ b/kernel/dma/swiotlb.c @@ -575,7 +575,10 @@ static void swiotlb_bounce(struct device *dev, phys_addr_t tlb_addr, size_t size } } -#define slot_addr(start, idx) ((start) + ((idx) << IO_TLB_SHIFT)) +static inline phys_addr_t slot_addr(phys_addr_t start, phys_addr_t idx) +{ + return start + (idx << IO_TLB_SHIFT); +} /* * Carefully handle integer overflow which can occur when boundary_mask == ~0UL. From 43b919017fe755ccd2b19afb2dc2d3da8f840038 Mon Sep 17 00:00:00 2001 From: Chao Gao Date: Fri, 26 Aug 2022 17:50:46 +0800 Subject: [PATCH 109/158] swiotlb: fix a typo "overwirte" isn't a word. It should be "overwrite". Signed-off-by: Chao Gao Signed-off-by: Christoph Hellwig --- kernel/dma/swiotlb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c index 1ce8977d911c..0ef6b12f961d 100644 --- a/kernel/dma/swiotlb.c +++ b/kernel/dma/swiotlb.c @@ -764,7 +764,7 @@ phys_addr_t swiotlb_tbl_map_single(struct device *dev, phys_addr_t orig_addr, /* * When dir == DMA_FROM_DEVICE we could omit the copy from the orig * to the tlb buffer, if we knew for sure the device will - * overwirte the entire current content. But we don't. Thus + * overwrite the entire current content. But we don't. Thus * unconditional bounce may prevent leaking swiotlb content (i.e. * kernel memory) to user-space. */ From 9fc18f6d56d5b79d527c17a8100a0965d18345cf Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 21 Aug 2022 16:06:44 +0200 Subject: [PATCH 110/158] dma-mapping: mark dma_supported static Now that the remaining users in drivers are gone, this function can be marked static. Signed-off-by: Christoph Hellwig --- include/linux/dma-mapping.h | 5 ----- kernel/dma/mapping.c | 3 +-- 2 files changed, 1 insertion(+), 7 deletions(-) diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h index 25a30906289d..0ee20b764000 100644 --- a/include/linux/dma-mapping.h +++ b/include/linux/dma-mapping.h @@ -139,7 +139,6 @@ int dma_mmap_attrs(struct device *dev, struct vm_area_struct *vma, void *cpu_addr, dma_addr_t dma_addr, size_t size, unsigned long attrs); bool dma_can_mmap(struct device *dev); -int dma_supported(struct device *dev, u64 mask); bool dma_pci_p2pdma_supported(struct device *dev); int dma_set_mask(struct device *dev, u64 mask); int dma_set_coherent_mask(struct device *dev, u64 mask); @@ -248,10 +247,6 @@ static inline bool dma_can_mmap(struct device *dev) { return false; } -static inline int dma_supported(struct device *dev, u64 mask) -{ - return 0; -} static inline bool dma_pci_p2pdma_supported(struct device *dev) { return false; diff --git a/kernel/dma/mapping.c b/kernel/dma/mapping.c index 49cbf3e33de7..27f272381cf2 100644 --- a/kernel/dma/mapping.c +++ b/kernel/dma/mapping.c @@ -707,7 +707,7 @@ int dma_mmap_noncontiguous(struct device *dev, struct vm_area_struct *vma, } EXPORT_SYMBOL_GPL(dma_mmap_noncontiguous); -int dma_supported(struct device *dev, u64 mask) +static int dma_supported(struct device *dev, u64 mask) { const struct dma_map_ops *ops = get_dma_ops(dev); @@ -721,7 +721,6 @@ int dma_supported(struct device *dev, u64 mask) return 1; return ops->dma_supported(dev, mask); } -EXPORT_SYMBOL(dma_supported); bool dma_pci_p2pdma_supported(struct device *dev) { From 94a568ce32038d8ff9257004bb4632e60eb43a49 Mon Sep 17 00:00:00 2001 From: John Sperbeck Date: Mon, 1 Aug 2022 19:22:29 +0000 Subject: [PATCH 111/158] iommu/amd: use full 64-bit value in build_completion_wait() We started using a 64 bit completion value. Unfortunately, we only stored the low 32-bits, so a very large completion value would never be matched in iommu_completion_wait(). Fixes: c69d89aff393 ("iommu/amd: Use 4K page for completion wait write-back semaphore") Signed-off-by: John Sperbeck Link: https://lore.kernel.org/r/20220801192229.3358786-1-jsperbeck@google.com Signed-off-by: Joerg Roedel --- drivers/iommu/amd/iommu.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/iommu/amd/iommu.c b/drivers/iommu/amd/iommu.c index 65b8e4fd8217..828672a46a3d 100644 --- a/drivers/iommu/amd/iommu.c +++ b/drivers/iommu/amd/iommu.c @@ -939,7 +939,8 @@ static void build_completion_wait(struct iommu_cmd *cmd, memset(cmd, 0, sizeof(*cmd)); cmd->data[0] = lower_32_bits(paddr) | CMD_COMPL_WAIT_STORE_MASK; cmd->data[1] = upper_32_bits(paddr); - cmd->data[2] = data; + cmd->data[2] = lower_32_bits(data); + cmd->data[3] = upper_32_bits(data); CMD_SET_TYPE(cmd, CMD_COMPL_WAIT); } From 809f44a0cc5ad4b1209467a6287f8ac0eb49d393 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Wed, 7 Sep 2022 12:04:21 +0200 Subject: [PATCH 112/158] ALSA: usb-audio: Clear fixed clock rate at closing EP The recent commit c11117b634f4 ("ALSA: usb-audio: Refcount multiple accesses on the single clock") tries to manage the clock rate shared by several endpoints. This was intended for avoiding the unmatched rate by a different endpoint, but unfortunately, it introduced a regression for PulseAudio and pipewire, too; those applications try to probe the multiple possible rates (44.1k and 48kHz) and setting up the normal rate fails but only the last rate is applied. The cause is that the last sample rate is still left to the clock reference even after closing the endpoint, and this value is still used at the next open. It happens only when applications set up via PCM prepare but don't start/stop the stream; the rate is reset when the stream is stopped, but it's not cleared at close. This patch addresses the issue above, simply by clearing the rate set in the clock reference at the last close of each endpoint. Fixes: c11117b634f4 ("ALSA: usb-audio: Refcount multiple accesses on the single clock") Reported-by: Jason A. Donenfeld Tested-by: Jason A. Donenfeld Cc: Link: https://lore.kernel.org/all/YxXIWv8dYmg1tnXP@zx2c4.com/ Link: https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/2620 Link: https://lore.kernel.org/r/20220907100421.6443-1-tiwai@suse.de Signed-off-by: Takashi Iwai --- sound/usb/endpoint.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/sound/usb/endpoint.c b/sound/usb/endpoint.c index a42f2ce19455..8c8f9a851f89 100644 --- a/sound/usb/endpoint.c +++ b/sound/usb/endpoint.c @@ -925,6 +925,8 @@ void snd_usb_endpoint_close(struct snd_usb_audio *chip, endpoint_set_interface(chip, ep, false); if (!--ep->opened) { + if (ep->clock_ref && !atomic_read(&ep->clock_ref->locked)) + ep->clock_ref->rate = 0; ep->iface = 0; ep->altsetting = 0; ep->cur_audiofmt = NULL; From 12dd19c159659ec9050f45dc8a2ff3c3917f4be3 Mon Sep 17 00:00:00 2001 From: Alexander Gordeev Date: Sat, 13 Aug 2022 19:45:21 +0200 Subject: [PATCH 113/158] s390/boot: fix absolute zero lowcore corruption on boot Crash dump always starts on CPU0. In case CPU0 is offline the prefix page is not installed and the absolute zero lowcore is used. However, struct lowcore::mcesad is never assigned and stays zero. That leads to __machine_kdump() -> save_vx_regs() call silently stores vector registers to the absolute lowcore at 0x11b0 offset. Fixes: a62bc0739253 ("s390/kdump: add support for vector extension") Reviewed-by: Heiko Carstens Signed-off-by: Alexander Gordeev Signed-off-by: Vasily Gorbik --- arch/s390/kernel/nmi.c | 2 +- arch/s390/kernel/setup.c | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c index 60ac66aab163..31cb9b00a36b 100644 --- a/arch/s390/kernel/nmi.c +++ b/arch/s390/kernel/nmi.c @@ -64,7 +64,7 @@ static inline unsigned long nmi_get_mcesa_size(void) * structure. The structure is required for machine check happening * early in the boot process. */ -static struct mcesa boot_mcesa __initdata __aligned(MCESA_MAX_SIZE); +static struct mcesa boot_mcesa __aligned(MCESA_MAX_SIZE); void __init nmi_alloc_mcesa_early(u64 *mcesad) { diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index ed4fbbbdd1b0..74adb433c257 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -479,6 +479,7 @@ static void __init setup_lowcore_dat_off(void) put_abs_lowcore(restart_data, lc->restart_data); put_abs_lowcore(restart_source, lc->restart_source); put_abs_lowcore(restart_psw, lc->restart_psw); + put_abs_lowcore(mcesad, lc->mcesad); mcck_stack = (unsigned long)memblock_alloc(THREAD_SIZE, THREAD_SIZE); if (!mcck_stack) From 8d96bba75a43ba564bf8732e955d9f519d2bbaec Mon Sep 17 00:00:00 2001 From: Alexander Gordeev Date: Wed, 20 Jul 2022 07:24:03 +0200 Subject: [PATCH 114/158] s390/smp: enforce lowcore protection on CPU restart As result of commit 915fea04f932 ("s390/smp: enable DAT before CPU restart callback is called") the low-address protection bit gets mistakenly unset in control register 0 save area of the absolute zero memory. That area is used when manual PSW restart happened to hit an offline CPU. In this case the low-address protection for that CPU will be dropped. Reviewed-by: Heiko Carstens Fixes: 915fea04f932 ("s390/smp: enable DAT before CPU restart callback is called") Signed-off-by: Alexander Gordeev Signed-off-by: Vasily Gorbik --- arch/s390/kernel/setup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index 74adb433c257..bbd4bde4f65d 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -508,8 +508,8 @@ static void __init setup_lowcore_dat_on(void) S390_lowcore.svc_new_psw.mask |= PSW_MASK_DAT; S390_lowcore.program_new_psw.mask |= PSW_MASK_DAT; S390_lowcore.io_new_psw.mask |= PSW_MASK_DAT; - __ctl_store(S390_lowcore.cregs_save_area, 0, 15); __ctl_set_bit(0, 28); + __ctl_store(S390_lowcore.cregs_save_area, 0, 15); put_abs_lowcore(restart_flags, RESTART_FLAG_CTLREGS); put_abs_lowcore(program_new_psw, lc->program_new_psw); for (cr = 0; cr < ARRAY_SIZE(lc->cregs_save_area); cr++) From 388f788341178b91d2a18e7d605126dab6732257 Mon Sep 17 00:00:00 2001 From: Alexander Sverdlin Date: Tue, 6 Sep 2022 10:32:39 +0200 Subject: [PATCH 115/158] MIPS: octeon: Get rid of preprocessor directives around RESERVE32 Some of them were pointless because CONFIG_CAVIUM_RESERVE32 is now always defined, some were not enough (Yu Zhao reported "Failed to allocate CAVIUM_RESERVE32 memory area" error). Removing the directives allows for compiler coverage of RESERVE32 code and replacing one of [always-true] "ifdef" with a compiler conditional fixes the [cosmetic] error message. Fixes: 3e3114ac460e ("MIPS: Introduce CAVIUM_RESERVE32 Kconfig option") Signed-off-by: Alexander Sverdlin Signed-off-by: Thomas Bogendoerfer --- .../cavium-octeon/executive/cvmx-cmd-queue.c | 4 --- arch/mips/cavium-octeon/setup.c | 27 ++++++++----------- 2 files changed, 11 insertions(+), 20 deletions(-) diff --git a/arch/mips/cavium-octeon/executive/cvmx-cmd-queue.c b/arch/mips/cavium-octeon/executive/cvmx-cmd-queue.c index bf13e35871b2..aa7bbf8d0df5 100644 --- a/arch/mips/cavium-octeon/executive/cvmx-cmd-queue.c +++ b/arch/mips/cavium-octeon/executive/cvmx-cmd-queue.c @@ -57,14 +57,11 @@ EXPORT_SYMBOL_GPL(__cvmx_cmd_queue_state_ptr); static cvmx_cmd_queue_result_t __cvmx_cmd_queue_init_state_ptr(void) { char *alloc_name = "cvmx_cmd_queues"; -#if defined(CONFIG_CAVIUM_RESERVE32) && CONFIG_CAVIUM_RESERVE32 extern uint64_t octeon_reserve32_memory; -#endif if (likely(__cvmx_cmd_queue_state_ptr)) return CVMX_CMD_QUEUE_SUCCESS; -#if defined(CONFIG_CAVIUM_RESERVE32) && CONFIG_CAVIUM_RESERVE32 if (octeon_reserve32_memory) __cvmx_cmd_queue_state_ptr = cvmx_bootmem_alloc_named_range(sizeof(*__cvmx_cmd_queue_state_ptr), @@ -73,7 +70,6 @@ static cvmx_cmd_queue_result_t __cvmx_cmd_queue_init_state_ptr(void) (CONFIG_CAVIUM_RESERVE32 << 20) - 1, 128, alloc_name); else -#endif __cvmx_cmd_queue_state_ptr = cvmx_bootmem_alloc_named(sizeof(*__cvmx_cmd_queue_state_ptr), 128, diff --git a/arch/mips/cavium-octeon/setup.c b/arch/mips/cavium-octeon/setup.c index cbd83205518d..e7f994393ae8 100644 --- a/arch/mips/cavium-octeon/setup.c +++ b/arch/mips/cavium-octeon/setup.c @@ -284,10 +284,8 @@ void octeon_crash_smp_send_stop(void) #endif /* CONFIG_KEXEC */ -#ifdef CONFIG_CAVIUM_RESERVE32 uint64_t octeon_reserve32_memory; EXPORT_SYMBOL(octeon_reserve32_memory); -#endif #ifdef CONFIG_KEXEC /* crashkernel cmdline parameter is parsed _after_ memory setup @@ -666,9 +664,6 @@ void __init prom_init(void) int i; u64 t; int argc; -#ifdef CONFIG_CAVIUM_RESERVE32 - int64_t addr = -1; -#endif /* * The bootloader passes a pointer to the boot descriptor in * $a3, this is available as fw_arg3. @@ -783,7 +778,7 @@ void __init prom_init(void) cvmx_write_csr(CVMX_LED_UDD_DATX(1), 0); cvmx_write_csr(CVMX_LED_EN, 1); } -#ifdef CONFIG_CAVIUM_RESERVE32 + /* * We need to temporarily allocate all memory in the reserve32 * region. This makes sure the kernel doesn't allocate this @@ -794,14 +789,16 @@ void __init prom_init(void) * Allocate memory for RESERVED32 aligned on 2MB boundary. This * is in case we later use hugetlb entries with it. */ - addr = cvmx_bootmem_phy_named_block_alloc(CONFIG_CAVIUM_RESERVE32 << 20, - 0, 0, 2 << 20, - "CAVIUM_RESERVE32", 0); - if (addr < 0) - pr_err("Failed to allocate CAVIUM_RESERVE32 memory area\n"); - else - octeon_reserve32_memory = addr; -#endif + if (CONFIG_CAVIUM_RESERVE32) { + int64_t addr = + cvmx_bootmem_phy_named_block_alloc(CONFIG_CAVIUM_RESERVE32 << 20, + 0, 0, 2 << 20, + "CAVIUM_RESERVE32", 0); + if (addr < 0) + pr_err("Failed to allocate CAVIUM_RESERVE32 memory area\n"); + else + octeon_reserve32_memory = addr; + } #ifdef CONFIG_CAVIUM_OCTEON_LOCK_L2 if (cvmx_read_csr(CVMX_L2D_FUS3) & (3ull << 34)) { @@ -1079,7 +1076,6 @@ void __init plat_mem_setup(void) cvmx_bootmem_unlock(); #endif /* CONFIG_CRASH_DUMP */ -#ifdef CONFIG_CAVIUM_RESERVE32 /* * Now that we've allocated the kernel memory it is safe to * free the reserved region. We free it here so that builtin @@ -1087,7 +1083,6 @@ void __init plat_mem_setup(void) */ if (octeon_reserve32_memory) cvmx_bootmem_free_named("CAVIUM_RESERVE32"); -#endif /* CONFIG_CAVIUM_RESERVE32 */ if (total == 0) panic("Unable to allocate memory from " From ba912afbd611d3a5f22af247721a071ad1d5b9e0 Mon Sep 17 00:00:00 2001 From: Alexander Sverdlin Date: Tue, 6 Sep 2022 11:59:43 +0200 Subject: [PATCH 116/158] MIPS: OCTEON: irq: Fix octeon_irq_force_ciu_mapping() For irq_domain_associate() to work the virq descriptor has to be pre-allocated in advance. Otherwise the following happens: WARNING: CPU: 0 PID: 0 at .../kernel/irq/irqdomain.c:527 irq_domain_associate+0x298/0x2e8 error: virq128 is not allocated Modules linked in: CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.19.78-... #1 ... Call Trace: [] show_stack+0x9c/0x130 [] dump_stack+0x90/0xd0 [] __warn+0x118/0x130 [] warn_slowpath_fmt+0x4c/0x70 [] irq_domain_associate+0x298/0x2e8 [] octeon_irq_init_ciu+0x4c8/0x53c [] of_irq_init+0x1e0/0x388 [] init_IRQ+0x4c/0xf4 [] start_kernel+0x404/0x698 Use irq_alloc_desc_at() to avoid the above problem. Signed-off-by: Alexander Sverdlin Signed-off-by: Thomas Bogendoerfer --- arch/mips/cavium-octeon/octeon-irq.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/mips/cavium-octeon/octeon-irq.c b/arch/mips/cavium-octeon/octeon-irq.c index 9cb9ed44bcaf..fd8043f6ff8a 100644 --- a/arch/mips/cavium-octeon/octeon-irq.c +++ b/arch/mips/cavium-octeon/octeon-irq.c @@ -127,6 +127,16 @@ static void octeon_irq_free_cd(struct irq_domain *d, unsigned int irq) static int octeon_irq_force_ciu_mapping(struct irq_domain *domain, int irq, int line, int bit) { + struct device_node *of_node; + int ret; + + of_node = irq_domain_get_of_node(domain); + if (!of_node) + return -EINVAL; + ret = irq_alloc_desc_at(irq, of_node_to_nid(of_node)); + if (ret < 0) + return ret; + return irq_domain_associate(domain, irq, line << 6 | bit); } From 0c5f6c0d8201a809a6585b07b6263e9db2c874a3 Mon Sep 17 00:00:00 2001 From: Lu Baolu Date: Tue, 23 Aug 2022 14:15:54 +0800 Subject: [PATCH 117/158] iommu/vt-d: Fix kdump kernels boot failure with scalable mode The translation table copying code for kdump kernels is currently based on the extended root/context entry formats of ECS mode defined in older VT-d v2.5, and doesn't handle the scalable mode formats. This causes the kexec capture kernel boot failure with DMAR faults if the IOMMU was enabled in scalable mode by the previous kernel. The ECS mode has already been deprecated by the VT-d spec since v3.0 and Intel IOMMU driver doesn't support this mode as there's no real hardware implementation. Hence this converts ECS checking in copying table code into scalable mode. The existing copying code consumes a bit in the context entry as a mark of copied entry. It needs to work for the old format as well as for the extended context entries. As it's hard to find such a common bit for both legacy and scalable mode context entries. This replaces it with a per- IOMMU bitmap. Fixes: 7373a8cc38197 ("iommu/vt-d: Setup context and enable RID2PASID support") Cc: stable@vger.kernel.org Reported-by: Jerry Snitselaar Tested-by: Wen Jin Signed-off-by: Lu Baolu Link: https://lore.kernel.org/r/20220817011035.3250131-1-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel --- drivers/iommu/intel/iommu.c | 100 ++++++++++++++++-------------------- drivers/iommu/intel/iommu.h | 9 ++-- 2 files changed, 50 insertions(+), 59 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 7cca030a508e..b9d058c27568 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -163,38 +163,6 @@ static phys_addr_t root_entry_uctp(struct root_entry *re) return re->hi & VTD_PAGE_MASK; } -static inline void context_clear_pasid_enable(struct context_entry *context) -{ - context->lo &= ~(1ULL << 11); -} - -static inline bool context_pasid_enabled(struct context_entry *context) -{ - return !!(context->lo & (1ULL << 11)); -} - -static inline void context_set_copied(struct context_entry *context) -{ - context->hi |= (1ull << 3); -} - -static inline bool context_copied(struct context_entry *context) -{ - return !!(context->hi & (1ULL << 3)); -} - -static inline bool __context_present(struct context_entry *context) -{ - return (context->lo & 1); -} - -bool context_present(struct context_entry *context) -{ - return context_pasid_enabled(context) ? - __context_present(context) : - __context_present(context) && !context_copied(context); -} - static inline void context_set_present(struct context_entry *context) { context->lo |= 1; @@ -242,6 +210,26 @@ static inline void context_clear_entry(struct context_entry *context) context->hi = 0; } +static inline bool context_copied(struct intel_iommu *iommu, u8 bus, u8 devfn) +{ + if (!iommu->copied_tables) + return false; + + return test_bit(((long)bus << 8) | devfn, iommu->copied_tables); +} + +static inline void +set_context_copied(struct intel_iommu *iommu, u8 bus, u8 devfn) +{ + set_bit(((long)bus << 8) | devfn, iommu->copied_tables); +} + +static inline void +clear_context_copied(struct intel_iommu *iommu, u8 bus, u8 devfn) +{ + clear_bit(((long)bus << 8) | devfn, iommu->copied_tables); +} + /* * This domain is a statically identity mapping domain. * 1. This domain creats a static 1:1 mapping to all usable memory. @@ -578,6 +566,13 @@ struct context_entry *iommu_context_addr(struct intel_iommu *iommu, u8 bus, struct context_entry *context; u64 *entry; + /* + * Except that the caller requested to allocate a new entry, + * returning a copied context entry makes no sense. + */ + if (!alloc && context_copied(iommu, bus, devfn)) + return NULL; + entry = &root->lo; if (sm_supported(iommu)) { if (devfn >= 0x80) { @@ -1688,6 +1683,11 @@ static void free_dmar_iommu(struct intel_iommu *iommu) iommu->domain_ids = NULL; } + if (iommu->copied_tables) { + bitmap_free(iommu->copied_tables); + iommu->copied_tables = NULL; + } + /* free context mapping */ free_context_table(iommu); @@ -1913,7 +1913,7 @@ static int domain_context_mapping_one(struct dmar_domain *domain, goto out_unlock; ret = 0; - if (context_present(context)) + if (context_present(context) && !context_copied(iommu, bus, devfn)) goto out_unlock; /* @@ -1925,7 +1925,7 @@ static int domain_context_mapping_one(struct dmar_domain *domain, * in-flight DMA will exist, and we don't need to worry anymore * hereafter. */ - if (context_copied(context)) { + if (context_copied(iommu, bus, devfn)) { u16 did_old = context_domain_id(context); if (did_old < cap_ndoms(iommu->cap)) { @@ -1936,6 +1936,8 @@ static int domain_context_mapping_one(struct dmar_domain *domain, iommu->flush.flush_iotlb(iommu, did_old, 0, 0, DMA_TLB_DSI_FLUSH); } + + clear_context_copied(iommu, bus, devfn); } context_clear_entry(context); @@ -2684,32 +2686,14 @@ static int copy_context_table(struct intel_iommu *iommu, /* Now copy the context entry */ memcpy(&ce, old_ce + idx, sizeof(ce)); - if (!__context_present(&ce)) + if (!context_present(&ce)) continue; did = context_domain_id(&ce); if (did >= 0 && did < cap_ndoms(iommu->cap)) set_bit(did, iommu->domain_ids); - /* - * We need a marker for copied context entries. This - * marker needs to work for the old format as well as - * for extended context entries. - * - * Bit 67 of the context entry is used. In the old - * format this bit is available to software, in the - * extended format it is the PGE bit, but PGE is ignored - * by HW if PASIDs are disabled (and thus still - * available). - * - * So disable PASIDs first and then mark the entry - * copied. This means that we don't copy PASID - * translations from the old kernel, but this is fine as - * faults there are not fatal. - */ - context_clear_pasid_enable(&ce); - context_set_copied(&ce); - + set_context_copied(iommu, bus, devfn); new_ce[idx] = ce; } @@ -2735,8 +2719,8 @@ static int copy_translation_tables(struct intel_iommu *iommu) bool new_ext, ext; rtaddr_reg = dmar_readq(iommu->reg + DMAR_RTADDR_REG); - ext = !!(rtaddr_reg & DMA_RTADDR_RTT); - new_ext = !!ecap_ecs(iommu->ecap); + ext = !!(rtaddr_reg & DMA_RTADDR_SMT); + new_ext = !!sm_supported(iommu); /* * The RTT bit can only be changed when translation is disabled, @@ -2747,6 +2731,10 @@ static int copy_translation_tables(struct intel_iommu *iommu) if (new_ext != ext) return -EINVAL; + iommu->copied_tables = bitmap_zalloc(BIT_ULL(16), GFP_KERNEL); + if (!iommu->copied_tables) + return -ENOMEM; + old_rt_phys = rtaddr_reg & VTD_PAGE_MASK; if (!old_rt_phys) return -EINVAL; diff --git a/drivers/iommu/intel/iommu.h b/drivers/iommu/intel/iommu.h index fae45bbb0c7f..74b0e19e23ee 100644 --- a/drivers/iommu/intel/iommu.h +++ b/drivers/iommu/intel/iommu.h @@ -197,7 +197,6 @@ #define ecap_dis(e) (((e) >> 27) & 0x1) #define ecap_nest(e) (((e) >> 26) & 0x1) #define ecap_mts(e) (((e) >> 25) & 0x1) -#define ecap_ecs(e) (((e) >> 24) & 0x1) #define ecap_iotlb_offset(e) ((((e) >> 8) & 0x3ff) * 16) #define ecap_max_iotlb_offset(e) (ecap_iotlb_offset(e) + 16) #define ecap_coherent(e) ((e) & 0x1) @@ -265,7 +264,6 @@ #define DMA_GSTS_CFIS (((u32)1) << 23) /* DMA_RTADDR_REG */ -#define DMA_RTADDR_RTT (((u64)1) << 11) #define DMA_RTADDR_SMT (((u64)1) << 10) /* CCMD_REG */ @@ -579,6 +577,7 @@ struct intel_iommu { #ifdef CONFIG_INTEL_IOMMU unsigned long *domain_ids; /* bitmap of domains */ + unsigned long *copied_tables; /* bitmap of copied tables */ spinlock_t lock; /* protect context, domain ids */ struct root_entry *root_entry; /* virtual address */ @@ -701,6 +700,11 @@ static inline int nr_pte_to_next_page(struct dma_pte *pte) (struct dma_pte *)ALIGN((unsigned long)pte, VTD_PAGE_SIZE) - pte; } +static inline bool context_present(struct context_entry *context) +{ + return (context->lo & 1); +} + extern struct dmar_drhd_unit * dmar_find_matched_drhd_unit(struct pci_dev *dev); extern int dmar_enable_qi(struct intel_iommu *iommu); @@ -784,7 +788,6 @@ static inline void intel_iommu_debugfs_init(void) {} #endif /* CONFIG_INTEL_IOMMU_DEBUGFS */ extern const struct attribute_group *intel_iommu_groups[]; -bool context_present(struct context_entry *context); struct context_entry *iommu_context_addr(struct intel_iommu *iommu, u8 bus, u8 devfn, int alloc); From 53fc7ad6edf210b497230ce74b61b322a202470c Mon Sep 17 00:00:00 2001 From: Lu Baolu Date: Tue, 23 Aug 2022 14:15:55 +0800 Subject: [PATCH 118/158] iommu/vt-d: Correctly calculate sagaw value of IOMMU The Intel IOMMU driver possibly selects between the first-level and the second-level translation tables for DMA address translation. However, the levels of page-table walks for the 4KB base page size are calculated from the SAGAW field of the capability register, which is only valid for the second-level page table. This causes the IOMMU driver to stop working if the hardware (or the emulated IOMMU) advertises only first-level translation capability and reports the SAGAW field as 0. This solves the above problem by considering both the first level and the second level when calculating the supported page table levels. Fixes: b802d070a52a1 ("iommu/vt-d: Use iova over first level") Cc: stable@vger.kernel.org Signed-off-by: Lu Baolu Link: https://lore.kernel.org/r/20220817023558.3253263-1-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel --- drivers/iommu/intel/iommu.c | 28 +++++++++++++++++++++++++--- 1 file changed, 25 insertions(+), 3 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index b9d058c27568..b155c7af7d15 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -390,14 +390,36 @@ static inline int domain_pfn_supported(struct dmar_domain *domain, return !(addr_width < BITS_PER_LONG && pfn >> addr_width); } +/* + * Calculate the Supported Adjusted Guest Address Widths of an IOMMU. + * Refer to 11.4.2 of the VT-d spec for the encoding of each bit of + * the returned SAGAW. + */ +static unsigned long __iommu_calculate_sagaw(struct intel_iommu *iommu) +{ + unsigned long fl_sagaw, sl_sagaw; + + fl_sagaw = BIT(2) | (cap_fl1gp_support(iommu->cap) ? BIT(3) : 0); + sl_sagaw = cap_sagaw(iommu->cap); + + /* Second level only. */ + if (!sm_supported(iommu) || !ecap_flts(iommu->ecap)) + return sl_sagaw; + + /* First level only. */ + if (!ecap_slts(iommu->ecap)) + return fl_sagaw; + + return fl_sagaw & sl_sagaw; +} + static int __iommu_calculate_agaw(struct intel_iommu *iommu, int max_gaw) { unsigned long sagaw; int agaw; - sagaw = cap_sagaw(iommu->cap); - for (agaw = width_to_agaw(max_gaw); - agaw >= 0; agaw--) { + sagaw = __iommu_calculate_sagaw(iommu); + for (agaw = width_to_agaw(max_gaw); agaw >= 0; agaw--) { if (test_bit(agaw, &sagaw)) break; } From a349ffcb4daf77a4040ab5a6c95295bb533bf384 Mon Sep 17 00:00:00 2001 From: Lu Baolu Date: Tue, 23 Aug 2022 14:15:56 +0800 Subject: [PATCH 119/158] iommu/vt-d: Fix recursive lock issue in iommu_flush_dev_iotlb() The per domain spinlock is acquired in iommu_flush_dev_iotlb(), which is possbile to be called in the interrupt context. For example, the drm-intel's CI system got completely blocked with below error: WARNING: inconsistent lock state 6.0.0-rc1-CI_DRM_11990-g6590d43d39b9+ #1 Not tainted -------------------------------- inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage. swapper/6/0 [HC0[0]:SC1[1]:HE1:SE0] takes: ffff88810440d678 (&domain->lock){+.?.}-{2:2}, at: iommu_flush_dev_iotlb.part.61+0x23/0x80 {SOFTIRQ-ON-W} state was registered at: lock_acquire+0xd3/0x310 _raw_spin_lock+0x2a/0x40 domain_update_iommu_cap+0x20b/0x2c0 intel_iommu_attach_device+0x5bd/0x860 __iommu_attach_device+0x18/0xe0 bus_iommu_probe+0x1f3/0x2d0 bus_set_iommu+0x82/0xd0 intel_iommu_init+0xe45/0x102a pci_iommu_init+0x9/0x31 do_one_initcall+0x53/0x2f0 kernel_init_freeable+0x18f/0x1e1 kernel_init+0x11/0x120 ret_from_fork+0x1f/0x30 irq event stamp: 162354 hardirqs last enabled at (162354): [] _raw_spin_unlock_irqrestore+0x54/0x70 hardirqs last disabled at (162353): [] _raw_spin_lock_irqsave+0x4b/0x50 softirqs last enabled at (162338): [] __do_softirq+0x323/0x48e softirqs last disabled at (162349): [] irq_exit_rcu+0xb8/0xe0 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&domain->lock); lock(&domain->lock); *** DEADLOCK *** 1 lock held by swapper/6/0: This coverts the spin_lock/unlock() into the irq save/restore varieties to fix the recursive locking issues. Fixes: ffd5869d93530 ("iommu/vt-d: Replace spin_lock_irqsave() with spin_lock()") Signed-off-by: Lu Baolu Acked-by: Lucas De Marchi Link: https://lore.kernel.org/r/20220817025650.3253959-1-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel --- drivers/iommu/intel/iommu.c | 39 ++++++++++++++++++++++--------------- 1 file changed, 23 insertions(+), 16 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index b155c7af7d15..e3fe1a148187 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -515,8 +515,9 @@ static int domain_update_device_node(struct dmar_domain *domain) { struct device_domain_info *info; int nid = NUMA_NO_NODE; + unsigned long flags; - spin_lock(&domain->lock); + spin_lock_irqsave(&domain->lock, flags); list_for_each_entry(info, &domain->devices, link) { /* * There could possibly be multiple device numa nodes as devices @@ -528,7 +529,7 @@ static int domain_update_device_node(struct dmar_domain *domain) if (nid != NUMA_NO_NODE) break; } - spin_unlock(&domain->lock); + spin_unlock_irqrestore(&domain->lock, flags); return nid; } @@ -1362,19 +1363,20 @@ iommu_support_dev_iotlb(struct dmar_domain *domain, struct intel_iommu *iommu, u8 bus, u8 devfn) { struct device_domain_info *info; + unsigned long flags; if (!iommu->qi) return NULL; - spin_lock(&domain->lock); + spin_lock_irqsave(&domain->lock, flags); list_for_each_entry(info, &domain->devices, link) { if (info->iommu == iommu && info->bus == bus && info->devfn == devfn) { - spin_unlock(&domain->lock); + spin_unlock_irqrestore(&domain->lock, flags); return info->ats_supported ? info : NULL; } } - spin_unlock(&domain->lock); + spin_unlock_irqrestore(&domain->lock, flags); return NULL; } @@ -1383,8 +1385,9 @@ static void domain_update_iotlb(struct dmar_domain *domain) { struct device_domain_info *info; bool has_iotlb_device = false; + unsigned long flags; - spin_lock(&domain->lock); + spin_lock_irqsave(&domain->lock, flags); list_for_each_entry(info, &domain->devices, link) { if (info->ats_enabled) { has_iotlb_device = true; @@ -1392,7 +1395,7 @@ static void domain_update_iotlb(struct dmar_domain *domain) } } domain->has_iotlb_device = has_iotlb_device; - spin_unlock(&domain->lock); + spin_unlock_irqrestore(&domain->lock, flags); } static void iommu_enable_dev_iotlb(struct device_domain_info *info) @@ -1484,14 +1487,15 @@ static void iommu_flush_dev_iotlb(struct dmar_domain *domain, u64 addr, unsigned mask) { struct device_domain_info *info; + unsigned long flags; if (!domain->has_iotlb_device) return; - spin_lock(&domain->lock); + spin_lock_irqsave(&domain->lock, flags); list_for_each_entry(info, &domain->devices, link) __iommu_flush_dev_iotlb(info, addr, mask); - spin_unlock(&domain->lock); + spin_unlock_irqrestore(&domain->lock, flags); } static void iommu_flush_iotlb_psi(struct intel_iommu *iommu, @@ -2453,6 +2457,7 @@ static int domain_add_dev_info(struct dmar_domain *domain, struct device *dev) { struct device_domain_info *info = dev_iommu_priv_get(dev); struct intel_iommu *iommu; + unsigned long flags; u8 bus, devfn; int ret; @@ -2464,9 +2469,9 @@ static int domain_add_dev_info(struct dmar_domain *domain, struct device *dev) if (ret) return ret; info->domain = domain; - spin_lock(&domain->lock); + spin_lock_irqsave(&domain->lock, flags); list_add(&info->link, &domain->devices); - spin_unlock(&domain->lock); + spin_unlock_irqrestore(&domain->lock, flags); /* PASID table is mandatory for a PCI device in scalable mode. */ if (sm_supported(iommu) && !dev_is_real_dma_subdevice(dev)) { @@ -4090,6 +4095,7 @@ static void dmar_remove_one_dev_info(struct device *dev) struct device_domain_info *info = dev_iommu_priv_get(dev); struct dmar_domain *domain = info->domain; struct intel_iommu *iommu = info->iommu; + unsigned long flags; if (!dev_is_real_dma_subdevice(info->dev)) { if (dev_is_pci(info->dev) && sm_supported(iommu)) @@ -4101,9 +4107,9 @@ static void dmar_remove_one_dev_info(struct device *dev) intel_pasid_free_table(info->dev); } - spin_lock(&domain->lock); + spin_lock_irqsave(&domain->lock, flags); list_del(&info->link); - spin_unlock(&domain->lock); + spin_unlock_irqrestore(&domain->lock, flags); domain_detach_iommu(domain, iommu); info->domain = NULL; @@ -4422,19 +4428,20 @@ static void domain_set_force_snooping(struct dmar_domain *domain) static bool intel_iommu_enforce_cache_coherency(struct iommu_domain *domain) { struct dmar_domain *dmar_domain = to_dmar_domain(domain); + unsigned long flags; if (dmar_domain->force_snooping) return true; - spin_lock(&dmar_domain->lock); + spin_lock_irqsave(&dmar_domain->lock, flags); if (!domain_support_force_snooping(dmar_domain)) { - spin_unlock(&dmar_domain->lock); + spin_unlock_irqrestore(&dmar_domain->lock, flags); return false; } domain_set_force_snooping(dmar_domain); dmar_domain->force_snooping = true; - spin_unlock(&dmar_domain->lock); + spin_unlock_irqrestore(&dmar_domain->lock, flags); return true; } From 35bf49e054214a0c7d3b0a0f8606f753f3f6ae96 Mon Sep 17 00:00:00 2001 From: Lu Baolu Date: Tue, 23 Aug 2022 14:15:57 +0800 Subject: [PATCH 120/158] iommu/vt-d: Fix lockdep splat due to klist iteration in atomic context With CONFIG_INTEL_IOMMU_DEBUGFS enabled, below lockdep splat are seen when an I/O fault occurs on a machine with an Intel IOMMU in it. DMAR: DRHD: handling fault status reg 3 DMAR: [DMA Write NO_PASID] Request device [00:1a.0] fault addr 0x0 [fault reason 0x05] PTE Write access is not set DMAR: Dump dmar0 table entries for IOVA 0x0 DMAR: root entry: 0x0000000127f42001 DMAR: context entry: hi 0x0000000000001502, low 0x000000012d8ab001 ================================ WARNING: inconsistent lock state 5.20.0-0.rc0.20220812git7ebfc85e2cd7.10.fc38.x86_64 #1 Not tainted -------------------------------- inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage. rngd/1006 [HC1[1]:SC0[0]:HE0:SE1] takes: ff177021416f2d78 (&k->k_lock){?.+.}-{2:2}, at: klist_next+0x1b/0x160 {HARDIRQ-ON-W} state was registered at: lock_acquire+0xce/0x2d0 _raw_spin_lock+0x33/0x80 klist_add_tail+0x46/0x80 bus_add_device+0xee/0x150 device_add+0x39d/0x9a0 add_memory_block+0x108/0x1d0 memory_dev_init+0xe1/0x117 driver_init+0x43/0x4d kernel_init_freeable+0x1c2/0x2cc kernel_init+0x16/0x140 ret_from_fork+0x1f/0x30 irq event stamp: 7812 hardirqs last enabled at (7811): [] asm_sysvec_apic_timer_interrupt+0x16/0x20 hardirqs last disabled at (7812): [] irqentry_enter+0x54/0x60 softirqs last enabled at (7794): [] __irq_exit_rcu+0xf9/0x170 softirqs last disabled at (7787): [] __irq_exit_rcu+0xf9/0x170 The klist iterator functions using spin_*lock_irq*() but the klist insertion functions using spin_*lock(), combined with the Intel DMAR IOMMU driver iterating over klists from atomic (hardirq) context, where pci_get_domain_bus_and_slot() calls into bus_find_device() which iterates over klists. As currently there's no plan to fix the klist to make it safe to use in atomic context, this fixes the lockdep splat by avoid calling pci_get_domain_bus_and_slot() in the hardirq context. Fixes: 8ac0b64b9735 ("iommu/vt-d: Use pci_get_domain_bus_and_slot() in pgtable_walk()") Reported-by: Lennert Buytenhek Link: https://lore.kernel.org/linux-iommu/Yvo2dfpEh%2FWC+Wrr@wantstofly.org/ Link: https://lore.kernel.org/linux-iommu/YvyBdPwrTuHHbn5X@wantstofly.org/ Signed-off-by: Lu Baolu Link: https://lore.kernel.org/r/20220819015949.4795-1-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel --- drivers/iommu/intel/iommu.c | 47 +++++++++++++++---------------------- 1 file changed, 19 insertions(+), 28 deletions(-) diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index e3fe1a148187..64d30895a4c8 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -813,32 +813,11 @@ static void free_context_table(struct intel_iommu *iommu) } #ifdef CONFIG_DMAR_DEBUG -static void pgtable_walk(struct intel_iommu *iommu, unsigned long pfn, u8 bus, u8 devfn) +static void pgtable_walk(struct intel_iommu *iommu, unsigned long pfn, + u8 bus, u8 devfn, struct dma_pte *parent, int level) { - struct device_domain_info *info; - struct dma_pte *parent, *pte; - struct dmar_domain *domain; - struct pci_dev *pdev; - int offset, level; - - pdev = pci_get_domain_bus_and_slot(iommu->segment, bus, devfn); - if (!pdev) - return; - - info = dev_iommu_priv_get(&pdev->dev); - if (!info || !info->domain) { - pr_info("device [%02x:%02x.%d] not probed\n", - bus, PCI_SLOT(devfn), PCI_FUNC(devfn)); - return; - } - - domain = info->domain; - level = agaw_to_level(domain->agaw); - parent = domain->pgd; - if (!parent) { - pr_info("no page table setup\n"); - return; - } + struct dma_pte *pte; + int offset; while (1) { offset = pfn_level_offset(pfn, level); @@ -865,9 +844,10 @@ void dmar_fault_dump_ptes(struct intel_iommu *iommu, u16 source_id, struct pasid_entry *entries, *pte; struct context_entry *ctx_entry; struct root_entry *rt_entry; + int i, dir_index, index, level; u8 devfn = source_id & 0xff; u8 bus = source_id >> 8; - int i, dir_index, index; + struct dma_pte *pgtable; pr_info("Dump %s table entries for IOVA 0x%llx\n", iommu->name, addr); @@ -895,8 +875,11 @@ void dmar_fault_dump_ptes(struct intel_iommu *iommu, u16 source_id, ctx_entry->hi, ctx_entry->lo); /* legacy mode does not require PASID entries */ - if (!sm_supported(iommu)) + if (!sm_supported(iommu)) { + level = agaw_to_level(ctx_entry->hi & 7); + pgtable = phys_to_virt(ctx_entry->lo & VTD_PAGE_MASK); goto pgtable_walk; + } /* get the pointer to pasid directory entry */ dir = phys_to_virt(ctx_entry->lo & VTD_PAGE_MASK); @@ -923,8 +906,16 @@ void dmar_fault_dump_ptes(struct intel_iommu *iommu, u16 source_id, for (i = 0; i < ARRAY_SIZE(pte->val); i++) pr_info("pasid table entry[%d]: 0x%016llx\n", i, pte->val[i]); + if (pasid_pte_get_pgtt(pte) == PASID_ENTRY_PGTT_FL_ONLY) { + level = pte->val[2] & BIT_ULL(2) ? 5 : 4; + pgtable = phys_to_virt(pte->val[2] & VTD_PAGE_MASK); + } else { + level = agaw_to_level((pte->val[0] >> 2) & 0x7); + pgtable = phys_to_virt(pte->val[0] & VTD_PAGE_MASK); + } + pgtable_walk: - pgtable_walk(iommu, addr >> VTD_PAGE_SHIFT, bus, devfn); + pgtable_walk(iommu, addr >> VTD_PAGE_SHIFT, bus, devfn, pgtable, level); } #endif From 91c98fe7941499e4127cdc359c30841b873dd43a Mon Sep 17 00:00:00 2001 From: Jean-Philippe Brucker Date: Thu, 25 Aug 2022 16:46:24 +0100 Subject: [PATCH 121/158] iommu/virtio: Fix interaction with VFIO Commit e8ae0e140c05 ("vfio: Require that devices support DMA cache coherence") requires IOMMU drivers to advertise IOMMU_CAP_CACHE_COHERENCY, in order to be used by VFIO. Since VFIO does not provide to userspace the ability to maintain coherency through cache invalidations, it requires hardware coherency. Advertise the capability in order to restore VFIO support. The meaning of IOMMU_CAP_CACHE_COHERENCY also changed from "IOMMU can enforce cache coherent DMA transactions" to "IOMMU_CACHE is supported". While virtio-iommu cannot enforce coherency (of PCIe no-snoop transactions), it does support IOMMU_CACHE. We can distinguish different cases of non-coherent DMA: (1) When accesses from a hardware endpoint are not coherent. The host would describe such a device using firmware methods ('dma-coherent' in device-tree, '_CCA' in ACPI), since they are also needed without a vIOMMU. In this case mappings are created without IOMMU_CACHE. virtio-iommu doesn't need any additional support. It sends the same requests as for coherent devices. (2) When the physical IOMMU supports non-cacheable mappings. Supporting those would require a new feature in virtio-iommu, new PROBE request property and MAP flags. Device drivers would use a new API to discover this since it depends on the architecture and the physical IOMMU. (3) When the hardware supports PCIe no-snoop. It is possible for assigned PCIe devices to issue no-snoop transactions, and the virtio-iommu specification is lacking any mention of this. Arm platforms don't necessarily support no-snoop, and those that do cannot enforce coherency of no-snoop transactions. Device drivers must be careful about assuming that no-snoop transactions won't end up cached; see commit e02f5c1bb228 ("drm: disable uncached DMA optimization for ARM and arm64"). On x86 platforms, the host may or may not enforce coherency of no-snoop transactions with the physical IOMMU. But according to the above commit, on x86 a driver which assumes that no-snoop DMA is compatible with uncached CPU mappings will also work if the host enforces coherency. Although these issues are not specific to virtio-iommu, it could be used to facilitate discovery and configuration of no-snoop. This would require a new feature bit, PROBE property and ATTACH/MAP flags. Cc: stable@vger.kernel.org Fixes: e8ae0e140c05 ("vfio: Require that devices support DMA cache coherence") Signed-off-by: Jean-Philippe Brucker Reviewed-by: Robin Murphy Link: https://lore.kernel.org/r/20220825154622.86759-1-jean-philippe@linaro.org Signed-off-by: Joerg Roedel --- drivers/iommu/virtio-iommu.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/iommu/virtio-iommu.c b/drivers/iommu/virtio-iommu.c index 08eeafc9529f..80151176ba12 100644 --- a/drivers/iommu/virtio-iommu.c +++ b/drivers/iommu/virtio-iommu.c @@ -1006,7 +1006,18 @@ static int viommu_of_xlate(struct device *dev, struct of_phandle_args *args) return iommu_fwspec_add_ids(dev, args->args, 1); } +static bool viommu_capable(enum iommu_cap cap) +{ + switch (cap) { + case IOMMU_CAP_CACHE_COHERENCY: + return true; + default: + return false; + } +} + static struct iommu_ops viommu_ops = { + .capable = viommu_capable, .domain_alloc = viommu_domain_alloc, .probe_device = viommu_probe_device, .probe_finalize = viommu_probe_finalize, From 5ac251c8a05ce074e5efac779debf82a15d870a3 Mon Sep 17 00:00:00 2001 From: Yicong Yang Date: Mon, 5 Sep 2022 20:26:15 +0800 Subject: [PATCH 122/158] arch_topology: Make cluster topology span at least SMT CPUs Currently cpu_clustergroup_mask() will return CPU mask if cluster span more or the same CPUs as cpu_coregroup_mask(). This will result topology borken on non-Cluster SMT machines when building with CONFIG_SCHED_CLUSTER=y. Test with: qemu-system-aarch64 -enable-kvm -machine virt \ -net none \ -cpu host \ -bios ./QEMU_EFI.fd \ -m 2G \ -smp 48,sockets=2,cores=12,threads=2 \ -kernel $Image \ -initrd $Rootfs \ -nographic -append "rdinit=init console=ttyAMA0 sched_verbose loglevel=8" We'll get below error: [ 3.084568] BUG: arch topology borken [ 3.084570] the SMT domain not a subset of the CLS domain Since cluster is a level higher than SMT, fix this by making cluster spans at least SMT CPUs. Fixes: bfcc4397435d ("arch_topology: Limit span of cpu_clustergroup_mask()") Cc: Sudeep Holla Cc: Vincent Guittot Cc: Ionela Voinescu Cc: Greg KH Reviewed-by: Sudeep Holla Signed-off-by: Yicong Yang Link: https://lore.kernel.org/r/20220905122615.12946-1-yangyicong@huawei.com Signed-off-by: Greg Kroah-Hartman --- drivers/base/arch_topology.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/base/arch_topology.c b/drivers/base/arch_topology.c index eaa1b8d2d39d..46cbe4471e78 100644 --- a/drivers/base/arch_topology.c +++ b/drivers/base/arch_topology.c @@ -724,7 +724,7 @@ const struct cpumask *cpu_clustergroup_mask(int cpu) */ if (cpumask_subset(cpu_coregroup_mask(cpu), &cpu_topology[cpu].cluster_sibling)) - return get_cpu_mask(cpu); + return topology_sibling_cpumask(cpu); return &cpu_topology[cpu].cluster_sibling; } From df6d3422d3eed27afa23df092b3ce147c558d1a8 Mon Sep 17 00:00:00 2001 From: Pavel Begunkov Date: Tue, 6 Sep 2022 17:11:16 +0100 Subject: [PATCH 123/158] io_uring/kbuf: fix not advancing READV kbuf ring When we don't recycle a selected ring buffer we should advance the head of the ring, so don't just skip io_kbuf_recycle() for IORING_OP_READV but adjust the ring. Fixes: 934447a603b22 ("io_uring: do not recycle buffer in READV") Signed-off-by: Pavel Begunkov Reviewed-by: Dylan Yudaken Link: https://lore.kernel.org/r/a6d85e2611471bcb5d5dcd63a8342077ddc2d73d.1662480490.git.asml.silence@gmail.com Signed-off-by: Jens Axboe --- io_uring/kbuf.h | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/io_uring/kbuf.h b/io_uring/kbuf.h index d6af208d109f..746fbf31a703 100644 --- a/io_uring/kbuf.h +++ b/io_uring/kbuf.h @@ -91,9 +91,13 @@ static inline void io_kbuf_recycle(struct io_kiocb *req, unsigned issue_flags) * buffer data. However if that buffer is recycled the original request * data stored in addr is lost. Therefore forbid recycling for now. */ - if (req->opcode == IORING_OP_READV) + if (req->opcode == IORING_OP_READV) { + if ((req->flags & REQ_F_BUFFER_RING) && req->buf_list) { + req->buf_list->head++; + req->buf_list = NULL; + } return; - + } if (req->flags & REQ_F_BUFFER_SELECTED) io_kbuf_recycle_legacy(req, issue_flags); if (req->flags & REQ_F_BUFFER_RING) From 336d28a8f38013a069f2d46e73aaa1880ef17a47 Mon Sep 17 00:00:00 2001 From: Pavel Begunkov Date: Tue, 6 Sep 2022 17:11:17 +0100 Subject: [PATCH 124/158] io_uring: recycle kbuf recycle on tw requeue When we queue a request via tw for execution it's not going to be executed immediately, so when io_queue_async() hits IO_APOLL_READY and queues a tw but doesn't try to recycle/consume the buffer some other request may try to use the the buffer. Fixes: c7fb19428d67 ("io_uring: add support for ring mapped supplied buffers") Signed-off-by: Pavel Begunkov Link: https://lore.kernel.org/r/a19bc9e211e3184215a58e129b62f440180e9212.1662480490.git.asml.silence@gmail.com Signed-off-by: Jens Axboe --- io_uring/io_uring.c | 1 + 1 file changed, 1 insertion(+) diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c index f9be9b7eb654..b9640ad5069f 100644 --- a/io_uring/io_uring.c +++ b/io_uring/io_uring.c @@ -1728,6 +1728,7 @@ static void io_queue_async(struct io_kiocb *req, int ret) switch (io_arm_poll_handler(req, 0)) { case IO_APOLL_READY: + io_kbuf_recycle(req, 0); io_req_task_queue(req); break; case IO_APOLL_ABORTED: From 2a2dfc869d3345ccdd91322b023f4b0da84acbe7 Mon Sep 17 00:00:00 2001 From: Joe Fradley Date: Tue, 23 Aug 2022 21:19:33 -0700 Subject: [PATCH 125/158] tools: Add new "test" taint to kernel-chktaint Commit c272612cb4a2 ("kunit: Taint the kernel when KUnit tests are run") added a new taint flag for when in-kernel tests run. This commit adds recognition of this new flag in kernel-chktaint. With this change the correct reason will be reported if the kernel is tainted because of a test run. Amended Commit log: Shuah Khan Reviewed-by: David Gow Signed-off-by: Joe Fradley Reviewed-by: Brendan Higgins Signed-off-by: Shuah Khan --- tools/debugging/kernel-chktaint | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/tools/debugging/kernel-chktaint b/tools/debugging/kernel-chktaint index f1af27ce9f20..279be06332be 100755 --- a/tools/debugging/kernel-chktaint +++ b/tools/debugging/kernel-chktaint @@ -187,6 +187,7 @@ else echo " * auxiliary taint, defined for and used by distros (#16)" fi + T=`expr $T / 2` if [ `expr $T % 2` -eq 0 ]; then addout " " @@ -195,6 +196,14 @@ else echo " * kernel was built with the struct randomization plugin (#17)" fi +T=`expr $T / 2` +if [ `expr $T % 2` -eq 0 ]; then + addout " " +else + addout "N" + echo " * an in-kernel test (such as a KUnit test) has been run (#18)" +fi + echo "For a more detailed explanation of the various taint flags see" echo " Documentation/admin-guide/tainted-kernels.rst in the Linux kernel sources" echo " or https://kernel.org/doc/html/latest/admin-guide/tainted-kernels.html" From a66de5283e16602b74658289360505ceeb308c90 Mon Sep 17 00:00:00 2001 From: Michael Ellerman Date: Wed, 7 Sep 2022 09:37:17 +1000 Subject: [PATCH 126/158] powerpc/pseries: Fix plpks crash on non-pseries MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit As reported[1] by Nathan, the recently added plpks driver will crash if it's built into the kernel and booted on a non-pseries machine, eg powernv: kernel BUG at arch/powerpc/kernel/syscall.c:39! Oops: Exception in kernel mode, sig: 5 [#1] LE PAGE_SIZE=64K MMU=Radix SMP NR_CPUS=2048 NUMA PowerNV ... NIP system_call_exception+0x90/0x3d0 LR system_call_common+0xec/0x250 Call Trace: 0xc0000000035c3e10 (unreliable) system_call_common+0xec/0x250 --- interrupt: c00 at plpar_hcall+0x38/0x60 NIP: c0000000000e4300 LR: c00000000202945c CTR: 0000000000000000 REGS: c0000000035c3e80 TRAP: 0c00 Not tainted (6.0.0-rc4) MSR: 9000000002009033 CR: 28000284 XER: 00000000 ... NIP plpar_hcall+0x38/0x60 LR pseries_plpks_init+0x64/0x23c --- interrupt: c00 On powernv Linux is the hypervisor, so a hypercall just ends up going to the syscall path, which BUGs if the syscall (hypercall) didn't come from userspace. The fix is simply to not probe the plpks driver on non-pseries machines. [1] https://lore.kernel.org/linuxppc-dev/Yxe06fbq18Wv9y3W@dev-arch.thelio-3990X/ Fixes: 2454a7af0f2a ("powerpc/pseries: define driver for Platform KeyStore") Reported-by: Nathan Chancellor Signed-off-by: Michael Ellerman Tested-by: Dan Horák Reviewed-by: Dan Horák Tested-by: Nathan Chancellor Link: https://lore.kernel.org/r/20220907065038.1604504-1-mpe@ellerman.id.au --- arch/powerpc/platforms/pseries/plpks.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/powerpc/platforms/pseries/plpks.c b/arch/powerpc/platforms/pseries/plpks.c index 52aaa2894606..f4b5b5a64db3 100644 --- a/arch/powerpc/platforms/pseries/plpks.c +++ b/arch/powerpc/platforms/pseries/plpks.c @@ -17,6 +17,7 @@ #include #include #include +#include #include "plpks.h" @@ -457,4 +458,4 @@ static __init int pseries_plpks_init(void) return rc; } -arch_initcall(pseries_plpks_init); +machine_arch_initcall(pseries, pseries_plpks_init); From 992db92b07736ec98432a97b61c37dfe7bd06dd9 Mon Sep 17 00:00:00 2001 From: Chengming Gui Date: Tue, 6 Sep 2022 09:26:37 +0800 Subject: [PATCH 127/158] drm/amd/amdgpu: add rlc_firmware_header_v2_4 to amdgpu_firmware_header Add missing structure to avoid incorrect size and version check. Signed-off-by: Chengming Gui Reviewed-by: Feifei Xu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h index ebed3f5226db..96b6cf4c4d54 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h @@ -390,6 +390,7 @@ union amdgpu_firmware_header { struct rlc_firmware_header_v2_1 rlc_v2_1; struct rlc_firmware_header_v2_2 rlc_v2_2; struct rlc_firmware_header_v2_3 rlc_v2_3; + struct rlc_firmware_header_v2_4 rlc_v2_4; struct sdma_firmware_header_v1_0 sdma; struct sdma_firmware_header_v1_1 sdma_v1_1; struct sdma_firmware_header_v2_0 sdma_v2_0; From bbcbd63231cbd1405ff8e987870b02abc6a4d669 Mon Sep 17 00:00:00 2001 From: Evan Quan Date: Thu, 1 Sep 2022 13:48:58 +0800 Subject: [PATCH 128/158] drm/amd/pm: add missing SetMGpuFanBoostLimitRpm mapping for SMU 13.0.7 Missing SetMGpuFanBoostLimitRpm mapping leads to loading failure for SMU 13.0.7. Signed-off-by: Evan Quan Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c index fcf24c540859..c422bf8a09b1 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c @@ -120,6 +120,7 @@ static struct cmn2asic_msg_mapping smu_v13_0_7_message_map[SMU_MSG_MAX_COUNT] = MSG_MAP(DisallowGfxOff, PPSMC_MSG_DisallowGfxOff, 0), MSG_MAP(Mode1Reset, PPSMC_MSG_Mode1Reset, 0), MSG_MAP(PrepareMp1ForUnload, PPSMC_MSG_PrepareMp1ForUnload, 0), + MSG_MAP(SetMGpuFanBoostLimitRpm, PPSMC_MSG_SetMGpuFanBoostLimitRpm, 0), }; static struct cmn2asic_mapping smu_v13_0_7_clk_map[SMU_CLK_COUNT] = { From cbfac7fa491651c57926c99edeb7495c6c1aeac2 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Fri, 2 Sep 2022 15:01:05 +0200 Subject: [PATCH 129/158] drm/amd/display: fix memory leak when using debugfs_lookup() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When calling debugfs_lookup() the result must have dput() called on it, otherwise the memory will leak over time. Fix this up by properly calling dput(). Cc: Harry Wentland Cc: Leo Li Cc: Rodrigo Siqueira Cc: Alex Deucher Cc: "Christian König" Cc: "Pan, Xinhui" Cc: David Airlie Cc: Daniel Vetter Cc: Wayne Lin Cc: hersen wu Cc: Wenjing Liu Cc: Patrik Jakobsson Cc: Thelford Williams Cc: Fangzhi Zuo Cc: Yongzhi Liu Cc: Mikita Lipski Cc: Jiapeng Chong Cc: Bhanuprakash Modem Cc: Sean Paul Cc: amd-gfx@lists.freedesktop.org Cc: dri-devel@lists.freedesktop.org Cc: stable@vger.kernel.org Reviewed-by: Rodrigo Siqueira Signed-off-by: Greg Kroah-Hartman Signed-off-by: Rodrigo Siqueira Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c index 0e48824f55e3..ee242d9d8b06 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c @@ -3288,6 +3288,7 @@ void crtc_debugfs_init(struct drm_crtc *crtc) &crc_win_y_end_fops); debugfs_create_file_unsafe("crc_win_update", 0644, dir, crtc, &crc_win_update_fops); + dput(dir); #endif debugfs_create_file("amdgpu_current_bpc", 0644, crtc->debugfs_entry, crtc, &amdgpu_current_bpc_fops); From ae0448bc88b7827babda73edfe587ccc222baccc Mon Sep 17 00:00:00 2001 From: Yifan Zhang Date: Sun, 4 Sep 2022 15:53:27 +0800 Subject: [PATCH 130/158] drm/amdkfd: print address in hex format rather than decimal Addresses should be printed in hex format. Signed-off-by: Yifan Zhang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index cbd593f7d553..2170db83e41d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -1728,7 +1728,7 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu( add_kgd_mem_to_kfd_bo_list(*mem, avm->process_info, user_addr); if (user_addr) { - pr_debug("creating userptr BO for user_addr = %llu\n", user_addr); + pr_debug("creating userptr BO for user_addr = %llx\n", user_addr); ret = init_user_pages(*mem, user_addr, criu_resume); if (ret) goto allocate_init_user_pages_failed; From d832db12af4a057582011a590f60f32853680dd4 Mon Sep 17 00:00:00 2001 From: Yifan Zhang Date: Tue, 6 Sep 2022 13:09:20 +0800 Subject: [PATCH 131/158] drm/amdgpu: correct doorbell range/size value for CSDMA_DOORBELL_RANGE current function mixes CSDMA_DOORBELL_RANGE and SDMA0_DOORBELL_RANGE range/size manipulation, while these 2 registers have difference size field mask. Remove range/size manipulation for SDMA0_DOORBELL_RANGE. Signed-off-by: Yifan Zhang Reviewed-by: Xiaojian Du Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/nbio_v7_7.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_7.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_7.c index 1dc95ef21da6..f30bc826a878 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_7.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_7.c @@ -68,12 +68,6 @@ static void nbio_v7_7_sdma_doorbell_range(struct amdgpu_device *adev, int instan doorbell_range = REG_SET_FIELD(doorbell_range, GDC0_BIF_CSDMA_DOORBELL_RANGE, SIZE, doorbell_size); - doorbell_range = REG_SET_FIELD(doorbell_range, - GDC0_BIF_SDMA0_DOORBELL_RANGE, - OFFSET, doorbell_index); - doorbell_range = REG_SET_FIELD(doorbell_range, - GDC0_BIF_SDMA0_DOORBELL_RANGE, - SIZE, doorbell_size); } else { doorbell_range = REG_SET_FIELD(doorbell_range, GDC0_BIF_SDMA0_DOORBELL_RANGE, From aac4cec1ec45d72bd03eaf3fd772c5a609f5ed26 Mon Sep 17 00:00:00 2001 From: Guchun Chen Date: Fri, 2 Sep 2022 14:08:55 +0800 Subject: [PATCH 132/158] drm/amdgpu: prevent toc firmware memory leak It's missed in psp fini. Signed-off-by: Guchun Chen Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 9f7a5e393f85..35738db12cd1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -486,11 +486,14 @@ static int psp_sw_fini(void *handle) release_firmware(psp->ta_fw); psp->ta_fw = NULL; } - if (adev->psp.cap_fw) { + if (psp->cap_fw) { release_firmware(psp->cap_fw); psp->cap_fw = NULL; } - + if (psp->toc_fw) { + release_firmware(psp->toc_fw); + psp->toc_fw = NULL; + } if (adev->ip_versions[MP0_HWIP][0] == IP_VERSION(11, 0, 0) || adev->ip_versions[MP0_HWIP][0] == IP_VERSION(11, 0, 7)) psp_sysfs_fini(adev); From e3d3fd1cea22ada5a06284d7bbb6cd5794d8545a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Wed, 7 Sep 2022 11:56:22 +0200 Subject: [PATCH 133/158] drm/ttm: cleanup the resource of ghost objects after locking them MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Otherwise lockdep will complain about cleaning up the bulk_move. Signed-off-by: Christian König Reviewed-by: Matthew Auld Link: https://patchwork.freedesktop.org/patch/msgid/20220907100051.570641-1-christian.koenig@amd.com Fixes: d91c411c744b ("drm/ttm: update bulk move object of ghost BO") --- drivers/gpu/drm/ttm/ttm_bo_util.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c index 57a27847206f..911141d16e95 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_util.c +++ b/drivers/gpu/drm/ttm/ttm_bo_util.c @@ -236,6 +236,11 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo, if (bo->type != ttm_bo_type_sg) fbo->base.base.resv = &fbo->base.base._resv; + dma_resv_init(&fbo->base.base._resv); + fbo->base.base.dev = NULL; + ret = dma_resv_trylock(&fbo->base.base._resv); + WARN_ON(!ret); + if (fbo->base.resource) { ttm_resource_set_bo(fbo->base.resource, &fbo->base); bo->resource = NULL; @@ -244,11 +249,6 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo, fbo->base.bulk_move = NULL; } - dma_resv_init(&fbo->base.base._resv); - fbo->base.base.dev = NULL; - ret = dma_resv_trylock(&fbo->base.base._resv); - WARN_ON(!ret); - ret = dma_resv_reserve_fences(&fbo->base.base._resv, 1); if (ret) { kfree(fbo); From d76034a427a2660b080bc155e4fd8f6393eefb48 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Cl=C3=A9ment=20P=C3=A9ron?= Date: Tue, 6 Sep 2022 17:30:33 +0200 Subject: [PATCH 134/158] drm/panfrost: devfreq: set opp to the recommended one to configure regulator MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Enabling panfrost GPU OPP with dynamic regulator will make OPP responsible to enable and configure it. Unfortunately OPP configure and enable the regulator when an OPP is asked to be set, which is not the case during panfrost_devfreq_init(). This leave the regulator unconfigured and if no GPU load is triggered, no OPP is asked to be set which make the regulator framework switching it off during regulator_late_cleanup() without noticing and therefore make the board hang as any access to GPU memory space make bus locks up. Call dev_pm_opp_set_opp() with the recommend OPP in panfrost_devfreq_init() to enable the regulator, this will properly configure and enable the regulator and will avoid any switch off by regulator_late_cleanup(). Suggested-by: Viresh Kumar Signed-off-by: Clément Péron Reviewed-by: Steven Price Signed-off-by: Steven Price Link: https://patchwork.freedesktop.org/patch/msgid/20220906153034.153321-5-peron.clem@gmail.com --- drivers/gpu/drm/panfrost/panfrost_devfreq.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/gpu/drm/panfrost/panfrost_devfreq.c b/drivers/gpu/drm/panfrost/panfrost_devfreq.c index 5110cd9b2425..fe5f12f16a63 100644 --- a/drivers/gpu/drm/panfrost/panfrost_devfreq.c +++ b/drivers/gpu/drm/panfrost/panfrost_devfreq.c @@ -131,6 +131,17 @@ int panfrost_devfreq_init(struct panfrost_device *pfdev) return PTR_ERR(opp); panfrost_devfreq_profile.initial_freq = cur_freq; + + /* + * Set the recommend OPP this will enable and configure the regulator + * if any and will avoid a switch off by regulator_late_cleanup() + */ + ret = dev_pm_opp_set_opp(dev, opp); + if (ret) { + DRM_DEV_ERROR(dev, "Couldn't set recommended OPP\n"); + return ret; + } + dev_pm_opp_put(opp); /* From d105d6920ec758125b69c0b097bf498348888a9d Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Fri, 2 Sep 2022 14:28:02 +0100 Subject: [PATCH 135/158] arm64/ptrace: Don't clear calling process' TIF_SME on OOM If allocating memory for the target SVE state in za_set() fails we clear TIF_SME for the ptracing task which is obviously not correct. If we are here we know that the target task already had neither TIF_SVE nor TIF_SME set since we only need to allocate if either the target had not used either SVE or SME and had no need to allocate state before or we just changed the vector length with vec_set_vector_length() which clears TIF_ for us on allocation failure so just remove the clear entirely. Reported-by: Wang ShaoBo Signed-off-by: Mark Brown Link: https://lore.kernel.org/r/20220902132802.39682-1-broonie@kernel.org Signed-off-by: Will Deacon --- arch/arm64/kernel/ptrace.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/arch/arm64/kernel/ptrace.c b/arch/arm64/kernel/ptrace.c index eb7c08dfb834..041d2ae5c30a 100644 --- a/arch/arm64/kernel/ptrace.c +++ b/arch/arm64/kernel/ptrace.c @@ -1084,7 +1084,6 @@ static int za_set(struct task_struct *target, if (!target->thread.sve_state) { sve_alloc(target, false); if (!target->thread.sve_state) { - clear_thread_flag(TIF_SME); ret = -ENOMEM; goto out; } @@ -1094,7 +1093,6 @@ static int za_set(struct task_struct *target, sme_alloc(target); if (!target->thread.za_state) { ret = -ENOMEM; - clear_tsk_thread_flag(target, TIF_SME); goto out; } From 3c8400532dd8305024ff6eea38707de20b1b9822 Mon Sep 17 00:00:00 2001 From: Pavel Begunkov Date: Thu, 8 Sep 2022 14:01:10 +0100 Subject: [PATCH 136/158] io_uring/net: copy addr for zc on POLL_FIRST Every time we return from an issue handler and expect the request to be retried we should also setup it for async exec ourselves. Do that when we return on IORING_RECVSEND_POLL_FIRST in io_sendzc(), otherwise it'll re-read the address, which might be a surprise for the userspace. Fixes: 092aeedb750a9 ("io_uring: allow to pass addr into sendzc") Signed-off-by: Pavel Begunkov Link: https://lore.kernel.org/r/ab1d0657890d6721339c56d2e161a4bba06f85d0.1662642013.git.asml.silence@gmail.com Signed-off-by: Jens Axboe --- io_uring/net.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/io_uring/net.c b/io_uring/net.c index 7047c1342541..e9efed40cf3d 100644 --- a/io_uring/net.c +++ b/io_uring/net.c @@ -1003,9 +1003,6 @@ int io_sendzc(struct io_kiocb *req, unsigned int issue_flags) unsigned msg_flags, cflags; int ret, min_ret = 0; - if (!(req->flags & REQ_F_POLLED) && - (zc->flags & IORING_RECVSEND_POLL_FIRST)) - return -EAGAIN; sock = sock_from_file(req->file); if (unlikely(!sock)) return -ENOTSOCK; @@ -1030,6 +1027,10 @@ int io_sendzc(struct io_kiocb *req, unsigned int issue_flags) msg.msg_namelen = zc->addr_len; } + if (!(req->flags & REQ_F_POLLED) && + (zc->flags & IORING_RECVSEND_POLL_FIRST)) + return io_setup_async_addr(req, addr, issue_flags); + if (zc->flags & IORING_RECVSEND_FIXED_BUF) { ret = io_import_fixed(WRITE, &msg.msg_iter, req->imu, (u64)(uintptr_t)zc->buf, zc->len); From 7864d8f7c088aad988c44c631f1ceed9179cf2cf Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Mon, 5 Sep 2022 14:42:09 +0300 Subject: [PATCH 137/158] libperf evlist: Fix per-thread mmaps for multi-threaded targets MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The offending commit removed mmap_per_thread(), which did not consider the different set-output rules for per-thread mmaps i.e. in the per-thread case set-output is used for file descriptors of the same thread not the same cpu. This was not immediately noticed because it only happens with multi-threaded targets and we do not have a test for that yet. Reinstate mmap_per_thread() expanding it to cover also system-wide per-cpu events i.e. to continue to allow the mixing of per-thread and per-cpu mmaps. Debug messages (with -vv) show the file descriptors that are opened with sys_perf_event_open. New debug messages are added (needs -vvv) that show also which file descriptors are mmapped and which are redirected with set-output. In the per-cpu case (cpu != -1) file descriptors for the same CPU are set-output to the first file descriptor for that CPU. In the per-thread case (cpu == -1) file descriptors for the same thread are set-output to the first file descriptor for that thread. Example (process 17489 has 2 threads): Before (but with new debug prints): $ perf record --no-bpf-event -vvv --per-thread -p 17489 sys_perf_event_open: pid 17489 cpu -1 group_fd -1 flags 0x8 = 5 sys_perf_event_open: pid 17490 cpu -1 group_fd -1 flags 0x8 = 6 libperf: idx 0: mmapping fd 5 libperf: idx 0: set output fd 6 -> 5 failed to mmap with 22 (Invalid argument) After: $ perf record --no-bpf-event -vvv --per-thread -p 17489 sys_perf_event_open: pid 17489 cpu -1 group_fd -1 flags 0x8 = 5 sys_perf_event_open: pid 17490 cpu -1 group_fd -1 flags 0x8 = 6 libperf: mmap_per_thread: nr cpu values (may include -1) 1 nr threads 2 libperf: idx 0: mmapping fd 5 libperf: idx 1: mmapping fd 6 [ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 0.018 MB perf.data (15 samples) ] Per-cpu example (process 20341 has 2 threads, same as above): $ perf record --no-bpf-event -vvv -p 20341 sys_perf_event_open: pid 20341 cpu 0 group_fd -1 flags 0x8 = 5 sys_perf_event_open: pid 20342 cpu 0 group_fd -1 flags 0x8 = 6 sys_perf_event_open: pid 20341 cpu 1 group_fd -1 flags 0x8 = 7 sys_perf_event_open: pid 20342 cpu 1 group_fd -1 flags 0x8 = 8 sys_perf_event_open: pid 20341 cpu 2 group_fd -1 flags 0x8 = 9 sys_perf_event_open: pid 20342 cpu 2 group_fd -1 flags 0x8 = 10 sys_perf_event_open: pid 20341 cpu 3 group_fd -1 flags 0x8 = 11 sys_perf_event_open: pid 20342 cpu 3 group_fd -1 flags 0x8 = 12 sys_perf_event_open: pid 20341 cpu 4 group_fd -1 flags 0x8 = 13 sys_perf_event_open: pid 20342 cpu 4 group_fd -1 flags 0x8 = 14 sys_perf_event_open: pid 20341 cpu 5 group_fd -1 flags 0x8 = 15 sys_perf_event_open: pid 20342 cpu 5 group_fd -1 flags 0x8 = 16 sys_perf_event_open: pid 20341 cpu 6 group_fd -1 flags 0x8 = 17 sys_perf_event_open: pid 20342 cpu 6 group_fd -1 flags 0x8 = 18 sys_perf_event_open: pid 20341 cpu 7 group_fd -1 flags 0x8 = 19 sys_perf_event_open: pid 20342 cpu 7 group_fd -1 flags 0x8 = 20 libperf: mmap_per_cpu: nr cpu values 8 nr threads 2 libperf: idx 0: mmapping fd 5 libperf: idx 0: set output fd 6 -> 5 libperf: idx 1: mmapping fd 7 libperf: idx 1: set output fd 8 -> 7 libperf: idx 2: mmapping fd 9 libperf: idx 2: set output fd 10 -> 9 libperf: idx 3: mmapping fd 11 libperf: idx 3: set output fd 12 -> 11 libperf: idx 4: mmapping fd 13 libperf: idx 4: set output fd 14 -> 13 libperf: idx 5: mmapping fd 15 libperf: idx 5: set output fd 16 -> 15 libperf: idx 6: mmapping fd 17 libperf: idx 6: set output fd 18 -> 17 libperf: idx 7: mmapping fd 19 libperf: idx 7: set output fd 20 -> 19 [ perf record: Woken up 7 times to write data ] [ perf record: Captured and wrote 0.020 MB perf.data (17 samples) ] Fixes: ae4f8ae16a078964 ("libperf evlist: Allow mixing per-thread and per-cpu mmaps") Reported-by: Tomáš Trnka Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=216441 Signed-off-by: Adrian Hunter Acked-by: Jiri Olsa Cc: Adrian Hunter Cc: Ian Rogers Cc: Ingo Molnar Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20220905114209.8389-1-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/lib/perf/evlist.c | 50 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 50 insertions(+) diff --git a/tools/lib/perf/evlist.c b/tools/lib/perf/evlist.c index e6c98a6e3908..6b1bafe267a4 100644 --- a/tools/lib/perf/evlist.c +++ b/tools/lib/perf/evlist.c @@ -486,6 +486,7 @@ mmap_per_evsel(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, if (ops->idx) ops->idx(evlist, evsel, mp, idx); + pr_debug("idx %d: mmapping fd %d\n", idx, *output); if (ops->mmap(map, mp, *output, evlist_cpu) < 0) return -1; @@ -494,6 +495,7 @@ mmap_per_evsel(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, if (!idx) perf_evlist__set_mmap_first(evlist, map, overwrite); } else { + pr_debug("idx %d: set output fd %d -> %d\n", idx, fd, *output); if (ioctl(fd, PERF_EVENT_IOC_SET_OUTPUT, *output) != 0) return -1; @@ -519,6 +521,48 @@ mmap_per_evsel(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, return 0; } +static int +mmap_per_thread(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, + struct perf_mmap_param *mp) +{ + int nr_threads = perf_thread_map__nr(evlist->threads); + int nr_cpus = perf_cpu_map__nr(evlist->all_cpus); + int cpu, thread, idx = 0; + int nr_mmaps = 0; + + pr_debug("%s: nr cpu values (may include -1) %d nr threads %d\n", + __func__, nr_cpus, nr_threads); + + /* per-thread mmaps */ + for (thread = 0; thread < nr_threads; thread++, idx++) { + int output = -1; + int output_overwrite = -1; + + if (mmap_per_evsel(evlist, ops, idx, mp, 0, thread, &output, + &output_overwrite, &nr_mmaps)) + goto out_unmap; + } + + /* system-wide mmaps i.e. per-cpu */ + for (cpu = 1; cpu < nr_cpus; cpu++, idx++) { + int output = -1; + int output_overwrite = -1; + + if (mmap_per_evsel(evlist, ops, idx, mp, cpu, 0, &output, + &output_overwrite, &nr_mmaps)) + goto out_unmap; + } + + if (nr_mmaps != evlist->nr_mmaps) + pr_err("Miscounted nr_mmaps %d vs %d\n", nr_mmaps, evlist->nr_mmaps); + + return 0; + +out_unmap: + perf_evlist__munmap(evlist); + return -1; +} + static int mmap_per_cpu(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, struct perf_mmap_param *mp) @@ -528,6 +572,8 @@ mmap_per_cpu(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, int nr_mmaps = 0; int cpu, thread; + pr_debug("%s: nr cpu values %d nr threads %d\n", __func__, nr_cpus, nr_threads); + for (cpu = 0; cpu < nr_cpus; cpu++) { int output = -1; int output_overwrite = -1; @@ -569,6 +615,7 @@ int perf_evlist__mmap_ops(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, struct perf_mmap_param *mp) { + const struct perf_cpu_map *cpus = evlist->all_cpus; struct perf_evsel *evsel; if (!ops || !ops->get || !ops->mmap) @@ -588,6 +635,9 @@ int perf_evlist__mmap_ops(struct perf_evlist *evlist, if (evlist->pollfd.entries == NULL && perf_evlist__alloc_pollfd(evlist) < 0) return -ENOMEM; + if (perf_cpu_map__empty(cpus)) + return mmap_per_thread(evlist, ops, mp); + return mmap_per_cpu(evlist, ops, mp); } From 1706623e940347ad23fdf77910eca4905dc37f91 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Mon, 5 Sep 2022 10:47:35 +0300 Subject: [PATCH 138/158] perf dlfilter dlfilter-show-cycles: Fix types for print format Avoid compiler warning about format %llu that expects long long unsigned int but argument has type __u64. Reported-by: Arnaldo Carvalho de Melo Fixes: c3afd6e50fce824f ("perf dlfilter: Add dlfilter-show-cycles") Signed-off-by: Adrian Hunter Cc: Adrian Hunter Cc: Ian Rogers Cc: Jiri Olsa Cc: Namhyung Kim Link: https://lore.kernel.org/r/20220905074735.4513-1-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/dlfilters/dlfilter-show-cycles.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/perf/dlfilters/dlfilter-show-cycles.c b/tools/perf/dlfilters/dlfilter-show-cycles.c index 9eccc97bff82..6d47298ebe9f 100644 --- a/tools/perf/dlfilters/dlfilter-show-cycles.c +++ b/tools/perf/dlfilters/dlfilter-show-cycles.c @@ -98,9 +98,9 @@ int filter_event_early(void *data, const struct perf_dlfilter_sample *sample, vo static void print_vals(__u64 cycles, __u64 delta) { if (delta) - printf("%10llu %10llu ", cycles, delta); + printf("%10llu %10llu ", (unsigned long long)cycles, (unsigned long long)delta); else - printf("%10llu %10s ", cycles, ""); + printf("%10llu %10s ", (unsigned long long)cycles, ""); } int filter_event(void *data, const struct perf_dlfilter_sample *sample, void *ctx) From 3705a6ef4068cbc9c2344561345d09630fd278ec Mon Sep 17 00:00:00 2001 From: Yang Jihong Date: Thu, 8 Sep 2022 09:48:54 +0800 Subject: [PATCH 139/158] perf lock: Remove redundant word 'contention' in help message Before: # perf lock -h Usage: perf lock [] {record|report|script|info|contention|contention} -D, --dump-raw-trace dump raw trace in ASCII -f, --force don't complain, do it -i, --input input file name -v, --verbose be more verbose (show symbol address, etc) --kallsyms kallsyms pathname --vmlinux vmlinux pathname After: # perf lock -h Usage: perf lock [] {record|report|script|info|contention} -D, --dump-raw-trace dump raw trace in ASCII -f, --force don't complain, do it -i, --input input file name -v, --verbose be more verbose (show symbol address, etc) --kallsyms kallsyms pathname --vmlinux vmlinux pathname Fixes: 528b9cab3b813a3b ("perf lock: Add 'contention' subcommand") Signed-off-by: Yang Jihong Acked-by: Namhyung Kim Cc: Alexander Shishkin Cc: Ingo Molnar Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20220908014854.151203-1-yangjihong1@huawei.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-lock.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/tools/perf/builtin-lock.c b/tools/perf/builtin-lock.c index dd11d3471baf..ea40ae52cd2c 100644 --- a/tools/perf/builtin-lock.c +++ b/tools/perf/builtin-lock.c @@ -1874,8 +1874,7 @@ int cmd_lock(int argc, const char **argv) NULL }; const char *const lock_subcommands[] = { "record", "report", "script", - "info", "contention", - "contention", NULL }; + "info", "contention", NULL }; const char *lock_usage[] = { NULL, NULL From 82b2425fad2dd47204b3da589b679220f8aacc0e Mon Sep 17 00:00:00 2001 From: Zhengjun Xing Date: Thu, 8 Sep 2022 15:00:30 +0800 Subject: [PATCH 140/158] perf script: Fix Cannot print 'iregs' field for hybrid systems Commit b91e5492f9d7ca89 ("perf record: Add a dummy event on hybrid systems to collect metadata records") adds a dummy event on hybrid systems to fix the symbol "unknown" issue when the workload is created in a P-core but runs on an E-core. The added dummy event will cause "perf script -F iregs" to fail. Dummy events do not have "iregs" attribute set, so when we do evsel__check_attr, the "iregs" attribute check will fail, so the issue happened. The following commit [1] has fixed a similar issue by skipping the attr check for the dummy event because it does not have any samples anyway. It works okay for the normal mode, but the issue still happened when running the test in the pipe mode. In the pipe mode, it calls process_attr() which still checks the attr for the dummy event. This commit fixed the issue by skipping the attr check for the dummy event in the API evsel__check_attr, Otherwise, we have to patch everywhere when evsel__check_attr() is called. Before: #./perf record -o - --intr-regs=di,r8,dx,cx -e br_inst_retired.near_call:p -c 1000 --per-thread true 2>/dev/null|./perf script -F iregs |head -5 Samples for 'dummy:HG' event do not have IREGS attribute set. Cannot print 'iregs' field. 0x120 [0x90]: failed to process type: 64 # After: # ./perf record -o - --intr-regs=di,r8,dx,cx -e br_inst_retired.near_call:p -c 1000 --per-thread true 2>/dev/null|./perf script -F iregs |head -5 ABI:2 CX:0x55b8efa87000 DX:0x55b8efa7e000 DI:0xffffba5e625efbb0 R8:0xffff90e51f8ae100 ABI:2 CX:0x7f1dae1e4000 DX:0xd0 DI:0xffff90e18c675ac0 R8:0x71 ABI:2 CX:0xcc0 DX:0x1 DI:0xffff90e199880240 R8:0x0 ABI:2 CX:0xffff90e180dd7500 DX:0xffff90e180dd7500 DI:0xffff90e180043500 R8:0x1 ABI:2 CX:0x50 DX:0xffff90e18c583bd0 DI:0xffff90e1998803c0 R8:0x58 # [1]https://lore.kernel.org/lkml/20220831124041.219925-1-jolsa@kernel.org/ Fixes: b91e5492f9d7ca89 ("perf record: Add a dummy event on hybrid systems to collect metadata records") Suggested-by: Namhyung Kim Signed-off-by: Xing Zhengjun Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Ian Rogers Cc: Ingo Molnar Cc: Kan Liang Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20220908070030.3455164-1-zhengjun.xing@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-script.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index 304d234d8e84..029b4330e59b 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -445,6 +445,9 @@ static int evsel__check_attr(struct evsel *evsel, struct perf_session *session) struct perf_event_attr *attr = &evsel->core.attr; bool allow_user_set; + if (evsel__is_dummy_event(evsel)) + return 0; + if (perf_header__has_feat(&session->header, HEADER_STAT)) return 0; From d0b34d5bf7ffc4b906f1da901376c3caf0f0d7c0 Mon Sep 17 00:00:00 2001 From: Eliav Farber Date: Thu, 8 Sep 2022 15:24:29 +0000 Subject: [PATCH 141/158] dt-bindings: hwmon: (mr75203) fix "intel,vm-map" property to be optional Change "intel,vm-map" property to be optional instead of required. The driver implementation indicates it is not mandatory to have "intel,vm-map" in the device tree: - probe doesn't fail in case it is absent. - explicit comment in code - "Incase intel,vm-map property is not defined, we assume incremental channel numbers". Fixes: 748022ef093f ("hwmon: Add DT bindings schema for PVT controller") Signed-off-by: Eliav Farber Acked-by: Rob Herring Link: https://lore.kernel.org/r/20220908152449.35457-2-farbere@amazon.com Signed-off-by: Guenter Roeck --- Documentation/devicetree/bindings/hwmon/moortec,mr75203.yaml | 1 - 1 file changed, 1 deletion(-) diff --git a/Documentation/devicetree/bindings/hwmon/moortec,mr75203.yaml b/Documentation/devicetree/bindings/hwmon/moortec,mr75203.yaml index b79f069a04c2..8ea97e774364 100644 --- a/Documentation/devicetree/bindings/hwmon/moortec,mr75203.yaml +++ b/Documentation/devicetree/bindings/hwmon/moortec,mr75203.yaml @@ -48,7 +48,6 @@ required: - compatible - reg - reg-names - - intel,vm-map - clocks - resets - "#thermal-sensor-cells" From 81114fc3d27bf5b06b2137d2fd2b63da656a8b90 Mon Sep 17 00:00:00 2001 From: Eliav Farber Date: Thu, 8 Sep 2022 15:24:30 +0000 Subject: [PATCH 142/158] hwmon: (mr75203) fix VM sensor allocation when "intel,vm-map" not defined Bug - in case "intel,vm-map" is missing in device-tree ,'num' is set to 0, and no voltage channel infos are allocated. The reason num is set to 0 when "intel,vm-map" is missing is to set the entire pvt->vm_idx[] with incremental channel numbers, but it didn't take into consideration that same num is used later in devm_kcalloc(). If "intel,vm-map" does exist there is no need to set the unspecified channels with incremental numbers, because the unspecified channels can't be accessed in pvt_read_in() which is the only other place besides the probe functions that uses pvt->vm_idx[]. This change fixes the bug by moving the incremental channel numbers setting to be done only if "intel,vm-map" property is defined (starting loop from 0), and removing 'num = 0'. Fixes: 9d823351a337 ("hwmon: Add hardware monitoring driver for Moortec MR75203 PVT controller") Signed-off-by: Eliav Farber Reviewed-by: Andy Shevchenko Link: https://lore.kernel.org/r/20220908152449.35457-3-farbere@amazon.com Signed-off-by: Guenter Roeck --- drivers/hwmon/mr75203.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/drivers/hwmon/mr75203.c b/drivers/hwmon/mr75203.c index 26278b0f17a9..8b72e8fe34c1 100644 --- a/drivers/hwmon/mr75203.c +++ b/drivers/hwmon/mr75203.c @@ -584,7 +584,12 @@ static int mr75203_probe(struct platform_device *pdev) ret = device_property_read_u8_array(dev, "intel,vm-map", pvt->vm_idx, vm_num); if (ret) { - num = 0; + /* + * Incase intel,vm-map property is not defined, we + * assume incremental channel numbers. + */ + for (i = 0; i < vm_num; i++) + pvt->vm_idx[i] = i; } else { for (i = 0; i < vm_num; i++) if (pvt->vm_idx[i] >= vm_num || @@ -594,13 +599,6 @@ static int mr75203_probe(struct platform_device *pdev) } } - /* - * Incase intel,vm-map property is not defined, we assume - * incremental channel numbers. - */ - for (i = num; i < vm_num; i++) - pvt->vm_idx[i] = i; - in_config = devm_kcalloc(dev, num + 1, sizeof(*in_config), GFP_KERNEL); if (!in_config) From bb9195bd6664d94d71647631593e09f705ff5edd Mon Sep 17 00:00:00 2001 From: Eliav Farber Date: Thu, 8 Sep 2022 15:24:31 +0000 Subject: [PATCH 143/158] hwmon: (mr75203) update pvt->v_num and vm_num to the actual number of used sensors This issue is relevant when "intel,vm-map" is set in device-tree, and defines a lower number of VMs than actually supported. This change is needed for all places that use pvt->v_num or vm_num later on in the code. Fixes: 9d823351a337 ("hwmon: Add hardware monitoring driver for Moortec MR75203 PVT controller") Signed-off-by: Eliav Farber Reviewed-by: Andy Shevchenko Link: https://lore.kernel.org/r/20220908152449.35457-4-farbere@amazon.com Signed-off-by: Guenter Roeck --- drivers/hwmon/mr75203.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/hwmon/mr75203.c b/drivers/hwmon/mr75203.c index 8b72e8fe34c1..be02f32bf143 100644 --- a/drivers/hwmon/mr75203.c +++ b/drivers/hwmon/mr75203.c @@ -595,6 +595,8 @@ static int mr75203_probe(struct platform_device *pdev) if (pvt->vm_idx[i] >= vm_num || pvt->vm_idx[i] == 0xff) { num = i; + pvt->v_num = i; + vm_num = i; break; } } From 227a3a2fc31d8e4bb9c88d4804e19530af245b1b Mon Sep 17 00:00:00 2001 From: Eliav Farber Date: Thu, 8 Sep 2022 15:24:32 +0000 Subject: [PATCH 144/158] hwmon: (mr75203) fix voltage equation for negative source input According to Moortec Embedded Voltage Monitor (MEVM) series 3 data sheet, the minimum input signal is -100mv and maximum input signal is +1000mv. The equation used to convert the digital word to voltage uses mixed types (*val signed and n unsigned), and on 64 bit machines also has different size, since sizeof(u32) = 4 and sizeof(long) = 8. So when measuring a negative input, n will be small enough, such that PVT_N_CONST * n < PVT_R_CONST, and the result of (PVT_N_CONST * n - PVT_R_CONST) will overflow to a very big positive 32 bit number. Then when storing the result in *val it will be the same value just in 64 bit (instead of it representing a negative number which will what happen when sizeof(long) = 4). When -1023 <= (PVT_N_CONST * n - PVT_R_CONST) <= -1 dividing the number by 1024 should result of in 0, but because ">> 10" is used, and the sign bit is used to fill the vacated bit positions, it results in -1 (0xf...fffff) which is wrong. This change fixes the sign problem and supports negative values by casting n to long and replacing the shift right with div operation. Fixes: 9d823351a337 ("hwmon: Add hardware monitoring driver for Moortec MR75203 PVT controller") Signed-off-by: Eliav Farber Reviewed-by: Andy Shevchenko Link: https://lore.kernel.org/r/20220908152449.35457-5-farbere@amazon.com Signed-off-by: Guenter Roeck --- drivers/hwmon/mr75203.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/drivers/hwmon/mr75203.c b/drivers/hwmon/mr75203.c index be02f32bf143..6d3b3c499ed8 100644 --- a/drivers/hwmon/mr75203.c +++ b/drivers/hwmon/mr75203.c @@ -201,8 +201,18 @@ static int pvt_read_in(struct device *dev, u32 attr, int channel, long *val) return ret; n &= SAMPLE_DATA_MSK; - /* Convert the N bitstream count into voltage */ - *val = (PVT_N_CONST * n - PVT_R_CONST) >> PVT_CONV_BITS; + /* + * Convert the N bitstream count into voltage. + * To support negative voltage calculation for 64bit machines + * n must be cast to long, since n and *val differ both in + * signedness and in size. + * Division is used instead of right shift, because for signed + * numbers, the sign bit is used to fill the vacated bit + * positions, and if the number is negative, 1 is used. + * BIT(x) may not be used instead of (1 << x) because it's + * unsigned. + */ + *val = (PVT_N_CONST * (long)n - PVT_R_CONST) / (1 << PVT_CONV_BITS); return 0; default: From 0a9eaf616f29ca32068d2d8fe04eeef67505720d Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Thu, 8 Sep 2022 08:04:26 +0200 Subject: [PATCH 145/158] perf tools: Don't install data files with x permissions install(1), by default, installs with rwxr-xr-x permissions. Modify perf's Makefile to pass '-m 644' when installing: * Documentation/tips.txt * examples/bpf/* * perf-completion.sh * perf_dlfilter.h header * scripts/perl/Perf-Trace-Util/lib/Perf/Trace/* * scripts/perl/*.pl * tests/attr/* * tests/attr.py * tests/shell/lib/*.sh * trace/strace/groups/* All those are supposed to be non-executable. Either they are not scripts at all, or they don't have shebang. Signed-off-by: Cc: Adrian Hunter Cc: Andi Kleen Cc: Ingo Molnar Cc: Jiri Olsa Cc: Kan Liang Cc: Leo Yan Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20220908060426.9619-1-jslaby@suse.cz Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/Makefile.perf | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/tools/perf/Makefile.perf b/tools/perf/Makefile.perf index e5921b347153..bd947885a639 100644 --- a/tools/perf/Makefile.perf +++ b/tools/perf/Makefile.perf @@ -954,11 +954,11 @@ ifndef NO_LIBBPF $(call QUIET_INSTALL, bpf-headers) \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perf_include_instdir_SQ)/bpf'; \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perf_include_instdir_SQ)/bpf/linux'; \ - $(INSTALL) include/bpf/*.h -t '$(DESTDIR_SQ)$(perf_include_instdir_SQ)/bpf'; \ - $(INSTALL) include/bpf/linux/*.h -t '$(DESTDIR_SQ)$(perf_include_instdir_SQ)/bpf/linux' + $(INSTALL) include/bpf/*.h -m 644 -t '$(DESTDIR_SQ)$(perf_include_instdir_SQ)/bpf'; \ + $(INSTALL) include/bpf/linux/*.h -m 644 -t '$(DESTDIR_SQ)$(perf_include_instdir_SQ)/bpf/linux' $(call QUIET_INSTALL, bpf-examples) \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perf_examples_instdir_SQ)/bpf'; \ - $(INSTALL) examples/bpf/*.c -t '$(DESTDIR_SQ)$(perf_examples_instdir_SQ)/bpf' + $(INSTALL) examples/bpf/*.c -m 644 -t '$(DESTDIR_SQ)$(perf_examples_instdir_SQ)/bpf' endif $(call QUIET_INSTALL, perf-archive) \ $(INSTALL) $(OUTPUT)perf-archive -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)' @@ -967,13 +967,13 @@ endif ifndef NO_LIBAUDIT $(call QUIET_INSTALL, strace/groups) \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(STRACE_GROUPS_INSTDIR_SQ)'; \ - $(INSTALL) trace/strace/groups/* -t '$(DESTDIR_SQ)$(STRACE_GROUPS_INSTDIR_SQ)' + $(INSTALL) trace/strace/groups/* -m 644 -t '$(DESTDIR_SQ)$(STRACE_GROUPS_INSTDIR_SQ)' endif ifndef NO_LIBPERL $(call QUIET_INSTALL, perl-scripts) \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/Perf-Trace-Util/lib/Perf/Trace'; \ - $(INSTALL) scripts/perl/Perf-Trace-Util/lib/Perf/Trace/* -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/Perf-Trace-Util/lib/Perf/Trace'; \ - $(INSTALL) scripts/perl/*.pl -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl'; \ + $(INSTALL) scripts/perl/Perf-Trace-Util/lib/Perf/Trace/* -m 644 -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/Perf-Trace-Util/lib/Perf/Trace'; \ + $(INSTALL) scripts/perl/*.pl -m 644 -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl'; \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/bin'; \ $(INSTALL) scripts/perl/bin/* -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/bin' endif @@ -990,23 +990,23 @@ endif $(INSTALL) $(DLFILTERS) '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/dlfilters'; $(call QUIET_INSTALL, perf_completion-script) \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(sysconfdir_SQ)/bash_completion.d'; \ - $(INSTALL) perf-completion.sh '$(DESTDIR_SQ)$(sysconfdir_SQ)/bash_completion.d/perf' + $(INSTALL) perf-completion.sh -m 644 '$(DESTDIR_SQ)$(sysconfdir_SQ)/bash_completion.d/perf' $(call QUIET_INSTALL, perf-tip) \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(tip_instdir_SQ)'; \ - $(INSTALL) Documentation/tips.txt -t '$(DESTDIR_SQ)$(tip_instdir_SQ)' + $(INSTALL) Documentation/tips.txt -m 644 -t '$(DESTDIR_SQ)$(tip_instdir_SQ)' install-tests: all install-gtk $(call QUIET_INSTALL, tests) \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests'; \ - $(INSTALL) tests/attr.py '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests'; \ + $(INSTALL) tests/attr.py -m 644 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests'; \ $(INSTALL) tests/pe-file.exe* '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests'; \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/attr'; \ - $(INSTALL) tests/attr/* '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/attr'; \ + $(INSTALL) tests/attr/* -m 644 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/attr'; \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/shell'; \ $(INSTALL) tests/shell/*.sh '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/shell'; \ $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/shell/lib'; \ - $(INSTALL) tests/shell/lib/*.sh '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/shell/lib'; \ - $(INSTALL) tests/shell/lib/*.py '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/shell/lib' + $(INSTALL) tests/shell/lib/*.sh -m 644 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/shell/lib'; \ + $(INSTALL) tests/shell/lib/*.py -m 644 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/shell/lib' install-bin: install-tools install-tests install-traceevent-plugins From 91a9e063cdcfca8fe642b078d6fae4ce49187975 Mon Sep 17 00:00:00 2001 From: Eliav Farber Date: Thu, 8 Sep 2022 15:24:33 +0000 Subject: [PATCH 146/158] hwmon: (mr75203) fix multi-channel voltage reading Fix voltage allocation and reading to support all channels in all VMs. Prior to this change allocation and reading were done only for the first channel in each VM. This change counts the total number of channels for allocation, and takes into account the channel offset when reading the sample data register. Fixes: 9d823351a337 ("hwmon: Add hardware monitoring driver for Moortec MR75203 PVT controller") Signed-off-by: Eliav Farber Reviewed-by: Andy Shevchenko Link: https://lore.kernel.org/r/20220908152449.35457-6-farbere@amazon.com Signed-off-by: Guenter Roeck --- drivers/hwmon/mr75203.c | 29 +++++++++++++++++------------ 1 file changed, 17 insertions(+), 12 deletions(-) diff --git a/drivers/hwmon/mr75203.c b/drivers/hwmon/mr75203.c index 6d3b3c499ed8..812ae40e7d74 100644 --- a/drivers/hwmon/mr75203.c +++ b/drivers/hwmon/mr75203.c @@ -68,8 +68,9 @@ /* VM Individual Macro Register */ #define VM_COM_REG_SIZE 0x200 -#define VM_SDIF_DONE(n) (VM_COM_REG_SIZE + 0x34 + 0x200 * (n)) -#define VM_SDIF_DATA(n) (VM_COM_REG_SIZE + 0x40 + 0x200 * (n)) +#define VM_SDIF_DONE(vm) (VM_COM_REG_SIZE + 0x34 + 0x200 * (vm)) +#define VM_SDIF_DATA(vm, ch) \ + (VM_COM_REG_SIZE + 0x40 + 0x200 * (vm) + 0x4 * (ch)) /* SDA Slave Register */ #define IP_CTRL 0x00 @@ -115,6 +116,7 @@ struct pvt_device { u32 t_num; u32 p_num; u32 v_num; + u32 c_num; u32 ip_freq; u8 *vm_idx; }; @@ -178,14 +180,15 @@ static int pvt_read_in(struct device *dev, u32 attr, int channel, long *val) { struct pvt_device *pvt = dev_get_drvdata(dev); struct regmap *v_map = pvt->v_map; + u8 vm_idx, ch_idx; u32 n, stat; - u8 vm_idx; int ret; - if (channel >= pvt->v_num) + if (channel >= pvt->v_num * pvt->c_num) return -EINVAL; - vm_idx = pvt->vm_idx[channel]; + vm_idx = pvt->vm_idx[channel / pvt->c_num]; + ch_idx = channel % pvt->c_num; switch (attr) { case hwmon_in_input: @@ -196,7 +199,7 @@ static int pvt_read_in(struct device *dev, u32 attr, int channel, long *val) if (ret) return ret; - ret = regmap_read(v_map, VM_SDIF_DATA(vm_idx), &n); + ret = regmap_read(v_map, VM_SDIF_DATA(vm_idx, ch_idx), &n); if(ret < 0) return ret; @@ -499,8 +502,8 @@ static int pvt_reset_control_deassert(struct device *dev, struct pvt_device *pvt static int mr75203_probe(struct platform_device *pdev) { + u32 ts_num, vm_num, pd_num, ch_num, val, index, i; const struct hwmon_channel_info **pvt_info; - u32 ts_num, vm_num, pd_num, val, index, i; struct device *dev = &pdev->dev; u32 *temp_config, *in_config; struct device *hwmon_dev; @@ -541,9 +544,11 @@ static int mr75203_probe(struct platform_device *pdev) ts_num = (val & TS_NUM_MSK) >> TS_NUM_SFT; pd_num = (val & PD_NUM_MSK) >> PD_NUM_SFT; vm_num = (val & VM_NUM_MSK) >> VM_NUM_SFT; + ch_num = (val & CH_NUM_MSK) >> CH_NUM_SFT; pvt->t_num = ts_num; pvt->p_num = pd_num; pvt->v_num = vm_num; + pvt->c_num = ch_num; val = 0; if (ts_num) val++; @@ -580,7 +585,7 @@ static int mr75203_probe(struct platform_device *pdev) } if (vm_num) { - u32 num = vm_num; + u32 total_ch; ret = pvt_get_regmap(pdev, "vm", pvt); if (ret) @@ -604,20 +609,20 @@ static int mr75203_probe(struct platform_device *pdev) for (i = 0; i < vm_num; i++) if (pvt->vm_idx[i] >= vm_num || pvt->vm_idx[i] == 0xff) { - num = i; pvt->v_num = i; vm_num = i; break; } } - in_config = devm_kcalloc(dev, num + 1, + total_ch = ch_num * vm_num; + in_config = devm_kcalloc(dev, total_ch + 1, sizeof(*in_config), GFP_KERNEL); if (!in_config) return -ENOMEM; - memset32(in_config, HWMON_I_INPUT, num); - in_config[num] = 0; + memset32(in_config, HWMON_I_INPUT, total_ch); + in_config[total_ch] = 0; pvt_in.config = in_config; pvt_info[index++] = &pvt_in; From e43212e0f55dc2d6b15d6c174cc0a64b25fab5e7 Mon Sep 17 00:00:00 2001 From: Eliav Farber Date: Thu, 8 Sep 2022 15:24:34 +0000 Subject: [PATCH 147/158] hwmon: (mr75203) enable polling for all VM channels Configure ip-polling register to enable polling for all voltage monitor channels. This enables reading the voltage values for all inputs other than just input 0. Fixes: 9d823351a337 ("hwmon: Add hardware monitoring driver for Moortec MR75203 PVT controller") Signed-off-by: Eliav Farber Reviewed-by: Andy Shevchenko Link: https://lore.kernel.org/r/20220908152449.35457-7-farbere@amazon.com Signed-off-by: Guenter Roeck --- drivers/hwmon/mr75203.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/drivers/hwmon/mr75203.c b/drivers/hwmon/mr75203.c index 812ae40e7d74..9259779cc2df 100644 --- a/drivers/hwmon/mr75203.c +++ b/drivers/hwmon/mr75203.c @@ -388,6 +388,19 @@ static int pvt_init(struct pvt_device *pvt) if (ret) return ret; + val = (BIT(pvt->c_num) - 1) | VM_CH_INIT | + IP_POLL << SDIF_ADDR_SFT | SDIF_WRN_W | SDIF_PROG; + ret = regmap_write(v_map, SDIF_W, val); + if (ret < 0) + return ret; + + ret = regmap_read_poll_timeout(v_map, SDIF_STAT, + val, !(val & SDIF_BUSY), + PVT_POLL_DELAY_US, + PVT_POLL_TIMEOUT_US); + if (ret) + return ret; + val = CFG1_VOL_MEAS_MODE | CFG1_PARALLEL_OUT | CFG1_14_BIT | IP_CFG << SDIF_ADDR_SFT | SDIF_WRN_W | SDIF_PROG; From faf59ec8c3c3708c64ff76b50e6f757c6b4a1054 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 7 Sep 2022 19:24:58 +0300 Subject: [PATCH 148/158] perf record: Fix synthesis failure warnings Some calls to synthesis functions set err < 0 but only warn about the failure and continue. However they do not set err back to zero, relying on subsequent code to do that. That changed with the introduction of option --synth. When --synth=no subsequent functions that set err back to zero are not called. Fix by setting err = 0 in those cases. Example: Before: $ perf record --no-bpf-event --synth=all -o /tmp/huh uname Couldn't synthesize bpf events. Linux [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.014 MB /tmp/huh (7 samples) ] $ perf record --no-bpf-event --synth=no -o /tmp/huh uname Couldn't synthesize bpf events. After: $ perf record --no-bpf-event --synth=no -o /tmp/huh uname Couldn't synthesize bpf events. Linux [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.014 MB /tmp/huh (7 samples) ] Fixes: 41b740b6e8a994e5 ("perf record: Add --synth option") Signed-off-by: Adrian Hunter Acked-by: Namhyung Kim Cc: Ian Rogers Cc: Jiri Olsa Cc: Namhyung Kim Link: https://lore.kernel.org/r/20220907162458.72817-1-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-record.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 09b68d76bbdc..f87ef43eb820 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -1906,14 +1906,18 @@ static int record__synthesize(struct record *rec, bool tail) err = perf_event__synthesize_bpf_events(session, process_synthesized_event, machine, opts); - if (err < 0) + if (err < 0) { pr_warning("Couldn't synthesize bpf events.\n"); + err = 0; + } if (rec->opts.synth & PERF_SYNTH_CGROUP) { err = perf_event__synthesize_cgroups(tool, process_synthesized_event, machine); - if (err < 0) + if (err < 0) { pr_warning("Couldn't synthesize cgroup events.\n"); + err = 0; + } } if (rec->opts.nr_threads_synthesize > 1) { From 20e0fbab16003ae23a9e86a64bcb93e3121587ca Mon Sep 17 00:00:00 2001 From: Sergey Matyukevich Date: Tue, 30 Aug 2022 18:53:04 +0300 Subject: [PATCH 149/158] perf: RISC-V: fix access beyond allocated array SBI firmware should report total number of firmware and hardware counters including unused ones or special ones. In this case the kernel doesn't need to make any assumptions about gaps in reported counters, e.g. excluded timer counter. That was fixed in OpenSBI v1.1 by commit 3f66465fb6bf ("lib: pmu: allow to use the highest available counter"). This kernel patch has no effect if SBI firmware behaves correctly. However it eliminates access beyond the allocated pmu_ctr_list if the kernel is used with OpenSBI older than v1.1. Fixes: e9991434596f ("RISC-V: Add perf platform driver based on SBI PMU extension") Signed-off-by: Sergey Matyukevich Reviewed-by: Atish Patra Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20220830155306.301714-2-geomatsi@gmail.com Signed-off-by: Palmer Dabbelt --- drivers/perf/riscv_pmu_sbi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/perf/riscv_pmu_sbi.c b/drivers/perf/riscv_pmu_sbi.c index 6f6681bbfd36..8de4ca2fef21 100644 --- a/drivers/perf/riscv_pmu_sbi.c +++ b/drivers/perf/riscv_pmu_sbi.c @@ -473,7 +473,7 @@ static int pmu_sbi_get_ctrinfo(int nctr) if (!pmu_ctr_list) return -ENOMEM; - for (i = 0; i <= nctr; i++) { + for (i = 0; i < nctr; i++) { ret = sbi_ecall(SBI_EXT_PMU, SBI_EXT_PMU_COUNTER_GET_INFO, i, 0, 0, 0, 0, 0); if (ret.error) /* The logical counter ids are not expected to be contiguous */ From 1efda38d6f9ba26ac88b359c6277f1172db03f1e Mon Sep 17 00:00:00 2001 From: "Christian A. Ehrhardt" Date: Wed, 7 Sep 2022 22:09:17 +0200 Subject: [PATCH 150/158] kprobes: Prohibit probes in gate area The system call gate area counts as kernel text but trying to install a kprobe in this area fails with an Oops later on. To fix this explicitly disallow the gate area for kprobes. Found by syzkaller with the following reproducer: perf_event_open$cgroup(&(0x7f00000001c0)={0x6, 0x80, 0x0, 0x0, 0x0, 0x0, 0x80ffff, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, @perf_config_ext={0x0, 0xffffffffff600000}}, 0xffffffffffffffff, 0x0, 0xffffffffffffffff, 0x0) Sample report: BUG: unable to handle page fault for address: fffffbfff3ac6000 PGD 6dfcb067 P4D 6dfcb067 PUD 6df8f067 PMD 6de4d067 PTE 0 Oops: 0000 [#1] PREEMPT SMP KASAN NOPTI CPU: 0 PID: 21978 Comm: syz-executor.2 Not tainted 6.0.0-rc3-00363-g7726d4c3e60b-dirty #6 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.15.0-1 04/01/2014 RIP: 0010:__insn_get_emulate_prefix arch/x86/lib/insn.c:91 [inline] RIP: 0010:insn_get_emulate_prefix arch/x86/lib/insn.c:106 [inline] RIP: 0010:insn_get_prefixes.part.0+0xa8/0x1110 arch/x86/lib/insn.c:134 Code: 49 be 00 00 00 00 00 fc ff df 48 8b 40 60 48 89 44 24 08 e9 81 00 00 00 e8 e5 4b 39 ff 4c 89 fa 4c 89 f9 48 c1 ea 03 83 e1 07 <42> 0f b6 14 32 38 ca 7f 08 84 d2 0f 85 06 10 00 00 48 89 d8 48 89 RSP: 0018:ffffc900088bf860 EFLAGS: 00010246 RAX: 0000000000040000 RBX: ffffffff9b9bebc0 RCX: 0000000000000000 RDX: 1ffffffff3ac6000 RSI: ffffc90002d82000 RDI: ffffc900088bf9e8 RBP: ffffffff9d630001 R08: 0000000000000000 R09: ffffc900088bf9e8 R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000000001 R13: ffffffff9d630000 R14: dffffc0000000000 R15: ffffffff9d630000 FS: 00007f63eef63640(0000) GS:ffff88806d000000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: fffffbfff3ac6000 CR3: 0000000029d90005 CR4: 0000000000770ef0 PKRU: 55555554 Call Trace: insn_get_prefixes arch/x86/lib/insn.c:131 [inline] insn_get_opcode arch/x86/lib/insn.c:272 [inline] insn_get_modrm+0x64a/0x7b0 arch/x86/lib/insn.c:343 insn_get_sib+0x29a/0x330 arch/x86/lib/insn.c:421 insn_get_displacement+0x350/0x6b0 arch/x86/lib/insn.c:464 insn_get_immediate arch/x86/lib/insn.c:632 [inline] insn_get_length arch/x86/lib/insn.c:707 [inline] insn_decode+0x43a/0x490 arch/x86/lib/insn.c:747 can_probe+0xfc/0x1d0 arch/x86/kernel/kprobes/core.c:282 arch_prepare_kprobe+0x79/0x1c0 arch/x86/kernel/kprobes/core.c:739 prepare_kprobe kernel/kprobes.c:1160 [inline] register_kprobe kernel/kprobes.c:1641 [inline] register_kprobe+0xb6e/0x1690 kernel/kprobes.c:1603 __register_trace_kprobe kernel/trace/trace_kprobe.c:509 [inline] __register_trace_kprobe+0x26a/0x2d0 kernel/trace/trace_kprobe.c:477 create_local_trace_kprobe+0x1f7/0x350 kernel/trace/trace_kprobe.c:1833 perf_kprobe_init+0x18c/0x280 kernel/trace/trace_event_perf.c:271 perf_kprobe_event_init+0xf8/0x1c0 kernel/events/core.c:9888 perf_try_init_event+0x12d/0x570 kernel/events/core.c:11261 perf_init_event kernel/events/core.c:11325 [inline] perf_event_alloc.part.0+0xf7f/0x36a0 kernel/events/core.c:11619 perf_event_alloc kernel/events/core.c:12059 [inline] __do_sys_perf_event_open+0x4a8/0x2a00 kernel/events/core.c:12157 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x38/0x90 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd RIP: 0033:0x7f63ef7efaed Code: 02 b8 ff ff ff ff c3 66 0f 1f 44 00 00 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b0 ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007f63eef63028 EFLAGS: 00000246 ORIG_RAX: 000000000000012a RAX: ffffffffffffffda RBX: 00007f63ef90ff80 RCX: 00007f63ef7efaed RDX: 0000000000000000 RSI: ffffffffffffffff RDI: 00000000200001c0 RBP: 00007f63ef86019c R08: 0000000000000000 R09: 0000000000000000 R10: ffffffffffffffff R11: 0000000000000246 R12: 0000000000000000 R13: 0000000000000002 R14: 00007f63ef90ff80 R15: 00007f63eef43000 Modules linked in: CR2: fffffbfff3ac6000 ---[ end trace 0000000000000000 ]--- RIP: 0010:__insn_get_emulate_prefix arch/x86/lib/insn.c:91 [inline] RIP: 0010:insn_get_emulate_prefix arch/x86/lib/insn.c:106 [inline] RIP: 0010:insn_get_prefixes.part.0+0xa8/0x1110 arch/x86/lib/insn.c:134 Code: 49 be 00 00 00 00 00 fc ff df 48 8b 40 60 48 89 44 24 08 e9 81 00 00 00 e8 e5 4b 39 ff 4c 89 fa 4c 89 f9 48 c1 ea 03 83 e1 07 <42> 0f b6 14 32 38 ca 7f 08 84 d2 0f 85 06 10 00 00 48 89 d8 48 89 RSP: 0018:ffffc900088bf860 EFLAGS: 00010246 RAX: 0000000000040000 RBX: ffffffff9b9bebc0 RCX: 0000000000000000 RDX: 1ffffffff3ac6000 RSI: ffffc90002d82000 RDI: ffffc900088bf9e8 RBP: ffffffff9d630001 R08: 0000000000000000 R09: ffffc900088bf9e8 R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000000001 R13: ffffffff9d630000 R14: dffffc0000000000 R15: ffffffff9d630000 FS: 00007f63eef63640(0000) GS:ffff88806d000000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: fffffbfff3ac6000 CR3: 0000000029d90005 CR4: 0000000000770ef0 PKRU: 55555554 ================================================================== Link: https://lkml.kernel.org/r/20220907200917.654103-1-lk@c--e.de cc: "Naveen N. Rao" cc: Anil S Keshavamurthy cc: "David S. Miller" Cc: stable@vger.kernel.org Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Acked-by: Masami Hiramatsu (Google) Signed-off-by: Christian A. Ehrhardt Signed-off-by: Steven Rostedt (Google) --- kernel/kprobes.c | 1 + 1 file changed, 1 insertion(+) diff --git a/kernel/kprobes.c b/kernel/kprobes.c index 08350e35aba2..ca9d834d0b84 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -1562,6 +1562,7 @@ static int check_kprobe_address_safe(struct kprobe *p, /* Ensure it is not in reserved area nor out of text */ if (!(core_kernel_text((unsigned long) p->addr) || is_module_text_address((unsigned long) p->addr)) || + in_gate_area_no_mm((unsigned long) p->addr) || within_kprobe_blacklist((unsigned long) p->addr) || jump_label_text_reserved(p->addr, p->addr) || static_call_text_reserved(p->addr, p->addr) || From 47311db8e8f33011d90dee76b39c8886120cdda4 Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Fri, 26 Aug 2022 17:44:17 -0700 Subject: [PATCH 151/158] tracefs: Only clobber mode/uid/gid on remount if asked Users may have explicitly configured their tracefs permissions; we shouldn't overwrite those just because a second mount appeared. Only clobber if the options were provided at mount time. Note: the previous behavior was especially surprising in the presence of automounted /sys/kernel/debug/tracing/. Existing behavior: ## Pre-existing status: tracefs is 0755. # stat -c '%A' /sys/kernel/tracing/ drwxr-xr-x ## (Re)trigger the automount. # umount /sys/kernel/debug/tracing # stat -c '%A' /sys/kernel/debug/tracing/. drwx------ ## Unexpected: the automount changed mode for other mount instances. # stat -c '%A' /sys/kernel/tracing/ drwx------ New behavior (after this change): ## Pre-existing status: tracefs is 0755. # stat -c '%A' /sys/kernel/tracing/ drwxr-xr-x ## (Re)trigger the automount. # umount /sys/kernel/debug/tracing # stat -c '%A' /sys/kernel/debug/tracing/. drwxr-xr-x ## Expected: the automount does not change other mount instances. # stat -c '%A' /sys/kernel/tracing/ drwxr-xr-x Link: https://lkml.kernel.org/r/20220826174353.2.Iab6e5ea57963d6deca5311b27fb7226790d44406@changeid Cc: stable@vger.kernel.org Fixes: 4282d60689d4f ("tracefs: Add new tracefs file system") Signed-off-by: Brian Norris Signed-off-by: Steven Rostedt (Google) --- fs/tracefs/inode.c | 31 +++++++++++++++++++++++-------- 1 file changed, 23 insertions(+), 8 deletions(-) diff --git a/fs/tracefs/inode.c b/fs/tracefs/inode.c index 81d26abf486f..da85b3979195 100644 --- a/fs/tracefs/inode.c +++ b/fs/tracefs/inode.c @@ -141,6 +141,8 @@ struct tracefs_mount_opts { kuid_t uid; kgid_t gid; umode_t mode; + /* Opt_* bitfield. */ + unsigned int opts; }; enum { @@ -241,6 +243,7 @@ static int tracefs_parse_options(char *data, struct tracefs_mount_opts *opts) kgid_t gid; char *p; + opts->opts = 0; opts->mode = TRACEFS_DEFAULT_MODE; while ((p = strsep(&data, ",")) != NULL) { @@ -275,24 +278,36 @@ static int tracefs_parse_options(char *data, struct tracefs_mount_opts *opts) * but traditionally tracefs has ignored all mount options */ } + + opts->opts |= BIT(token); } return 0; } -static int tracefs_apply_options(struct super_block *sb) +static int tracefs_apply_options(struct super_block *sb, bool remount) { struct tracefs_fs_info *fsi = sb->s_fs_info; struct inode *inode = d_inode(sb->s_root); struct tracefs_mount_opts *opts = &fsi->mount_opts; - inode->i_mode &= ~S_IALLUGO; - inode->i_mode |= opts->mode; + /* + * On remount, only reset mode/uid/gid if they were provided as mount + * options. + */ - inode->i_uid = opts->uid; + if (!remount || opts->opts & BIT(Opt_mode)) { + inode->i_mode &= ~S_IALLUGO; + inode->i_mode |= opts->mode; + } - /* Set all the group ids to the mount option */ - set_gid(sb->s_root, opts->gid); + if (!remount || opts->opts & BIT(Opt_uid)) + inode->i_uid = opts->uid; + + if (!remount || opts->opts & BIT(Opt_gid)) { + /* Set all the group ids to the mount option */ + set_gid(sb->s_root, opts->gid); + } return 0; } @@ -307,7 +322,7 @@ static int tracefs_remount(struct super_block *sb, int *flags, char *data) if (err) goto fail; - tracefs_apply_options(sb); + tracefs_apply_options(sb, true); fail: return err; @@ -359,7 +374,7 @@ static int trace_fill_super(struct super_block *sb, void *data, int silent) sb->s_op = &tracefs_super_operations; - tracefs_apply_options(sb); + tracefs_apply_options(sb, false); return 0; From 745ed37277c5a7202180aa276c87db1362c90fe5 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 8 Sep 2022 17:26:59 -0600 Subject: [PATCH 152/158] block: add missing request flags to debugfs code We're missing TIMED_OUT and RESV. Particularly the former is handy for debugging, let's get them added. Reviewed-by: Bart Van Assche Signed-off-by: Jens Axboe --- block/blk-mq-debugfs.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/block/blk-mq-debugfs.c b/block/blk-mq-debugfs.c index 8559cea7f300..dee789f2f98f 100644 --- a/block/blk-mq-debugfs.c +++ b/block/blk-mq-debugfs.c @@ -283,7 +283,9 @@ static const char *const rqf_name[] = { RQF_NAME(SPECIAL_PAYLOAD), RQF_NAME(ZONE_WRITE_LOCKED), RQF_NAME(MQ_POLL_SLEPT), + RQF_NAME(TIMED_OUT), RQF_NAME(ELV), + RQF_NAME(RESV), }; #undef RQF_NAME From 4d9cb92ca41dd8e905a4569ceba4716c2f39c75a Mon Sep 17 00:00:00 2001 From: Pavel Begunkov Date: Fri, 9 Sep 2022 12:11:49 +0100 Subject: [PATCH 153/158] io_uring/rw: fix short rw error handling We have a couple of problems, first reports of unexpected link breakage for reads when cqe->res indicates that the IO was done in full. The reason here is partial IO with retries. TL;DR; we compare the result in __io_complete_rw_common() against req->cqe.res, but req->cqe.res doesn't store the full length but rather the length left to be done. So, when we pass the full corrected result via kiocb_done() -> __io_complete_rw_common(), it fails. The second problem is that we don't try to correct res in io_complete_rw(), which, for instance, might be a problem for O_DIRECT but when a prefix of data was cached in the page cache. We also definitely don't want to pass a corrected result into io_rw_done(). The fix here is to leave __io_complete_rw_common() alone, always pass not corrected result into it and fix it up as the last step just before actually finishing the I/O. Cc: stable@vger.kernel.org Signed-off-by: Pavel Begunkov Link: https://github.com/axboe/liburing/issues/643 Reported-by: Beld Zhang Signed-off-by: Jens Axboe --- io_uring/rw.c | 30 ++++++++++++++++++------------ 1 file changed, 18 insertions(+), 12 deletions(-) diff --git a/io_uring/rw.c b/io_uring/rw.c index 1babd77da79c..1e18a44adcf5 100644 --- a/io_uring/rw.c +++ b/io_uring/rw.c @@ -206,6 +206,20 @@ static bool __io_complete_rw_common(struct io_kiocb *req, long res) return false; } +static inline unsigned io_fixup_rw_res(struct io_kiocb *req, unsigned res) +{ + struct io_async_rw *io = req->async_data; + + /* add previously done IO, if any */ + if (req_has_async_data(req) && io->bytes_done > 0) { + if (res < 0) + res = io->bytes_done; + else + res += io->bytes_done; + } + return res; +} + static void io_complete_rw(struct kiocb *kiocb, long res) { struct io_rw *rw = container_of(kiocb, struct io_rw, kiocb); @@ -213,7 +227,7 @@ static void io_complete_rw(struct kiocb *kiocb, long res) if (__io_complete_rw_common(req, res)) return; - io_req_set_res(req, res, 0); + io_req_set_res(req, io_fixup_rw_res(req, res), 0); req->io_task_work.func = io_req_task_complete; io_req_task_work_add(req); } @@ -240,22 +254,14 @@ static void io_complete_rw_iopoll(struct kiocb *kiocb, long res) static int kiocb_done(struct io_kiocb *req, ssize_t ret, unsigned int issue_flags) { - struct io_async_rw *io = req->async_data; struct io_rw *rw = io_kiocb_to_cmd(req, struct io_rw); - - /* add previously done IO, if any */ - if (req_has_async_data(req) && io->bytes_done > 0) { - if (ret < 0) - ret = io->bytes_done; - else - ret += io->bytes_done; - } + unsigned final_ret = io_fixup_rw_res(req, ret); if (req->flags & REQ_F_CUR_POS) req->file->f_pos = rw->kiocb.ki_pos; if (ret >= 0 && (rw->kiocb.ki_complete == io_complete_rw)) { if (!__io_complete_rw_common(req, ret)) { - io_req_set_res(req, req->cqe.res, + io_req_set_res(req, final_ret, io_put_kbuf(req, issue_flags)); return IOU_OK; } @@ -268,7 +274,7 @@ static int kiocb_done(struct io_kiocb *req, ssize_t ret, if (io_resubmit_prep(req)) io_req_task_queue_reissue(req); else - io_req_task_queue_fail(req, ret); + io_req_task_queue_fail(req, final_ret); } return IOU_ISSUE_SKIP_COMPLETE; } From 727488e305b223ca69205ca5a3b99ace21bbbf5f Mon Sep 17 00:00:00 2001 From: Alexander Sverdlin Date: Fri, 9 Sep 2022 11:30:42 +0200 Subject: [PATCH 154/158] mips: Select SPARSEMEM_EXTREME Commit c46173183657 ("MIPS: Add NUMA support for Loongson-3") has increased .bss size of the Octeon kernel from 16k to 16M. Providing the conditions for SPARSEMEM_EXTREME avoids the waste of memory. Thomas has tested the loogsoon64 kernel, where .bss is being reduced by this patch from 16.5M to 515k. Cc: Thomas Bogendoerfer Signed-off-by: Alexander Sverdlin Signed-off-by: Thomas Bogendoerfer --- arch/mips/Kconfig | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index ec21f8999249..25dd4c5a8ef5 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -2669,7 +2669,6 @@ config ARCH_FLATMEM_ENABLE config ARCH_SPARSEMEM_ENABLE bool - select SPARSEMEM_STATIC if !SGI_IP27 config NUMA bool "NUMA Support" From 88700d1396bae72d9a4c23a48bbd98c1c2f53f3d Mon Sep 17 00:00:00 2001 From: Eugene Shalygin Date: Fri, 9 Sep 2022 17:56:53 +0200 Subject: [PATCH 155/158] hwmon: (asus-ec-sensors) autoload module via DMI data Replace autoloading data based on the ACPI EC device with the DMI records for motherboards models. The ACPI method created a bug that when this driver returns error from the probe function because of the unsupported motherboard model, the ACPI subsystem concludes that the EC device does not work properly. Fixes: 5cd29012028d ("hwmon: (asus-ec-sensors) introduce ec_board_info struct for board data") Bug: https://bugzilla.kernel.org/show_bug.cgi?id=216412 Bug: https://bugzilla.redhat.com/show_bug.cgi?id=2121844 Signed-off-by: Eugene Shalygin Link: https://lore.kernel.org/r/20220909155654.123398-2-eugene.shalygin@gmail.com Cc: stable@vger.kernel.org Signed-off-by: Guenter Roeck --- drivers/hwmon/asus-ec-sensors.c | 408 +++++++++++++++++--------------- 1 file changed, 222 insertions(+), 186 deletions(-) diff --git a/drivers/hwmon/asus-ec-sensors.c b/drivers/hwmon/asus-ec-sensors.c index 61a4684fc020..81e688975c6a 100644 --- a/drivers/hwmon/asus-ec-sensors.c +++ b/drivers/hwmon/asus-ec-sensors.c @@ -266,9 +266,7 @@ static const struct ec_sensor_info sensors_family_intel_600[] = { #define SENSOR_SET_WATER_BLOCK \ (SENSOR_TEMP_WATER_BLOCK_IN | SENSOR_TEMP_WATER_BLOCK_OUT) - struct ec_board_info { - const char *board_names[MAX_IDENTICAL_BOARD_VARIATIONS]; unsigned long sensors; /* * Defines which mutex to use for guarding access to the state and the @@ -281,152 +279,194 @@ struct ec_board_info { enum board_family family; }; -static const struct ec_board_info board_info[] = { - { - .board_names = {"PRIME X470-PRO"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | - SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM | - SENSOR_FAN_CPU_OPT | - SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE, - .mutex_path = ACPI_GLOBAL_LOCK_PSEUDO_PATH, - .family = family_amd_400_series, - }, - { - .board_names = {"PRIME X570-PRO"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | SENSOR_TEMP_VRM | - SENSOR_TEMP_T_SENSOR | SENSOR_FAN_CHIPSET, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = {"ProArt X570-CREATOR WIFI"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | SENSOR_TEMP_VRM | - SENSOR_TEMP_T_SENSOR | SENSOR_FAN_CPU_OPT | - SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE, - }, - { - .board_names = {"Pro WS X570-ACE"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | SENSOR_TEMP_VRM | - SENSOR_TEMP_T_SENSOR | SENSOR_FAN_CHIPSET | - SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = {"ROG CROSSHAIR VIII DARK HERO"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | - SENSOR_TEMP_T_SENSOR | - SENSOR_TEMP_VRM | SENSOR_SET_TEMP_WATER | - SENSOR_FAN_CPU_OPT | SENSOR_FAN_WATER_FLOW | - SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = { - "ROG CROSSHAIR VIII FORMULA", - "ROG CROSSHAIR VIII HERO", - "ROG CROSSHAIR VIII HERO (WI-FI)", - }, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | - SENSOR_TEMP_T_SENSOR | - SENSOR_TEMP_VRM | SENSOR_SET_TEMP_WATER | - SENSOR_FAN_CPU_OPT | SENSOR_FAN_CHIPSET | - SENSOR_FAN_WATER_FLOW | SENSOR_CURR_CPU | - SENSOR_IN_CPU_CORE, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = { - "ROG MAXIMUS XI HERO", - "ROG MAXIMUS XI HERO (WI-FI)", - }, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | - SENSOR_TEMP_T_SENSOR | - SENSOR_TEMP_VRM | SENSOR_SET_TEMP_WATER | - SENSOR_FAN_CPU_OPT | SENSOR_FAN_WATER_FLOW, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_intel_300_series, - }, - { - .board_names = {"ROG CROSSHAIR VIII IMPACT"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | - SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM | - SENSOR_FAN_CHIPSET | SENSOR_CURR_CPU | - SENSOR_IN_CPU_CORE, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = {"ROG STRIX B550-E GAMING"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | - SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM | - SENSOR_FAN_CPU_OPT, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = {"ROG STRIX B550-I GAMING"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | - SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM | - SENSOR_FAN_VRM_HS | SENSOR_CURR_CPU | - SENSOR_IN_CPU_CORE, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = {"ROG STRIX X570-E GAMING"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | - SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM | - SENSOR_FAN_CHIPSET | SENSOR_CURR_CPU | - SENSOR_IN_CPU_CORE, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = {"ROG STRIX X570-E GAMING WIFI II"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | - SENSOR_TEMP_T_SENSOR | SENSOR_CURR_CPU | - SENSOR_IN_CPU_CORE, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = {"ROG STRIX X570-F GAMING"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | - SENSOR_TEMP_T_SENSOR | SENSOR_FAN_CHIPSET, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = {"ROG STRIX X570-I GAMING"}, - .sensors = SENSOR_TEMP_CHIPSET | SENSOR_TEMP_VRM | - SENSOR_TEMP_T_SENSOR | - SENSOR_FAN_VRM_HS | SENSOR_FAN_CHIPSET | - SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE, - .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, - .family = family_amd_500_series, - }, - { - .board_names = {"ROG STRIX Z690-A GAMING WIFI D4"}, - .sensors = SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM, - .mutex_path = ASUS_HW_ACCESS_MUTEX_RMTW_ASMX, - .family = family_intel_600_series, - }, - { - .board_names = {"ROG ZENITH II EXTREME"}, - .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | SENSOR_TEMP_T_SENSOR | - SENSOR_TEMP_VRM | SENSOR_SET_TEMP_WATER | - SENSOR_FAN_CPU_OPT | SENSOR_FAN_CHIPSET | SENSOR_FAN_VRM_HS | - SENSOR_FAN_WATER_FLOW | SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE | - SENSOR_SET_WATER_BLOCK | - SENSOR_TEMP_T_SENSOR_2 | SENSOR_TEMP_SENSOR_EXTRA_1 | - SENSOR_TEMP_SENSOR_EXTRA_2 | SENSOR_TEMP_SENSOR_EXTRA_3, - .mutex_path = ASUS_HW_ACCESS_MUTEX_SB_PCI0_SBRG_SIO1_MUT0, - .family = family_amd_500_series, - }, - {} +static const struct ec_board_info board_info_prime_x470_pro = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | + SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM | + SENSOR_FAN_CPU_OPT | + SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE, + .mutex_path = ACPI_GLOBAL_LOCK_PSEUDO_PATH, + .family = family_amd_400_series, +}; + +static const struct ec_board_info board_info_prime_x570_pro = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | SENSOR_TEMP_VRM | + SENSOR_TEMP_T_SENSOR | SENSOR_FAN_CHIPSET, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_pro_art_x570_creator_wifi = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | SENSOR_TEMP_VRM | + SENSOR_TEMP_T_SENSOR | SENSOR_FAN_CPU_OPT | + SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_pro_ws_x570_ace = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | SENSOR_TEMP_VRM | + SENSOR_TEMP_T_SENSOR | SENSOR_FAN_CHIPSET | + SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_crosshair_viii_dark_hero = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | + SENSOR_TEMP_T_SENSOR | + SENSOR_TEMP_VRM | SENSOR_SET_TEMP_WATER | + SENSOR_FAN_CPU_OPT | SENSOR_FAN_WATER_FLOW | + SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_crosshair_viii_hero = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | + SENSOR_TEMP_T_SENSOR | + SENSOR_TEMP_VRM | SENSOR_SET_TEMP_WATER | + SENSOR_FAN_CPU_OPT | SENSOR_FAN_CHIPSET | + SENSOR_FAN_WATER_FLOW | SENSOR_CURR_CPU | + SENSOR_IN_CPU_CORE, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_maximus_xi_hero = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | + SENSOR_TEMP_T_SENSOR | + SENSOR_TEMP_VRM | SENSOR_SET_TEMP_WATER | + SENSOR_FAN_CPU_OPT | SENSOR_FAN_WATER_FLOW, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_intel_300_series, +}; + +static const struct ec_board_info board_info_crosshair_viii_impact = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | + SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM | + SENSOR_FAN_CHIPSET | SENSOR_CURR_CPU | + SENSOR_IN_CPU_CORE, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_strix_b550_e_gaming = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | + SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM | + SENSOR_FAN_CPU_OPT, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_strix_b550_i_gaming = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | + SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM | + SENSOR_FAN_VRM_HS | SENSOR_CURR_CPU | + SENSOR_IN_CPU_CORE, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_strix_x570_e_gaming = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | + SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM | + SENSOR_FAN_CHIPSET | SENSOR_CURR_CPU | + SENSOR_IN_CPU_CORE, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_strix_x570_e_gaming_wifi_ii = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | + SENSOR_TEMP_T_SENSOR | SENSOR_CURR_CPU | + SENSOR_IN_CPU_CORE, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_strix_x570_f_gaming = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | + SENSOR_TEMP_T_SENSOR | SENSOR_FAN_CHIPSET, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_strix_x570_i_gaming = { + .sensors = SENSOR_TEMP_CHIPSET | SENSOR_TEMP_VRM | + SENSOR_TEMP_T_SENSOR | + SENSOR_FAN_VRM_HS | SENSOR_FAN_CHIPSET | + SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE, + .mutex_path = ASUS_HW_ACCESS_MUTEX_ASMX, + .family = family_amd_500_series, +}; + +static const struct ec_board_info board_info_strix_z690_a_gaming_wifi_d4 = { + .sensors = SENSOR_TEMP_T_SENSOR | SENSOR_TEMP_VRM, + .mutex_path = ASUS_HW_ACCESS_MUTEX_RMTW_ASMX, + .family = family_intel_600_series, +}; + +static const struct ec_board_info board_info_zenith_ii_extreme = { + .sensors = SENSOR_SET_TEMP_CHIPSET_CPU_MB | SENSOR_TEMP_T_SENSOR | + SENSOR_TEMP_VRM | SENSOR_SET_TEMP_WATER | + SENSOR_FAN_CPU_OPT | SENSOR_FAN_CHIPSET | SENSOR_FAN_VRM_HS | + SENSOR_FAN_WATER_FLOW | SENSOR_CURR_CPU | SENSOR_IN_CPU_CORE | + SENSOR_SET_WATER_BLOCK | + SENSOR_TEMP_T_SENSOR_2 | SENSOR_TEMP_SENSOR_EXTRA_1 | + SENSOR_TEMP_SENSOR_EXTRA_2 | SENSOR_TEMP_SENSOR_EXTRA_3, + .mutex_path = ASUS_HW_ACCESS_MUTEX_SB_PCI0_SBRG_SIO1_MUT0, + .family = family_amd_500_series, +}; + +#define DMI_EXACT_MATCH_ASUS_BOARD_NAME(name, board_info) \ + { \ + .matches = { \ + DMI_EXACT_MATCH(DMI_BOARD_VENDOR, \ + "ASUSTeK COMPUTER INC."), \ + DMI_EXACT_MATCH(DMI_BOARD_NAME, name), \ + }, \ + .driver_data = (void *)board_info, \ + } + +static const struct dmi_system_id dmi_table[] = { + DMI_EXACT_MATCH_ASUS_BOARD_NAME("PRIME X470-PRO", + &board_info_prime_x470_pro), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("PRIME X570-PRO", + &board_info_prime_x570_pro), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ProArt X570-CREATOR WIFI", + &board_info_pro_art_x570_creator_wifi), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("Pro WS X570-ACE", + &board_info_pro_ws_x570_ace), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG CROSSHAIR VIII DARK HERO", + &board_info_crosshair_viii_dark_hero), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG CROSSHAIR VIII FORMULA", + &board_info_crosshair_viii_hero), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG CROSSHAIR VIII HERO", + &board_info_crosshair_viii_hero), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG CROSSHAIR VIII HERO (WI-FI)", + &board_info_crosshair_viii_hero), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG MAXIMUS XI HERO", + &board_info_maximus_xi_hero), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG MAXIMUS XI HERO (WI-FI)", + &board_info_maximus_xi_hero), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG CROSSHAIR VIII IMPACT", + &board_info_crosshair_viii_impact), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG STRIX B550-E GAMING", + &board_info_strix_b550_e_gaming), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG STRIX B550-I GAMING", + &board_info_strix_b550_i_gaming), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG STRIX X570-E GAMING", + &board_info_strix_x570_e_gaming), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG STRIX X570-E GAMING WIFI II", + &board_info_strix_x570_e_gaming_wifi_ii), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG STRIX X570-F GAMING", + &board_info_strix_x570_f_gaming), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG STRIX X570-I GAMING", + &board_info_strix_x570_i_gaming), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG STRIX Z690-A GAMING WIFI D4", + &board_info_strix_z690_a_gaming_wifi_d4), + DMI_EXACT_MATCH_ASUS_BOARD_NAME("ROG ZENITH II EXTREME", + &board_info_zenith_ii_extreme), + {}, }; struct ec_sensor { @@ -537,12 +577,12 @@ static int find_ec_sensor_index(const struct ec_sensors_data *ec, return -ENOENT; } -static int __init bank_compare(const void *a, const void *b) +static int bank_compare(const void *a, const void *b) { return *((const s8 *)a) - *((const s8 *)b); } -static void __init setup_sensor_data(struct ec_sensors_data *ec) +static void setup_sensor_data(struct ec_sensors_data *ec) { struct ec_sensor *s = ec->sensors; bool bank_found; @@ -574,7 +614,7 @@ static void __init setup_sensor_data(struct ec_sensors_data *ec) sort(ec->banks, ec->nr_banks, 1, bank_compare, NULL); } -static void __init fill_ec_registers(struct ec_sensors_data *ec) +static void fill_ec_registers(struct ec_sensors_data *ec) { const struct ec_sensor_info *si; unsigned int i, j, register_idx = 0; @@ -589,7 +629,7 @@ static void __init fill_ec_registers(struct ec_sensors_data *ec) } } -static int __init setup_lock_data(struct device *dev) +static int setup_lock_data(struct device *dev) { const char *mutex_path; int status; @@ -812,7 +852,7 @@ static umode_t asus_ec_hwmon_is_visible(const void *drvdata, return find_ec_sensor_index(state, type, channel) >= 0 ? S_IRUGO : 0; } -static int __init +static int asus_ec_hwmon_add_chan_info(struct hwmon_channel_info *asus_ec_hwmon_chan, struct device *dev, int num, enum hwmon_sensor_types type, u32 config) @@ -841,27 +881,15 @@ static struct hwmon_chip_info asus_ec_chip_info = { .ops = &asus_ec_hwmon_ops, }; -static const struct ec_board_info * __init get_board_info(void) +static const struct ec_board_info *get_board_info(void) { - const char *dmi_board_vendor = dmi_get_system_info(DMI_BOARD_VENDOR); - const char *dmi_board_name = dmi_get_system_info(DMI_BOARD_NAME); - const struct ec_board_info *board; + const struct dmi_system_id *dmi_entry; - if (!dmi_board_vendor || !dmi_board_name || - strcasecmp(dmi_board_vendor, "ASUSTeK COMPUTER INC.")) - return NULL; - - for (board = board_info; board->sensors; board++) { - if (match_string(board->board_names, - MAX_IDENTICAL_BOARD_VARIATIONS, - dmi_board_name) >= 0) - return board; - } - - return NULL; + dmi_entry = dmi_first_match(dmi_table); + return dmi_entry ? dmi_entry->driver_data : NULL; } -static int __init asus_ec_probe(struct platform_device *pdev) +static int asus_ec_probe(struct platform_device *pdev) { const struct hwmon_channel_info **ptr_asus_ec_ci; int nr_count[hwmon_max] = { 0 }, nr_types = 0; @@ -970,29 +998,37 @@ static int __init asus_ec_probe(struct platform_device *pdev) return PTR_ERR_OR_ZERO(hwdev); } - -static const struct acpi_device_id acpi_ec_ids[] = { - /* Embedded Controller Device */ - { "PNP0C09", 0 }, - {} -}; +MODULE_DEVICE_TABLE(dmi, dmi_table); static struct platform_driver asus_ec_sensors_platform_driver = { .driver = { .name = "asus-ec-sensors", - .acpi_match_table = acpi_ec_ids, }, + .probe = asus_ec_probe, }; -MODULE_DEVICE_TABLE(acpi, acpi_ec_ids); -/* - * we use module_platform_driver_probe() rather than module_platform_driver() - * because the probe function (and its dependants) are marked with __init, which - * means we can't put it into the .probe member of the platform_driver struct - * above, and we can't mark the asus_ec_sensors_platform_driver object as __init - * because the object is referenced from the module exit code. - */ -module_platform_driver_probe(asus_ec_sensors_platform_driver, asus_ec_probe); +static struct platform_device *asus_ec_sensors_platform_device; + +static int __init asus_ec_init(void) +{ + asus_ec_sensors_platform_device = + platform_create_bundle(&asus_ec_sensors_platform_driver, + asus_ec_probe, NULL, 0, NULL, 0); + + if (IS_ERR(asus_ec_sensors_platform_device)) + return PTR_ERR(asus_ec_sensors_platform_device); + + return 0; +} + +static void __exit asus_ec_exit(void) +{ + platform_device_unregister(asus_ec_sensors_platform_device); + platform_driver_unregister(&asus_ec_sensors_platform_driver); +} + +module_init(asus_ec_init); +module_exit(asus_ec_exit); module_param_named(mutex_path, mutex_path_override, charp, 0); MODULE_PARM_DESC(mutex_path, From 3fe3fd5f30720b4afd3345cc186808125e7f5848 Mon Sep 17 00:00:00 2001 From: Joey Gouly Date: Fri, 9 Sep 2022 13:43:11 +0100 Subject: [PATCH 156/158] arm64: mm: fix resume for 52-bit enabled builds __cpu_setup() was changed to take the actual number of VA bits in x0, however the resume path was not updated at the same time. Load `vabits_actual` in the resume path, to ensure that the correct number of VA bits is used. This fixes booting v6.0-rc kernels on my Juno. Signed-off-by: Joey Gouly Fixes: 0aaa68532e9d ("arm64: mm: fix booting with 52-bit address space") Cc: Catalin Marinas Cc: Will Deacon Cc: Ard Biesheuvel Acked-by: Ard Biesheuvel Link: https://lore.kernel.org/r/20220909124311.38489-1-joey.gouly@arm.com Signed-off-by: Will Deacon --- arch/arm64/kernel/sleep.S | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/arm64/kernel/sleep.S b/arch/arm64/kernel/sleep.S index 617f78ad43a1..97c9de57725d 100644 --- a/arch/arm64/kernel/sleep.S +++ b/arch/arm64/kernel/sleep.S @@ -101,6 +101,9 @@ SYM_FUNC_END(__cpu_suspend_enter) SYM_CODE_START(cpu_resume) bl init_kernel_el bl finalise_el2 +#if VA_BITS > 48 + ldr_l x0, vabits_actual +#endif bl __cpu_setup /* enable the MMU early - so we can access sleep_save_stash by va */ adrp x1, swapper_pg_dir From 9cd4f1434479f1ac25c440c421fbf52069079914 Mon Sep 17 00:00:00 2001 From: Lu Baolu Date: Sun, 11 Sep 2022 11:18:45 +0800 Subject: [PATCH 157/158] iommu/vt-d: Fix possible recursive locking in intel_iommu_init() The global rwsem dmar_global_lock was introduced by commit 3a5670e8ac932 ("iommu/vt-d: Introduce a rwsem to protect global data structures"). It is used to protect DMAR related global data from DMAR hotplug operations. The dmar_global_lock used in the intel_iommu_init() might cause recursive locking issue, for example, intel_iommu_get_resv_regions() is taking the dmar_global_lock from within a section where intel_iommu_init() already holds it via probe_acpi_namespace_devices(). Using dmar_global_lock in intel_iommu_init() could be relaxed since it is unlikely that any IO board must be hot added before the IOMMU subsystem is initialized. This eliminates the possible recursive locking issue by moving down DMAR hotplug support after the IOMMU is initialized and removing the uses of dmar_global_lock in intel_iommu_init(). Fixes: d5692d4af08cd ("iommu/vt-d: Fix suspicious RCU usage in probe_acpi_namespace_devices()") Reported-by: Robin Murphy Signed-off-by: Lu Baolu Reviewed-by: Kevin Tian Link: https://lore.kernel.org/r/894db0ccae854b35c73814485569b634237b5538.1657034828.git.robin.murphy@arm.com Link: https://lore.kernel.org/r/20220718235325.3952426-1-baolu.lu@linux.intel.com Signed-off-by: Joerg Roedel --- drivers/iommu/intel/dmar.c | 7 +++++++ drivers/iommu/intel/iommu.c | 27 ++------------------------- include/linux/dmar.h | 4 +++- 3 files changed, 12 insertions(+), 26 deletions(-) diff --git a/drivers/iommu/intel/dmar.c b/drivers/iommu/intel/dmar.c index 5a8f780e7ffd..497c912ad9e1 100644 --- a/drivers/iommu/intel/dmar.c +++ b/drivers/iommu/intel/dmar.c @@ -2349,6 +2349,13 @@ static int dmar_device_hotplug(acpi_handle handle, bool insert) if (!dmar_in_use()) return 0; + /* + * It's unlikely that any I/O board is hot added before the IOMMU + * subsystem is initialized. + */ + if (IS_ENABLED(CONFIG_INTEL_IOMMU) && !intel_iommu_enabled) + return -EOPNOTSUPP; + if (dmar_detect_dsm(handle, DMAR_DSM_FUNC_DRHD)) { tmp = handle; } else { diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index 64d30895a4c8..1f2cd43cf9bc 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -3019,13 +3019,7 @@ static int __init init_dmars(void) #ifdef CONFIG_INTEL_IOMMU_SVM if (pasid_supported(iommu) && ecap_prs(iommu->ecap)) { - /* - * Call dmar_alloc_hwirq() with dmar_global_lock held, - * could cause possible lock race condition. - */ - up_write(&dmar_global_lock); ret = intel_svm_enable_prq(iommu); - down_write(&dmar_global_lock); if (ret) goto free_iommu; } @@ -3938,7 +3932,6 @@ int __init intel_iommu_init(void) force_on = (!intel_iommu_tboot_noforce && tboot_force_iommu()) || platform_optin_force_iommu(); - down_write(&dmar_global_lock); if (dmar_table_init()) { if (force_on) panic("tboot: Failed to initialize DMAR table\n"); @@ -3951,16 +3944,6 @@ int __init intel_iommu_init(void) goto out_free_dmar; } - up_write(&dmar_global_lock); - - /* - * The bus notifier takes the dmar_global_lock, so lockdep will - * complain later when we register it under the lock. - */ - dmar_register_bus_notifier(); - - down_write(&dmar_global_lock); - if (!no_iommu) intel_iommu_debugfs_init(); @@ -4005,11 +3988,9 @@ int __init intel_iommu_init(void) pr_err("Initialization failed\n"); goto out_free_dmar; } - up_write(&dmar_global_lock); init_iommu_pm_ops(); - down_read(&dmar_global_lock); for_each_active_iommu(iommu, drhd) { /* * The flush queue implementation does not perform @@ -4027,13 +4008,11 @@ int __init intel_iommu_init(void) "%s", iommu->name); iommu_device_register(&iommu->iommu, &intel_iommu_ops, NULL); } - up_read(&dmar_global_lock); bus_set_iommu(&pci_bus_type, &intel_iommu_ops); if (si_domain && !hw_pass_through) register_memory_notifier(&intel_iommu_memory_nb); - down_read(&dmar_global_lock); if (probe_acpi_namespace_devices()) pr_warn("ACPI name space devices didn't probe correctly\n"); @@ -4044,17 +4023,15 @@ int __init intel_iommu_init(void) iommu_disable_protect_mem_regions(iommu); } - up_read(&dmar_global_lock); - - pr_info("Intel(R) Virtualization Technology for Directed I/O\n"); intel_iommu_enabled = 1; + dmar_register_bus_notifier(); + pr_info("Intel(R) Virtualization Technology for Directed I/O\n"); return 0; out_free_dmar: intel_iommu_free_dmars(); - up_write(&dmar_global_lock); return ret; } diff --git a/include/linux/dmar.h b/include/linux/dmar.h index d81a51978d01..8917a32173c4 100644 --- a/include/linux/dmar.h +++ b/include/linux/dmar.h @@ -65,6 +65,7 @@ struct dmar_pci_notify_info { extern struct rw_semaphore dmar_global_lock; extern struct list_head dmar_drhd_units; +extern int intel_iommu_enabled; #define for_each_drhd_unit(drhd) \ list_for_each_entry_rcu(drhd, &dmar_drhd_units, list, \ @@ -88,7 +89,8 @@ extern struct list_head dmar_drhd_units; static inline bool dmar_rcu_check(void) { return rwsem_is_locked(&dmar_global_lock) || - system_state == SYSTEM_BOOTING; + system_state == SYSTEM_BOOTING || + (IS_ENABLED(CONFIG_INTEL_IOMMU) && !intel_iommu_enabled); } #define dmar_rcu_dereference(p) rcu_dereference_check((p), dmar_rcu_check()) From 2380f1e8195ef612deea1dc7a3d611c5d2b9b56a Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Fri, 9 Sep 2022 16:46:31 -0300 Subject: [PATCH 158/158] iommu: Fix false ownership failure on AMD systems with PASID activated The AMD IOMMU driver cannot activate PASID mode on a RID without the RID's translation being set to IDENTITY. Further it requires changing the RID's page table layout from the normal v1 IOMMU_DOMAIN_IDENTITY layout to a different v2 layout. It does this by creating a new iommu_domain, configuring that domain for v2 identity operation and then attaching it to the group, from within the driver. This logic assumes the group is already set to the IDENTITY domain and is being used by the DMA API. However, since the ownership logic is based on the group's domain pointer equaling the default domain to detect DMA API ownership, this causes it to look like the group is not attached to the DMA API any more. This blocks attaching drivers to any other devices in the group. In a real system this manifests itself as the HD-audio devices on some AMD platforms losing their device drivers. Work around this unique behavior of the AMD driver by checking for equality of IDENTITY domains based on their type, not their pointer value. This allows the AMD driver to have two IDENTITY domains for internal purposes without breaking the check. Have the AMD driver properly declare that the special domain it created is actually an IDENTITY domain. Cc: Robin Murphy Cc: stable@vger.kernel.org Fixes: 512881eacfa7 ("bus: platform,amba,fsl-mc,PCI: Add device DMA ownership management") Reported-by: Takashi Iwai Tested-by: Takashi Iwai Signed-off-by: Jason Gunthorpe Reviewed-by: Robin Murphy Link: https://lore.kernel.org/r/0-v1-ea566e16b06b+811-amd_owner_jgg@nvidia.com Signed-off-by: Joerg Roedel --- drivers/iommu/amd/iommu_v2.c | 2 ++ drivers/iommu/iommu.c | 21 +++++++++++++++++++-- 2 files changed, 21 insertions(+), 2 deletions(-) diff --git a/drivers/iommu/amd/iommu_v2.c b/drivers/iommu/amd/iommu_v2.c index 696d5555be57..6a1f02c62dff 100644 --- a/drivers/iommu/amd/iommu_v2.c +++ b/drivers/iommu/amd/iommu_v2.c @@ -777,6 +777,8 @@ int amd_iommu_init_device(struct pci_dev *pdev, int pasids) if (dev_state->domain == NULL) goto out_free_states; + /* See iommu_is_default_domain() */ + dev_state->domain->type = IOMMU_DOMAIN_IDENTITY; amd_iommu_domain_direct_map(dev_state->domain); ret = amd_iommu_domain_enable_v2(dev_state->domain, pasids); diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c index 780fb7071577..3a808146b50f 100644 --- a/drivers/iommu/iommu.c +++ b/drivers/iommu/iommu.c @@ -3076,6 +3076,24 @@ out: return ret; } +static bool iommu_is_default_domain(struct iommu_group *group) +{ + if (group->domain == group->default_domain) + return true; + + /* + * If the default domain was set to identity and it is still an identity + * domain then we consider this a pass. This happens because of + * amd_iommu_init_device() replacing the default idenytity domain with an + * identity domain that has a different configuration for AMDGPU. + */ + if (group->default_domain && + group->default_domain->type == IOMMU_DOMAIN_IDENTITY && + group->domain && group->domain->type == IOMMU_DOMAIN_IDENTITY) + return true; + return false; +} + /** * iommu_device_use_default_domain() - Device driver wants to handle device * DMA through the kernel DMA API. @@ -3094,8 +3112,7 @@ int iommu_device_use_default_domain(struct device *dev) mutex_lock(&group->mutex); if (group->owner_cnt) { - if (group->domain != group->default_domain || - group->owner) { + if (group->owner || !iommu_is_default_domain(group)) { ret = -EBUSY; goto unlock_out; }