In case of MCC_Q creation failed, the MCCQ info memory is freed
from be_mcc_queues_destroy and be_mcc_queues_create. This caused
kernel to panic because of double free.
Signed-off-by: John Soni Jose <sony.john-n@emulex.com>
Signed-off-by: Jayamohan Kallickal <jayamohan.kallickal@emulex.com>
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
The WRB and SGL Handle allocated for Login task were not freed
back to the pool after the login process was done. This code
releases the WRB and SGL Handle after the login process.
v2:
- Fix up locking so bh calls are not done when not needed.
- Make beiscsi_cleanup_task static.
Signed-off-by: Jayamohan Kallickal <jayamohan.kallickal@emulex.com>
[various fixes]
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Removing code duplication during the WRB_Handle and WRB
initialization.
Added memory allocation failure handling code during WRB
initialization.
Signed-off-by: Jayamohan Kallickal <jayamohan.kallickal@emulex.com>
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Set the ASYNC PDU Handle pBuffer for Data ring with the VA/PA
of the allocated memory for it.
To get the correct ASYNC PDY Handle iterate the list and compare
the PA set during initialization with the passed PHY Address.
The buffer_size and num_enteries are common for HDR and Data ring
Signed-off-by: Jayamohan Kallickal <jayamohan.kallickal@emulex.com>
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
A previous commit changed the mfs checking to ensure the new
mfs is less or equal to the mfs supported by the FCF. This
doesn't work for BRDCM cards as they set an mfs of 2048 regardless
of whether the switch returns a larger mfs.
This patch validates the new mfs against the upper and lower spec
defined boundries for a FCoE mfs.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
The patch "ath6kl: support fw reporting phy capabilities" gave the
firmware the ability to disable certain wiphy supported bands. Check if
this is the case in ath6kl_wmi_beginscan_cmd to avoid dereferencing a
NULL pointer.
kvalo: change the patch so that there's no code between declarations
Signed-off-by: Thomas Pedersen <c_tpeder@qca.qualcomm.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
Based on the original patch from Ying Cai <ycai@google.com>
This change ensures that the itr/itr_setting adjustment logic is used,
even for the default/compiled-in value.
Context:
When we changed the default InterruptThrottleRate value from default
(3 = dynamic mode) to 8000 for example, only adapter->itr_setting
(which controls interrupt coalescing mode) was set to 8000, but
adapter->itr (which controls the value set in NIC register) was not
updated accordingly. So from ethtool, it seemed the interrupt
throttling is enabled at 8000 intr/s, but the NIC actually was
running in dynamic mode which has lower CPU efficiency especially
when throughput is not high.
CC: Ying Cai <ycai@google.com>
CC: David Decotigny <david.decotigny@google.com>
Signed-off-by: Jeff Kirsher <jeffrey.kirsher@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Following logs where seen on Systems with multiple NICs,
while using MSI interrupts as shown below:
Feb 16 15:09:32 (none) user.notice kernel: 0000:00:0d.0: lan0_0: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:32 (none) user.notice kernel: 0000:40:0d.0: wan0_1: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:32 (none) user.notice kernel: 0000:40:0d.0: lan0_1: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:32 (none) user.warn kernel: 0000:40:0e.0: wan4_0: MSI interrupt
test failed, using legacy interrupt.
Feb 16 15:09:32 (none) user.notice kernel: 0000:00:0e.0: wan1_0: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:33 (none) user.notice kernel: 0000:00:0e.0: lan1_0: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:33 (none) user.notice kernel: 0000:00:0f.0: wan2_0: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:33 (none) user.notice kernel: 0000:00:0f.0: lan2_0: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:33 (none) user.notice kernel: 0000:40:0a.0: wan3_0: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:33 (none) user.notice kernel: 0000:40:0a.0: lan3_0: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:34 (none) user.notice kernel: 0000:40:0e.0: lan4_0: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:34 (none) user.notice kernel: 0000:40:0f.0: wan5_0: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 16 15:09:34 (none) user.notice kernel: 0000:40:0f.0: lan5_0: NIC Link is Up
1000 Mbps Full Duplex, Flow Control: RX/TX
This patch fixes this problem by increasing the msleep from 50 to 100.
Signed-off-by: Prasanna S Panchamukhi <ppanchamukhi@riverbed.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
The rcutorture initialization code ignored the error returns from
rcu_torture_onoff_init() and rcu_torture_stall_init(). The rcutorture
cleanup code failed to NULL out a number of pointers. These bugs will
normally have no effect, but this commit fixes them nevertheless.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Both Steven Rostedt's new idle-capable trace macros and the RCU_NONIDLE()
macro can cause RCU to momentarily pause out of idle without the rest
of the system being involved. This can cause rcu_prepare_for_idle()
to run through its state machine too quickly, which can in turn result
in needless scheduling-clock interrupts.
This commit therefore adds code to enable rcu_prepare_for_idle() to
distinguish between an initial entry to idle on the one hand (which needs
to advance the rcu_prepare_for_idle() state machine) and an idle reentry
due to idle-capable trace macros and RCU_NONIDLE() on the other hand
(which should avoid advancing the rcu_prepare_for_idle() state machine).
Additional state is maintained to allow the timer to be correctly reposted
when returning after a momentary pause out of idle, and even more state
is maintained to detect when new non-lazy callbacks have been enqueued
(which may require re-evaluation of the approach to idleness).
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The RCU_FAST_NO_HZ facility uses an hrtimer to wake up a CPU when
it is allowed to go into dyntick-idle mode, which is almost always
cancelled soon after. This is not what hrtimers are good at, so
this commit switches to the timer wheel.
Reported-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Traces of rcu_prep_idle events can be confusing because
rcu_cleanup_after_idle() does no tracing. This commit therefore adds
this tracing.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The rcu_blocking_is_gp() function tests to see if there is only one
online CPU, and if so, synchronize_sched() and friends become no-ops.
However, for larger systems, num_online_cpus() scans a large vector,
and might be preempted while doing so. While preempted, any number
of CPUs might come online and go offline, potentially resulting in
num_online_cpus() returning 1 when there never had only been one
CPU online. This could result in a too-short RCU grace period, which
could in turn result in total failure, except that the only way that
the grace period is too short is if there is an RCU read-side critical
section spanning it. For RCU-sched and RCU-bh (which are the only
cases using rcu_blocking_is_gp()), RCU read-side critical sections
have either preemption or bh disabled, which prevents CPUs from going
offline. This in turn prevents actual failures from occurring.
This commit therefore adds a large block comment to rcu_blocking_is_gp()
documenting why it is safe. This commit also moves rcu_blocking_is_gp()
into kernel/rcutree.c, which should help prevent unwary developers from
mistaking it for a generally useful function.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Commit #0209f649 (rcu: limit rcu_node leaf-level fanout) set an upper
limit of 16 on the leaf-level fanout for the rcu_node tree. This was
needed to reduce lock contention that was induced by the synchronization
of scheduling-clock interrupts, which was in turn needed to improve
energy efficiency for moderate-sized lightly loaded servers.
However, reducing the leaf-level fanout means that there are more
leaf-level rcu_node structures in the tree, which in turn means that
RCU's grace-period initialization incurs more cache misses. This is
not a problem on moderate-sized servers with only a few tens of CPUs,
but becomes a major source of real-time latency spikes on systems with
many hundreds of CPUs. In addition, the workloads running on these large
systems tend to be CPU-bound, which eliminates the energy-efficiency
advantages of synchronizing scheduling-clock interrupts. Therefore,
these systems need maximal values for the rcu_node leaf-level fanout.
This commit addresses this problem by introducing a new kernel parameter
named RCU_FANOUT_LEAF that directly controls the leaf-level fanout.
This parameter defaults to 16 to handle the common case of a moderate
sized lightly loaded servers, but may be set higher on larger systems.
Reported-by: Mike Galbraith <efault@gmx.de>
Reported-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Currently, __kfree_rcu() is implemented as an inline function, and
contains a BUILD_BUG_ON() that malfunctions if __kfree_rcu() is compiled
as an out-of-line function. Unfortunately, there are compiler settings
(e.g., -O0) that can result in __kfree_rcu() being compiled out of line,
resulting in annoying build breakage. This commit therefore converts
both __kfree_rcu() and __is_kfree_rcu_offset() from inline functions to
macros to prevent such misbehavior on the part of the compiler.
Signed-off-by: Jan Engelhardt <jengelh@medozas.de>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
The old text confused real-time applications with real-time threads, so
that you pretty much needed to understand how this kernel configuration
parameter worked to understand the help text. This commit therefore
attempts to make the help text human-readable.
Reported-by: Jörn Engel <joern@purestorage.com>
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The list_first_entry_rcu() macro is inherently unsafe because it cannot
be applied to an empty list. But because RCU readers do not exclude
updaters, a list might become empty between the time that list_empty()
claimed it was non-empty and the time that list_first_entry_rcu() is
invoked. Therefore, the list_empty() test cannot be separated from the
list_first_entry_rcu() call. This commit therefore combines these to
macros to create a new list_first_or_null_rcu() macro that replaces
the old (and unsafe) list_first_entry_rcu() macro.
This patch incorporates Paul's review comments on the previous version of
this patch available here:
https://lkml.org/lkml/2012/4/2/536
This patch cannot break any upstream code because list_first_entry_rcu()
is not being used anywhere in the kernel (tested with grep(1)), and any
external code using it is probably broken as a result of using it.
Signed-off-by: Michel Machado <michel@digirati.com.br>
CC: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
CC: Dipankar Sarma <dipankar@in.ibm.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
* Make __list_add_rcu check the next->prev and prev->next pointers
just like __list_add does.
* Make list_del_rcu use __list_del_entry, which does the same checking
at deletion time.
Has been running for a week here without anything being tripped up,
but it seems worth adding for completeness just in case something
ever does corrupt those lists.
Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Newer SoCs from the S3C24XX line, namely S3C2416/2443/2450 contain
hsspi-controllers compatible with the s3c64xx type.
The previous patches enabled platform support for it, so make the
driver also usable for the S3C24xx architecture.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Acked-by: Grant Likely <grant.likely@secretlab.ca>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
This lets the s3c64xx-spi driver know the specifics of the controller-
variant and also setups the gpios and the misccr bit.
This setup is valid for all S3C24XX SoCs containing a HSSPI controller
(i.e. S3C2416/2450 and S3C2443)
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
The S3C_PA_SPIx constants are only used by the s3c64xx-spi driver
and don't conflict with the older SoCs as they don't support hsspi.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
SoCs starting with the S3C2443 contain SPI controllers compatible with the
spi-s3c64xx driver and therefore need separate dma channels for rx and tx.
This patch introduces dma channel declarations for these and changes the
dma-s3c2443.c accordingly.
None of the older SoCs use the spi-dma at all. Most boards bitbang their
spi use and the spi-s3c24xx driver also does not use the dma system.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
When the hsmmc clock lookup was changed to be lookup tables based,
it seems the S3C2443 was forgotten.
As subsequent patches will want to add more lookups to it, this
patch adds the base table with the missing hsmmc lookup.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
Commit 6e8201f57c "mmc: core: add the capability for broken voltage"
introduced a new quirk to indicate that MMC core should ignore voltage
change errors reported by the regulators core. This is required to get
SDHCI working on UniversalC210, NURI and GONI boards again after commit
ceb6143b2d ("mmc: sdhci: fix vmmc handling").
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
Using the common clock infrastructure without the common clkdev code makes
little sense, so select CLKDEV_LOOKUP for COMMON_CLK.
Signed-off-by: Rob Herring <rob.herring@calxeda.com>
if (!clk->ops->round_rate && (clk->flags & CLK_SET_RATE_PARENT)) is true, then
we don't need to set clk->new_rate here, as we will call clk_calc_subtree()
afterwards and it also sets clk->new_rate.
Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
All macros used for creating different kind of clocks have similar code for
initializing struct clk. This patch removes those redundant lines and create
another macro DEFINE_CLK.
Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
This patch tries to remove duplicate code for clk_gate clocks. This creates
another routine clk_gate_endisable() which will take care of enable/disable
clock with knowledge of CLK_GATE_SET_TO_DISABLE flag.
It works on following logic:
For enabling clock, enable = 1
set2dis = 1 -> clear bit -> set = 0
set2dis = 0 -> set bit -> set = 1
For disabling clock, enable = 0
set2dis = 1 -> set bit -> set = 1
set2dis = 0 -> clear bit -> set = 0
So, result is always: enable xor set2dis.
Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
CLK_MUX_INDEX_BIT is mistakenly written as CLK_MUX_INDEX_BITWISE in comment. Fix
it.
CLK_GATE_SET_TO_DISABLE is mistakenly written as CLK_GATE_SET_DISABLE in
comment. Fix it.
Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
Need to propagate round_rate call for the clk that has no .round_rate
operation but with flag CLK_SET_RATE_PARENT set.
For example, clk_mux is a clk with no .round_rate operation. However,
it could likely be in a clk_set_rate propagation path, saying it has
parent clk who has .round_rate and .set_rate operations.
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
For most of .set_rate implementation, parent_rate will be used, so just
like passing parent_rate into .recalc_rate, let's pass parent_rate into
.set_rate too.
It also updates the kernel doc for .set_rate ops.
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
The parent_rate will likely be used by most .round_rate implementation
no matter whether flag CLK_SET_RATE_PARENT is set or not, so let's
always pass parent_rate into .round_rate.
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
This patch is the basic clk version of 'clk: core: copy parent_names &
return error codes'.
The registration functions are changed to allow the core code to copy
the array of strings and allow platforms to declare those arrays as
__initdata.
This patch also converts all of the basic clk registration functions to
return error codes which better aligns them with the existing clk.h api.
Signed-off-by: Mike Turquette <mturquette@linaro.org>
This patch cleans up clk_register and solves a few bugs by teaching
clk_register and __clk_init to return error codes (instead of just NULL)
to better align with the existing clk.h api.
Along with that change this patch also introduces a new behavior whereby
clk_register copies the parent_names array, thus allowing platforms to
declare their parent_names arrays as __initdata.
Signed-off-by: Mike Turquette <mturquette@linaro.org>
Drivers should be able to declare their arrays of parent names as const
so the APIs need to accept const arguments.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
[mturquette@linaro.org: constified gate]
Signed-off-by: Mike Turquette <mturquette@linaro.org>
