This one isn't related to previous patch. If online cpus are
below NUM_INVALIDATE_TLB_VECTORS, we don't need the lock. The
comments in the code declares we don't need the check, but a hot
lock still needs an atomic operation and expensive, so add the
check here.
Uses nr_cpu_ids here as suggested by Eric Dumazet.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Andi Kleen <andi@firstfloor.org>
LKML-Reference: <1295232730.1949.710.camel@sli10-conroe>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Make the maxium TLB invalidate vectors depend on NR_CPUS linearly,
with a maximum of 32 vectors.
We currently only have 8 vectors for TLB invalidation and that is clearly
inadequate. If we have a lot of CPUs, the CPUs need share the 8 vectors and
tlbstate_lock is used to protect them. flush_tlb_page() is
heavily used in page reclaim, which will cause a lot of lock
contention for tlbstate_lock.
Andi Kleen suggested increasing the vectors number to 32, which should be
good for current typical systems to reduce the tlbstate_lock contention.
My test system has 4 sockets and 64G memory, and 64 CPUs. My
workload creates 64 processes. Each process mmap reads a big
empty sparse file. The total size of the files are 2*total_mem,
so this will cause a lot of page reclaim.
Below is the result I get from perf call-graph profiling:
without the patch:
------------------
24.25% usemem [kernel] [k] _raw_spin_lock
|
--- _raw_spin_lock
|
|--42.15%-- native_flush_tlb_others
with the patch:
------------------
14.96% usemem [kernel] [k] _raw_spin_lock
|
--- _raw_spin_lock
|--13.89%-- native_flush_tlb_others
So this heavily reduces the tlbstate_lock contention.
Suggested-by: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1295232727.1949.709.camel@sli10-conroe>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Add up to 32 invalidate_interrupt handlers. How many handlers are
added depends on NUM_INVALIDATE_TLB_VECTORS. So if
NUM_INVALIDATE_TLB_VECTORS is smaller than 32, we reduce code
size.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
LKML-Reference: <1295232725.1949.708.camel@sli10-conroe>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cleanup the vector usage and make them continuous if possible.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
LKML-Reference: <1295232722.1949.707.camel@sli10-conroe>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Add NULL check for avoiding NULL pointer deref.
This bug has been introduced by:
1ff511e35e: tracing/kprobes: Add bitfield type
which causes a null pointer dereference bug when kprobe-tracer
parses an argument without type.
Reported-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: 2nddept-manager@sdl.hitachi.co.jp
Cc: Peter Zijlstra <peterz@infradead.org>
LKML-Reference: <20110214054807.8919.69740.stgit@ltc236.sdl.hitachi.co.jp>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Reported-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
We use it in non __cpuinit code now too so drop marker.
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
LKML-Reference: <20110211171754.GA21047@aftab>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Conflicts:
arch/x86/mm/numa_64.c
Merge reason: fix the conflict, update to latest -rc and pick up this
dependent fix from Yinghai:
e6d2e2b2b1: memblock: don't adjust size in memblock_find_base()
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Slave-dma has become the predominant usage model for dmaengine and needs
special attention. Memory-to-memory dma usage cases will continue to be
maintained by Dan.
Cc: Alan Cox <alan@linux.intel.com>
Acked-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Currently when two or more buffers are queued by the camera driver
and so the double buffering is enabled in the idmac, we lose one
frame comming from CSI since the reporting of arrival of the first
frame is deferred by the DMAIC_7_EOF interrupt handler and reporting
of the arrival of the last frame is not done at all. So when requesting
N frames from the image sensor we actually receive N - 1 frames in
user space.
The reason for this behaviour is that the DMAIC_7_EOF interrupt
handler misleadingly assumes that the CUR_BUF flag is pointing to the
buffer used by the IDMAC. Actually it is not the case since the
CUR_BUF flag will be flipped by the FSU when the FSU is sending the
<TASK>_NEW_FRM_RDY signal when new frame data is delivered by the CSI.
When sending this singal, FSU updates the DMA_CUR_BUF and the
DMA_BUFx_RDY flags: the DMA_CUR_BUF is flipped, the DMA_BUFx_RDY
is cleared, indicating that the frame data is beeing written by
the IDMAC to the pointed buffer. DMA_BUFx_RDY is supposed to be
set to the ready state again by the MCU, when it has handled the
received data. DMAIC_7_CUR_BUF flag won't be flipped here by the
IPU, so waiting for this event in the EOF interrupt handler is wrong.
Actually there is no spurious interrupt as described in the comments,
this is the valid DMAIC_7_EOF interrupt indicating reception of the
frame from CSI.
The patch removes code that waits for flipping of the DMAIC_7_CUR_BUF
flag in the DMAIC_7_EOF interrupt handler. As the comment in the
current code denotes, this waiting doesn't help anyway. As a result
of this removal the reporting of the first arrived frame is not
deferred to the time of arrival of the next frame and the drivers
software flag 'ichan->active_buffer' is in sync with DMAIC_7_CUR_BUF
flag, so the reception of all requested frames works.
This has been verified on the hardware which is triggering the
image sensor by the programmable state machine, allowing to
obtain exact number of frames. On this hardware we do not tolerate
losing frames.
This patch also removes resetting the DMA_BUFx_RDY flags of
all channels in ipu_disable_channel() since transfers on other
DMA channels might be triggered by other running tasks and the
buffers should always be ready for data sending or reception.
Signed-off-by: Anatolij Gustschin <agust@denx.de>
Reviewed-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Tested-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Bosch C_CAN controller is a full-CAN implementation which is compliant
to CAN protocol version 2.0 part A and B. Bosch C_CAN user manual can be
obtained from:
http://www.semiconductors.bosch.de/media/en/pdf/ipmodules_1/c_can/users_manual_c_can.pdf
This patch adds the support for this controller.
The following are the design choices made while writing the controller
driver:
1. Interface Register set IF1 has be used only in the current design.
2. Out of the 32 Message objects available, 16 are kept aside for RX
purposes and the rest for TX purposes.
3. NAPI implementation is such that both the TX and RX paths function
in polling mode.
Signed-off-by: Bhupesh Sharma <bhupesh.sharma@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This PCH_GBE driver had an issue that the receiving data is not normal.
This driver had not removed correctly the padding data
which the DMA include in receiving data.
This patch fixed this issue.
Signed-off-by: Toshiharu Okada <toshiharu-linux@dsn.okisemi.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some JMicron Chip treat 0 as error checksum for UDP packets.
Which should be "No checksum needed".
Reported-by: Adam Swift <Adam.Swift@omnitude.net>
Confirmed-by: "Aries Lee" <arieslee@jmicron.com>
Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The value of the register which holds receive Unicast MAC Address
sometimes get messed-up after resume.
This patch refill it before enabling the hardware filter.
Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adding control to clk_rx, and makes the control of clk_{rx|tx|tcp}
with safer sequence.
This sequence is provided by JMicron.
Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Clear Transmit Timer/Retry setting while full-duplex.
Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
After main chip rev 5, the hardware support more power saving
control registers.
Some Non-Linux drivers might turn off the phy power with new
interfaces, this patch makes it possible for Linux to turn it
on again.
Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Get the main and sub chip revision for later workaround use.
Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch enables it by default when the driver starts.
This has been required by many people and seems to actually be
useful on STB.
At any rate, the WoL modes can be selected and turned-on/off
by using the ethtool at run-time by users.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch allows userspace to enslave/release slave devices via netlink
interface using IFLA_MASTER. This introduces generic way to add/remove
underling devices.
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
Acked-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
If alloc_skb() fails to allocate memory and returns NULL then we want to
return -ENOMEM from drivers/atm/solos-pci.c::popen() regardless of the
value of net_ratelimit(). The way the code is today, we may not return if
net_ratelimit() returns 0, then we'll proceed to pass a NULL pointer to
skb_put() which will blow up in our face.
This patch ensures that we always return -ENOMEM on alloc_skb() failure
and only let the dev_warn() be controlled by the value of net_ratelimit().
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
In drivers/net/usb/hso.c::hso_create_bulk_serial_device() we have this
code:
...
serial = kzalloc(sizeof(*serial), GFP_KERNEL);
if (!serial)
goto exit;
...
exit:
hso_free_tiomget(serial);
...
hso_free_tiomget() directly dereferences its argument, which in the
example above is a NULL pointer, ouch.
I could just add a 'if (serial)' test at the 'exit' label, but since most
freeing functions in the kernel accept NULL pointers (and it seems like
this was also assumed here) I opted to instead change 'hso_free_tiomget()'
so that it is safe to call it with a NULL argument. I also modified the
function to get rid of a pointles conditional before the call to
'usb_free_urb()' since that function already tests for NULL itself -
besides fixing the NULL deref this change also buys us a few bytes in
size.
Before:
$ size drivers/net/usb/hso.o
text data bss dec hex filename
32200 592 9960 42752 a700 drivers/net/usb/hso.o
After:
$ size drivers/net/usb/hso.o
text data bss dec hex filename
32196 592 9960 42748 a6fc drivers/net/usb/hso.o
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit c0e69a5bbc ("klist.c: bit 0 in pointer can't be used as flag")
intended to make sure that all klist objects were at least pointer size
aligned, but used the constant "4" which only works on 32-bit.
Use "sizeof(void *)" which is correct in all cases.
Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Jesper Nilsson <jesper.nilsson@axis.com>
Cc: stable <stable@kernel.org>
Cc: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jesper Juhl noticed that l2_pull_iqueue() does not
check to see if alloc_skb() fails.
Fix this by first trying to reallocate the headroom
if necessary, rather than later after we've made hard
to undo state changes.
Reported-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
* 'intel/drm-intel-fixes' of /ssd/git/drm-next:
drm/i915: Fix resume regression from 5d1d0cc
drm/i915/tv: Use polling rather than interrupt-based hotplug
drm/i915: Trigger modesetting if force-audio changes
drm/i915/sdvo: If we have an EDID confirm it matches the mode of the connection
drm/i915: Disable RC6 on Ironlake
drm/i915/lvds: Restore dithering on native modes for gen2/3
drm/i915: Invalidate TLB caches on SNB BLT/BSD rings
Makes debugging CS rejections much easier.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
This is an important security fix because we allowed arbitrary values
to be passed to AARESOLVE_OFFSET. This also puts the right buffer address
in the register.
Signed-off-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Not only is linear aligned supposedly more performant,
linear general is only supported by the CB in single
slice mode. The texture hardware doesn't support
linear general, but I think the hw automatically
upgrades it to linear aligned.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Not only is linear aligned supposedly more performant,
linear general is only supported by the CB in single
slice mode. The texture hardware doesn't support
linear general, but I think the hw automatically
upgrades it to linear aligned.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
My evergreen has been in a remote PC for week and reset has never once
saved me from certain doom, I finally relocated to the box with a
serial cable and noticed an oops when the GPU resets, and the TTM
delayed delete thread tries to remove something from the GTT.
This stops the delayed delete thread from executing across the GPU
reset handler, and woot I can GPU reset now.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Based on 6xx/7xx endian fixes from Cédric Cano.
v2: fix typo in shader
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
agd5f: minor cleanups
Signed-off-by: Cédric Cano <ccano@interfaceconcept.com>
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
agd5f: minor cleanups
Signed-off-by: Cédric Cano <ccano@interfaceconcept.com>
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The colorbuffer, zbuffer, and texture states are checked only once when
they get changed. This improves performance in the apps which emit
lots of draw packets and few state changes.
This drops performance in glxgears by a 1% or so, but glxgears is not
a benchmark we care about.
The time spent in the kernel when running Torcs dropped from 33% to 23%
and the frame rate is higher, which is a good thing.
r600 might need something like this as well.
Signed-off-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
In the continuing effort to avoid kernel addresses leaking to unprivileged
users, this patch switches to %pK for /proc/dri/*/vma.
Signed-off-by: Kees Cook <kees.cook@canonical.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
PPC Mac cards do not provide connector tables in
their vbios. Their connector/encoder configurations
must be hardcoded in the driver.
verified by nyef on #radeon
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
drivers/gpu/drm/radeon/mkregtable.c:parser_auth() almost always remembers
to fclose(file) before returning, but it misses two spots.
This is not really important since the process will exit shortly after and
thus close the file for us, but being explicit prevents static analysis
tools from complaining about leaked memory and missing fclose() calls and
it also seems to be the prefered style of the existing code to explicitly
close the file.
So, here's a patch to add the two missing fclose() calls.
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Dave Airlie <airlied@redhat.com>
- set scaler table clears the interleave bit, need to
reset it in encoder quirks, this was already done for
pre-dce4.
- remove the interleave settings from set_base() functions
this is now handled in the encoder quirks functions, and
isn't technically part of the display base setup.
- rename evergreen_do_set_base() to dce4_do_set_base() since
it's used on both evergreen and NI asics.
Fixes:
https://bugzilla.kernel.org/show_bug.cgi?id=28182
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
The old code dereferenced a value, the new code just needs to pass
the ptr.
fixes an oops looking at files in debugfs.
cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
Takeover raid1->raid0 not succeded. Kernel message is shown:
"md/raid0:md126: too few disks (1 of 2) - aborting!"
Problem was that we weren't updating ->raid_disks for that
takeover, unlike all the others.
Signed-off-by: Krzysztof Wojcik <krzysztof.wojcik@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
