The netburst PMU is way different from the "architectural
perfomance monitoring" specification that current CPUs use.
P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle
perfomance monitoring events.
A few implementational details:
1) We need a separate x86_pmu::hw_config helper in struct
x86_pmu since register bit-fields are quite different from P6,
Core and later cpu series.
2) For the same reason is a x86_pmu::schedule_events helper
introduced.
3) hw_perf_event::config consists of packed ESCR+CCCR values.
It's allowed since in reality both registers only use a half
of their size. Of course before making a real write into a
particular MSR we need to unpack the value and extend it to
a proper size.
4) The tuple of packed ESCR+CCCR in hw_perf_event::config
doesn't describe the memory address of ESCR MSR register
so that we need to keep a mapping between these tuples
used and available ESCR (various P4 events may use same
ESCRs but not simultaneously), for this sake every active
event has a per-cpu map of hw_perf_event::idx <--> ESCR
addresses.
5) Since hw_perf_event::idx is an offset to counter/control register
we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel
strips it down to 8 registers and event armed may never be turned
off (ie the bit in active_mask is set but the loop never reaches
this index to check), thanks to Peter Zijlstra
Restrictions:
- No cascaded counters support (do we ever need them?)
- No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS
doesn't work for now)
- There are events with same counters which can't work simultaneously
(need to use intersected ones due to broken counter 1)
- No PERF_COUNT_HW_CACHE_ events yet
Todo:
- Implement dependent events
- Need proper hashing for event opcodes (no linear search, good for
debugging stage but not in real loads)
- Some events counted during a clock cycle -- need to set threshold
for them and count every clock cycle just to get summary statistics
(ie to behave the same way as other PMUs do)
- Need to swicth to use event_constraints
- To support RAW events we need to encode a global list of P4 events
into p4_templates
- Cache events need to be added
Event support status matrix:
Event status
-----------------------------
cycles works
cache-references works
cache-misses works
branch-misses works
bus-cycles partially (does not work on 64bit cpu with HT enabled)
instruction doesnt work (needs dependent event [mop tagging])
branches doesnt work
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Lin Ming <ming.m.lin@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Robert Richter <robert.richter@amd.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <20100311165439.GB5129@lenovo>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch moves omap_smc1 function to a seperate omap44xx-smc.S file
and sets compile flags as -Wa,-march=armv7-a.
This fix was suggested by Tony Lindgren <tony@atomide.com>
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
[tony@atomide.com: otherwise multi-omap build with V6 and V7 breaks]
Signed-off-by: Tony Lindgren <tony@atomide.com>
Disabling affine wakeups is too horrible to contemplate. Remove the feature flag.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301890.6785.50.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This features has been enabled for quite a while, after testing showed that
easing preemption for light tasks was harmful to high priority threads.
Remove the feature flag.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301675.6785.44.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Sync wakeups are critical functionality with a long history. Remove it, we don't
need the branch or icache footprint.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301817.6785.47.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This feature never earned its keep, remove it.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301591.6785.42.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Our preemption model relies too heavily on sleeper fairness to disable it
without dire consequences. Remove the feature, and save a branch or two.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301520.6785.40.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This feature hasn't been enabled in a long time, remove effectively dead code.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301447.6785.38.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Don't bother with selection when the current cpu is idle. Recent load
balancing changes also make it no longer necessary to check wake_affine()
success before returning the selected sibling, so we now always use it.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301369.6785.36.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Allow LAST_BUDDY to kick in sooner, improving cache utilization as soon as
a second buddy pair arrives on scene. The cost is latency starting to climb
sooner, the tbenefit for tbench 8 on my Q6600 box is ~2%. No detrimental
effects noted in normal idesktop usage.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301285.6785.34.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Now that we no longer depend on the clock being updated prior to enqueueing
on migratory wakeup, we can clean up a bit, placing calls to update_rq_clock()
exactly where they are needed, ie on enqueue, dequeue and schedule events.
In the case of a freshly enqueued task immediately preempting, we can skip the
update during preemption, as the clock was just updated by the enqueue event.
We also save an unneeded call during a migratory wakeup by not updating the
previous runqueue, where update_curr() won't be invoked.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301199.6785.32.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Both avg_overlap and avg_wakeup had an inherent problem in that their accuracy
was detrimentally affected by cross-cpu wakeups, this because we are missing
the necessary call to update_curr(). This can't be fixed without increasing
overhead in our already too fat fastpath.
Additionally, with recent load balancing changes making us prefer to place tasks
in an idle cache domain (which is good for compute bound loads), communicating
tasks suffer when a sync wakeup, which would enable affine placement, is turned
into a non-sync wakeup by SYNC_LESS. With one task on the runqueue, wake_affine()
rejects the affine wakeup request, leaving the unfortunate where placed, taking
frequent cache misses.
Remove it, and recover some fastpath cycles.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301121.6785.30.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Testing the load which led to this heuristic (nfs4 kbuild) shows that it has
outlived it's usefullness. With intervening load balancing changes, I cannot
see any difference with/without, so recover there fastpath cycles.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301062.6785.29.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Entering nohz code on every micro-idle is costing ~10% throughput for netperf
TCP_RR when scheduling cross-cpu. Rate limiting entry fixes this, but raises
ticks a bit. On my Q6600, an idle box goes from ~85 interrupts/sec to 128.
The higher the context switch rate, the more nohz entry costs. With this patch
and some cycle recovery patches in my tree, max cross cpu context switch rate is
improved by ~16%, a large portion of which of which is this ratelimiting.
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268301003.6785.28.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
the early_param() call in board-omap3touchbook.c expands to:
static const char __setup_str_early_touchbook_revision[]
__section(.init.rodata) _aligned(1) = tbr;
[...]
and we have a non-const variable being added to the
same section:
static struct ehci_hcd_omap_platform_data ehci_pdata
__section(.init.rodata);
because of that, gcc generates a section type conflict
which can (and actually should) be avoided by marking
const every variable marked with __initconst.
This patch fixes that for the ehci_hdc_omap_platform_data.
Signed-off-by: Felipe Balbi <felipe.balbi@nokia.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
... instead of comparing with DMA_ERROR_CODE, which will only work on
powerpc/sparc/x86.
Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
r852 fails to build when CONFIG_PCI is not enabled since it uses
pci_*() calls and is a PCI driver, so it should depend on PCI
to prevent build errors.
It should also #include <linux/pci.h>.
drivers/mtd/nand/r852.c:1053: error: implicit declaration of function 'pci_prepare_to_sleep'
drivers/mtd/nand/r852.c:1062: error: implicit declaration of function 'pci_back_from_sleep'
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Maxim Levitsky <maximlevitsky@gmail.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
gfs2_lock() will skip locks on file which have mode set to 02666. This is a problem in cases where the mode of the file is changed after a process has obtained a lock on the file. Such a lock will be skipped and will result in a BUG in locks_remove_flock().
gfs2_lock() should skip the check for mandatory locks when unlocking a file.
Signed-off-by: Sachin Prabhu <sprabhu@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Up to now, applying the in-core activity-log to the on-disk
bitmap did not care for logical_block_size.
On logical_block_size != 512 byte, this very likely results
in misalligned block access and spurious "io errors".
We now simply always submit aligned whole 4k blocks, fixing this
for logical block sizes of 512, 1024, 2048 and 4096.
For even larger logical block sizes, this won't work.
But I'm not aware of devices with such properties being available.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Up to now this only worked for Outdated and Inconsistent disks, that
it did not worked for Consistent disks was an inconsistent omission.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
This is a race condition that existed for ages.
The previous commit reduces the window, this one closes it.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
cmdname() should map command number to its human readable
representation. The string table was incomplete, though.
Maybe rather do a switch() block, and let the compiler help us
to keep it complete?
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Fixes a race and potential kernel panic if e.g. the worker was just
about to send a few P_RS_IS_IN_SYNC via the meta socket for checksum
based resync, while the receiver destroys the sockets in
drbd_disconnect.
To make sure no-one is using the meta socket,
it is not enough to stop the asender...
Grab the meta socket mutex before destroying it.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Depending on resync request size,
we need to account for more than one bit.
Impact: cosmetic
If SyncTarget reported correctly 100% equal checksums,
the SyncSource usually reported 12% equal checksums instead,
because it only counted requests, we typically do 32k resync requests,
and the bitmap granularity is still 4k.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Situation:
we have diverging data sets, i.e. we had a split brain somewhen,
but currently are connected, one node diskless.
Then we try to attach that disk, figure it is consistent,
but has a diverging data set, we refuse to attach.
This led to strange state changes:
22:18:35 bb drbd1: peer( Unknown -> Primary ) conn( WFReportParams -> Connected) pdsk( DUnknown -> UpToDate )
22:19:30 bb drbd1: disk( Diskless -> Attaching )
22:19:30 bb drbd1: disk( Attaching -> Negotiating )
22:19:30 bb drbd1: drbd_sync_handshake:
22:19:30 bb drbd1: self 97BF25798B9D5222:F33D1F62ADE698DD:4269796F9D027C83:AC45D8B5C3C1BF93 bits:19449 flags:0
22:19:30 bb drbd1: peer 280DFB6E125465D3:F33D1F62ADE698DC:4269796F9D027C82:AC45D8B5C3C1BF93 bits:2575806 flags:0
22:19:30 bb drbd1: uuid_compare()=100 by rule 90
22:19:30 bb drbd1: Split-Brain detected, dropping connection!
22:19:30 bb drbd1: disk( Negotiating -> Diskless )
while the other side says:
22:19:30 aa drbd1: Split-Brain detected, dropping connection!
22:19:30 aa drbd1: Disk attach process on the peer node was aborted.
22:19:30 aa drbd1: conn( Connected -> TOO_LARGE ) pdsk( Diskless -> Consistent )
This should be fixed now.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
we still don't support 4k 'physical' sectors 'natively',
but use a read-modify-write workaround.
And we even tried to use the extra page before we allocated it :(
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
This patch is an optimization in perf_event_task_sched_in() to avoid
scheduling the events twice in a row.
Without it, the perf_disable()/perf_enable() pair is invoked twice,
thereby pinned events counts while scheduling flexible events and we go
throuh hw_perf_enable() twice.
By encapsulating, the whole sequence into perf_disable()/perf_enable() we
ensure, hw_perf_enable() is going to be invoked only once because of the
refcount protection.
Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268288765-5326-1-git-send-email-eranian@google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Put all statistic fields of sched_entity in one struct, sched_statistics,
and embed it into sched_entity.
This change allows to memset the sched_statistics to 0 when needed (for
instance when forking), avoiding bugs of non initialized fields.
Signed-off-by: Lucas De Marchi <lucas.de.marchi@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268275065-18542-1-git-send-email-lucas.de.marchi@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Since pick_next_highest_task_rt() already iterates all the cgroups and
is really only interested in tasks, skip over the !task entries.
Reported-by: Dhaval Giani <dhaval.giani@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Tested-by: Dhaval Giani <dhaval.giani@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Export perf_trace_regs and perf_arch_fetch_caller_regs since module will
use these.
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
[ use EXPORT_PER_CPU_SYMBOL_GPL() ]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <4B989C1B.2090407@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
What happens is that we schedule badly like:
<...>-1987 [019] 280.252808: x86_pmu_start: event-46/1300c0: idx: 0
<...>-1987 [019] 280.252811: x86_pmu_start: event-47/1300c0: idx: 1
<...>-1987 [019] 280.252812: x86_pmu_start: event-48/1300c0: idx: 2
<...>-1987 [019] 280.252813: x86_pmu_start: event-49/1300c0: idx: 3
<...>-1987 [019] 280.252814: x86_pmu_start: event-50/1300c0: idx: 32
<...>-1987 [019] 280.252825: x86_pmu_stop: event-46/1300c0: idx: 0
<...>-1987 [019] 280.252826: x86_pmu_stop: event-47/1300c0: idx: 1
<...>-1987 [019] 280.252827: x86_pmu_stop: event-48/1300c0: idx: 2
<...>-1987 [019] 280.252828: x86_pmu_stop: event-49/1300c0: idx: 3
<...>-1987 [019] 280.252829: x86_pmu_stop: event-50/1300c0: idx: 32
<...>-1987 [019] 280.252834: x86_pmu_start: event-47/1300c0: idx: 1
<...>-1987 [019] 280.252834: x86_pmu_start: event-48/1300c0: idx: 2
<...>-1987 [019] 280.252835: x86_pmu_start: event-49/1300c0: idx: 3
<...>-1987 [019] 280.252836: x86_pmu_start: event-50/1300c0: idx: 32
<...>-1987 [019] 280.252837: x86_pmu_start: event-51/1300c0: idx: 32 *FAIL*
This happens because we only iterate the n_running events in the first
pass, and reset their index to -1 if they don't match to force a
re-assignment.
Now, in our RR example, n_running == 0 because we fully unscheduled, so
event-50 will retain its idx==32, even though in scheduling it will have
gotten idx=0, and we don't trigger the re-assign path.
The easiest way to fix this is the below patch, which simply validates
the full assignment in the second pass.
Reported-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1268311069.5037.31.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Fix:
arch/powerpc/kernel/perf_event.c:1334: error: 'power_pmu_notifier' undeclared (first use in this function)
arch/powerpc/kernel/perf_event.c:1334: error: (Each undeclared identifier is reported only once
arch/powerpc/kernel/perf_event.c:1334: error: for each function it appears in.)
arch/powerpc/kernel/perf_event.c:1334: error: implicit declaration of function 'power_pmu_notifier'
arch/powerpc/kernel/perf_event.c:1334: error: implicit declaration of function 'register_cpu_notifier'
Due to commit 3f6da390 (perf: Rework and fix the arch CPU-hotplug hooks).
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Reduce warning message output to one line only instead of per
cpu.
Signed-of-by: Mike Travis <travis@sgi.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: x86@kernel.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
There is possible to save r3/r4 at the beggining of user part
before calling handlers and at the end restore it.
Signed-off-by: Michal Simek <monstr@monstr.eu>
We have to use consistent code to be able to do coherent dma
function. In consistent code is used cache inhibit page mapping.
Xilinx reported that there is bug in Microblaze for WB and
d-cache_always use option. Microblaze 7.30.a should be first version
where is this bug removed.
Signed-off-by: Michal Simek <monstr@monstr.eu>
Remove ancient Kconfig option for consistent code.
MMU uses cache inhibit pages.
noMMU uses UNCACHE SHADOW feature where is used double ram size.
For example:
Physical ram is 256MB and cache are setup to cover the same size.
But if you setup in HW that size is 512MB and cache covers 256MB
than you can use adresses from 256-512MB without caches and
correspond with 0-256MB with cache. That's why I am using
dcache base/high addresses to find out uncache area.
Signed-off-by: Michal Simek <monstr@monstr.eu>
Update UV mmr definitions header file. Eliminate definitions no
longer needed. Move 2 definitions from tlb_uv.c into the header
file where they belong.
Signed-off-by: Jack Steiner <steiner@sgi.com>
LKML-Reference: <20100310204458.GA28835@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
I found several problems for ll_temac driver and on system with WB.
This early fix should fix it. I will clean this patch before I will add
it to mainline
Signed-off-by: Michal Simek <monstr@monstr.eu>