| Age | Commit message (Collapse) | Author |
|---|
|
When internal users want VRAM we shouldn't return GART memory instead.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Driver internal users shouldn't be limited in their allocation size.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
When allocating more than 2GB of GART the array of pages
gets to big for kzalloc, use vzalloc instead.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
GART and VRAM size limits need to be a power of two.
Fix values greater than 1GB and simplify those checks a bit.
v2: also fix radeon_vram_limit usage, and simplify test even more.
v3: agd5f: fix spelling as noticed by Klaus Schnass
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Instead of the current whitelist which accepts duplicates
only for the quiet and vendor IEs, use a blacklist of all
IEs (that we currently parse) that can't be duplicated.
This avoids detecting a beacon as corrupt in the future
when new IEs are added that can be duplicated.
Cc: stable@vger.kernel.org
Signed-off-by: Paul Stewart <pstew@chromium.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
|
|
BugLink: http://bugs.launchpad.net/bugs/1056078
Profile replacement can cause long chains of profiles to build up when
the profile being replaced is pinned. When the pinned profile is finally
freed, it puts the reference to its replacement, which may in turn nest
another call to free_profile on the stack. Because this may happen for
each profile in the replacedby chain this can result in a recusion that
causes the stack to overflow.
Break this nesting by directly walking the chain of replacedby profiles
(ie. use iteration instead of recursion to free the list). This results
in at most 2 levels of free_profile being called, while freeing a
replacedby chain.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: James Morris <james.l.morris@oracle.com>
|
|
Fix a deadly typo in macro definition.
Cc: stable@vger.kernel.org
Signed-off-by: Hiro Sugawara <hsugawara@nvidia.com>
Signed-off-by: Hiroshi Doyu <hdoyu@nvidia.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
|
|
The current code is clearing it in all cases _except_ when zero.
Reported-by: Stanislav Kinsbursky <skinsbursky@parallels.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org
|
|
Commit e9406db20fecbfcab646bad157b4cfdc7cadddfb (lockd: per-net
NSM client creation and destruction helpers introduced) contains
a nasty race on initialisation of the per-net NSM client because
it doesn't check whether or not the client is set after grabbing
the nsm_create_mutex.
Reported-by: Nix <nix@esperi.org.uk>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org
|
|
Chris Perl reports that we're seeing races between the wakeup call in
xs_error_report and the connect attempts. Basically, Chris has shown
that in certain circumstances, the call to xs_error_report causes the
rpc_task that is responsible for reconnecting to wake up early, thus
triggering a disconnect and retry.
Since the sk->sk_error_report() calls in the socket layer are always
followed by a tcp_done() in the cases where we care about waking up
the rpc_tasks, just let the state_change callbacks take responsibility
for those wake ups.
Reported-by: Chris Perl <chris.perl@gmail.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org
Tested-by: Chris Perl <chris.perl@gmail.com>
|
|
The call to xprt_disconnect_done() that is triggered by a successful
connection reset will trigger another automatic wakeup of all tasks
on the xprt->pending rpc_wait_queue. In particular it will cause an
early wake up of the task that called xprt_connect().
All we really want to do here is clear all the socket-specific state
flags, so we split that functionality out of xs_sock_mark_closed()
into a helper that can be called by xs_abort_connection()
Reported-by: Chris Perl <chris.perl@gmail.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org
Tested-by: Chris Perl <chris.perl@gmail.com>
|
|
This reverts commit 55420c24a0d4d1fce70ca713f84aa00b6b74a70e.
Now that we clear the connected flag when entering TCP_CLOSE_WAIT,
the deadlock described in this commit is no longer possible.
Instead, the resulting call to xs_tcp_shutdown() can interfere
with pending reconnection attempts.
Reported-by: Chris Perl <chris.perl@gmail.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org
Tested-by: Chris Perl <chris.perl@gmail.com>
|
|
This is needed to ensure that we call xprt_connect() upon the next
call to call_connect().
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org
Tested-by: Chris Perl <chris.perl@gmail.com>
|
|
If none of the elements in scrubrates[] matches, this loop will cause
__amd64_set_scrub_rate() to incorrectly use the n+1th element.
As the function is designed to use the final scrubrates[] element in the
case of no match, we can fix this bug by simply terminating the array
search at the n-1th element.
Boris: this code is fragile anyway, see here why:
http://marc.info/?l=linux-kernel&m=135102834131236&w=2
It will be rewritten more robustly soonish.
Reported-by: Denis Kirjanov <kirjanov@gmail.com>
Cc: stable@vger.kernel.org
Cc: Doug Thompson <dougthompson@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
|
|
corruption
cryptd_queue_worker attempts to prevent simultaneous accesses to crypto
workqueue by cryptd_enqueue_request using preempt_disable/preempt_enable.
However cryptd_enqueue_request might be called from softirq context,
so add local_bh_disable/local_bh_enable to prevent data corruption and
panics.
Bug report at http://marc.info/?l=linux-crypto-vger&m=134858649616319&w=2
v2:
- Disable software interrupts instead of hardware interrupts
Cc: stable@vger.kernel.org
Reported-by: Gurucharan Shetty <gurucharan.shetty@gmail.com>
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
|
|
Like in the case of native hdmi, which is fixed already in
commit adf00b26d18e1b3570451296e03bcb20e4798cdd
Author: Paulo Zanoni <paulo.r.zanoni@intel.com>
Date: Tue Sep 25 13:23:34 2012 -0300
drm/i915: make sure we write all the DIP data bytes
we need to clear the entire sdvo buffer to avoid upsetting the
display.
Since infoframe buffer writing is now a bit more elaborate, extract it
into it's own function. This will be useful if we ever get around to
properly update the ELD for sdvo. Also #define proper names for the
two buffer indexes with fixed usage.
v2: Cite the right commit above, spotted by Paulo Zanoni.
v3: I'm too stupid to paste the right commit.
v4: Ben Hutchings noticed that I've failed to handle an underflow in
my loop logic, breaking it for i >= length + 8. Since I've just lost C
programmer license, use his solution. Also, make the frustrated 0-base
buffer size a notch more clear.
Reported-and-tested-by: Jürg Billeter <j@bitron.ch>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=25732
Cc: stable@vger.kernel.org
Cc: Paulo Zanoni <przanoni@gmail.com>
Cc: Ben Hutchings <ben@decadent.org.uk>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
|
|
Rather than always assuming the maximum possible BCLK rate will be
required generate BCLKs for stereo if either one or two channels is
enabled. In order to support this we also need to ensure that only
the relevant channels are enabled.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
|
|
list_move_tail(&schan->queued, &schan->active) makes the list_empty(schan->queued)
undefined, we either should change it to:
list_move_tail(schan->queued.next, &schan->active)
or
list_move_tail(&sdesc->node, &schan->active)
Signed-off-by: Barry Song <Baohua.Song@csr.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Vinod Koul <vinod.koul@linux.intel.com>
|
|
either DEV_TO_MEM or MEM_TO_DEV is supported, so change
OR to AND.
Signed-off-by: Barry Song <Baohua.Song@csr.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Vinod Koul <vinod.koul@linux.intel.com>
|
|
smp_irq_move_cleanup_interrupt
Posting this patch to fix an issue concerning sparse irq's that
I raised a while back. There was discussion about adding
refcounting to sparse irqs (to fix other potential race
conditions), but that does not appear to have been addressed
yet. This covers the only issue of this type that I've
encountered in this area.
A NULL pointer dereference can occur in
smp_irq_move_cleanup_interrupt() if we haven't yet setup the
irq_cfg pointer in the irq_desc.irq_data.chip_data.
In create_irq_nr() there is a window where we have set
vector_irq in __assign_irq_vector(), but not yet called
irq_set_chip_data() to set the irq_cfg pointer.
Should an IRQ_MOVE_CLEANUP_VECTOR hit the cpu in question during
this time, smp_irq_move_cleanup_interrupt() will attempt to
process the aforementioned irq, but panic when accessing
irq_cfg.
Only continue processing the irq if irq_cfg is non-NULL.
Signed-off-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Joerg Roedel <joerg.roedel@amd.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Alexander Gordeev <agordeev@redhat.com>
Link: http://lkml.kernel.org/r/20121016125021.GA22935@sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
The variable port is initialized but never used
otherwise, so remove the unused variable.
dpatch engine is used to auto generate this patch.
(https://github.com/weiyj/dpatch)
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Cc: Yan, Zheng <zheng.z.yan@intel.com>
Cc: a.p.zijlstra@chello.nl
Cc: paulus@samba.org
Cc: acme@ghostprotocols.net
Link: http://lkml.kernel.org/r/CAPgLHd8NZkYSkZm22FpZxiEh6HcA0q-V%3D29vdnheiDhgrJZ%2Byw@mail.gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Calling __pa() with an ioremap'd address is invalid. If we
encounter an efi_memory_desc_t without EFI_MEMORY_WB set in
->attribute we currently call set_memory_uc(), which in turn
calls __pa() on a potentially ioremap'd address.
On CONFIG_X86_32 this results in the following oops:
BUG: unable to handle kernel paging request at f7f22280
IP: [<c10257b9>] reserve_ram_pages_type+0x89/0x210
*pdpt = 0000000001978001 *pde = 0000000001ffb067 *pte = 0000000000000000
Oops: 0000 [#1] PREEMPT SMP
Modules linked in:
Pid: 0, comm: swapper Not tainted 3.0.0-acpi-efi-0805 #3
EIP: 0060:[<c10257b9>] EFLAGS: 00010202 CPU: 0
EIP is at reserve_ram_pages_type+0x89/0x210
EAX: 0070e280 EBX: 38714000 ECX: f7814000 EDX: 00000000
ESI: 00000000 EDI: 38715000 EBP: c189fef0 ESP: c189fea8
DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068
Process swapper (pid: 0, ti=c189e000 task=c18bbe60 task.ti=c189e000)
Stack:
80000200 ff108000 00000000 c189ff00 00038714 00000000 00000000 c189fed0
c104f8ca 00038714 00000000 00038715 00000000 00000000 00038715 00000000
00000010 38715000 c189ff48 c1025aff 38715000 00000000 00000010 00000000
Call Trace:
[<c104f8ca>] ? page_is_ram+0x1a/0x40
[<c1025aff>] reserve_memtype+0xdf/0x2f0
[<c1024dc9>] set_memory_uc+0x49/0xa0
[<c19334d0>] efi_enter_virtual_mode+0x1c2/0x3aa
[<c19216d4>] start_kernel+0x291/0x2f2
[<c19211c7>] ? loglevel+0x1b/0x1b
[<c19210bf>] i386_start_kernel+0xbf/0xc8
The only time we can call set_memory_uc() for a memory region is
when it is part of the direct kernel mapping. For the case where
we ioremap a memory region we must leave it alone.
This patch reimplements the fix from e8c7106280a3 ("x86, efi:
Calling __pa() with an ioremap()ed address is invalid") which
was reverted in e1ad783b12ec because it caused a regression on
some MacBooks (they hung at boot). The regression was caused
because the commit only marked EFI_RUNTIME_SERVICES_DATA as
E820_RESERVED_EFI, when it should have marked all regions that
have the EFI_MEMORY_RUNTIME attribute.
Despite first impressions, it's not possible to use
ioremap_cache() to map all cached memory regions on
CONFIG_X86_64 because of the way that the memory map might be
configured as detailed in the following bug report,
https://bugzilla.redhat.com/show_bug.cgi?id=748516
e.g. some of the EFI memory regions *need* to be mapped as part
of the direct kernel mapping.
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
Cc: Matthew Garrett <mjg@redhat.com>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Huang Ying <huang.ying.caritas@gmail.com>
Cc: Keith Packard <keithp@keithp.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/1350649546-23541-1-git-send-email-matt@console-pimps.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Emphasis the way tree_mod_log_insert_move avoids adding
MOD_LOG_KEY_REMOVE_WHILE_MOVING operations, depending on the direction of
the move operation.
Signed-off-by: Jan Schmidt <list.btrfs@jan-o-sch.net>
|
|
In get_old_root we grab a lock on the extent buffer before we obtain a
reference on that buffer. That order is changed now.
Signed-off-by: Jan Schmidt <list.btrfs@jan-o-sch.net>
|
|
In btrfs_find_all_roots' termination condition, we compare the level of the
old buffer we got from btrfs_search_old_slot to the level of the current
root node. We'd better compare it to the level of the rewinded root node.
Signed-off-by: Jan Schmidt <list.btrfs@jan-o-sch.net>
|
|
Tree mod log treated old root buffers as always empty buffers when starting
the rewind operations. However, the old root may still be part of the
current tree at a lower level, with still some valid entries.
Signed-off-by: Jan Schmidt <list.btrfs@jan-o-sch.net>
|
|
Although based on the Intel P6 design, the interrupt mechnanism
for KNC more closely resembles the Intel architectural
perfmon one.
We can't just re-use that code though, because KNC has different
MSR numbers for the status and ack registers.
In this case we just cut-and paste from perf_event_intel.c
with some minor changes, as it looks like it would not be
worth the trouble to change that code to be MSR-configurable.
Signed-off-by: Vince Weaver <vincent.weaver@maine.edu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Cc: eranian@gmail.com
Cc: Meadows Lawrence F <lawrence.f.meadows@intel.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1210171304410.23243@vincent-weaver-1.um.maine.edu
[ Small stylistic edits. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
x86_pmu.enable() is called from x86_pmu_enable() with
cpuc->enabled set to 0. This means we weren't re-enabling the
counters after a context switch.
This patch just removes the check, as it should't be necessary
(and the equivelent x86_ generic code does not have the checks).
The origin of this problem is the KNC driver being based on the
P6 one. The P6 driver also has this issue, but works anyway
due to various lucky accidents.
Signed-off-by: Vince Weaver <vincent.weaver@maine.edu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Cc: eranian@gmail.com
Cc: Meadows
Cc: Lawrence F <lawrence.f.meadows@intel.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1210171303290.23243@vincent-weaver-1.um.maine.edu
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Early versions of Intel KNC chips have a bug where bits above 32
were not properly set. We worked around this by only using the
bottom 32 bits (out of 40 that should be available).
It turns out this workaround breaks overflow handling.
The buggy silicon will in theory never be used in production
systems, so remove this workaround so we get proper overflow
support.
Signed-off-by: Vince Weaver <vincent.weaver@maine.edu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Cc: eranian@gmail.com
Cc: Meadows Lawrence F <lawrence.f.meadows@intel.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1210171302140.23243@vincent-weaver-1.um.maine.edu
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
This, beyond handling corner cases, also fixes some build warnings:
arch/x86/kernel/cpu/perf_event_intel_uncore.c: In function ‘snbep_uncore_pci_disable_box’:
arch/x86/kernel/cpu/perf_event_intel_uncore.c:124:9: warning: ‘config’ is used uninitialized in this function [-Wuninitialized]
arch/x86/kernel/cpu/perf_event_intel_uncore.c: In function ‘snbep_uncore_pci_enable_box’:
arch/x86/kernel/cpu/perf_event_intel_uncore.c:135:9: warning: ‘config’ is used uninitialized in this function [-Wuninitialized]
arch/x86/kernel/cpu/perf_event_intel_uncore.c: In function ‘snbep_uncore_pci_read_counter’:
arch/x86/kernel/cpu/perf_event_intel_uncore.c:164:2: warning: ‘count’ is used uninitialized in this function [-Wuninitialized]
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
Cc: a.p.zijlstra@chello.nl
Link: http://lkml.kernel.org/r/1351068140-13456-1-git-send-email-zheng.z.yan@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
This patch adds the support for 4 new PCI boards based on the SJA1000 CAN
controller, from PEAK-System Technik:
. PCAN-miniPCIe (PCI-Express Mini slots, available as single or dual-channel)
. PCAN-cPCI (CompactPCI format, available as single or dual-channel)
. PCAN-PC/104-Plus (PC/104-Plus system, av. as 1, 2 or 4 channels)
. PCAN-PCI/104-Express (PCI/104-Express system, av. as 1 or 2 channels)
This patch also fixes a typo in existing "PEAK_MPCI_DEVICE_ID" identifier
(missing "e" for Express). Finally, it also changes the author as well as it
updates the module supported devices list.
Signed-off-by: Stephane Grosjean <s.grosjean@peak-system.com>
Acked-by: Oliver Hartkopp <socketcan@hartkopp.net>
Acked-by: Wolfgang Grandegger <wg@grandegger.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
|
|
Commit 20167d3421a089a1bf1bd680b150dc69c9506810 ("x86-64: Fix
accounting in kernel_physical_mapping_init()") went a little too
far by entirely removing the counting of pre-populated page
tables: this should be done at boot time (to cover the page
tables set up in early boot code), but shouldn't be done during
memory hot add.
Hence, re-add the removed increments of "pages", but make them
and the one in phys_pte_init() conditional upon !after_bootmem.
Reported-Acked-and-Tested-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Cc: <stable@kernel.org>
Link: http://lkml.kernel.org/r/506DAFBA020000780009FA8C@nat28.tlf.novell.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Between 2.6.33 and 2.6.34 the PMU code was made modular.
The x86_pmu_enable() call was extended to disable cpuc->enabled
and iterate the counters, enabling one at a time, before calling
enable_all() at the end, followed by re-enabling cpuc->enabled.
Since cpuc->enabled was set to 0, that change effectively caused
the "val |= ARCH_PERFMON_EVENTSEL_ENABLE;" code in p6_pmu_enable_event()
and p6_pmu_disable_event() to be dead code that was never called.
This change removes this code (which was confusing) and adds some
extra commentary to make it more clear what is going on.
Signed-off-by: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1210191732000.14552@vincent-weaver-1.um.maine.edu
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
This patch updates the generic events on p6, including some new
extended cache events.
Values for these events were taken from the equivelant PAPI
predefined events.
Tested on a Pentium II.
Signed-off-by: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1210191730080.14552@vincent-weaver-1.um.maine.edu
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
According to Intel SDM Volume 3B, FP_ASSIST is limited to Counter 1 only,
not Counter 0.
Tested on a Pentium II.
Signed-off-by: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1210191728570.14552@vincent-weaver-1.um.maine.edu
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Add some scribbles on how and why the load-balancer works..
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1341316406.23484.64.camel@twins
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
While per-entity load-tracking is generally useful, beyond computing shares
distribution, e.g. runnable based load-balance (in progress), governors,
power-management, etc.
These facilities are not yet consumers of this data. This may be trivially
reverted when the information is required; but avoid paying the overhead for
calculations we will not use until then.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141507.422162369@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
__update_entity_runnable_avg forms the core of maintaining an entity's runnable
load average. In this function we charge the accumulated run-time since last
update and handle appropriate decay. In some cases, e.g. a waking task, this
time interval may be much larger than our period unit.
Fortunately we can exploit some properties of our series to perform decay for a
blocked update in constant time and account the contribution for a running
update in essentially-constant* time.
[*]: For any running entity they should be performing updates at the tick which
gives us a soft limit of 1 jiffy between updates, and we can compute up to a
32 jiffy update in a single pass.
C program to generate the magic constants in the arrays:
#include <math.h>
#include <stdio.h>
#define N 32
#define WMULT_SHIFT 32
const long WMULT_CONST = ((1UL << N) - 1);
double y;
long runnable_avg_yN_inv[N];
void calc_mult_inv() {
int i;
double yn = 0;
printf("inverses\n");
for (i = 0; i < N; i++) {
yn = (double)WMULT_CONST * pow(y, i);
runnable_avg_yN_inv[i] = yn;
printf("%2d: 0x%8lx\n", i, runnable_avg_yN_inv[i]);
}
printf("\n");
}
long mult_inv(long c, int n) {
return (c * runnable_avg_yN_inv[n]) >> WMULT_SHIFT;
}
void calc_yn_sum(int n)
{
int i;
double sum = 0, sum_fl = 0, diff = 0;
/*
* We take the floored sum to ensure the sum of partial sums is never
* larger than the actual sum.
*/
printf("sum y^n\n");
printf(" %8s %8s %8s\n", "exact", "floor", "error");
for (i = 1; i <= n; i++) {
sum = (y * sum + y * 1024);
sum_fl = floor(y * sum_fl+ y * 1024);
printf("%2d: %8.0f %8.0f %8.0f\n", i, sum, sum_fl,
sum_fl - sum);
}
printf("\n");
}
void calc_conv(long n) {
long old_n;
int i = -1;
printf("convergence (LOAD_AVG_MAX, LOAD_AVG_MAX_N)\n");
do {
old_n = n;
n = mult_inv(n, 1) + 1024;
i++;
} while (n != old_n);
printf("%d> %ld\n", i - 1, n);
printf("\n");
}
void main() {
y = pow(0.5, 1/(double)N);
calc_mult_inv();
calc_conv(1024);
calc_yn_sum(N);
}
[ Compile with -lm ]
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141507.277808946@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Now that our measurement intervals are small (~1ms) we can amortize the posting
of update_shares() to be about each period overflow. This is a large cost
saving for frequently switching tasks.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141507.200772172@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Now that running entities maintain their own load-averages the work we must do
in update_shares() is largely restricted to the periodic decay of blocked
entities. This allows us to be a little less pessimistic regarding our
occupancy on rq->lock and the associated rq->clock updates required.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141507.133999170@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Now that the machinery in place is in place to compute contributed load in a
bottom up fashion; replace the shares distribution code within update_shares()
accordingly.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141507.061208672@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
With bandwidth control tracked entities may cease execution according to user
specified bandwidth limits. Charging this time as either throttled or blocked
however, is incorrect and would falsely skew in either direction.
What we actually want is for any throttled periods to be "invisible" to
load-tracking as they are removed from the system for that interval and
contribute normally otherwise.
Do this by moderating the progression of time to omit any periods in which the
entity belonged to a throttled hierarchy.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141506.998912151@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Entities of equal weight should receive equitable distribution of cpu time.
This is challenging in the case of a task_group's shares as execution may be
occurring on multiple cpus simultaneously.
To handle this we divide up the shares into weights proportionate with the load
on each cfs_rq. This does not however, account for the fact that the sum of
the parts may be less than one cpu and so we need to normalize:
load(tg) = min(runnable_avg(tg), 1) * tg->shares
Where runnable_avg is the aggregate time in which the task_group had runnable
children.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141506.930124292@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Unlike task entities who have a fixed weight, group entities instead own a
fraction of their parenting task_group's shares as their contributed weight.
Compute this fraction so that we can correctly account hierarchies and shared
entity nodes.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141506.855074415@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Maintain a global running sum of the average load seen on each cfs_rq belonging
to each task group so that it may be used in calculating an appropriate
shares:weight distribution.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141506.792901086@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
When a running entity blocks we migrate its tracked load to
cfs_rq->blocked_runnable_avg. In the sleep case this occurs while holding
rq->lock and so is a natural transition. Wake-ups however, are potentially
asynchronous in the presence of migration and so special care must be taken.
We use an atomic counter to track such migrated load, taking care to match this
with the previously introduced decay counters so that we don't migrate too much
load.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141506.726077467@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Since we are now doing bottom up load accumulation we need explicit
notification when a task has been re-parented so that the old hierarchy can be
updated.
Adds: migrate_task_rq(struct task_struct *p, int next_cpu)
(The alternative is to do this out of __set_task_cpu, but it was suggested that
this would be a cleaner encapsulation.)
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141506.660023400@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
We are currently maintaining:
runnable_load(cfs_rq) = \Sum task_load(t)
For all running children t of cfs_rq. While this can be naturally updated for
tasks in a runnable state (as they are scheduled); this does not account for
the load contributed by blocked task entities.
This can be solved by introducing a separate accounting for blocked load:
blocked_load(cfs_rq) = \Sum runnable(b) * weight(b)
Obviously we do not want to iterate over all blocked entities to account for
their decay, we instead observe that:
runnable_load(t) = \Sum p_i*y^i
and that to account for an additional idle period we only need to compute:
y*runnable_load(t).
This means that we can compute all blocked entities at once by evaluating:
blocked_load(cfs_rq)` = y * blocked_load(cfs_rq)
Finally we maintain a decay counter so that when a sleeping entity re-awakens
we can determine how much of its load should be removed from the blocked sum.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141506.585389902@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
For a given task t, we can compute its contribution to load as:
task_load(t) = runnable_avg(t) * weight(t)
On a parenting cfs_rq we can then aggregate:
runnable_load(cfs_rq) = \Sum task_load(t), for all runnable children t
Maintain this bottom up, with task entities adding their contributed load to
the parenting cfs_rq sum. When a task entity's load changes we add the same
delta to the maintained sum.
Signed-off-by: Paul Turner <pjt@google.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141506.514678907@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Since runqueues do not have a corresponding sched_entity we instead embed a
sched_avg structure directly.
Signed-off-by: Ben Segall <bsegall@google.com>
Reviewed-by: Paul Turner <pjt@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120823141506.442637130@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
