If the previous CPU is cache affine and idle, select it.
The current implementation simply traverses the sd_llc domain,
taking the first idle CPU encountered, which walks buddy pairs
hand in hand over the package, inflicting excruciating pain.
1 tbench pair (worst case) in a 10 core + SMT package:
pre 15.22 MB/sec 1 procs
post 252.01 MB/sec 1 procs
Signed-off-by: Mike Galbraith <bitbucket@online.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1359371965.5783.127.camel@marge.simpson.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Function next_prio() has been removed and pull_rt_task() is the
only user of pick_next_highest_task_rt() at the moment.
pull_rt_task is not interested in p->nr_cpus_allowed, its only
interest is the fact that cpu is allowed to execute p. If
nr_cpus_allowed == 1, cpu != task_cpu(p) and cpu is allowed then
it means that task p is in the middle of the migration
techniques; the task waits until it is moved by migration
thread. So, lets pull it earlier.
Signed-off-by: Kirill V Tkhai <tkhai@yandex.ru>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Peter Zijlstra <peterz@infradead.org>
CC: linux-rt-users <linux-rt-users@vger.kernel.org>
Link: http://lkml.kernel.org/r/70871359644177@web16d.yandex.ru
Signed-off-by: Ingo Molnar <mingo@kernel.org>
There are several places of consecutive calls of
dequeue_task_rt() and put_prev_task_rt() in the scheduler.
For example, function rt_mutex_setprio() does it.
The both calls lead to update_curr_rt(), the second of it
receives zeroed delta_exec. The only effective action in this
case is call of sched_rt_avg_update(), which can change
rq->age_stamp and rq->rt_avg. But it is possible in case of
""floating"" rq->clock. This fact is not reasonable to be
accounted. Another actions do nothing.
Signed-off-by: Kirill V Tkhai <tkhai@yandex.ru>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Peter Zijlstra <peterz@infradead.org>
CC: linux-rt-users <linux-rt-users@vger.kernel.org>
Link: http://lkml.kernel.org/r/931541359550236@web1g.yandex.ru
Signed-off-by: Ingo Molnar <mingo@kernel.org>
We scale stime, utime values based on rtime (sum_exec_runtime
converted to jiffies). During scaling we multiple rtime * utime,
which seems to be fine, since both values are converted to u64,
but it's not.
Let assume HZ is 1000 - 1ms tick. Process consist of 64 threads,
run for 1 day, threads utilize 100% cpu on user space. Machine
has 64 cpus.
Process rtime = utime will be 64 * 24 * 60 * 60 * 1000 jiffies,
which is 0x149970000. Multiplication rtime * utime result is
0x1a855771100000000, which can not be covered in 64 bits.
Result of overflow is stall of utime values visible in user
space (prev_utime in kernel), even if application still consume
lot of CPU time.
A solution to solve this is to perform the multiplication on
stime instead of utime. It's easy to grow the utime value fast
with a CPU bound thread in userspace for example. Now we assume
that doing so with stime is much harder. In most cases a task
shouldn't ever spend much time in kernel space as it tends to
sleep waiting for jobs completion when they take long to
achieve. IO is the typical example of that.
Hence scaling the cputime by performing the multiplication on
stime instead of utime should considerably reduce the chances of
an overflow on most workloads.
This is largely inspired by a patch from Stanislaw Gruszka:
http://lkml.kernel.org/r/20130107113144.GA7544@redhat.com
Inspired-by: Stanislaw Gruszka <sgruszka@redhat.com>
Reported-by: Stanislaw Gruszka <sgruszka@redhat.com>
Acked-by: Stanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/1359217182-25184-1-git-send-email-fweisbec@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The issue below was found in 2.6.34-rt rather than mainline rt
kernel, but the issue still exists upstream as well.
So please let me describe how it was noticed on 2.6.34-rt:
On this version, each softirq has its own thread, it means there
is at least one RT FIFO task per cpu. The priority of these
tasks is set to 49 by default. If user launches an RT FIFO task
with priority lower than 49 of softirq RT tasks, it's possible
there are two RT FIFO tasks enqueued one cpu runqueue at one
moment. By current strategy of balancing RT tasks, when it comes
to RT tasks, we really need to put them off to a CPU that they
can run on as soon as possible. Even if it means a bit of cache
line flushing, we want RT tasks to be run with the least latency.
When the user RT FIFO task which just launched before is
running, the sched timer tick of the current cpu happens. In this
tick period, the timeout value of the user RT task will be
updated once. Subsequently, we try to wake up one softirq RT
task on its local cpu. As the priority of current user RT task
is lower than the softirq RT task, the current task will be
preempted by the higher priority softirq RT task. Before
preemption, we check to see if current can readily move to a
different cpu. If so, we will reschedule to allow the RT push logic
to try to move current somewhere else. Whenever the woken
softirq RT task runs, it first tries to migrate the user FIFO RT
task over to a cpu that is running a task of lesser priority. If
migration is done, it will send a reschedule request to the found
cpu by IPI interrupt. Once the target cpu responds the IPI
interrupt, it will pick the migrated user RT task to preempt its
current task. When the user RT task is running on the new cpu,
the sched timer tick of the cpu fires. So it will tick the user
RT task again. This also means the RT task timeout value will be
updated again. As the migration may be done in one tick period,
it means the user RT task timeout value will be updated twice
within one tick.
If we set a limit on the amount of cpu time for the user RT task
by setrlimit(RLIMIT_RTTIME), the SIGXCPU signal should be posted
upon reaching the soft limit.
But exactly when the SIGXCPU signal should be sent depends on the
RT task timeout value. In fact the timeout mechanism of sending
the SIGXCPU signal assumes the RT task timeout is increased once
every tick.
However, currently the timeout value may be added twice per
tick. So it results in the SIGXCPU signal being sent earlier
than expected.
To solve this issue, we prevent the timeout value from increasing
twice within one tick time by remembering the jiffies value of
last updating the timeout. As long as the RT task's jiffies is
different with the global jiffies value, we allow its timeout to
be updated.
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Fan Du <fan.du@windriver.com>
Reviewed-by: Yong Zhang <yong.zhang0@gmail.com>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1342508623-2887-1-git-send-email-ying.xue@windriver.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The caller of sched_sliced() should pass se.cfs_rq and se as the
arguments, however in sched_rr_get_interval() we gave it
rq.cfs_rq and se, which made the following computation obviously
wrong.
The change was introduced by commit:
77034937dc sched: fix crash in sys_sched_rr_get_interval()
... 5 years ago, while it had been the correct 'cfs_rq_of' before
the commit. The change seems to be irrelevant to the commit
msg, which was to return a 0 timeslice for tasks that are on an
idle runqueue. So I believe that was just a plain typo.
Signed-off-by: Zhu Yanhai <gaoyang.zyh@taobao.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Paul Turner <pjt@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/1357621012-15039-1-git-send-email-gaoyang.zyh@taobao.com
[ Since this is an ABI and an old bug, we'll test this via a
slow upstream route, to hopefully discover any app breakage. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull more USB fixes from Greg Kroah-Hartman:
"Here are some more USB fixes for the 3.8-rc4 tree.
Some gadget driver fixes, and finally resolved the ehci-mxc driver
build issues (it's just some code moving around and being deleted)."
* tag 'usb-3.8-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
USB: EHCI: fix build error in ehci-mxc
USB: EHCI: add a name for the platform-private field
USB: EHCI: fix incorrect configuration test
USB: EHCI: Move definition of EHCI_STATS to ehci.h
USB: UHCI: fix IRQ race during initialization
usb: gadget: FunctionFS: Fix missing braces in parse_opts
usb: dwc3: gadget: fix ep->maxburst for ep0
ARM: i.MX clock: Change the connection-id for fsl-usb2-udc
usb: gadget: fsl_mxc_udc: replace MX35_IO_ADDRESS to ioremap
usb: gadget: fsl-mxc-udc: replace cpu_is_xxx() with platform_device_id
usb: musb: cppi_dma: drop '__init' annotation
Pull drivers/misc fix from Greg Kroah-Hartman:
"Here is a single revert for the ti-st misc driver, fixing problem that
was introduced in 3.7-rc1 that has been bothering people."
* tag 'char-misc-3.8-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
Revert "drivers/misc/ti-st: remove gpio handling"
Pull a TTY maintainer patch from Greg Kroah-Hartman:
"Just a MAINTAINERS update, now that Alan has left for a bit, I'll
continue to watch over the serial drivers."
* tag 'tty-3.8-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
MAINTAINERS: Someone needs to watch over the serial drivers
Pull media fixes from Mauro Carvalho Chehab:
- gspca: add needed delay for I2C traffic for sonixb/sonixj cameras
- gspca: add one missing Kinect USB ID
- usbvideo: some regression fixes
- omap3isp: fix some build issues
- videobuf2: fix video output handling
- exynos s5p/m5mols: a few regression fixes.
* 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
[media] uvcvideo: Set error_idx properly for S_EXT_CTRLS failures
[media] uvcvideo: Cleanup leftovers of partial revert
[media] uvcvideo: Return -EACCES when trying to set a read-only control
[media] omap3isp: Don't include <plat/cpu.h>
[media] s5p-mfc: Fix interrupt error handling routine
[media] s5p-fimc: Fix return value of __fimc_md_create_flite_source_links()
[media] m5mols: Fix typo in get_fmt callback
[media] v4l: vb2: Set data_offset to 0 for single-plane output buffers
[media] [FOR,v3.8] omap3isp: Don't include deleted OMAP plat/ header files
[media] gspca_sonixj: Add a small delay after i2c_w1
[media] gspca_sonixb: Properly wait between i2c writes
[media] gspca_kinect: add Kinect for Windows USB id
Pull m68k fixes from Geert Uytterhoeven:
"The asm-generic changeset has been ack'ed by Arnd."
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k:
m68k: Wire up finit_module
asm-generic/dma-mapping-broken.h: Provide dma_alloc_attrs()/dma_free_attrs()
m68k: Provide dma_alloc_attrs()/dma_free_attrs()
Pull arm64 fixes from Catalin Marinas:
- ELF coredump fix (more registers dumped than what user space expects)
- SUBARCH name generation (s/aarch64/arm64/)
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/cmarinas/linux-aarch64:
arm64: makefile: fix uname munging when setting ARCH on native machine
arm64: elf: fix core dumping to match what glibc expects
This patch (as1643b) fixes a build error in ehci-hcd when compiling for
ARM with allmodconfig:
drivers/usb/host/ehci-hcd.c:1285:0: warning: "PLATFORM_DRIVER" redefined [enabled by default]
drivers/usb/host/ehci-hcd.c:1255:0: note: this is the location of the previous definition
drivers/usb/host/ehci-mxc.c:280:31: warning: 'ehci_mxc_driver' defined but not used [-Wunused-variable]
drivers/usb/host/ehci-hcd.c:1285:0: warning: "PLATFORM_DRIVER" redefined [enabled by default]
drivers/usb/host/ehci-hcd.c:1255:0: note: this is the location of the previous definition
The fix is to convert ehci-mxc over to the new "ehci-hcd is a library"
scheme so that it can coexist peacefully with the ehci-platform
driver. As part of the conversion the ehci_mxc_priv data structure,
which was allocated dynamically, is now placed where it belongs: in
the private area at the end of struct ehci_hcd.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Tested-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Pull sound fixes from Takashi Iwai:
"Only a few small HD-audio fixes:
- Addition of new Conexant codec IDs
- Two one-liners to add fixups for Realtek codecs
- A last-minute regression fix for auto-mute with power-saving mode
(regressed since 3.8-rc1)"
* tag 'sound-3.8' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
ALSA: hda - Fix inconsistent pin states after resume
ALSA: hda - Add Conexant CX20755/20756/20757 codec IDs
ALSA: hda - Add fixup for Acer AO725 laptop
ALSA: hda - Fix mute led for another HP machine
The commit [26a6cb6c: ALSA: hda - Implement a poll loop for jacks as a
module parameter] introduced the polling jack detection code, but it
also moved the call of snd_hda_jack_set_dirty_all() in the resume path
after resume/init ops call. This caused a regression when the jack
state has been changed during power-down (e.g. in the power save
mode). Since the driver doesn't probe the new jack state but keeps
using the cached value due to no dirty flag, the pin state remains
also as if the jack is still plugged.
The fix is simply moving snd_hda_jack_set_dirty_all() to the original
position.
Reported-by: Manolo Díaz <diaz.manolo@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This reverts commit eccf2979b2.
The reason is that it broke TI WiLink shared transport on Panda.
Also, callback functions should not be added to board files anymore,
so revert to implementing the power functions in the driver itself.
Additionally, changed a variable name ('status' to 'err') so that this
revert compiles properly.
Cc: stable <stable@vger.kernel.org> [3.7]
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Luciano Coelho <coelho@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Pull PCI updates from Bjorn Helgaas:
"The most important is a fix for a pciehp deadlock that occurs when
unplugging a Thunderbolt adapter. We also applied the same fix to
shpchp, removed CONFIG_EXPERIMENTAL dependencies, fixed a
pcie_aspm=force problem, and fixed a refcount leak.
Details:
- Hotplug
PCI: pciehp: Use per-slot workqueues to avoid deadlock
PCI: shpchp: Make shpchp_wq non-ordered
PCI: shpchp: Handle push button event asynchronously
PCI: shpchp: Use per-slot workqueues to avoid deadlock
- Power management
PCI: Allow pcie_aspm=force even when FADT indicates it is unsupported
- Misc
PCI/AER: pci_get_domain_bus_and_slot() call missing required pci_dev_put()
PCI: remove depends on CONFIG_EXPERIMENTAL"
* tag '3.8-pci-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
PCI: remove depends on CONFIG_EXPERIMENTAL
PCI: Allow pcie_aspm=force even when FADT indicates it is unsupported
PCI: shpchp: Use per-slot workqueues to avoid deadlock
PCI: shpchp: Handle push button event asynchronously
PCI: shpchp: Make shpchp_wq non-ordered
PCI/AER: pci_get_domain_bus_and_slot() call missing required pci_dev_put()
PCI: pciehp: Use per-slot workqueues to avoid deadlock
Commit 083b804c4d ("async: use workqueue for worker pool") made it
possible that async jobs are moved from pending to running out-of-order.
While pending async jobs will be queued and dispatched for execution in
the same order, nothing guarantees they'll enter "1) move self to the
running queue" of async_run_entry_fn() in the same order.
Before the conversion, async implemented its own worker pool. An async
worker, upon being woken up, fetches the first item from the pending
list, which kept the executing lists sorted. The conversion to
workqueue was done by adding work_struct to each async_entry and async
just schedules the work item. The queueing and dispatching of such work
items are still in order but now each worker thread is associated with a
specific async_entry and moves that specific async_entry to the
executing list. So, depending on which worker reaches that point
earlier, which is non-deterministic, we may end up moving an async_entry
with larger cookie before one with smaller one.
This broke __lowest_in_progress(). running->domain may not be properly
sorted and is not guaranteed to contain lower cookies than pending list
when not empty. Fix it by ensuring sort-inserting to the running list
and always looking at both pending and running when trying to determine
the lowest cookie.
Over time, the async synchronization implementation became quite messy.
We better restructure it such that each async_entry is linked to two
lists - one global and one per domain - and not move it when execution
starts. There's no reason to distinguish pending and running. They
behave the same for synchronization purposes.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull perf/urgent fixes from Arnaldo Carvalho de Melo:
. revert 20b279 - require exclude_guest to use PEBS - kernel side, now
older binaries will continue working for things like cycles:pp
without needing to pass extra modifiers, from David Ahern.
. Fix building from 'make perf-*-src-pkg' tarballs, broken by UAPI,
from Sebastian Andrzej Siewior
[ Pulling directly, Ingo would normally pull but has been unresponsive ]
* tag 'perf-urgent-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
perf tools: Fix building from 'make perf-*-src-pkg' tarballs
perf x86: revert 20b279 - require exclude_guest to use PEBS - kernel side
Pull parisc fixes from Helge Deller:
"Improve the stability of the linux kernel on the parisc architecture"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:
parisc: sigaltstack doesn't round ss.ss_sp as required
parisc: improve ptrace support for gdb single-step
parisc: don't claim cpu irqs more than once
parisc: avoid undefined shift in cnv_float.h
Pull fuse fixes from Miklos Szeredi:
"This contain a bugfix for CUSE and miscellaneous small fixes"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:
fuse: remove unused variable in fuse_try_move_page()
fuse: make fuse_file_fallocate() static
fuse: Move CUSE Kconfig entry from fs/Kconfig into fs/fuse/Kconfig
cuse: fix uninitialized variable warnings
cuse: do not register multiple devices with identical names
cuse: use mutex as registration lock instead of spinlocks
Mike Galbraith
Kirill Tkhai
Frederic Weisbecker
Ying Xue
Viresh Kumar
Zhu Yanhai
Linus Torvalds
Greg Kroah-Hartman
Alan Stern
Alan Cox
Takashi Iwai
Luciano Coelho
Tejun Heo