Pull scheduler changes from Ingo Molnar:
- Add the initial implementation of SCHED_DEADLINE support: a real-time
scheduling policy where tasks that meet their deadlines and
periodically execute their instances in less than their runtime quota
see real-time scheduling and won't miss any of their deadlines.
Tasks that go over their quota get delayed (Available to privileged
users for now)
- Clean up and fix preempt_enable_no_resched() abuse all around the
tree
- Do sched_clock() performance optimizations on x86 and elsewhere
- Fix and improve auto-NUMA balancing
- Fix and clean up the idle loop
- Apply various cleanups and fixes
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (60 commits)
sched: Fix __sched_setscheduler() nice test
sched: Move SCHED_RESET_ON_FORK into attr::sched_flags
sched: Fix up attr::sched_priority warning
sched: Fix up scheduler syscall LTP fails
sched: Preserve the nice level over sched_setscheduler() and sched_setparam() calls
sched/core: Fix htmldocs warnings
sched/deadline: No need to check p if dl_se is valid
sched/deadline: Remove unused variables
sched/deadline: Fix sparse static warnings
m68k: Fix build warning in mac_via.h
sched, thermal: Clean up preempt_enable_no_resched() abuse
sched, net: Fixup busy_loop_us_clock()
sched, net: Clean up preempt_enable_no_resched() abuse
sched/preempt: Fix up missed PREEMPT_NEED_RESCHED folding
sched/preempt, locking: Rework local_bh_{dis,en}able()
sched/clock, x86: Avoid a runtime condition in native_sched_clock()
sched/clock: Fix up clear_sched_clock_stable()
sched/clock, x86: Use a static_key for sched_clock_stable
sched/clock: Remove local_irq_disable() from the clocks
sched/clock, x86: Rewrite cyc2ns() to avoid the need to disable IRQs
...
Pull perf updates from Ingo Molnar:
"Kernel side changes:
- Add Intel RAPL energy counter support (Stephane Eranian)
- Clean up uprobes (Oleg Nesterov)
- Optimize ring-buffer writes (Peter Zijlstra)
Tooling side changes, user visible:
- 'perf diff':
- Add column colouring improvements (Ramkumar Ramachandra)
- 'perf kvm':
- Add guest related improvements, including allowing to specify a
directory with guest specific /proc information (Dongsheng Yang)
- Add shell completion support (Ramkumar Ramachandra)
- Add '-v' option (Dongsheng Yang)
- Support --guestmount (Dongsheng Yang)
- 'perf probe':
- Support showing source code, asking for variables to be collected
at probe time and other 'perf probe' operations that use DWARF
information.
This supports only binaries with debugging information at this
time, detached debuginfo (aka debuginfo packages) support should
come in later patches (Masami Hiramatsu)
- 'perf record':
- Rename --no-delay option to --no-buffering, better reflecting its
purpose and freeing up '--delay' to take the place of
'--initial-delay', so that 'record' and 'stat' are consistent
(Arnaldo Carvalho de Melo)
- Default the -t/--thread option to no inheritance (Adrian Hunter)
- Make per-cpu mmaps the default (Adrian Hunter)
- 'perf report':
- Improve callchain processing performance (Frederic Weisbecker)
- Retain bfd reference to lookup source line numbers, greatly
optimizing, among other use cases, 'perf report -s srcline'
(Adrian Hunter)
- Improve callchain processing performance even more (Namhyung Kim)
- Add a perf.data file header window in the 'perf report' TUI,
associated with the 'i' hotkey, providing a counterpart to the
--header option in the stdio UI (Namhyung Kim)
- 'perf script':
- Add an option in 'perf script' to print the source line number
(Adrian Hunter)
- Add --header/--header-only options to 'script' and 'report', the
default is not tho show the header info, but as this has been the
default for some time, leave a single line explaining how to
obtain that information (Jiri Olsa)
- Add options to show comm, fork, exit and mmap PERF_RECORD_ events
(Namhyung Kim)
- Print callchains and symbols if they exist (David Ahern)
- 'perf timechart'
- Add backtrace support to CPU info
- Print pid along the name
- Add support for CPU topology
- Add new option --highlight'ing threads, be it by name or, if a
numeric value is provided, that run more than given duration
(Stanislav Fomichev)
- 'perf top':
- Make 'perf top -g' refer to callchains, for consistency with
other tools (David Ahern)
- 'perf trace':
- Handle old kernels where the "raw_syscalls" tracepoints were
called plain "syscalls" (David Ahern)
- Remove thread summary coloring, by Pekka Enberg.
- Honour -m option in 'trace', the tool was offering the option to
set the mmap size, but wasn't using it when doing the actual mmap
on the events file descriptors (Jiri Olsa)
- generic:
- Backport libtraceevent plugin support (trace-cmd repository, with
plugins for jbd2, hrtimer, kmem, kvm, mac80211, sched_switch,
function, xen, scsi, cfg80211 (Jiri Olsa)
- Print session information only if --stdio is given (Namhyung Kim)
Tooling side changes, developer visible (plumbing):
- Improve 'perf probe' exit path, release resources (Masami
Hiramatsu)
- Improve libtraceevent plugins exit path, allowing the registering
of an unregister handler to be called at exit time (Namhyung Kim)
- Add an alias to the build test makefile (make -C tools/perf
build-test) (Namhyung Kim)
- Get rid of die() and friends (good riddance!) in libtraceevent
(Namhyung Kim)
- Fix cross build problems related to pkgconfig and CROSS_COMPILE not
being propagated to the feature tests, leading to features being
tested in the host and then being enabled on the target (Mark
Rutland)
- Improve forked workload error reporting by sending the errno in the
signal data queueing integer field, using sigqueue and by doing the
signal setup in the evlist methods, removing open coded equivalents
in various tools (Arnaldo Carvalho de Melo)
- Do more auto exit cleanup chores in the 'evlist' destructor, so
that the tools don't have to all do that sequence (Arnaldo Carvalho
de Melo)
- Pack 'struct perf_session_env' and 'struct trace' (Arnaldo Carvalho
de Melo)
- Add test for building detached source tarballs (Arnaldo Carvalho de
Melo)
- Move some header files (tools/perf/ to tools/include/ to make them
available to other tools/ dwelling codebases (Namhyung Kim)
- Move logic to warn about kptr_restrict'ed kernels to separate
function in 'report' (Arnaldo Carvalho de Melo)
- Move hist browser selection code to separate function (Arnaldo
Carvalho de Melo)
- Move histogram entries collapsing to separate function (Arnaldo
Carvalho de Melo)
- Introduce evlist__for_each() & friends (Arnaldo Carvalho de Melo)
- Automate setup of FEATURE_CHECK_(C|LD)FLAGS-all variables (Jiri
Olsa)
- Move arch setup into seprate Makefile (Jiri Olsa)
- Make libtraceevent install target quieter (Jiri Olsa)
- Make tests/make output more compact (Jiri Olsa)
- Ignore generated files in feature-checks (Chunwei Chen)
- Introduce pevent_filter_strerror() in libtraceevent, similar in
purpose to libc's strerror() function (Namhyung Kim)
- Use perf_data_file methods to write output file in 'record' and
'inject' (Jiri Olsa)
- Use pr_*() functions where applicable in 'report' (Namhyumg Kim)
- Add 'machine' 'addr_location' struct to have full picture (machine,
thread, map, symbol, addr) for a (partially) resolved address,
reducing function signatures (Arnaldo Carvalho de Melo)
- Reduce code duplication in the histogram entry creation/insertion
(Arnaldo Carvalho de Melo)
- Auto allocate annotation histogram data structures (Arnaldo
Carvalho de Melo)
- No need to test against NULL before calling free, also set freed
memory in struct pointers to NULL, to help fixing use after free
bugs (Arnaldo Carvalho de Melo)
- Rename some struct DSO binary_type related members and methods, to
clarify its purpose and need for differentiation (symtab_type, ie
one is about the files .text, CFI, etc, i.e. its binary contents,
and the other is about where the symbol table came from (Arnaldo
Carvalho de Melo)
- Convert to new topic libraries, starting with an API one (sysfs,
debugfs, etc), renaming liblk in the process (Borislav Petkov)
- Get rid of some more panic() like error handling in libtraceevent.
(Namhyung Kim)
- Get rid of panic() like calls in libtraceevent (Namyung Kim)
- Start carving out symbol parsing routines (perf, just moving
routines to topic files in tools/lib/symbol/, tools that want to
use it need to integrate it directly, ie no
tools/lib/symbol/Makefile is provided (Arnaldo Carvalho de Melo)
- Assorted refactoring patches, moving code around and adding utility
evlist methods that will be used in the IPT patchset (Adrian
Hunter)
- Assorted mmap_pages handling fixes (Adrian Hunter)
- Several man pages typo fixes (Dongsheng Yang)
- Get rid of several die() calls in libtraceevent (Namhyung Kim)
- Use basename() in a more robust way, to avoid problems related to
different system library implementations for that function
(Stephane Eranian)
- Remove open coded management of short_name_allocated member (Adrian
Hunter)
- Several cleanups in the "dso" methods, constifying some parameters
and renaming some fields to clarify its purpose (Arnaldo Carvalho
de Melo)
- Add per-feature check flags, fixing libunwind related build
problems on some architectures (Jean Pihet)
- Do not disable source line lookup just because of one failure.
(Adrian Hunter)
- Several 'perf kvm' man page corrections (Dongsheng Yang)
- Correct the message in feature-libnuma checking, swowing the right
devel package names for various distros (Dongsheng Yang)
- Polish 'readn()' function and introduce its counterpart,
'writen()' (Jiri Olsa)
- Start moving timechart state from global variables to a 'perf_tool'
derived 'timechart' struct (Arnaldo Carvalho de Melo)
... and lots of fixes and improvements I forgot to list"
* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (282 commits)
perf tools: Remove unnecessary callchain cursor state restore on unmatch
perf callchain: Spare double comparison of callchain first entry
perf tools: Do proper comm override error handling
perf symbols: Export elf_section_by_name and reuse
perf probe: Release all dynamically allocated parameters
perf probe: Release allocated probe_trace_event if failed
perf tools: Add 'build-test' make target
tools lib traceevent: Unregister handler when xen plugin is unloaded
tools lib traceevent: Unregister handler when scsi plugin is unloaded
tools lib traceevent: Unregister handler when jbd2 plugin is is unloaded
tools lib traceevent: Unregister handler when cfg80211 plugin is unloaded
tools lib traceevent: Unregister handler when mac80211 plugin is unloaded
tools lib traceevent: Unregister handler when sched_switch plugin is unloaded
tools lib traceevent: Unregister handler when kvm plugin is unloaded
tools lib traceevent: Unregister handler when kmem plugin is unloaded
tools lib traceevent: Unregister handler when hrtimer plugin is unloaded
tools lib traceevent: Unregister handler when function plugin is unloaded
tools lib traceevent: Add pevent_unregister_print_function()
tools lib traceevent: Add pevent_unregister_event_handler()
tools lib traceevent: fix pointer-integer size mismatch
...
Pull RCU updates from Ingo Molnar:
- add RCU torture scripts/tooling
- static analysis improvements
- update RCU documentation
- miscellaneous fixes
* 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (52 commits)
rcu: Remove "extern" from function declarations in kernel/rcu/rcu.h
rcu: Remove "extern" from function declarations in include/linux/*rcu*.h
rcu/torture: Dynamically allocate SRCU output buffer to avoid overflow
rcu: Don't activate RCU core on NO_HZ_FULL CPUs
rcu: Warn on allegedly impossible rcu_read_unlock_special() from irq
rcu: Add an RCU_INITIALIZER for global RCU-protected pointers
rcu: Make rcu_assign_pointer's assignment volatile and type-safe
bonding: Use RCU_INIT_POINTER() for better overhead and for sparse
rcu: Add comment on evaluate-once properties of rcu_assign_pointer().
rcu: Provide better diagnostics for blocking in RCU callback functions
rcu: Improve SRCU's grace-period comments
rcu: Fix CONFIG_RCU_FANOUT_EXACT for odd fanout/leaf values
rcu: Fix coccinelle warnings
rcutorture: Stop tracking FSF's postal address
rcutorture: Move checkarg to functions.sh
rcutorture: Flag errors and warnings with color coding
rcutorture: Record results from repeated runs of the same test scenario
rcutorture: Test summary at end of run with less chattiness
rcutorture: Update comment in kvm.sh listing typical RCU trace events
rcutorture: Add tracing-enabled version of TREE08
...
Pull core locking changes from Ingo Molnar:
- futex performance increases: larger hashes, smarter wakeups
- mutex debugging improvements
- lots of SMP ordering documentation updates
- introduce the smp_load_acquire(), smp_store_release() primitives.
(There are WIP patches that make use of them - not yet merged)
- lockdep micro-optimizations
- lockdep improvement: better cover IRQ contexts
- liblockdep at last. We'll continue to monitor how useful this is
* 'core-locking-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (34 commits)
futexes: Fix futex_hashsize initialization
arch: Re-sort some Kbuild files to hopefully help avoid some conflicts
futexes: Avoid taking the hb->lock if there's nothing to wake up
futexes: Document multiprocessor ordering guarantees
futexes: Increase hash table size for better performance
futexes: Clean up various details
arch: Introduce smp_load_acquire(), smp_store_release()
arch: Clean up asm/barrier.h implementations using asm-generic/barrier.h
arch: Move smp_mb__{before,after}_atomic_{inc,dec}.h into asm/atomic.h
locking/doc: Rename LOCK/UNLOCK to ACQUIRE/RELEASE
mutexes: Give more informative mutex warning in the !lock->owner case
powerpc: Full barrier for smp_mb__after_unlock_lock()
rcu: Apply smp_mb__after_unlock_lock() to preserve grace periods
Documentation/memory-barriers.txt: Downgrade UNLOCK+BLOCK
locking: Add an smp_mb__after_unlock_lock() for UNLOCK+BLOCK barrier
Documentation/memory-barriers.txt: Document ACCESS_ONCE()
Documentation/memory-barriers.txt: Prohibit speculative writes
Documentation/memory-barriers.txt: Add long atomic examples to memory-barriers.txt
Documentation/memory-barriers.txt: Add needed ACCESS_ONCE() calls to memory-barriers.txt
Revert "smp/cpumask: Make CONFIG_CPUMASK_OFFSTACK=y usable without debug dependency"
...
Pull core debug changes from Ingo Molnar:
"Currently there are two methods to set the panic_timeout: via
'panic=X' boot commandline option, or via /proc/sys/kernel/panic.
This tree adds a third panic_timeout configuration method:
configuration via Kconfig, via CONFIG_PANIC_TIMEOUT=X - useful to
distros that generally want their kernel defaults to come with the
.config.
CONFIG_PANIC_TIMEOUT defaults to 0, which was the previous default
value of panic_timeout.
Doing that unearthed a few arch trickeries regarding arch-special
panic_timeout values and related complications - hopefully all
resolved to the satisfaction of everyone"
* 'core-debug-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
powerpc: Clean up panic_timeout usage
MIPS: Remove panic_timeout settings
panic: Make panic_timeout configurable
Pull namespace fixes from Eric Biederman:
"This is a set of 3 regression fixes.
This fixes /proc/mounts when using "ip netns add <netns>" to display
the actual mount point.
This fixes a regression in clone that broke lxc-attach.
This fixes a regression in the permission checks for mounting /proc
that made proc unmountable if binfmt_misc was in use. Oops.
My apologies for sending this pull request so late. Al Viro gave
interesting review comments about the d_path fix that I wanted to
address in detail before I sent this pull request. Unfortunately a
bad round of colds kept from addressing that in detail until today.
The executive summary of the review was:
Al: Is patching d_path really sufficient?
The prepend_path, d_path, d_absolute_path, and __d_path family of
functions is a really mess.
Me: Yes, patching d_path is really sufficient. Yes, the code is mess.
No it is not appropriate to rewrite all of d_path for a regression
that has existed for entirely too long already, when a two line
change will do"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace:
vfs: Fix a regression in mounting proc
fork: Allow CLONE_PARENT after setns(CLONE_NEWPID)
vfs: In d_path don't call d_dname on a mount point
With the introduction of sched_attr::sched_nice we need to check
if we've got permission to actually change the nice value.
Daniel found that can_nice() would always fail; and upon
inspection it turns out that can_nice() only tests to see if we
can lower the nice value, but it doesn't validate if we're
lowering or not.
Therefore amend the test to only call can_nice() when we lower
the nice value.
Reported-and-Tested-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: raistlin@linux.it
Cc: juri.lelli@gmail.com
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Fixes: d50dde5a10 ("sched: Add new scheduler syscalls to support an extended scheduling parameters ABI")
Link: http://lkml.kernel.org/r/20140116165425.GA9481@laptop.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
"futexes: Increase hash table size for better performance"
introduces a new alloc_large_system_hash() call.
alloc_large_system_hash() however may allocate less memory than
requested, e.g. limited by MAX_ORDER.
Hence pass a pointer to alloc_large_system_hash() which will
contain the hash shift when the function returns. Afterwards
correctly set futex_hashsize.
Fixes a crash on s390 where the requested allocation size was
4MB but only 1MB was allocated.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Darren Hart <dvhart@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Waiman Long <Waiman.Long@hp.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Davidlohr Bueso <davidlohr@hp.com>
Link: http://lkml.kernel.org/r/20140116135450.GA4345@osiris
Signed-off-by: Ingo Molnar <mingo@kernel.org>
I noticed the new sched_{set,get}attr() calls didn't properly deal
with the SCHED_RESET_ON_FORK hack.
Instead of propagating the flags in high bits nonsense use the brand
spanking new attr::sched_flags field.
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Juri Lelli <juri.lelli@gmail.com>
Cc: Dario Faggioli <raistlin@linux.it>
Link: http://lkml.kernel.org/r/20140115162242.GJ31570@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Fengguang Wu reported the following build warning:
> kernel/sched/core.c:3067 __sched_setscheduler() warn: unsigned 'attr->sched_priority' is never less than zero.
Since it doesn't make sense for attr::sched_priority to be negative,
remove the check, since we already test for an upper limit any actual
negative values passed in through the old param::sched_priority field
will still be detected.
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Juri Lelli <juri.lelli@gmail.com>
Cc: Dario Faggioli <raistlin@linux.it>
Fixes: d50dde5a10 ("sched: Add new scheduler syscalls to support an extended scheduling parameters ABI")
Link: http://lkml.kernel.org/n/tip-fid9nalzii2r5voxtf4eh5kz@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Wu reported LTP failures:
> ltp.sched_setparam02.1.TFAIL
> ltp.sched_setparam02.2.TFAIL
> ltp.sched_setparam02.3.TFAIL
> ltp.sched_setparam03.1.TFAIL
There were 2 things wrong; firstly __setscheduler() failed on
sched_setparam()'s policy = -1, fix that by reading from p->policy in
that case.
Secondly, getparam() (and getattr()) would still report !0
sched_priority for !FIFO/RR tasks after having been such. So
unconditionally set p->rt_priority.
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Juri Lelli <juri.lelli@gmail.com>
Cc: Dario Faggioli <raistlin@linux.it>
Fixes: d50dde5a10 ("sched: Add new scheduler syscalls to support an extended scheduling parameters ABI")
Link: http://lkml.kernel.org/r/20140115153320.GH31570@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Previously sched_setscheduler() and sched_setparam() would not affect
the nice value of a task, restore this behaviour.
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: raistlin@linux.it
Cc: juri.lelli@gmail.com
Cc: Michael wang <wangyun@linux.vnet.ibm.com>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Fixes: d50dde5a10 ("sched: Add new scheduler syscalls to support an extended scheduling parameters ABI")
Link: http://lkml.kernel.org/r/20140115113015.GB31570@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Fengguang Wu's kbuild test robot reported the following new htmldocs warnings:
>>> Warning(kernel/sched/core.c:3380): No description found for parameter 'uattr'
>>> Warning(kernel/sched/core.c:3380): Excess function parameter 'attr' description in 'sys_sched_setattr'
>>> Warning(kernel/sched/core.c:3520): No description found for parameter 'uattr'
>>> Warning(kernel/sched/core.c:3520): Excess function parameter 'attr' description in 'sys_sched_getattr'
The second argument to sys_sched_{setattr,getattr}() is named uattr (not attr).
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dario Faggioli <raistlin@linux.it>
Fixes: d50dde5a10 ("sched: Add new scheduler syscalls to support an extended scheduling parameters ABI")
Link: http://lkml.kernel.org/r/52D5552D.5000102@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Dan Carpenter reported new 'Smatch' warnings:
> tree: git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git sched/core
> head: 130816ce4d
> commit: 1baca4ce16 [17/50] sched/deadline: Add SCHED_DEADLINE SMP-related data structures & logic
>
> kernel/sched/deadline.c:937 pick_next_task_dl() warn: variable dereferenced before check 'p' (see line 934)
BUG_ON() already fires if pick_next_dl_entity() doesn't return a valid
dl_se. No need to check if p is valid afterward.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Fixes: 1baca4ce16 ("sched/deadline: Add SCHED_DEADLINE SMP-related data structures & logic")
Link: http://lkml.kernel.org/r/52D54E25.6060100@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
fix these new sparse warnings:
>> kernel/sched/core.c:305:14: sparse: symbol 'sysctl_sched_dl_period' was not declared. Should it be static?
>> kernel/sched/core.c:306:5: sparse: symbol 'sysctl_sched_dl_runtime' was not declared. Should it be static?
Better still, they're completely unused so remove them.
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Juri Lelli <juri.lelli@gmail.com>
Link: http://lkml.kernel.org/n/tip-ke0shkG7vMnzmcdqhhiymyem@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
new sparse warnings:
>> kernel/sched/cpudeadline.c:38:6: sparse: symbol 'cpudl_exchange' was not declared. Should it be static?
>> kernel/sched/cpudeadline.c:46:6: sparse: symbol 'cpudl_heapify' was not declared. Should it be static?
>> kernel/sched/cpudeadline.c:71:6: sparse: symbol 'cpudl_change_key' was not declared. Should it be static?
>> kernel/sched/cpudeadline.c:195:15: sparse: memset with byte count of 163928
Signed-off-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Juri Lelli <juri.lelli@gmail.com>
Fixes: 6bfd6d72f5 ("sched/deadline: speed up SCHED_DEADLINE pushes with a push-heap")
Link: http://lkml.kernel.org/r/52d47f8c.EYJsA5+mELPBk4t6\%fengguang.wu@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull scheduler and timer fixes from Ingo Molnar:
"Contains a fix for a scheduler bug that manifested itself as a 3D
performance regression and a crash fix for the ARM Cadence TTC clock
driver"
* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched: Calculate effective load even if local weight is 0
* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
clocksource: cadence_ttc: Fix mutex taken inside interrupt context
With various drivers wanting to inject idle time; we get people
calling idle routines outside of the idle loop proper.
Therefore we need to be extra careful about not missing
TIF_NEED_RESCHED -> PREEMPT_NEED_RESCHED propagations.
While looking at this, I also realized there's a small window in the
existing idle loop where we can miss TIF_NEED_RESCHED; when it hits
right after the tif_need_resched() test at the end of the loop but
right before the need_resched() test at the start of the loop.
So move preempt_fold_need_resched() out of the loop where we're
guaranteed to have TIF_NEED_RESCHED set.
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/n/tip-x9jgh45oeayzajz2mjt0y7d6@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Currently local_bh_disable() is out-of-line for no apparent reason.
So inline it to save a few cycles on call/return nonsense, the
function body is a single add on x86 (a few loads and store extra on
load/store archs).
Also expose two new local_bh functions:
__local_bh_{dis,en}able_ip(unsigned long ip, unsigned int cnt);
Which implement the actual local_bh_{dis,en}able() behaviour.
The next patch uses the exposed @cnt argument to optimize bh lock
functions.
With build fixes from Jacob Pan.
Cc: rjw@rjwysocki.net
Cc: rui.zhang@intel.com
Cc: jacob.jun.pan@linux.intel.com
Cc: Mike Galbraith <bitbucket@online.de>
Cc: hpa@zytor.com
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: lenb@kernel.org
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20131119151338.GF3694@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Currently all _bh_ lock functions do two preempt_count operations:
local_bh_disable();
preempt_disable();
and for the unlock:
preempt_enable_no_resched();
local_bh_enable();
Since its a waste of perfectly good cycles to modify the same variable
twice when you can do it in one go; use the new
__local_bh_{dis,en}able_ip() functions that allow us to provide a
preempt_count value to add/sub.
So define SOFTIRQ_LOCK_OFFSET as the offset a _bh_ lock needs to
add/sub to be done in one go.
As a bonus it gets rid of the preempt_enable_no_resched() usage.
This reduces a 1000 loops of:
spin_lock_bh(&bh_lock);
spin_unlock_bh(&bh_lock);
from 53596 cycles to 51995 cycles. I didn't do enough measurements to
say for absolute sure that the result is significant but the the few
runs I did for each suggest it is so.
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: jacob.jun.pan@linux.intel.com
Cc: Mike Galbraith <bitbucket@online.de>
Cc: hpa@zytor.com
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: lenb@kernel.org
Cc: rjw@rjwysocki.net
Cc: rui.zhang@intel.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/20131119151338.GF3694@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The test on_null_domain is done twice in the trigger_load_balance function.
Move the test at the begin of the function, so there is only one check.
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1389008085-9069-9-git-send-email-daniel.lezcano@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The cpu information is stored in the struct rq. Pass the struct rq to
nohz_idle_balance, so all the functions called in run_rebalance_domains have
the same parameters and the 'this_cpu' variable becomes pointless.
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
[ Added !SMP build fix. ]
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1389008085-9069-8-git-send-email-daniel.lezcano@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The cpu information is stored in the struct rq and the caller of the
rebalance_domains function pass the cpu to retrieve the struct rq but
it already has the struct rq info. Replace the cpu parameter with the
struct rq.
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1389008085-9069-7-git-send-email-daniel.lezcano@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The cpu parameter is no longer needed in nohz_balancer_kick, let's remove
the parameter.
Reviewed-by: Preeti U Murthy <preeti@linux.vnet.ibm.com>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1389008085-9069-6-git-send-email-daniel.lezcano@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The 'call_cpu' is never used in the function. Remove it.
Reviewed-by: Preeti U Murthy <preeti@linux.vnet.ibm.com>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1389008085-9069-5-git-send-email-daniel.lezcano@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The on_null_domain() function is getting the cpu to retrieve the struct rq
associated with it.
Pass 'struct rq' directly to the function as the caller already has the info.
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1389008085-9069-4-git-send-email-daniel.lezcano@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The cpu information is already stored in the struct rq, so no need to pass it
as parameter to the nohz_kick_needed function.
The caller of this function just called idle_cpu() before to fill the
rq->idle_balance field.
Use rq->cpu and rq->idle_balance.
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1389008085-9069-3-git-send-email-daniel.lezcano@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The current hotplug admission control is broken because:
CPU_DYING -> migration_call() -> migrate_tasks() -> __migrate_task()
cannot fail and hard assumes it _will_ move all tasks off of the dying
cpu, failing this will break hotplug.
The much simpler solution is a DOWN_PREPARE handler that fails when
removing one CPU gets us below the total allocated bandwidth.
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20131220171343.GL2480@laptop.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Remove the deadline specific sysctls for now. The problem with them is
that the interaction with the exisiting rt knobs is nearly impossible
to get right.
The current (as per before this patch) situation is that the rt and dl
bandwidth is completely separate and we enforce rt+dl < 100%. This is
undesirable because this means that the rt default of 95% leaves us
hardly any room, even though dl tasks are saver than rt tasks.
Another proposed solution was (a discarted patch) to have the dl
bandwidth be a fraction of the rt bandwidth. This is highly
confusing imo.
Furthermore neither proposal is consistent with the situation we
actually want; which is rt tasks ran from a dl server. In which case
the rt bandwidth is a direct subset of dl.
So whichever way we go, the introduction of dl controls at this point
is painful. Therefore remove them and instead share the rt budget.
This means that for now the rt knobs are used for dl admission control
and the dl runtime is accounted against the rt runtime. I realise that
this isn't entirely desirable either; but whatever we do we appear to
need to change the interface later, so better have a small interface
for now.
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/n/tip-zpyqbqds1r0vyxtxza1e7rdc@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
For now deadline tasks are not allowed to set smp affinity; however
the current tests are wrong, cure this.
The test in __sched_setscheduler() also uses an on-stack cpumask_t
which is a no-no.
Change both tests to use cpumask_subset() such that we test the root
domain span to be a subset of the cpus_allowed mask. This way we're
sure the tasks can always run on all CPUs they can be balanced over,
and have no effective affinity constraints.
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/n/tip-fyqtb1lapxca3lhsxv9cumdc@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Data from tests confirmed that the original active load balancing
logic didn't scale neither in the number of CPU nor in the number of
tasks (as sched_rt does).
Here we provide a global data structure to keep track of deadlines
of the running tasks in the system. The structure is composed by
a bitmask showing the free CPUs and a max-heap, needed when the system
is heavily loaded.
The implementation and concurrent access scheme are kept simple by
design. However, our measurements show that we can compete with sched_rt
on large multi-CPUs machines [1].
Only the push path is addressed, the extension to use this structure
also for pull decisions is straightforward. However, we are currently
evaluating different (in order to decrease/avoid contention) data
structures to solve possibly both problems. We are also going to re-run
tests considering recent changes inside cpupri [2].
[1] http://retis.sssup.it/~jlelli/papers/Ospert11Lelli.pdf
[2] http://www.spinics.net/lists/linux-rt-users/msg06778.html
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-14-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
In order of deadline scheduling to be effective and useful, it is
important that some method of having the allocation of the available
CPU bandwidth to tasks and task groups under control.
This is usually called "admission control" and if it is not performed
at all, no guarantee can be given on the actual scheduling of the
-deadline tasks.
Since when RT-throttling has been introduced each task group have a
bandwidth associated to itself, calculated as a certain amount of
runtime over a period. Moreover, to make it possible to manipulate
such bandwidth, readable/writable controls have been added to both
procfs (for system wide settings) and cgroupfs (for per-group
settings).
Therefore, the same interface is being used for controlling the
bandwidth distrubution to -deadline tasks and task groups, i.e.,
new controls but with similar names, equivalent meaning and with
the same usage paradigm are added.
However, more discussion is needed in order to figure out how
we want to manage SCHED_DEADLINE bandwidth at the task group level.
Therefore, this patch adds a less sophisticated, but actually
very sensible, mechanism to ensure that a certain utilization
cap is not overcome per each root_domain (the single rq for !SMP
configurations).
Another main difference between deadline bandwidth management and
RT-throttling is that -deadline tasks have bandwidth on their own
(while -rt ones doesn't!), and thus we don't need an higher level
throttling mechanism to enforce the desired bandwidth.
This patch, therefore:
- adds system wide deadline bandwidth management by means of:
* /proc/sys/kernel/sched_dl_runtime_us,
* /proc/sys/kernel/sched_dl_period_us,
that determine (i.e., runtime / period) the total bandwidth
available on each CPU of each root_domain for -deadline tasks;
- couples the RT and deadline bandwidth management, i.e., enforces
that the sum of how much bandwidth is being devoted to -rt
-deadline tasks to stay below 100%.
This means that, for a root_domain comprising M CPUs, -deadline tasks
can be created until the sum of their bandwidths stay below:
M * (sched_dl_runtime_us / sched_dl_period_us)
It is also possible to disable this bandwidth management logic, and
be thus free of oversubscribing the system up to any arbitrary level.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-12-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Some method to deal with rt-mutexes and make sched_dl interact with
the current PI-coded is needed, raising all but trivial issues, that
needs (according to us) to be solved with some restructuring of
the pi-code (i.e., going toward a proxy execution-ish implementation).
This is under development, in the meanwhile, as a temporary solution,
what this commits does is:
- ensure a pi-lock owner with waiters is never throttled down. Instead,
when it runs out of runtime, it immediately gets replenished and it's
deadline is postponed;
- the scheduling parameters (relative deadline and default runtime)
used for that replenishments --during the whole period it holds the
pi-lock-- are the ones of the waiting task with earliest deadline.
Acting this way, we provide some kind of boosting to the lock-owner,
still by using the existing (actually, slightly modified by the previous
commit) pi-architecture.
We would stress the fact that this is only a surely needed, all but
clean solution to the problem. In the end it's only a way to re-start
discussion within the community. So, as always, comments, ideas, rants,
etc.. are welcome! :-)
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
[ Added !RT_MUTEXES build fix. ]
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-11-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Turn the pi-chains from plist to rb-tree, in the rt_mutex code,
and provide a proper comparison function for -deadline and
-priority tasks.
This is done mainly because:
- classical prio field of the plist is just an int, which might
not be enough for representing a deadline;
- manipulating such a list would become O(nr_deadline_tasks),
which might be to much, as the number of -deadline task increases.
Therefore, an rb-tree is used, and tasks are queued in it according
to the following logic:
- among two -priority (i.e., SCHED_BATCH/OTHER/RR/FIFO) tasks, the
one with the higher (lower, actually!) prio wins;
- among a -priority and a -deadline task, the latter always wins;
- among two -deadline tasks, the one with the earliest deadline
wins.
Queueing and dequeueing functions are changed accordingly, for both
the list of a task's pi-waiters and the list of tasks blocked on
a pi-lock.
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-again-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-10-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
It is very likely that systems that wants/needs to use the new
SCHED_DEADLINE policy also want to have the scheduling latency of
the -deadline tasks under control.
For this reason a new version of the scheduling wakeup latency,
called "wakeup_dl", is introduced.
As a consequence of applying this patch there will be three wakeup
latency tracer:
* "wakeup", that deals with all tasks in the system;
* "wakeup_rt", that deals with -rt and -deadline tasks only;
* "wakeup_dl", that deals with -deadline tasks only.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-9-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Make it possible to specify a period (different or equal than
deadline) for -deadline tasks. Relative deadlines (D_i) are used on
task arrivals to generate new scheduling (absolute) deadlines as "d =
t + D_i", and periods (P_i) to postpone the scheduling deadlines as "d
= d + P_i" when the budget is zero.
This is in general useful to model (and schedule) tasks that have slow
activation rates (long periods), but have to be scheduled soon once
activated (short deadlines).
Signed-off-by: Harald Gustafsson <harald.gustafsson@ericsson.com>
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-7-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Make the core scheduler and load balancer aware of the load
produced by -deadline tasks, by updating the moving average
like for sched_rt.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-6-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Introduces data structures relevant for implementing dynamic
migration of -deadline tasks and the logic for checking if
runqueues are overloaded with -deadline tasks and for choosing
where a task should migrate, when it is the case.
Adds also dynamic migrations to SCHED_DEADLINE, so that tasks can
be moved among CPUs when necessary. It is also possible to bind a
task to a (set of) CPU(s), thus restricting its capability of
migrating, or forbidding migrations at all.
The very same approach used in sched_rt is utilised:
- -deadline tasks are kept into CPU-specific runqueues,
- -deadline tasks are migrated among runqueues to achieve the
following:
* on an M-CPU system the M earliest deadline ready tasks
are always running;
* affinity/cpusets settings of all the -deadline tasks is
always respected.
Therefore, this very special form of "load balancing" is done with
an active method, i.e., the scheduler pushes or pulls tasks between
runqueues when they are woken up and/or (de)scheduled.
IOW, every time a preemption occurs, the descheduled task might be sent
to some other CPU (depending on its deadline) to continue executing
(push). On the other hand, every time a CPU becomes idle, it might pull
the second earliest deadline ready task from some other CPU.
To enforce this, a pull operation is always attempted before taking any
scheduling decision (pre_schedule()), as well as a push one after each
scheduling decision (post_schedule()). In addition, when a task arrives
or wakes up, the best CPU where to resume it is selected taking into
account its affinity mask, the system topology, but also its deadline.
E.g., from the scheduling point of view, the best CPU where to wake
up (and also where to push) a task is the one which is running the task
with the latest deadline among the M executing ones.
In order to facilitate these decisions, per-runqueue "caching" of the
deadlines of the currently running and of the first ready task is used.
Queued but not running tasks are also parked in another rb-tree to
speed-up pushes.
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-5-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Introduces the data structures, constants and symbols needed for
SCHED_DEADLINE implementation.
Core data structure of SCHED_DEADLINE are defined, along with their
initializers. Hooks for checking if a task belong to the new policy
are also added where they are needed.
Adds a scheduling class, in sched/dl.c and a new policy called
SCHED_DEADLINE. It is an implementation of the Earliest Deadline
First (EDF) scheduling algorithm, augmented with a mechanism (called
Constant Bandwidth Server, CBS) that makes it possible to isolate
the behaviour of tasks between each other.
The typical -deadline task will be made up of a computation phase
(instance) which is activated on a periodic or sporadic fashion. The
expected (maximum) duration of such computation is called the task's
runtime; the time interval by which each instance need to be completed
is called the task's relative deadline. The task's absolute deadline
is dynamically calculated as the time instant a task (better, an
instance) activates plus the relative deadline.
The EDF algorithms selects the task with the smallest absolute
deadline as the one to be executed first, while the CBS ensures each
task to run for at most its runtime every (relative) deadline
length time interval, avoiding any interference between different
tasks (bandwidth isolation).
Thanks to this feature, also tasks that do not strictly comply with
the computational model sketched above can effectively use the new
policy.
To summarize, this patch:
- introduces the data structures, constants and symbols needed;
- implements the core logic of the scheduling algorithm in the new
scheduling class file;
- provides all the glue code between the new scheduling class and
the core scheduler and refines the interactions between sched/dl
and the other existing scheduling classes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Fabio Checconi <fchecconi@gmail.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-4-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Add the syscalls needed for supporting scheduling algorithms
with extended scheduling parameters (e.g., SCHED_DEADLINE).
In general, it makes possible to specify a periodic/sporadic task,
that executes for a given amount of runtime at each instance, and is
scheduled according to the urgency of their own timing constraints,
i.e.:
- a (maximum/typical) instance execution time,
- a minimum interval between consecutive instances,
- a time constraint by which each instance must be completed.
Thus, both the data structure that holds the scheduling parameters of
the tasks and the system calls dealing with it must be extended.
Unfortunately, modifying the existing struct sched_param would break
the ABI and result in potentially serious compatibility issues with
legacy binaries.
For these reasons, this patch:
- defines the new struct sched_attr, containing all the fields
that are necessary for specifying a task in the computational
model described above;
- defines and implements the new scheduling related syscalls that
manipulate it, i.e., sched_setattr() and sched_getattr().
Syscalls are introduced for x86 (32 and 64 bits) and ARM only, as a
proof of concept and for developing and testing purposes. Making them
available on other architectures is straightforward.
Since no "user" for these new parameters is introduced in this patch,
the implementation of the new system calls is just identical to their
already existing counterpart. Future patches that implement scheduling
policies able to exploit the new data structure must also take care of
modifying the sched_*attr() calls accordingly with their own purposes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
[ Rewrote to use sched_attr. ]
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
[ Removed sched_setscheduler2() for now. ]
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-3-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
In futex_wake() there is clearly no point in taking the hb->lock
if we know beforehand that there are no tasks to be woken. While
the hash bucket's plist head is a cheap way of knowing this, we
cannot rely 100% on it as there is a racy window between the
futex_wait call and when the task is actually added to the
plist. To this end, we couple it with the spinlock check as
tasks trying to enter the critical region are most likely
potential waiters that will be added to the plist, thus
preventing tasks sleeping forever if wakers don't acknowledge
all possible waiters.
Furthermore, the futex ordering guarantees are preserved,
ensuring that waiters either observe the changed user space
value before blocking or is woken by a concurrent waker. For
wakers, this is done by relying on the barriers in
get_futex_key_refs() -- for archs that do not have implicit mb
in atomic_inc(), we explicitly add them through a new
futex_get_mm function. For waiters we rely on the fact that
spin_lock calls already update the head counter, so spinners
are visible even if the lock hasn't been acquired yet.
For more details please refer to the updated comments in the
code and related discussion:
https://lkml.org/lkml/2013/11/26/556
Special thanks to tglx for careful review and feedback.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Davidlohr Bueso <davidlohr@hp.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Jeff Mahoney <jeffm@suse.com>
Cc: Scott Norton <scott.norton@hp.com>
Cc: Tom Vaden <tom.vaden@hp.com>
Cc: Aswin Chandramouleeswaran <aswin@hp.com>
Cc: Waiman Long <Waiman.Long@hp.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/1389569486-25487-5-git-send-email-davidlohr@hp.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
That's essential, if you want to hack on futexes.
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Reviewed-by: Peter Zijlstra <peterz@infradead.org>
Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Davidlohr Bueso <davidlohr@hp.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Jeff Mahoney <jeffm@suse.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Randy Dunlap <rdunlap@infradead.org>
Cc: Scott Norton <scott.norton@hp.com>
Cc: Tom Vaden <tom.vaden@hp.com>
Cc: Aswin Chandramouleeswaran <aswin@hp.com>
Cc: Waiman Long <Waiman.Long@hp.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/1389569486-25487-4-git-send-email-davidlohr@hp.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Currently, the futex global hash table suffers from its fixed,
smallish (for today's standards) size of 256 entries, as well as
its lack of NUMA awareness. Large systems, using many futexes,
can be prone to high amounts of collisions; where these futexes
hash to the same bucket and lead to extra contention on the same
hb->lock. Furthermore, cacheline bouncing is a reality when we
have multiple hb->locks residing on the same cacheline and
different futexes hash to adjacent buckets.
This patch keeps the current static size of 16 entries for small
systems, or otherwise, 256 * ncpus (or larger as we need to
round the number to a power of 2). Note that this number of CPUs
accounts for all CPUs that can ever be available in the system,
taking into consideration things like hotpluging. While we do
impose extra overhead at bootup by making the hash table larger,
this is a one time thing, and does not shadow the benefits of
this patch.
Furthermore, as suggested by tglx, by cache aligning the hash
buckets we can avoid access across cacheline boundaries and also
avoid massive cache line bouncing if multiple cpus are hammering
away at different hash buckets which happen to reside in the
same cache line.
Also, similar to other core kernel components (pid, dcache,
tcp), by using alloc_large_system_hash() we benefit from its
NUMA awareness and thus the table is distributed among the nodes
instead of in a single one.
For a custom microbenchmark that pounds on the uaddr hashing --
making the wait path fail at futex_wait_setup() returning
-EWOULDBLOCK for large amounts of futexes, we can see the
following benefits on a 80-core, 8-socket 1Tb server:
+---------+--------------------+------------------------+-----------------------+-------------------------------+
| threads | baseline (ops/sec) | aligned-only (ops/sec) | large table (ops/sec) | large table+aligned (ops/sec) |
+---------+--------------------+------------------------+-----------------------+-------------------------------+
| 512 | 32426 | 50531 (+55.8%) | 255274 (+687.2%) | 292553 (+802.2%) |
| 256 | 65360 | 99588 (+52.3%) | 443563 (+578.6%) | 508088 (+677.3%) |
| 128 | 125635 | 200075 (+59.2%) | 742613 (+491.1%) | 835452 (+564.9%) |
| 80 | 193559 | 323425 (+67.1%) | 1028147 (+431.1%) | 1130304 (+483.9%) |
| 64 | 247667 | 443740 (+79.1%) | 997300 (+302.6%) | 1145494 (+362.5%) |
| 32 | 628412 | 721401 (+14.7%) | 965996 (+53.7%) | 1122115 (+78.5%) |
+---------+--------------------+------------------------+-----------------------+-------------------------------+
Reviewed-by: Darren Hart <dvhart@linux.intel.com>
Reviewed-by: Peter Zijlstra <peterz@infradead.org>
Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Waiman Long <Waiman.Long@hp.com>
Reviewed-and-tested-by: Jason Low <jason.low2@hp.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Davidlohr Bueso <davidlohr@hp.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Jeff Mahoney <jeffm@suse.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Scott Norton <scott.norton@hp.com>
Cc: Tom Vaden <tom.vaden@hp.com>
Cc: Aswin Chandramouleeswaran <aswin@hp.com>
Link: http://lkml.kernel.org/r/1389569486-25487-3-git-send-email-davidlohr@hp.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
