Node distance either used direct node comparison, ACPI PXM comparison
or ACPI SLIT table lookup. This patch implements generic node
distance handling. NUMA init methods can call numa_set_distance() to
set distance between nodes and the common __node_distance()
implementation will report the set distance.
Due to the way NUMA emulation is implemented, the generic node
distance handling is used only when emulation is not used. Later
patches will update NUMA emulation to use the generic distance
mechanism.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Cyrill Gorcunov <gorcunov@gmail.com>
Cc: Shaohui Zheng <shaohui.zheng@intel.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: H. Peter Anvin <hpa@linux.intel.com>
Unlike 64bit, 32bit has been using its own cpu_to_node_map[] for
CPU -> NUMA node mapping. Replace it with early_percpu variable
x86_cpu_to_node_map and share the mapping code with 64bit.
* USE_PERCPU_NUMA_NODE_ID is now enabled for 32bit too.
* x86_cpu_to_node_map and numa_set/clear_node() are moved from
numa_64 to numa. For now, on 32bit, x86_cpu_to_node_map is initialized
with 0 instead of NUMA_NO_NODE. This is to avoid introducing unexpected
behavior change and will be updated once init path is unified.
* srat_detect_node() is now enabled for x86_32 too. It calls
numa_set_node() and initializes the mapping making explicit
cpu_to_node_map[] updates from map/unmap_cpu_to_node() unnecessary.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: eric.dumazet@gmail.com
Cc: yinghai@kernel.org
Cc: brgerst@gmail.com
Cc: gorcunov@gmail.com
Cc: penberg@kernel.org
Cc: shaohui.zheng@intel.com
Cc: rientjes@google.com
LKML-Reference: <1295789862-25482-15-git-send-email-tj@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: David Rientjes <rientjes@google.com>
x86 arch specific changes to use generic numa_node_id() based on generic
percpu variable infrastructure. Back out x86's custom version of
numa_node_id()
Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Nick Piggin <npiggin@suse.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Eric Whitney <eric.whitney@hp.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: <linux-arch@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Rework the generic version of the numa_node_id() function to use the new
generic percpu variable infrastructure.
Guard the new implementation with a new config option:
CONFIG_USE_PERCPU_NUMA_NODE_ID.
Archs which support this new implemention will default this option to 'y'
when NUMA is configured. This config option could be removed if/when all
archs switch over to the generic percpu implementation of numa_node_id().
Arch support involves:
1) converting any existing per cpu variable implementations to use
this implementation. x86_64 is an instance of such an arch.
2) archs that don't use a per cpu variable for numa_node_id() will
need to initialize the new per cpu variable "numa_node" as cpus
are brought on-line. ia64 is an example.
3) Defining USE_PERCPU_NUMA_NODE_ID in arch dependent Kconfig--e.g.,
when NUMA is configured. This is required because I have
retained the old implementation by default to allow archs to
be modified incrementally, as desired.
Subsequent patches will convert x86_64 and ia64 to use this implemenation.
Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Reviewed-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Nick Piggin <npiggin@suse.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Eric Whitney <eric.whitney@hp.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: <linux-arch@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Move definition of NUMA_NO_NODE from ia64 and x86_64 arch specific headers
to generic header 'linux/numa.h' for use in generic code. NUMA_NO_NODE
replaces bare '-1' where it's used in this series to indicate "no node id
specified". Ultimately, it can be used to replace the -1 elsewhere where
it is used similarly.
Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com>
Acked-by: David Rientjes <rientjes@google.com>
Acked-by: Mel Gorman <mel@csn.ul.ie>
Reviewed-by: Andi Kleen <andi@firstfloor.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Nishanth Aravamudan <nacc@us.ibm.com>
Cc: Adam Litke <agl@us.ibm.com>
Cc: Andy Whitcroft <apw@canonical.com>
Cc: Eric Whitney <eric.whitney@hp.com>
Cc: Christoph Lameter <cl@linux-foundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Yanmin Zhang reported that SD_PREFER_LOCAL induces an order of
magnitude increase in select_task_rq_fair() overhead while
running heavy wakeup benchmarks (tbench and vmark).
Since SD_BALANCE_WAKE is off at node level, turn SD_PREFER_LOCAL
off as well pending further investigation.
Reported-by: Zhang, Yanmin <yanmin_zhang@linux.intel.com>
Signed-off-by: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Yanmin reported that both tbench and hackbench were significantly
hurt by trying to keep tasks local on these domains, esp on small
cache machines.
So disable it in order to promote spreading outside of the cache
domains.
Reported-by: "Zhang, Yanmin" <yanmin_zhang@linux.intel.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
CC: Mike Galbraith <efault@gmx.de>
LKML-Reference: <1255083400.8802.15.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
arch/x86/include/asm/topology.h declares inline fns cpu_to_node and
cpumask_of_node for !NUMA, even though they are then declared as
macros by asm-generic/topology.h, which is #included just below.
The macros (which are the same) end up being used; these functions
are just confusing.
Noticed-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Cc: "Greg Kroah-Hartman" <gregkh@suse.de>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Tejun Heo <tj@kernel.org>
LKML-Reference: <200909241748.45629.rusty@rustcorp.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Sysbench thinks SD_BALANCE_WAKE is too agressive and kbuild doesn't
really mind too much, SD_BALANCE_NEWIDLE picks up most of the
slack.
On a dual socket, quad core, dual thread nehalem system:
sysbench (--num_threads=16):
SD_BALANCE_WAKE-: 13982 tx/s
SD_BALANCE_WAKE+: 15688 tx/s
kbuild (-j16):
SD_BALANCE_WAKE-: 47.648295846 seconds time elapsed ( +- 0.312% )
SD_BALANCE_WAKE+: 47.608607360 seconds time elapsed ( +- 0.026% )
(same within noise)
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
If we're looking to place a new task, we might as well find the
idlest position _now_, not 1 tick ago.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Make the idle balancer more agressive, to improve a
x264 encoding workload provided by Jason Garrett-Glaser:
NEXT_BUDDY NO_LB_BIAS
encoded 600 frames, 252.82 fps, 22096.60 kb/s
encoded 600 frames, 250.69 fps, 22096.60 kb/s
encoded 600 frames, 245.76 fps, 22096.60 kb/s
NO_NEXT_BUDDY LB_BIAS
encoded 600 frames, 344.44 fps, 22096.60 kb/s
encoded 600 frames, 346.66 fps, 22096.60 kb/s
encoded 600 frames, 352.59 fps, 22096.60 kb/s
NO_NEXT_BUDDY NO_LB_BIAS
encoded 600 frames, 425.75 fps, 22096.60 kb/s
encoded 600 frames, 425.45 fps, 22096.60 kb/s
encoded 600 frames, 422.49 fps, 22096.60 kb/s
Peter pointed out that this is better done via newidle_idx,
not via LB_BIAS, newidle balancing should look for where
there is load _now_, not where there was load 2 ticks ago.
Worst-case latencies are improved as well as no buddies
means less vruntime spread. (as per prior lkml discussions)
This change improves kbuild-peak parallelism as well.
Reported-by: Jason Garrett-Glaser <darkshikari@gmail.com>
Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1253011667.9128.16.camel@marge.simson.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
When merging select_task_rq_fair() and sched_balance_self() we lost
the use of wake_idx, restore that and set them to 0 to make wake
balancing more aggressive.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The problem with wake_idle() is that is doesn't respect things like
cpu_power, which means it doesn't deal well with SMT nor the recent
RT interaction.
To cure this, it needs to do what sched_balance_self() does, which
leads to the possibility of merging select_task_rq_fair() and
sched_balance_self().
Modify sched_balance_self() to:
- update_shares() when walking up the domain tree,
(it only called it for the top domain, but it should
have done this anyway), which allows us to remove
this ugly bit from try_to_wake_up().
- do wake_affine() on the smallest domain that contains
both this (the waking) and the prev (the wakee) cpu for
WAKE invocations.
Then use the top-down balance steps it had to replace wake_idle().
This leads to the dissapearance of SD_WAKE_BALANCE and
SD_WAKE_IDLE_FAR, with SD_WAKE_IDLE replaced with SD_BALANCE_WAKE.
SD_WAKE_AFFINE needs SD_BALANCE_WAKE to be effective.
Touch all topology bits to replace the old with new SD flags --
platforms might need re-tuning, enabling SD_BALANCE_WAKE
conditionally on a NUMA distance seems like a good additional
feature, magny-core and small nehalem systems would want this
enabled, systems with slow interconnects would not.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Now that SD_WAKE_IDLE doesn't make pipe-test suck anymore,
enable it by default for MC, CPU and NUMA domains.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Re-organize the flag settings so that it's visible at a glance
which sched-domains flags are set and which not.
With the new balancer code we'll need to re-tune these details
anyway, so make it cleaner to make fewer mistakes down the
road ;-)
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Andreas Herrmann <andreas.herrmann3@amd.com>
Cc: Andreas Herrmann <andreas.herrmann3@amd.com>
Cc: Gautham R Shenoy <ego@in.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Fix to prevent sched_mc_power_saving from being exported through sysfs
for multi-scoket single core system. Max cores should be always greater than
one (1). My earlier patch that introduced fix for not exporting
'sched_mc_power_saving' on laptops broke it on multi-socket single core
system. This fix addresses issue on both laptop and multi-socket single
core system.
Below are the Test results:
1. Single socket - multi-core
Before Patch: Does not export 'sched_mc_power_saving'
After Patch: Does not export 'sched_mc_power_saving'
Result: Pass
2. Multi Socket - single core
Before Patch: exports 'sched_mc_power_saving'
After Patch: Does not export 'sched_mc_power_saving'
Result: Pass
3. Multi Socket - Multi core
Before Patch: exports 'sched_mc_power_saving'
After Patch: exports 'sched_mc_power_saving'
[ Impact: make the sched_mc_power_saving control available more consistently ]
Signed-off-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Cc: Suresh B Siddha <suresh.b.siddha@intel.com>
Cc: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20090511143914.GB4853@dirshya.in.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Rename set_pci_bus_resources_arch_default to x86_pci_root_bus_res_quirks, move
the weak version from common.c to i386.c, and before calling, make sure it's a
root bus.
Reviewed-by: Matthew Wilcox <willy@linux.intel.com>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Everyone defines it, and only one person uses it
(arch/mips/sgi-ip27/ip27-nmi.c). So just open code it there.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: linux-mips@linux-mips.org
Impact: cleanup
We are removing cpumask_t in favour of struct cpumask: mainly as a
marker of what code is now CONFIG_CPUMASK_OFFSTACK-safe.
The only non-trivial change here is vector_allocation_domain():
explicitly clear the mask and set the first word, rather than using
assignment.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Impact: reduce kernel memory usage when CONFIG_CPUMASK_OFFSTACK=y
Straightforward conversion: done for 32 and 64 bit kernels.
node_to_cpumask_map is now a cpumask_var_t array.
64-bit used to be a dynamic cpumask_t array, and 32-bit used to be a
static cpumask_t array.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Impact: cleanup
We take the 64-bit code and use it on 32-bit as well. The new file
is called mm/numa.c.
In a minor cleanup, we use cpu_none_mask instead of declaring a local
cpu_mask_none.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Impact: reduce per-cpu size for CONFIG_CPUMASK_OFFSTACK=y
In most places it's cleaner to use the accessors cpu_sibling_mask()
and cpu_core_mask() wrappers which already exist.
I couldn't avoid cleaning up the access in oprofile, either.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Impact: Code movement, no functional change.
Move the 64-bit NUMA code from setup_percpu.c to numa_64.c
Signed-off-by: Brian Gerst <brgerst@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Conflicts:
arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
arch/x86/kernel/tlb_32.c
Merge it here because both the cpumask changes and the ongoing percpu
work is touching the TLB code. The percpu changes take precedence, as
they eliminate tlb_32.c altogether.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Make early_per_cpu() a lvalue as per_cpu() is and use it where
applicable.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: fix build warning
The macro cpu_to_node did not reference it's argument, and instead
simply returned a 0. This causes a "unused variable" warning if
it's the only reference in a function (show_cache_disable).
Replace it with the more correct inline function.
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: New API
Like cpu_coregroup_map, but returns a (const) pointer.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: Ingo Molnar <mingo@redhat.com>
Impact: New APIs
The old node_to_cpumask/node_to_pcibus returned a cpumask_t: these
return a pointer to a struct cpumask. Part of removing cpumasks from
the stack.
Also makes __pcibus_to_node take a const pointer.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Ingo Molnar <mingo@elte.hu>
Impact: new API
The old topology_core_siblings() and topology_thread_siblings() return
a cpumask_t; these new ones return a (const) struct cpumask *.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Mike Travis <travis@sgi.com>
Impact: do not expose a control that has no effect
Fix to prevent sched_mc_power_saving from being exported through sysfs
on single-socket systems. (Say multicore single socket (Laptop))
CPU core map of the boot cpu should be equal to possible number
of cpus for single socket system.
This fix has been developed at FOSS.in kernel workout.
Signed-off-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: improve wakeup affinity on NUMA systems, tweak SMP systems
Given the fixes+tweaks to the wakeup-buddy code, re-tweak the domain
balancing defaults on NUMA and SMP systems.
Turn on SD_WAKE_AFFINE which was off on x86 NUMA - there's no reason
why we would not want to have wakeup affinity across nodes as well.
(we already do this in the standard NUMA template.)
lat_ctx on a NUMA box is particularly happy about this change:
before:
| phoenix:~/l> ./lat_ctx -s 0 2
| "size=0k ovr=2.60
| 2 5.70
after:
| phoenix:~/l> ./lat_ctx -s 0 2
| "size=0k ovr=2.65
| 2 2.07
a 2.75x speedup.
pipe-test is similarly happy about it too:
| phoenix:~/sched-tests> ./pipe-test
| 18.26 usecs/loop.
| 14.70 usecs/loop.
| 14.38 usecs/loop.
| 10.55 usecs/loop. # +WAKE_AFFINE on domain0+domain1
| 8.63 usecs/loop.
| 8.59 usecs/loop.
| 9.03 usecs/loop.
| 8.94 usecs/loop.
| 8.96 usecs/loop.
| 8.63 usecs/loop.
Also:
- disable SD_BALANCE_NEWIDLE on NUMA and SMP domains (keep it for siblings)
- enable SD_WAKE_BALANCE on SMP domains
Sysbench+postgresql improves all around the board, quite significantly:
.28-rc3-11474e2c .28-rc3-11474e2c-tune
-------------------------------------------------
1: 571 688 +17.08%
2: 1236 1206 -2.55%
4: 2381 2642 +9.89%
8: 4958 5164 +3.99%
16: 9580 9574 -0.07%
32: 7128 8118 +12.20%
64: 7342 8266 +11.18%
128: 7342 8064 +8.95%
256: 7519 7884 +4.62%
512: 7350 7731 +4.93%
-------------------------------------------------
SUM: 55412 59341 +6.62%
So it's a win both for the runup portion, the peak area and the tail.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Change header guards named "ASM_X86__*" to "_ASM_X86_*" since:
a. the double underscore is ugly and pointless.
b. no leading underscore violates namespace constraints.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
