linux/arch/x86
Robert Richter bc1738f6ee perf, x86: Fix event scheduler for constraints with overlapping counters
The current x86 event scheduler fails to resolve scheduling problems
of certain combinations of events and constraints. This happens if the
counter mask of such an event is not a subset of any other counter
mask of a constraint with an equal or higher weight, e.g. constraints
of the AMD family 15h pmu:

                        counter mask    weight

 amd_f15_PMC30          0x09            2  <--- overlapping counters
 amd_f15_PMC20          0x07            3
 amd_f15_PMC53          0x38            3

The scheduler does not find then an existing solution. Here is an
example:

 event code     counter         failure         possible solution

 0x02E          PMC[3,0]        0               3
 0x043          PMC[2:0]        1               0
 0x045          PMC[2:0]        2               1
 0x046          PMC[2:0]        FAIL            2

The event scheduler may not select the correct counter in the first
cycle because it needs to know which subsequent events will be
scheduled. It may fail to schedule the events then.

To solve this, we now save the scheduler state of events with
overlapping counter counstraints.  If we fail to schedule the events
we rollback to those states and try to use another free counter.

Constraints with overlapping counters are marked with a new introduced
overlap flag. We set the overlap flag for such constraints to give the
scheduler a hint which events to select for counter rescheduling. The
EVENT_CONSTRAINT_OVERLAP() macro can be used for this.

Care must be taken as the rescheduling algorithm is O(n!) which will
increase scheduling cycles for an over-commited system dramatically.
The number of such EVENT_CONSTRAINT_OVERLAP() macros and its counter
masks must be kept at a minimum. Thus, the current stack is limited to
2 states to limit the number of loops the algorithm takes in the worst
case.

On systems with no overlapping-counter constraints, this
implementation does not increase the loop count compared to the
previous algorithm.

V2:
* Renamed redo -> overlap.
* Reimplementation using perf scheduling helper functions.

V3:
* Added WARN_ON_ONCE() if out of save states.
* Changed function interface of perf_sched_restore_state() to use bool
  as return value.

Signed-off-by: Robert Richter <robert.richter@amd.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1321616122-1533-3-git-send-email-robert.richter@amd.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-12-06 08:33:56 +01:00
..
boot doc: fix broken references 2011-09-27 18:08:04 +02:00
configs iommu: Rename the DMAR and INTR_REMAP config options 2011-09-21 10:22:03 +02:00
crypto Merge branch 'modsplit-Oct31_2011' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux 2011-11-06 19:44:47 -08:00
ia32 Cross Memory Attach 2011-10-31 17:30:44 -07:00
include/asm x86, perf: Add a build-time sanity test to the x86 decoder 2011-11-10 12:38:51 +01:00
kernel perf, x86: Fix event scheduler for constraints with overlapping counters 2011-12-06 08:33:56 +01:00
kvm Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu 2011-10-30 15:46:19 -07:00
lguest lguest: add export.h to lguest files for THIS_MODULE/EXPORT_SYMBOL 2011-10-31 19:32:13 -04:00
lib x86: Update instruction decoder to support new AVX formats 2011-12-05 14:53:21 +01:00
math-emu
mm thp: share get_huge_page_tail() 2011-11-02 16:06:58 -07:00
net net: filter: Just In Time compiler for x86-64 2011-04-27 23:05:08 -07:00
oprofile Merge branch 'core' of git://amd64.org/linux/rric into perf/core 2011-11-15 11:05:18 +01:00
pci x86: Fix files explicitly requiring export.h for EXPORT_SYMBOL/THIS_MODULE 2011-10-31 19:30:35 -04:00
platform bma023: Add SFI translation for this device 2011-11-11 23:58:58 -02:00
power x86: Fix files explicitly requiring export.h for EXPORT_SYMBOL/THIS_MODULE 2011-10-31 19:30:35 -04:00
tools x86/tools: Add decoded instruction dump mode 2011-12-05 14:53:23 +01:00
um Merge branch 'trivial' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild 2011-11-06 18:54:53 -08:00
vdso Merge branch 'x86-cpu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2011-10-28 05:03:12 -07:00
video x86: fix up files really needing to include module.h 2011-10-31 19:30:36 -04:00
xen Merge branch 'upstream/xen-settime' of git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen 2011-11-06 20:15:05 -08:00
.gitignore
Kbuild net: filter: Just In Time compiler for x86-64 2011-04-27 23:05:08 -07:00
Kconfig Merge branch 'trivial' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild 2011-11-06 18:54:53 -08:00
Kconfig.cpu x86: Add support for cmpxchg_double 2011-06-25 12:17:32 -07:00
Kconfig.debug doc: fix broken references 2011-09-27 18:08:04 +02:00
Makefile
Makefile_32.cpu x86, cpu: Move AMD Elan Kconfig under "Processor family" 2011-04-08 13:01:25 -07:00
Makefile.um um: take arch/um/sys-x86 to arch/x86/um 2011-11-02 14:15:05 +01:00