linux_dsm_epyc7002/arch/x86/kernel/cpu/perf_event.c

1910 lines
43 KiB
C
Raw Normal View History

/*
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
* Performance events x86 architecture code
*
* Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de>
* Copyright (C) 2008-2009 Red Hat, Inc., Ingo Molnar
* Copyright (C) 2009 Jaswinder Singh Rajput
* Copyright (C) 2009 Advanced Micro Devices, Inc., Robert Richter
* Copyright (C) 2008-2009 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com>
* Copyright (C) 2009 Intel Corporation, <markus.t.metzger@intel.com>
* Copyright (C) 2009 Google, Inc., Stephane Eranian
*
* For licencing details see kernel-base/COPYING
*/
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
#include <linux/perf_event.h>
#include <linux/capability.h>
#include <linux/notifier.h>
#include <linux/hardirq.h>
#include <linux/kprobes.h>
#include <linux/module.h>
#include <linux/kdebug.h>
#include <linux/sched.h>
#include <linux/uaccess.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 15:04:11 +07:00
#include <linux/slab.h>
perf_counter: x86: Fix call-chain support to use NMI-safe methods __copy_from_user_inatomic() isn't NMI safe in that it can trigger the page fault handler which is another trap and its return path invokes IRET which will also close the NMI context. Therefore use a GUP based approach to copy the stack frames over. We tried an alternative solution as well: we used a forward ported version of Mathieu Desnoyers's "NMI safe INT3 and Page Fault" patch that modifies the exception return path to use an open-coded IRET with explicit stack unrolling and TF checking. This didnt work as it interacted with faulting user-space instructions, causing them not to restart properly, which corrupts user-space registers. Solving that would probably involve disassembling those instructions and backtracing the RIP. But even without that, the code was deemed rather complex to the already non-trivial x86 entry assembly code, so instead we went for this GUP based method that does a software-walk of the pagetables. Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Nick Piggin <npiggin@suse.de> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Cc: Vegard Nossum <vegard.nossum@gmail.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-06-15 18:07:24 +07:00
#include <linux/highmem.h>
#include <linux/cpu.h>
#include <linux/bitops.h>
#include <asm/apic.h>
#include <asm/stacktrace.h>
#include <asm/nmi.h>
#include <asm/compat.h>
#include <asm/smp.h>
#if 0
#undef wrmsrl
#define wrmsrl(msr, val) \
do { \
trace_printk("wrmsrl(%lx, %lx)\n", (unsigned long)(msr),\
(unsigned long)(val)); \
native_write_msr((msr), (u32)((u64)(val)), \
(u32)((u64)(val) >> 32)); \
} while (0)
#endif
/*
* best effort, GUP based copy_from_user() that assumes IRQ or NMI context
*/
static unsigned long
copy_from_user_nmi(void *to, const void __user *from, unsigned long n)
{
unsigned long offset, addr = (unsigned long)from;
unsigned long size, len = 0;
struct page *page;
void *map;
int ret;
do {
ret = __get_user_pages_fast(addr, 1, 0, &page);
if (!ret)
break;
offset = addr & (PAGE_SIZE - 1);
size = min(PAGE_SIZE - offset, n - len);
map = kmap_atomic(page);
memcpy(to, map+offset, size);
kunmap_atomic(map);
put_page(page);
len += size;
to += size;
addr += size;
} while (len < n);
return len;
}
struct event_constraint {
union {
unsigned long idxmsk[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
u64 idxmsk64;
};
u64 code;
u64 cmask;
int weight;
};
struct amd_nb {
int nb_id; /* NorthBridge id */
int refcnt; /* reference count */
struct perf_event *owners[X86_PMC_IDX_MAX];
struct event_constraint event_constraints[X86_PMC_IDX_MAX];
};
perf: Add support for supplementary event registers Change logs against Andi's original version: - Extends perf_event_attr:config to config{,1,2} (Peter Zijlstra) - Fixed a major event scheduling issue. There cannot be a ref++ on an event that has already done ref++ once and without calling put_constraint() in between. (Stephane Eranian) - Use thread_cpumask for percore allocation. (Lin Ming) - Use MSR names in the extra reg lists. (Lin Ming) - Remove redundant "c = NULL" in intel_percore_constraints - Fix comment of perf_event_attr::config1 Intel Nehalem/Westmere have a special OFFCORE_RESPONSE event that can be used to monitor any offcore accesses from a core. This is a very useful event for various tunings, and it's also needed to implement the generic LLC-* events correctly. Unfortunately this event requires programming a mask in a separate register. And worse this separate register is per core, not per CPU thread. This patch: - Teaches perf_events that OFFCORE_RESPONSE needs extra parameters. The extra parameters are passed by user space in the perf_event_attr::config1 field. - Adds support to the Intel perf_event core to schedule per core resources. This adds fairly generic infrastructure that can be also used for other per core resources. The basic code has is patterned after the similar AMD northbridge constraints code. Thanks to Stephane Eranian who pointed out some problems in the original version and suggested improvements. Signed-off-by: Andi Kleen <ak@linux.intel.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1299119690-13991-2-git-send-email-ming.m.lin@intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-03-03 09:34:47 +07:00
struct intel_percore;
#define MAX_LBR_ENTRIES 16
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
struct cpu_hw_events {
/*
* Generic x86 PMC bits
*/
struct perf_event *events[X86_PMC_IDX_MAX]; /* in counter order */
unsigned long active_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
unsigned long running[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
int enabled;
int n_events;
int n_added;
perf_events: Fix event scheduling issues introduced by transactional API The transactional API patch between the generic and model-specific code introduced several important bugs with event scheduling, at least on X86. If you had pinned events, e.g., watchdog, and were over-committing the PMU, you would get bogus counts. The bug was showing up on Intel CPU because events would move around more often that on AMD. But the problem also existed on AMD, though harder to expose. The issues were: - group_sched_in() was missing a cancel_txn() in the error path - cpuc->n_added was not properly maintained, leading to missing actions in hw_perf_enable(), i.e., n_running being 0. You cannot update n_added until you know the transaction has succeeded. In case of failed transaction n_added was not adjusted back. - in case of failed transactions, event_sched_out() was called and eventually invoked x86_disable_event() to touch the HW reg. But with transactions, on X86, event_sched_in() does not touch HW registers, it simply collects events into a list. Thus, you could end up calling x86_disable_event() on a counter which did not correspond to the current event when idx != -1. The patch modifies the generic and X86 code to avoid all those problems. First, we keep track of the number of events added last. In case the transaction fails, we substract them from n_added. This approach is necessary (as opposed to delaying updates to n_added) because not all event updates use the transaction API, e.g., single events. Second, we encapsulate the event_sched_in() and event_sched_out() in group_sched_in() inside the transaction. That makes the operations symmetrical and you can also detect that you are inside a transaction and skip the HW reg access by checking cpuc->group_flag. With this patch, you can now overcommit the PMU even with pinned system-wide events present and still get valid counts. Signed-off-by: Stephane Eranian <eranian@google.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1274796225.5882.1389.camel@twins> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-05-25 21:23:10 +07:00
int n_txn;
int assign[X86_PMC_IDX_MAX]; /* event to counter assignment */
u64 tags[X86_PMC_IDX_MAX];
struct perf_event *event_list[X86_PMC_IDX_MAX]; /* in enabled order */
unsigned int group_flag;
/*
* Intel DebugStore bits
*/
struct debug_store *ds;
u64 pebs_enabled;
/*
* Intel LBR bits
*/
int lbr_users;
void *lbr_context;
struct perf_branch_stack lbr_stack;
struct perf_branch_entry lbr_entries[MAX_LBR_ENTRIES];
perf: Add support for supplementary event registers Change logs against Andi's original version: - Extends perf_event_attr:config to config{,1,2} (Peter Zijlstra) - Fixed a major event scheduling issue. There cannot be a ref++ on an event that has already done ref++ once and without calling put_constraint() in between. (Stephane Eranian) - Use thread_cpumask for percore allocation. (Lin Ming) - Use MSR names in the extra reg lists. (Lin Ming) - Remove redundant "c = NULL" in intel_percore_constraints - Fix comment of perf_event_attr::config1 Intel Nehalem/Westmere have a special OFFCORE_RESPONSE event that can be used to monitor any offcore accesses from a core. This is a very useful event for various tunings, and it's also needed to implement the generic LLC-* events correctly. Unfortunately this event requires programming a mask in a separate register. And worse this separate register is per core, not per CPU thread. This patch: - Teaches perf_events that OFFCORE_RESPONSE needs extra parameters. The extra parameters are passed by user space in the perf_event_attr::config1 field. - Adds support to the Intel perf_event core to schedule per core resources. This adds fairly generic infrastructure that can be also used for other per core resources. The basic code has is patterned after the similar AMD northbridge constraints code. Thanks to Stephane Eranian who pointed out some problems in the original version and suggested improvements. Signed-off-by: Andi Kleen <ak@linux.intel.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1299119690-13991-2-git-send-email-ming.m.lin@intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-03-03 09:34:47 +07:00
/*
* Intel percore register state.
* Coordinate shared resources between HT threads.
*/
int percore_used; /* Used by this CPU? */
struct intel_percore *per_core;
/*
* AMD specific bits
*/
struct amd_nb *amd_nb;
};
#define __EVENT_CONSTRAINT(c, n, m, w) {\
{ .idxmsk64 = (n) }, \
.code = (c), \
.cmask = (m), \
.weight = (w), \
}
#define EVENT_CONSTRAINT(c, n, m) \
__EVENT_CONSTRAINT(c, n, m, HWEIGHT(n))
/*
* Constraint on the Event code.
*/
#define INTEL_EVENT_CONSTRAINT(c, n) \
EVENT_CONSTRAINT(c, n, ARCH_PERFMON_EVENTSEL_EVENT)
/*
* Constraint on the Event code + UMask + fixed-mask
*
* filter mask to validate fixed counter events.
* the following filters disqualify for fixed counters:
* - inv
* - edge
* - cnt-mask
* The other filters are supported by fixed counters.
* The any-thread option is supported starting with v3.
*/
#define FIXED_EVENT_CONSTRAINT(c, n) \
EVENT_CONSTRAINT(c, (1ULL << (32+n)), X86_RAW_EVENT_MASK)
/*
* Constraint on the Event code + UMask
*/
#define INTEL_UEVENT_CONSTRAINT(c, n) \
EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK)
#define EVENT_CONSTRAINT_END \
EVENT_CONSTRAINT(0, 0, 0)
#define for_each_event_constraint(e, c) \
for ((e) = (c); (e)->weight; (e)++)
perf: Add support for supplementary event registers Change logs against Andi's original version: - Extends perf_event_attr:config to config{,1,2} (Peter Zijlstra) - Fixed a major event scheduling issue. There cannot be a ref++ on an event that has already done ref++ once and without calling put_constraint() in between. (Stephane Eranian) - Use thread_cpumask for percore allocation. (Lin Ming) - Use MSR names in the extra reg lists. (Lin Ming) - Remove redundant "c = NULL" in intel_percore_constraints - Fix comment of perf_event_attr::config1 Intel Nehalem/Westmere have a special OFFCORE_RESPONSE event that can be used to monitor any offcore accesses from a core. This is a very useful event for various tunings, and it's also needed to implement the generic LLC-* events correctly. Unfortunately this event requires programming a mask in a separate register. And worse this separate register is per core, not per CPU thread. This patch: - Teaches perf_events that OFFCORE_RESPONSE needs extra parameters. The extra parameters are passed by user space in the perf_event_attr::config1 field. - Adds support to the Intel perf_event core to schedule per core resources. This adds fairly generic infrastructure that can be also used for other per core resources. The basic code has is patterned after the similar AMD northbridge constraints code. Thanks to Stephane Eranian who pointed out some problems in the original version and suggested improvements. Signed-off-by: Andi Kleen <ak@linux.intel.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1299119690-13991-2-git-send-email-ming.m.lin@intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-03-03 09:34:47 +07:00
/*
* Extra registers for specific events.
* Some events need large masks and require external MSRs.
* Define a mapping to these extra registers.
*/
struct extra_reg {
unsigned int event;
unsigned int msr;
u64 config_mask;
u64 valid_mask;
};
#define EVENT_EXTRA_REG(e, ms, m, vm) { \
.event = (e), \
.msr = (ms), \
.config_mask = (m), \
.valid_mask = (vm), \
}
#define INTEL_EVENT_EXTRA_REG(event, msr, vm) \
EVENT_EXTRA_REG(event, msr, ARCH_PERFMON_EVENTSEL_EVENT, vm)
#define EVENT_EXTRA_END EVENT_EXTRA_REG(0, 0, 0, 0)
union perf_capabilities {
struct {
u64 lbr_format : 6;
u64 pebs_trap : 1;
u64 pebs_arch_reg : 1;
u64 pebs_format : 4;
u64 smm_freeze : 1;
};
u64 capabilities;
};
/*
* struct x86_pmu - generic x86 pmu
*/
struct x86_pmu {
/*
* Generic x86 PMC bits
*/
const char *name;
int version;
int (*handle_irq)(struct pt_regs *);
void (*disable_all)(void);
void (*enable_all)(int added);
void (*enable)(struct perf_event *);
void (*disable)(struct perf_event *);
int (*hw_config)(struct perf_event *event);
perf, x86: Implement initial P4 PMU driver The netburst PMU is way different from the "architectural perfomance monitoring" specification that current CPUs use. P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle perfomance monitoring events. A few implementational details: 1) We need a separate x86_pmu::hw_config helper in struct x86_pmu since register bit-fields are quite different from P6, Core and later cpu series. 2) For the same reason is a x86_pmu::schedule_events helper introduced. 3) hw_perf_event::config consists of packed ESCR+CCCR values. It's allowed since in reality both registers only use a half of their size. Of course before making a real write into a particular MSR we need to unpack the value and extend it to a proper size. 4) The tuple of packed ESCR+CCCR in hw_perf_event::config doesn't describe the memory address of ESCR MSR register so that we need to keep a mapping between these tuples used and available ESCR (various P4 events may use same ESCRs but not simultaneously), for this sake every active event has a per-cpu map of hw_perf_event::idx <--> ESCR addresses. 5) Since hw_perf_event::idx is an offset to counter/control register we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel strips it down to 8 registers and event armed may never be turned off (ie the bit in active_mask is set but the loop never reaches this index to check), thanks to Peter Zijlstra Restrictions: - No cascaded counters support (do we ever need them?) - No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS doesn't work for now) - There are events with same counters which can't work simultaneously (need to use intersected ones due to broken counter 1) - No PERF_COUNT_HW_CACHE_ events yet Todo: - Implement dependent events - Need proper hashing for event opcodes (no linear search, good for debugging stage but not in real loads) - Some events counted during a clock cycle -- need to set threshold for them and count every clock cycle just to get summary statistics (ie to behave the same way as other PMUs do) - Need to swicth to use event_constraints - To support RAW events we need to encode a global list of P4 events into p4_templates - Cache events need to be added Event support status matrix: Event status ----------------------------- cycles works cache-references works cache-misses works branch-misses works bus-cycles partially (does not work on 64bit cpu with HT enabled) instruction doesnt work (needs dependent event [mop tagging]) branches doesnt work Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Robert Richter <robert.richter@amd.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20100311165439.GB5129@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 23:54:39 +07:00
int (*schedule_events)(struct cpu_hw_events *cpuc, int n, int *assign);
unsigned eventsel;
unsigned perfctr;
u64 (*event_map)(int);
int max_events;
int num_counters;
int num_counters_fixed;
int cntval_bits;
u64 cntval_mask;
int apic;
u64 max_period;
struct event_constraint *
(*get_event_constraints)(struct cpu_hw_events *cpuc,
struct perf_event *event);
void (*put_event_constraints)(struct cpu_hw_events *cpuc,
struct perf_event *event);
struct event_constraint *event_constraints;
perf: Add support for supplementary event registers Change logs against Andi's original version: - Extends perf_event_attr:config to config{,1,2} (Peter Zijlstra) - Fixed a major event scheduling issue. There cannot be a ref++ on an event that has already done ref++ once and without calling put_constraint() in between. (Stephane Eranian) - Use thread_cpumask for percore allocation. (Lin Ming) - Use MSR names in the extra reg lists. (Lin Ming) - Remove redundant "c = NULL" in intel_percore_constraints - Fix comment of perf_event_attr::config1 Intel Nehalem/Westmere have a special OFFCORE_RESPONSE event that can be used to monitor any offcore accesses from a core. This is a very useful event for various tunings, and it's also needed to implement the generic LLC-* events correctly. Unfortunately this event requires programming a mask in a separate register. And worse this separate register is per core, not per CPU thread. This patch: - Teaches perf_events that OFFCORE_RESPONSE needs extra parameters. The extra parameters are passed by user space in the perf_event_attr::config1 field. - Adds support to the Intel perf_event core to schedule per core resources. This adds fairly generic infrastructure that can be also used for other per core resources. The basic code has is patterned after the similar AMD northbridge constraints code. Thanks to Stephane Eranian who pointed out some problems in the original version and suggested improvements. Signed-off-by: Andi Kleen <ak@linux.intel.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1299119690-13991-2-git-send-email-ming.m.lin@intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-03-03 09:34:47 +07:00
struct event_constraint *percore_constraints;
void (*quirks)(void);
int perfctr_second_write;
int (*cpu_prepare)(int cpu);
void (*cpu_starting)(int cpu);
void (*cpu_dying)(int cpu);
void (*cpu_dead)(int cpu);
/*
* Intel Arch Perfmon v2+
*/
u64 intel_ctrl;
union perf_capabilities intel_cap;
/*
* Intel DebugStore bits
*/
int bts, pebs;
int bts_active, pebs_active;
int pebs_record_size;
void (*drain_pebs)(struct pt_regs *regs);
struct event_constraint *pebs_constraints;
/*
* Intel LBR
*/
unsigned long lbr_tos, lbr_from, lbr_to; /* MSR base regs */
int lbr_nr; /* hardware stack size */
perf: Add support for supplementary event registers Change logs against Andi's original version: - Extends perf_event_attr:config to config{,1,2} (Peter Zijlstra) - Fixed a major event scheduling issue. There cannot be a ref++ on an event that has already done ref++ once and without calling put_constraint() in between. (Stephane Eranian) - Use thread_cpumask for percore allocation. (Lin Ming) - Use MSR names in the extra reg lists. (Lin Ming) - Remove redundant "c = NULL" in intel_percore_constraints - Fix comment of perf_event_attr::config1 Intel Nehalem/Westmere have a special OFFCORE_RESPONSE event that can be used to monitor any offcore accesses from a core. This is a very useful event for various tunings, and it's also needed to implement the generic LLC-* events correctly. Unfortunately this event requires programming a mask in a separate register. And worse this separate register is per core, not per CPU thread. This patch: - Teaches perf_events that OFFCORE_RESPONSE needs extra parameters. The extra parameters are passed by user space in the perf_event_attr::config1 field. - Adds support to the Intel perf_event core to schedule per core resources. This adds fairly generic infrastructure that can be also used for other per core resources. The basic code has is patterned after the similar AMD northbridge constraints code. Thanks to Stephane Eranian who pointed out some problems in the original version and suggested improvements. Signed-off-by: Andi Kleen <ak@linux.intel.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1299119690-13991-2-git-send-email-ming.m.lin@intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-03-03 09:34:47 +07:00
/*
* Extra registers for events
*/
struct extra_reg *extra_regs;
};
static struct x86_pmu x86_pmu __read_mostly;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
static DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events) = {
.enabled = 1,
};
static int x86_perf_event_set_period(struct perf_event *event);
/*
* Generalized hw caching related hw_event table, filled
* in on a per model basis. A value of 0 means
* 'not supported', -1 means 'hw_event makes no sense on
* this CPU', any other value means the raw hw_event
* ID.
*/
#define C(x) PERF_COUNT_HW_CACHE_##x
static u64 __read_mostly hw_cache_event_ids
[PERF_COUNT_HW_CACHE_MAX]
[PERF_COUNT_HW_CACHE_OP_MAX]
[PERF_COUNT_HW_CACHE_RESULT_MAX];
static u64 __read_mostly hw_cache_extra_regs
[PERF_COUNT_HW_CACHE_MAX]
[PERF_COUNT_HW_CACHE_OP_MAX]
[PERF_COUNT_HW_CACHE_RESULT_MAX];
/*
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
* Propagate event elapsed time into the generic event.
* Can only be executed on the CPU where the event is active.
* Returns the delta events processed.
*/
static u64
x86_perf_event_update(struct perf_event *event)
{
struct hw_perf_event *hwc = &event->hw;
int shift = 64 - x86_pmu.cntval_bits;
u64 prev_raw_count, new_raw_count;
int idx = hwc->idx;
s64 delta;
if (idx == X86_PMC_IDX_FIXED_BTS)
return 0;
/*
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
* Careful: an NMI might modify the previous event value.
*
* Our tactic to handle this is to first atomically read and
* exchange a new raw count - then add that new-prev delta
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
* count to the generic event atomically:
*/
again:
prev_raw_count = local64_read(&hwc->prev_count);
rdmsrl(hwc->event_base, new_raw_count);
if (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
new_raw_count) != prev_raw_count)
goto again;
/*
* Now we have the new raw value and have updated the prev
* timestamp already. We can now calculate the elapsed delta
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
* (event-)time and add that to the generic event.
*
* Careful, not all hw sign-extends above the physical width
* of the count.
*/
delta = (new_raw_count << shift) - (prev_raw_count << shift);
delta >>= shift;
local64_add(delta, &event->count);
local64_sub(delta, &hwc->period_left);
return new_raw_count;
}
/* using X86_FEATURE_PERFCTR_CORE to later implement ALTERNATIVE() here */
static inline int x86_pmu_addr_offset(int index)
{
if (boot_cpu_has(X86_FEATURE_PERFCTR_CORE))
return index << 1;
return index;
}
static inline unsigned int x86_pmu_config_addr(int index)
{
return x86_pmu.eventsel + x86_pmu_addr_offset(index);
}
static inline unsigned int x86_pmu_event_addr(int index)
{
return x86_pmu.perfctr + x86_pmu_addr_offset(index);
}
perf: Add support for supplementary event registers Change logs against Andi's original version: - Extends perf_event_attr:config to config{,1,2} (Peter Zijlstra) - Fixed a major event scheduling issue. There cannot be a ref++ on an event that has already done ref++ once and without calling put_constraint() in between. (Stephane Eranian) - Use thread_cpumask for percore allocation. (Lin Ming) - Use MSR names in the extra reg lists. (Lin Ming) - Remove redundant "c = NULL" in intel_percore_constraints - Fix comment of perf_event_attr::config1 Intel Nehalem/Westmere have a special OFFCORE_RESPONSE event that can be used to monitor any offcore accesses from a core. This is a very useful event for various tunings, and it's also needed to implement the generic LLC-* events correctly. Unfortunately this event requires programming a mask in a separate register. And worse this separate register is per core, not per CPU thread. This patch: - Teaches perf_events that OFFCORE_RESPONSE needs extra parameters. The extra parameters are passed by user space in the perf_event_attr::config1 field. - Adds support to the Intel perf_event core to schedule per core resources. This adds fairly generic infrastructure that can be also used for other per core resources. The basic code has is patterned after the similar AMD northbridge constraints code. Thanks to Stephane Eranian who pointed out some problems in the original version and suggested improvements. Signed-off-by: Andi Kleen <ak@linux.intel.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1299119690-13991-2-git-send-email-ming.m.lin@intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-03-03 09:34:47 +07:00
/*
* Find and validate any extra registers to set up.
*/
static int x86_pmu_extra_regs(u64 config, struct perf_event *event)
{
struct extra_reg *er;
event->hw.extra_reg = 0;
event->hw.extra_config = 0;
if (!x86_pmu.extra_regs)
return 0;
for (er = x86_pmu.extra_regs; er->msr; er++) {
if (er->event != (config & er->config_mask))
continue;
if (event->attr.config1 & ~er->valid_mask)
return -EINVAL;
event->hw.extra_reg = er->msr;
event->hw.extra_config = event->attr.config1;
break;
}
return 0;
}
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
static atomic_t active_events;
static DEFINE_MUTEX(pmc_reserve_mutex);
#ifdef CONFIG_X86_LOCAL_APIC
static bool reserve_pmc_hardware(void)
{
int i;
for (i = 0; i < x86_pmu.num_counters; i++) {
if (!reserve_perfctr_nmi(x86_pmu_event_addr(i)))
goto perfctr_fail;
}
for (i = 0; i < x86_pmu.num_counters; i++) {
if (!reserve_evntsel_nmi(x86_pmu_config_addr(i)))
goto eventsel_fail;
}
return true;
eventsel_fail:
for (i--; i >= 0; i--)
release_evntsel_nmi(x86_pmu_config_addr(i));
i = x86_pmu.num_counters;
perfctr_fail:
for (i--; i >= 0; i--)
release_perfctr_nmi(x86_pmu_event_addr(i));
return false;
}
static void release_pmc_hardware(void)
{
int i;
for (i = 0; i < x86_pmu.num_counters; i++) {
release_perfctr_nmi(x86_pmu_event_addr(i));
release_evntsel_nmi(x86_pmu_config_addr(i));
}
}
#else
static bool reserve_pmc_hardware(void) { return true; }
static void release_pmc_hardware(void) {}
#endif
static bool check_hw_exists(void)
{
u64 val, val_new = 0;
int i, reg, ret = 0;
/*
* Check to see if the BIOS enabled any of the counters, if so
* complain and bail.
*/
for (i = 0; i < x86_pmu.num_counters; i++) {
reg = x86_pmu_config_addr(i);
ret = rdmsrl_safe(reg, &val);
if (ret)
goto msr_fail;
if (val & ARCH_PERFMON_EVENTSEL_ENABLE)
goto bios_fail;
}
if (x86_pmu.num_counters_fixed) {
reg = MSR_ARCH_PERFMON_FIXED_CTR_CTRL;
ret = rdmsrl_safe(reg, &val);
if (ret)
goto msr_fail;
for (i = 0; i < x86_pmu.num_counters_fixed; i++) {
if (val & (0x03 << i*4))
goto bios_fail;
}
}
/*
* Now write a value and read it back to see if it matches,
* this is needed to detect certain hardware emulators (qemu/kvm)
* that don't trap on the MSR access and always return 0s.
*/
val = 0xabcdUL;
ret = checking_wrmsrl(x86_pmu_event_addr(0), val);
ret |= rdmsrl_safe(x86_pmu_event_addr(0), &val_new);
if (ret || val != val_new)
goto msr_fail;
return true;
bios_fail:
/*
* We still allow the PMU driver to operate:
*/
printk(KERN_CONT "Broken BIOS detected, complain to your hardware vendor.\n");
printk(KERN_ERR FW_BUG "the BIOS has corrupted hw-PMU resources (MSR %x is %Lx)\n", reg, val);
return true;
msr_fail:
printk(KERN_CONT "Broken PMU hardware detected, using software events only.\n");
return false;
}
static void reserve_ds_buffers(void);
static void release_ds_buffers(void);
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
static void hw_perf_event_destroy(struct perf_event *event)
{
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
if (atomic_dec_and_mutex_lock(&active_events, &pmc_reserve_mutex)) {
release_pmc_hardware();
release_ds_buffers();
mutex_unlock(&pmc_reserve_mutex);
}
}
static inline int x86_pmu_initialized(void)
{
return x86_pmu.handle_irq != NULL;
}
static inline int
set_ext_hw_attr(struct hw_perf_event *hwc, struct perf_event *event)
{
struct perf_event_attr *attr = &event->attr;
unsigned int cache_type, cache_op, cache_result;
u64 config, val;
config = attr->config;
cache_type = (config >> 0) & 0xff;
if (cache_type >= PERF_COUNT_HW_CACHE_MAX)
return -EINVAL;
cache_op = (config >> 8) & 0xff;
if (cache_op >= PERF_COUNT_HW_CACHE_OP_MAX)
return -EINVAL;
cache_result = (config >> 16) & 0xff;
if (cache_result >= PERF_COUNT_HW_CACHE_RESULT_MAX)
return -EINVAL;
val = hw_cache_event_ids[cache_type][cache_op][cache_result];
if (val == 0)
return -ENOENT;
if (val == -1)
return -EINVAL;
hwc->config |= val;
attr->config1 = hw_cache_extra_regs[cache_type][cache_op][cache_result];
return x86_pmu_extra_regs(val, event);
}
static int x86_setup_perfctr(struct perf_event *event)
{
struct perf_event_attr *attr = &event->attr;
struct hw_perf_event *hwc = &event->hw;
u64 config;
if (!is_sampling_event(event)) {
hwc->sample_period = x86_pmu.max_period;
hwc->last_period = hwc->sample_period;
local64_set(&hwc->period_left, hwc->sample_period);
} else {
/*
* If we have a PMU initialized but no APIC
* interrupts, we cannot sample hardware
* events (user-space has to fall back and
* sample via a hrtimer based software event):
*/
if (!x86_pmu.apic)
return -EOPNOTSUPP;
}
x86, perf event: Turn off unstructured raw event access to offcore registers Andi Kleen pointed out that the Intel offcore support patches were merged without user-space tool support to the functionality: | | The offcore_msr perf kernel code was merged into 2.6.39-rc*, but the | user space bits were not. This made it impossible to set the extra mask | and actually do the OFFCORE profiling | Andi submitted a preliminary patch for user-space support, as an extension to perf's raw event syntax: | | Some raw events -- like the Intel OFFCORE events -- support additional | parameters. These can be appended after a ':'. | | For example on a multi socket Intel Nehalem: | | perf stat -e r1b7:20ff -a sleep 1 | | Profile the OFFCORE_RESPONSE.ANY_REQUEST with event mask REMOTE_DRAM_0 | that measures any access to DRAM on another socket. | But this kind of usability is absolutely unacceptable - users should not be expected to type in magic, CPU and model specific incantations to get access to useful hardware functionality. The proper solution is to expose useful offcore functionality via generalized events - that way users do not have to care which specific CPU model they are using, they can use the conceptual event and not some model specific quirky hexa number. We already have such generalization in place for CPU cache events, and it's all very extensible. "Offcore" events measure general DRAM access patters along various parameters. They are particularly useful in NUMA systems. We want to support them via generalized DRAM events: either as the fourth level of cache (after the last-level cache), or as a separate generalization category. That way user-space support would be very obvious, memory access profiling could be done via self-explanatory commands like: perf record -e dram ./myapp perf record -e dram-remote ./myapp ... to measure DRAM accesses or more expensive cross-node NUMA DRAM accesses. These generalized events would work on all CPUs and architectures that have comparable PMU features. ( Note, these are just examples: actual implementation could have more sophistication and more parameter - as long as they center around similarly simple usecases. ) Now we do not want to revert *all* of the current offcore bits, as they are still somewhat useful for generic last-level-cache events, implemented in this commit: e994d7d23a0b: perf: Fix LLC-* events on Intel Nehalem/Westmere But we definitely do not yet want to expose the unstructured raw events to user-space, until better generalization and usability is implemented for these hardware event features. ( Note: after generalization has been implemented raw offcore events can be supported as well: there can always be an odd event that is marginally useful but not useful enough to generalize. DRAM profiling is definitely *not* such a category so generalization must be done first. ) Furthermore, PERF_TYPE_RAW access to these registers was not intended to go upstream without proper support - it was a side-effect of the above e994d7d23a0b commit, not mentioned in the changelog. As v2.6.39 is nearing release we go for the simplest approach: disable the PERF_TYPE_RAW offcore hack for now, before it escapes into a released kernel and becomes an ABI. Once proper structure is implemented for these hardware events and users are offered usable solutions we can revisit this issue. Reported-by: Andi Kleen <ak@linux.intel.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Link: http://lkml.kernel.org/r/1302658203-4239-1-git-send-email-andi@firstfloor.org Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-04-22 13:44:38 +07:00
/*
* Do not allow config1 (extended registers) to propagate,
* there's no sane user-space generalization yet:
*/
if (attr->type == PERF_TYPE_RAW)
x86, perf event: Turn off unstructured raw event access to offcore registers Andi Kleen pointed out that the Intel offcore support patches were merged without user-space tool support to the functionality: | | The offcore_msr perf kernel code was merged into 2.6.39-rc*, but the | user space bits were not. This made it impossible to set the extra mask | and actually do the OFFCORE profiling | Andi submitted a preliminary patch for user-space support, as an extension to perf's raw event syntax: | | Some raw events -- like the Intel OFFCORE events -- support additional | parameters. These can be appended after a ':'. | | For example on a multi socket Intel Nehalem: | | perf stat -e r1b7:20ff -a sleep 1 | | Profile the OFFCORE_RESPONSE.ANY_REQUEST with event mask REMOTE_DRAM_0 | that measures any access to DRAM on another socket. | But this kind of usability is absolutely unacceptable - users should not be expected to type in magic, CPU and model specific incantations to get access to useful hardware functionality. The proper solution is to expose useful offcore functionality via generalized events - that way users do not have to care which specific CPU model they are using, they can use the conceptual event and not some model specific quirky hexa number. We already have such generalization in place for CPU cache events, and it's all very extensible. "Offcore" events measure general DRAM access patters along various parameters. They are particularly useful in NUMA systems. We want to support them via generalized DRAM events: either as the fourth level of cache (after the last-level cache), or as a separate generalization category. That way user-space support would be very obvious, memory access profiling could be done via self-explanatory commands like: perf record -e dram ./myapp perf record -e dram-remote ./myapp ... to measure DRAM accesses or more expensive cross-node NUMA DRAM accesses. These generalized events would work on all CPUs and architectures that have comparable PMU features. ( Note, these are just examples: actual implementation could have more sophistication and more parameter - as long as they center around similarly simple usecases. ) Now we do not want to revert *all* of the current offcore bits, as they are still somewhat useful for generic last-level-cache events, implemented in this commit: e994d7d23a0b: perf: Fix LLC-* events on Intel Nehalem/Westmere But we definitely do not yet want to expose the unstructured raw events to user-space, until better generalization and usability is implemented for these hardware event features. ( Note: after generalization has been implemented raw offcore events can be supported as well: there can always be an odd event that is marginally useful but not useful enough to generalize. DRAM profiling is definitely *not* such a category so generalization must be done first. ) Furthermore, PERF_TYPE_RAW access to these registers was not intended to go upstream without proper support - it was a side-effect of the above e994d7d23a0b commit, not mentioned in the changelog. As v2.6.39 is nearing release we go for the simplest approach: disable the PERF_TYPE_RAW offcore hack for now, before it escapes into a released kernel and becomes an ABI. Once proper structure is implemented for these hardware events and users are offered usable solutions we can revisit this issue. Reported-by: Andi Kleen <ak@linux.intel.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Link: http://lkml.kernel.org/r/1302658203-4239-1-git-send-email-andi@firstfloor.org Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-04-22 13:44:38 +07:00
return 0;
if (attr->type == PERF_TYPE_HW_CACHE)
return set_ext_hw_attr(hwc, event);
if (attr->config >= x86_pmu.max_events)
return -EINVAL;
/*
* The generic map:
*/
config = x86_pmu.event_map(attr->config);
if (config == 0)
return -ENOENT;
if (config == -1LL)
return -EINVAL;
/*
* Branch tracing:
*/
if (attr->config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS &&
!attr->freq && hwc->sample_period == 1) {
/* BTS is not supported by this architecture. */
if (!x86_pmu.bts_active)
return -EOPNOTSUPP;
/* BTS is currently only allowed for user-mode. */
if (!attr->exclude_kernel)
return -EOPNOTSUPP;
}
hwc->config |= config;
return 0;
}
static int x86_pmu_hw_config(struct perf_event *event)
perf, x86: Implement initial P4 PMU driver The netburst PMU is way different from the "architectural perfomance monitoring" specification that current CPUs use. P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle perfomance monitoring events. A few implementational details: 1) We need a separate x86_pmu::hw_config helper in struct x86_pmu since register bit-fields are quite different from P6, Core and later cpu series. 2) For the same reason is a x86_pmu::schedule_events helper introduced. 3) hw_perf_event::config consists of packed ESCR+CCCR values. It's allowed since in reality both registers only use a half of their size. Of course before making a real write into a particular MSR we need to unpack the value and extend it to a proper size. 4) The tuple of packed ESCR+CCCR in hw_perf_event::config doesn't describe the memory address of ESCR MSR register so that we need to keep a mapping between these tuples used and available ESCR (various P4 events may use same ESCRs but not simultaneously), for this sake every active event has a per-cpu map of hw_perf_event::idx <--> ESCR addresses. 5) Since hw_perf_event::idx is an offset to counter/control register we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel strips it down to 8 registers and event armed may never be turned off (ie the bit in active_mask is set but the loop never reaches this index to check), thanks to Peter Zijlstra Restrictions: - No cascaded counters support (do we ever need them?) - No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS doesn't work for now) - There are events with same counters which can't work simultaneously (need to use intersected ones due to broken counter 1) - No PERF_COUNT_HW_CACHE_ events yet Todo: - Implement dependent events - Need proper hashing for event opcodes (no linear search, good for debugging stage but not in real loads) - Some events counted during a clock cycle -- need to set threshold for them and count every clock cycle just to get summary statistics (ie to behave the same way as other PMUs do) - Need to swicth to use event_constraints - To support RAW events we need to encode a global list of P4 events into p4_templates - Cache events need to be added Event support status matrix: Event status ----------------------------- cycles works cache-references works cache-misses works branch-misses works bus-cycles partially (does not work on 64bit cpu with HT enabled) instruction doesnt work (needs dependent event [mop tagging]) branches doesnt work Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Robert Richter <robert.richter@amd.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20100311165439.GB5129@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 23:54:39 +07:00
{
if (event->attr.precise_ip) {
int precise = 0;
/* Support for constant skid */
if (x86_pmu.pebs_active) {
precise++;
/* Support for IP fixup */
if (x86_pmu.lbr_nr)
precise++;
}
if (event->attr.precise_ip > precise)
return -EOPNOTSUPP;
}
perf, x86: Implement initial P4 PMU driver The netburst PMU is way different from the "architectural perfomance monitoring" specification that current CPUs use. P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle perfomance monitoring events. A few implementational details: 1) We need a separate x86_pmu::hw_config helper in struct x86_pmu since register bit-fields are quite different from P6, Core and later cpu series. 2) For the same reason is a x86_pmu::schedule_events helper introduced. 3) hw_perf_event::config consists of packed ESCR+CCCR values. It's allowed since in reality both registers only use a half of their size. Of course before making a real write into a particular MSR we need to unpack the value and extend it to a proper size. 4) The tuple of packed ESCR+CCCR in hw_perf_event::config doesn't describe the memory address of ESCR MSR register so that we need to keep a mapping between these tuples used and available ESCR (various P4 events may use same ESCRs but not simultaneously), for this sake every active event has a per-cpu map of hw_perf_event::idx <--> ESCR addresses. 5) Since hw_perf_event::idx is an offset to counter/control register we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel strips it down to 8 registers and event armed may never be turned off (ie the bit in active_mask is set but the loop never reaches this index to check), thanks to Peter Zijlstra Restrictions: - No cascaded counters support (do we ever need them?) - No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS doesn't work for now) - There are events with same counters which can't work simultaneously (need to use intersected ones due to broken counter 1) - No PERF_COUNT_HW_CACHE_ events yet Todo: - Implement dependent events - Need proper hashing for event opcodes (no linear search, good for debugging stage but not in real loads) - Some events counted during a clock cycle -- need to set threshold for them and count every clock cycle just to get summary statistics (ie to behave the same way as other PMUs do) - Need to swicth to use event_constraints - To support RAW events we need to encode a global list of P4 events into p4_templates - Cache events need to be added Event support status matrix: Event status ----------------------------- cycles works cache-references works cache-misses works branch-misses works bus-cycles partially (does not work on 64bit cpu with HT enabled) instruction doesnt work (needs dependent event [mop tagging]) branches doesnt work Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Robert Richter <robert.richter@amd.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20100311165439.GB5129@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 23:54:39 +07:00
/*
* Generate PMC IRQs:
* (keep 'enabled' bit clear for now)
*/
event->hw.config = ARCH_PERFMON_EVENTSEL_INT;
perf, x86: Implement initial P4 PMU driver The netburst PMU is way different from the "architectural perfomance monitoring" specification that current CPUs use. P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle perfomance monitoring events. A few implementational details: 1) We need a separate x86_pmu::hw_config helper in struct x86_pmu since register bit-fields are quite different from P6, Core and later cpu series. 2) For the same reason is a x86_pmu::schedule_events helper introduced. 3) hw_perf_event::config consists of packed ESCR+CCCR values. It's allowed since in reality both registers only use a half of their size. Of course before making a real write into a particular MSR we need to unpack the value and extend it to a proper size. 4) The tuple of packed ESCR+CCCR in hw_perf_event::config doesn't describe the memory address of ESCR MSR register so that we need to keep a mapping between these tuples used and available ESCR (various P4 events may use same ESCRs but not simultaneously), for this sake every active event has a per-cpu map of hw_perf_event::idx <--> ESCR addresses. 5) Since hw_perf_event::idx is an offset to counter/control register we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel strips it down to 8 registers and event armed may never be turned off (ie the bit in active_mask is set but the loop never reaches this index to check), thanks to Peter Zijlstra Restrictions: - No cascaded counters support (do we ever need them?) - No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS doesn't work for now) - There are events with same counters which can't work simultaneously (need to use intersected ones due to broken counter 1) - No PERF_COUNT_HW_CACHE_ events yet Todo: - Implement dependent events - Need proper hashing for event opcodes (no linear search, good for debugging stage but not in real loads) - Some events counted during a clock cycle -- need to set threshold for them and count every clock cycle just to get summary statistics (ie to behave the same way as other PMUs do) - Need to swicth to use event_constraints - To support RAW events we need to encode a global list of P4 events into p4_templates - Cache events need to be added Event support status matrix: Event status ----------------------------- cycles works cache-references works cache-misses works branch-misses works bus-cycles partially (does not work on 64bit cpu with HT enabled) instruction doesnt work (needs dependent event [mop tagging]) branches doesnt work Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Robert Richter <robert.richter@amd.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20100311165439.GB5129@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 23:54:39 +07:00
/*
* Count user and OS events unless requested not to
*/
if (!event->attr.exclude_user)
event->hw.config |= ARCH_PERFMON_EVENTSEL_USR;
if (!event->attr.exclude_kernel)
event->hw.config |= ARCH_PERFMON_EVENTSEL_OS;
perf, x86: Implement initial P4 PMU driver The netburst PMU is way different from the "architectural perfomance monitoring" specification that current CPUs use. P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle perfomance monitoring events. A few implementational details: 1) We need a separate x86_pmu::hw_config helper in struct x86_pmu since register bit-fields are quite different from P6, Core and later cpu series. 2) For the same reason is a x86_pmu::schedule_events helper introduced. 3) hw_perf_event::config consists of packed ESCR+CCCR values. It's allowed since in reality both registers only use a half of their size. Of course before making a real write into a particular MSR we need to unpack the value and extend it to a proper size. 4) The tuple of packed ESCR+CCCR in hw_perf_event::config doesn't describe the memory address of ESCR MSR register so that we need to keep a mapping between these tuples used and available ESCR (various P4 events may use same ESCRs but not simultaneously), for this sake every active event has a per-cpu map of hw_perf_event::idx <--> ESCR addresses. 5) Since hw_perf_event::idx is an offset to counter/control register we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel strips it down to 8 registers and event armed may never be turned off (ie the bit in active_mask is set but the loop never reaches this index to check), thanks to Peter Zijlstra Restrictions: - No cascaded counters support (do we ever need them?) - No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS doesn't work for now) - There are events with same counters which can't work simultaneously (need to use intersected ones due to broken counter 1) - No PERF_COUNT_HW_CACHE_ events yet Todo: - Implement dependent events - Need proper hashing for event opcodes (no linear search, good for debugging stage but not in real loads) - Some events counted during a clock cycle -- need to set threshold for them and count every clock cycle just to get summary statistics (ie to behave the same way as other PMUs do) - Need to swicth to use event_constraints - To support RAW events we need to encode a global list of P4 events into p4_templates - Cache events need to be added Event support status matrix: Event status ----------------------------- cycles works cache-references works cache-misses works branch-misses works bus-cycles partially (does not work on 64bit cpu with HT enabled) instruction doesnt work (needs dependent event [mop tagging]) branches doesnt work Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Robert Richter <robert.richter@amd.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20100311165439.GB5129@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 23:54:39 +07:00
if (event->attr.type == PERF_TYPE_RAW)
event->hw.config |= event->attr.config & X86_RAW_EVENT_MASK;
perf, x86: Implement initial P4 PMU driver The netburst PMU is way different from the "architectural perfomance monitoring" specification that current CPUs use. P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle perfomance monitoring events. A few implementational details: 1) We need a separate x86_pmu::hw_config helper in struct x86_pmu since register bit-fields are quite different from P6, Core and later cpu series. 2) For the same reason is a x86_pmu::schedule_events helper introduced. 3) hw_perf_event::config consists of packed ESCR+CCCR values. It's allowed since in reality both registers only use a half of their size. Of course before making a real write into a particular MSR we need to unpack the value and extend it to a proper size. 4) The tuple of packed ESCR+CCCR in hw_perf_event::config doesn't describe the memory address of ESCR MSR register so that we need to keep a mapping between these tuples used and available ESCR (various P4 events may use same ESCRs but not simultaneously), for this sake every active event has a per-cpu map of hw_perf_event::idx <--> ESCR addresses. 5) Since hw_perf_event::idx is an offset to counter/control register we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel strips it down to 8 registers and event armed may never be turned off (ie the bit in active_mask is set but the loop never reaches this index to check), thanks to Peter Zijlstra Restrictions: - No cascaded counters support (do we ever need them?) - No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS doesn't work for now) - There are events with same counters which can't work simultaneously (need to use intersected ones due to broken counter 1) - No PERF_COUNT_HW_CACHE_ events yet Todo: - Implement dependent events - Need proper hashing for event opcodes (no linear search, good for debugging stage but not in real loads) - Some events counted during a clock cycle -- need to set threshold for them and count every clock cycle just to get summary statistics (ie to behave the same way as other PMUs do) - Need to swicth to use event_constraints - To support RAW events we need to encode a global list of P4 events into p4_templates - Cache events need to be added Event support status matrix: Event status ----------------------------- cycles works cache-references works cache-misses works branch-misses works bus-cycles partially (does not work on 64bit cpu with HT enabled) instruction doesnt work (needs dependent event [mop tagging]) branches doesnt work Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Robert Richter <robert.richter@amd.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20100311165439.GB5129@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 23:54:39 +07:00
return x86_setup_perfctr(event);
}
/*
* Setup the hardware configuration for a given attr_type
*/
static int __x86_pmu_event_init(struct perf_event *event)
{
int err;
if (!x86_pmu_initialized())
return -ENODEV;
err = 0;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
if (!atomic_inc_not_zero(&active_events)) {
mutex_lock(&pmc_reserve_mutex);
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
if (atomic_read(&active_events) == 0) {
if (!reserve_pmc_hardware())
err = -EBUSY;
else
reserve_ds_buffers();
}
if (!err)
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
atomic_inc(&active_events);
mutex_unlock(&pmc_reserve_mutex);
}
if (err)
return err;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
event->destroy = hw_perf_event_destroy;
event->hw.idx = -1;
event->hw.last_cpu = -1;
event->hw.last_tag = ~0ULL;
return x86_pmu.hw_config(event);
}
static void x86_pmu_disable_all(void)
{
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
int idx;
for (idx = 0; idx < x86_pmu.num_counters; idx++) {
u64 val;
if (!test_bit(idx, cpuc->active_mask))
continue;
rdmsrl(x86_pmu_config_addr(idx), val);
if (!(val & ARCH_PERFMON_EVENTSEL_ENABLE))
continue;
val &= ~ARCH_PERFMON_EVENTSEL_ENABLE;
wrmsrl(x86_pmu_config_addr(idx), val);
}
}
2010-06-16 19:37:10 +07:00
static void x86_pmu_disable(struct pmu *pmu)
{
struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
if (!x86_pmu_initialized())
return;
if (!cpuc->enabled)
return;
cpuc->n_added = 0;
cpuc->enabled = 0;
barrier();
x86_pmu.disable_all();
}
static inline void __x86_pmu_enable_event(struct hw_perf_event *hwc,
u64 enable_mask)
{
perf: Add support for supplementary event registers Change logs against Andi's original version: - Extends perf_event_attr:config to config{,1,2} (Peter Zijlstra) - Fixed a major event scheduling issue. There cannot be a ref++ on an event that has already done ref++ once and without calling put_constraint() in between. (Stephane Eranian) - Use thread_cpumask for percore allocation. (Lin Ming) - Use MSR names in the extra reg lists. (Lin Ming) - Remove redundant "c = NULL" in intel_percore_constraints - Fix comment of perf_event_attr::config1 Intel Nehalem/Westmere have a special OFFCORE_RESPONSE event that can be used to monitor any offcore accesses from a core. This is a very useful event for various tunings, and it's also needed to implement the generic LLC-* events correctly. Unfortunately this event requires programming a mask in a separate register. And worse this separate register is per core, not per CPU thread. This patch: - Teaches perf_events that OFFCORE_RESPONSE needs extra parameters. The extra parameters are passed by user space in the perf_event_attr::config1 field. - Adds support to the Intel perf_event core to schedule per core resources. This adds fairly generic infrastructure that can be also used for other per core resources. The basic code has is patterned after the similar AMD northbridge constraints code. Thanks to Stephane Eranian who pointed out some problems in the original version and suggested improvements. Signed-off-by: Andi Kleen <ak@linux.intel.com> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1299119690-13991-2-git-send-email-ming.m.lin@intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-03-03 09:34:47 +07:00
if (hwc->extra_reg)
wrmsrl(hwc->extra_reg, hwc->extra_config);
wrmsrl(hwc->config_base, hwc->config | enable_mask);
}
static void x86_pmu_enable_all(int added)
{
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
int idx;
for (idx = 0; idx < x86_pmu.num_counters; idx++) {
struct hw_perf_event *hwc = &cpuc->events[idx]->hw;
if (!test_bit(idx, cpuc->active_mask))
continue;
__x86_pmu_enable_event(hwc, ARCH_PERFMON_EVENTSEL_ENABLE);
}
}
static struct pmu pmu;
static inline int is_x86_event(struct perf_event *event)
{
return event->pmu == &pmu;
}
static int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign)
{
struct event_constraint *c, *constraints[X86_PMC_IDX_MAX];
unsigned long used_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
int i, j, w, wmax, num = 0;
struct hw_perf_event *hwc;
bitmap_zero(used_mask, X86_PMC_IDX_MAX);
for (i = 0; i < n; i++) {
c = x86_pmu.get_event_constraints(cpuc, cpuc->event_list[i]);
constraints[i] = c;
}
/*
* fastpath, try to reuse previous register
*/
for (i = 0; i < n; i++) {
hwc = &cpuc->event_list[i]->hw;
c = constraints[i];
/* never assigned */
if (hwc->idx == -1)
break;
/* constraint still honored */
if (!test_bit(hwc->idx, c->idxmsk))
break;
/* not already used */
if (test_bit(hwc->idx, used_mask))
break;
__set_bit(hwc->idx, used_mask);
if (assign)
assign[i] = hwc->idx;
}
if (i == n)
goto done;
/*
* begin slow path
*/
bitmap_zero(used_mask, X86_PMC_IDX_MAX);
/*
* weight = number of possible counters
*
* 1 = most constrained, only works on one counter
* wmax = least constrained, works on any counter
*
* assign events to counters starting with most
* constrained events.
*/
wmax = x86_pmu.num_counters;
/*
* when fixed event counters are present,
* wmax is incremented by 1 to account
* for one more choice
*/
if (x86_pmu.num_counters_fixed)
wmax++;
for (w = 1, num = n; num && w <= wmax; w++) {
/* for each event */
for (i = 0; num && i < n; i++) {
c = constraints[i];
hwc = &cpuc->event_list[i]->hw;
if (c->weight != w)
continue;
for_each_set_bit(j, c->idxmsk, X86_PMC_IDX_MAX) {
if (!test_bit(j, used_mask))
break;
}
if (j == X86_PMC_IDX_MAX)
break;
__set_bit(j, used_mask);
if (assign)
assign[i] = j;
num--;
}
}
done:
/*
* scheduling failed or is just a simulation,
* free resources if necessary
*/
if (!assign || num) {
for (i = 0; i < n; i++) {
if (x86_pmu.put_event_constraints)
x86_pmu.put_event_constraints(cpuc, cpuc->event_list[i]);
}
}
return num ? -ENOSPC : 0;
}
/*
* dogrp: true if must collect siblings events (group)
* returns total number of events and error code
*/
static int collect_events(struct cpu_hw_events *cpuc, struct perf_event *leader, bool dogrp)
{
struct perf_event *event;
int n, max_count;
max_count = x86_pmu.num_counters + x86_pmu.num_counters_fixed;
/* current number of events already accepted */
n = cpuc->n_events;
if (is_x86_event(leader)) {
if (n >= max_count)
return -ENOSPC;
cpuc->event_list[n] = leader;
n++;
}
if (!dogrp)
return n;
list_for_each_entry(event, &leader->sibling_list, group_entry) {
if (!is_x86_event(event) ||
event->state <= PERF_EVENT_STATE_OFF)
continue;
if (n >= max_count)
return -ENOSPC;
cpuc->event_list[n] = event;
n++;
}
return n;
}
static inline void x86_assign_hw_event(struct perf_event *event,
struct cpu_hw_events *cpuc, int i)
{
struct hw_perf_event *hwc = &event->hw;
hwc->idx = cpuc->assign[i];
hwc->last_cpu = smp_processor_id();
hwc->last_tag = ++cpuc->tags[i];
if (hwc->idx == X86_PMC_IDX_FIXED_BTS) {
hwc->config_base = 0;
hwc->event_base = 0;
} else if (hwc->idx >= X86_PMC_IDX_FIXED) {
hwc->config_base = MSR_ARCH_PERFMON_FIXED_CTR_CTRL;
hwc->event_base = MSR_ARCH_PERFMON_FIXED_CTR0 + (hwc->idx - X86_PMC_IDX_FIXED);
} else {
hwc->config_base = x86_pmu_config_addr(hwc->idx);
hwc->event_base = x86_pmu_event_addr(hwc->idx);
}
}
static inline int match_prev_assignment(struct hw_perf_event *hwc,
struct cpu_hw_events *cpuc,
int i)
{
return hwc->idx == cpuc->assign[i] &&
hwc->last_cpu == smp_processor_id() &&
hwc->last_tag == cpuc->tags[i];
}
2010-06-16 19:37:10 +07:00
static void x86_pmu_start(struct perf_event *event, int flags);
static void x86_pmu_stop(struct perf_event *event, int flags);
2010-06-16 19:37:10 +07:00
static void x86_pmu_enable(struct pmu *pmu)
{
struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
struct perf_event *event;
struct hw_perf_event *hwc;
int i, added = cpuc->n_added;
if (!x86_pmu_initialized())
return;
if (cpuc->enabled)
return;
if (cpuc->n_added) {
int n_running = cpuc->n_events - cpuc->n_added;
/*
* apply assignment obtained either from
* hw_perf_group_sched_in() or x86_pmu_enable()
*
* step1: save events moving to new counters
* step2: reprogram moved events into new counters
*/
for (i = 0; i < n_running; i++) {
event = cpuc->event_list[i];
hwc = &event->hw;
/*
* we can avoid reprogramming counter if:
* - assigned same counter as last time
* - running on same CPU as last time
* - no other event has used the counter since
*/
if (hwc->idx == -1 ||
match_prev_assignment(hwc, cpuc, i))
continue;
2010-06-16 19:37:10 +07:00
/*
* Ensure we don't accidentally enable a stopped
* counter simply because we rescheduled.
*/
if (hwc->state & PERF_HES_STOPPED)
hwc->state |= PERF_HES_ARCH;
x86_pmu_stop(event, PERF_EF_UPDATE);
}
for (i = 0; i < cpuc->n_events; i++) {
event = cpuc->event_list[i];
hwc = &event->hw;
perf, x86: Fix hw_perf_enable() event assignment What happens is that we schedule badly like: <...>-1987 [019] 280.252808: x86_pmu_start: event-46/1300c0: idx: 0 <...>-1987 [019] 280.252811: x86_pmu_start: event-47/1300c0: idx: 1 <...>-1987 [019] 280.252812: x86_pmu_start: event-48/1300c0: idx: 2 <...>-1987 [019] 280.252813: x86_pmu_start: event-49/1300c0: idx: 3 <...>-1987 [019] 280.252814: x86_pmu_start: event-50/1300c0: idx: 32 <...>-1987 [019] 280.252825: x86_pmu_stop: event-46/1300c0: idx: 0 <...>-1987 [019] 280.252826: x86_pmu_stop: event-47/1300c0: idx: 1 <...>-1987 [019] 280.252827: x86_pmu_stop: event-48/1300c0: idx: 2 <...>-1987 [019] 280.252828: x86_pmu_stop: event-49/1300c0: idx: 3 <...>-1987 [019] 280.252829: x86_pmu_stop: event-50/1300c0: idx: 32 <...>-1987 [019] 280.252834: x86_pmu_start: event-47/1300c0: idx: 1 <...>-1987 [019] 280.252834: x86_pmu_start: event-48/1300c0: idx: 2 <...>-1987 [019] 280.252835: x86_pmu_start: event-49/1300c0: idx: 3 <...>-1987 [019] 280.252836: x86_pmu_start: event-50/1300c0: idx: 32 <...>-1987 [019] 280.252837: x86_pmu_start: event-51/1300c0: idx: 32 *FAIL* This happens because we only iterate the n_running events in the first pass, and reset their index to -1 if they don't match to force a re-assignment. Now, in our RR example, n_running == 0 because we fully unscheduled, so event-50 will retain its idx==32, even though in scheduling it will have gotten idx=0, and we don't trigger the re-assign path. The easiest way to fix this is the below patch, which simply validates the full assignment in the second pass. Reported-by: Stephane Eranian <eranian@google.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1268311069.5037.31.camel@laptop> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 19:40:30 +07:00
if (!match_prev_assignment(hwc, cpuc, i))
x86_assign_hw_event(event, cpuc, i);
perf, x86: Fix hw_perf_enable() event assignment What happens is that we schedule badly like: <...>-1987 [019] 280.252808: x86_pmu_start: event-46/1300c0: idx: 0 <...>-1987 [019] 280.252811: x86_pmu_start: event-47/1300c0: idx: 1 <...>-1987 [019] 280.252812: x86_pmu_start: event-48/1300c0: idx: 2 <...>-1987 [019] 280.252813: x86_pmu_start: event-49/1300c0: idx: 3 <...>-1987 [019] 280.252814: x86_pmu_start: event-50/1300c0: idx: 32 <...>-1987 [019] 280.252825: x86_pmu_stop: event-46/1300c0: idx: 0 <...>-1987 [019] 280.252826: x86_pmu_stop: event-47/1300c0: idx: 1 <...>-1987 [019] 280.252827: x86_pmu_stop: event-48/1300c0: idx: 2 <...>-1987 [019] 280.252828: x86_pmu_stop: event-49/1300c0: idx: 3 <...>-1987 [019] 280.252829: x86_pmu_stop: event-50/1300c0: idx: 32 <...>-1987 [019] 280.252834: x86_pmu_start: event-47/1300c0: idx: 1 <...>-1987 [019] 280.252834: x86_pmu_start: event-48/1300c0: idx: 2 <...>-1987 [019] 280.252835: x86_pmu_start: event-49/1300c0: idx: 3 <...>-1987 [019] 280.252836: x86_pmu_start: event-50/1300c0: idx: 32 <...>-1987 [019] 280.252837: x86_pmu_start: event-51/1300c0: idx: 32 *FAIL* This happens because we only iterate the n_running events in the first pass, and reset their index to -1 if they don't match to force a re-assignment. Now, in our RR example, n_running == 0 because we fully unscheduled, so event-50 will retain its idx==32, even though in scheduling it will have gotten idx=0, and we don't trigger the re-assign path. The easiest way to fix this is the below patch, which simply validates the full assignment in the second pass. Reported-by: Stephane Eranian <eranian@google.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1268311069.5037.31.camel@laptop> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 19:40:30 +07:00
else if (i < n_running)
continue;
2010-06-16 19:37:10 +07:00
if (hwc->state & PERF_HES_ARCH)
continue;
x86_pmu_start(event, PERF_EF_RELOAD);
}
cpuc->n_added = 0;
perf_events_lapic_init();
}
cpuc->enabled = 1;
barrier();
x86_pmu.enable_all(added);
}
static inline void x86_pmu_disable_event(struct perf_event *event)
{
struct hw_perf_event *hwc = &event->hw;
wrmsrl(hwc->config_base, hwc->config);
}
static DEFINE_PER_CPU(u64 [X86_PMC_IDX_MAX], pmc_prev_left);
/*
* Set the next IRQ period, based on the hwc->period_left value.
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
* To be called with the event disabled in hw:
*/
static int
x86_perf_event_set_period(struct perf_event *event)
{
struct hw_perf_event *hwc = &event->hw;
s64 left = local64_read(&hwc->period_left);
s64 period = hwc->sample_period;
int ret = 0, idx = hwc->idx;
if (idx == X86_PMC_IDX_FIXED_BTS)
return 0;
/*
* If we are way outside a reasonable range then just skip forward:
*/
if (unlikely(left <= -period)) {
left = period;
local64_set(&hwc->period_left, left);
hwc->last_period = period;
ret = 1;
}
if (unlikely(left <= 0)) {
left += period;
local64_set(&hwc->period_left, left);
hwc->last_period = period;
ret = 1;
}
/*
* Quirk: certain CPUs dont like it if just 1 hw_event is left:
*/
if (unlikely(left < 2))
left = 2;
if (left > x86_pmu.max_period)
left = x86_pmu.max_period;
per_cpu(pmc_prev_left[idx], smp_processor_id()) = left;
/*
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
* The hw event starts counting from this event offset,
* mark it to be able to extra future deltas:
*/
local64_set(&hwc->prev_count, (u64)-left);
wrmsrl(hwc->event_base, (u64)(-left) & x86_pmu.cntval_mask);
/*
* Due to erratum on certan cpu we need
* a second write to be sure the register
* is updated properly
*/
if (x86_pmu.perfctr_second_write) {
wrmsrl(hwc->event_base,
(u64)(-left) & x86_pmu.cntval_mask);
}
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
perf_event_update_userpage(event);
return ret;
}
static void x86_pmu_enable_event(struct perf_event *event)
{
if (__this_cpu_read(cpu_hw_events.enabled))
__x86_pmu_enable_event(&event->hw,
ARCH_PERFMON_EVENTSEL_ENABLE);
}
/*
2010-06-16 19:37:10 +07:00
* Add a single event to the PMU.
*
* The event is added to the group of enabled events
* but only if it can be scehduled with existing events.
*/
2010-06-16 19:37:10 +07:00
static int x86_pmu_add(struct perf_event *event, int flags)
{
struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
struct hw_perf_event *hwc;
int assign[X86_PMC_IDX_MAX];
int n, n0, ret;
hwc = &event->hw;
perf_pmu_disable(event->pmu);
n0 = cpuc->n_events;
ret = n = collect_events(cpuc, event, false);
if (ret < 0)
goto out;
2010-06-16 19:37:10 +07:00
hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
if (!(flags & PERF_EF_START))
hwc->state |= PERF_HES_ARCH;
/*
* If group events scheduling transaction was started,
* skip the schedulability test here, it will be performed
2010-06-16 19:37:10 +07:00
* at commit time (->commit_txn) as a whole
*/
if (cpuc->group_flag & PERF_EVENT_TXN)
goto done_collect;
perf, x86: Implement initial P4 PMU driver The netburst PMU is way different from the "architectural perfomance monitoring" specification that current CPUs use. P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle perfomance monitoring events. A few implementational details: 1) We need a separate x86_pmu::hw_config helper in struct x86_pmu since register bit-fields are quite different from P6, Core and later cpu series. 2) For the same reason is a x86_pmu::schedule_events helper introduced. 3) hw_perf_event::config consists of packed ESCR+CCCR values. It's allowed since in reality both registers only use a half of their size. Of course before making a real write into a particular MSR we need to unpack the value and extend it to a proper size. 4) The tuple of packed ESCR+CCCR in hw_perf_event::config doesn't describe the memory address of ESCR MSR register so that we need to keep a mapping between these tuples used and available ESCR (various P4 events may use same ESCRs but not simultaneously), for this sake every active event has a per-cpu map of hw_perf_event::idx <--> ESCR addresses. 5) Since hw_perf_event::idx is an offset to counter/control register we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel strips it down to 8 registers and event armed may never be turned off (ie the bit in active_mask is set but the loop never reaches this index to check), thanks to Peter Zijlstra Restrictions: - No cascaded counters support (do we ever need them?) - No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS doesn't work for now) - There are events with same counters which can't work simultaneously (need to use intersected ones due to broken counter 1) - No PERF_COUNT_HW_CACHE_ events yet Todo: - Implement dependent events - Need proper hashing for event opcodes (no linear search, good for debugging stage but not in real loads) - Some events counted during a clock cycle -- need to set threshold for them and count every clock cycle just to get summary statistics (ie to behave the same way as other PMUs do) - Need to swicth to use event_constraints - To support RAW events we need to encode a global list of P4 events into p4_templates - Cache events need to be added Event support status matrix: Event status ----------------------------- cycles works cache-references works cache-misses works branch-misses works bus-cycles partially (does not work on 64bit cpu with HT enabled) instruction doesnt work (needs dependent event [mop tagging]) branches doesnt work Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Robert Richter <robert.richter@amd.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20100311165439.GB5129@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 23:54:39 +07:00
ret = x86_pmu.schedule_events(cpuc, n, assign);
if (ret)
goto out;
/*
* copy new assignment, now we know it is possible
* will be used by hw_perf_enable()
*/
memcpy(cpuc->assign, assign, n*sizeof(int));
done_collect:
cpuc->n_events = n;
cpuc->n_added += n - n0;
perf_events: Fix event scheduling issues introduced by transactional API The transactional API patch between the generic and model-specific code introduced several important bugs with event scheduling, at least on X86. If you had pinned events, e.g., watchdog, and were over-committing the PMU, you would get bogus counts. The bug was showing up on Intel CPU because events would move around more often that on AMD. But the problem also existed on AMD, though harder to expose. The issues were: - group_sched_in() was missing a cancel_txn() in the error path - cpuc->n_added was not properly maintained, leading to missing actions in hw_perf_enable(), i.e., n_running being 0. You cannot update n_added until you know the transaction has succeeded. In case of failed transaction n_added was not adjusted back. - in case of failed transactions, event_sched_out() was called and eventually invoked x86_disable_event() to touch the HW reg. But with transactions, on X86, event_sched_in() does not touch HW registers, it simply collects events into a list. Thus, you could end up calling x86_disable_event() on a counter which did not correspond to the current event when idx != -1. The patch modifies the generic and X86 code to avoid all those problems. First, we keep track of the number of events added last. In case the transaction fails, we substract them from n_added. This approach is necessary (as opposed to delaying updates to n_added) because not all event updates use the transaction API, e.g., single events. Second, we encapsulate the event_sched_in() and event_sched_out() in group_sched_in() inside the transaction. That makes the operations symmetrical and you can also detect that you are inside a transaction and skip the HW reg access by checking cpuc->group_flag. With this patch, you can now overcommit the PMU even with pinned system-wide events present and still get valid counts. Signed-off-by: Stephane Eranian <eranian@google.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1274796225.5882.1389.camel@twins> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-05-25 21:23:10 +07:00
cpuc->n_txn += n - n0;
ret = 0;
out:
perf_pmu_enable(event->pmu);
return ret;
}
2010-06-16 19:37:10 +07:00
static void x86_pmu_start(struct perf_event *event, int flags)
{
struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
int idx = event->hw.idx;
2010-06-16 19:37:10 +07:00
if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED)))
return;
if (WARN_ON_ONCE(idx == -1))
return;
if (flags & PERF_EF_RELOAD) {
WARN_ON_ONCE(!(event->hw.state & PERF_HES_UPTODATE));
x86_perf_event_set_period(event);
}
event->hw.state = 0;
cpuc->events[idx] = event;
__set_bit(idx, cpuc->active_mask);
__set_bit(idx, cpuc->running);
x86_pmu.enable(event);
perf_event_update_userpage(event);
}
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
void perf_event_print_debug(void)
{
u64 ctrl, status, overflow, pmc_ctrl, pmc_count, prev_left, fixed;
u64 pebs;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
struct cpu_hw_events *cpuc;
unsigned long flags;
int cpu, idx;
if (!x86_pmu.num_counters)
return;
local_irq_save(flags);
cpu = smp_processor_id();
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
cpuc = &per_cpu(cpu_hw_events, cpu);
if (x86_pmu.version >= 2) {
rdmsrl(MSR_CORE_PERF_GLOBAL_CTRL, ctrl);
rdmsrl(MSR_CORE_PERF_GLOBAL_STATUS, status);
rdmsrl(MSR_CORE_PERF_GLOBAL_OVF_CTRL, overflow);
rdmsrl(MSR_ARCH_PERFMON_FIXED_CTR_CTRL, fixed);
rdmsrl(MSR_IA32_PEBS_ENABLE, pebs);
pr_info("\n");
pr_info("CPU#%d: ctrl: %016llx\n", cpu, ctrl);
pr_info("CPU#%d: status: %016llx\n", cpu, status);
pr_info("CPU#%d: overflow: %016llx\n", cpu, overflow);
pr_info("CPU#%d: fixed: %016llx\n", cpu, fixed);
pr_info("CPU#%d: pebs: %016llx\n", cpu, pebs);
}
pr_info("CPU#%d: active: %016llx\n", cpu, *(u64 *)cpuc->active_mask);
for (idx = 0; idx < x86_pmu.num_counters; idx++) {
rdmsrl(x86_pmu_config_addr(idx), pmc_ctrl);
rdmsrl(x86_pmu_event_addr(idx), pmc_count);
prev_left = per_cpu(pmc_prev_left[idx], cpu);
pr_info("CPU#%d: gen-PMC%d ctrl: %016llx\n",
cpu, idx, pmc_ctrl);
pr_info("CPU#%d: gen-PMC%d count: %016llx\n",
cpu, idx, pmc_count);
pr_info("CPU#%d: gen-PMC%d left: %016llx\n",
cpu, idx, prev_left);
}
for (idx = 0; idx < x86_pmu.num_counters_fixed; idx++) {
rdmsrl(MSR_ARCH_PERFMON_FIXED_CTR0 + idx, pmc_count);
pr_info("CPU#%d: fixed-PMC%d count: %016llx\n",
cpu, idx, pmc_count);
}
local_irq_restore(flags);
}
2010-06-16 19:37:10 +07:00
static void x86_pmu_stop(struct perf_event *event, int flags)
{
struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
struct hw_perf_event *hwc = &event->hw;
2010-06-16 19:37:10 +07:00
if (__test_and_clear_bit(hwc->idx, cpuc->active_mask)) {
x86_pmu.disable(event);
cpuc->events[hwc->idx] = NULL;
WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED);
hwc->state |= PERF_HES_STOPPED;
}
2010-06-16 19:37:10 +07:00
if ((flags & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) {
/*
* Drain the remaining delta count out of a event
* that we are disabling:
*/
x86_perf_event_update(event);
hwc->state |= PERF_HES_UPTODATE;
}
}
2010-06-16 19:37:10 +07:00
static void x86_pmu_del(struct perf_event *event, int flags)
{
struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
int i;
perf_events: Fix event scheduling issues introduced by transactional API The transactional API patch between the generic and model-specific code introduced several important bugs with event scheduling, at least on X86. If you had pinned events, e.g., watchdog, and were over-committing the PMU, you would get bogus counts. The bug was showing up on Intel CPU because events would move around more often that on AMD. But the problem also existed on AMD, though harder to expose. The issues were: - group_sched_in() was missing a cancel_txn() in the error path - cpuc->n_added was not properly maintained, leading to missing actions in hw_perf_enable(), i.e., n_running being 0. You cannot update n_added until you know the transaction has succeeded. In case of failed transaction n_added was not adjusted back. - in case of failed transactions, event_sched_out() was called and eventually invoked x86_disable_event() to touch the HW reg. But with transactions, on X86, event_sched_in() does not touch HW registers, it simply collects events into a list. Thus, you could end up calling x86_disable_event() on a counter which did not correspond to the current event when idx != -1. The patch modifies the generic and X86 code to avoid all those problems. First, we keep track of the number of events added last. In case the transaction fails, we substract them from n_added. This approach is necessary (as opposed to delaying updates to n_added) because not all event updates use the transaction API, e.g., single events. Second, we encapsulate the event_sched_in() and event_sched_out() in group_sched_in() inside the transaction. That makes the operations symmetrical and you can also detect that you are inside a transaction and skip the HW reg access by checking cpuc->group_flag. With this patch, you can now overcommit the PMU even with pinned system-wide events present and still get valid counts. Signed-off-by: Stephane Eranian <eranian@google.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1274796225.5882.1389.camel@twins> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-05-25 21:23:10 +07:00
/*
* If we're called during a txn, we don't need to do anything.
* The events never got scheduled and ->cancel_txn will truncate
* the event_list.
*/
if (cpuc->group_flag & PERF_EVENT_TXN)
perf_events: Fix event scheduling issues introduced by transactional API The transactional API patch between the generic and model-specific code introduced several important bugs with event scheduling, at least on X86. If you had pinned events, e.g., watchdog, and were over-committing the PMU, you would get bogus counts. The bug was showing up on Intel CPU because events would move around more often that on AMD. But the problem also existed on AMD, though harder to expose. The issues were: - group_sched_in() was missing a cancel_txn() in the error path - cpuc->n_added was not properly maintained, leading to missing actions in hw_perf_enable(), i.e., n_running being 0. You cannot update n_added until you know the transaction has succeeded. In case of failed transaction n_added was not adjusted back. - in case of failed transactions, event_sched_out() was called and eventually invoked x86_disable_event() to touch the HW reg. But with transactions, on X86, event_sched_in() does not touch HW registers, it simply collects events into a list. Thus, you could end up calling x86_disable_event() on a counter which did not correspond to the current event when idx != -1. The patch modifies the generic and X86 code to avoid all those problems. First, we keep track of the number of events added last. In case the transaction fails, we substract them from n_added. This approach is necessary (as opposed to delaying updates to n_added) because not all event updates use the transaction API, e.g., single events. Second, we encapsulate the event_sched_in() and event_sched_out() in group_sched_in() inside the transaction. That makes the operations symmetrical and you can also detect that you are inside a transaction and skip the HW reg access by checking cpuc->group_flag. With this patch, you can now overcommit the PMU even with pinned system-wide events present and still get valid counts. Signed-off-by: Stephane Eranian <eranian@google.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1274796225.5882.1389.camel@twins> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-05-25 21:23:10 +07:00
return;
2010-06-16 19:37:10 +07:00
x86_pmu_stop(event, PERF_EF_UPDATE);
for (i = 0; i < cpuc->n_events; i++) {
if (event == cpuc->event_list[i]) {
if (x86_pmu.put_event_constraints)
x86_pmu.put_event_constraints(cpuc, event);
while (++i < cpuc->n_events)
cpuc->event_list[i-1] = cpuc->event_list[i];
--cpuc->n_events;
break;
}
}
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
perf_event_update_userpage(event);
}
static int x86_pmu_handle_irq(struct pt_regs *regs)
{
struct perf_sample_data data;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
struct cpu_hw_events *cpuc;
struct perf_event *event;
int idx, handled = 0;
u64 val;
perf_sample_data_init(&data, 0);
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
cpuc = &__get_cpu_var(cpu_hw_events);
for (idx = 0; idx < x86_pmu.num_counters; idx++) {
if (!test_bit(idx, cpuc->active_mask)) {
/*
* Though we deactivated the counter some cpus
* might still deliver spurious interrupts still
* in flight. Catch them:
*/
if (__test_and_clear_bit(idx, cpuc->running))
handled++;
continue;
}
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
event = cpuc->events[idx];
val = x86_perf_event_update(event);
if (val & (1ULL << (x86_pmu.cntval_bits - 1)))
continue;
/*
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
* event overflow
*/
perf, x86: Try to handle unknown nmis with an enabled PMU When the PMU is enabled it is valid to have unhandled nmis, two events could trigger 'simultaneously' raising two back-to-back NMIs. If the first NMI handles both, the latter will be empty and daze the CPU. The solution to avoid an 'unknown nmi' massage in this case was simply to stop the nmi handler chain when the PMU is enabled by stating the nmi was handled. This has the drawback that a) we can not detect unknown nmis anymore, and b) subsequent nmi handlers are not called. This patch addresses this. Now, we check this unknown NMI if it could be a PMU back-to-back NMI. Otherwise we pass it and let the kernel handle the unknown nmi. This is a debug log: cpu #6, nmi #32333, skip_nmi #32330, handled = 1, time = 1934364430 cpu #6, nmi #32334, skip_nmi #32330, handled = 1, time = 1934704616 cpu #6, nmi #32335, skip_nmi #32336, handled = 2, time = 1936032320 cpu #6, nmi #32336, skip_nmi #32336, handled = 0, time = 1936034139 cpu #6, nmi #32337, skip_nmi #32336, handled = 1, time = 1936120100 cpu #6, nmi #32338, skip_nmi #32336, handled = 1, time = 1936404607 cpu #6, nmi #32339, skip_nmi #32336, handled = 1, time = 1937983416 cpu #6, nmi #32340, skip_nmi #32341, handled = 2, time = 1938201032 cpu #6, nmi #32341, skip_nmi #32341, handled = 0, time = 1938202830 cpu #6, nmi #32342, skip_nmi #32341, handled = 1, time = 1938443743 cpu #6, nmi #32343, skip_nmi #32341, handled = 1, time = 1939956552 cpu #6, nmi #32344, skip_nmi #32341, handled = 1, time = 1940073224 cpu #6, nmi #32345, skip_nmi #32341, handled = 1, time = 1940485677 cpu #6, nmi #32346, skip_nmi #32347, handled = 2, time = 1941947772 cpu #6, nmi #32347, skip_nmi #32347, handled = 1, time = 1941949818 cpu #6, nmi #32348, skip_nmi #32347, handled = 0, time = 1941951591 Uhhuh. NMI received for unknown reason 00 on CPU 6. Do you have a strange power saving mode enabled? Dazed and confused, but trying to continue Deltas: nmi #32334 340186 nmi #32335 1327704 nmi #32336 1819 <<<< back-to-back nmi [1] nmi #32337 85961 nmi #32338 284507 nmi #32339 1578809 nmi #32340 217616 nmi #32341 1798 <<<< back-to-back nmi [2] nmi #32342 240913 nmi #32343 1512809 nmi #32344 116672 nmi #32345 412453 nmi #32346 1462095 <<<< 1st nmi (standard) handling 2 counters nmi #32347 2046 <<<< 2nd nmi (back-to-back) handling one counter nmi #32348 1773 <<<< 3rd nmi (back-to-back) handling no counter! [3] For back-to-back nmi detection there are the following rules: The PMU nmi handler was handling more than one counter and no counter was handled in the subsequent nmi (see [1] and [2] above). There is another case if there are two subsequent back-to-back nmis [3]. The 2nd is detected as back-to-back because the first handled more than one counter. If the second handles one counter and the 3rd handles nothing, we drop the 3rd nmi because it could be a back-to-back nmi. Signed-off-by: Robert Richter <robert.richter@amd.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> [ renamed nmi variable to pmu_nmi to avoid clash with .nmi in entry.S ] Signed-off-by: Don Zickus <dzickus@redhat.com> Cc: peterz@infradead.org Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-3-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-09-03 02:07:48 +07:00
handled++;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
data.period = event->hw.last_period;
if (!x86_perf_event_set_period(event))
continue;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
if (perf_event_overflow(event, 1, &data, regs))
2010-06-16 19:37:10 +07:00
x86_pmu_stop(event, 0);
}
if (handled)
inc_irq_stat(apic_perf_irqs);
return handled;
}
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
void perf_events_lapic_init(void)
{
if (!x86_pmu.apic || !x86_pmu_initialized())
return;
/*
* Always use NMI for PMU
*/
apic_write(APIC_LVTPC, APIC_DM_NMI);
}
perf, x86: Try to handle unknown nmis with an enabled PMU When the PMU is enabled it is valid to have unhandled nmis, two events could trigger 'simultaneously' raising two back-to-back NMIs. If the first NMI handles both, the latter will be empty and daze the CPU. The solution to avoid an 'unknown nmi' massage in this case was simply to stop the nmi handler chain when the PMU is enabled by stating the nmi was handled. This has the drawback that a) we can not detect unknown nmis anymore, and b) subsequent nmi handlers are not called. This patch addresses this. Now, we check this unknown NMI if it could be a PMU back-to-back NMI. Otherwise we pass it and let the kernel handle the unknown nmi. This is a debug log: cpu #6, nmi #32333, skip_nmi #32330, handled = 1, time = 1934364430 cpu #6, nmi #32334, skip_nmi #32330, handled = 1, time = 1934704616 cpu #6, nmi #32335, skip_nmi #32336, handled = 2, time = 1936032320 cpu #6, nmi #32336, skip_nmi #32336, handled = 0, time = 1936034139 cpu #6, nmi #32337, skip_nmi #32336, handled = 1, time = 1936120100 cpu #6, nmi #32338, skip_nmi #32336, handled = 1, time = 1936404607 cpu #6, nmi #32339, skip_nmi #32336, handled = 1, time = 1937983416 cpu #6, nmi #32340, skip_nmi #32341, handled = 2, time = 1938201032 cpu #6, nmi #32341, skip_nmi #32341, handled = 0, time = 1938202830 cpu #6, nmi #32342, skip_nmi #32341, handled = 1, time = 1938443743 cpu #6, nmi #32343, skip_nmi #32341, handled = 1, time = 1939956552 cpu #6, nmi #32344, skip_nmi #32341, handled = 1, time = 1940073224 cpu #6, nmi #32345, skip_nmi #32341, handled = 1, time = 1940485677 cpu #6, nmi #32346, skip_nmi #32347, handled = 2, time = 1941947772 cpu #6, nmi #32347, skip_nmi #32347, handled = 1, time = 1941949818 cpu #6, nmi #32348, skip_nmi #32347, handled = 0, time = 1941951591 Uhhuh. NMI received for unknown reason 00 on CPU 6. Do you have a strange power saving mode enabled? Dazed and confused, but trying to continue Deltas: nmi #32334 340186 nmi #32335 1327704 nmi #32336 1819 <<<< back-to-back nmi [1] nmi #32337 85961 nmi #32338 284507 nmi #32339 1578809 nmi #32340 217616 nmi #32341 1798 <<<< back-to-back nmi [2] nmi #32342 240913 nmi #32343 1512809 nmi #32344 116672 nmi #32345 412453 nmi #32346 1462095 <<<< 1st nmi (standard) handling 2 counters nmi #32347 2046 <<<< 2nd nmi (back-to-back) handling one counter nmi #32348 1773 <<<< 3rd nmi (back-to-back) handling no counter! [3] For back-to-back nmi detection there are the following rules: The PMU nmi handler was handling more than one counter and no counter was handled in the subsequent nmi (see [1] and [2] above). There is another case if there are two subsequent back-to-back nmis [3]. The 2nd is detected as back-to-back because the first handled more than one counter. If the second handles one counter and the 3rd handles nothing, we drop the 3rd nmi because it could be a back-to-back nmi. Signed-off-by: Robert Richter <robert.richter@amd.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> [ renamed nmi variable to pmu_nmi to avoid clash with .nmi in entry.S ] Signed-off-by: Don Zickus <dzickus@redhat.com> Cc: peterz@infradead.org Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-3-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-09-03 02:07:48 +07:00
struct pmu_nmi_state {
unsigned int marked;
int handled;
};
static DEFINE_PER_CPU(struct pmu_nmi_state, pmu_nmi);
static int __kprobes
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
perf_event_nmi_handler(struct notifier_block *self,
unsigned long cmd, void *__args)
{
struct die_args *args = __args;
perf, x86: Try to handle unknown nmis with an enabled PMU When the PMU is enabled it is valid to have unhandled nmis, two events could trigger 'simultaneously' raising two back-to-back NMIs. If the first NMI handles both, the latter will be empty and daze the CPU. The solution to avoid an 'unknown nmi' massage in this case was simply to stop the nmi handler chain when the PMU is enabled by stating the nmi was handled. This has the drawback that a) we can not detect unknown nmis anymore, and b) subsequent nmi handlers are not called. This patch addresses this. Now, we check this unknown NMI if it could be a PMU back-to-back NMI. Otherwise we pass it and let the kernel handle the unknown nmi. This is a debug log: cpu #6, nmi #32333, skip_nmi #32330, handled = 1, time = 1934364430 cpu #6, nmi #32334, skip_nmi #32330, handled = 1, time = 1934704616 cpu #6, nmi #32335, skip_nmi #32336, handled = 2, time = 1936032320 cpu #6, nmi #32336, skip_nmi #32336, handled = 0, time = 1936034139 cpu #6, nmi #32337, skip_nmi #32336, handled = 1, time = 1936120100 cpu #6, nmi #32338, skip_nmi #32336, handled = 1, time = 1936404607 cpu #6, nmi #32339, skip_nmi #32336, handled = 1, time = 1937983416 cpu #6, nmi #32340, skip_nmi #32341, handled = 2, time = 1938201032 cpu #6, nmi #32341, skip_nmi #32341, handled = 0, time = 1938202830 cpu #6, nmi #32342, skip_nmi #32341, handled = 1, time = 1938443743 cpu #6, nmi #32343, skip_nmi #32341, handled = 1, time = 1939956552 cpu #6, nmi #32344, skip_nmi #32341, handled = 1, time = 1940073224 cpu #6, nmi #32345, skip_nmi #32341, handled = 1, time = 1940485677 cpu #6, nmi #32346, skip_nmi #32347, handled = 2, time = 1941947772 cpu #6, nmi #32347, skip_nmi #32347, handled = 1, time = 1941949818 cpu #6, nmi #32348, skip_nmi #32347, handled = 0, time = 1941951591 Uhhuh. NMI received for unknown reason 00 on CPU 6. Do you have a strange power saving mode enabled? Dazed and confused, but trying to continue Deltas: nmi #32334 340186 nmi #32335 1327704 nmi #32336 1819 <<<< back-to-back nmi [1] nmi #32337 85961 nmi #32338 284507 nmi #32339 1578809 nmi #32340 217616 nmi #32341 1798 <<<< back-to-back nmi [2] nmi #32342 240913 nmi #32343 1512809 nmi #32344 116672 nmi #32345 412453 nmi #32346 1462095 <<<< 1st nmi (standard) handling 2 counters nmi #32347 2046 <<<< 2nd nmi (back-to-back) handling one counter nmi #32348 1773 <<<< 3rd nmi (back-to-back) handling no counter! [3] For back-to-back nmi detection there are the following rules: The PMU nmi handler was handling more than one counter and no counter was handled in the subsequent nmi (see [1] and [2] above). There is another case if there are two subsequent back-to-back nmis [3]. The 2nd is detected as back-to-back because the first handled more than one counter. If the second handles one counter and the 3rd handles nothing, we drop the 3rd nmi because it could be a back-to-back nmi. Signed-off-by: Robert Richter <robert.richter@amd.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> [ renamed nmi variable to pmu_nmi to avoid clash with .nmi in entry.S ] Signed-off-by: Don Zickus <dzickus@redhat.com> Cc: peterz@infradead.org Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-3-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-09-03 02:07:48 +07:00
unsigned int this_nmi;
int handled;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
if (!atomic_read(&active_events))
return NOTIFY_DONE;
switch (cmd) {
case DIE_NMI:
break;
perf, x86: Try to handle unknown nmis with an enabled PMU When the PMU is enabled it is valid to have unhandled nmis, two events could trigger 'simultaneously' raising two back-to-back NMIs. If the first NMI handles both, the latter will be empty and daze the CPU. The solution to avoid an 'unknown nmi' massage in this case was simply to stop the nmi handler chain when the PMU is enabled by stating the nmi was handled. This has the drawback that a) we can not detect unknown nmis anymore, and b) subsequent nmi handlers are not called. This patch addresses this. Now, we check this unknown NMI if it could be a PMU back-to-back NMI. Otherwise we pass it and let the kernel handle the unknown nmi. This is a debug log: cpu #6, nmi #32333, skip_nmi #32330, handled = 1, time = 1934364430 cpu #6, nmi #32334, skip_nmi #32330, handled = 1, time = 1934704616 cpu #6, nmi #32335, skip_nmi #32336, handled = 2, time = 1936032320 cpu #6, nmi #32336, skip_nmi #32336, handled = 0, time = 1936034139 cpu #6, nmi #32337, skip_nmi #32336, handled = 1, time = 1936120100 cpu #6, nmi #32338, skip_nmi #32336, handled = 1, time = 1936404607 cpu #6, nmi #32339, skip_nmi #32336, handled = 1, time = 1937983416 cpu #6, nmi #32340, skip_nmi #32341, handled = 2, time = 1938201032 cpu #6, nmi #32341, skip_nmi #32341, handled = 0, time = 1938202830 cpu #6, nmi #32342, skip_nmi #32341, handled = 1, time = 1938443743 cpu #6, nmi #32343, skip_nmi #32341, handled = 1, time = 1939956552 cpu #6, nmi #32344, skip_nmi #32341, handled = 1, time = 1940073224 cpu #6, nmi #32345, skip_nmi #32341, handled = 1, time = 1940485677 cpu #6, nmi #32346, skip_nmi #32347, handled = 2, time = 1941947772 cpu #6, nmi #32347, skip_nmi #32347, handled = 1, time = 1941949818 cpu #6, nmi #32348, skip_nmi #32347, handled = 0, time = 1941951591 Uhhuh. NMI received for unknown reason 00 on CPU 6. Do you have a strange power saving mode enabled? Dazed and confused, but trying to continue Deltas: nmi #32334 340186 nmi #32335 1327704 nmi #32336 1819 <<<< back-to-back nmi [1] nmi #32337 85961 nmi #32338 284507 nmi #32339 1578809 nmi #32340 217616 nmi #32341 1798 <<<< back-to-back nmi [2] nmi #32342 240913 nmi #32343 1512809 nmi #32344 116672 nmi #32345 412453 nmi #32346 1462095 <<<< 1st nmi (standard) handling 2 counters nmi #32347 2046 <<<< 2nd nmi (back-to-back) handling one counter nmi #32348 1773 <<<< 3rd nmi (back-to-back) handling no counter! [3] For back-to-back nmi detection there are the following rules: The PMU nmi handler was handling more than one counter and no counter was handled in the subsequent nmi (see [1] and [2] above). There is another case if there are two subsequent back-to-back nmis [3]. The 2nd is detected as back-to-back because the first handled more than one counter. If the second handles one counter and the 3rd handles nothing, we drop the 3rd nmi because it could be a back-to-back nmi. Signed-off-by: Robert Richter <robert.richter@amd.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> [ renamed nmi variable to pmu_nmi to avoid clash with .nmi in entry.S ] Signed-off-by: Don Zickus <dzickus@redhat.com> Cc: peterz@infradead.org Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-3-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-09-03 02:07:48 +07:00
case DIE_NMIUNKNOWN:
this_nmi = percpu_read(irq_stat.__nmi_count);
if (this_nmi != __this_cpu_read(pmu_nmi.marked))
perf, x86: Try to handle unknown nmis with an enabled PMU When the PMU is enabled it is valid to have unhandled nmis, two events could trigger 'simultaneously' raising two back-to-back NMIs. If the first NMI handles both, the latter will be empty and daze the CPU. The solution to avoid an 'unknown nmi' massage in this case was simply to stop the nmi handler chain when the PMU is enabled by stating the nmi was handled. This has the drawback that a) we can not detect unknown nmis anymore, and b) subsequent nmi handlers are not called. This patch addresses this. Now, we check this unknown NMI if it could be a PMU back-to-back NMI. Otherwise we pass it and let the kernel handle the unknown nmi. This is a debug log: cpu #6, nmi #32333, skip_nmi #32330, handled = 1, time = 1934364430 cpu #6, nmi #32334, skip_nmi #32330, handled = 1, time = 1934704616 cpu #6, nmi #32335, skip_nmi #32336, handled = 2, time = 1936032320 cpu #6, nmi #32336, skip_nmi #32336, handled = 0, time = 1936034139 cpu #6, nmi #32337, skip_nmi #32336, handled = 1, time = 1936120100 cpu #6, nmi #32338, skip_nmi #32336, handled = 1, time = 1936404607 cpu #6, nmi #32339, skip_nmi #32336, handled = 1, time = 1937983416 cpu #6, nmi #32340, skip_nmi #32341, handled = 2, time = 1938201032 cpu #6, nmi #32341, skip_nmi #32341, handled = 0, time = 1938202830 cpu #6, nmi #32342, skip_nmi #32341, handled = 1, time = 1938443743 cpu #6, nmi #32343, skip_nmi #32341, handled = 1, time = 1939956552 cpu #6, nmi #32344, skip_nmi #32341, handled = 1, time = 1940073224 cpu #6, nmi #32345, skip_nmi #32341, handled = 1, time = 1940485677 cpu #6, nmi #32346, skip_nmi #32347, handled = 2, time = 1941947772 cpu #6, nmi #32347, skip_nmi #32347, handled = 1, time = 1941949818 cpu #6, nmi #32348, skip_nmi #32347, handled = 0, time = 1941951591 Uhhuh. NMI received for unknown reason 00 on CPU 6. Do you have a strange power saving mode enabled? Dazed and confused, but trying to continue Deltas: nmi #32334 340186 nmi #32335 1327704 nmi #32336 1819 <<<< back-to-back nmi [1] nmi #32337 85961 nmi #32338 284507 nmi #32339 1578809 nmi #32340 217616 nmi #32341 1798 <<<< back-to-back nmi [2] nmi #32342 240913 nmi #32343 1512809 nmi #32344 116672 nmi #32345 412453 nmi #32346 1462095 <<<< 1st nmi (standard) handling 2 counters nmi #32347 2046 <<<< 2nd nmi (back-to-back) handling one counter nmi #32348 1773 <<<< 3rd nmi (back-to-back) handling no counter! [3] For back-to-back nmi detection there are the following rules: The PMU nmi handler was handling more than one counter and no counter was handled in the subsequent nmi (see [1] and [2] above). There is another case if there are two subsequent back-to-back nmis [3]. The 2nd is detected as back-to-back because the first handled more than one counter. If the second handles one counter and the 3rd handles nothing, we drop the 3rd nmi because it could be a back-to-back nmi. Signed-off-by: Robert Richter <robert.richter@amd.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> [ renamed nmi variable to pmu_nmi to avoid clash with .nmi in entry.S ] Signed-off-by: Don Zickus <dzickus@redhat.com> Cc: peterz@infradead.org Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-3-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-09-03 02:07:48 +07:00
/* let the kernel handle the unknown nmi */
return NOTIFY_DONE;
/*
* This one is a PMU back-to-back nmi. Two events
* trigger 'simultaneously' raising two back-to-back
* NMIs. If the first NMI handles both, the latter
* will be empty and daze the CPU. So, we drop it to
* avoid false-positive 'unknown nmi' messages.
*/
return NOTIFY_STOP;
default:
return NOTIFY_DONE;
}
apic_write(APIC_LVTPC, APIC_DM_NMI);
perf, x86: Try to handle unknown nmis with an enabled PMU When the PMU is enabled it is valid to have unhandled nmis, two events could trigger 'simultaneously' raising two back-to-back NMIs. If the first NMI handles both, the latter will be empty and daze the CPU. The solution to avoid an 'unknown nmi' massage in this case was simply to stop the nmi handler chain when the PMU is enabled by stating the nmi was handled. This has the drawback that a) we can not detect unknown nmis anymore, and b) subsequent nmi handlers are not called. This patch addresses this. Now, we check this unknown NMI if it could be a PMU back-to-back NMI. Otherwise we pass it and let the kernel handle the unknown nmi. This is a debug log: cpu #6, nmi #32333, skip_nmi #32330, handled = 1, time = 1934364430 cpu #6, nmi #32334, skip_nmi #32330, handled = 1, time = 1934704616 cpu #6, nmi #32335, skip_nmi #32336, handled = 2, time = 1936032320 cpu #6, nmi #32336, skip_nmi #32336, handled = 0, time = 1936034139 cpu #6, nmi #32337, skip_nmi #32336, handled = 1, time = 1936120100 cpu #6, nmi #32338, skip_nmi #32336, handled = 1, time = 1936404607 cpu #6, nmi #32339, skip_nmi #32336, handled = 1, time = 1937983416 cpu #6, nmi #32340, skip_nmi #32341, handled = 2, time = 1938201032 cpu #6, nmi #32341, skip_nmi #32341, handled = 0, time = 1938202830 cpu #6, nmi #32342, skip_nmi #32341, handled = 1, time = 1938443743 cpu #6, nmi #32343, skip_nmi #32341, handled = 1, time = 1939956552 cpu #6, nmi #32344, skip_nmi #32341, handled = 1, time = 1940073224 cpu #6, nmi #32345, skip_nmi #32341, handled = 1, time = 1940485677 cpu #6, nmi #32346, skip_nmi #32347, handled = 2, time = 1941947772 cpu #6, nmi #32347, skip_nmi #32347, handled = 1, time = 1941949818 cpu #6, nmi #32348, skip_nmi #32347, handled = 0, time = 1941951591 Uhhuh. NMI received for unknown reason 00 on CPU 6. Do you have a strange power saving mode enabled? Dazed and confused, but trying to continue Deltas: nmi #32334 340186 nmi #32335 1327704 nmi #32336 1819 <<<< back-to-back nmi [1] nmi #32337 85961 nmi #32338 284507 nmi #32339 1578809 nmi #32340 217616 nmi #32341 1798 <<<< back-to-back nmi [2] nmi #32342 240913 nmi #32343 1512809 nmi #32344 116672 nmi #32345 412453 nmi #32346 1462095 <<<< 1st nmi (standard) handling 2 counters nmi #32347 2046 <<<< 2nd nmi (back-to-back) handling one counter nmi #32348 1773 <<<< 3rd nmi (back-to-back) handling no counter! [3] For back-to-back nmi detection there are the following rules: The PMU nmi handler was handling more than one counter and no counter was handled in the subsequent nmi (see [1] and [2] above). There is another case if there are two subsequent back-to-back nmis [3]. The 2nd is detected as back-to-back because the first handled more than one counter. If the second handles one counter and the 3rd handles nothing, we drop the 3rd nmi because it could be a back-to-back nmi. Signed-off-by: Robert Richter <robert.richter@amd.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> [ renamed nmi variable to pmu_nmi to avoid clash with .nmi in entry.S ] Signed-off-by: Don Zickus <dzickus@redhat.com> Cc: peterz@infradead.org Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-3-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-09-03 02:07:48 +07:00
handled = x86_pmu.handle_irq(args->regs);
if (!handled)
return NOTIFY_DONE;
this_nmi = percpu_read(irq_stat.__nmi_count);
if ((handled > 1) ||
/* the next nmi could be a back-to-back nmi */
((__this_cpu_read(pmu_nmi.marked) == this_nmi) &&
(__this_cpu_read(pmu_nmi.handled) > 1))) {
perf, x86: Try to handle unknown nmis with an enabled PMU When the PMU is enabled it is valid to have unhandled nmis, two events could trigger 'simultaneously' raising two back-to-back NMIs. If the first NMI handles both, the latter will be empty and daze the CPU. The solution to avoid an 'unknown nmi' massage in this case was simply to stop the nmi handler chain when the PMU is enabled by stating the nmi was handled. This has the drawback that a) we can not detect unknown nmis anymore, and b) subsequent nmi handlers are not called. This patch addresses this. Now, we check this unknown NMI if it could be a PMU back-to-back NMI. Otherwise we pass it and let the kernel handle the unknown nmi. This is a debug log: cpu #6, nmi #32333, skip_nmi #32330, handled = 1, time = 1934364430 cpu #6, nmi #32334, skip_nmi #32330, handled = 1, time = 1934704616 cpu #6, nmi #32335, skip_nmi #32336, handled = 2, time = 1936032320 cpu #6, nmi #32336, skip_nmi #32336, handled = 0, time = 1936034139 cpu #6, nmi #32337, skip_nmi #32336, handled = 1, time = 1936120100 cpu #6, nmi #32338, skip_nmi #32336, handled = 1, time = 1936404607 cpu #6, nmi #32339, skip_nmi #32336, handled = 1, time = 1937983416 cpu #6, nmi #32340, skip_nmi #32341, handled = 2, time = 1938201032 cpu #6, nmi #32341, skip_nmi #32341, handled = 0, time = 1938202830 cpu #6, nmi #32342, skip_nmi #32341, handled = 1, time = 1938443743 cpu #6, nmi #32343, skip_nmi #32341, handled = 1, time = 1939956552 cpu #6, nmi #32344, skip_nmi #32341, handled = 1, time = 1940073224 cpu #6, nmi #32345, skip_nmi #32341, handled = 1, time = 1940485677 cpu #6, nmi #32346, skip_nmi #32347, handled = 2, time = 1941947772 cpu #6, nmi #32347, skip_nmi #32347, handled = 1, time = 1941949818 cpu #6, nmi #32348, skip_nmi #32347, handled = 0, time = 1941951591 Uhhuh. NMI received for unknown reason 00 on CPU 6. Do you have a strange power saving mode enabled? Dazed and confused, but trying to continue Deltas: nmi #32334 340186 nmi #32335 1327704 nmi #32336 1819 <<<< back-to-back nmi [1] nmi #32337 85961 nmi #32338 284507 nmi #32339 1578809 nmi #32340 217616 nmi #32341 1798 <<<< back-to-back nmi [2] nmi #32342 240913 nmi #32343 1512809 nmi #32344 116672 nmi #32345 412453 nmi #32346 1462095 <<<< 1st nmi (standard) handling 2 counters nmi #32347 2046 <<<< 2nd nmi (back-to-back) handling one counter nmi #32348 1773 <<<< 3rd nmi (back-to-back) handling no counter! [3] For back-to-back nmi detection there are the following rules: The PMU nmi handler was handling more than one counter and no counter was handled in the subsequent nmi (see [1] and [2] above). There is another case if there are two subsequent back-to-back nmis [3]. The 2nd is detected as back-to-back because the first handled more than one counter. If the second handles one counter and the 3rd handles nothing, we drop the 3rd nmi because it could be a back-to-back nmi. Signed-off-by: Robert Richter <robert.richter@amd.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> [ renamed nmi variable to pmu_nmi to avoid clash with .nmi in entry.S ] Signed-off-by: Don Zickus <dzickus@redhat.com> Cc: peterz@infradead.org Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-3-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-09-03 02:07:48 +07:00
/*
* We could have two subsequent back-to-back nmis: The
* first handles more than one counter, the 2nd
* handles only one counter and the 3rd handles no
* counter.
*
* This is the 2nd nmi because the previous was
* handling more than one counter. We will mark the
* next (3rd) and then drop it if unhandled.
*/
__this_cpu_write(pmu_nmi.marked, this_nmi + 1);
__this_cpu_write(pmu_nmi.handled, handled);
perf, x86: Try to handle unknown nmis with an enabled PMU When the PMU is enabled it is valid to have unhandled nmis, two events could trigger 'simultaneously' raising two back-to-back NMIs. If the first NMI handles both, the latter will be empty and daze the CPU. The solution to avoid an 'unknown nmi' massage in this case was simply to stop the nmi handler chain when the PMU is enabled by stating the nmi was handled. This has the drawback that a) we can not detect unknown nmis anymore, and b) subsequent nmi handlers are not called. This patch addresses this. Now, we check this unknown NMI if it could be a PMU back-to-back NMI. Otherwise we pass it and let the kernel handle the unknown nmi. This is a debug log: cpu #6, nmi #32333, skip_nmi #32330, handled = 1, time = 1934364430 cpu #6, nmi #32334, skip_nmi #32330, handled = 1, time = 1934704616 cpu #6, nmi #32335, skip_nmi #32336, handled = 2, time = 1936032320 cpu #6, nmi #32336, skip_nmi #32336, handled = 0, time = 1936034139 cpu #6, nmi #32337, skip_nmi #32336, handled = 1, time = 1936120100 cpu #6, nmi #32338, skip_nmi #32336, handled = 1, time = 1936404607 cpu #6, nmi #32339, skip_nmi #32336, handled = 1, time = 1937983416 cpu #6, nmi #32340, skip_nmi #32341, handled = 2, time = 1938201032 cpu #6, nmi #32341, skip_nmi #32341, handled = 0, time = 1938202830 cpu #6, nmi #32342, skip_nmi #32341, handled = 1, time = 1938443743 cpu #6, nmi #32343, skip_nmi #32341, handled = 1, time = 1939956552 cpu #6, nmi #32344, skip_nmi #32341, handled = 1, time = 1940073224 cpu #6, nmi #32345, skip_nmi #32341, handled = 1, time = 1940485677 cpu #6, nmi #32346, skip_nmi #32347, handled = 2, time = 1941947772 cpu #6, nmi #32347, skip_nmi #32347, handled = 1, time = 1941949818 cpu #6, nmi #32348, skip_nmi #32347, handled = 0, time = 1941951591 Uhhuh. NMI received for unknown reason 00 on CPU 6. Do you have a strange power saving mode enabled? Dazed and confused, but trying to continue Deltas: nmi #32334 340186 nmi #32335 1327704 nmi #32336 1819 <<<< back-to-back nmi [1] nmi #32337 85961 nmi #32338 284507 nmi #32339 1578809 nmi #32340 217616 nmi #32341 1798 <<<< back-to-back nmi [2] nmi #32342 240913 nmi #32343 1512809 nmi #32344 116672 nmi #32345 412453 nmi #32346 1462095 <<<< 1st nmi (standard) handling 2 counters nmi #32347 2046 <<<< 2nd nmi (back-to-back) handling one counter nmi #32348 1773 <<<< 3rd nmi (back-to-back) handling no counter! [3] For back-to-back nmi detection there are the following rules: The PMU nmi handler was handling more than one counter and no counter was handled in the subsequent nmi (see [1] and [2] above). There is another case if there are two subsequent back-to-back nmis [3]. The 2nd is detected as back-to-back because the first handled more than one counter. If the second handles one counter and the 3rd handles nothing, we drop the 3rd nmi because it could be a back-to-back nmi. Signed-off-by: Robert Richter <robert.richter@amd.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> [ renamed nmi variable to pmu_nmi to avoid clash with .nmi in entry.S ] Signed-off-by: Don Zickus <dzickus@redhat.com> Cc: peterz@infradead.org Cc: gorcunov@gmail.com Cc: fweisbec@gmail.com Cc: ying.huang@intel.com Cc: ming.m.lin@intel.com Cc: eranian@google.com LKML-Reference: <1283454469-1909-3-git-send-email-dzickus@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-09-03 02:07:48 +07:00
}
return NOTIFY_STOP;
}
static __read_mostly struct notifier_block perf_event_nmi_notifier = {
.notifier_call = perf_event_nmi_handler,
.next = NULL,
.priority = NMI_LOCAL_LOW_PRIOR,
};
static struct event_constraint unconstrained;
static struct event_constraint emptyconstraint;
static struct event_constraint *
x86_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event)
{
struct event_constraint *c;
if (x86_pmu.event_constraints) {
for_each_event_constraint(c, x86_pmu.event_constraints) {
if ((event->hw.config & c->cmask) == c->code)
return c;
}
}
return &unconstrained;
}
#include "perf_event_amd.c"
#include "perf_event_p6.c"
perf, x86: Implement initial P4 PMU driver The netburst PMU is way different from the "architectural perfomance monitoring" specification that current CPUs use. P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle perfomance monitoring events. A few implementational details: 1) We need a separate x86_pmu::hw_config helper in struct x86_pmu since register bit-fields are quite different from P6, Core and later cpu series. 2) For the same reason is a x86_pmu::schedule_events helper introduced. 3) hw_perf_event::config consists of packed ESCR+CCCR values. It's allowed since in reality both registers only use a half of their size. Of course before making a real write into a particular MSR we need to unpack the value and extend it to a proper size. 4) The tuple of packed ESCR+CCCR in hw_perf_event::config doesn't describe the memory address of ESCR MSR register so that we need to keep a mapping between these tuples used and available ESCR (various P4 events may use same ESCRs but not simultaneously), for this sake every active event has a per-cpu map of hw_perf_event::idx <--> ESCR addresses. 5) Since hw_perf_event::idx is an offset to counter/control register we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel strips it down to 8 registers and event armed may never be turned off (ie the bit in active_mask is set but the loop never reaches this index to check), thanks to Peter Zijlstra Restrictions: - No cascaded counters support (do we ever need them?) - No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS doesn't work for now) - There are events with same counters which can't work simultaneously (need to use intersected ones due to broken counter 1) - No PERF_COUNT_HW_CACHE_ events yet Todo: - Implement dependent events - Need proper hashing for event opcodes (no linear search, good for debugging stage but not in real loads) - Some events counted during a clock cycle -- need to set threshold for them and count every clock cycle just to get summary statistics (ie to behave the same way as other PMUs do) - Need to swicth to use event_constraints - To support RAW events we need to encode a global list of P4 events into p4_templates - Cache events need to be added Event support status matrix: Event status ----------------------------- cycles works cache-references works cache-misses works branch-misses works bus-cycles partially (does not work on 64bit cpu with HT enabled) instruction doesnt work (needs dependent event [mop tagging]) branches doesnt work Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Robert Richter <robert.richter@amd.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20100311165439.GB5129@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 23:54:39 +07:00
#include "perf_event_p4.c"
#include "perf_event_intel_lbr.c"
#include "perf_event_intel_ds.c"
#include "perf_event_intel.c"
static int __cpuinit
x86_pmu_notifier(struct notifier_block *self, unsigned long action, void *hcpu)
{
unsigned int cpu = (long)hcpu;
int ret = NOTIFY_OK;
switch (action & ~CPU_TASKS_FROZEN) {
case CPU_UP_PREPARE:
if (x86_pmu.cpu_prepare)
ret = x86_pmu.cpu_prepare(cpu);
break;
case CPU_STARTING:
if (x86_pmu.cpu_starting)
x86_pmu.cpu_starting(cpu);
break;
case CPU_DYING:
if (x86_pmu.cpu_dying)
x86_pmu.cpu_dying(cpu);
break;
case CPU_UP_CANCELED:
case CPU_DEAD:
if (x86_pmu.cpu_dead)
x86_pmu.cpu_dead(cpu);
break;
default:
break;
}
return ret;
}
static void __init pmu_check_apic(void)
{
if (cpu_has_apic)
return;
x86_pmu.apic = 0;
pr_info("no APIC, boot with the \"lapic\" boot parameter to force-enable it.\n");
pr_info("no hardware sampling interrupt available.\n");
}
static int __init init_hw_perf_events(void)
{
struct event_constraint *c;
int err;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
pr_info("Performance Events: ");
switch (boot_cpu_data.x86_vendor) {
case X86_VENDOR_INTEL:
err = intel_pmu_init();
break;
case X86_VENDOR_AMD:
err = amd_pmu_init();
break;
default:
return 0;
}
if (err != 0) {
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
pr_cont("no PMU driver, software events only.\n");
return 0;
}
pmu_check_apic();
/* sanity check that the hardware exists or is emulated */
if (!check_hw_exists())
return 0;
pr_cont("%s PMU driver.\n", x86_pmu.name);
if (x86_pmu.quirks)
x86_pmu.quirks();
if (x86_pmu.num_counters > X86_PMC_MAX_GENERIC) {
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
WARN(1, KERN_ERR "hw perf events %d > max(%d), clipping!",
x86_pmu.num_counters, X86_PMC_MAX_GENERIC);
x86_pmu.num_counters = X86_PMC_MAX_GENERIC;
}
x86_pmu.intel_ctrl = (1 << x86_pmu.num_counters) - 1;
if (x86_pmu.num_counters_fixed > X86_PMC_MAX_FIXED) {
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
WARN(1, KERN_ERR "hw perf events fixed %d > max(%d), clipping!",
x86_pmu.num_counters_fixed, X86_PMC_MAX_FIXED);
x86_pmu.num_counters_fixed = X86_PMC_MAX_FIXED;
}
x86_pmu.intel_ctrl |=
((1LL << x86_pmu.num_counters_fixed)-1) << X86_PMC_IDX_FIXED;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
perf_events_lapic_init();
register_die_notifier(&perf_event_nmi_notifier);
unconstrained = (struct event_constraint)
__EVENT_CONSTRAINT(0, (1ULL << x86_pmu.num_counters) - 1,
0, x86_pmu.num_counters);
if (x86_pmu.event_constraints) {
for_each_event_constraint(c, x86_pmu.event_constraints) {
if (c->cmask != X86_RAW_EVENT_MASK)
continue;
c->idxmsk64 |= (1ULL << x86_pmu.num_counters) - 1;
c->weight += x86_pmu.num_counters;
}
}
pr_info("... version: %d\n", x86_pmu.version);
pr_info("... bit width: %d\n", x86_pmu.cntval_bits);
pr_info("... generic registers: %d\n", x86_pmu.num_counters);
pr_info("... value mask: %016Lx\n", x86_pmu.cntval_mask);
pr_info("... max period: %016Lx\n", x86_pmu.max_period);
pr_info("... fixed-purpose events: %d\n", x86_pmu.num_counters_fixed);
pr_info("... event mask: %016Lx\n", x86_pmu.intel_ctrl);
perf_pmu_register(&pmu, "cpu", PERF_TYPE_RAW);
perf_cpu_notifier(x86_pmu_notifier);
return 0;
}
early_initcall(init_hw_perf_events);
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
static inline void x86_pmu_read(struct perf_event *event)
{
x86_perf_event_update(event);
}
/*
* Start group events scheduling transaction
* Set the flag to make pmu::enable() not perform the
* schedulability test, it will be performed at commit time
*/
static void x86_pmu_start_txn(struct pmu *pmu)
{
perf_pmu_disable(pmu);
__this_cpu_or(cpu_hw_events.group_flag, PERF_EVENT_TXN);
__this_cpu_write(cpu_hw_events.n_txn, 0);
}
/*
* Stop group events scheduling transaction
* Clear the flag and pmu::enable() will perform the
* schedulability test.
*/
static void x86_pmu_cancel_txn(struct pmu *pmu)
{
__this_cpu_and(cpu_hw_events.group_flag, ~PERF_EVENT_TXN);
perf_events: Fix event scheduling issues introduced by transactional API The transactional API patch between the generic and model-specific code introduced several important bugs with event scheduling, at least on X86. If you had pinned events, e.g., watchdog, and were over-committing the PMU, you would get bogus counts. The bug was showing up on Intel CPU because events would move around more often that on AMD. But the problem also existed on AMD, though harder to expose. The issues were: - group_sched_in() was missing a cancel_txn() in the error path - cpuc->n_added was not properly maintained, leading to missing actions in hw_perf_enable(), i.e., n_running being 0. You cannot update n_added until you know the transaction has succeeded. In case of failed transaction n_added was not adjusted back. - in case of failed transactions, event_sched_out() was called and eventually invoked x86_disable_event() to touch the HW reg. But with transactions, on X86, event_sched_in() does not touch HW registers, it simply collects events into a list. Thus, you could end up calling x86_disable_event() on a counter which did not correspond to the current event when idx != -1. The patch modifies the generic and X86 code to avoid all those problems. First, we keep track of the number of events added last. In case the transaction fails, we substract them from n_added. This approach is necessary (as opposed to delaying updates to n_added) because not all event updates use the transaction API, e.g., single events. Second, we encapsulate the event_sched_in() and event_sched_out() in group_sched_in() inside the transaction. That makes the operations symmetrical and you can also detect that you are inside a transaction and skip the HW reg access by checking cpuc->group_flag. With this patch, you can now overcommit the PMU even with pinned system-wide events present and still get valid counts. Signed-off-by: Stephane Eranian <eranian@google.com> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <1274796225.5882.1389.camel@twins> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-05-25 21:23:10 +07:00
/*
* Truncate the collected events.
*/
__this_cpu_sub(cpu_hw_events.n_added, __this_cpu_read(cpu_hw_events.n_txn));
__this_cpu_sub(cpu_hw_events.n_events, __this_cpu_read(cpu_hw_events.n_txn));
perf_pmu_enable(pmu);
}
/*
* Commit group events scheduling transaction
* Perform the group schedulability test as a whole
* Return 0 if success
*/
static int x86_pmu_commit_txn(struct pmu *pmu)
{
struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
int assign[X86_PMC_IDX_MAX];
int n, ret;
n = cpuc->n_events;
if (!x86_pmu_initialized())
return -EAGAIN;
ret = x86_pmu.schedule_events(cpuc, n, assign);
if (ret)
return ret;
/*
* copy new assignment, now we know it is possible
* will be used by hw_perf_enable()
*/
memcpy(cpuc->assign, assign, n*sizeof(int));
cpuc->group_flag &= ~PERF_EVENT_TXN;
perf_pmu_enable(pmu);
return 0;
}
/*
* validate that we can schedule this event
*/
static int validate_event(struct perf_event *event)
{
struct cpu_hw_events *fake_cpuc;
struct event_constraint *c;
int ret = 0;
fake_cpuc = kmalloc(sizeof(*fake_cpuc), GFP_KERNEL | __GFP_ZERO);
if (!fake_cpuc)
return -ENOMEM;
c = x86_pmu.get_event_constraints(fake_cpuc, event);
if (!c || !c->weight)
ret = -ENOSPC;
if (x86_pmu.put_event_constraints)
x86_pmu.put_event_constraints(fake_cpuc, event);
kfree(fake_cpuc);
return ret;
}
/*
* validate a single event group
*
* validation include:
* - check events are compatible which each other
* - events do not compete for the same counter
* - number of events <= number of counters
*
* validation ensures the group can be loaded onto the
* PMU if it was the only group available.
*/
static int validate_group(struct perf_event *event)
{
struct perf_event *leader = event->group_leader;
struct cpu_hw_events *fake_cpuc;
int ret, n;
ret = -ENOMEM;
fake_cpuc = kmalloc(sizeof(*fake_cpuc), GFP_KERNEL | __GFP_ZERO);
if (!fake_cpuc)
goto out;
/*
* the event is not yet connected with its
* siblings therefore we must first collect
* existing siblings, then add the new event
* before we can simulate the scheduling
*/
ret = -ENOSPC;
n = collect_events(fake_cpuc, leader, true);
if (n < 0)
goto out_free;
fake_cpuc->n_events = n;
n = collect_events(fake_cpuc, event, false);
if (n < 0)
goto out_free;
fake_cpuc->n_events = n;
perf, x86: Implement initial P4 PMU driver The netburst PMU is way different from the "architectural perfomance monitoring" specification that current CPUs use. P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle perfomance monitoring events. A few implementational details: 1) We need a separate x86_pmu::hw_config helper in struct x86_pmu since register bit-fields are quite different from P6, Core and later cpu series. 2) For the same reason is a x86_pmu::schedule_events helper introduced. 3) hw_perf_event::config consists of packed ESCR+CCCR values. It's allowed since in reality both registers only use a half of their size. Of course before making a real write into a particular MSR we need to unpack the value and extend it to a proper size. 4) The tuple of packed ESCR+CCCR in hw_perf_event::config doesn't describe the memory address of ESCR MSR register so that we need to keep a mapping between these tuples used and available ESCR (various P4 events may use same ESCRs but not simultaneously), for this sake every active event has a per-cpu map of hw_perf_event::idx <--> ESCR addresses. 5) Since hw_perf_event::idx is an offset to counter/control register we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel strips it down to 8 registers and event armed may never be turned off (ie the bit in active_mask is set but the loop never reaches this index to check), thanks to Peter Zijlstra Restrictions: - No cascaded counters support (do we ever need them?) - No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS doesn't work for now) - There are events with same counters which can't work simultaneously (need to use intersected ones due to broken counter 1) - No PERF_COUNT_HW_CACHE_ events yet Todo: - Implement dependent events - Need proper hashing for event opcodes (no linear search, good for debugging stage but not in real loads) - Some events counted during a clock cycle -- need to set threshold for them and count every clock cycle just to get summary statistics (ie to behave the same way as other PMUs do) - Need to swicth to use event_constraints - To support RAW events we need to encode a global list of P4 events into p4_templates - Cache events need to be added Event support status matrix: Event status ----------------------------- cycles works cache-references works cache-misses works branch-misses works bus-cycles partially (does not work on 64bit cpu with HT enabled) instruction doesnt work (needs dependent event [mop tagging]) branches doesnt work Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by: Lin Ming <ming.m.lin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Robert Richter <robert.richter@amd.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <20100311165439.GB5129@lenovo> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-03-11 23:54:39 +07:00
ret = x86_pmu.schedule_events(fake_cpuc, n, NULL);
out_free:
kfree(fake_cpuc);
out:
return ret;
}
static int x86_pmu_event_init(struct perf_event *event)
{
struct pmu *tmp;
int err;
switch (event->attr.type) {
case PERF_TYPE_RAW:
case PERF_TYPE_HARDWARE:
case PERF_TYPE_HW_CACHE:
break;
default:
return -ENOENT;
}
err = __x86_pmu_event_init(event);
if (!err) {
/*
* we temporarily connect event to its pmu
* such that validate_group() can classify
* it as an x86 event using is_x86_event()
*/
tmp = event->pmu;
event->pmu = &pmu;
if (event->group_leader != event)
err = validate_group(event);
else
err = validate_event(event);
event->pmu = tmp;
}
if (err) {
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 17:02:48 +07:00
if (event->destroy)
event->destroy(event);
}
return err;
}
static struct pmu pmu = {
2010-06-16 19:37:10 +07:00
.pmu_enable = x86_pmu_enable,
.pmu_disable = x86_pmu_disable,
.event_init = x86_pmu_event_init,
2010-06-16 19:37:10 +07:00
.add = x86_pmu_add,
.del = x86_pmu_del,
.start = x86_pmu_start,
.stop = x86_pmu_stop,
.read = x86_pmu_read,
2010-06-16 19:37:10 +07:00
.start_txn = x86_pmu_start_txn,
.cancel_txn = x86_pmu_cancel_txn,
.commit_txn = x86_pmu_commit_txn,
};
/*
* callchain support
*/
static void
backtrace_warning_symbol(void *data, char *msg, unsigned long symbol)
{
/* Ignore warnings */
}
static void backtrace_warning(void *data, char *msg)
{
/* Ignore warnings */
}
static int backtrace_stack(void *data, char *name)
{
return 0;
}
static void backtrace_address(void *data, unsigned long addr, int reliable)
{
struct perf_callchain_entry *entry = data;
perf_callchain_store(entry, addr);
}
static const struct stacktrace_ops backtrace_ops = {
.warning = backtrace_warning,
.warning_symbol = backtrace_warning_symbol,
.stack = backtrace_stack,
.address = backtrace_address,
.walk_stack = print_context_stack_bp,
};
void
perf_callchain_kernel(struct perf_callchain_entry *entry, struct pt_regs *regs)
{
if (perf_guest_cbs && perf_guest_cbs->is_in_guest()) {
/* TODO: We don't support guest os callchain now */
return;
}
perf_callchain_store(entry, regs->ip);
x86, dumpstack: Correct stack dump info when frame pointer is available Current stack dump code scans entire stack and check each entry contains a pointer to kernel code. If CONFIG_FRAME_POINTER=y it could mark whether the pointer is valid or not based on value of the frame pointer. Invalid entries could be preceded by '?' sign. However this was not going to happen because scan start point was always higher than the frame pointer so that they could not meet. Commit 9c0729dc8062 ("x86: Eliminate bp argument from the stack tracing routines") delayed bp acquisition point, so the bp was read in lower frame, thus all of the entries were marked invalid. This patch fixes this by reverting above commit while retaining stack_frame() helper as suggested by Frederic Weisbecker. End result looks like below: before: [ 3.508329] Call Trace: [ 3.508551] [<ffffffff814f35c9>] ? panic+0x91/0x199 [ 3.508662] [<ffffffff814f3739>] ? printk+0x68/0x6a [ 3.508770] [<ffffffff81a981b2>] ? mount_block_root+0x257/0x26e [ 3.508876] [<ffffffff81a9821f>] ? mount_root+0x56/0x5a [ 3.508975] [<ffffffff81a98393>] ? prepare_namespace+0x170/0x1a9 [ 3.509216] [<ffffffff81a9772b>] ? kernel_init+0x1d2/0x1e2 [ 3.509335] [<ffffffff81003894>] ? kernel_thread_helper+0x4/0x10 [ 3.509442] [<ffffffff814f6880>] ? restore_args+0x0/0x30 [ 3.509542] [<ffffffff81a97559>] ? kernel_init+0x0/0x1e2 [ 3.509641] [<ffffffff81003890>] ? kernel_thread_helper+0x0/0x10 after: [ 3.522991] Call Trace: [ 3.523351] [<ffffffff814f35b9>] panic+0x91/0x199 [ 3.523468] [<ffffffff814f3729>] ? printk+0x68/0x6a [ 3.523576] [<ffffffff81a981b2>] mount_block_root+0x257/0x26e [ 3.523681] [<ffffffff81a9821f>] mount_root+0x56/0x5a [ 3.523780] [<ffffffff81a98393>] prepare_namespace+0x170/0x1a9 [ 3.523885] [<ffffffff81a9772b>] kernel_init+0x1d2/0x1e2 [ 3.523987] [<ffffffff81003894>] kernel_thread_helper+0x4/0x10 [ 3.524228] [<ffffffff814f6880>] ? restore_args+0x0/0x30 [ 3.524345] [<ffffffff81a97559>] ? kernel_init+0x0/0x1e2 [ 3.524445] [<ffffffff81003890>] ? kernel_thread_helper+0x0/0x10 -v5: * fix build breakage with oprofile -v4: * use 0 instead of regs->bp * separate out printk changes -v3: * apply comment from Frederic * add a couple of printk fixes Signed-off-by: Namhyung Kim <namhyung@gmail.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Soren Sandmann <ssp@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Robert Richter <robert.richter@amd.com> LKML-Reference: <1300416006-3163-1-git-send-email-namhyung@gmail.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-03-18 09:40:06 +07:00
dump_trace(NULL, regs, NULL, 0, &backtrace_ops, entry);
}
#ifdef CONFIG_COMPAT
static inline int
perf_callchain_user32(struct pt_regs *regs, struct perf_callchain_entry *entry)
perf_counter: x86: Fix call-chain support to use NMI-safe methods __copy_from_user_inatomic() isn't NMI safe in that it can trigger the page fault handler which is another trap and its return path invokes IRET which will also close the NMI context. Therefore use a GUP based approach to copy the stack frames over. We tried an alternative solution as well: we used a forward ported version of Mathieu Desnoyers's "NMI safe INT3 and Page Fault" patch that modifies the exception return path to use an open-coded IRET with explicit stack unrolling and TF checking. This didnt work as it interacted with faulting user-space instructions, causing them not to restart properly, which corrupts user-space registers. Solving that would probably involve disassembling those instructions and backtracing the RIP. But even without that, the code was deemed rather complex to the already non-trivial x86 entry assembly code, so instead we went for this GUP based method that does a software-walk of the pagetables. Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Nick Piggin <npiggin@suse.de> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Cc: Vegard Nossum <vegard.nossum@gmail.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-06-15 18:07:24 +07:00
{
/* 32-bit process in 64-bit kernel. */
struct stack_frame_ia32 frame;
const void __user *fp;
perf_counter: x86: Fix call-chain support to use NMI-safe methods __copy_from_user_inatomic() isn't NMI safe in that it can trigger the page fault handler which is another trap and its return path invokes IRET which will also close the NMI context. Therefore use a GUP based approach to copy the stack frames over. We tried an alternative solution as well: we used a forward ported version of Mathieu Desnoyers's "NMI safe INT3 and Page Fault" patch that modifies the exception return path to use an open-coded IRET with explicit stack unrolling and TF checking. This didnt work as it interacted with faulting user-space instructions, causing them not to restart properly, which corrupts user-space registers. Solving that would probably involve disassembling those instructions and backtracing the RIP. But even without that, the code was deemed rather complex to the already non-trivial x86 entry assembly code, so instead we went for this GUP based method that does a software-walk of the pagetables. Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Nick Piggin <npiggin@suse.de> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Cc: Vegard Nossum <vegard.nossum@gmail.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-06-15 18:07:24 +07:00
if (!test_thread_flag(TIF_IA32))
return 0;
fp = compat_ptr(regs->bp);
while (entry->nr < PERF_MAX_STACK_DEPTH) {
unsigned long bytes;
frame.next_frame = 0;
frame.return_address = 0;
bytes = copy_from_user_nmi(&frame, fp, sizeof(frame));
if (bytes != sizeof(frame))
break;
perf_counter: x86: Fix call-chain support to use NMI-safe methods __copy_from_user_inatomic() isn't NMI safe in that it can trigger the page fault handler which is another trap and its return path invokes IRET which will also close the NMI context. Therefore use a GUP based approach to copy the stack frames over. We tried an alternative solution as well: we used a forward ported version of Mathieu Desnoyers's "NMI safe INT3 and Page Fault" patch that modifies the exception return path to use an open-coded IRET with explicit stack unrolling and TF checking. This didnt work as it interacted with faulting user-space instructions, causing them not to restart properly, which corrupts user-space registers. Solving that would probably involve disassembling those instructions and backtracing the RIP. But even without that, the code was deemed rather complex to the already non-trivial x86 entry assembly code, so instead we went for this GUP based method that does a software-walk of the pagetables. Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Nick Piggin <npiggin@suse.de> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Cc: Vegard Nossum <vegard.nossum@gmail.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-06-15 18:07:24 +07:00
if (fp < compat_ptr(regs->sp))
break;
perf_counter: x86: Fix call-chain support to use NMI-safe methods __copy_from_user_inatomic() isn't NMI safe in that it can trigger the page fault handler which is another trap and its return path invokes IRET which will also close the NMI context. Therefore use a GUP based approach to copy the stack frames over. We tried an alternative solution as well: we used a forward ported version of Mathieu Desnoyers's "NMI safe INT3 and Page Fault" patch that modifies the exception return path to use an open-coded IRET with explicit stack unrolling and TF checking. This didnt work as it interacted with faulting user-space instructions, causing them not to restart properly, which corrupts user-space registers. Solving that would probably involve disassembling those instructions and backtracing the RIP. But even without that, the code was deemed rather complex to the already non-trivial x86 entry assembly code, so instead we went for this GUP based method that does a software-walk of the pagetables. Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Nick Piggin <npiggin@suse.de> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Cc: Vegard Nossum <vegard.nossum@gmail.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-06-15 18:07:24 +07:00
perf_callchain_store(entry, frame.return_address);
fp = compat_ptr(frame.next_frame);
}
return 1;
}
#else
static inline int
perf_callchain_user32(struct pt_regs *regs, struct perf_callchain_entry *entry)
{
return 0;
}
#endif
void
perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs)
{
struct stack_frame frame;
const void __user *fp;
if (perf_guest_cbs && perf_guest_cbs->is_in_guest()) {
/* TODO: We don't support guest os callchain now */
return;
}
perf_counter: x86: Fix call-chain support to use NMI-safe methods __copy_from_user_inatomic() isn't NMI safe in that it can trigger the page fault handler which is another trap and its return path invokes IRET which will also close the NMI context. Therefore use a GUP based approach to copy the stack frames over. We tried an alternative solution as well: we used a forward ported version of Mathieu Desnoyers's "NMI safe INT3 and Page Fault" patch that modifies the exception return path to use an open-coded IRET with explicit stack unrolling and TF checking. This didnt work as it interacted with faulting user-space instructions, causing them not to restart properly, which corrupts user-space registers. Solving that would probably involve disassembling those instructions and backtracing the RIP. But even without that, the code was deemed rather complex to the already non-trivial x86 entry assembly code, so instead we went for this GUP based method that does a software-walk of the pagetables. Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Nick Piggin <npiggin@suse.de> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Cc: Vegard Nossum <vegard.nossum@gmail.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-06-15 18:07:24 +07:00
fp = (void __user *)regs->bp;
perf_callchain_store(entry, regs->ip);
if (perf_callchain_user32(regs, entry))
return;
while (entry->nr < PERF_MAX_STACK_DEPTH) {
unsigned long bytes;
frame.next_frame = NULL;
frame.return_address = 0;
bytes = copy_from_user_nmi(&frame, fp, sizeof(frame));
if (bytes != sizeof(frame))
break;
if ((unsigned long)fp < regs->sp)
break;
perf_callchain_store(entry, frame.return_address);
fp = frame.next_frame;
}
}
unsigned long perf_instruction_pointer(struct pt_regs *regs)
{
unsigned long ip;
if (perf_guest_cbs && perf_guest_cbs->is_in_guest())
ip = perf_guest_cbs->get_guest_ip();
else
ip = instruction_pointer(regs);
return ip;
}
unsigned long perf_misc_flags(struct pt_regs *regs)
{
int misc = 0;
if (perf_guest_cbs && perf_guest_cbs->is_in_guest()) {
if (perf_guest_cbs->is_user_mode())
misc |= PERF_RECORD_MISC_GUEST_USER;
else
misc |= PERF_RECORD_MISC_GUEST_KERNEL;
} else {
if (user_mode(regs))
misc |= PERF_RECORD_MISC_USER;
else
misc |= PERF_RECORD_MISC_KERNEL;
}
if (regs->flags & PERF_EFLAGS_EXACT)
misc |= PERF_RECORD_MISC_EXACT_IP;
return misc;
}