perf tools: Add support for some spe events and precise ip
v2 cover perf tools: Add support for some spe events and precise ip
After the commit ffd3d18c20b8 ("perf tools: Add ARM Statistical Profiling Extensions (SPE) support") was merged, "perf record" and "perf report --dump-raw-trace" are supported. However, the raw data that is dumped cannot be used without parsing.
This patchset is to improve the "perf report" support for SPE, and further process the data. Currently, support for the three events of llc-miss, tlb-miss, branch-miss and remote-access is added.
The macro definition was modified under Jeremy's suggestion, and the "event:pp" approach was used under James' suggestion to achieve the precise ip of some events on arm64. Currently, only branch-misses are implemented, and other event support will be added later.
In addition, we also found that when recording large multi-threaded programs, ctrl + c could not end recording, so it was fixed and two patches were added.
James Clark (1):
perf tools: Unset precise_ip when using SPE
Tan Xiaojun (4):
perf tools: Move arm-spe-pkt-decoder.h/c to the new dir perf tools: Add support for "report" for some spe events perf report: Add --spe options for arm-spe
perf tools: Support "branch-misses:pp" on arm64
Wei Li (2):
perf tools: add perf_evlist__terminate() for terminate perf tools: arm-spe: fix record hang after being terminated
tools/perf/Documentation/perf-report.txt | 10 +
tools/perf/arch/arm64/util/arm-spe.c | 10 +-
tools/perf/builtin-record.c | 1 +
tools/perf/builtin-report.c | 5 +
tools/perf/util/Build | 2 +-
tools/perf/util/arm-spe-decoder/Build | 1 +
.../util/arm-spe-decoder/arm-spe-decoder.c | 225 +++++
.../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++
.../arm-spe-pkt-decoder.c | 0
.../arm-spe-pkt-decoder.h | 2 +
tools/perf/util/arm-spe.c | 789 +++++++++++++++++-
tools/perf/util/arm-spe.h | 3 +
tools/perf/util/auxtrace.c | 49 ++
tools/perf/util/auxtrace.h | 29 +
tools/perf/util/evlist.c | 16 +
tools/perf/util/evlist.h | 1 +
tools/perf/util/evsel.h | 1 +
tools/perf/util/session.h | 2 +
18 files changed, 1170 insertions(+), 42 deletions(-)
create mode 100644 tools/perf/util/arm-spe-decoder/Build
create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c
create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%)
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (96%)Signature · 1 line
2.25.0
PATCH v2 1/7 perf tools: Move arm-spe-pkt-decoder.h/c to the new dir
From: Tan Xiaojun <[email protected]>
Create a new arm-spe-decoder directory for subsequent extensions and move arm-spe-pkt-decoder.h/c to this directory. No code changes.
---
tools/perf/util/Build | 2 +-
tools/perf/util/arm-spe-decoder/Build | 1 +
tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c | 0
tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h | 0
tools/perf/util/arm-spe.c | 2 +-
5 files changed, 3 insertions(+), 2 deletions(-)
create mode 100644 tools/perf/util/arm-spe-decoder/Build
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%)
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (100%)tools/perf/util/Build +1 -1
@@ -104,7 +104,7 @@ perf-$(CONFIG_AUXTRACE) += intel-pt-decoder/ perf-$(CONFIG_AUXTRACE) += intel-pt.o perf-$(CONFIG_AUXTRACE) += intel-bts.o perf-$(CONFIG_AUXTRACE) += arm-spe.o -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-decoder/ perf-$(CONFIG_AUXTRACE) += s390-cpumsf.o ifdef CONFIG_LIBOPENCSD
tools/perf/util/arm-spe-decoder/Build new file +1 -0
@@ -0,0 +1 @@ +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.c renamed from tools/perf/util/arm-spe-pkt-decoder.c +0 -0
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h renamed from tools/perf/util/arm-spe-pkt-decoder.h +0 -0
tools/perf/util/arm-spe.c +1 -1
@@ -23,7 +23,7 @@#include "debug.h" #include "auxtrace.h" #include "arm-spe.h" -#include "arm-spe-pkt-decoder.h" +#include "arm-spe-decoder/arm-spe-pkt-decoder.h" struct arm_spe {struct auxtrace auxtrace;
Signature · 1 line
2.25.0
PATCH v2 2/7 perf tools: Add support for "report" for some spe events
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
After the commit ffd3d18c20b8 ("perf tools: Add ARM Statistical Profiling Extensions (SPE) support") is merged, "perf record" and "perf report --dump-raw-trace" have been supported. However, the raw data that is dumped cannot be used without parsing.
This patch is to improve the "perf report" support for spe, and further process the data. Currently, support for the four events of llc-miss, tlb-miss, branch-miss, and remote-access is added.
Example usage:
$ ./perf record -c 1024 -e arm_spe_0/branch_filter=1,ts_enable=1,pct_enable=1,pa_enable=1,load_filter=1,jitter=1,store_filter=1,min_latency=0/ -o perf-armspe-dd.data dd if=/dev/zero of=/dev/null count=10000
$ ./perf report -i perf-armspe-dd.data --stdio
--------------------------------------------------------------------
...
# Samples: 23 of event 'llc-miss'
# Event count (approx.): 23
...
33.33% 33.33% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64
12.12% 12.12% dd [kernel.kallsyms] [k] perf_event_mmap
6.06% 6.06% dd [kernel.kallsyms] [k] copy_page
6.06% 6.06% dd ld-2.28.so [.] _dl_relocate_object
3.03% 3.03% dd [kernel.kallsyms] [k] change_protection_range
3.03% 3.03% dd [kernel.kallsyms] [k] filemap_map_pages
3.03% 3.03% dd [kernel.kallsyms] [k] free_pages_and_swap_cache
3.03% 3.03% dd [kernel.kallsyms] [k] generic_permission
3.03% 3.03% dd [kernel.kallsyms] [k] kmem_cache_alloc
3.03% 3.03% dd [kernel.kallsyms] [k] lookup_fast
3.03% 3.03% dd [kernel.kallsyms] [k] perf_event_exec
3.03% 3.03% dd [kernel.kallsyms] [k] radix_tree_next_chunk
3.03% 3.03% dd [kernel.kallsyms] [k] ring_buffer_record_is_on
3.03% 3.03% dd ld-2.28.so [.] _dl_lookup_symbol_x
3.03% 3.03% dd ld-2.28.so [.] _dl_start
3.03% 3.03% dd ld-2.28.so [.] dl_main
3.03% 3.03% dd ld-2.28.so [.] strcmp
3.03% 3.03% dd libc-2.28.so [.] _dl_addr
...
# Samples: 3 of event 'tlb-miss'
# Event count (approx.): 3
...
33.33% 33.33% dd [kernel.kallsyms] [k] filemap_map_pages
33.33% 33.33% dd ld-2.28.so [.] _dl_start
33.33% 33.33% dd ld-2.28.so [.] dl_main
...
# Samples: 20 of event 'branch-miss'
# Event count (approx.): 20
...
15.38% 15.38% dd [kernel.kallsyms] [k] __fput
7.69% 7.69% dd [kernel.kallsyms] [k] do_el0_ia_bp_hardening
7.69% 7.69% dd [kernel.kallsyms] [k] filemap_map_pages
7.69% 7.69% dd [kernel.kallsyms] [k] pagevec_lru_move_fn
7.69% 7.69% dd [kernel.kallsyms] [k] perf_event_mmap_output
7.69% 7.69% dd [kernel.kallsyms] [k] task_work_run
7.69% 7.69% dd [kernel.kallsyms] [k] unmap_single_vma
7.69% 7.69% dd libc-2.28.so [.] _IO_flush_all_lockp
7.69% 7.69% dd libc-2.28.so [.] __memcpy_generic
7.69% 7.69% dd libc-2.28.so [.] _dl_addr
7.69% 7.69% dd libc-2.28.so [.] msort_with_tmp.part.0
7.69% 7.69% dd libc-2.28.so [.] read_alias_file
...
# Samples: 5 of event 'remote-access'
# Event count (approx.): 5
...
27.78% 27.78% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64
16.67% 16.67% dd [kernel.kallsyms] [k] perf_event_mmap
5.56% 5.56% dd [kernel.kallsyms] [k] change_protection_range
5.56% 5.56% dd [kernel.kallsyms] [k] filemap_map_pages
5.56% 5.56% dd [kernel.kallsyms] [k] free_pages_and_swap_cache
5.56% 5.56% dd [kernel.kallsyms] [k] generic_permission
5.56% 5.56% dd [kernel.kallsyms] [k] lookup_fast
5.56% 5.56% dd [kernel.kallsyms] [k] perf_event_exec
5.56% 5.56% dd [kernel.kallsyms] [k] radix_tree_next_chunk
5.56% 5.56% dd ld-2.28.so [.] _dl_relocate_object
5.56% 5.56% dd ld-2.28.so [.] _dl_start
5.56% 5.56% dd ld-2.28.so [.] dl_main--------------------------------------------------------------------
After that, more analysis and processing of the raw data of spe will be done.
--- tools/perf/builtin-report.c | 5 + tools/perf/util/arm-spe-decoder/Build | 2 +- .../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++ .../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++ .../arm-spe-decoder/arm-spe-pkt-decoder.h | 2 + tools/perf/util/arm-spe.c | 745 +++++++++++++++++- tools/perf/util/auxtrace.c | 49 ++ tools/perf/util/auxtrace.h | 29 + tools/perf/util/session.h | 2 + 9 files changed, 1087 insertions(+), 38 deletions(-) create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
tools/perf/builtin-report.c +5 -0
@@ -1073,6 +1073,7 @@ int cmd_report(int argc, const char **argv) {struct perf_session *session;struct itrace_synth_opts itrace_synth_opts = { .set =0 , }; +struct arm_spe_synth_opts arm_spe_synth_opts = { .set =0 , };struct stat st;bool has_br_stack = false;int branch_mode =-1 ;
@@ -1237,6 +1238,9 @@ int cmd_report(int argc, const char **argv)OPT_CALLBACK_OPTARG (0 ,"itrace" , &itrace_synth_opts,NULL ,"opts" ,"Instruction Tracing options\n" ITRACE_HELP , itrace_parse_synth_opts), +OPT_CALLBACK_OPTARG (0 ,"spe" , &arm_spe_synth_opts,NULL ,"spe opts" , +"ARM SPE Tracing options" , + arm_spe_parse_synth_opts),OPT_BOOLEAN (0 ,"full-source-path" , &srcline_full_filename,"Show full source file name path for source lines" ), OPT_BOOLEAN(0, "show-ref-call-graph", &symbol_conf.show_ref_callgraph,
@@ -1348,6 +1352,7 @@ int cmd_report(int argc, const char **argv) } session->itrace_synth_opts = &itrace_synth_opts; + session->arm_spe_synth_opts = &arm_spe_synth_opts; report.session = session;
tools/perf/util/arm-spe-decoder/Build +1 -1
@@ -1 +1 @@ -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o arm-spe-decoder.o
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +225 -0
225 lines not on this page · raw message
tools/perf/util/arm-spe-decoder/arm-spe-decoder.h new file +66 -0
66 lines not on this page · raw message
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h +2 -0
@@ -15,6 +15,8 @@#define ARM_SPE_NEED_MORE_BYTES -1#define ARM_SPE_BAD_PACKET -2 +#define ARM_SPE_PKT_MAX_SZ 16 +enum arm_spe_pkt_type {ARM_SPE_BAD ,ARM_SPE_PAD ,
tools/perf/util/arm-spe.c +708 -37
819 lines not on this page · raw message
tools/perf/util/auxtrace.c +49 -0
55 lines not on this page · raw message
tools/perf/util/auxtrace.h +29 -0
@@ -111,6 +111,22 @@ struct itrace_synth_opts {int range_num; }; +/** +* struct arm_spe_synth_opts - ARM SPE tracing synthesis options. +* @set: indicates whether or not options have been set +* @llc_miss: whether to synthesize last level cache miss events +* @tlb_miss: whether to synthesize TLB miss events +* @branch_miss: whether to synthesize Branch miss events +* @remote_access: whether to synthesize Remote access events +*/ +struct arm_spe_synth_opts { +bool set; +bool llc_miss; +bool tlb_miss; +bool branch_miss; +bool remote_access; +}; +/** * struct auxtrace_index_entry - indexes a AUX area tracing event within a * perf.data file.
@@ -560,6 +576,10 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str,void itrace_synth_opts__set_default (struct itrace_synth_opts *synth_opts,bool no_sample); +int arm_spe_parse_synth_opts (const struct option *opt,const char *str, +int unset); +void arm_spe_synth_opts__set_default (struct arm_spe_synth_opts *synth_opts); +size_t perf_event__fprintf_auxtrace_error (union perf_event *event,FILE *fp);void perf_session__auxtrace_error_inc (struct perf_session *session,union perf_event *event);
@@ -662,6 +682,15 @@ int itrace_parse_synth_opts(const struct option *opt __maybe_unused,return -EINVAL ; } +static inline +int arm_spe_parse_synth_opts (const struct option *opt__maybe_unused , +const char *str__maybe_unused , +int unset__maybe_unused ) +{ +pr_err ("ARM SPE area tracing not supported\n" ); +return -EINVAL ; +} + static inline int auxtrace_parse_snapshot_options(struct auxtrace_record *itr __maybe_unused, struct record_opts *opts __maybe_unused,
tools/perf/util/session.h +2 -0
@@ -19,6 +19,7 @@ struct thread;struct auxtrace ;struct itrace_synth_opts ; +struct arm_spe_synth_opts ;struct perf_session {struct perf_header header;
@@ -26,6 +27,7 @@ struct perf_session {struct evlist *evlist;struct auxtrace *auxtrace;struct itrace_synth_opts *itrace_synth_opts; +struct arm_spe_synth_opts *arm_spe_synth_opts;struct list_head auxtrace_index;struct trace_event tevent;struct perf_record_time_conv time_conv;
Signature · 1 line
2.25.0
Re: [PATCH v2 2/7] perf tools: Add support for "report" for some spe events
On Thu, Jan 23, 2020 at 04:07:29PM +0000, James Clark wrote:
SNIP
Quoted 12 lines
tools/perf/util/auxtrace.h +4
-0@@ -111,6 +111,22 @@ struct itrace_synth_opts {int range_num; }; +/** +* struct arm_spe_synth_opts - ARM SPE tracing synthesis options. +* @set: indicates whether or not options have been set +* @llc_miss: whether to synthesize last level cache miss events
tools/perf/util/auxtrace.h struct itrace_synth_opts { · line 127
Quoted 10 lines
tools/perf/util/auxtrace.h +10
-0+* @tlb_miss: whether to synthesize TLB miss events +* @branch_miss: whether to synthesize Branch miss events +* @remote_access: whether to synthesize Remote access events +*/ +struct arm_spe_synth_opts { +bool set; +bool llc_miss; +bool tlb_miss; +bool branch_miss; +bool remote_access;
hum, why don't you add that to itrace_synth_opts instead? seems generic enough
I don't follow the code much, but I assume itrace_synth_opts is generic object, as it's already used by some of the s390 and x86 code.. also I don't like new pointer to synth_ops in perf_session
jirka
Quoted 23 lines
tools/perf/util/session.h +2
-0@@ -19,6 +19,7 @@ struct thread;struct auxtrace ;struct itrace_synth_opts ; +struct arm_spe_synth_opts ;struct perf_session {struct perf_header header;@@ -26,6 +27,7 @@ struct perf_session {struct evlist *evlist;struct auxtrace *auxtrace;struct itrace_synth_opts *itrace_synth_opts; +struct arm_spe_synth_opts *arm_spe_synth_opts;struct list_head auxtrace_index;struct trace_event tevent;struct perf_record_time_conv time_conv;Signature · 1 line
2.25.0
PATCH v2 3/7 perf report: Add --spe options for arm-spe
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
The previous patch added support in "perf report" for some arm-spe events(llc-miss, tlb-miss, branch-miss, remote_access). This patch adds their help instructions.
--- tools/perf/Documentation/perf-report.txt | 10 ++++++++++ 1 file changed, 10 insertions(+)
tools/perf/Documentation/perf-report.txt +10 -0
@@ -468,6 +468,16 @@ include::itrace.txt[] To disable decoding entirely, use --no-itrace. +--spe:: + Options for decoding arm-spe tracing data. The options are: + + l synthesize llc miss events + t synthesize tlb miss events + b synthesize branch miss events + r synthesize remote access events + + The default is all events i.e. the same as --spe=ltbr + --full-source-path:: Show the full path for source files for srcline output.
Signature · 1 line
2.25.0
PATCH v2 4/7 perf tools: Support "branch-misses:pp" on arm64
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
At the suggestion of James Clark, use spe to support the precise ip of some events. Currently its support event is:
branch-misses.
Example usage:
$ ./perf record -e branch-misses:pp dd if=/dev/zero of=/dev/null count=10000 (:p/pp/ppp is same for this case.)
$ ./perf report --stdio
("--stdio is not necessary")--------------------------------------------------------------------
...
# Samples: 14 of event 'branch-misses:pp' # Event count (approx.): 14
#
# Children Self Command Shared Object Symbol
# ........ ........ ....... ................. ..........................
#
14.29% 14.29% dd [kernel.kallsyms] [k] __arch_copy_from_user
14.29% 14.29% dd libc-2.28.so [.] _dl_addr
7.14% 7.14% dd [kernel.kallsyms] [k] __free_pages
7.14% 7.14% dd [kernel.kallsyms] [k] __pi_memcpy
7.14% 7.14% dd [kernel.kallsyms] [k] pagecache_get_page
7.14% 7.14% dd [kernel.kallsyms] [k] unmap_single_vma
7.14% 7.14% dd dd [.] 0x00000000000025ec
7.14% 7.14% dd ld-2.28.so [.] _dl_lookup_symbol_x
7.14% 7.14% dd ld-2.28.so [.] check_match
7.14% 7.14% dd libc-2.28.so [.] __mpn_rshift
7.14% 7.14% dd libc-2.28.so [.] _nl_intern_locale_data
7.14% 7.14% dd libc-2.28.so [.] read_alias_file...
--------------------------------------------------------------------
--- tools/perf/util/arm-spe.c | 41 +++++++++++++++++++++++++++++++++++++++ tools/perf/util/arm-spe.h | 3 +++ tools/perf/util/evlist.c | 2 ++ 3 files changed, 46 insertions(+)
tools/perf/util/arm-spe.c +41 -0
@@ -35,6 +35,19 @@#define MAX_TIMESTAMP (~0ULL) +#define SPE_ATTR_TS_ENABLE BIT(0) +#define SPE_ATTR_PA_ENABLE BIT(1) +#define SPE_ATTR_PCT_ENABLE BIT(2) +#define SPE_ATTR_JITTER BIT(16) +#define SPE_ATTR_BRANCH_FILTER BIT(32) +#define SPE_ATTR_LOAD_FILTER BIT(33) +#define SPE_ATTR_STORE_FILTER BIT(34) + +#define SPE_ATTR_EV_RETIRED BIT(1) +#define SPE_ATTR_EV_CACHE BIT(3) +#define SPE_ATTR_EV_TLB BIT(5) +#define SPE_ATTR_EV_BRANCH BIT(7) +struct arm_spe {struct auxtrace auxtrace;struct auxtrace_queues queues;
@@ -778,6 +791,15 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session) attr.sample_id_all = evsel->core.attr.sample_id_all; attr.read_format = evsel->core.attr.read_format; +/* If it is in the precise ip mode, there is no need to +* synthesize new events. */ +if (!strncmp (evsel->name,"branch-misses" ,13 )) { + spe->sample_branch_miss = true; + spe->branch_miss_id = evsel->core.id[0 ]; + +return 0 ; + } +/* create new id val to be a fixed offset from evsel id */ id = evsel->core.id[0 ] +1000000000 ;
@@ -899,3 +921,22 @@ int arm_spe_process_auxtrace_info(union perf_event *event,free (spe);return err; } + +void arm_spe_precise_ip_support (struct evlist *evlist,struct evsel *evsel) +{ +struct perf_pmu *pmu; + +/* Currently only supports precise_ip for branch-misses on arm64 */ +if (!strcmp (perf_env__arch (evlist->env),"arm64" ) + && evsel->core.attr.config ==PERF_COUNT_HW_BRANCH_MISSES + && evsel->core.attr.precise_ip) { + pmu =perf_pmu__find ("arm_spe_0" ); +if (pmu) { + evsel->pmu_name = pmu->name; + evsel->core.attr.type = pmu->type; + evsel->core.attr.config =SPE_ATTR_TS_ENABLE + |SPE_ATTR_BRANCH_FILTER ; + evsel->core.attr.config1 =SPE_ATTR_EV_BRANCH ; + } + } +}
tools/perf/util/arm-spe.h +3 -0
@@ -20,6 +20,8 @@ enum {union perf_event ;struct perf_session ;struct perf_pmu ; +struct evlist ; +struct evsel ;struct auxtrace_record *arm_spe_recording_init (int *err,struct perf_pmu *arm_spe_pmu);
@@ -28,4 +30,5 @@ int arm_spe_process_auxtrace_info(union perf_event *event, struct perf_session *session);struct perf_event_attr *arm_spe_pmu_default_config (struct perf_pmu *arm_spe_pmu); +void arm_spe_precise_ip_support (struct evlist *evlist,struct evsel *evsel);#endif
tools/perf/util/evlist.c +2 -0
@@ -9,6 +9,7 @@#include <errno.h> #include <inttypes.h> #include <poll.h> +#include "arm-spe.h" #include "cpumap.h" #include "util/mmap.h" #include "thread_map.h"
@@ -179,6 +180,7 @@ void perf_evlist__splice_list_tail(struct evlist *evlist,struct evsel *evsel, *temp;__evlist__for_each_entry_safe (list, temp, evsel) { +arm_spe_precise_ip_support (evlist, evsel);list_del_init (&evsel->core.node);evlist__add (evlist, evsel); }
Signature · 1 line
2.25.0
Re: [PATCH v2 4/7] perf tools: Support "branch-misses:pp" on arm64
On Thu, Jan 23, 2020 at 04:07:31PM +0000, James Clark wrote:
SNIP
Quoted 13 lines
tools/perf/util/evlist.c +1
-0@@ -9,6 +9,7 @@#include <errno.h> #include <inttypes.h> #include <poll.h> +#include "arm-spe.h" #include "cpumap.h" #include "util/mmap.h" #include "thread_map.h"
tools/perf/util/evlist.c void perf_evlist__splice_list_tail(struct evlist *evlist, · line 183
Quoted 4 lines
tools/perf/util/evlist.c +1
-0@@ -179,6 +180,7 @@ void perf_evlist__splice_list_tail(struct evlist *evlist,struct evsel *evsel, *temp;__evlist__for_each_entry_safe (list, temp, evsel) { +arm_spe_precise_ip_support (evlist, evsel);
this is 'splice' function, you can't configure precise in here
do you need this 'config thing' to be executed on arm only?
if yes, please add something like arch_evsel__config, make it weak for generic code and define it for arm
if no, just add the call at the end perf_evsel__config I guess
thanks,
jirka
Quoted 6 lines
tools/perf/util/evlist.c +0
-0list_del_init (&evsel->core.node);evlist__add (evlist, evsel); }Signature · 1 line
2.25.0
PATCH v2 5/7 perf tools: add perf_evlist__terminate() for terminate
In reply to James ClarkFrom: Wei Li <[email protected]>
In __cmd_record(), when receiving SIGINT(ctrl + c), a done flag will be set and the event list will be disabled by perf_evlist__disable() once.
While in auxtrace_record.read_finish(), the related events will be enabled again, if they are continuous, the recording seems to be endless.
Mark the evlist's state as terminated, preparing for the following fix.
--- tools/perf/builtin-record.c | 1 + tools/perf/util/evlist.c | 14 ++++++++++++++ tools/perf/util/evlist.h | 1 + tools/perf/util/evsel.h | 1 + 4 files changed, 17 insertions(+)
tools/perf/builtin-record.c +1 -0
@@ -1722,6 +1722,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)if (done && !disabled && !target__none (&opts->target)) {trigger_off (&auxtrace_snapshot_trigger);evlist__disable (rec->evlist); +evlist__terminate (rec->evlist); disabled = true; } }
tools/perf/util/evlist.c +14 -0
@@ -377,6 +377,20 @@ bool evsel__cpu_iter_skip(struct evsel *ev, int cpu)return true; } +void evlist__terminate (struct evlist *evlist) +{ +struct evsel *pos; + +evlist__for_each_entry (evlist, pos) { +if (pos->disabled || !perf_evsel__is_group_leader (pos) || !pos->core.fd) +continue ; +evsel__disable (pos); + pos->terminated = true; + } + + evlist->enabled = false; +} +void evlist__disable (struct evlist *evlist) {struct evsel *pos;
tools/perf/util/evlist.h +1 -0
@@ -206,6 +206,7 @@ void evlist__munmap(struct evlist *evlist);size_t evlist__mmap_size (unsigned long pages); +void evlist__terminate (struct evlist *evlist);void evlist__disable (struct evlist *evlist);void evlist__enable (struct evlist *evlist);void perf_evlist__toggle_enable (struct evlist *evlist);
tools/perf/util/evsel.h +1 -0
@@ -104,6 +104,7 @@ struct evsel {perf_evsel__sb_cb_t *cb;void *data; } side_band; +bool terminated; };struct perf_missing_features {
Signature · 1 line
2.25.0
Re: [PATCH v2 5/7] perf tools: add perf_evlist__terminate() for terminate
Quoted 51 lines · James Clark
From: Wei Li <[email protected]>
In
__cmd_record(), when receiving SIGINT(ctrl + c), a done flag will be set and the event list will be disabled byperf_evlist__disable()once.While in
auxtrace_record.read_finish(), the related events will be enabled again, if they are continuous, the recording seems to be endless.Mark the evlist's state as terminated, preparing for the following fix.
Signed-off-by Wei Li <[email protected]>Tested-by Qi Liu <[email protected]>Signed-off-by James Clark <[email protected]>Cc Will Deacon <[email protected]>Cc Mark Rutland <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Ingo Molnar <[email protected]>Cc Arnaldo Carvalho de Melo <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Jiri Olsa <[email protected]>Cc Tan Xiaojun <[email protected]>Cc Al Grant <[email protected]>Cc Namhyung Kim <[email protected]>--- tools/perf/builtin-record.c | 1 + tools/perf/util/evlist.c | 14 ++++++++++++++ tools/perf/util/evlist.h | 1 + tools/perf/util/evsel.h | 1 + 4 files changed, 17 insertions(+)tools/perf/builtin-record.c +1
-0@@ -1722,6 +1722,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)if (done && !disabled && !target__none (&opts->target)) {trigger_off (&auxtrace_snapshot_trigger);evlist__disable (rec->evlist); +evlist__terminate (rec->evlist); disabled = true; } }tools/perf/util/evlist.c +0
-0@@ -377,6 +377,20 @@ bool evsel__cpu_iter_skip(struct evsel *ev, int cpu)return true; }
tools/perf/util/evlist.c bool evsel__cpu_iter_skip(struct evsel *ev, int cpu) · line 389
Quoted 10 lines
tools/perf/util/evlist.c +10
-0+void evlist__terminate (struct evlist *evlist) +{ +struct evsel *pos; + +evlist__for_each_entry (evlist, pos) { +if (pos->disabled || !perf_evsel__is_group_leader (pos) || !pos->core.fd) +continue ; +evsel__disable (pos); + pos->terminated = true; + }
how is this different from evlist__disable? other than it does not follow the cpu affinity ;-) can't you just call evlist__disable and check later on evlist->enabled instead of evlist->terminated?
jirka
v3 cover perf tools: Add support for some spe events and precise ip
Hi Jirka,
Thanks for the feedback. I've made the following changes:
Removed the arm_spe_synth_opts struct and added the new types into itrace_synth_opts. I could re-use branch but the other ones are new. And the --trace argument documentation has been updated accordingly.
I've removed the processing of the evlist from perf_evlist__splice_list_tail and put it in a weak function "auxtrace__preprocess_evlist" that is only built on Arm.
For the 2 patches about the hang on termination, I have removed them because I haven't been able to reproduce it and everything is working ok for me. @Wei Li, are you able to post steps required to reproduce?
Thanks
James
Tan Xiaojun (4):
perf tools: Move arm-spe-pkt-decoder.h/c to the new dir perf tools: Add support for "report" for some spe events perf report: Add SPE options to --itrace argument perf tools: Support "branch-misses:pp" on arm64
tools/perf/Documentation/itrace.txt | 5 +-
tools/perf/arch/arm/util/auxtrace.c | 38 +
tools/perf/builtin-record.c | 5 +
tools/perf/util/Build | 2 +-
tools/perf/util/arm-spe-decoder/Build | 1 +
.../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++
.../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++
.../arm-spe-pkt-decoder.c | 0
.../arm-spe-pkt-decoder.h | 2 +
tools/perf/util/arm-spe.c | 756 +++++++++++++++++-
tools/perf/util/arm-spe.h | 3 +
tools/perf/util/auxtrace.c | 13 +
tools/perf/util/auxtrace.h | 14 +-
tools/perf/util/evlist.c | 1 +
tools/perf/util/evsel.h | 1 -
15 files changed, 1090 insertions(+), 42 deletions(-)
create mode 100644 tools/perf/util/arm-spe-decoder/Build
create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c
create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%)
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (96%)Signature · 1 line
2.17.1
PATCH v3 1/4 perf tools: Move arm-spe-pkt-decoder.h/c to the new dir
From: Tan Xiaojun <[email protected]>
Create a new arm-spe-decoder directory for subsequent extensions and move arm-spe-pkt-decoder.h/c to this directory. No code changes.
---
tools/perf/util/Build | 2 +-
tools/perf/util/arm-spe-decoder/Build | 1 +
tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c | 0
tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h | 0
tools/perf/util/arm-spe.c | 2 +-
5 files changed, 3 insertions(+), 2 deletions(-)
create mode 100644 tools/perf/util/arm-spe-decoder/Build
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%)
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (100%)tools/perf/util/Build +1 -1
@@ -104,7 +104,7 @@ perf-$(CONFIG_AUXTRACE) += intel-pt-decoder/ perf-$(CONFIG_AUXTRACE) += intel-pt.o perf-$(CONFIG_AUXTRACE) += intel-bts.o perf-$(CONFIG_AUXTRACE) += arm-spe.o -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-decoder/ perf-$(CONFIG_AUXTRACE) += s390-cpumsf.o ifdef CONFIG_LIBOPENCSD
tools/perf/util/arm-spe-decoder/Build new file +1 -0
@@ -0,0 +1 @@ +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.c renamed from tools/perf/util/arm-spe-pkt-decoder.c +0 -0
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h renamed from tools/perf/util/arm-spe-pkt-decoder.h +0 -0
tools/perf/util/arm-spe.c +1 -1
@@ -23,7 +23,7 @@#include "debug.h" #include "auxtrace.h" #include "arm-spe.h" -#include "arm-spe-pkt-decoder.h" +#include "arm-spe-decoder/arm-spe-pkt-decoder.h" struct arm_spe {struct auxtrace auxtrace;
Signature · 1 line
2.17.1
PATCH v3 2/4 perf tools: Add support for "report" for some spe events
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
After the commit ffd3d18c20b8 ("perf tools: Add ARM Statistical Profiling Extensions (SPE) support") is merged, "perf record" and "perf report --dump-raw-trace" have been supported. However, the raw data that is dumped cannot be used without parsing.
This patch is to improve the "perf report" support for spe, and further process the data. Currently, support for the four events of llc-miss, tlb-miss, branch-miss, and remote-access is added.
Example usage:
$ ./perf record -c 1024 -e arm_spe_0/branch_filter=1,ts_enable=1,pct_enable=1,pa_enable=1,load_filter=1,jitter=1,store_filter=1,min_latency=0/ -o perf-armspe-dd.data dd if=/dev/zero of=/dev/null count=10000
$ ./perf report -i perf-armspe-dd.data --stdio
--------------------------------------------------------------------
...
# Samples: 23 of event 'llc-miss'
# Event count (approx.): 23
...
33.33% 33.33% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64
12.12% 12.12% dd [kernel.kallsyms] [k] perf_event_mmap
6.06% 6.06% dd [kernel.kallsyms] [k] copy_page
6.06% 6.06% dd ld-2.28.so [.] _dl_relocate_object
3.03% 3.03% dd [kernel.kallsyms] [k] change_protection_range
3.03% 3.03% dd [kernel.kallsyms] [k] filemap_map_pages
3.03% 3.03% dd [kernel.kallsyms] [k] free_pages_and_swap_cache
3.03% 3.03% dd [kernel.kallsyms] [k] generic_permission
3.03% 3.03% dd [kernel.kallsyms] [k] kmem_cache_alloc
3.03% 3.03% dd [kernel.kallsyms] [k] lookup_fast
3.03% 3.03% dd [kernel.kallsyms] [k] perf_event_exec
3.03% 3.03% dd [kernel.kallsyms] [k] radix_tree_next_chunk
3.03% 3.03% dd [kernel.kallsyms] [k] ring_buffer_record_is_on
3.03% 3.03% dd ld-2.28.so [.] _dl_lookup_symbol_x
3.03% 3.03% dd ld-2.28.so [.] _dl_start
3.03% 3.03% dd ld-2.28.so [.] dl_main
3.03% 3.03% dd ld-2.28.so [.] strcmp
3.03% 3.03% dd libc-2.28.so [.] _dl_addr
...
# Samples: 3 of event 'tlb-miss'
# Event count (approx.): 3
...
33.33% 33.33% dd [kernel.kallsyms] [k] filemap_map_pages
33.33% 33.33% dd ld-2.28.so [.] _dl_start
33.33% 33.33% dd ld-2.28.so [.] dl_main
...
# Samples: 20 of event 'branch-miss'
# Event count (approx.): 20
...
15.38% 15.38% dd [kernel.kallsyms] [k] __fput
7.69% 7.69% dd [kernel.kallsyms] [k] do_el0_ia_bp_hardening
7.69% 7.69% dd [kernel.kallsyms] [k] filemap_map_pages
7.69% 7.69% dd [kernel.kallsyms] [k] pagevec_lru_move_fn
7.69% 7.69% dd [kernel.kallsyms] [k] perf_event_mmap_output
7.69% 7.69% dd [kernel.kallsyms] [k] task_work_run
7.69% 7.69% dd [kernel.kallsyms] [k] unmap_single_vma
7.69% 7.69% dd libc-2.28.so [.] _IO_flush_all_lockp
7.69% 7.69% dd libc-2.28.so [.] __memcpy_generic
7.69% 7.69% dd libc-2.28.so [.] _dl_addr
7.69% 7.69% dd libc-2.28.so [.] msort_with_tmp.part.0
7.69% 7.69% dd libc-2.28.so [.] read_alias_file
...
# Samples: 5 of event 'remote-access'
# Event count (approx.): 5
...
27.78% 27.78% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64
16.67% 16.67% dd [kernel.kallsyms] [k] perf_event_mmap
5.56% 5.56% dd [kernel.kallsyms] [k] change_protection_range
5.56% 5.56% dd [kernel.kallsyms] [k] filemap_map_pages
5.56% 5.56% dd [kernel.kallsyms] [k] free_pages_and_swap_cache
5.56% 5.56% dd [kernel.kallsyms] [k] generic_permission
5.56% 5.56% dd [kernel.kallsyms] [k] lookup_fast
5.56% 5.56% dd [kernel.kallsyms] [k] perf_event_exec
5.56% 5.56% dd [kernel.kallsyms] [k] radix_tree_next_chunk
5.56% 5.56% dd ld-2.28.so [.] _dl_relocate_object
5.56% 5.56% dd ld-2.28.so [.] _dl_start
5.56% 5.56% dd ld-2.28.so [.] dl_main--------------------------------------------------------------------
After that, more analysis and processing of the raw data of spe will be done.
--- tools/perf/util/arm-spe-decoder/Build | 2 +- .../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++ .../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++ .../arm-spe-decoder/arm-spe-pkt-decoder.h | 2 + tools/perf/util/arm-spe.c | 745 +++++++++++++++++- tools/perf/util/auxtrace.c | 13 + tools/perf/util/auxtrace.h | 8 +- 7 files changed, 1022 insertions(+), 39 deletions(-) create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
tools/perf/util/arm-spe-decoder/Build +1 -1
@@ -1 +1 @@ -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o arm-spe-decoder.o
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +225 -0
225 lines not on this page · raw message
tools/perf/util/arm-spe-decoder/arm-spe-decoder.h new file +66 -0
66 lines not on this page · raw message
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h +2 -0
@@ -15,6 +15,8 @@#define ARM_SPE_NEED_MORE_BYTES -1#define ARM_SPE_BAD_PACKET -2 +#define ARM_SPE_PKT_MAX_SZ 16 +enum arm_spe_pkt_type {ARM_SPE_BAD ,ARM_SPE_PAD ,
tools/perf/util/arm-spe.c +708 -37
819 lines not on this page · raw message
tools/perf/util/auxtrace.c +13 -0
@@ -1279,6 +1279,10 @@ void itrace_synth_opts__set_default(struct itrace_synth_opts *synth_opts, synth_opts->pwr_events = true; synth_opts->other_events = true; synth_opts->errors = true; + synth_opts->llc_miss = true; + synth_opts->tlb_miss = true; + synth_opts->remote_access = true; +if (no_sample) { synth_opts->period_type =PERF_ITRACE_PERIOD_INSTRUCTIONS ; synth_opts->period =1 ;
@@ -1431,6 +1435,15 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str, goto out_err; p = endptr;break ; +case 'm' : + synth_opts->llc_miss = true; +break ; +case 't' : + synth_opts->tlb_miss = true; +break ; +case 'a' : + synth_opts->remote_access = true; +break ;case ' ' :case ',' :break ;
tools/perf/util/auxtrace.h +7 -1
@@ -60,7 +60,7 @@ enum itrace_period_type {* @inject: indicates the event (not just the sample) must be fully synthesized * because 'perf inject' will write it out * @instructions: whether to synthesize 'instructions' events -* @branches: whether to synthesize 'branches' events +* @branches: whether to synthesize 'branches' events (branch misses only on Arm) * @transactions: whether to synthesize events for transactions * @ptwrites: whether to synthesize events for ptwrites * @pwr_events: whether to synthesize power events
@@ -74,6 +74,9 @@ enum itrace_period_type {* @callchain: add callchain to 'instructions' events * @thread_stack: feed branches to the thread_stack * @last_branch: add branch context to 'instruction' events +* @llc_miss: whether to synthesize last level cache miss events +* @tlb_miss: whether to synthesize TLB miss events +* @remote_access: whether to synthesize Remote access events * @callchain_sz: maximum callchain size * @last_branch_sz: branch context size * @period: 'instructions' events period
@@ -101,6 +104,9 @@ struct itrace_synth_opts {bool callchain;bool thread_stack;bool last_branch; +bool llc_miss; +bool tlb_miss; +bool remote_access;unsigned int callchain_sz;unsigned int last_branch_sz;unsigned long long period;
Signature · 1 line
2.17.1
PATCH v3 3/4 perf report: Add SPE options to --itrace argument
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
The previous patch added support in "perf report" for some arm-spe events(llc-miss, tlb-miss, branch-miss, remote_access). This patch adds their help instructions.
--- tools/perf/Documentation/itrace.txt | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
tools/perf/Documentation/itrace.txt +4 -1
@@ -1,5 +1,5 @@ i synthesize instructions events - b synthesize branches events + b synthesize branches events (branch misses on Arm) c synthesize branches events (calls only) r synthesize branches events (returns only) x synthesize transactions events
@@ -12,6 +12,9 @@ g synthesize a call chain (use with i or x) l synthesize last branch entries (use with i or x) s skip initial number of events + m synthesize LLC miss events + t synthesize TLB miss events + a synthesize remote access events The default is all events i.e. the same as --itrace=ibxwpe, except for perf script where it is --itrace=ce
Signature · 1 line
2.17.1
PATCH v3 4/4 perf tools: Support "branch-misses:pp" on arm64
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
At the suggestion of James Clark, use spe to support the precise ip of some events. Currently its support event is:
branch-misses.
Example usage:
$ ./perf record -e branch-misses:pp dd if=/dev/zero of=/dev/null count=10000 (:p/pp/ppp is same for this case.)
$ ./perf report --stdio
("--stdio is not necessary")--------------------------------------------------------------------
...
# Samples: 14 of event 'branch-misses:pp' # Event count (approx.): 14
#
# Children Self Command Shared Object Symbol
# ........ ........ ....... ................. ..........................
#
14.29% 14.29% dd [kernel.kallsyms] [k] __arch_copy_from_user
14.29% 14.29% dd libc-2.28.so [.] _dl_addr
7.14% 7.14% dd [kernel.kallsyms] [k] __free_pages
7.14% 7.14% dd [kernel.kallsyms] [k] __pi_memcpy
7.14% 7.14% dd [kernel.kallsyms] [k] pagecache_get_page
7.14% 7.14% dd [kernel.kallsyms] [k] unmap_single_vma
7.14% 7.14% dd dd [.] 0x00000000000025ec
7.14% 7.14% dd ld-2.28.so [.] _dl_lookup_symbol_x
7.14% 7.14% dd ld-2.28.so [.] check_match
7.14% 7.14% dd libc-2.28.so [.] __mpn_rshift
7.14% 7.14% dd libc-2.28.so [.] _nl_intern_locale_data
7.14% 7.14% dd libc-2.28.so [.] read_alias_file...
--------------------------------------------------------------------
--- tools/perf/arch/arm/util/auxtrace.c | 38 +++++++++++++++++++++++++++++ tools/perf/builtin-record.c | 5 ++++ tools/perf/util/arm-spe.c | 9 +++++++ tools/perf/util/arm-spe.h | 3 +++ tools/perf/util/auxtrace.h | 6 +++++ tools/perf/util/evlist.c | 1 + tools/perf/util/evsel.h | 1 - 7 files changed, 62 insertions(+), 1 deletion(-)
tools/perf/arch/arm/util/auxtrace.c +38 -0
@@ -10,11 +10,25 @@#include "../../util/auxtrace.h" #include "../../util/debug.h" +#include "../../util/env.h" #include "../../util/evlist.h" #include "../../util/pmu.h" #include "cs-etm.h" #include "arm-spe.h" +#define SPE_ATTR_TS_ENABLE BIT(0) +#define SPE_ATTR_PA_ENABLE BIT(1) +#define SPE_ATTR_PCT_ENABLE BIT(2) +#define SPE_ATTR_JITTER BIT(16) +#define SPE_ATTR_BRANCH_FILTER BIT(32) +#define SPE_ATTR_LOAD_FILTER BIT(33) +#define SPE_ATTR_STORE_FILTER BIT(34) + +#define SPE_ATTR_EV_RETIRED BIT(1) +#define SPE_ATTR_EV_CACHE BIT(3) +#define SPE_ATTR_EV_TLB BIT(5) +#define SPE_ATTR_EV_BRANCH BIT(7) +static struct perf_pmu **find_all_arm_spe_pmus (int *nr_spes,int *err) {struct perf_pmu **arm_spe_pmus =NULL ;
@@ -108,3 +122,27 @@ struct auxtrace_record *err =0 ;return NULL ; } + +void auxtrace__preprocess_evlist (struct evlist *evlist) +{ +struct evsel *evsel; +struct perf_pmu *pmu; + +evlist__for_each_entry (evlist, evsel) { +/* Currently only supports precise_ip for branch-misses on arm64 */ +if (!strcmp (perf_env__arch (evlist->env),"arm64" ) + && evsel->core.attr.config ==PERF_COUNT_HW_BRANCH_MISSES + && evsel->core.attr.precise_ip) + { + pmu =perf_pmu__find ("arm_spe_0" ); +if (pmu) { + evsel->pmu_name = pmu->name; + evsel->core.attr.type = pmu->type; + evsel->core.attr.config =SPE_ATTR_TS_ENABLE + |SPE_ATTR_BRANCH_FILTER ; + evsel->core.attr.config1 =SPE_ATTR_EV_BRANCH ; + evsel->core.attr.precise_ip =0 ; + } + } + } +}
tools/perf/builtin-record.c +5 -0
@@ -2451,6 +2451,11 @@ int cmd_record(int argc, const char **argv) argc =parse_options (argc, argv, record_options, record_usage,PARSE_OPT_STOP_AT_NON_OPTION ); + +if (auxtrace__preprocess_evlist) { +auxtrace__preprocess_evlist (rec->evlist); + } +if (quiet)perf_quiet_option ();
tools/perf/util/arm-spe.c +9 -0
@@ -778,6 +778,15 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session) attr.sample_id_all = evsel->core.attr.sample_id_all; attr.read_format = evsel->core.attr.read_format; +/* If it is in the precise ip mode, there is no need to +* synthesize new events. */ +if (!strncmp (evsel->name,"branch-misses" ,13 )) { + spe->sample_branch_miss = true; + spe->branch_miss_id = evsel->core.id[0 ]; + +return 0 ; + } +/* create new id val to be a fixed offset from evsel id */ id = evsel->core.id[0 ] +1000000000 ;
tools/perf/util/arm-spe.h +3 -0
@@ -20,6 +20,8 @@ enum {union perf_event ;struct perf_session ;struct perf_pmu ; +struct evlist ; +struct evsel ;struct auxtrace_record *arm_spe_recording_init (int *err,struct perf_pmu *arm_spe_pmu);
@@ -28,4 +30,5 @@ int arm_spe_process_auxtrace_info(union perf_event *event, struct perf_session *session);struct perf_event_attr *arm_spe_pmu_default_config (struct perf_pmu *arm_spe_pmu); +void arm_spe_precise_ip_support (struct evlist *evlist,struct evsel *evsel);#endif
tools/perf/util/auxtrace.h +6 -0
@@ -584,6 +584,7 @@ void auxtrace__dump_auxtrace_sample(struct perf_session *session,int auxtrace__flush_events (struct perf_session *session,struct perf_tool *tool);void auxtrace__free_events (struct perf_session *session);void auxtrace__free (struct perf_session *session); +void auxtrace__preprocess_evlist (struct evlist *evlist) __attribute__((weak));#define ITRACE_HELP \ " i: synthesize instructions events\n" \
@@ -728,6 +729,11 @@ void auxtrace__free(struct perf_session *session __maybe_unused) { } +static inline +void auxtrace__preprocess_evlist (struct evlist *evlist__maybe_unused ) +{ +} +static inline int auxtrace_index__write (int fd__maybe_unused ,struct list_head *head__maybe_unused )
tools/perf/util/evlist.c +1 -0
@@ -9,6 +9,7 @@#include <errno.h> #include <inttypes.h> #include <poll.h> +#include "arm-spe.h" #include "cpumap.h" #include "util/mmap.h" #include "thread_map.h"
tools/perf/util/evsel.h +0 -1
@@ -174,7 +174,6 @@ void perf_evsel__exit(struct evsel *evsel);void evsel__delete (struct evsel *evsel);struct callchain_param ; -void perf_evsel__config (struct evsel *evsel,struct record_opts *opts,struct callchain_param *callchain);
Signature · 1 line
2.17.1
Re: [PATCH v3 4/4] perf tools: Support "branch-misses:pp" on arm64
On Fri, Feb 07, 2020 at 03:21:42PM +0000, James Clark wrote:
SNIP
Quoted 32 lines
#define
ITRACE_HELP\
" i: synthesize instructions events\n" \tools/perf/util/auxtrace.h +5
-0@@ -728,6 +729,11 @@ void auxtrace__free(struct perf_session *session __maybe_unused) { } +static inline +void auxtrace__preprocess_evlist (struct evlist *evlist__maybe_unused ) +{ +} +static inline int auxtrace_index__write (int fd__maybe_unused ,struct list_head *head__maybe_unused )tools/perf/util/evlist.c +1
-0@@ -9,6 +9,7 @@#include <errno.h> #include <inttypes.h> #include <poll.h> +#include "arm-spe.h" #include "cpumap.h" #include "util/mmap.h" #include "thread_map.h" tools/perf/util/evsel.h +0
-0
tools/perf/util/evsel.h void perf_evsel__exit(struct evsel *evsel); · line 177
Quoted 4 lines
tools/perf/util/evsel.h +0
-1@@ -174,7 +174,6 @@ void perf_evsel__exit(struct evsel *evsel);void evsel__delete (struct evsel *evsel);struct callchain_param ; -
hum? ;-)
jirka
Quoted 6 lines
tools/perf/util/evsel.h +0
-0void perf_evsel__config (struct evsel *evsel,struct record_opts *opts,struct callchain_param *callchain);Signature · 1 line
2.17.1
v4 cover perf tools: Add support for some spe events and precise ip
Hi Jirka,
Oops. I've removed all the changes to evlist.c and evsel.h
James
Tan Xiaojun (4):
perf tools: Move arm-spe-pkt-decoder.h/c to the new dir perf tools: Add support for "report" for some spe events perf report: Add SPE options to --itrace argument perf tools: Support "branch-misses:pp" on arm64
tools/perf/Documentation/itrace.txt | 5 +-
tools/perf/arch/arm/util/auxtrace.c | 38 +
tools/perf/builtin-record.c | 5 +
tools/perf/util/Build | 2 +-
tools/perf/util/arm-spe-decoder/Build | 1 +
.../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++
.../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++
.../arm-spe-pkt-decoder.c | 0
.../arm-spe-pkt-decoder.h | 2 +
tools/perf/util/arm-spe.c | 756 +++++++++++++++++-
tools/perf/util/arm-spe.h | 3 +
tools/perf/util/auxtrace.c | 13 +
tools/perf/util/auxtrace.h | 14 +-
13 files changed, 1089 insertions(+), 41 deletions(-)
create mode 100644 tools/perf/util/arm-spe-decoder/Build
create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c
create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%)
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (96%)Signature · 1 line
2.17.1
PATCH v4 1/4 perf tools: Move arm-spe-pkt-decoder.h/c to the new dir
From: Tan Xiaojun <[email protected]>
Create a new arm-spe-decoder directory for subsequent extensions and move arm-spe-pkt-decoder.h/c to this directory. No code changes.
---
tools/perf/util/Build | 2 +-
tools/perf/util/arm-spe-decoder/Build | 1 +
tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c | 0
tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h | 0
tools/perf/util/arm-spe.c | 2 +-
5 files changed, 3 insertions(+), 2 deletions(-)
create mode 100644 tools/perf/util/arm-spe-decoder/Build
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%)
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (100%)tools/perf/util/Build +1 -1
@@ -104,7 +104,7 @@ perf-$(CONFIG_AUXTRACE) += intel-pt-decoder/ perf-$(CONFIG_AUXTRACE) += intel-pt.o perf-$(CONFIG_AUXTRACE) += intel-bts.o perf-$(CONFIG_AUXTRACE) += arm-spe.o -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-decoder/ perf-$(CONFIG_AUXTRACE) += s390-cpumsf.o ifdef CONFIG_LIBOPENCSD
tools/perf/util/arm-spe-decoder/Build new file +1 -0
@@ -0,0 +1 @@ +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.c renamed from tools/perf/util/arm-spe-pkt-decoder.c +0 -0
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h renamed from tools/perf/util/arm-spe-pkt-decoder.h +0 -0
tools/perf/util/arm-spe.c +1 -1
@@ -23,7 +23,7 @@#include "debug.h" #include "auxtrace.h" #include "arm-spe.h" -#include "arm-spe-pkt-decoder.h" +#include "arm-spe-decoder/arm-spe-pkt-decoder.h" struct arm_spe {struct auxtrace auxtrace;
Signature · 1 line
2.17.1
PATCH v4 2/4 perf tools: Add support for "report" for some spe events
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
After the commit ffd3d18c20b8 ("perf tools: Add ARM Statistical Profiling Extensions (SPE) support") is merged, "perf record" and "perf report --dump-raw-trace" have been supported. However, the raw data that is dumped cannot be used without parsing.
This patch is to improve the "perf report" support for spe, and further process the data. Currently, support for the four events of llc-miss, tlb-miss, branch-miss, and remote-access is added.
Example usage:
$ ./perf record -c 1024 -e arm_spe_0/branch_filter=1,ts_enable=1,pct_enable=1,pa_enable=1,load_filter=1,jitter=1,store_filter=1,min_latency=0/ -o perf-armspe-dd.data dd if=/dev/zero of=/dev/null count=10000
$ ./perf report -i perf-armspe-dd.data --stdio
--------------------------------------------------------------------
...
# Samples: 23 of event 'llc-miss'
# Event count (approx.): 23
...
33.33% 33.33% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64
12.12% 12.12% dd [kernel.kallsyms] [k] perf_event_mmap
6.06% 6.06% dd [kernel.kallsyms] [k] copy_page
6.06% 6.06% dd ld-2.28.so [.] _dl_relocate_object
3.03% 3.03% dd [kernel.kallsyms] [k] change_protection_range
3.03% 3.03% dd [kernel.kallsyms] [k] filemap_map_pages
3.03% 3.03% dd [kernel.kallsyms] [k] free_pages_and_swap_cache
3.03% 3.03% dd [kernel.kallsyms] [k] generic_permission
3.03% 3.03% dd [kernel.kallsyms] [k] kmem_cache_alloc
3.03% 3.03% dd [kernel.kallsyms] [k] lookup_fast
3.03% 3.03% dd [kernel.kallsyms] [k] perf_event_exec
3.03% 3.03% dd [kernel.kallsyms] [k] radix_tree_next_chunk
3.03% 3.03% dd [kernel.kallsyms] [k] ring_buffer_record_is_on
3.03% 3.03% dd ld-2.28.so [.] _dl_lookup_symbol_x
3.03% 3.03% dd ld-2.28.so [.] _dl_start
3.03% 3.03% dd ld-2.28.so [.] dl_main
3.03% 3.03% dd ld-2.28.so [.] strcmp
3.03% 3.03% dd libc-2.28.so [.] _dl_addr
...
# Samples: 3 of event 'tlb-miss'
# Event count (approx.): 3
...
33.33% 33.33% dd [kernel.kallsyms] [k] filemap_map_pages
33.33% 33.33% dd ld-2.28.so [.] _dl_start
33.33% 33.33% dd ld-2.28.so [.] dl_main
...
# Samples: 20 of event 'branch-miss'
# Event count (approx.): 20
...
15.38% 15.38% dd [kernel.kallsyms] [k] __fput
7.69% 7.69% dd [kernel.kallsyms] [k] do_el0_ia_bp_hardening
7.69% 7.69% dd [kernel.kallsyms] [k] filemap_map_pages
7.69% 7.69% dd [kernel.kallsyms] [k] pagevec_lru_move_fn
7.69% 7.69% dd [kernel.kallsyms] [k] perf_event_mmap_output
7.69% 7.69% dd [kernel.kallsyms] [k] task_work_run
7.69% 7.69% dd [kernel.kallsyms] [k] unmap_single_vma
7.69% 7.69% dd libc-2.28.so [.] _IO_flush_all_lockp
7.69% 7.69% dd libc-2.28.so [.] __memcpy_generic
7.69% 7.69% dd libc-2.28.so [.] _dl_addr
7.69% 7.69% dd libc-2.28.so [.] msort_with_tmp.part.0
7.69% 7.69% dd libc-2.28.so [.] read_alias_file
...
# Samples: 5 of event 'remote-access'
# Event count (approx.): 5
...
27.78% 27.78% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64
16.67% 16.67% dd [kernel.kallsyms] [k] perf_event_mmap
5.56% 5.56% dd [kernel.kallsyms] [k] change_protection_range
5.56% 5.56% dd [kernel.kallsyms] [k] filemap_map_pages
5.56% 5.56% dd [kernel.kallsyms] [k] free_pages_and_swap_cache
5.56% 5.56% dd [kernel.kallsyms] [k] generic_permission
5.56% 5.56% dd [kernel.kallsyms] [k] lookup_fast
5.56% 5.56% dd [kernel.kallsyms] [k] perf_event_exec
5.56% 5.56% dd [kernel.kallsyms] [k] radix_tree_next_chunk
5.56% 5.56% dd ld-2.28.so [.] _dl_relocate_object
5.56% 5.56% dd ld-2.28.so [.] _dl_start
5.56% 5.56% dd ld-2.28.so [.] dl_main--------------------------------------------------------------------
After that, more analysis and processing of the raw data of spe will be done.
--- tools/perf/util/arm-spe-decoder/Build | 2 +- .../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++ .../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++ .../arm-spe-decoder/arm-spe-pkt-decoder.h | 2 + tools/perf/util/arm-spe.c | 745 +++++++++++++++++- tools/perf/util/auxtrace.c | 13 + tools/perf/util/auxtrace.h | 8 +- 7 files changed, 1022 insertions(+), 39 deletions(-) create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
tools/perf/util/arm-spe-decoder/Build +1 -1
@@ -1 +1 @@ -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o arm-spe-decoder.o
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +225 -0
225 lines not on this page · raw message
tools/perf/util/arm-spe-decoder/arm-spe-decoder.h new file +66 -0
66 lines not on this page · raw message
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h +2 -0
@@ -15,6 +15,8 @@#define ARM_SPE_NEED_MORE_BYTES -1#define ARM_SPE_BAD_PACKET -2 +#define ARM_SPE_PKT_MAX_SZ 16 +enum arm_spe_pkt_type {ARM_SPE_BAD ,ARM_SPE_PAD ,
tools/perf/util/arm-spe.c +708 -37
819 lines not on this page · raw message
tools/perf/util/auxtrace.c +13 -0
@@ -1279,6 +1279,10 @@ void itrace_synth_opts__set_default(struct itrace_synth_opts *synth_opts, synth_opts->pwr_events = true; synth_opts->other_events = true; synth_opts->errors = true; + synth_opts->llc_miss = true; + synth_opts->tlb_miss = true; + synth_opts->remote_access = true; +if (no_sample) { synth_opts->period_type =PERF_ITRACE_PERIOD_INSTRUCTIONS ; synth_opts->period =1 ;
@@ -1431,6 +1435,15 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str, goto out_err; p = endptr;break ; +case 'm' : + synth_opts->llc_miss = true; +break ; +case 't' : + synth_opts->tlb_miss = true; +break ; +case 'a' : + synth_opts->remote_access = true; +break ;case ' ' :case ',' :break ;
tools/perf/util/auxtrace.h +7 -1
@@ -60,7 +60,7 @@ enum itrace_period_type {* @inject: indicates the event (not just the sample) must be fully synthesized * because 'perf inject' will write it out * @instructions: whether to synthesize 'instructions' events -* @branches: whether to synthesize 'branches' events +* @branches: whether to synthesize 'branches' events (branch misses only on Arm) * @transactions: whether to synthesize events for transactions * @ptwrites: whether to synthesize events for ptwrites * @pwr_events: whether to synthesize power events
@@ -74,6 +74,9 @@ enum itrace_period_type {* @callchain: add callchain to 'instructions' events * @thread_stack: feed branches to the thread_stack * @last_branch: add branch context to 'instruction' events +* @llc_miss: whether to synthesize last level cache miss events +* @tlb_miss: whether to synthesize TLB miss events +* @remote_access: whether to synthesize Remote access events * @callchain_sz: maximum callchain size * @last_branch_sz: branch context size * @period: 'instructions' events period
@@ -101,6 +104,9 @@ struct itrace_synth_opts {bool callchain;bool thread_stack;bool last_branch; +bool llc_miss; +bool tlb_miss; +bool remote_access;unsigned int callchain_sz;unsigned int last_branch_sz;unsigned long long period;
Signature · 1 line
2.17.1
Re: [PATCH v4 2/4] perf tools: Add support for "report" for some spe events
Quoted 83 lines · James Clark
From: Tan Xiaojun <[email protected]>
After the commit
ffd3d18c20b8("perf tools: Add ARM Statistical Profiling Extensions (SPE) support") is merged, "perf record" and "perf report --dump-raw-trace" have been supported. However, the raw data that is dumped cannot be used without parsing.This patch is to improve the "perf report" support for spe, and further process the data. Currently, support for the four events of llc-miss, tlb-miss, branch-miss, and remote-access is added.
Example usage:
$ ./perf record -c 1024 -e arm_spe_0/branch_filter=1,ts_enable=1,pct_enable=1,pa_enable=1,load_filter=1,jitter=1,store_filter=1,min_latency=0/ -o perf-armspe-dd.data dd if=/dev/zero of=/dev/null count=10000$ ./perf report -i perf-armspe-dd.data --stdio -------------------------------------------------------------------- ... # Samples: 23 of event 'llc-miss' # Event count (approx.): 23 ... 33.33% 33.33% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64 12.12% 12.12% dd [kernel.kallsyms] [k] perf_event_mmap 6.06% 6.06% dd [kernel.kallsyms] [k] copy_page 6.06% 6.06% dd ld-2.28.so [.] _dl_relocate_object 3.03% 3.03% dd [kernel.kallsyms] [k] change_protection_range 3.03% 3.03% dd [kernel.kallsyms] [k] filemap_map_pages 3.03% 3.03% dd [kernel.kallsyms] [k] free_pages_and_swap_cache 3.03% 3.03% dd [kernel.kallsyms] [k] generic_permission 3.03% 3.03% dd [kernel.kallsyms] [k] kmem_cache_alloc 3.03% 3.03% dd [kernel.kallsyms] [k] lookup_fast 3.03% 3.03% dd [kernel.kallsyms] [k] perf_event_exec 3.03% 3.03% dd [kernel.kallsyms] [k] radix_tree_next_chunk 3.03% 3.03% dd [kernel.kallsyms] [k] ring_buffer_record_is_on 3.03% 3.03% dd ld-2.28.so [.] _dl_lookup_symbol_x 3.03% 3.03% dd ld-2.28.so [.] _dl_start 3.03% 3.03% dd ld-2.28.so [.] dl_main 3.03% 3.03% dd ld-2.28.so [.] strcmp 3.03% 3.03% dd libc-2.28.so [.] _dl_addr ... # Samples: 3 of event 'tlb-miss' # Event count (approx.): 3 ... 33.33% 33.33% dd [kernel.kallsyms] [k] filemap_map_pages 33.33% 33.33% dd ld-2.28.so [.] _dl_start 33.33% 33.33% dd ld-2.28.so [.] dl_main ... # Samples: 20 of event 'branch-miss' # Event count (approx.): 20 ... 15.38% 15.38% dd [kernel.kallsyms] [k] __fput 7.69% 7.69% dd [kernel.kallsyms] [k] do_el0_ia_bp_hardening 7.69% 7.69% dd [kernel.kallsyms] [k] filemap_map_pages 7.69% 7.69% dd [kernel.kallsyms] [k] pagevec_lru_move_fn 7.69% 7.69% dd [kernel.kallsyms] [k] perf_event_mmap_output 7.69% 7.69% dd [kernel.kallsyms] [k] task_work_run 7.69% 7.69% dd [kernel.kallsyms] [k] unmap_single_vma 7.69% 7.69% dd libc-2.28.so [.] _IO_flush_all_lockp 7.69% 7.69% dd libc-2.28.so [.] __memcpy_generic 7.69% 7.69% dd libc-2.28.so [.] _dl_addr 7.69% 7.69% dd libc-2.28.so [.] msort_with_tmp.part.0 7.69% 7.69% dd libc-2.28.so [.] read_alias_file ... # Samples: 5 of event 'remote-access' # Event count (approx.): 5 ... 27.78% 27.78% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64 16.67% 16.67% dd [kernel.kallsyms] [k] perf_event_mmap 5.56% 5.56% dd [kernel.kallsyms] [k] change_protection_range 5.56% 5.56% dd [kernel.kallsyms] [k] filemap_map_pages 5.56% 5.56% dd [kernel.kallsyms] [k] free_pages_and_swap_cache 5.56% 5.56% dd [kernel.kallsyms] [k] generic_permission 5.56% 5.56% dd [kernel.kallsyms] [k] lookup_fast 5.56% 5.56% dd [kernel.kallsyms] [k] perf_event_exec 5.56% 5.56% dd [kernel.kallsyms] [k] radix_tree_next_chunk 5.56% 5.56% dd ld-2.28.so [.] _dl_relocate_object 5.56% 5.56% dd ld-2.28.so [.] _dl_start 5.56% 5.56% dd ld-2.28.so [.] dl_main--------------------------------------------------------------------
After that, more analysis and processing of the raw data of spe will be done.
Quoted 13 lines
Signed-off-by Tan Xiaojun <[email protected]>Tested-by Qi Liu <[email protected]>Signed-off-by James Clark <[email protected]>Cc Will Deacon <[email protected]>Cc Mark Rutland <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Ingo Molnar <[email protected]>Cc Arnaldo Carvalho de Melo <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Jiri Olsa <[email protected]>Cc Tan Xiaojun <[email protected]>Cc Al Grant <[email protected]>Cc Namhyung Kim <[email protected]>
For auxtrace:
Quoted 1,230 lines
--- tools/perf/util/arm-spe-decoder/Build | 2 +- .../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++ .../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++ .../arm-spe-decoder/arm-spe-pkt-decoder.h | 2 + tools/perf/util/arm-spe.c | 745 +++++++++++++++++- tools/perf/util/auxtrace.c | 13 + tools/perf/util/auxtrace.h | 8 +- 7 files changed, 1022 insertions(+), 39 deletions(-) create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.htools/perf/util/arm-spe-decoder/Build +1
-1@@ -1 +1 @@ -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o arm-spe-decoder.otools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +225
-0@@ -0,0 +1,225 @@ +// SPDX-License-Identifier: GPL-2.0 +/* +* arm_spe_decoder.c: ARM SPE support +*/ + +#ifndef _GNU_SOURCE +#define _GNU_SOURCE +#endif +#include <stdlib.h> +#include <stdbool.h> +#include <string.h> +#include <errno.h> +#include <stdint.h> +#include <inttypes.h> +#include <linux/compiler.h> +#include <linux/zalloc.h> + +#include "../util.h" +#include "../debug.h" +#include "../auxtrace.h" + +#include "arm-spe-pkt-decoder.h" +#include "arm-spe-decoder.h" + +#ifndef BIT +#define BIT (n) (1UL << (n)) +#endif + +struct arm_spe_decoder { +int (*get_trace)(struct arm_spe_buffer *buffer,void *data); +void *data; +struct arm_spe_state state; +const unsigned char *buf; +size_t len; +uint64_t pos; +struct arm_spe_pkt packet; +int pkt_step; +int pkt_len; +int last_packet_type; + +uint64_t last_ip; +uint64_t ip; +uint64_t timestamp; +uint64_t sample_timestamp; +const unsigned char *next_buf; +size_t next_len; +unsigned char temp_buf[ARM_SPE_PKT_MAX_SZ ]; +}; + +static uint64_t arm_spe_calc_ip (uint64_t payload) +{ +uint64_t ip = (payload & ~(0xffULL <<56 )); + +/* fill high 8 bits for kernel virtual address */ +/* In Armv8 Architecture Reference Manual: Xn[55] determines +* whether the address lies in the upper or lower address range +* for the purpose of determining whether address tagging is +* used */ +if (ip &BIT (55 )) + ip |= (uint64_t )(0xffULL <<56 ); + +return ip; +} + +struct arm_spe_decoder *arm_spe_decoder_new (struct arm_spe_params *params) +{ +struct arm_spe_decoder *decoder; + +if (!params->get_trace) +return NULL ; + + decoder =zalloc (sizeof (struct arm_spe_decoder )); +if (!decoder) +return NULL ; + + decoder->get_trace = params->get_trace; + decoder->data = params->data; + +return decoder; +} + +void arm_spe_decoder_free (struct arm_spe_decoder *decoder) +{ +free (decoder); +} + +static int arm_spe_bad_packet (struct arm_spe_decoder *decoder) +{ + decoder->pkt_len =1 ; + decoder->pkt_step =1 ; +pr_debug ("ERROR: Bad packet\n" ); + +return -EBADMSG ; +} + + +static int arm_spe_get_data (struct arm_spe_decoder *decoder) +{ +struct arm_spe_buffer buffer = { .buf =0 , }; +int ret; + + decoder->pkt_step =0 ; + +pr_debug ("Getting more data\n" ); + ret = decoder->get_trace (&buffer, decoder->data); +if (ret) +return ret; + + decoder->buf = buffer.buf; + decoder->len = buffer.len; +if (!decoder->len) { +pr_debug ("No more data\n" ); +return -ENODATA ; + } + +return 0 ; +} + +static int arm_spe_get_next_data (struct arm_spe_decoder *decoder) +{ +return arm_spe_get_data (decoder); +} + +static int arm_spe_get_next_packet (struct arm_spe_decoder *decoder) +{ +int ret; + + decoder->last_packet_type = decoder->packet.type; + +do { + decoder->pos += decoder->pkt_step; + decoder->buf += decoder->pkt_step; + decoder->len -= decoder->pkt_step; + + +if (!decoder->len) { + ret =arm_spe_get_next_data (decoder); +if (ret) +return ret; + } + + ret =arm_spe_get_packet (decoder->buf, decoder->len, + &decoder->packet); +if (ret <=0 ) +return arm_spe_bad_packet (decoder); + + decoder->pkt_len = ret; + decoder->pkt_step = ret; + }while (decoder->packet.type ==ARM_SPE_PAD ); + +return 0 ; +} + +static int arm_spe_walk_trace (struct arm_spe_decoder *decoder) +{ +int err; +int idx; +uint64_t payload; + +while (1 ) { + err =arm_spe_get_next_packet (decoder); +if (err) +return err; + + idx = decoder->packet.index; + payload = decoder->packet.payload; + +switch (decoder->packet.type) { +case ARM_SPE_TIMESTAMP : + decoder->sample_timestamp = payload; +return 0 ; +case ARM_SPE_END : + decoder->sample_timestamp =0 ; +return 0 ; +case ARM_SPE_ADDRESS : + decoder->ip =arm_spe_calc_ip (payload); +if (idx ==0 ) + decoder->state.from_ip = decoder->ip; +else if (idx ==1 ) + decoder->state.to_ip = decoder->ip; +break ; +case ARM_SPE_COUNTER : +break ; +case ARM_SPE_CONTEXT : +break ; +case ARM_SPE_OP_TYPE : +break ; +case ARM_SPE_EVENTS : +if (payload &BIT (EV_TLB_REFILL )) + decoder->state.type |=ARM_SPE_TLB_MISS ; +if (payload &BIT (EV_MISPRED )) + decoder->state.type |=ARM_SPE_BRANCH_MISS ; +if (idx >1 && (payload &BIT (EV_LLC_REFILL ))) + decoder->state.type |=ARM_SPE_LLC_MISS ; +if (idx >1 && (payload &BIT (EV_REMOTE_ACCESS ))) + decoder->state.type |=ARM_SPE_REMOTE_ACCESS ; + +break ; +case ARM_SPE_DATA_SOURCE : +break ; +case ARM_SPE_BAD : +break ; +case ARM_SPE_PAD : +break ; +default : +pr_err ("Get Packet Error!\n" ); +return -ENOSYS ; + } + } +} + +const struct arm_spe_state *arm_spe_decode (struct arm_spe_decoder *decoder) +{ +int err; + + decoder->state.type =0 ; + + err =arm_spe_walk_trace (decoder); +if (err) + decoder->state.err = err; + + decoder->state.timestamp = decoder->sample_timestamp; + +return &decoder->state; +}tools/perf/util/arm-spe-decoder/arm-spe-decoder.h new file +66
-0@@ -0,0 +1,66 @@ +// SPDX-License-Identifier: GPL-2.0 +/* +* arm_spe_decoder.c: ARM SPE support +*/ + +#ifndef INCLUDE__ARM_SPE_DECODER_H__ +#define INCLUDE__ARM_SPE_DECODER_H__ + +#include <stdint.h> +#include <stddef.h> +#include <stdbool.h> + +enum arm_spe_events { +EV_EXCEPTION_GEN , +EV_RETIRED , +EV_L1D_ACCESS , +EV_L1D_REFILL , +EV_TLB_ACCESS , +EV_TLB_REFILL , +EV_NOT_TAKEN , +EV_MISPRED , +EV_LLC_ACCESS , +EV_LLC_REFILL , +EV_REMOTE_ACCESS , +}; + +enum arm_spe_sample_type { +ARM_SPE_LLC_MISS =1 <<0 , +ARM_SPE_TLB_MISS =1 <<1 , +ARM_SPE_BRANCH_MISS =1 <<2 , +ARM_SPE_REMOTE_ACCESS =1 <<3 , +ARM_SPE_EX_STOP =1 <<6 , +}; + +struct arm_spe_state { +enum arm_spe_sample_type type; +int err; +uint64_t from_ip; +uint64_t to_ip; +uint64_t timestamp; +}; + +struct arm_spe_insn ; + +struct arm_spe_buffer { +const unsigned char *buf; +size_t len; +u64 offset; +bool consecutive; +uint64_t ref_timestamp; +uint64_t trace_nr; +}; + +struct arm_spe_params { +int (*get_trace)(struct arm_spe_buffer *buffer,void *data); +void *data; +}; + +struct arm_spe_decoder ; + +struct arm_spe_decoder *arm_spe_decoder_new (struct arm_spe_params *params); +void arm_spe_decoder_free (struct arm_spe_decoder *decoder); + +const struct arm_spe_state *arm_spe_decode (struct arm_spe_decoder *decoder); + +#endif tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h +2
-0@@ -15,6 +15,8 @@#define ARM_SPE_NEED_MORE_BYTES -1#define ARM_SPE_BAD_PACKET -2 +#define ARM_SPE_PKT_MAX_SZ 16 +enum arm_spe_pkt_type {ARM_SPE_BAD ,ARM_SPE_PAD ,tools/perf/util/arm-spe.c +708
-37@@ -16,34 +16,68 @@#include <linux/log2.h> #include <linux/zalloc.h> +#include "auxtrace.h" #include "color.h" +#include "debug.h" #include "evsel.h" +#include "evlist.h" #include "machine.h" #include "session.h" -#include "debug.h" -#include "auxtrace.h" +#include "symbol.h" +#include "thread.h" +#include "thread-stack.h" +#include "tool.h" +#include "util/synthetic-events.h" +#include "arm-spe.h" +#include "arm-spe-decoder/arm-spe-decoder.h" #include "arm-spe-decoder/arm-spe-pkt-decoder.h" +#define MAX_TIMESTAMP (~0ULL) +struct arm_spe {struct auxtrace auxtrace;struct auxtrace_queues queues;struct auxtrace_heap heap; +struct itrace_synth_opts synth_opts;u32 auxtrace_type;struct perf_session *session;struct machine *machine;u32 pmu_type; + +u8 timeless_decoding; +u8 data_queued; + +u8 sample_llc_miss; +u8 sample_tlb_miss; +u8 sample_branch_miss; +u8 sample_remote_access; +u64 llc_miss_id; +u64 tlb_miss_id; +u64 branch_miss_id; +u64 remote_access_id; +u64 kernel_start; + +unsigned long num_events; };struct arm_spe_queue { -struct arm_spe *spe; -unsigned int queue_nr; -struct auxtrace_buffer *buffer; -bool on_heap; -bool done; -pid_t pid; -pid_t tid; -int cpu; +struct arm_spe *spe; +unsigned int queue_nr; +struct auxtrace_buffer *buffer; +struct auxtrace_buffer *old_buffer; +union perf_event *event_buf; +bool on_heap; +bool done; +pid_t pid; +pid_t tid; +int cpu; +void *decoder; +const struct arm_spe_state *state; +u64 time; +u64 timestamp; +struct thread *thread; +bool have_sample; }; static void arm_spe_dump(struct arm_spe *spe __maybe_unused,@@ -92,44 +126,494 @@ static void arm_spe_dump_event(struct arm_spe *spe, unsigned char *buf,arm_spe_dump (spe, buf, len); } -static int arm_spe_process_event (struct perf_session *session__maybe_unused , -union perf_event *event__maybe_unused , -struct perf_sample *sample__maybe_unused , -struct perf_tool *tool__maybe_unused ) +static int arm_spe_get_trace (struct arm_spe_buffer *b,void *data) +{ +struct arm_spe_queue *speq = data; +struct auxtrace_buffer *buffer = speq->buffer; +struct auxtrace_buffer *old_buffer = speq->old_buffer; +struct auxtrace_queue *queue; + + queue = &speq->spe->queues.queue_array[speq->queue_nr]; + + buffer =auxtrace_buffer__next (queue, buffer); +/* If no more data, drop the previous auxtrace_buffer and return */ +if (!buffer) { +if (old_buffer) +auxtrace_buffer__drop_data (old_buffer); + b->len =0 ; +return 0 ; + } + + speq->buffer = buffer; + +/* If the aux_buffer doesn't have data associated, try to load it */ +if (!buffer->data) { +/* get the file desc associated with the perf data file */ +int fd =perf_data__fd (speq->spe->session->data); + + buffer->data =auxtrace_buffer__get_data (buffer, fd); +if (!buffer->data) +return -ENOMEM ; + } + +if (buffer->use_data) { + b->len = buffer->use_size; + b->buf = buffer->use_data; + }else { + b->len = buffer->size; + b->buf = buffer->data; + } + + b->ref_timestamp = buffer->reference; + +if (b->len) { +if (old_buffer) +auxtrace_buffer__drop_data (old_buffer); + speq->old_buffer = buffer; + }else { +auxtrace_buffer__drop_data (buffer); +return arm_spe_get_trace (b, data); + } + +return 0 ; +} + +static struct arm_spe_queue *arm_spe__alloc_queue (struct arm_spe *spe, +unsigned int queue_nr) +{ +struct arm_spe_params params = { .get_trace =0 , }; +struct arm_spe_queue *speq; + + speq =zalloc (sizeof (*speq)); +if (!speq) +return NULL ; + + speq->event_buf =malloc (PERF_SAMPLE_MAX_SIZE ); +if (!speq->event_buf) + goto out_free; + + speq->spe = spe; + speq->queue_nr = queue_nr; + speq->pid =-1 ; + speq->tid =-1 ; + speq->cpu =-1 ; + +/* params set */ + params.get_trace = arm_spe_get_trace; + params.data = speq; + +/* create new decoder */ + speq->decoder =arm_spe_decoder_new (¶ms); +if (!speq->decoder) + goto out_free; + +return speq; + +out_free: +zfree (&speq->event_buf); +free (speq); + +return NULL ; +} + +static inline u8 arm_spe_cpumode (struct arm_spe *spe,uint64_t ip) +{ +return ip >= spe->kernel_start ? +PERF_RECORD_MISC_KERNEL : +PERF_RECORD_MISC_USER ; +} + +static void arm_spe_prep_sample (struct arm_spe *spe, +struct arm_spe_queue *speq, +union perf_event *event, +struct perf_sample *sample) +{ +if (!spe->timeless_decoding) + sample->time = speq->timestamp; + + sample->ip = speq->state->from_ip; + sample->cpumode =arm_spe_cpumode (spe, sample->ip); + sample->pid = speq->pid; + sample->tid = speq->tid; + sample->addr = speq->state->to_ip; + sample->period =1 ; + sample->cpu = speq->cpu; + + event->sample.header.type =PERF_RECORD_SAMPLE ; + event->sample.header.misc = sample->cpumode; + event->sample.header.size =sizeof (struct perf_event_header ); +} + +static inline int +arm_spe_deliver_synth_event (struct arm_spe *spe, +struct arm_spe_queue *speq__maybe_unused , +union perf_event *event, +struct perf_sample *sample) +{ +int ret; + + ret =perf_session__deliver_synth_event (spe->session, event, sample); +if (ret) +pr_err ("ARM SPE: failed to deliver event, error %d\n" , ret); + +return ret; +} + +static int +arm_spe_synth_spe_events_sample (struct arm_spe_queue *speq, +u64 spe_events_id) +{ +struct arm_spe *spe = speq->spe; +union perf_event *event = speq->event_buf; +struct perf_sample sample = { .ip =0 , }; + +arm_spe_prep_sample (spe, speq, event, &sample); + + sample.id = spe_events_id; + sample.stream_id = spe_events_id; + +return arm_spe_deliver_synth_event (spe, speq, event, &sample); +} + +static int arm_spe_sample (struct arm_spe_queue *speq) +{ +const struct arm_spe_state *state = speq->state; +struct arm_spe *spe = speq->spe; +int err; + +if (!speq->have_sample) +return 0 ; + + speq->have_sample = false; + +if (spe->sample_llc_miss && (state->type &ARM_SPE_LLC_MISS )) { + err =arm_spe_synth_spe_events_sample (speq, spe->llc_miss_id); +if (err) +return err; + } + +if (spe->sample_tlb_miss && (state->type &ARM_SPE_TLB_MISS )) { + err =arm_spe_synth_spe_events_sample (speq, spe->tlb_miss_id); +if (err) +return err; + } + +if (spe->sample_branch_miss && (state->type &ARM_SPE_BRANCH_MISS )) { + err =arm_spe_synth_spe_events_sample (speq, + spe->branch_miss_id); +if (err) +return err; + } + +if (spe->sample_remote_access && (state->type &ARM_SPE_REMOTE_ACCESS )) { + err =arm_spe_synth_spe_events_sample (speq, spe->remote_access_id); +if (err) +return err; + } + +return 0 ; +} + +static int arm_spe_run_decoder (struct arm_spe_queue *speq,u64 *timestamp) +{ +const struct arm_spe_state *state = speq->state; +struct arm_spe *spe = speq->spe; +int err; + +if (!spe->kernel_start) + spe->kernel_start =machine__kernel_start (spe->machine); + +while (1 ) { + err =arm_spe_sample (speq); +if (err) +return err; + + state =arm_spe_decode (speq->decoder); +if (state->err) { +if (state->err == -ENODATA ) { +pr_debug ("No data or all data has been processed.\n" ); +return 1 ; + } +continue ; + } + + speq->state = state; + speq->have_sample = true; + +if (!spe->timeless_decoding && speq->timestamp >= *timestamp) { + *timestamp = speq->timestamp; +return 0 ; + } + } + +return 0 ; +} + +static int arm_spe__setup_queue (struct arm_spe *spe, +struct auxtrace_queue *queue, +unsigned int queue_nr) +{ +struct arm_spe_queue *speq = queue->priv; + +if (list_empty (&queue->head) || speq) +return 0 ; + + speq =arm_spe__alloc_queue (spe, queue_nr); + +if (!speq) +return -ENOMEM ; + + queue->priv = speq; + +if (queue->cpu !=-1 ) + speq->cpu = queue->cpu; + +if (!speq->on_heap) { +const struct arm_spe_state *state; +int ret; + +if (spe->timeless_decoding) +return 0 ; + +retry: + state =arm_spe_decode (speq->decoder); +if (state->err) { +if (state->err == -ENODATA ) { +pr_debug ("queue %u has no timestamp\n" , + queue_nr); +return 0 ; + } + goto retry; + } + + speq->timestamp = state->timestamp; + speq->state = state; + speq->have_sample = true; + ret =auxtrace_heap__add (&spe->heap, queue_nr, speq->timestamp); +if (ret) +return ret; + speq->on_heap = true; + } + +return 0 ; +} + +static int arm_spe__setup_queues (struct arm_spe *spe) { +unsigned int i; +int ret; + +for (i =0 ; i < spe->queues.nr_queues; i++) { + ret =arm_spe__setup_queue (spe, &spe->queues.queue_array[i], i); +if (ret) +return ret; + } + +return 0 ; +} + +static int arm_spe__update_queues (struct arm_spe *spe) +{ +if (spe->queues.new_data) { + spe->queues.new_data = false; +return arm_spe__setup_queues (spe); + } +return 0 ; } +static bool arm_spe__is_timeless_decoding (struct arm_spe *spe) +{ +struct evsel *evsel; +struct evlist *evlist = spe->session->evlist; +bool timeless_decoding = true; + +/* +* Circle through the list of event and complain if we find one +* with the time bit set. +*/ +evlist__for_each_entry (evlist, evsel) { +if ((evsel->core.attr.sample_type &PERF_SAMPLE_TIME )) + timeless_decoding = false; + } + +return timeless_decoding; +} + +static void arm_spe_set_pid_tid_cpu (struct arm_spe *spe, +struct auxtrace_queue *queue) +{ +struct arm_spe_queue *speq = queue->priv; +pid_t tid; + + tid =machine__get_current_tid (spe->machine, speq->cpu); +if (tid !=-1 ) { + speq->tid = tid; +thread__zput (speq->thread); + }else + speq->tid = queue->tid; + +if ((!speq->thread) && (speq->tid !=-1 )) { + speq->thread =machine__find_thread (spe->machine,-1 , + speq->tid); + } + +if (speq->thread) { + speq->pid = speq->thread->pid_; +if (queue->cpu ==-1 ) + speq->cpu = speq->thread->cpu; + } +} + +static int arm_spe_process_queues (struct arm_spe *spe,u64 timestamp) +{ +unsigned int queue_nr; +u64 ts; +int ret; + +while (1 ) { +struct auxtrace_queue *queue; +struct arm_spe_queue *speq; + +if (!spe->heap.heap_cnt) +return 0 ; + +if (spe->heap.heap_array[0 ].ordinal >= timestamp) +return 0 ; + + queue_nr = spe->heap.heap_array[0 ].queue_nr; + queue = &spe->queues.queue_array[queue_nr]; + speq = queue->priv; + +auxtrace_heap__pop (&spe->heap); + +if (spe->heap.heap_cnt) { + ts = spe->heap.heap_array[0 ].ordinal +1 ; +if (ts > timestamp) + ts = timestamp; + }else { + ts = timestamp; + } + +arm_spe_set_pid_tid_cpu (spe, queue); + + ret =arm_spe_run_decoder (speq, &ts); +if (ret <0 ) { +auxtrace_heap__add (&spe->heap, queue_nr, ts); +return ret; + } + +if (!ret) { + ret =auxtrace_heap__add (&spe->heap, queue_nr, ts); +if (ret <0 ) +return ret; + }else { + speq->on_heap = false; + } + } + +return 0 ; +} + +static int arm_spe_process_timeless_queues (struct arm_spe *spe,pid_t tid, +u64 time_) +{ +struct auxtrace_queues *queues = &spe->queues; +unsigned int i; +u64 ts =0 ; + +for (i =0 ; i < queues->nr_queues; i++) { +struct auxtrace_queue *queue = &spe->queues.queue_array[i]; +struct arm_spe_queue *speq = queue->priv; + +if (speq && (tid ==-1 || speq->tid == tid)) { + speq->time = time_; +arm_spe_set_pid_tid_cpu (spe, queue); +arm_spe_run_decoder (speq, &ts); + } + } +return 0 ; +} + +static int arm_spe_process_event (struct perf_session *session, +union perf_event *event, +struct perf_sample *sample, +struct perf_tool *tool) +{ +int err =0 ; +u64 timestamp; +struct arm_spe *spe =container_of (session->auxtrace, + struct arm_spe, auxtrace); + +if (dump_trace) +return 0 ; + +if (!tool->ordered_events) { +pr_err ("CoreSight SPE Trace requires ordered events\n" ); +return -EINVAL ; + } + +if (sample->time && (sample->time != (u64) -1 )) + timestamp = sample->time; +else + timestamp =0 ; + +if (timestamp || spe->timeless_decoding) { + err =arm_spe__update_queues (spe); +if (err) +return err; + } + +if (spe->timeless_decoding) { +if (event->header.type ==PERF_RECORD_EXIT ) { + err =arm_spe_process_timeless_queues (spe, + event->fork.tid, + sample->time); + } + }else if (timestamp) { +if (event->header.type ==PERF_RECORD_EXIT ) { + err =arm_spe_process_queues (spe, timestamp); +if (err) +return err; + } + } + +return err; +} +static int arm_spe_process_auxtrace_event (struct perf_session *session,union perf_event *event,struct perf_tool *tool__maybe_unused ) { struct arm_spe *spe = container_of(session->auxtrace, struct arm_spe, auxtrace); -struct auxtrace_buffer *buffer; -off_t data_offset; -int fd =perf_data__fd (session->data); -int err; -if (perf_data__is_pipe (session->data)) { - data_offset =0 ; - }else { - data_offset =lseek (fd,0 ,SEEK_CUR ); -if (data_offset ==-1 ) -return -errno; - } +if (!spe->data_queued) { +struct auxtrace_buffer *buffer; +off_t data_offset; +int fd =perf_data__fd (session->data); +int err; - err =auxtrace_queues__add_event (&spe->queues, session, event, - data_offset, &buffer); -if (err) -return err; - -/* Dump here now we have copied a piped trace out of the pipe */ -if (dump_trace) { -if (auxtrace_buffer__get_data (buffer, fd)) { -arm_spe_dump_event (spe, buffer->data, - buffer->size); -auxtrace_buffer__put_data (buffer); +if (perf_data__is_pipe (session->data)) { + data_offset =0 ; + }else { + data_offset =lseek (fd,0 ,SEEK_CUR ); +if (data_offset ==-1 ) +return -errno; + } + + err =auxtrace_queues__add_event (&spe->queues, session, event, + data_offset, &buffer); +if (err) +return err; + +/* Dump here now we have copied a piped trace out of the pipe */ +if (dump_trace) { +if (auxtrace_buffer__get_data (buffer, fd)) { +arm_spe_dump_event (spe, buffer->data, + buffer->size); +auxtrace_buffer__put_data (buffer); + } } }@@ -139,7 +623,25 @@ static int arm_spe_process_auxtrace_event(struct perf_session *session,static int arm_spe_flush (struct perf_session *session__maybe_unused ,struct perf_tool *tool__maybe_unused ) { -return 0 ; + struct arm_spe *spe = container_of(session->auxtrace, struct arm_spe, + auxtrace); +int ret; + +if (dump_trace) +return 0 ; + +if (!tool->ordered_events) +return -EINVAL ; + + ret =arm_spe__update_queues (spe); +if (ret <0 ) +return ret; + +if (spe->timeless_decoding) +return arm_spe_process_timeless_queues (spe,-1 , +MAX_TIMESTAMP -1 ); + +return arm_spe_process_queues (spe,MAX_TIMESTAMP ); }static void arm_spe_free_queue (void *priv)@@ -148,6 +650,9 @@ static void arm_spe_free_queue(void *priv)if (!speq)return ; +thread__zput (speq->thread); +arm_spe_decoder_free (speq->decoder); +zfree (&speq->event_buf);free (speq); }@@ -188,6 +693,149 @@ static void arm_spe_print_info(__u64 *arr)fprintf (stdout, arm_spe_info_fmts[ARM_SPE_PMU_TYPE ], arr[ARM_SPE_PMU_TYPE ]); } +struct arm_spe_synth { +struct perf_tool dummy_tool; +struct perf_session *session; +}; + +static int arm_spe_event_synth (struct perf_tool *tool, +union perf_event *event, +struct perf_sample *sample__maybe_unused , +struct machine *machine__maybe_unused ) +{ +struct arm_spe_synth *arm_spe_synth = + container_of(tool, struct arm_spe_synth, dummy_tool); + +return perf_session__deliver_synth_event (arm_spe_synth->session, + event,NULL ); +} + +static int arm_spe_synth_event (struct perf_session *session, +struct perf_event_attr *attr,u64 id) +{ +struct arm_spe_synth arm_spe_synth; + +memset (&arm_spe_synth,0 ,sizeof (struct arm_spe_synth )); + arm_spe_synth.session = session; + +return perf_event__synthesize_attr (&arm_spe_synth.dummy_tool, attr,1 , + &id, arm_spe_event_synth); +} + +static void arm_spe_set_event_name (struct evlist *evlist,u64 id, +const char *name) +{ +struct evsel *evsel; + +evlist__for_each_entry (evlist, evsel) { +if (evsel->core.id && evsel->core.id[0 ] == id) { +if (evsel->name) +zfree (&evsel->name); + evsel->name =strdup (name); +break ; + } + } +} + +static int +arm_spe_synth_events (struct arm_spe *spe,struct perf_session *session) +{ +struct evlist *evlist = session->evlist; +struct evsel *evsel; +struct perf_event_attr attr; +bool found = false; +u64 id; +int err; + +evlist__for_each_entry (evlist, evsel) { +if (evsel->core.attr.type == spe->pmu_type) { + found = true; +break ; + } + } + +if (!found) { +pr_debug ("No selected events with CoreSight Trace data\n" ); +return 0 ; + } + +memset (&attr,0 ,sizeof (struct perf_event_attr )); + attr.size =sizeof (struct perf_event_attr ); + attr.type =PERF_TYPE_HARDWARE ; + attr.sample_type = evsel->core.attr.sample_type &PERF_SAMPLE_MASK ; + attr.sample_type |=PERF_SAMPLE_IP |PERF_SAMPLE_TID | +PERF_SAMPLE_PERIOD ; +if (spe->timeless_decoding) + attr.sample_type &= ~(u64 )PERF_SAMPLE_TIME ; +else + attr.sample_type |=PERF_SAMPLE_TIME ; + + attr.exclude_user = evsel->core.attr.exclude_user; + attr.exclude_kernel = evsel->core.attr.exclude_kernel; + attr.exclude_hv = evsel->core.attr.exclude_hv; + attr.exclude_host = evsel->core.attr.exclude_host; + attr.exclude_guest = evsel->core.attr.exclude_guest; + attr.sample_id_all = evsel->core.attr.sample_id_all; + attr.read_format = evsel->core.attr.read_format; + +/* create new id val to be a fixed offset from evsel id */ + id = evsel->core.id[0 ] +1000000000 ; + +if (!id) + id =1 ; + +/* spe events set */ +if (spe->synth_opts.llc_miss) { + spe->sample_llc_miss = true; + +/* llc-miss */ + err =arm_spe_synth_event (session, &attr, id); +if (err) +return err; + spe->llc_miss_id = id; +arm_spe_set_event_name (evlist, id,"llc-miss" ); + id +=1 ; + } + +if (spe->synth_opts.tlb_miss) { + spe->sample_tlb_miss = true; + +/* tlb-miss */ + err =arm_spe_synth_event (session, &attr, id); +if (err) +return err; + spe->tlb_miss_id = id; +arm_spe_set_event_name (evlist, id,"tlb-miss" ); + id +=1 ; + } + +if (spe->synth_opts.branches) { + spe->sample_branch_miss = true; + +/* branch-miss */ + err =arm_spe_synth_event (session, &attr, id); +if (err) +return err; + spe->branch_miss_id = id; +arm_spe_set_event_name (evlist, id,"branch-miss" ); + id +=1 ; + } + +if (spe->synth_opts.remote_access) { + spe->sample_remote_access = true; + +/* remote-access */ + err =arm_spe_synth_event (session, &attr, id); +if (err) +return err; + spe->remote_access_id = id; +arm_spe_set_event_name (evlist, id,"remote-access" ); + id +=1 ; + } + +return 0 ; +} +int arm_spe_process_auxtrace_info (union perf_event *event,struct perf_session *session) {@@ -213,6 +861,7 @@ int arm_spe_process_auxtrace_info(union perf_event *event, spe->auxtrace_type = auxtrace_info->type; spe->pmu_type = auxtrace_info->priv[ARM_SPE_PMU_TYPE ]; + spe->timeless_decoding =arm_spe__is_timeless_decoding (spe); spe->auxtrace.process_event = arm_spe_process_event; spe->auxtrace.process_auxtrace_event = arm_spe_process_auxtrace_event; spe->auxtrace.flush_events = arm_spe_flush;@@ -222,8 +871,30 @@ int arm_spe_process_auxtrace_info(union perf_event *event,arm_spe_print_info (&auxtrace_info->priv[0 ]); +if (dump_trace) +return 0 ; + +if (session->itrace_synth_opts && session->itrace_synth_opts->set) + spe->synth_opts = *session->itrace_synth_opts; +else +itrace_synth_opts__set_default (&spe->synth_opts, false); + + err =arm_spe_synth_events (spe, session); +if (err) + goto err_free_queues; + + err =auxtrace_queues__process_index (&spe->queues, session); +if (err) + goto err_free_queues; + +if (spe->queues.populated) + spe->data_queued = true; +return 0 ; +err_free_queues: +auxtrace_queues__free (&spe->queues); + session->auxtrace =NULL ; err_free:free (spe);return err;tools/perf/util/auxtrace.c +13
-0@@ -1279,6 +1279,10 @@ void itrace_synth_opts__set_default(struct itrace_synth_opts *synth_opts, synth_opts->pwr_events = true; synth_opts->other_events = true; synth_opts->errors = true; + synth_opts->llc_miss = true; + synth_opts->tlb_miss = true; + synth_opts->remote_access = true; +if (no_sample) { synth_opts->period_type =PERF_ITRACE_PERIOD_INSTRUCTIONS ; synth_opts->period =1 ;@@ -1431,6 +1435,15 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str, goto out_err; p = endptr;break ; +case 'm' : + synth_opts->llc_miss = true; +break ; +case 't' : + synth_opts->tlb_miss = true; +break ; +case 'a' : + synth_opts->remote_access = true; +break ;case ' ' :case ',' :break ;tools/perf/util/auxtrace.h +7
-1@@ -60,7 +60,7 @@ enum itrace_period_type {* @inject: indicates the event (not just the sample) must be fully synthesized * because 'perf inject' will write it out * @instructions: whether to synthesize 'instructions' events -* @branches: whether to synthesize 'branches' events +* @branches: whether to synthesize 'branches' events (branch misses only on Arm) * @transactions: whether to synthesize events for transactions * @ptwrites: whether to synthesize events for ptwrites * @pwr_events: whether to synthesize power events @@ -74,6 +74,9 @@ enum itrace_period_type {* @callchain: add callchain to 'instructions' events * @thread_stack: feed branches to the thread_stack * @last_branch: add branch context to 'instruction' events +* @llc_miss: whether to synthesize last level cache miss events +* @tlb_miss: whether to synthesize TLB miss events +* @remote_access: whether to synthesize Remote access events * @callchain_sz: maximum callchain size * @last_branch_sz: branch context size * @period: 'instructions' events period @@ -101,6 +104,9 @@ struct itrace_synth_opts {bool callchain;bool thread_stack;bool last_branch; +bool llc_miss; +bool tlb_miss; +bool remote_access;unsigned int callchain_sz;unsigned int last_branch_sz;unsigned long long period;
PATCH v4 3/4 perf report: Add SPE options to --itrace argument
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
The previous patch added support in "perf report" for some arm-spe events(llc-miss, tlb-miss, branch-miss, remote_access). This patch adds their help instructions.
--- tools/perf/Documentation/itrace.txt | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
tools/perf/Documentation/itrace.txt +4 -1
@@ -1,5 +1,5 @@ i synthesize instructions events - b synthesize branches events + b synthesize branches events (branch misses on Arm) c synthesize branches events (calls only) r synthesize branches events (returns only) x synthesize transactions events
@@ -12,6 +12,9 @@ g synthesize a call chain (use with i or x) l synthesize last branch entries (use with i or x) s skip initial number of events + m synthesize LLC miss events + t synthesize TLB miss events + a synthesize remote access events The default is all events i.e. the same as --itrace=ibxwpe, except for perf script where it is --itrace=ce
Signature · 1 line
2.17.1
Re: [PATCH v4 3/4] perf report: Add SPE options to --itrace argument
Quoted 6 lines · James Clark
From: Tan Xiaojun <[email protected]>
The previous patch added support in "perf report" for some arm-spe
events(llc-miss, tlb-miss, branch-miss, remote_access). This patch adds their help instructions.
Quoted 13 lines
Signed-off-by Tan Xiaojun <[email protected]>Tested-by Qi Liu <[email protected]>Signed-off-by James Clark <[email protected]>Cc Will Deacon <[email protected]>Cc Mark Rutland <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Ingo Molnar <[email protected]>Cc Arnaldo Carvalho de Melo <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Jiri Olsa <[email protected]>Cc Tan Xiaojun <[email protected]>Cc Al Grant <[email protected]>Cc Namhyung Kim <[email protected]>
There is also ITRACE_HELP in tools/perf/util/auxtrace.h
Otherwise for auxtrace:
Quoted 26 lines
--- tools/perf/Documentation/itrace.txt | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)tools/perf/Documentation/itrace.txt +4
-1@@ -1,5 +1,5 @@ i synthesize instructions events - b synthesize branches events + b synthesize branches events (branch misses on Arm) c synthesize branches events (calls only) r synthesize branches events (returns only) x synthesize transactions events@@ -12,6 +12,9 @@ g synthesize a call chain (use with i or x) l synthesize last branch entries (use with i or x) s skip initial number of events + m synthesize LLC miss events + t synthesize TLB miss events + a synthesize remote access events The default is all events i.e. the same as --itrace=ibxwpe, except for perf script where it is --itrace=ce
v5 cover perf tools: Add support for some spe events and precise ip
Hi Adrian,
I've added the itrace arguments to ITRACE_HELP and also added the evsel->core.attr.type == PERF_TYPE_HARDWARE comparison.
Thanks
James
Tan Xiaojun (4):
perf tools: Move arm-spe-pkt-decoder.h/c to the new dir perf tools: Add support for "report" for some spe events perf report: Add SPE options to --itrace argument perf tools: Support "branch-misses:pp" on arm64
tools/perf/Documentation/itrace.txt | 5 +-
tools/perf/arch/arm/util/auxtrace.c | 39 +
tools/perf/builtin-record.c | 5 +
tools/perf/util/Build | 2 +-
tools/perf/util/arm-spe-decoder/Build | 1 +
.../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++
.../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++
.../arm-spe-pkt-decoder.c | 0
.../arm-spe-pkt-decoder.h | 2 +
tools/perf/util/arm-spe.c | 756 +++++++++++++++++-
tools/perf/util/arm-spe.h | 3 +
tools/perf/util/auxtrace.c | 13 +
tools/perf/util/auxtrace.h | 19 +-
13 files changed, 1094 insertions(+), 42 deletions(-)
create mode 100644 tools/perf/util/arm-spe-decoder/Build
create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c
create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%)
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (96%)Signature · 7 lines
2.17.1 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH v5 1/4 perf tools: Move arm-spe-pkt-decoder.h/c to the new dir
From: Tan Xiaojun <[email protected]>
Create a new arm-spe-decoder directory for subsequent extensions and move arm-spe-pkt-decoder.h/c to this directory. No code changes.
---
tools/perf/util/Build | 2 +-
tools/perf/util/arm-spe-decoder/Build | 1 +
tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c | 0
tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h | 0
tools/perf/util/arm-spe.c | 2 +-
5 files changed, 3 insertions(+), 2 deletions(-)
create mode 100644 tools/perf/util/arm-spe-decoder/Build
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%)
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (100%)tools/perf/util/Build +1 -1
@@ -104,7 +104,7 @@ perf-$(CONFIG_AUXTRACE) += intel-pt-decoder/ perf-$(CONFIG_AUXTRACE) += intel-pt.o perf-$(CONFIG_AUXTRACE) += intel-bts.o perf-$(CONFIG_AUXTRACE) += arm-spe.o -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-decoder/ perf-$(CONFIG_AUXTRACE) += s390-cpumsf.o ifdef CONFIG_LIBOPENCSD
tools/perf/util/arm-spe-decoder/Build new file +1 -0
@@ -0,0 +1 @@ +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.c renamed from tools/perf/util/arm-spe-pkt-decoder.c +0 -0
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h renamed from tools/perf/util/arm-spe-pkt-decoder.h +0 -0
tools/perf/util/arm-spe.c +1 -1
@@ -23,7 +23,7 @@#include "debug.h" #include "auxtrace.h" #include "arm-spe.h" -#include "arm-spe-pkt-decoder.h" +#include "arm-spe-decoder/arm-spe-pkt-decoder.h" struct arm_spe {struct auxtrace auxtrace;
Signature · 7 lines
2.17.1 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH v5 2/4 perf tools: Add support for "report" for some spe events
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
After the commit ffd3d18c20b8 ("perf tools: Add ARM Statistical Profiling Extensions (SPE) support") is merged, "perf record" and "perf report --dump-raw-trace" have been supported. However, the raw data that is dumped cannot be used without parsing.
This patch is to improve the "perf report" support for spe, and further process the data. Currently, support for the four events of llc-miss, tlb-miss, branch-miss, and remote-access is added.
Example usage:
$ ./perf record -c 1024 -e arm_spe_0/branch_filter=1,ts_enable=1,pct_enable=1,pa_enable=1,load_filter=1,jitter=1,store_filter=1,min_latency=0/ -o perf-armspe-dd.data dd if=/dev/zero of=/dev/null count=10000
$ ./perf report -i perf-armspe-dd.data --stdio
--------------------------------------------------------------------
...
# Samples: 23 of event 'llc-miss'
# Event count (approx.): 23
...
33.33% 33.33% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64
12.12% 12.12% dd [kernel.kallsyms] [k] perf_event_mmap
6.06% 6.06% dd [kernel.kallsyms] [k] copy_page
6.06% 6.06% dd ld-2.28.so [.] _dl_relocate_object
3.03% 3.03% dd [kernel.kallsyms] [k] change_protection_range
3.03% 3.03% dd [kernel.kallsyms] [k] filemap_map_pages
3.03% 3.03% dd [kernel.kallsyms] [k] free_pages_and_swap_cache
3.03% 3.03% dd [kernel.kallsyms] [k] generic_permission
3.03% 3.03% dd [kernel.kallsyms] [k] kmem_cache_alloc
3.03% 3.03% dd [kernel.kallsyms] [k] lookup_fast
3.03% 3.03% dd [kernel.kallsyms] [k] perf_event_exec
3.03% 3.03% dd [kernel.kallsyms] [k] radix_tree_next_chunk
3.03% 3.03% dd [kernel.kallsyms] [k] ring_buffer_record_is_on
3.03% 3.03% dd ld-2.28.so [.] _dl_lookup_symbol_x
3.03% 3.03% dd ld-2.28.so [.] _dl_start
3.03% 3.03% dd ld-2.28.so [.] dl_main
3.03% 3.03% dd ld-2.28.so [.] strcmp
3.03% 3.03% dd libc-2.28.so [.] _dl_addr
...
# Samples: 3 of event 'tlb-miss'
# Event count (approx.): 3
...
33.33% 33.33% dd [kernel.kallsyms] [k] filemap_map_pages
33.33% 33.33% dd ld-2.28.so [.] _dl_start
33.33% 33.33% dd ld-2.28.so [.] dl_main
...
# Samples: 20 of event 'branch-miss'
# Event count (approx.): 20
...
15.38% 15.38% dd [kernel.kallsyms] [k] __fput
7.69% 7.69% dd [kernel.kallsyms] [k] do_el0_ia_bp_hardening
7.69% 7.69% dd [kernel.kallsyms] [k] filemap_map_pages
7.69% 7.69% dd [kernel.kallsyms] [k] pagevec_lru_move_fn
7.69% 7.69% dd [kernel.kallsyms] [k] perf_event_mmap_output
7.69% 7.69% dd [kernel.kallsyms] [k] task_work_run
7.69% 7.69% dd [kernel.kallsyms] [k] unmap_single_vma
7.69% 7.69% dd libc-2.28.so [.] _IO_flush_all_lockp
7.69% 7.69% dd libc-2.28.so [.] __memcpy_generic
7.69% 7.69% dd libc-2.28.so [.] _dl_addr
7.69% 7.69% dd libc-2.28.so [.] msort_with_tmp.part.0
7.69% 7.69% dd libc-2.28.so [.] read_alias_file
...
# Samples: 5 of event 'remote-access'
# Event count (approx.): 5
...
27.78% 27.78% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64
16.67% 16.67% dd [kernel.kallsyms] [k] perf_event_mmap
5.56% 5.56% dd [kernel.kallsyms] [k] change_protection_range
5.56% 5.56% dd [kernel.kallsyms] [k] filemap_map_pages
5.56% 5.56% dd [kernel.kallsyms] [k] free_pages_and_swap_cache
5.56% 5.56% dd [kernel.kallsyms] [k] generic_permission
5.56% 5.56% dd [kernel.kallsyms] [k] lookup_fast
5.56% 5.56% dd [kernel.kallsyms] [k] perf_event_exec
5.56% 5.56% dd [kernel.kallsyms] [k] radix_tree_next_chunk
5.56% 5.56% dd ld-2.28.so [.] _dl_relocate_object
5.56% 5.56% dd ld-2.28.so [.] _dl_start
5.56% 5.56% dd ld-2.28.so [.] dl_main--------------------------------------------------------------------
After that, more analysis and processing of the raw data of spe will be done.
--- tools/perf/util/arm-spe-decoder/Build | 2 +- .../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++ .../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++ .../arm-spe-decoder/arm-spe-pkt-decoder.h | 2 + tools/perf/util/arm-spe.c | 745 +++++++++++++++++- tools/perf/util/auxtrace.c | 13 + tools/perf/util/auxtrace.h | 8 +- 7 files changed, 1022 insertions(+), 39 deletions(-) create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
tools/perf/util/arm-spe-decoder/Build +1 -1
@@ -1 +1 @@ -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o arm-spe-decoder.o
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +225 -0
225 lines not on this page · raw message
tools/perf/util/arm-spe-decoder/arm-spe-decoder.h new file +66 -0
66 lines not on this page · raw message
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h +2 -0
@@ -15,6 +15,8 @@#define ARM_SPE_NEED_MORE_BYTES -1#define ARM_SPE_BAD_PACKET -2 +#define ARM_SPE_PKT_MAX_SZ 16 +enum arm_spe_pkt_type {ARM_SPE_BAD ,ARM_SPE_PAD ,
tools/perf/util/arm-spe.c +708 -37
819 lines not on this page · raw message
tools/perf/util/auxtrace.c +13 -0
@@ -1279,6 +1279,10 @@ void itrace_synth_opts__set_default(struct itrace_synth_opts *synth_opts, synth_opts->pwr_events = true; synth_opts->other_events = true; synth_opts->errors = true; + synth_opts->llc_miss = true; + synth_opts->tlb_miss = true; + synth_opts->remote_access = true; +if (no_sample) { synth_opts->period_type =PERF_ITRACE_PERIOD_INSTRUCTIONS ; synth_opts->period =1 ;
@@ -1431,6 +1435,15 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str, goto out_err; p = endptr;break ; +case 'm' : + synth_opts->llc_miss = true; +break ; +case 't' : + synth_opts->tlb_miss = true; +break ; +case 'a' : + synth_opts->remote_access = true; +break ;case ' ' :case ',' :break ;
tools/perf/util/auxtrace.h +7 -1
@@ -60,7 +60,7 @@ enum itrace_period_type {* @inject: indicates the event (not just the sample) must be fully synthesized * because 'perf inject' will write it out * @instructions: whether to synthesize 'instructions' events -* @branches: whether to synthesize 'branches' events +* @branches: whether to synthesize 'branches' events (branch misses only on Arm) * @transactions: whether to synthesize events for transactions * @ptwrites: whether to synthesize events for ptwrites * @pwr_events: whether to synthesize power events
@@ -74,6 +74,9 @@ enum itrace_period_type {* @callchain: add callchain to 'instructions' events * @thread_stack: feed branches to the thread_stack * @last_branch: add branch context to 'instruction' events +* @llc_miss: whether to synthesize last level cache miss events +* @tlb_miss: whether to synthesize TLB miss events +* @remote_access: whether to synthesize Remote access events * @callchain_sz: maximum callchain size * @last_branch_sz: branch context size * @period: 'instructions' events period
@@ -101,6 +104,9 @@ struct itrace_synth_opts {bool callchain;bool thread_stack;bool last_branch; +bool llc_miss; +bool tlb_miss; +bool remote_access;unsigned int callchain_sz;unsigned int last_branch_sz;unsigned long long period;
Signature · 1 line
2.17.1
Re: [PATCH v5 2/4] perf tools: Add support for "report" for some spe events
Hi James, Xiaojun,
Quoted 8 lines · James Clark
From: Tan Xiaojun <[email protected]>
After the commit
ffd3d18c20b8("perf tools: Add ARM Statistical Profiling Extensions (SPE) support") is merged, "perf record" and "perf report --dump-raw-trace" have been supported. However, the raw data that is dumped cannot be used without parsing.This patch is to improve the "perf report" support for spe, and
Usually the capital letters are used for abbreviation, so s/spe/SPE.
Quoted 2 lines
further process the data. Currently, support for the four events of llc-miss, tlb-miss, branch-miss, and remote-access is added.
checkpatch.pl report 1 error and 10 warnings at my side, please consider to fix them.
Quoted 3 lines
Example usage:
$ ./perf record -c 1024 -e arm_spe_0/branch_filter=1,ts_enable=1,pct_enable=1,pa_enable=1,load_filter=1,jitter=1,store_filter=1,min_latency=0/ -o perf-armspe-dd.data dd if=/dev/zero of=/dev/null count=10000
If we need to input many configurations when use SPE, it might be not friendly for users. It's good to use default values as possible, and I'd like to suggest to write a document in Documentation/trace/ folder.
Quoted 112 lines
$ ./perf report -i perf-armspe-dd.data --stdio -------------------------------------------------------------------- ... # Samples: 23 of event 'llc-miss' # Event count (approx.): 23 ... 33.33% 33.33% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64 12.12% 12.12% dd [kernel.kallsyms] [k] perf_event_mmap 6.06% 6.06% dd [kernel.kallsyms] [k] copy_page 6.06% 6.06% dd ld-2.28.so [.] _dl_relocate_object 3.03% 3.03% dd [kernel.kallsyms] [k] change_protection_range 3.03% 3.03% dd [kernel.kallsyms] [k] filemap_map_pages 3.03% 3.03% dd [kernel.kallsyms] [k] free_pages_and_swap_cache 3.03% 3.03% dd [kernel.kallsyms] [k] generic_permission 3.03% 3.03% dd [kernel.kallsyms] [k] kmem_cache_alloc 3.03% 3.03% dd [kernel.kallsyms] [k] lookup_fast 3.03% 3.03% dd [kernel.kallsyms] [k] perf_event_exec 3.03% 3.03% dd [kernel.kallsyms] [k] radix_tree_next_chunk 3.03% 3.03% dd [kernel.kallsyms] [k] ring_buffer_record_is_on 3.03% 3.03% dd ld-2.28.so [.] _dl_lookup_symbol_x 3.03% 3.03% dd ld-2.28.so [.] _dl_start 3.03% 3.03% dd ld-2.28.so [.] dl_main 3.03% 3.03% dd ld-2.28.so [.] strcmp 3.03% 3.03% dd libc-2.28.so [.] _dl_addr ... # Samples: 3 of event 'tlb-miss' # Event count (approx.): 3 ... 33.33% 33.33% dd [kernel.kallsyms] [k] filemap_map_pages 33.33% 33.33% dd ld-2.28.so [.] _dl_start 33.33% 33.33% dd ld-2.28.so [.] dl_main ... # Samples: 20 of event 'branch-miss' # Event count (approx.): 20 ... 15.38% 15.38% dd [kernel.kallsyms] [k] __fput 7.69% 7.69% dd [kernel.kallsyms] [k] do_el0_ia_bp_hardening 7.69% 7.69% dd [kernel.kallsyms] [k] filemap_map_pages 7.69% 7.69% dd [kernel.kallsyms] [k] pagevec_lru_move_fn 7.69% 7.69% dd [kernel.kallsyms] [k] perf_event_mmap_output 7.69% 7.69% dd [kernel.kallsyms] [k] task_work_run 7.69% 7.69% dd [kernel.kallsyms] [k] unmap_single_vma 7.69% 7.69% dd libc-2.28.so [.] _IO_flush_all_lockp 7.69% 7.69% dd libc-2.28.so [.] __memcpy_generic 7.69% 7.69% dd libc-2.28.so [.] _dl_addr 7.69% 7.69% dd libc-2.28.so [.] msort_with_tmp.part.0 7.69% 7.69% dd libc-2.28.so [.] read_alias_file ... # Samples: 5 of event 'remote-access' # Event count (approx.): 5 ... 27.78% 27.78% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64 16.67% 16.67% dd [kernel.kallsyms] [k] perf_event_mmap 5.56% 5.56% dd [kernel.kallsyms] [k] change_protection_range 5.56% 5.56% dd [kernel.kallsyms] [k] filemap_map_pages 5.56% 5.56% dd [kernel.kallsyms] [k] free_pages_and_swap_cache 5.56% 5.56% dd [kernel.kallsyms] [k] generic_permission 5.56% 5.56% dd [kernel.kallsyms] [k] lookup_fast 5.56% 5.56% dd [kernel.kallsyms] [k] perf_event_exec 5.56% 5.56% dd [kernel.kallsyms] [k] radix_tree_next_chunk 5.56% 5.56% dd ld-2.28.so [.] _dl_relocate_object 5.56% 5.56% dd ld-2.28.so [.] _dl_start 5.56% 5.56% dd ld-2.28.so [.] dl_main--------------------------------------------------------------------
After that, more analysis and processing of the raw data of spe will be done.Signed-off-by Tan Xiaojun <[email protected]>Tested-by Qi Liu <[email protected]>Signed-off-by James Clark <[email protected]>Cc Will Deacon <[email protected]>Cc Mark Rutland <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Ingo Molnar <[email protected]>Cc Arnaldo Carvalho de Melo <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Jiri Olsa <[email protected]>Cc Tan Xiaojun <[email protected]>Cc Al Grant <[email protected]>Cc Namhyung Kim <[email protected]>--- tools/perf/util/arm-spe-decoder/Build | 2 +- .../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++ .../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++ .../arm-spe-decoder/arm-spe-pkt-decoder.h | 2 + tools/perf/util/arm-spe.c | 745 +++++++++++++++++- tools/perf/util/auxtrace.c | 13 + tools/perf/util/auxtrace.h | 8 +- 7 files changed, 1022 insertions(+), 39 deletions(-) create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.htools/perf/util/arm-spe-decoder/Build +1
-1@@ -1 +1 @@ -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o arm-spe-decoder.otools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +6
-0@@ -0,0 +1,225 @@ +// SPDX-License-Identifier: GPL-2.0 +/* +* arm_spe_decoder.c: ARM SPE support +*/ + +#ifndef _GNU_SOURCE
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 16
Quoted 10 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +10
-0+#define _GNU_SOURCE +#endif +#include <stdlib.h> +#include <stdbool.h> +#include <string.h> +#include <errno.h> +#include <stdint.h> +#include <inttypes.h> +#include <linux/compiler.h> +#include <linux/zalloc.h>
List headers with alphabetical order.
Quoted 9 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +9
-0+ +#include "../util.h" +#include "../debug.h" +#include "../auxtrace.h" + +#include "arm-spe-pkt-decoder.h" +#include "arm-spe-decoder.h" + +#ifndef BIT
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 35
Quoted 10 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +10
-0+#define BIT (n) (1UL << (n)) +#endif + +struct arm_spe_decoder { +int (*get_trace)(struct arm_spe_buffer *buffer,void *data); +void *data; +struct arm_spe_state state; +const unsigned char *buf; +size_t len; +uint64_t pos;
It's good to use U64 as type rather than uint64_t.
Quoted 10 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +10
-0+struct arm_spe_pkt packet; +int pkt_step; +int pkt_len; +int last_packet_type; + +uint64_t last_ip; +uint64_t ip; +uint64_t timestamp; +uint64_t sample_timestamp; +const unsigned char *next_buf;
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 55
Quoted 10 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +10
-0+size_t next_len; +unsigned char temp_buf[ARM_SPE_PKT_MAX_SZ ]; +}; + +static uint64_t arm_spe_calc_ip (uint64_t payload) +{ +uint64_t ip = (payload & ~(0xffULL <<56 )); + +/* fill high 8 bits for kernel virtual address */ +/* In Armv8 Architecture Reference Manual: Xn[55] determines
If refer to ARMv8-ARM, it's good to give out the exactly document number, e.g. ARM DDI 0487E.a.
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 58
Quoted 3 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +3
-0+* whether the address lies in the upper or lower address range +* for the purpose of determining whether address tagging is +* used */
Multiple lines comments use the fashion like:
/*
* Comments ...
* ... end
*/
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 60
Quoted 2 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +2
-0+if (ip &BIT (55 )) + ip |= (uint64_t )(0xffULL <<56 );
Sorry I might miss something at here when I searched the spec.
Please give more detailed section for the packet format. I read the section D10.2.1 'Address packet' and sub section 'Address packet payload', but doesn't see any description for BIT 55.
I also don't see any handling for below sub types:
- Data access physical address;
- Data access virtual address;
- Instruction virtual address.
Quoted 7 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +7
-0+ +return ip; +} + +struct arm_spe_decoder *arm_spe_decoder_new (struct arm_spe_params *params) +{ +struct arm_spe_decoder *decoder;
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 77
Quoted 10 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +10
-0+ +if (!params->get_trace) +return NULL ; + + decoder =zalloc (sizeof (struct arm_spe_decoder )); +if (!decoder) +return NULL ; + + decoder->get_trace = params->get_trace; + decoder->data = params->data;
Don't use indent before assignment.
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 90
Quoted 13 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +13
-0+ +return decoder; +} + +void arm_spe_decoder_free (struct arm_spe_decoder *decoder) +{ +free (decoder); +} + +static int arm_spe_bad_packet (struct arm_spe_decoder *decoder) +{ + decoder->pkt_len =1 ; + decoder->pkt_step =1 ;
I don't find decoder->pkt_len is used in any place.
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 91
Quoted 1 line
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +1
-0+pr_debug ("ERROR: Bad packet\n" );
For error, it's good to use pr_err() rather than pr_debug().
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 96
Quoted 5 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +5
-0+ +return -EBADMSG ; +} + +
Duplicate new lines.
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 104
Quoted 8 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +8
-0+static int arm_spe_get_data (struct arm_spe_decoder *decoder) +{ +struct arm_spe_buffer buffer = { .buf =0 , }; +int ret; + + decoder->pkt_step =0 ; + +pr_debug ("Getting more data\n" );
I'd like to remove the debugging info without any concrete info, if this log is used for debugging flow, we can use GDB alternatively.
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 113
Quoted 9 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +9
-0+ ret = decoder->get_trace (&buffer, decoder->data); +if (ret) +return ret; + + decoder->buf = buffer.buf; + decoder->len = buffer.len; +if (!decoder->len) { +pr_debug ("No more data\n" ); +return -ENODATA ;
This is the normal end of trace data, I don't think we need to return error number for this case.
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 121
Quoted 8 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +8
-0+ } + +return 0 ; +} + +static int arm_spe_get_next_data (struct arm_spe_decoder *decoder) +{ +return arm_spe_get_data (decoder);
The two functions arm_spe_get_next_data() and arm_spe_get_data() do the exactly same thing, so remove arm_spe_get_data()?
Quoted 4 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +4
-0+} + +static int arm_spe_get_next_packet (struct arm_spe_decoder *decoder) +{
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 135
Quoted 10 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +10
-0+int ret; + + decoder->last_packet_type = decoder->packet.type; + + do { + decoder->pos += decoder->pkt_step; + decoder->buf += decoder->pkt_step; + decoder->len -= decoder->pkt_step; + +
Redundant new line.
Quoted 15 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +15
-0+if (!decoder->len) { + ret =arm_spe_get_next_data (decoder); +if (ret) +return ret; + } + + ret =arm_spe_get_packet (decoder->buf, decoder->len, + &decoder->packet); +if (ret <=0 ) +return arm_spe_bad_packet (decoder); + + decoder->pkt_len = ret; + decoder->pkt_step = ret; + }while (decoder->packet.type ==ARM_SPE_PAD ); +
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 160
Quoted 10 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +10
-0+return 0 ; +} + +static int arm_spe_walk_trace (struct arm_spe_decoder *decoder) +{ +int err; +int idx; +uint64_t payload; + +while (1 ) {
I am confused by why here it needs to use 'while (1)' to traverse all packets.
Let's see below logic, if arm_spe_walk_trace() uses 'while (1)' to parse all packets, and then return to up layer to generate samples. Seems to me, the more reasonable logic is to parse one packet and directly return to up layer for samples synthesizing.
arm_spe_run_decoder() {
while (1) {
arm_spe_sample() => synthesize sample.
arm_spe_decode()
`-> arm_spe_walk_trace() => go through all packets.
}
}Quoted 7 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +7
-0+ err =arm_spe_get_next_packet (decoder); +if (err) +return err; + + idx = decoder->packet.index; + payload = decoder->packet.payload; +
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 177
Quoted 10 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +10
-0+switch (decoder->packet.type) { +case ARM_SPE_TIMESTAMP : + decoder->sample_timestamp = payload; +return 0 ; +case ARM_SPE_END : + decoder->sample_timestamp =0 ; +return 0 ; +case ARM_SPE_ADDRESS : + decoder->ip =arm_spe_calc_ip (payload); +if (idx ==0 )
Define macros for idx's 0 and 1, this would be more readable.
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 184
Quoted 7 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +7
-0+ decoder->state.from_ip = decoder->ip; + else if (idx == 1) + decoder->state.to_ip = decoder->ip; + break; + case ARM_SPE_COUNTER: + break; + case ARM_SPE_CONTEXT:
I think it misses to read out process ID.
Quoted 30 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +30
-0+break ; +case ARM_SPE_OP_TYPE : +break ; +case ARM_SPE_EVENTS : +if (payload &BIT (EV_TLB_REFILL )) + decoder->state.type |=ARM_SPE_TLB_MISS ; +if (payload &BIT (EV_MISPRED )) + decoder->state.type |=ARM_SPE_BRANCH_MISS ; +if (idx >1 && (payload &BIT (EV_LLC_REFILL ))) + decoder->state.type |=ARM_SPE_LLC_MISS ; +if (idx >1 && (payload &BIT (EV_REMOTE_ACCESS ))) + decoder->state.type |=ARM_SPE_REMOTE_ACCESS ; + +break ; +case ARM_SPE_DATA_SOURCE : +break ; +case ARM_SPE_BAD : +break ; +case ARM_SPE_PAD : +break ; +default : +pr_err ("Get Packet Error!\n" ); +return -ENOSYS ; + } + } +} + +const struct arm_spe_state *arm_spe_decode (struct arm_spe_decoder *decoder) +{ +int err;
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c line 224
Quoted 10 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +10
-0+ + decoder->state.type =0 ; + + err =arm_spe_walk_trace (decoder); +if (err) + decoder->state.err = err; + + decoder->state.timestamp = decoder->sample_timestamp; + +return &decoder->state;
Since decoder::state can be fetched by the caller, it's pointless to return &decoder->state. I think it's better to return error code for the function rather than return a structure pointer.
Quoted 91 lines
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +1
-0+}tools/perf/util/arm-spe-decoder/arm-spe-decoder.h new file +66
-0@@ -0,0 +1,66 @@ +// SPDX-License-Identifier: GPL-2.0 +/* +* arm_spe_decoder.c: ARM SPE support +*/ + +#ifndef INCLUDE__ARM_SPE_DECODER_H__ +#define INCLUDE__ARM_SPE_DECODER_H__ + +#include <stdint.h> +#include <stddef.h> +#include <stdbool.h> + +enum arm_spe_events { +EV_EXCEPTION_GEN , +EV_RETIRED , +EV_L1D_ACCESS , +EV_L1D_REFILL , +EV_TLB_ACCESS , +EV_TLB_REFILL , +EV_NOT_TAKEN , +EV_MISPRED , +EV_LLC_ACCESS , +EV_LLC_REFILL , +EV_REMOTE_ACCESS , +}; + +enum arm_spe_sample_type { +ARM_SPE_LLC_MISS =1 <<0 , +ARM_SPE_TLB_MISS =1 <<1 , +ARM_SPE_BRANCH_MISS =1 <<2 , +ARM_SPE_REMOTE_ACCESS =1 <<3 , +ARM_SPE_EX_STOP =1 <<6 , +}; + +struct arm_spe_state { +enum arm_spe_sample_type type; +int err; +uint64_t from_ip; +uint64_t to_ip; +uint64_t timestamp; +}; + +struct arm_spe_insn ; + +struct arm_spe_buffer { +const unsigned char *buf; +size_t len; +u64 offset; +bool consecutive; +uint64_t ref_timestamp; +uint64_t trace_nr; +}; + +struct arm_spe_params { +int (*get_trace)(struct arm_spe_buffer *buffer,void *data); +void *data; +}; + +struct arm_spe_decoder ; + +struct arm_spe_decoder *arm_spe_decoder_new (struct arm_spe_params *params); +void arm_spe_decoder_free (struct arm_spe_decoder *decoder); + +const struct arm_spe_state *arm_spe_decode (struct arm_spe_decoder *decoder); + +#endif tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h +2
-0@@ -15,6 +15,8 @@#define ARM_SPE_NEED_MORE_BYTES -1#define ARM_SPE_BAD_PACKET -2 +#define ARM_SPE_PKT_MAX_SZ 16 +enum arm_spe_pkt_type {ARM_SPE_BAD ,ARM_SPE_PAD ,tools/perf/util/arm-spe.c +0
-0
tools/perf/util/arm-spe.c line 23
Quoted 8 lines
tools/perf/util/arm-spe.c +3
-0@@ -16,34 +16,68 @@#include <linux/log2.h> #include <linux/zalloc.h> +#include "auxtrace.h" #include "color.h" +#include "debug.h" #include "evsel.h" +#include "evlist.h"
Alphabetical order.
Quoted 11 lines
tools/perf/util/arm-spe.c +6
-2#include "machine.h" #include "session.h" -#include "debug.h" -#include "auxtrace.h" +#include "symbol.h" +#include "thread.h" +#include "thread-stack.h" +#include "tool.h" +#include "util/synthetic-events.h" +#include "arm-spe.h"
tools/perf/util/arm-spe.c line 42
Quoted 10 lines
tools/perf/util/arm-spe.c +4
-0+#include "arm-spe-decoder/arm-spe-decoder.h" #include "arm-spe-decoder/arm-spe-pkt-decoder.h" +#define MAX_TIMESTAMP (~0ULL) +struct arm_spe {struct auxtrace auxtrace;struct auxtrace_queues queues;struct auxtrace_heap heap; +struct itrace_synth_opts synth_opts;
Tab indent.
Quoted 248 lines
tools/perf/util/arm-spe.c +222
-12u32 auxtrace_type;struct perf_session *session;struct machine *machine;u32 pmu_type; + +u8 timeless_decoding; +u8 data_queued; + +u8 sample_llc_miss; +u8 sample_tlb_miss; +u8 sample_branch_miss; +u8 sample_remote_access; +u64 llc_miss_id; +u64 tlb_miss_id; +u64 branch_miss_id; +u64 remote_access_id; +u64 kernel_start; + +unsigned long num_events; };struct arm_spe_queue { -struct arm_spe *spe; -unsigned int queue_nr; -struct auxtrace_buffer *buffer; -bool on_heap; -bool done; -pid_t pid; -pid_t tid; -int cpu; +struct arm_spe *spe; +unsigned int queue_nr; +struct auxtrace_buffer *buffer; +struct auxtrace_buffer *old_buffer; +union perf_event *event_buf; +bool on_heap; +bool done; +pid_t pid; +pid_t tid; +int cpu; +void *decoder; +const struct arm_spe_state *state; +u64 time; +u64 timestamp; +struct thread *thread; +bool have_sample; }; static void arm_spe_dump(struct arm_spe *spe __maybe_unused,@@ -92,44 +126,494 @@ static void arm_spe_dump_event(struct arm_spe *spe, unsigned char *buf,arm_spe_dump (spe, buf, len); } -static int arm_spe_process_event (struct perf_session *session__maybe_unused , -union perf_event *event__maybe_unused , -struct perf_sample *sample__maybe_unused , -struct perf_tool *tool__maybe_unused ) +static int arm_spe_get_trace (struct arm_spe_buffer *b,void *data) +{ +struct arm_spe_queue *speq = data; +struct auxtrace_buffer *buffer = speq->buffer; +struct auxtrace_buffer *old_buffer = speq->old_buffer; +struct auxtrace_queue *queue; + + queue = &speq->spe->queues.queue_array[speq->queue_nr]; + + buffer =auxtrace_buffer__next (queue, buffer); +/* If no more data, drop the previous auxtrace_buffer and return */ +if (!buffer) { +if (old_buffer) +auxtrace_buffer__drop_data (old_buffer); + b->len =0 ; +return 0 ; + } + + speq->buffer = buffer; + +/* If the aux_buffer doesn't have data associated, try to load it */ +if (!buffer->data) { +/* get the file desc associated with the perf data file */ +int fd =perf_data__fd (speq->spe->session->data); + + buffer->data =auxtrace_buffer__get_data (buffer, fd); +if (!buffer->data) +return -ENOMEM ; + } + +if (buffer->use_data) { + b->len = buffer->use_size; + b->buf = buffer->use_data; + }else { + b->len = buffer->size; + b->buf = buffer->data; + } + + b->ref_timestamp = buffer->reference; + +if (b->len) { +if (old_buffer) +auxtrace_buffer__drop_data (old_buffer); + speq->old_buffer = buffer; + }else { +auxtrace_buffer__drop_data (buffer); +return arm_spe_get_trace (b, data); + } + +return 0 ; +} + +static struct arm_spe_queue *arm_spe__alloc_queue (struct arm_spe *spe, +unsigned int queue_nr) +{ +struct arm_spe_params params = { .get_trace =0 , }; +struct arm_spe_queue *speq; + + speq =zalloc (sizeof (*speq)); +if (!speq) +return NULL ; + + speq->event_buf =malloc (PERF_SAMPLE_MAX_SIZE ); +if (!speq->event_buf) + goto out_free; + + speq->spe = spe; + speq->queue_nr = queue_nr; + speq->pid =-1 ; + speq->tid =-1 ; + speq->cpu =-1 ; + +/* params set */ + params.get_trace = arm_spe_get_trace; + params.data = speq; + +/* create new decoder */ + speq->decoder =arm_spe_decoder_new (¶ms); +if (!speq->decoder) + goto out_free; + +return speq; + +out_free: +zfree (&speq->event_buf); +free (speq); + +return NULL ; +} + +static inline u8 arm_spe_cpumode (struct arm_spe *spe,uint64_t ip) +{ +return ip >= spe->kernel_start ? +PERF_RECORD_MISC_KERNEL : +PERF_RECORD_MISC_USER ; +} + +static void arm_spe_prep_sample (struct arm_spe *spe, +struct arm_spe_queue *speq, +union perf_event *event, +struct perf_sample *sample) +{ +if (!spe->timeless_decoding) + sample->time = speq->timestamp; + + sample->ip = speq->state->from_ip; + sample->cpumode =arm_spe_cpumode (spe, sample->ip); + sample->pid = speq->pid; + sample->tid = speq->tid; + sample->addr = speq->state->to_ip; + sample->period =1 ; + sample->cpu = speq->cpu; + + event->sample.header.type =PERF_RECORD_SAMPLE ; + event->sample.header.misc = sample->cpumode; + event->sample.header.size =sizeof (struct perf_event_header ); +} + +static inline int +arm_spe_deliver_synth_event (struct arm_spe *spe, +struct arm_spe_queue *speq__maybe_unused , +union perf_event *event, +struct perf_sample *sample) +{ +int ret; + + ret =perf_session__deliver_synth_event (spe->session, event, sample); +if (ret) +pr_err ("ARM SPE: failed to deliver event, error %d\n" , ret); + +return ret; +} + +static int +arm_spe_synth_spe_events_sample (struct arm_spe_queue *speq, +u64 spe_events_id) +{ +struct arm_spe *spe = speq->spe; +union perf_event *event = speq->event_buf; +struct perf_sample sample = { .ip =0 , }; + +arm_spe_prep_sample (spe, speq, event, &sample); + + sample.id = spe_events_id; + sample.stream_id = spe_events_id; + +return arm_spe_deliver_synth_event (spe, speq, event, &sample); +} + +static int arm_spe_sample (struct arm_spe_queue *speq) +{ +const struct arm_spe_state *state = speq->state; +struct arm_spe *spe = speq->spe; +int err; + +if (!speq->have_sample) +return 0 ; + + speq->have_sample = false; + +if (spe->sample_llc_miss && (state->type &ARM_SPE_LLC_MISS )) { + err =arm_spe_synth_spe_events_sample (speq, spe->llc_miss_id); +if (err) +return err; + } + +if (spe->sample_tlb_miss && (state->type &ARM_SPE_TLB_MISS )) { + err =arm_spe_synth_spe_events_sample (speq, spe->tlb_miss_id); +if (err) +return err; + } + +if (spe->sample_branch_miss && (state->type &ARM_SPE_BRANCH_MISS )) { + err =arm_spe_synth_spe_events_sample (speq, + spe->branch_miss_id); +if (err) +return err; + } + +if (spe->sample_remote_access && (state->type &ARM_SPE_REMOTE_ACCESS )) { + err =arm_spe_synth_spe_events_sample (speq, spe->remote_access_id); +if (err) +return err; + } + +return 0 ; +} + +static int arm_spe_run_decoder (struct arm_spe_queue *speq,u64 *timestamp) +{ +const struct arm_spe_state *state = speq->state;
tools/perf/util/arm-spe.c static void arm_spe_dump_event(struct arm_spe *spe, unsigned char *buf, · line 329
Quoted 10 lines
tools/perf/util/arm-spe.c +10
-0+struct arm_spe *spe = speq->spe; +int err; + +if (!spe->kernel_start) + spe->kernel_start =machine__kernel_start (spe->machine); + +while (1 ) { + err =arm_spe_sample (speq); +if (err) +return err;
Should reverse the flow between arm_spe_sample() and arm_spe_decode().
Quoted 159 lines
tools/perf/util/arm-spe.c +155
-0+ + state =arm_spe_decode (speq->decoder); +if (state->err) { +if (state->err == -ENODATA ) { +pr_debug ("No data or all data has been processed.\n" ); +return 1 ; + } +continue ; + } + + speq->state = state; + speq->have_sample = true; + +if (!spe->timeless_decoding && speq->timestamp >= *timestamp) { + *timestamp = speq->timestamp; +return 0 ; + } + } + +return 0 ; +} + +static int arm_spe__setup_queue (struct arm_spe *spe, +struct auxtrace_queue *queue, +unsigned int queue_nr) +{ +struct arm_spe_queue *speq = queue->priv; + +if (list_empty (&queue->head) || speq) +return 0 ; + + speq =arm_spe__alloc_queue (spe, queue_nr); + +if (!speq) +return -ENOMEM ; + + queue->priv = speq; + +if (queue->cpu !=-1 ) + speq->cpu = queue->cpu; + +if (!speq->on_heap) { +const struct arm_spe_state *state; +int ret; + +if (spe->timeless_decoding) +return 0 ; + +retry: + state =arm_spe_decode (speq->decoder); +if (state->err) { +if (state->err == -ENODATA ) { +pr_debug ("queue %u has no timestamp\n" , + queue_nr); +return 0 ; + } + goto retry; + } + + speq->timestamp = state->timestamp; + speq->state = state; + speq->have_sample = true; + ret =auxtrace_heap__add (&spe->heap, queue_nr, speq->timestamp); +if (ret) +return ret; + speq->on_heap = true; + } + +return 0 ; +} + +static int arm_spe__setup_queues (struct arm_spe *spe) { +unsigned int i; +int ret; + +for (i =0 ; i < spe->queues.nr_queues; i++) { + ret =arm_spe__setup_queue (spe, &spe->queues.queue_array[i], i); +if (ret) +return ret; + } + +return 0 ; +} + +static int arm_spe__update_queues (struct arm_spe *spe) +{ +if (spe->queues.new_data) { + spe->queues.new_data = false; +return arm_spe__setup_queues (spe); + } +return 0 ; } +static bool arm_spe__is_timeless_decoding (struct arm_spe *spe) +{ +struct evsel *evsel; +struct evlist *evlist = spe->session->evlist; +bool timeless_decoding = true; + +/* +* Circle through the list of event and complain if we find one +* with the time bit set. +*/ +evlist__for_each_entry (evlist, evsel) { +if ((evsel->core.attr.sample_type &PERF_SAMPLE_TIME )) + timeless_decoding = false; + } + +return timeless_decoding; +} + +static void arm_spe_set_pid_tid_cpu (struct arm_spe *spe, +struct auxtrace_queue *queue) +{ +struct arm_spe_queue *speq = queue->priv; +pid_t tid; + + tid =machine__get_current_tid (spe->machine, speq->cpu); +if (tid !=-1 ) { + speq->tid = tid; +thread__zput (speq->thread); + }else + speq->tid = queue->tid; + +if ((!speq->thread) && (speq->tid !=-1 )) { + speq->thread =machine__find_thread (spe->machine,-1 , + speq->tid); + } + +if (speq->thread) { + speq->pid = speq->thread->pid_; +if (queue->cpu ==-1 ) + speq->cpu = speq->thread->cpu; + } +} + +static int arm_spe_process_queues (struct arm_spe *spe,u64 timestamp) +{ +unsigned int queue_nr; +u64 ts; +int ret; + +while (1 ) { +struct auxtrace_queue *queue; +struct arm_spe_queue *speq; + +if (!spe->heap.heap_cnt) +return 0 ; + +if (spe->heap.heap_array[0 ].ordinal >= timestamp) +return 0 ; + + queue_nr = spe->heap.heap_array[0 ].queue_nr; + queue = &spe->queues.queue_array[queue_nr]; + speq = queue->priv; + +auxtrace_heap__pop (&spe->heap);
tools/perf/util/arm-spe.c static void arm_spe_dump_event(struct arm_spe *spe, unsigned char *buf, · line 498
Quoted 10 lines
tools/perf/util/arm-spe.c +10
-0+ +if (spe->heap.heap_cnt) { + ts = spe->heap.heap_array[0 ].ordinal +1 ; +if (ts > timestamp) + ts = timestamp; + }else { + ts = timestamp; + } + +arm_spe_set_pid_tid_cpu (spe, queue);
I don't think this is right.
arm_spe_set_pid_tid_cpu() should be invoked by SPE decoder when SPE decoder finds CONTEXT packet.
I will look into more detailed implementation at my side when I can run the code on a test platform, and might give more comments after get some trying.
Thanks,
Leo
Quoted 436 lines
tools/perf/util/arm-spe.c +298
-23+ + ret =arm_spe_run_decoder (speq, &ts); +if (ret <0 ) { +auxtrace_heap__add (&spe->heap, queue_nr, ts); +return ret; + } + +if (!ret) { + ret =auxtrace_heap__add (&spe->heap, queue_nr, ts); +if (ret <0 ) +return ret; + }else { + speq->on_heap = false; + } + } + +return 0 ; +} + +static int arm_spe_process_timeless_queues (struct arm_spe *spe,pid_t tid, +u64 time_) +{ +struct auxtrace_queues *queues = &spe->queues; +unsigned int i; +u64 ts =0 ; + +for (i =0 ; i < queues->nr_queues; i++) { +struct auxtrace_queue *queue = &spe->queues.queue_array[i]; +struct arm_spe_queue *speq = queue->priv; + +if (speq && (tid ==-1 || speq->tid == tid)) { + speq->time = time_; +arm_spe_set_pid_tid_cpu (spe, queue); +arm_spe_run_decoder (speq, &ts); + } + } +return 0 ; +} + +static int arm_spe_process_event (struct perf_session *session, +union perf_event *event, +struct perf_sample *sample, +struct perf_tool *tool) +{ +int err =0 ; +u64 timestamp; +struct arm_spe *spe =container_of (session->auxtrace, + struct arm_spe, auxtrace); + +if (dump_trace) +return 0 ; + +if (!tool->ordered_events) { +pr_err ("CoreSight SPE Trace requires ordered events\n" ); +return -EINVAL ; + } + +if (sample->time && (sample->time != (u64) -1 )) + timestamp = sample->time; +else + timestamp =0 ; + +if (timestamp || spe->timeless_decoding) { + err =arm_spe__update_queues (spe); +if (err) +return err; + } + +if (spe->timeless_decoding) { +if (event->header.type ==PERF_RECORD_EXIT ) { + err =arm_spe_process_timeless_queues (spe, + event->fork.tid, + sample->time); + } + }else if (timestamp) { +if (event->header.type ==PERF_RECORD_EXIT ) { + err =arm_spe_process_queues (spe, timestamp); +if (err) +return err; + } + } + +return err; +} +static int arm_spe_process_auxtrace_event (struct perf_session *session,union perf_event *event,struct perf_tool *tool__maybe_unused ) { struct arm_spe *spe = container_of(session->auxtrace, struct arm_spe, auxtrace); -struct auxtrace_buffer *buffer; -off_t data_offset; -int fd =perf_data__fd (session->data); -int err; -if (perf_data__is_pipe (session->data)) { - data_offset =0 ; - }else { - data_offset =lseek (fd,0 ,SEEK_CUR ); -if (data_offset ==-1 ) -return -errno; - } +if (!spe->data_queued) { +struct auxtrace_buffer *buffer; +off_t data_offset; +int fd =perf_data__fd (session->data); +int err; - err =auxtrace_queues__add_event (&spe->queues, session, event, - data_offset, &buffer); -if (err) -return err; - -/* Dump here now we have copied a piped trace out of the pipe */ -if (dump_trace) { -if (auxtrace_buffer__get_data (buffer, fd)) { -arm_spe_dump_event (spe, buffer->data, - buffer->size); -auxtrace_buffer__put_data (buffer); +if (perf_data__is_pipe (session->data)) { + data_offset =0 ; + }else { + data_offset =lseek (fd,0 ,SEEK_CUR ); +if (data_offset ==-1 ) +return -errno; + } + + err =auxtrace_queues__add_event (&spe->queues, session, event, + data_offset, &buffer); +if (err) +return err; + +/* Dump here now we have copied a piped trace out of the pipe */ +if (dump_trace) { +if (auxtrace_buffer__get_data (buffer, fd)) { +arm_spe_dump_event (spe, buffer->data, + buffer->size); +auxtrace_buffer__put_data (buffer); + } } }@@ -139,7 +623,25 @@ static int arm_spe_process_auxtrace_event(struct perf_session *session,static int arm_spe_flush (struct perf_session *session__maybe_unused ,struct perf_tool *tool__maybe_unused ) { -return 0 ; + struct arm_spe *spe = container_of(session->auxtrace, struct arm_spe, + auxtrace); +int ret; + +if (dump_trace) +return 0 ; + +if (!tool->ordered_events) +return -EINVAL ; + + ret =arm_spe__update_queues (spe); +if (ret <0 ) +return ret; + +if (spe->timeless_decoding) +return arm_spe_process_timeless_queues (spe,-1 , +MAX_TIMESTAMP -1 ); + +return arm_spe_process_queues (spe,MAX_TIMESTAMP ); }static void arm_spe_free_queue (void *priv)@@ -148,6 +650,9 @@ static void arm_spe_free_queue(void *priv)if (!speq)return ; +thread__zput (speq->thread); +arm_spe_decoder_free (speq->decoder); +zfree (&speq->event_buf);free (speq); }@@ -188,6 +693,149 @@ static void arm_spe_print_info(__u64 *arr)fprintf (stdout, arm_spe_info_fmts[ARM_SPE_PMU_TYPE ], arr[ARM_SPE_PMU_TYPE ]); } +struct arm_spe_synth { +struct perf_tool dummy_tool; +struct perf_session *session; +}; + +static int arm_spe_event_synth (struct perf_tool *tool, +union perf_event *event, +struct perf_sample *sample__maybe_unused , +struct machine *machine__maybe_unused ) +{ +struct arm_spe_synth *arm_spe_synth = + container_of(tool, struct arm_spe_synth, dummy_tool); + +return perf_session__deliver_synth_event (arm_spe_synth->session, + event,NULL ); +} + +static int arm_spe_synth_event (struct perf_session *session, +struct perf_event_attr *attr,u64 id) +{ +struct arm_spe_synth arm_spe_synth; + +memset (&arm_spe_synth,0 ,sizeof (struct arm_spe_synth )); + arm_spe_synth.session = session; + +return perf_event__synthesize_attr (&arm_spe_synth.dummy_tool, attr,1 , + &id, arm_spe_event_synth); +} + +static void arm_spe_set_event_name (struct evlist *evlist,u64 id, +const char *name) +{ +struct evsel *evsel; + +evlist__for_each_entry (evlist, evsel) { +if (evsel->core.id && evsel->core.id[0 ] == id) { +if (evsel->name) +zfree (&evsel->name); + evsel->name =strdup (name); +break ; + } + } +} + +static int +arm_spe_synth_events (struct arm_spe *spe,struct perf_session *session) +{ +struct evlist *evlist = session->evlist; +struct evsel *evsel; +struct perf_event_attr attr; +bool found = false; +u64 id; +int err; + +evlist__for_each_entry (evlist, evsel) { +if (evsel->core.attr.type == spe->pmu_type) { + found = true; +break ; + } + } + +if (!found) { +pr_debug ("No selected events with CoreSight Trace data\n" ); +return 0 ; + } + +memset (&attr,0 ,sizeof (struct perf_event_attr )); + attr.size =sizeof (struct perf_event_attr ); + attr.type =PERF_TYPE_HARDWARE ; + attr.sample_type = evsel->core.attr.sample_type &PERF_SAMPLE_MASK ; + attr.sample_type |=PERF_SAMPLE_IP |PERF_SAMPLE_TID | +PERF_SAMPLE_PERIOD ; +if (spe->timeless_decoding) + attr.sample_type &= ~(u64 )PERF_SAMPLE_TIME ; +else + attr.sample_type |=PERF_SAMPLE_TIME ; + + attr.exclude_user = evsel->core.attr.exclude_user; + attr.exclude_kernel = evsel->core.attr.exclude_kernel; + attr.exclude_hv = evsel->core.attr.exclude_hv; + attr.exclude_host = evsel->core.attr.exclude_host; + attr.exclude_guest = evsel->core.attr.exclude_guest; + attr.sample_id_all = evsel->core.attr.sample_id_all; + attr.read_format = evsel->core.attr.read_format; + +/* create new id val to be a fixed offset from evsel id */ + id = evsel->core.id[0 ] +1000000000 ; + +if (!id) + id =1 ; + +/* spe events set */ +if (spe->synth_opts.llc_miss) { + spe->sample_llc_miss = true; + +/* llc-miss */ + err =arm_spe_synth_event (session, &attr, id); +if (err) +return err; + spe->llc_miss_id = id; +arm_spe_set_event_name (evlist, id,"llc-miss" ); + id +=1 ; + } + +if (spe->synth_opts.tlb_miss) { + spe->sample_tlb_miss = true; + +/* tlb-miss */ + err =arm_spe_synth_event (session, &attr, id); +if (err) +return err; + spe->tlb_miss_id = id; +arm_spe_set_event_name (evlist, id,"tlb-miss" ); + id +=1 ; + } + +if (spe->synth_opts.branches) { + spe->sample_branch_miss = true; + +/* branch-miss */ + err =arm_spe_synth_event (session, &attr, id); +if (err) +return err; + spe->branch_miss_id = id; +arm_spe_set_event_name (evlist, id,"branch-miss" ); + id +=1 ; + } + +if (spe->synth_opts.remote_access) { + spe->sample_remote_access = true; + +/* remote-access */ + err =arm_spe_synth_event (session, &attr, id); +if (err) +return err; + spe->remote_access_id = id; +arm_spe_set_event_name (evlist, id,"remote-access" ); + id +=1 ; + } + +return 0 ; +} +int arm_spe_process_auxtrace_info (union perf_event *event,struct perf_session *session) {@@ -213,6 +861,7 @@ int arm_spe_process_auxtrace_info(union perf_event *event, spe->auxtrace_type = auxtrace_info->type; spe->pmu_type = auxtrace_info->priv[ARM_SPE_PMU_TYPE ]; + spe->timeless_decoding =arm_spe__is_timeless_decoding (spe); spe->auxtrace.process_event = arm_spe_process_event; spe->auxtrace.process_auxtrace_event = arm_spe_process_auxtrace_event; spe->auxtrace.flush_events = arm_spe_flush;@@ -222,8 +871,30 @@ int arm_spe_process_auxtrace_info(union perf_event *event,arm_spe_print_info (&auxtrace_info->priv[0 ]); +if (dump_trace) +return 0 ; + +if (session->itrace_synth_opts && session->itrace_synth_opts->set) + spe->synth_opts = *session->itrace_synth_opts; +else +itrace_synth_opts__set_default (&spe->synth_opts, false); + + err =arm_spe_synth_events (spe, session); +if (err) + goto err_free_queues; + + err =auxtrace_queues__process_index (&spe->queues, session); +if (err) + goto err_free_queues; + +if (spe->queues.populated) + spe->data_queued = true; +return 0 ; +err_free_queues: +auxtrace_queues__free (&spe->queues); + session->auxtrace =NULL ; err_free:free (spe);return err;tools/perf/util/auxtrace.c +13
-0@@ -1279,6 +1279,10 @@ void itrace_synth_opts__set_default(struct itrace_synth_opts *synth_opts, synth_opts->pwr_events = true; synth_opts->other_events = true; synth_opts->errors = true; + synth_opts->llc_miss = true; + synth_opts->tlb_miss = true; + synth_opts->remote_access = true; +if (no_sample) { synth_opts->period_type =PERF_ITRACE_PERIOD_INSTRUCTIONS ; synth_opts->period =1 ;@@ -1431,6 +1435,15 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str, goto out_err; p = endptr;break ; +case 'm' : + synth_opts->llc_miss = true; +break ; +case 't' : + synth_opts->tlb_miss = true; +break ; +case 'a' : + synth_opts->remote_access = true; +break ;case ' ' :case ',' :break ;tools/perf/util/auxtrace.h +7
-1@@ -60,7 +60,7 @@ enum itrace_period_type {* @inject: indicates the event (not just the sample) must be fully synthesized * because 'perf inject' will write it out * @instructions: whether to synthesize 'instructions' events -* @branches: whether to synthesize 'branches' events +* @branches: whether to synthesize 'branches' events (branch misses only on Arm) * @transactions: whether to synthesize events for transactions * @ptwrites: whether to synthesize events for ptwrites * @pwr_events: whether to synthesize power events @@ -74,6 +74,9 @@ enum itrace_period_type {* @callchain: add callchain to 'instructions' events * @thread_stack: feed branches to the thread_stack * @last_branch: add branch context to 'instruction' events +* @llc_miss: whether to synthesize last level cache miss events +* @tlb_miss: whether to synthesize TLB miss events +* @remote_access: whether to synthesize Remote access events * @callchain_sz: maximum callchain size * @last_branch_sz: branch context size * @period: 'instructions' events period @@ -101,6 +104,9 @@ struct itrace_synth_opts {bool callchain;bool thread_stack;bool last_branch; +bool llc_miss; +bool tlb_miss; +bool remote_access;unsigned int callchain_sz;unsigned int last_branch_sz;unsigned long long period;Signature · 1 line
2.17.1
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH v5 3/4 perf report: Add SPE options to --itrace argument
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
The previous patch added support in "perf report" for some arm-spe events(llc-miss, tlb-miss, branch-miss, remote_access). This patch adds their help instructions.
--- tools/perf/Documentation/itrace.txt | 5 ++++- tools/perf/util/auxtrace.h | 5 ++++- 2 files changed, 8 insertions(+), 2 deletions(-)
tools/perf/Documentation/itrace.txt +4 -1
@@ -1,5 +1,5 @@ i synthesize instructions events - b synthesize branches events + b synthesize branches events (branch misses on Arm) c synthesize branches events (calls only) r synthesize branches events (returns only) x synthesize transactions events
@@ -9,6 +9,9 @@ of aux-output (refer to perf record) e synthesize error events d create a debug log + m synthesize LLC miss events + t synthesize TLB miss events + a synthesize remote access events g synthesize a call chain (use with i or x) l synthesize last branch entries (use with i or x) s skip initial number of events
tools/perf/util/auxtrace.h +4 -1
@@ -587,7 +587,7 @@ void auxtrace__free(struct perf_session *session);#define ITRACE_HELP \ " i: synthesize instructions events\n" \ -" b: synthesize branches events\n" \ +" b: synthesize branches events (branch misses on Arm)\n" \ " c: synthesize branches events (calls only)\n" \ " r: synthesize branches events (returns only)\n" \ " x: synthesize transactions events\n" \
@@ -595,6 +595,9 @@ void auxtrace__free(struct perf_session *session);" p: synthesize power events\n" \" e: synthesize error events\n" \" d: create a debug log\n" \ +" m: synthesize LLC miss events\n" \ +" t: synthesize TLB miss events\n" \ +" a: synthesize remote access events\n" \" g[len]: synthesize a call chain (use with i or x)\n" \" l[len]: synthesize last branch entries (use with i or x)\n" \" sNUMBER: skip initial number of events\n" \
Signature · 7 lines
2.17.1 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH v5 4/4 perf tools: Support "branch-misses:pp" on arm64
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
At the suggestion of James Clark, use spe to support the precise ip of some events. Currently its support event is:
branch-misses.
Example usage:
$ ./perf record -e branch-misses:pp dd if=/dev/zero of=/dev/null count=10000 (:p/pp/ppp is same for this case.)
$ ./perf report --stdio
("--stdio is not necessary")--------------------------------------------------------------------
...
# Samples: 14 of event 'branch-misses:pp' # Event count (approx.): 14
#
# Children Self Command Shared Object Symbol
# ........ ........ ....... ................. ..........................
#
14.29% 14.29% dd [kernel.kallsyms] [k] __arch_copy_from_user
14.29% 14.29% dd libc-2.28.so [.] _dl_addr
7.14% 7.14% dd [kernel.kallsyms] [k] __free_pages
7.14% 7.14% dd [kernel.kallsyms] [k] __pi_memcpy
7.14% 7.14% dd [kernel.kallsyms] [k] pagecache_get_page
7.14% 7.14% dd [kernel.kallsyms] [k] unmap_single_vma
7.14% 7.14% dd dd [.] 0x00000000000025ec
7.14% 7.14% dd ld-2.28.so [.] _dl_lookup_symbol_x
7.14% 7.14% dd ld-2.28.so [.] check_match
7.14% 7.14% dd libc-2.28.so [.] __mpn_rshift
7.14% 7.14% dd libc-2.28.so [.] _nl_intern_locale_data
7.14% 7.14% dd libc-2.28.so [.] read_alias_file...
--------------------------------------------------------------------
--- tools/perf/arch/arm/util/auxtrace.c | 39 +++++++++++++++++++++++++++++ tools/perf/builtin-record.c | 5 ++++ tools/perf/util/arm-spe.c | 9 +++++++ tools/perf/util/arm-spe.h | 3 +++ tools/perf/util/auxtrace.h | 6 +++++ 5 files changed, 62 insertions(+)
tools/perf/arch/arm/util/auxtrace.c +39 -0
@@ -10,11 +10,25 @@#include "../../util/auxtrace.h" #include "../../util/debug.h" +#include "../../util/env.h" #include "../../util/evlist.h" #include "../../util/pmu.h" #include "cs-etm.h" #include "arm-spe.h" +#define SPE_ATTR_TS_ENABLE BIT(0) +#define SPE_ATTR_PA_ENABLE BIT(1) +#define SPE_ATTR_PCT_ENABLE BIT(2) +#define SPE_ATTR_JITTER BIT(16) +#define SPE_ATTR_BRANCH_FILTER BIT(32) +#define SPE_ATTR_LOAD_FILTER BIT(33) +#define SPE_ATTR_STORE_FILTER BIT(34) + +#define SPE_ATTR_EV_RETIRED BIT(1) +#define SPE_ATTR_EV_CACHE BIT(3) +#define SPE_ATTR_EV_TLB BIT(5) +#define SPE_ATTR_EV_BRANCH BIT(7) +static struct perf_pmu **find_all_arm_spe_pmus (int *nr_spes,int *err) {struct perf_pmu **arm_spe_pmus =NULL ;
@@ -108,3 +122,28 @@ struct auxtrace_record *err =0 ;return NULL ; } + +void auxtrace__preprocess_evlist (struct evlist *evlist) +{ +struct evsel *evsel; +struct perf_pmu *pmu; + +evlist__for_each_entry (evlist, evsel) { +/* Currently only supports precise_ip for branch-misses on arm64 */ +if (!strcmp (perf_env__arch (evlist->env),"arm64" ) + && evsel->core.attr.config ==PERF_COUNT_HW_BRANCH_MISSES + && evsel->core.attr.type ==PERF_TYPE_HARDWARE + && evsel->core.attr.precise_ip) + { + pmu =perf_pmu__find ("arm_spe_0" ); +if (pmu) { + evsel->pmu_name = pmu->name; + evsel->core.attr.type = pmu->type; + evsel->core.attr.config =SPE_ATTR_TS_ENABLE + |SPE_ATTR_BRANCH_FILTER ; + evsel->core.attr.config1 =SPE_ATTR_EV_BRANCH ; + evsel->core.attr.precise_ip =0 ; + } + } + } +}
tools/perf/builtin-record.c +5 -0
@@ -2451,6 +2451,11 @@ int cmd_record(int argc, const char **argv) argc =parse_options (argc, argv, record_options, record_usage,PARSE_OPT_STOP_AT_NON_OPTION ); + +if (auxtrace__preprocess_evlist) { +auxtrace__preprocess_evlist (rec->evlist); + } +if (quiet)perf_quiet_option ();
tools/perf/util/arm-spe.c +9 -0
@@ -778,6 +778,15 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session) attr.sample_id_all = evsel->core.attr.sample_id_all; attr.read_format = evsel->core.attr.read_format; +/* If it is in the precise ip mode, there is no need to +* synthesize new events. */ +if (!strncmp (evsel->name,"branch-misses" ,13 )) { + spe->sample_branch_miss = true; + spe->branch_miss_id = evsel->core.id[0 ]; + +return 0 ; + } +/* create new id val to be a fixed offset from evsel id */ id = evsel->core.id[0 ] +1000000000 ;
tools/perf/util/arm-spe.h +3 -0
@@ -20,6 +20,8 @@ enum {union perf_event ;struct perf_session ;struct perf_pmu ; +struct evlist ; +struct evsel ;struct auxtrace_record *arm_spe_recording_init (int *err,struct perf_pmu *arm_spe_pmu);
@@ -28,4 +30,5 @@ int arm_spe_process_auxtrace_info(union perf_event *event, struct perf_session *session);struct perf_event_attr *arm_spe_pmu_default_config (struct perf_pmu *arm_spe_pmu); +void arm_spe_precise_ip_support (struct evlist *evlist,struct evsel *evsel);#endif
tools/perf/util/auxtrace.h +6 -0
@@ -584,6 +584,7 @@ void auxtrace__dump_auxtrace_sample(struct perf_session *session,int auxtrace__flush_events (struct perf_session *session,struct perf_tool *tool);void auxtrace__free_events (struct perf_session *session);void auxtrace__free (struct perf_session *session); +void auxtrace__preprocess_evlist (struct evlist *evlist) __attribute__((weak));#define ITRACE_HELP \ " i: synthesize instructions events\n" \
@@ -731,6 +732,11 @@ void auxtrace__free(struct perf_session *session __maybe_unused) { } +static inline +void auxtrace__preprocess_evlist (struct evlist *evlist__maybe_unused ) +{ +} +static inline int auxtrace_index__write (int fd__maybe_unused ,struct list_head *head__maybe_unused )
Signature · 1 line
2.17.1
Mark Rutland
reviewerRe: [PATCH v5 4/4] perf tools: Support "branch-misses:pp" on arm64
Hi James,
Sorry, I missed this v5 when replying to v4 just now, but my comments there equally apply here: I don't think that we should be silently overriding the event requested by the user, and I think that we can make that request explicit without being too painful for the user.
Thanks,
Mark.
Quoted 201 lines · James Clark
From: Tan Xiaojun <[email protected]>
At the suggestion of James Clark, use spe to support the precise ip of some events. Currently its support event is:
branch-misses.Example usage:
$ ./perf record -e branch-misses:pp dd if=/dev/zero of=/dev/null count=10000 (:p/pp/ppp is same for this case.)$ ./perf report --stdio ("--stdio is not necessary")--------------------------------------------------------------------
...
# Samples: 14 of event 'branch-misses:pp' # Event count (approx.): 14
## Children Self Command Shared Object Symbol # ........ ........ ....... ................. .......................... # 14.29% 14.29% dd [kernel.kallsyms] [k] __arch_copy_from_user 14.29% 14.29% dd libc-2.28.so [.] _dl_addr 7.14% 7.14% dd [kernel.kallsyms] [k] __free_pages 7.14% 7.14% dd [kernel.kallsyms] [k] __pi_memcpy 7.14% 7.14% dd [kernel.kallsyms] [k] pagecache_get_page 7.14% 7.14% dd [kernel.kallsyms] [k] unmap_single_vma 7.14% 7.14% dd dd [.] 0x00000000000025ec 7.14% 7.14% dd ld-2.28.so [.] _dl_lookup_symbol_x 7.14% 7.14% dd ld-2.28.so [.] check_match 7.14% 7.14% dd libc-2.28.so [.] __mpn_rshift 7.14% 7.14% dd libc-2.28.so [.] _nl_intern_locale_data 7.14% 7.14% dd libc-2.28.so [.] read_alias_file...
--------------------------------------------------------------------Signed-off-by Tan Xiaojun <[email protected]>Suggested-by James Clark <[email protected]>Tested-by Qi Liu <[email protected]>Signed-off-by James Clark <[email protected]>Cc Will Deacon <[email protected]>Cc Mark Rutland <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Ingo Molnar <[email protected]>Cc Arnaldo Carvalho de Melo <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Jiri Olsa <[email protected]>Cc Tan Xiaojun <[email protected]>Cc Al Grant <[email protected]>Cc Namhyung Kim <[email protected]>--- tools/perf/arch/arm/util/auxtrace.c | 39 +++++++++++++++++++++++++++++ tools/perf/builtin-record.c | 5 ++++ tools/perf/util/arm-spe.c | 9 +++++++ tools/perf/util/arm-spe.h | 3 +++ tools/perf/util/auxtrace.h | 6 +++++ 5 files changed, 62 insertions(+)tools/perf/arch/arm/util/auxtrace.c +39
-0@@ -10,11 +10,25 @@#include "../../util/auxtrace.h" #include "../../util/debug.h" +#include "../../util/env.h" #include "../../util/evlist.h" #include "../../util/pmu.h" #include "cs-etm.h" #include "arm-spe.h" +#define SPE_ATTR_TS_ENABLE BIT(0) +#define SPE_ATTR_PA_ENABLE BIT(1) +#define SPE_ATTR_PCT_ENABLE BIT(2) +#define SPE_ATTR_JITTER BIT(16) +#define SPE_ATTR_BRANCH_FILTER BIT(32) +#define SPE_ATTR_LOAD_FILTER BIT(33) +#define SPE_ATTR_STORE_FILTER BIT(34) + +#define SPE_ATTR_EV_RETIRED BIT(1) +#define SPE_ATTR_EV_CACHE BIT(3) +#define SPE_ATTR_EV_TLB BIT(5) +#define SPE_ATTR_EV_BRANCH BIT(7) +static struct perf_pmu **find_all_arm_spe_pmus (int *nr_spes,int *err) {struct perf_pmu **arm_spe_pmus =NULL ;@@ -108,3 +122,28 @@ struct auxtrace_record *err =0 ;return NULL ; } + +void auxtrace__preprocess_evlist (struct evlist *evlist) +{ +struct evsel *evsel; +struct perf_pmu *pmu; + +evlist__for_each_entry (evlist, evsel) { +/* Currently only supports precise_ip for branch-misses on arm64 */ +if (!strcmp (perf_env__arch (evlist->env),"arm64" ) + && evsel->core.attr.config ==PERF_COUNT_HW_BRANCH_MISSES + && evsel->core.attr.type ==PERF_TYPE_HARDWARE + && evsel->core.attr.precise_ip) + { + pmu =perf_pmu__find ("arm_spe_0" ); +if (pmu) { + evsel->pmu_name = pmu->name; + evsel->core.attr.type = pmu->type; + evsel->core.attr.config =SPE_ATTR_TS_ENABLE + |SPE_ATTR_BRANCH_FILTER ; + evsel->core.attr.config1 =SPE_ATTR_EV_BRANCH ; + evsel->core.attr.precise_ip =0 ; + } + } + } +}tools/perf/builtin-record.c +5
-0@@ -2451,6 +2451,11 @@ int cmd_record(int argc, const char **argv) argc =parse_options (argc, argv, record_options, record_usage,PARSE_OPT_STOP_AT_NON_OPTION ); + +if (auxtrace__preprocess_evlist) { +auxtrace__preprocess_evlist (rec->evlist); + } +if (quiet)perf_quiet_option ();tools/perf/util/arm-spe.c +9
-0@@ -778,6 +778,15 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session) attr.sample_id_all = evsel->core.attr.sample_id_all; attr.read_format = evsel->core.attr.read_format; +/* If it is in the precise ip mode, there is no need to +* synthesize new events. */ +if (!strncmp (evsel->name,"branch-misses" ,13 )) { + spe->sample_branch_miss = true; + spe->branch_miss_id = evsel->core.id[0 ]; + +return 0 ; + } +/* create new id val to be a fixed offset from evsel id */ id = evsel->core.id[0 ] +1000000000 ;tools/perf/util/arm-spe.h +3
-0@@ -20,6 +20,8 @@ enum {union perf_event ;struct perf_session ;struct perf_pmu ; +struct evlist ; +struct evsel ;struct auxtrace_record *arm_spe_recording_init (int *err,struct perf_pmu *arm_spe_pmu);@@ -28,4 +30,5 @@ int arm_spe_process_auxtrace_info(union perf_event *event, struct perf_session *session);struct perf_event_attr *arm_spe_pmu_default_config (struct perf_pmu *arm_spe_pmu); +void arm_spe_precise_ip_support (struct evlist *evlist,struct evsel *evsel);#endif tools/perf/util/auxtrace.h +6
-0@@ -584,6 +584,7 @@ void auxtrace__dump_auxtrace_sample(struct perf_session *session,int auxtrace__flush_events (struct perf_session *session,struct perf_tool *tool);void auxtrace__free_events (struct perf_session *session);void auxtrace__free (struct perf_session *session); +void auxtrace__preprocess_evlist (struct evlist *evlist) __attribute__((weak));#define ITRACE_HELP \ " i: synthesize instructions events\n" \@@ -731,6 +732,11 @@ void auxtrace__free(struct perf_session *session __maybe_unused) { } +static inline +void auxtrace__preprocess_evlist (struct evlist *evlist__maybe_unused ) +{ +} +static inline int auxtrace_index__write (int fd__maybe_unused ,struct list_head *head__maybe_unused )Signature · 1 line
2.17.1
PATCH v4 4/4 perf tools: Support "branch-misses:pp" on arm64
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
At the suggestion of James Clark, use spe to support the precise ip of some events. Currently its support event is:
branch-misses.
Example usage:
$ ./perf record -e branch-misses:pp dd if=/dev/zero of=/dev/null count=10000 (:p/pp/ppp is same for this case.)
$ ./perf report --stdio
("--stdio is not necessary")--------------------------------------------------------------------
...
# Samples: 14 of event 'branch-misses:pp' # Event count (approx.): 14
#
# Children Self Command Shared Object Symbol
# ........ ........ ....... ................. ..........................
#
14.29% 14.29% dd [kernel.kallsyms] [k] __arch_copy_from_user
14.29% 14.29% dd libc-2.28.so [.] _dl_addr
7.14% 7.14% dd [kernel.kallsyms] [k] __free_pages
7.14% 7.14% dd [kernel.kallsyms] [k] __pi_memcpy
7.14% 7.14% dd [kernel.kallsyms] [k] pagecache_get_page
7.14% 7.14% dd [kernel.kallsyms] [k] unmap_single_vma
7.14% 7.14% dd dd [.] 0x00000000000025ec
7.14% 7.14% dd ld-2.28.so [.] _dl_lookup_symbol_x
7.14% 7.14% dd ld-2.28.so [.] check_match
7.14% 7.14% dd libc-2.28.so [.] __mpn_rshift
7.14% 7.14% dd libc-2.28.so [.] _nl_intern_locale_data
7.14% 7.14% dd libc-2.28.so [.] read_alias_file...
--------------------------------------------------------------------
--- tools/perf/arch/arm/util/auxtrace.c | 38 +++++++++++++++++++++++++++++ tools/perf/builtin-record.c | 5 ++++ tools/perf/util/arm-spe.c | 9 +++++++ tools/perf/util/arm-spe.h | 3 +++ tools/perf/util/auxtrace.h | 6 +++++ 5 files changed, 61 insertions(+)
tools/perf/arch/arm/util/auxtrace.c +38 -0
@@ -10,11 +10,25 @@#include "../../util/auxtrace.h" #include "../../util/debug.h" +#include "../../util/env.h" #include "../../util/evlist.h" #include "../../util/pmu.h" #include "cs-etm.h" #include "arm-spe.h" +#define SPE_ATTR_TS_ENABLE BIT(0) +#define SPE_ATTR_PA_ENABLE BIT(1) +#define SPE_ATTR_PCT_ENABLE BIT(2) +#define SPE_ATTR_JITTER BIT(16) +#define SPE_ATTR_BRANCH_FILTER BIT(32) +#define SPE_ATTR_LOAD_FILTER BIT(33) +#define SPE_ATTR_STORE_FILTER BIT(34) + +#define SPE_ATTR_EV_RETIRED BIT(1) +#define SPE_ATTR_EV_CACHE BIT(3) +#define SPE_ATTR_EV_TLB BIT(5) +#define SPE_ATTR_EV_BRANCH BIT(7) +static struct perf_pmu **find_all_arm_spe_pmus (int *nr_spes,int *err) {struct perf_pmu **arm_spe_pmus =NULL ;
@@ -108,3 +122,27 @@ struct auxtrace_record *err =0 ;return NULL ; } + +void auxtrace__preprocess_evlist (struct evlist *evlist) +{ +struct evsel *evsel; +struct perf_pmu *pmu; + +evlist__for_each_entry (evlist, evsel) { +/* Currently only supports precise_ip for branch-misses on arm64 */ +if (!strcmp (perf_env__arch (evlist->env),"arm64" ) + && evsel->core.attr.config ==PERF_COUNT_HW_BRANCH_MISSES + && evsel->core.attr.precise_ip) + { + pmu =perf_pmu__find ("arm_spe_0" ); +if (pmu) { + evsel->pmu_name = pmu->name; + evsel->core.attr.type = pmu->type; + evsel->core.attr.config =SPE_ATTR_TS_ENABLE + |SPE_ATTR_BRANCH_FILTER ; + evsel->core.attr.config1 =SPE_ATTR_EV_BRANCH ; + evsel->core.attr.precise_ip =0 ; + } + } + } +}
tools/perf/builtin-record.c +5 -0
@@ -2451,6 +2451,11 @@ int cmd_record(int argc, const char **argv) argc =parse_options (argc, argv, record_options, record_usage,PARSE_OPT_STOP_AT_NON_OPTION ); + +if (auxtrace__preprocess_evlist) { +auxtrace__preprocess_evlist (rec->evlist); + } +if (quiet)perf_quiet_option ();
tools/perf/util/arm-spe.c +9 -0
@@ -778,6 +778,15 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session) attr.sample_id_all = evsel->core.attr.sample_id_all; attr.read_format = evsel->core.attr.read_format; +/* If it is in the precise ip mode, there is no need to +* synthesize new events. */ +if (!strncmp (evsel->name,"branch-misses" ,13 )) { + spe->sample_branch_miss = true; + spe->branch_miss_id = evsel->core.id[0 ]; + +return 0 ; + } +/* create new id val to be a fixed offset from evsel id */ id = evsel->core.id[0 ] +1000000000 ;
tools/perf/util/arm-spe.h +3 -0
@@ -20,6 +20,8 @@ enum {union perf_event ;struct perf_session ;struct perf_pmu ; +struct evlist ; +struct evsel ;struct auxtrace_record *arm_spe_recording_init (int *err,struct perf_pmu *arm_spe_pmu);
@@ -28,4 +30,5 @@ int arm_spe_process_auxtrace_info(union perf_event *event, struct perf_session *session);struct perf_event_attr *arm_spe_pmu_default_config (struct perf_pmu *arm_spe_pmu); +void arm_spe_precise_ip_support (struct evlist *evlist,struct evsel *evsel);#endif
tools/perf/util/auxtrace.h +6 -0
@@ -584,6 +584,7 @@ void auxtrace__dump_auxtrace_sample(struct perf_session *session,int auxtrace__flush_events (struct perf_session *session,struct perf_tool *tool);void auxtrace__free_events (struct perf_session *session);void auxtrace__free (struct perf_session *session); +void auxtrace__preprocess_evlist (struct evlist *evlist) __attribute__((weak));#define ITRACE_HELP \ " i: synthesize instructions events\n" \
@@ -728,6 +729,11 @@ void auxtrace__free(struct perf_session *session __maybe_unused) { } +static inline +void auxtrace__preprocess_evlist (struct evlist *evlist__maybe_unused ) +{ +} +static inline int auxtrace_index__write (int fd__maybe_unused ,struct list_head *head__maybe_unused )
Signature · 1 line
2.17.1
Adrian Hunter
reviewerRe: [PATCH v4 4/4] perf tools: Support "branch-misses:pp" on arm64
Quoted 92 lines · James Clark
From: Tan Xiaojun <[email protected]>
At the suggestion of James Clark, use spe to support the precise ip of some events. Currently its support event is:
branch-misses.Example usage:
$ ./perf record -e branch-misses:pp dd if=/dev/zero of=/dev/null count=10000 (:p/pp/ppp is same for this case.)$ ./perf report --stdio ("--stdio is not necessary")--------------------------------------------------------------------
...
# Samples: 14 of event 'branch-misses:pp' # Event count (approx.): 14
## Children Self Command Shared Object Symbol # ........ ........ ....... ................. .......................... # 14.29% 14.29% dd [kernel.kallsyms] [k] __arch_copy_from_user 14.29% 14.29% dd libc-2.28.so [.] _dl_addr 7.14% 7.14% dd [kernel.kallsyms] [k] __free_pages 7.14% 7.14% dd [kernel.kallsyms] [k] __pi_memcpy 7.14% 7.14% dd [kernel.kallsyms] [k] pagecache_get_page 7.14% 7.14% dd [kernel.kallsyms] [k] unmap_single_vma 7.14% 7.14% dd dd [.] 0x00000000000025ec 7.14% 7.14% dd ld-2.28.so [.] _dl_lookup_symbol_x 7.14% 7.14% dd ld-2.28.so [.] check_match 7.14% 7.14% dd libc-2.28.so [.] __mpn_rshift 7.14% 7.14% dd libc-2.28.so [.] _nl_intern_locale_data 7.14% 7.14% dd libc-2.28.so [.] read_alias_file...
--------------------------------------------------------------------Signed-off-by Tan Xiaojun <[email protected]>Suggested-by James Clark <[email protected]>Tested-by Qi Liu <[email protected]>Signed-off-by James Clark <[email protected]>Cc Will Deacon <[email protected]>Cc Mark Rutland <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Ingo Molnar <[email protected]>Cc Arnaldo Carvalho de Melo <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Jiri Olsa <[email protected]>Cc Tan Xiaojun <[email protected]>Cc Al Grant <[email protected]>Cc Namhyung Kim <[email protected]>--- tools/perf/arch/arm/util/auxtrace.c | 38 +++++++++++++++++++++++++++++ tools/perf/builtin-record.c | 5 ++++ tools/perf/util/arm-spe.c | 9 +++++++ tools/perf/util/arm-spe.h | 3 +++ tools/perf/util/auxtrace.h | 6 +++++ 5 files changed, 61 insertions(+)tools/perf/arch/arm/util/auxtrace.c +14
-0@@ -10,11 +10,25 @@#include "../../util/auxtrace.h" #include "../../util/debug.h" +#include "../../util/env.h" #include "../../util/evlist.h" #include "../../util/pmu.h" #include "cs-etm.h" #include "arm-spe.h" +#define SPE_ATTR_TS_ENABLE BIT(0) +#define SPE_ATTR_PA_ENABLE BIT(1) +#define SPE_ATTR_PCT_ENABLE BIT(2) +#define SPE_ATTR_JITTER BIT(16) +#define SPE_ATTR_BRANCH_FILTER BIT(32) +#define SPE_ATTR_LOAD_FILTER BIT(33) +#define SPE_ATTR_STORE_FILTER BIT(34) + +#define SPE_ATTR_EV_RETIRED BIT(1) +#define SPE_ATTR_EV_CACHE BIT(3) +#define SPE_ATTR_EV_TLB BIT(5) +#define SPE_ATTR_EV_BRANCH BIT(7) +static struct perf_pmu **find_all_arm_spe_pmus (int *nr_spes,int *err) {struct perf_pmu **arm_spe_pmus =NULL ;@@ -108,3 +122,27 @@ struct auxtrace_record *err =0 ;return NULL ;
tools/perf/arch/arm/util/auxtrace.c struct auxtrace_record · line 133
Quoted 10 lines
tools/perf/arch/arm/util/auxtrace.c +9
-0} + +void auxtrace__preprocess_evlist (struct evlist *evlist) +{ +struct evsel *evsel; +struct perf_pmu *pmu; + +evlist__for_each_entry (evlist, evsel) { +/* Currently only supports precise_ip for branch-misses on arm64 */ + if (!strcmp(perf_env__arch(evlist->env), "arm64")
Isn't config ambiguous unless you also check type i.e.
&& evsel->core.attr.type == PERF_TYPE_HARDWARE
tools/perf/arch/arm/util/auxtrace.c struct auxtrace_record · line 138
Quoted 5 lines
tools/perf/arch/arm/util/auxtrace.c +5
-0+ && evsel->core.attr.config == PERF_COUNT_HW_BRANCH_MISSES + && evsel->core.attr.precise_ip) + { + pmu =perf_pmu__find ("arm_spe_0" ); +if (pmu) {
Changing the event seems a bit weird.
Quoted 91 lines
tools/perf/arch/arm/util/auxtrace.c +10
-0+ evsel->pmu_name = pmu->name; + evsel->core.attr.type = pmu->type; + evsel->core.attr.config =SPE_ATTR_TS_ENABLE + |SPE_ATTR_BRANCH_FILTER ; + evsel->core.attr.config1 =SPE_ATTR_EV_BRANCH ; + evsel->core.attr.precise_ip =0 ; + } + } + } +}tools/perf/builtin-record.c +5
-0@@ -2451,6 +2451,11 @@ int cmd_record(int argc, const char **argv) argc =parse_options (argc, argv, record_options, record_usage,PARSE_OPT_STOP_AT_NON_OPTION ); + +if (auxtrace__preprocess_evlist) { +auxtrace__preprocess_evlist (rec->evlist); + } +if (quiet)perf_quiet_option ();tools/perf/util/arm-spe.c +9
-0@@ -778,6 +778,15 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session) attr.sample_id_all = evsel->core.attr.sample_id_all; attr.read_format = evsel->core.attr.read_format; +/* If it is in the precise ip mode, there is no need to +* synthesize new events. */ +if (!strncmp (evsel->name,"branch-misses" ,13 )) { + spe->sample_branch_miss = true; + spe->branch_miss_id = evsel->core.id[0 ]; + +return 0 ; + } +/* create new id val to be a fixed offset from evsel id */ id = evsel->core.id[0 ] +1000000000 ;tools/perf/util/arm-spe.h +3
-0@@ -20,6 +20,8 @@ enum {union perf_event ;struct perf_session ;struct perf_pmu ; +struct evlist ; +struct evsel ;struct auxtrace_record *arm_spe_recording_init (int *err,struct perf_pmu *arm_spe_pmu);@@ -28,4 +30,5 @@ int arm_spe_process_auxtrace_info(union perf_event *event, struct perf_session *session);struct perf_event_attr *arm_spe_pmu_default_config (struct perf_pmu *arm_spe_pmu); +void arm_spe_precise_ip_support (struct evlist *evlist,struct evsel *evsel);#endif tools/perf/util/auxtrace.h +6
-0@@ -584,6 +584,7 @@ void auxtrace__dump_auxtrace_sample(struct perf_session *session,int auxtrace__flush_events (struct perf_session *session,struct perf_tool *tool);void auxtrace__free_events (struct perf_session *session);void auxtrace__free (struct perf_session *session); +void auxtrace__preprocess_evlist (struct evlist *evlist) __attribute__((weak));#define ITRACE_HELP \ " i: synthesize instructions events\n" \@@ -728,6 +729,11 @@ void auxtrace__free(struct perf_session *session __maybe_unused) { } +static inline +void auxtrace__preprocess_evlist (struct evlist *evlist__maybe_unused ) +{ +} +static inline int auxtrace_index__write (int fd__maybe_unused ,struct list_head *head__maybe_unused )
Re: [PATCH v4 4/4] perf tools: Support "branch-misses:pp" on arm64
Hi Adrian,
Quoted 106 lines · Adrian Hunter
Quoted 92 lines · James Clark
From: Tan Xiaojun <[email protected]>
At the suggestion of James Clark, use spe to support the precise ip of some events. Currently its support event is:
branch-misses.Example usage:
$ ./perf record -e branch-misses:pp dd if=/dev/zero of=/dev/null count=10000 (:p/pp/ppp is same for this case.)$ ./perf report --stdio ("--stdio is not necessary")--------------------------------------------------------------------
...
# Samples: 14 of event 'branch-misses:pp' # Event count (approx.): 14
## Children Self Command Shared Object Symbol # ........ ........ ....... ................. .......................... # 14.29% 14.29% dd [kernel.kallsyms] [k] __arch_copy_from_user 14.29% 14.29% dd libc-2.28.so [.] _dl_addr 7.14% 7.14% dd [kernel.kallsyms] [k] __free_pages 7.14% 7.14% dd [kernel.kallsyms] [k] __pi_memcpy 7.14% 7.14% dd [kernel.kallsyms] [k] pagecache_get_page 7.14% 7.14% dd [kernel.kallsyms] [k] unmap_single_vma 7.14% 7.14% dd dd [.] 0x00000000000025ec 7.14% 7.14% dd ld-2.28.so [.] _dl_lookup_symbol_x 7.14% 7.14% dd ld-2.28.so [.] check_match 7.14% 7.14% dd libc-2.28.so [.] __mpn_rshift 7.14% 7.14% dd libc-2.28.so [.] _nl_intern_locale_data 7.14% 7.14% dd libc-2.28.so [.] read_alias_file...
--------------------------------------------------------------------Signed-off-by Tan Xiaojun <[email protected]>Suggested-by James Clark <[email protected]>Tested-by Qi Liu <[email protected]>Signed-off-by James Clark <[email protected]>Cc Will Deacon <[email protected]>Cc Mark Rutland <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Ingo Molnar <[email protected]>Cc Arnaldo Carvalho de Melo <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Jiri Olsa <[email protected]>Cc Tan Xiaojun <[email protected]>Cc Al Grant <[email protected]>Cc Namhyung Kim <[email protected]>--- tools/perf/arch/arm/util/auxtrace.c | 38 +++++++++++++++++++++++++++++ tools/perf/builtin-record.c | 5 ++++ tools/perf/util/arm-spe.c | 9 +++++++ tools/perf/util/arm-spe.h | 3 +++ tools/perf/util/auxtrace.h | 6 +++++ 5 files changed, 61 insertions(+)tools/perf/arch/arm/util/auxtrace.c +14
-0@@ -10,11 +10,25 @@#include "../../util/auxtrace.h" #include "../../util/debug.h" +#include "../../util/env.h" #include "../../util/evlist.h" #include "../../util/pmu.h" #include "cs-etm.h" #include "arm-spe.h" +#define SPE_ATTR_TS_ENABLE BIT(0) +#define SPE_ATTR_PA_ENABLE BIT(1) +#define SPE_ATTR_PCT_ENABLE BIT(2) +#define SPE_ATTR_JITTER BIT(16) +#define SPE_ATTR_BRANCH_FILTER BIT(32) +#define SPE_ATTR_LOAD_FILTER BIT(33) +#define SPE_ATTR_STORE_FILTER BIT(34) + +#define SPE_ATTR_EV_RETIRED BIT(1) +#define SPE_ATTR_EV_CACHE BIT(3) +#define SPE_ATTR_EV_TLB BIT(5) +#define SPE_ATTR_EV_BRANCH BIT(7) +static struct perf_pmu **find_all_arm_spe_pmus (int *nr_spes,int *err) {struct perf_pmu **arm_spe_pmus =NULL ;@@ -108,3 +122,27 @@ struct auxtrace_record *err =0 ;return NULL ;tools/perf/arch/arm/util/auxtrace.c struct auxtrace_record · line 133
Quoted 10 lines
tools/perf/arch/arm/util/auxtrace.c +9
-0} + +void auxtrace__preprocess_evlist (struct evlist *evlist) +{ +struct evsel *evsel; +struct perf_pmu *pmu; + +evlist__for_each_entry (evlist, evsel) { +/* Currently only supports precise_ip for branch-misses on arm64 */ + if (!strcmp(perf_env__arch(evlist->env), "arm64")Isn't config ambiguous unless you also check type i.e.
Quoted 2 lines
&& evsel->core.attr.type == PERF_TYPE_HARDWARE
Yes you're right I will add this.
Quoted 8 lines
tools/perf/arch/arm/util/auxtrace.c struct auxtrace_record · line 138
Quoted 5 lines
tools/perf/arch/arm/util/auxtrace.c +5
-0+ && evsel->core.attr.config == PERF_COUNT_HW_BRANCH_MISSES + && evsel->core.attr.precise_ip) + { + pmu =perf_pmu__find ("arm_spe_0" ); +if (pmu) {Changing the event seems a bit weird.
This is because there is no support in the kernel for the precise_ip attribute on Arm. SPE can give you precise ip data for the same event, but changing the event is required.
Quoted 92 lines
Quoted 91 lines
tools/perf/arch/arm/util/auxtrace.c +10
-0+ evsel->pmu_name = pmu->name; + evsel->core.attr.type = pmu->type; + evsel->core.attr.config =SPE_ATTR_TS_ENABLE + |SPE_ATTR_BRANCH_FILTER ; + evsel->core.attr.config1 =SPE_ATTR_EV_BRANCH ; + evsel->core.attr.precise_ip =0 ; + } + } + } +}tools/perf/builtin-record.c +5
-0@@ -2451,6 +2451,11 @@ int cmd_record(int argc, const char **argv) argc =parse_options (argc, argv, record_options, record_usage,PARSE_OPT_STOP_AT_NON_OPTION ); + +if (auxtrace__preprocess_evlist) { +auxtrace__preprocess_evlist (rec->evlist); + } +if (quiet)perf_quiet_option ();tools/perf/util/arm-spe.c +9
-0@@ -778,6 +778,15 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session) attr.sample_id_all = evsel->core.attr.sample_id_all; attr.read_format = evsel->core.attr.read_format; +/* If it is in the precise ip mode, there is no need to +* synthesize new events. */ +if (!strncmp (evsel->name,"branch-misses" ,13 )) { + spe->sample_branch_miss = true; + spe->branch_miss_id = evsel->core.id[0 ]; + +return 0 ; + } +/* create new id val to be a fixed offset from evsel id */ id = evsel->core.id[0 ] +1000000000 ;tools/perf/util/arm-spe.h +3
-0@@ -20,6 +20,8 @@ enum {union perf_event ;struct perf_session ;struct perf_pmu ; +struct evlist ; +struct evsel ;struct auxtrace_record *arm_spe_recording_init (int *err,struct perf_pmu *arm_spe_pmu);@@ -28,4 +30,5 @@ int arm_spe_process_auxtrace_info(union perf_event *event, struct perf_session *session);struct perf_event_attr *arm_spe_pmu_default_config (struct perf_pmu *arm_spe_pmu); +void arm_spe_precise_ip_support (struct evlist *evlist,struct evsel *evsel);#endif tools/perf/util/auxtrace.h +6
-0@@ -584,6 +584,7 @@ void auxtrace__dump_auxtrace_sample(struct perf_session *session,int auxtrace__flush_events (struct perf_session *session,struct perf_tool *tool);void auxtrace__free_events (struct perf_session *session);void auxtrace__free (struct perf_session *session); +void auxtrace__preprocess_evlist (struct evlist *evlist) __attribute__((weak));#define ITRACE_HELP \ " i: synthesize instructions events\n" \@@ -728,6 +729,11 @@ void auxtrace__free(struct perf_session *session __maybe_unused) { } +static inline +void auxtrace__preprocess_evlist (struct evlist *evlist__maybe_unused ) +{ +} +static inline int auxtrace_index__write (int fd__maybe_unused ,struct list_head *head__maybe_unused )
Mark Rutland
reviewerRe: [PATCH v4 4/4] perf tools: Support "branch-misses:pp" on arm64
Hi James,
Quoted 121 lines · James Clark
Quoted 106 lines · Adrian Hunter
Quoted 92 lines · James Clark
From: Tan Xiaojun <[email protected]>
At the suggestion of James Clark, use spe to support the precise ip of some events. Currently its support event is:
branch-misses.Example usage:
$ ./perf record -e branch-misses:pp dd if=/dev/zero of=/dev/null count=10000 (:p/pp/ppp is same for this case.)$ ./perf report --stdio ("--stdio is not necessary")--------------------------------------------------------------------
...
# Samples: 14 of event 'branch-misses:pp' # Event count (approx.): 14
## Children Self Command Shared Object Symbol # ........ ........ ....... ................. .......................... # 14.29% 14.29% dd [kernel.kallsyms] [k] __arch_copy_from_user 14.29% 14.29% dd libc-2.28.so [.] _dl_addr 7.14% 7.14% dd [kernel.kallsyms] [k] __free_pages 7.14% 7.14% dd [kernel.kallsyms] [k] __pi_memcpy 7.14% 7.14% dd [kernel.kallsyms] [k] pagecache_get_page 7.14% 7.14% dd [kernel.kallsyms] [k] unmap_single_vma 7.14% 7.14% dd dd [.] 0x00000000000025ec 7.14% 7.14% dd ld-2.28.so [.] _dl_lookup_symbol_x 7.14% 7.14% dd ld-2.28.so [.] check_match 7.14% 7.14% dd libc-2.28.so [.] __mpn_rshift 7.14% 7.14% dd libc-2.28.so [.] _nl_intern_locale_data 7.14% 7.14% dd libc-2.28.so [.] read_alias_file...
--------------------------------------------------------------------Signed-off-by Tan Xiaojun <[email protected]>Suggested-by James Clark <[email protected]>Tested-by Qi Liu <[email protected]>Signed-off-by James Clark <[email protected]>Cc Will Deacon <[email protected]>Cc Mark Rutland <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Ingo Molnar <[email protected]>Cc Arnaldo Carvalho de Melo <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Jiri Olsa <[email protected]>Cc Tan Xiaojun <[email protected]>Cc Al Grant <[email protected]>Cc Namhyung Kim <[email protected]>--- tools/perf/arch/arm/util/auxtrace.c | 38 +++++++++++++++++++++++++++++ tools/perf/builtin-record.c | 5 ++++ tools/perf/util/arm-spe.c | 9 +++++++ tools/perf/util/arm-spe.h | 3 +++ tools/perf/util/auxtrace.h | 6 +++++ 5 files changed, 61 insertions(+)tools/perf/arch/arm/util/auxtrace.c +14
-0@@ -10,11 +10,25 @@#include "../../util/auxtrace.h" #include "../../util/debug.h" +#include "../../util/env.h" #include "../../util/evlist.h" #include "../../util/pmu.h" #include "cs-etm.h" #include "arm-spe.h" +#define SPE_ATTR_TS_ENABLE BIT(0) +#define SPE_ATTR_PA_ENABLE BIT(1) +#define SPE_ATTR_PCT_ENABLE BIT(2) +#define SPE_ATTR_JITTER BIT(16) +#define SPE_ATTR_BRANCH_FILTER BIT(32) +#define SPE_ATTR_LOAD_FILTER BIT(33) +#define SPE_ATTR_STORE_FILTER BIT(34) + +#define SPE_ATTR_EV_RETIRED BIT(1) +#define SPE_ATTR_EV_CACHE BIT(3) +#define SPE_ATTR_EV_TLB BIT(5) +#define SPE_ATTR_EV_BRANCH BIT(7) +static struct perf_pmu **find_all_arm_spe_pmus (int *nr_spes,int *err) {struct perf_pmu **arm_spe_pmus =NULL ;@@ -108,3 +122,27 @@ struct auxtrace_record *err =0 ;return NULL ;tools/perf/arch/arm/util/auxtrace.c struct auxtrace_record · line 133
Quoted 10 lines
tools/perf/arch/arm/util/auxtrace.c +9
-0} + +void auxtrace__preprocess_evlist (struct evlist *evlist) +{ +struct evsel *evsel; +struct perf_pmu *pmu; + +evlist__for_each_entry (evlist, evsel) { +/* Currently only supports precise_ip for branch-misses on arm64 */ + if (!strcmp(perf_env__arch(evlist->env), "arm64")Isn't config ambiguous unless you also check type i.e.
Quoted 2 lines
&& evsel->core.attr.type == PERF_TYPE_HARDWAREYes you're right I will add this.
Quoted 8 lines
tools/perf/arch/arm/util/auxtrace.c struct auxtrace_record · line 138
Quoted 5 lines
tools/perf/arch/arm/util/auxtrace.c +5
-0+ && evsel->core.attr.config == PERF_COUNT_HW_BRANCH_MISSES + && evsel->core.attr.precise_ip) + { + pmu =perf_pmu__find ("arm_spe_0" ); +if (pmu) {Changing the event seems a bit weird.
Quoted 2 lines
This is because there is no support in the kernel for the
precise_ipattribute on Arm. SPE can give you precise ip data for the same event, but changing the event is required.
I don't think this is the right level to override the event.
It's true that contemporary CPU PMUs can't generate synchronous events, and hence the kernel doesn't have a precise IP, but that's not necessarily going to be the case in future. I'd rather we didn't silently override the event requested by the user, as I think that is going to cause more problems for us in future.
When the SPE buffer overflows, events will be silently dropped, which I don't believe is in keeping with the usual semantics of precise events. Additionally, hard-coding the "arm_spe_0" name means that this will not work reliably on big.LITTLE systems.
Instead, can we have the user explicitly request to use the SPE PMU in this way? If the perf tool could be smart with the "_%d" suffix, and collate PMUs differing only by that, the user would only need to do something like:
perf record -e arm_spe/branch-misses/pp
... which doesn't seem to onerous.
Is something like that possible?
Thanks,
Mark.
v6 cover perf tools: Add support for some spe events
Hi Mark,
Yes I think this is something I can look into. For now I have removed that last patch because the current patch set already works very similarly anyway and allows people to use SPE in perf:
./perf record -e arm_spe_0/branch_filter=1/
vs
./perf record -e arm_spe/branch-misses/pp
Also I don't have access to any big.LITTLE hardware with SPE so wouldn't be able to test collating all the SPE PMUs.
Thanks
James
Tan Xiaojun (3):
perf tools: Move arm-spe-pkt-decoder.h/c to the new dir perf tools: Add support for "report" for some spe events perf report: Add SPE options to --itrace argument
tools/perf/Documentation/itrace.txt | 5 +-
tools/perf/util/Build | 2 +-
tools/perf/util/arm-spe-decoder/Build | 1 +
.../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++
.../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++
.../arm-spe-pkt-decoder.c | 0
.../arm-spe-pkt-decoder.h | 2 +
tools/perf/util/arm-spe.c | 747 +++++++++++++++++-
tools/perf/util/auxtrace.c | 13 +
tools/perf/util/auxtrace.h | 13 +-
10 files changed, 1032 insertions(+), 42 deletions(-)
create mode 100644 tools/perf/util/arm-spe-decoder/Build
create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c
create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%)
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (96%)Signature · 1 line
2.17.1
PATCH v6 1/3 perf tools: Move arm-spe-pkt-decoder.h/c to the new dir
From: Tan Xiaojun <[email protected]>
Create a new arm-spe-decoder directory for subsequent extensions and move arm-spe-pkt-decoder.h/c to this directory. No code changes.
---
tools/perf/util/Build | 2 +-
tools/perf/util/arm-spe-decoder/Build | 1 +
tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c | 0
tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h | 0
tools/perf/util/arm-spe.c | 2 +-
5 files changed, 3 insertions(+), 2 deletions(-)
create mode 100644 tools/perf/util/arm-spe-decoder/Build
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%)
rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (100%)tools/perf/util/Build +1 -1
@@ -104,7 +104,7 @@ perf-$(CONFIG_AUXTRACE) += intel-pt-decoder/ perf-$(CONFIG_AUXTRACE) += intel-pt.o perf-$(CONFIG_AUXTRACE) += intel-bts.o perf-$(CONFIG_AUXTRACE) += arm-spe.o -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-decoder/ perf-$(CONFIG_AUXTRACE) += s390-cpumsf.o ifdef CONFIG_LIBOPENCSD
tools/perf/util/arm-spe-decoder/Build new file +1 -0
@@ -0,0 +1 @@ +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.c renamed from tools/perf/util/arm-spe-pkt-decoder.c +0 -0
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h renamed from tools/perf/util/arm-spe-pkt-decoder.h +0 -0
tools/perf/util/arm-spe.c +1 -1
@@ -23,7 +23,7 @@#include "debug.h" #include "auxtrace.h" #include "arm-spe.h" -#include "arm-spe-pkt-decoder.h" +#include "arm-spe-decoder/arm-spe-pkt-decoder.h" struct arm_spe {struct auxtrace auxtrace;
Signature · 1 line
2.17.1
PATCH v6 2/3 perf tools: Add support for "report" for some spe events
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
After the commit ffd3d18c20b8 ("perf tools: Add ARM Statistical Profiling Extensions (SPE) support") is merged, "perf record" and "perf report --dump-raw-trace" have been supported. However, the raw data that is dumped cannot be used without parsing.
This patch is to improve the "perf report" support for spe, and further process the data. Currently, support for the four events of llc-miss, tlb-miss, branch-miss, and remote-access is added.
Example usage:
$ ./perf record -c 1024 -e arm_spe_0/branch_filter=1,ts_enable=1,pct_enable=1,pa_enable=1,load_filter=1,jitter=1,store_filter=1,min_latency=0/ -o perf-armspe-dd.data dd if=/dev/zero of=/dev/null count=10000
$ ./perf report -i perf-armspe-dd.data --stdio
--------------------------------------------------------------------
...
# Samples: 23 of event 'llc-miss'
# Event count (approx.): 23
...
33.33% 33.33% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64
12.12% 12.12% dd [kernel.kallsyms] [k] perf_event_mmap
6.06% 6.06% dd [kernel.kallsyms] [k] copy_page
6.06% 6.06% dd ld-2.28.so [.] _dl_relocate_object
3.03% 3.03% dd [kernel.kallsyms] [k] change_protection_range
3.03% 3.03% dd [kernel.kallsyms] [k] filemap_map_pages
3.03% 3.03% dd [kernel.kallsyms] [k] free_pages_and_swap_cache
3.03% 3.03% dd [kernel.kallsyms] [k] generic_permission
3.03% 3.03% dd [kernel.kallsyms] [k] kmem_cache_alloc
3.03% 3.03% dd [kernel.kallsyms] [k] lookup_fast
3.03% 3.03% dd [kernel.kallsyms] [k] perf_event_exec
3.03% 3.03% dd [kernel.kallsyms] [k] radix_tree_next_chunk
3.03% 3.03% dd [kernel.kallsyms] [k] ring_buffer_record_is_on
3.03% 3.03% dd ld-2.28.so [.] _dl_lookup_symbol_x
3.03% 3.03% dd ld-2.28.so [.] _dl_start
3.03% 3.03% dd ld-2.28.so [.] dl_main
3.03% 3.03% dd ld-2.28.so [.] strcmp
3.03% 3.03% dd libc-2.28.so [.] _dl_addr
...
# Samples: 3 of event 'tlb-miss'
# Event count (approx.): 3
...
33.33% 33.33% dd [kernel.kallsyms] [k] filemap_map_pages
33.33% 33.33% dd ld-2.28.so [.] _dl_start
33.33% 33.33% dd ld-2.28.so [.] dl_main
...
# Samples: 20 of event 'branch-miss'
# Event count (approx.): 20
...
15.38% 15.38% dd [kernel.kallsyms] [k] __fput
7.69% 7.69% dd [kernel.kallsyms] [k] do_el0_ia_bp_hardening
7.69% 7.69% dd [kernel.kallsyms] [k] filemap_map_pages
7.69% 7.69% dd [kernel.kallsyms] [k] pagevec_lru_move_fn
7.69% 7.69% dd [kernel.kallsyms] [k] perf_event_mmap_output
7.69% 7.69% dd [kernel.kallsyms] [k] task_work_run
7.69% 7.69% dd [kernel.kallsyms] [k] unmap_single_vma
7.69% 7.69% dd libc-2.28.so [.] _IO_flush_all_lockp
7.69% 7.69% dd libc-2.28.so [.] __memcpy_generic
7.69% 7.69% dd libc-2.28.so [.] _dl_addr
7.69% 7.69% dd libc-2.28.so [.] msort_with_tmp.part.0
7.69% 7.69% dd libc-2.28.so [.] read_alias_file
...
# Samples: 5 of event 'remote-access'
# Event count (approx.): 5
...
27.78% 27.78% dd [kernel.kallsyms] [k] perf_iterate_ctx.constprop.64
16.67% 16.67% dd [kernel.kallsyms] [k] perf_event_mmap
5.56% 5.56% dd [kernel.kallsyms] [k] change_protection_range
5.56% 5.56% dd [kernel.kallsyms] [k] filemap_map_pages
5.56% 5.56% dd [kernel.kallsyms] [k] free_pages_and_swap_cache
5.56% 5.56% dd [kernel.kallsyms] [k] generic_permission
5.56% 5.56% dd [kernel.kallsyms] [k] lookup_fast
5.56% 5.56% dd [kernel.kallsyms] [k] perf_event_exec
5.56% 5.56% dd [kernel.kallsyms] [k] radix_tree_next_chunk
5.56% 5.56% dd ld-2.28.so [.] _dl_relocate_object
5.56% 5.56% dd ld-2.28.so [.] _dl_start
5.56% 5.56% dd ld-2.28.so [.] dl_main--------------------------------------------------------------------
After that, more analysis and processing of the raw data of spe will be done.
--- tools/perf/util/arm-spe-decoder/Build | 2 +- .../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++ .../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++ .../arm-spe-decoder/arm-spe-pkt-decoder.h | 2 + tools/perf/util/arm-spe.c | 745 +++++++++++++++++- tools/perf/util/auxtrace.c | 13 + tools/perf/util/auxtrace.h | 8 +- 7 files changed, 1022 insertions(+), 39 deletions(-) create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h
tools/perf/util/arm-spe-decoder/Build +1 -1
@@ -1 +1 @@ -perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o +perf-$(CONFIG_AUXTRACE) += arm-spe-pkt-decoder.o arm-spe-decoder.o
tools/perf/util/arm-spe-decoder/arm-spe-decoder.c new file +225 -0
225 lines not on this page · raw message
tools/perf/util/arm-spe-decoder/arm-spe-decoder.h new file +66 -0
66 lines not on this page · raw message
tools/perf/util/arm-spe-decoder/arm-spe-pkt-decoder.h +2 -0
@@ -15,6 +15,8 @@#define ARM_SPE_NEED_MORE_BYTES -1#define ARM_SPE_BAD_PACKET -2 +#define ARM_SPE_PKT_MAX_SZ 16 +enum arm_spe_pkt_type {ARM_SPE_BAD ,ARM_SPE_PAD ,
tools/perf/util/arm-spe.c +708 -37
819 lines not on this page · raw message
tools/perf/util/auxtrace.c +13 -0
@@ -1279,6 +1279,10 @@ void itrace_synth_opts__set_default(struct itrace_synth_opts *synth_opts, synth_opts->pwr_events = true; synth_opts->other_events = true; synth_opts->errors = true; + synth_opts->llc_miss = true; + synth_opts->tlb_miss = true; + synth_opts->remote_access = true; +if (no_sample) { synth_opts->period_type =PERF_ITRACE_PERIOD_INSTRUCTIONS ; synth_opts->period =1 ;
@@ -1431,6 +1435,15 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str, goto out_err; p = endptr;break ; +case 'm' : + synth_opts->llc_miss = true; +break ; +case 't' : + synth_opts->tlb_miss = true; +break ; +case 'a' : + synth_opts->remote_access = true; +break ;case ' ' :case ',' :break ;
tools/perf/util/auxtrace.h +7 -1
@@ -60,7 +60,7 @@ enum itrace_period_type {* @inject: indicates the event (not just the sample) must be fully synthesized * because 'perf inject' will write it out * @instructions: whether to synthesize 'instructions' events -* @branches: whether to synthesize 'branches' events +* @branches: whether to synthesize 'branches' events (branch misses only on Arm) * @transactions: whether to synthesize events for transactions * @ptwrites: whether to synthesize events for ptwrites * @pwr_events: whether to synthesize power events
@@ -74,6 +74,9 @@ enum itrace_period_type {* @callchain: add callchain to 'instructions' events * @thread_stack: feed branches to the thread_stack * @last_branch: add branch context to 'instruction' events +* @llc_miss: whether to synthesize last level cache miss events +* @tlb_miss: whether to synthesize TLB miss events +* @remote_access: whether to synthesize Remote access events * @callchain_sz: maximum callchain size * @last_branch_sz: branch context size * @period: 'instructions' events period
@@ -101,6 +104,9 @@ struct itrace_synth_opts {bool callchain;bool thread_stack;bool last_branch; +bool llc_miss; +bool tlb_miss; +bool remote_access;unsigned int callchain_sz;unsigned int last_branch_sz;unsigned long long period;
Signature · 1 line
2.17.1
PATCH v6 3/3 perf report: Add SPE options to --itrace argument
In reply to James ClarkFrom: Tan Xiaojun <[email protected]>
The previous patch added support in "perf report" for some arm-spe events(llc-miss, tlb-miss, branch-miss, remote_access). This patch adds their help instructions.
--- tools/perf/Documentation/itrace.txt | 5 ++++- tools/perf/util/auxtrace.h | 5 ++++- 2 files changed, 8 insertions(+), 2 deletions(-)
tools/perf/Documentation/itrace.txt +4 -1
@@ -1,5 +1,5 @@ i synthesize instructions events - b synthesize branches events + b synthesize branches events (branch misses on Arm) c synthesize branches events (calls only) r synthesize branches events (returns only) x synthesize transactions events
@@ -9,6 +9,9 @@ of aux-output (refer to perf record) e synthesize error events d create a debug log + m synthesize LLC miss events + t synthesize TLB miss events + a synthesize remote access events g synthesize a call chain (use with i or x) l synthesize last branch entries (use with i or x) s skip initial number of events
tools/perf/util/auxtrace.h +4 -1
@@ -587,7 +587,7 @@ void auxtrace__free(struct perf_session *session);#define ITRACE_HELP \ " i: synthesize instructions events\n" \ -" b: synthesize branches events\n" \ +" b: synthesize branches events (branch misses on Arm)\n" \ " c: synthesize branches events (calls only)\n" \ " r: synthesize branches events (returns only)\n" \ " x: synthesize transactions events\n" \
@@ -595,6 +595,9 @@ void auxtrace__free(struct perf_session *session);" p: synthesize power events\n" \" e: synthesize error events\n" \" d: create a debug log\n" \ +" m: synthesize LLC miss events\n" \ +" t: synthesize TLB miss events\n" \ +" a: synthesize remote access events\n" \" g[len]: synthesize a call chain (use with i or x)\n" \" l[len]: synthesize last branch entries (use with i or x)\n" \" sNUMBER: skip initial number of events\n" \
Signature · 1 line
2.17.1
Re: [PATCH v6 3/3] perf report: Add SPE options to --itrace argument
Hi James,
Quoted 29 lines · James Clark
From: Tan Xiaojun <[email protected]>
The previous patch added support in "perf report" for some arm-spe
events(llc-miss, tlb-miss, branch-miss, remote_access). This patch adds their help instructions.Signed-off-by Tan Xiaojun <[email protected]>Tested-by Qi Liu <[email protected]>Signed-off-by James Clark <[email protected]>Cc Will Deacon <[email protected]>Cc Mark Rutland <[email protected]>Cc Peter Zijlstra <[email protected]>Cc Ingo Molnar <[email protected]>Cc Arnaldo Carvalho de Melo <[email protected]>Cc Alexander Shishkin <[email protected]>Cc Jiri Olsa <[email protected]>Cc Tan Xiaojun <[email protected]>Cc Al Grant <[email protected]>Cc Namhyung Kim <[email protected]>--- tools/perf/Documentation/itrace.txt | 5 ++++- tools/perf/util/auxtrace.h | 5 ++++- 2 files changed, 8 insertions(+), 2 deletions(-)tools/perf/Documentation/itrace.txt +0
-0
tools/perf/Documentation/itrace.txt line 2
Quoted 3 lines
tools/perf/Documentation/itrace.txt +1
-1@@ -1,5 +1,5 @@ i synthesize instructions events - b synthesize branches events + b synthesize branches events (branch misses on Arm)
This is not valid for Arm CoreSight actually. Arm CoreSight can use option 'b' to inject branch samples. For this reason, suggest to change as "(branch misses for Arm SPE)".
Thanks,
Leo
Quoted 39 lines
tools/perf/Documentation/itrace.txt +3
-0c synthesize branches events (calls only) r synthesize branches events (returns only) x synthesize transactions events@@ -9,6 +9,9 @@ of aux-output (refer to perf record) e synthesize error events d create a debug log + m synthesize LLC miss events + t synthesize TLB miss events + a synthesize remote access events g synthesize a call chain (use with i or x) l synthesize last branch entries (use with i or x) s skip initial number of eventstools/perf/util/auxtrace.h +4
-1@@ -587,7 +587,7 @@ void auxtrace__free(struct perf_session *session);#define ITRACE_HELP \ " i: synthesize instructions events\n" \ -" b: synthesize branches events\n" \ +" b: synthesize branches events (branch misses on Arm)\n" \ " c: synthesize branches events (calls only)\n" \ " r: synthesize branches events (returns only)\n" \ " x: synthesize transactions events\n" \@@ -595,6 +595,9 @@ void auxtrace__free(struct perf_session *session);" p: synthesize power events\n" \" e: synthesize error events\n" \" d: create a debug log\n" \ +" m: synthesize LLC miss events\n" \ +" t: synthesize TLB miss events\n" \ +" a: synthesize remote access events\n" \" g[len]: synthesize a call chain (use with i or x)\n" \" l[len]: synthesize last branch entries (use with i or x)\n" \" sNUMBER: skip initial number of events\n" \Signature · 1 line
2.17.1
Re: [PATCH v6 0/3] perf tools: Add support for some spe events
In reply to James ClarkQuoted 9 lines · James Clark
Hi Mark,
Yes I think this is something I can look into. For now I have removed that last patch because the current patch set already works very similarly anyway and allows people to use SPE in perf:
./perf record -e arm_spe_0/branch_filter=1/vs
./perf record -e arm_spe/branch-misses/pp
Thanks, FWIW that looks fine to me.
Quoted 2 lines
Also I don't have access to any big.LITTLE hardware with SPE so wouldn't be able to test collating all the SPE PMUs.
Likewise, I just want to make sure we don't back ourselves into a corner.
Otherwise, I have no comments on these patches, so feel free to take that as:
Mark.
Quoted 29 lines
Thanks
JamesTan Xiaojun (3):
perf tools: Move arm-spe-pkt-decoder.h/c to the new dir perf tools: Add support for "report" for some spe events perf report: Add SPE options to --itrace argumenttools/perf/Documentation/itrace.txt | 5 +- tools/perf/util/Build | 2 +- tools/perf/util/arm-spe-decoder/Build | 1 + .../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++ .../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++ .../arm-spe-pkt-decoder.c | 0 .../arm-spe-pkt-decoder.h | 2 + tools/perf/util/arm-spe.c | 747 +++++++++++++++++- tools/perf/util/auxtrace.c | 13 + tools/perf/util/auxtrace.h | 13 +- 10 files changed, 1032 insertions(+), 42 deletions(-) create mode 100644 tools/perf/util/arm-spe-decoder/Build create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%) rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (96%)Signature · 1 line
2.17.1
Quoted 3 lines · James Clark
Hi Jirka,
Oops. I've removed all the changes to
evlist.candevsel.h
hi,
it looks ok from my POV, but I don't follow auxtrace that much
Adrian,
it's changing some generic bits of the auxtrace framework, could you please check?
thanks,
jirka
Quoted 33 lines
James
Tan Xiaojun (4):
perf tools: Move arm-spe-pkt-decoder.h/c to the new dir perf tools: Add support for "report" for some spe events perf report: Add SPE options to --itrace argument perf tools: Support "branch-misses:pp" on arm64tools/perf/Documentation/itrace.txt | 5 +- tools/perf/arch/arm/util/auxtrace.c | 38 + tools/perf/builtin-record.c | 5 + tools/perf/util/Build | 2 +- tools/perf/util/arm-spe-decoder/Build | 1 + .../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++ .../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++ .../arm-spe-pkt-decoder.c | 0 .../arm-spe-pkt-decoder.h | 2 + tools/perf/util/arm-spe.c | 756 +++++++++++++++++- tools/perf/util/arm-spe.h | 3 + tools/perf/util/auxtrace.c | 13 + tools/perf/util/auxtrace.h | 14 +- 13 files changed, 1089 insertions(+), 41 deletions(-) create mode 100644 tools/perf/util/arm-spe-decoder/Build create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%) rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (96%)Signature · 1 line
2.17.1
Adrian Hunter
reviewerQuoted 11 lines · Jiri Olsa
Quoted 3 lines · James Clark
Hi Jirka,
Oops. I've removed all the changes to
evlist.candevsel.hhi,
it looks ok from my POV, but I don't follow auxtrace that muchAdrian,
it's changing some generic bits of the auxtrace framework, could you please check?
Sure, in the next few days.
Quoted 38 lines
thanks,
jirkaQuoted 33 lines
James
Tan Xiaojun (4):
perf tools: Move arm-spe-pkt-decoder.h/c to the new dir perf tools: Add support for "report" for some spe events perf report: Add SPE options to --itrace argument perf tools: Support "branch-misses:pp" on arm64tools/perf/Documentation/itrace.txt | 5 +- tools/perf/arch/arm/util/auxtrace.c | 38 + tools/perf/builtin-record.c | 5 + tools/perf/util/Build | 2 +- tools/perf/util/arm-spe-decoder/Build | 1 + .../util/arm-spe-decoder/arm-spe-decoder.c | 225 ++++++ .../util/arm-spe-decoder/arm-spe-decoder.h | 66 ++ .../arm-spe-pkt-decoder.c | 0 .../arm-spe-pkt-decoder.h | 2 + tools/perf/util/arm-spe.c | 756 +++++++++++++++++- tools/perf/util/arm-spe.h | 3 + tools/perf/util/auxtrace.c | 13 + tools/perf/util/auxtrace.h | 14 +- 13 files changed, 1089 insertions(+), 41 deletions(-) create mode 100644 tools/perf/util/arm-spe-decoder/Build create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.c create mode 100644 tools/perf/util/arm-spe-decoder/arm-spe-decoder.h rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.c (100%) rename tools/perf/util/{ => arm-spe-decoder}/arm-spe-pkt-decoder.h (96%)Signature · 1 line
2.17.1
PATCH v2 6/7 perf tools: arm-spe: fix record hang after being terminated
In reply to James ClarkFrom: Wei Li <[email protected]>
If the spe event is terminated, we don't enable it again here.
--- tools/perf/arch/arm64/util/arm-spe.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-)
tools/perf/arch/arm64/util/arm-spe.c +7 -3
@@ -165,9 +165,13 @@ static int arm_spe_read_finish(struct auxtrace_record *itr, int idx)struct evsel *evsel;evlist__for_each_entry (sper->evlist, evsel) { -if (evsel->core.attr.type == sper->arm_spe_pmu->type) -return perf_evlist__enable_event_idx (sper->evlist, - evsel, idx); +if (evsel->core.attr.type == sper->arm_spe_pmu->type) { +if (evsel->terminated) +return 0 ; +else +return perf_evlist__enable_event_idx ( + sper->evlist, evsel, idx); + } }return -EINVAL ; }
Signature · 1 line
2.25.0
PATCH v2 7/7 perf tools: Unset precise_ip when using SPE
In reply to James Clarkprecise_ip is not supported on Arm and the kernel may be updated to reflect this. So unset it when we know we can use SPE to get precise data instead.
--- tools/perf/util/arm-spe.c | 1 + 1 file changed, 1 insertion(+)
tools/perf/util/arm-spe.c +1 -0
@@ -937,6 +937,7 @@ void arm_spe_precise_ip_support(struct evlist *evlist, struct evsel *evsel) evsel->core.attr.config =SPE_ATTR_TS_ENABLE |SPE_ATTR_BRANCH_FILTER ; evsel->core.attr.config1 =SPE_ATTR_EV_BRANCH ; + evsel->core.attr.precise_ip =0 ; } } }
Signature · 1 line
2.25.0