[PATCH V2 1/3] perf parse-regs: Split parse_regs

[PATCH V2 1/3] perf parse-regs: Split parse_regs

[kan.liang]

From: Kan Liang <kan.liang@linux.intel.com>

The available registers for --int-regs and --user-regs may be different,
e.g. XMM registers.

Split parse_regs into two dedicated functions for --int-regs and
--user-regs respectively.

Modify the warning message. "--user-regs=?" should be applied to show
the available registers for --user-regs.

Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
---

No changes since V1

 tools/perf/builtin-record.c          |  4 ++--
 tools/perf/util/parse-regs-options.c | 19 ++++++++++++++++---
 tools/perf/util/parse-regs-options.h |  3 ++-
 3 files changed, 20 insertions(+), 6 deletions(-)

diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 386e665..45172bb 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -2029,10 +2029,10 @@ static struct option __record_options[] = {
 		    "use per-thread mmaps"),
 	OPT_CALLBACK_OPTARG('I', "intr-regs", &record.opts.sample_intr_regs, NULL, "any register",
 		    "sample selected machine registers on interrupt,"
-		    " use '-I?' to list register names", parse_regs),
+		    " use '-I?' to list register names", parse_intr_regs),
 	OPT_CALLBACK_OPTARG(0, "user-regs", &record.opts.sample_user_regs, NULL, "any register",
 		    "sample selected machine registers on interrupt,"
-		    " use '-I?' to list register names", parse_regs),
+		    " use '--user-regs=?' to list register names", parse_user_regs),
 	OPT_BOOLEAN(0, "running-time", &record.opts.running_time,
 		    "Record running/enabled time of read (:S) events"),
 	OPT_CALLBACK('k', "clockid", &record.opts,
diff --git a/tools/perf/util/parse-regs-options.c b/tools/perf/util/parse-regs-options.c
index 9cb187a..b21617f 100644
--- a/tools/perf/util/parse-regs-options.c
+++ b/tools/perf/util/parse-regs-options.c
@@ -5,8 +5,8 @@
 #include <subcmd/parse-options.h>
 #include "util/parse-regs-options.h"
 
-int
-parse_regs(const struct option *opt, const char *str, int unset)
+static int
+__parse_regs(const struct option *opt, const char *str, int unset, bool intr)
 {
 	uint64_t *mode = (uint64_t *)opt->value;
 	const struct sample_reg *r;
@@ -48,7 +48,8 @@ parse_regs(const struct option *opt, const char *str, int unset)
 					break;
 			}
 			if (!r->name) {
-				ui__warning("Unknown register \"%s\", check man page or run \"perf record -I?\"\n", s);
+				ui__warning("Unknown register \"%s\", check man page or run \"perf record %s?\"\n",
+					    s, intr ? "-I" : "--user-regs=");
 				goto error;
 			}
 
@@ -69,3 +70,15 @@ parse_regs(const struct option *opt, const char *str, int unset)
 	free(os);
 	return ret;
 }
+
+int
+parse_user_regs(const struct option *opt, const char *str, int unset)
+{
+	return __parse_regs(opt, str, unset, false);
+}
+
+int
+parse_intr_regs(const struct option *opt, const char *str, int unset)
+{
+	return __parse_regs(opt, str, unset, true);
+}
diff --git a/tools/perf/util/parse-regs-options.h b/tools/perf/util/parse-regs-options.h
index cdefb1a..2b23d25 100644
--- a/tools/perf/util/parse-regs-options.h
+++ b/tools/perf/util/parse-regs-options.h
@@ -2,5 +2,6 @@
 #ifndef _PERF_PARSE_REGS_OPTIONS_H
 #define _PERF_PARSE_REGS_OPTIONS_H 1
 struct option;
-int parse_regs(const struct option *opt, const char *str, int unset);
+int parse_user_regs(const struct option *opt, const char *str, int unset);
+int parse_intr_regs(const struct option *opt, const char *str, int unset);
 #endif /* _PERF_PARSE_REGS_OPTIONS_H */
-- 
2.7.4

[PATCH V2 2/3] perf parse-regs: Add generic support for arch__intr/user_reg_mask()

[kan.liang]

From: Kan Liang <kan.liang@linux.intel.com>

There may be different register mask for use with intr or user on some
platforms, e.g. Icelake.

Add weak functions arch__intr_reg_mask() and arch__user_reg_mask() to
return intr and user register mask respectively.

Check mask before printing or comparing the register name.

Generic code always return PERF_REGS_MASK. No functional change.

Suggested-by: Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com>
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
---

Changes since V1:
- Drop has_non_gprs_support() and non_gprs_mask()
  Use arch__intr/user_reg_mask()

 tools/perf/util/parse-regs-options.c | 13 ++++++++++---
 tools/perf/util/perf_regs.c          | 10 ++++++++++
 tools/perf/util/perf_regs.h          |  2 ++
 3 files changed, 22 insertions(+), 3 deletions(-)

diff --git a/tools/perf/util/parse-regs-options.c b/tools/perf/util/parse-regs-options.c
index b21617f..08581e2 100644
--- a/tools/perf/util/parse-regs-options.c
+++ b/tools/perf/util/parse-regs-options.c
@@ -12,6 +12,7 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)
 	const struct sample_reg *r;
 	char *s, *os = NULL, *p;
 	int ret = -1;
+	uint64_t mask;
 
 	if (unset)
 		return 0;
@@ -22,6 +23,11 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)
 	if (*mode)
 		return -1;
 
+	if (intr)
+		mask = arch__intr_reg_mask();
+	else
+		mask = arch__user_reg_mask();
+
 	/* str may be NULL in case no arg is passed to -I */
 	if (str) {
 		/* because str is read-only */
@@ -37,14 +43,15 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)
 			if (!strcmp(s, "?")) {
 				fprintf(stderr, "available registers: ");
 				for (r = sample_reg_masks; r->name; r++) {
-					fprintf(stderr, "%s ", r->name);
+					if (r->mask & mask)
+						fprintf(stderr, "%s ", r->name);
 				}
 				fputc('\n', stderr);
 				/* just printing available regs */
 				return -1;
 			}
 			for (r = sample_reg_masks; r->name; r++) {
-				if (!strcasecmp(s, r->name))
+				if ((r->mask & mask) && !strcasecmp(s, r->name))
 					break;
 			}
 			if (!r->name) {
@@ -65,7 +72,7 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)
 
 	/* default to all possible regs */
 	if (*mode == 0)
-		*mode = PERF_REGS_MASK;
+		*mode = mask;
 error:
 	free(os);
 	return ret;
diff --git a/tools/perf/util/perf_regs.c b/tools/perf/util/perf_regs.c
index 2acfcc5..2774cec 100644
--- a/tools/perf/util/perf_regs.c
+++ b/tools/perf/util/perf_regs.c
@@ -13,6 +13,16 @@ int __weak arch_sdt_arg_parse_op(char *old_op __maybe_unused,
 	return SDT_ARG_SKIP;
 }
 
+uint64_t __weak arch__intr_reg_mask(void)
+{
+	return PERF_REGS_MASK;
+}
+
+uint64_t __weak arch__user_reg_mask(void)
+{
+	return PERF_REGS_MASK;
+}
+
 #ifdef HAVE_PERF_REGS_SUPPORT
 int perf_reg_value(u64 *valp, struct regs_dump *regs, int id)
 {
diff --git a/tools/perf/util/perf_regs.h b/tools/perf/util/perf_regs.h
index 1a15a4b..cb9c246 100644
--- a/tools/perf/util/perf_regs.h
+++ b/tools/perf/util/perf_regs.h
@@ -23,6 +23,8 @@ enum {
 };
 
 int arch_sdt_arg_parse_op(char *old_op, char **new_op);
+uint64_t arch__intr_reg_mask(void);
+uint64_t arch__user_reg_mask(void);
 
 #ifdef HAVE_PERF_REGS_SUPPORT
 #include <perf_regs.h>
-- 
2.7.4

[tip:perf/core] perf parse-regs: Add generic support for arch__intr/user_reg_mask()

[tip-bot for Kan Liang]

Commit-ID:  af785e75bf616704cab031e66403b6adcf5b700a
Gitweb:     https://git.kernel.org/tip/af785e75bf616704cab031e66403b6adcf5b700a
Author:     Kan Liang <kan.liang@linux.intel.com>
AuthorDate: Tue, 14 May 2019 13:19:33 -0700
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Thu, 16 May 2019 14:17:12 -0300

perf parse-regs: Add generic support for arch__intr/user_reg_mask()

There may be different register mask for use with intr or user on some
platforms, e.g. Icelake.

Add weak functions arch__intr_reg_mask() and arch__user_reg_mask() to
return intr and user register mask respectively.

Check mask before printing or comparing the register name.

Generic code always return PERF_REGS_MASK. No functional change.

Suggested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Tested-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/1557865174-56264-2-git-send-email-kan.liang@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/parse-regs-options.c | 13 ++++++++++---
 tools/perf/util/perf_regs.c          | 10 ++++++++++
 tools/perf/util/perf_regs.h          |  2 ++
 3 files changed, 22 insertions(+), 3 deletions(-)

diff --git a/tools/perf/util/parse-regs-options.c b/tools/perf/util/parse-regs-options.c
index b21617f2bec1..08581e276225 100644
--- a/tools/perf/util/parse-regs-options.c
+++ b/tools/perf/util/parse-regs-options.c
@@ -12,6 +12,7 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)
 	const struct sample_reg *r;
 	char *s, *os = NULL, *p;
 	int ret = -1;
+	uint64_t mask;
 
 	if (unset)
 		return 0;
@@ -22,6 +23,11 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)
 	if (*mode)
 		return -1;
 
+	if (intr)
+		mask = arch__intr_reg_mask();
+	else
+		mask = arch__user_reg_mask();
+
 	/* str may be NULL in case no arg is passed to -I */
 	if (str) {
 		/* because str is read-only */
@@ -37,14 +43,15 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)
 			if (!strcmp(s, "?")) {
 				fprintf(stderr, "available registers: ");
 				for (r = sample_reg_masks; r->name; r++) {
-					fprintf(stderr, "%s ", r->name);
+					if (r->mask & mask)
+						fprintf(stderr, "%s ", r->name);
 				}
 				fputc('\n', stderr);
 				/* just printing available regs */
 				return -1;
 			}
 			for (r = sample_reg_masks; r->name; r++) {
-				if (!strcasecmp(s, r->name))
+				if ((r->mask & mask) && !strcasecmp(s, r->name))
 					break;
 			}
 			if (!r->name) {
@@ -65,7 +72,7 @@ __parse_regs(const struct option *opt, const char *str, int unset, bool intr)
 
 	/* default to all possible regs */
 	if (*mode == 0)
-		*mode = PERF_REGS_MASK;
+		*mode = mask;
 error:
 	free(os);
 	return ret;
diff --git a/tools/perf/util/perf_regs.c b/tools/perf/util/perf_regs.c
index 2acfcc527cac..2774cec1f15f 100644
--- a/tools/perf/util/perf_regs.c
+++ b/tools/perf/util/perf_regs.c
@@ -13,6 +13,16 @@ int __weak arch_sdt_arg_parse_op(char *old_op __maybe_unused,
 	return SDT_ARG_SKIP;
 }
 
+uint64_t __weak arch__intr_reg_mask(void)
+{
+	return PERF_REGS_MASK;
+}
+
+uint64_t __weak arch__user_reg_mask(void)
+{
+	return PERF_REGS_MASK;
+}
+
 #ifdef HAVE_PERF_REGS_SUPPORT
 int perf_reg_value(u64 *valp, struct regs_dump *regs, int id)
 {
diff --git a/tools/perf/util/perf_regs.h b/tools/perf/util/perf_regs.h
index 1a15a4bfc28d..cb9c246c8962 100644
--- a/tools/perf/util/perf_regs.h
+++ b/tools/perf/util/perf_regs.h
@@ -23,6 +23,8 @@ enum {
 };
 
 int arch_sdt_arg_parse_op(char *old_op, char **new_op);
+uint64_t arch__intr_reg_mask(void);
+uint64_t arch__user_reg_mask(void);
 
 #ifdef HAVE_PERF_REGS_SUPPORT
 #include <perf_regs.h>

[PATCH V2 3/3] perf regs x86: Add X86 specific arch__intr_reg_mask()

[kan.liang]

From: Kan Liang <kan.liang@linux.intel.com>

XMM registers can be collected on Icelake and later platforms.

Add specific arch__intr_reg_mask(), which creating an event to check if
the kernel and hardware can collect XMM registers.

Test on Skylake which doesn't support XMM registers collection. There is
nothing changed.

   #perf record -I?
   available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9
   R10 R11 R12 R13 R14 R15

   Usage: perf record [<options>] [<command>]
    or: perf record [<options>] -- <command> [<options>]

    -I, --intr-regs[=<any register>]
                          sample selected machine registers on
   interrupt, use '-I?' to list register names

   #perf record -I
   [ perf record: Woken up 1 times to write data ]
   [ perf record: Captured and wrote 0.905 MB perf.data (2520 samples) ]

   #perf evlist -v
   cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type:
   IP|TID|TIME|CPU|PERIOD|REGS_INTR, read_format: ID, disabled: 1,
   inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3,
   sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol:
   1, bpf_event: 1, sample_regs_intr: 0xff0fff

Test on Icelake which support XMM registers collection.

   #perf record -I?
   available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9 R10
   R11 R12 R13 R14 R15 XMM0 XMM1 XMM2 XMM3 XMM4 XMM5 XMM6 XMM7 XMM8 XMM9
   XMM10 XMM11 XMM12 XMM13 XMM14 XMM15

   Usage: perf record [<options>] [<command>]
    or: perf record [<options>] -- <command> [<options>]

    -I, --intr-regs[=<any register>]
                          sample selected machine registers on
   interrupt, use '-I?' to list register names

   #perf record -I
   [ perf record: Woken up 1 times to write data ]
   [ perf record: Captured and wrote 0.800 MB perf.data (318 samples) ]

   #perf evlist -v
   cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type:
   IP|TID|TIME|CPU|PERIOD|REGS_INTR, read_format: ID, disabled: 1,
   inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3,
   sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol:
   1, bpf_event: 1, sample_regs_intr: 0xffffffff00ff0fff

Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
---

Changes since V1:
- Add specific arch__intr_reg_mask() support
  Drop specific has_non_gprs_support() and non_gprs_mask()

 tools/perf/arch/x86/include/perf_regs.h |  1 +
 tools/perf/arch/x86/util/perf_regs.c    | 25 +++++++++++++++++++++++++
 2 files changed, 26 insertions(+)

diff --git a/tools/perf/arch/x86/include/perf_regs.h b/tools/perf/arch/x86/include/perf_regs.h
index b732133..b7cd91a 100644
--- a/tools/perf/arch/x86/include/perf_regs.h
+++ b/tools/perf/arch/x86/include/perf_regs.h
@@ -9,6 +9,7 @@
 void perf_regs_load(u64 *regs);
 
 #define PERF_REGS_MAX PERF_REG_X86_XMM_MAX
+#define PERF_XMM_REGS_MASK	(~((1ULL << PERF_REG_X86_XMM0) - 1))
 #ifndef HAVE_ARCH_X86_64_SUPPORT
 #define PERF_REGS_MASK ((1ULL << PERF_REG_X86_32_MAX) - 1)
 #define PERF_SAMPLE_REGS_ABI PERF_SAMPLE_REGS_ABI_32
diff --git a/tools/perf/arch/x86/util/perf_regs.c b/tools/perf/arch/x86/util/perf_regs.c
index 71d7604..c3d7479 100644
--- a/tools/perf/arch/x86/util/perf_regs.c
+++ b/tools/perf/arch/x86/util/perf_regs.c
@@ -270,3 +270,28 @@ int arch_sdt_arg_parse_op(char *old_op, char **new_op)
 
 	return SDT_ARG_VALID;
 }
+
+uint64_t arch__intr_reg_mask(void)
+{
+	struct perf_event_attr attr = {
+		.type			= PERF_TYPE_HARDWARE,
+		.config			= PERF_COUNT_HW_CPU_CYCLES,
+		.sample_period		= 1,
+		.sample_type		= PERF_SAMPLE_REGS_INTR,
+		.sample_regs_intr	= PERF_XMM_REGS_MASK,
+		.precise_ip		= 1,
+		.disabled 		= 1,
+		.exclude_kernel		= 1,
+	};
+	int fd;
+
+	event_attr_init(&attr);
+
+	fd = sys_perf_event_open(&attr, 0, -1, -1, 0);
+	if (fd != -1) {
+		close(fd);
+		return (PERF_XMM_REGS_MASK | PERF_REGS_MASK);
+	}
+
+	return PERF_REGS_MASK;
+}
-- 
2.7.4

Re: [PATCH V2 3/3] perf regs x86: Add X86 specific arch__intr_reg_mask()

[Arnaldo Carvalho de Melo]

Em Tue, May 14, 2019 at 01:19:34PM -0700, kan.liang@linux.intel.com escreveu:
> From: Kan Liang <kan.liang@linux.intel.com>
> 
> XMM registers can be collected on Icelake and later platforms.
> 
> Add specific arch__intr_reg_mask(), which creating an event to check if
> the kernel and hardware can collect XMM registers.
> 
> Test on Skylake which doesn't support XMM registers collection. There is
> nothing changed.

Thanks a lot for doing this and tested on both a machine without these
registers as well as on one with it.

Applied, together with Ravi's tested-by for the first two and the change
in the --user-regs doc,

Regards,

- Arnaldo
 
>    #perf record -I?
>    available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9
>    R10 R11 R12 R13 R14 R15
> 
>    Usage: perf record [<options>] [<command>]
>     or: perf record [<options>] -- <command> [<options>]
> 
>     -I, --intr-regs[=<any register>]
>                           sample selected machine registers on
>    interrupt, use '-I?' to list register names
> 
>    #perf record -I
>    [ perf record: Woken up 1 times to write data ]
>    [ perf record: Captured and wrote 0.905 MB perf.data (2520 samples) ]
> 
>    #perf evlist -v
>    cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type:
>    IP|TID|TIME|CPU|PERIOD|REGS_INTR, read_format: ID, disabled: 1,
>    inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3,
>    sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol:
>    1, bpf_event: 1, sample_regs_intr: 0xff0fff
> 
> Test on Icelake which support XMM registers collection.
> 
>    #perf record -I?
>    available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9 R10
>    R11 R12 R13 R14 R15 XMM0 XMM1 XMM2 XMM3 XMM4 XMM5 XMM6 XMM7 XMM8 XMM9
>    XMM10 XMM11 XMM12 XMM13 XMM14 XMM15
> 
>    Usage: perf record [<options>] [<command>]
>     or: perf record [<options>] -- <command> [<options>]
> 
>     -I, --intr-regs[=<any register>]
>                           sample selected machine registers on
>    interrupt, use '-I?' to list register names
> 
>    #perf record -I
>    [ perf record: Woken up 1 times to write data ]
>    [ perf record: Captured and wrote 0.800 MB perf.data (318 samples) ]
> 
>    #perf evlist -v
>    cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type:
>    IP|TID|TIME|CPU|PERIOD|REGS_INTR, read_format: ID, disabled: 1,
>    inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3,
>    sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol:
>    1, bpf_event: 1, sample_regs_intr: 0xffffffff00ff0fff
> 
> Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
> ---
> 
> Changes since V1:
> - Add specific arch__intr_reg_mask() support
>   Drop specific has_non_gprs_support() and non_gprs_mask()
> 
>  tools/perf/arch/x86/include/perf_regs.h |  1 +
>  tools/perf/arch/x86/util/perf_regs.c    | 25 +++++++++++++++++++++++++
>  2 files changed, 26 insertions(+)
> 
> diff --git a/tools/perf/arch/x86/include/perf_regs.h b/tools/perf/arch/x86/include/perf_regs.h
> index b732133..b7cd91a 100644
> --- a/tools/perf/arch/x86/include/perf_regs.h
> +++ b/tools/perf/arch/x86/include/perf_regs.h
> @@ -9,6 +9,7 @@
>  void perf_regs_load(u64 *regs);
>  
>  #define PERF_REGS_MAX PERF_REG_X86_XMM_MAX
> +#define PERF_XMM_REGS_MASK	(~((1ULL << PERF_REG_X86_XMM0) - 1))
>  #ifndef HAVE_ARCH_X86_64_SUPPORT
>  #define PERF_REGS_MASK ((1ULL << PERF_REG_X86_32_MAX) - 1)
>  #define PERF_SAMPLE_REGS_ABI PERF_SAMPLE_REGS_ABI_32
> diff --git a/tools/perf/arch/x86/util/perf_regs.c b/tools/perf/arch/x86/util/perf_regs.c
> index 71d7604..c3d7479 100644
> --- a/tools/perf/arch/x86/util/perf_regs.c
> +++ b/tools/perf/arch/x86/util/perf_regs.c
> @@ -270,3 +270,28 @@ int arch_sdt_arg_parse_op(char *old_op, char **new_op)
>  
>  	return SDT_ARG_VALID;
>  }
> +
> +uint64_t arch__intr_reg_mask(void)
> +{
> +	struct perf_event_attr attr = {
> +		.type			= PERF_TYPE_HARDWARE,
> +		.config			= PERF_COUNT_HW_CPU_CYCLES,
> +		.sample_period		= 1,
> +		.sample_type		= PERF_SAMPLE_REGS_INTR,
> +		.sample_regs_intr	= PERF_XMM_REGS_MASK,
> +		.precise_ip		= 1,
> +		.disabled 		= 1,
> +		.exclude_kernel		= 1,
> +	};
> +	int fd;
> +
> +	event_attr_init(&attr);
> +
> +	fd = sys_perf_event_open(&attr, 0, -1, -1, 0);
> +	if (fd != -1) {
> +		close(fd);
> +		return (PERF_XMM_REGS_MASK | PERF_REGS_MASK);
> +	}
> +
> +	return PERF_REGS_MASK;
> +}
> -- 
> 2.7.4

-- 

- Arnaldo

[tip:perf/core] perf regs x86: Add X86 specific arch__intr_reg_mask()

[tip-bot for Kan Liang]

Commit-ID:  6466ec14aaf44ff14a05369dcf0929d0f01171c6
Gitweb:     https://git.kernel.org/tip/6466ec14aaf44ff14a05369dcf0929d0f01171c6
Author:     Kan Liang <kan.liang@linux.intel.com>
AuthorDate: Tue, 14 May 2019 13:19:34 -0700
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Thu, 16 May 2019 14:17:23 -0300

perf regs x86: Add X86 specific arch__intr_reg_mask()

XMM registers can be collected on Icelake and later platforms.

Add specific arch__intr_reg_mask(), which creating an event to check if
the kernel and hardware can collect XMM registers.

Test on Skylake which doesn't support XMM registers collection. There is
nothing changed.

   #perf record -I?
   available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9
   R10 R11 R12 R13 R14 R15

   Usage: perf record [<options>] [<command>]
    or: perf record [<options>] -- <command> [<options>]

    -I, --intr-regs[=<any register>]
                          sample selected machine registers on
   interrupt, use '-I?' to list register names

   #perf record -I
   [ perf record: Woken up 1 times to write data ]
   [ perf record: Captured and wrote 0.905 MB perf.data (2520 samples) ]

   #perf evlist -v
   cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type:
   IP|TID|TIME|CPU|PERIOD|REGS_INTR, read_format: ID, disabled: 1,
   inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3,
   sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol:
   1, bpf_event: 1, sample_regs_intr: 0xff0fff

Test on Icelake which support XMM registers collection.

   #perf record -I?
   available registers: AX BX CX DX SI DI BP SP IP FLAGS CS SS R8 R9 R10
   R11 R12 R13 R14 R15 XMM0 XMM1 XMM2 XMM3 XMM4 XMM5 XMM6 XMM7 XMM8 XMM9
   XMM10 XMM11 XMM12 XMM13 XMM14 XMM15

   Usage: perf record [<options>] [<command>]
    or: perf record [<options>] -- <command> [<options>]

    -I, --intr-regs[=<any register>]
                          sample selected machine registers on
   interrupt, use '-I?' to list register names

   #perf record -I
   [ perf record: Woken up 1 times to write data ]
   [ perf record: Captured and wrote 0.800 MB perf.data (318 samples) ]

   #perf evlist -v
   cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type:
   IP|TID|TIME|CPU|PERIOD|REGS_INTR, read_format: ID, disabled: 1,
   inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3,
   sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol:
   1, bpf_event: 1, sample_regs_intr: 0xffffffff00ff0fff

Committer notes:

Don't set attr.sample_period as a named struct init, as it is part of an
unnamed union in 'struct perf_event_attr', and doing so breaks the build
on older gcc versions, such as:

  gcc version 4.1.2 20080704 (Red Hat 4.1.2-55)
  gcc version 4.4.7 20120313 (Red Hat 4.4.7-23) (GCC)

  arch/x86/util/perf_regs.c: In function 'arch__intr_reg_mask':
  arch/x86/util/perf_regs.c:279: error: unknown field 'sample_period' specified in initializer
  cc1: warnings being treated as errors
  arch/x86/util/perf_regs.c:279: warning: missing braces around initializer
  arch/x86/util/perf_regs.c:279: warning: (near initialization for 'attr.<anonymous>')

Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
[ Only on a lenovo t480s, a skylake machine, where the XMM registers didn't show up in -I?/--user-regs=? as expected ]
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/1557865174-56264-3-git-send-email-kan.liang@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/arch/x86/include/perf_regs.h |  1 +
 tools/perf/arch/x86/util/perf_regs.c    | 28 ++++++++++++++++++++++++++++
 2 files changed, 29 insertions(+)

diff --git a/tools/perf/arch/x86/include/perf_regs.h b/tools/perf/arch/x86/include/perf_regs.h
index b7321337d100..b7cd91a9014f 100644
--- a/tools/perf/arch/x86/include/perf_regs.h
+++ b/tools/perf/arch/x86/include/perf_regs.h
@@ -9,6 +9,7 @@
 void perf_regs_load(u64 *regs);
 
 #define PERF_REGS_MAX PERF_REG_X86_XMM_MAX
+#define PERF_XMM_REGS_MASK	(~((1ULL << PERF_REG_X86_XMM0) - 1))
 #ifndef HAVE_ARCH_X86_64_SUPPORT
 #define PERF_REGS_MASK ((1ULL << PERF_REG_X86_32_MAX) - 1)
 #define PERF_SAMPLE_REGS_ABI PERF_SAMPLE_REGS_ABI_32
diff --git a/tools/perf/arch/x86/util/perf_regs.c b/tools/perf/arch/x86/util/perf_regs.c
index 71d7604dbf0b..7886ca5263e3 100644
--- a/tools/perf/arch/x86/util/perf_regs.c
+++ b/tools/perf/arch/x86/util/perf_regs.c
@@ -270,3 +270,31 @@ int arch_sdt_arg_parse_op(char *old_op, char **new_op)
 
 	return SDT_ARG_VALID;
 }
+
+uint64_t arch__intr_reg_mask(void)
+{
+	struct perf_event_attr attr = {
+		.type			= PERF_TYPE_HARDWARE,
+		.config			= PERF_COUNT_HW_CPU_CYCLES,
+		.sample_type		= PERF_SAMPLE_REGS_INTR,
+		.sample_regs_intr	= PERF_XMM_REGS_MASK,
+		.precise_ip		= 1,
+		.disabled 		= 1,
+		.exclude_kernel		= 1,
+	};
+	int fd;
+	/*
+	 * In an unnamed union, init it here to build on older gcc versions
+	 */
+	attr.sample_period = 1;
+
+	event_attr_init(&attr);
+
+	fd = sys_perf_event_open(&attr, 0, -1, -1, 0);
+	if (fd != -1) {
+		close(fd);
+		return (PERF_XMM_REGS_MASK | PERF_REGS_MASK);
+	}
+
+	return PERF_REGS_MASK;
+}

Re: [PATCH V2 1/3] perf parse-regs: Split parse_regs

[Ravi Bangoria]

On 5/15/19 1:49 AM, kan.liang@linux.intel.com wrote:
> From: Kan Liang <kan.liang@linux.intel.com>
> 
> The available registers for --int-regs and --user-regs may be different,
> e.g. XMM registers.
> 
> Split parse_regs into two dedicated functions for --int-regs and
> --user-regs respectively.
> 
> Modify the warning message. "--user-regs=?" should be applied to show
> the available registers for --user-regs.
> 
> Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
> ---

For patch 1 and 2,
Tested-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>

Minor neat. Should we update document as well? May be something like:

  tools/perf/Documentation/perf-record.txt

  --user-regs::
  Similar to -I, but capture user registers at sample time. To list the available
  user registers use --user-regs=\?.

Ravi

Re: [PATCH V2 1/3] perf parse-regs: Split parse_regs

[Liang, Kan]

On 5/15/2019 2:49 AM, Ravi Bangoria wrote:
> 
> On 5/15/19 1:49 AM, kan.liang@linux.intel.com wrote:
>> From: Kan Liang <kan.liang@linux.intel.com>
>>
>> The available registers for --int-regs and --user-regs may be different,
>> e.g. XMM registers.
>>
>> Split parse_regs into two dedicated functions for --int-regs and
>> --user-regs respectively.
>>
>> Modify the warning message. "--user-regs=?" should be applied to show
>> the available registers for --user-regs.
>>
>> Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
>> ---
> 
> For patch 1 and 2,
> Tested-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
> 
> Minor neat. Should we update document as well? May be something like:
> 
>    tools/perf/Documentation/perf-record.txt
> 
>    --user-regs::
>    Similar to -I, but capture user registers at sample time. To list the available
>    user registers use --user-regs=\?.
>
Hi Ravi,

Thanks for test.

The change of document looks good.

Thanks,
Kan

[tip:perf/core] perf parse-regs: Split parse_regs

[tip-bot for Kan Liang]

Commit-ID:  aeea9062d949584ac1f2f9a20f0e5ed306539a3e
Gitweb:     https://git.kernel.org/tip/aeea9062d949584ac1f2f9a20f0e5ed306539a3e
Author:     Kan Liang <kan.liang@linux.intel.com>
AuthorDate: Tue, 14 May 2019 13:19:32 -0700
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 15 May 2019 16:36:49 -0300

perf parse-regs: Split parse_regs

The available registers for --int-regs and --user-regs may be different,
e.g. XMM registers.

Split parse_regs into two dedicated functions for --int-regs and
--user-regs respectively.

Modify the warning message. "--user-regs=?" should be applied to show
the available registers for --user-regs.

Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Tested-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/1557865174-56264-1-git-send-email-kan.liang@linux.intel.com
[ Changed docs as suggested by Ravi and agreed by Kan ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Documentation/perf-record.txt |  3 ++-
 tools/perf/builtin-record.c              |  4 ++--
 tools/perf/util/parse-regs-options.c     | 19 ++++++++++++++++---
 tools/perf/util/parse-regs-options.h     |  3 ++-
 4 files changed, 22 insertions(+), 7 deletions(-)

diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt
index 27b37624c376..de269430720a 100644
--- a/tools/perf/Documentation/perf-record.txt
+++ b/tools/perf/Documentation/perf-record.txt
@@ -406,7 +406,8 @@ symbolic names, e.g. on x86, ax, si. To list the available registers use
 --intr-regs=ax,bx. The list of register is architecture dependent.
 
 --user-regs::
-Capture user registers at sample time. Same arguments as -I.
+Similar to -I, but capture user registers at sample time. To list the available
+user registers use --user-regs=\?.
 
 --running-time::
 Record running and enabled time for read events (:S)
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 861395753c25..e2c3a585a61e 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -2168,10 +2168,10 @@ static struct option __record_options[] = {
 		    "use per-thread mmaps"),
 	OPT_CALLBACK_OPTARG('I', "intr-regs", &record.opts.sample_intr_regs, NULL, "any register",
 		    "sample selected machine registers on interrupt,"
-		    " use '-I?' to list register names", parse_regs),
+		    " use '-I?' to list register names", parse_intr_regs),
 	OPT_CALLBACK_OPTARG(0, "user-regs", &record.opts.sample_user_regs, NULL, "any register",
 		    "sample selected machine registers on interrupt,"
-		    " use '-I?' to list register names", parse_regs),
+		    " use '--user-regs=?' to list register names", parse_user_regs),
 	OPT_BOOLEAN(0, "running-time", &record.opts.running_time,
 		    "Record running/enabled time of read (:S) events"),
 	OPT_CALLBACK('k', "clockid", &record.opts,
diff --git a/tools/perf/util/parse-regs-options.c b/tools/perf/util/parse-regs-options.c
index 9cb187a20fe2..b21617f2bec1 100644
--- a/tools/perf/util/parse-regs-options.c
+++ b/tools/perf/util/parse-regs-options.c
@@ -5,8 +5,8 @@
 #include <subcmd/parse-options.h>
 #include "util/parse-regs-options.h"
 
-int
-parse_regs(const struct option *opt, const char *str, int unset)
+static int
+__parse_regs(const struct option *opt, const char *str, int unset, bool intr)
 {
 	uint64_t *mode = (uint64_t *)opt->value;
 	const struct sample_reg *r;
@@ -48,7 +48,8 @@ parse_regs(const struct option *opt, const char *str, int unset)
 					break;
 			}
 			if (!r->name) {
-				ui__warning("Unknown register \"%s\", check man page or run \"perf record -I?\"\n", s);
+				ui__warning("Unknown register \"%s\", check man page or run \"perf record %s?\"\n",
+					    s, intr ? "-I" : "--user-regs=");
 				goto error;
 			}
 
@@ -69,3 +70,15 @@ error:
 	free(os);
 	return ret;
 }
+
+int
+parse_user_regs(const struct option *opt, const char *str, int unset)
+{
+	return __parse_regs(opt, str, unset, false);
+}
+
+int
+parse_intr_regs(const struct option *opt, const char *str, int unset)
+{
+	return __parse_regs(opt, str, unset, true);
+}
diff --git a/tools/perf/util/parse-regs-options.h b/tools/perf/util/parse-regs-options.h
index cdefb1acf6be..2b23d25c6394 100644
--- a/tools/perf/util/parse-regs-options.h
+++ b/tools/perf/util/parse-regs-options.h
@@ -2,5 +2,6 @@
 #ifndef _PERF_PARSE_REGS_OPTIONS_H
 #define _PERF_PARSE_REGS_OPTIONS_H 1
 struct option;
-int parse_regs(const struct option *opt, const char *str, int unset);
+int parse_user_regs(const struct option *opt, const char *str, int unset);
+int parse_intr_regs(const struct option *opt, const char *str, int unset);
 #endif /* _PERF_PARSE_REGS_OPTIONS_H */