Skip to contentSkip to conversation

perf tools: Fix cross compile for ARM64

RFC PATCH Applied John Garry 7 participants 17 messages Nov 6, 2019 – Dec 11, 2019

Subsystems (2)
lore
Share & tools
lore

John Garry

author
#0 lore raw

RFC PATCH 1/1 perf tools: Fix cross compile for ARM64

Currently when cross compiling perf tool for ARM64 on my x86 machine I get this error:
arch/arm64/util/sym-handling.c:9:10: fatal error: gelf.h: No such file or directory
#include <gelf.h>

For the build, libelf is reported off:
Auto-detecting system features:
...
... libelf: [ OFF ]

Indeed, test-libelf is not built successfully:
more ./build/feature/test-libelf.make.output test-libelf.c:2:10: fatal error: libelf.h: No such file or directory

 #include <libelf.h>
          ^~~~~~~~~~

compilation terminated.

I have no such problems natively compiling on ARM64, and I did not previously have this issue for cross compiling. Fix by relocating the gelf.h include.

---

I marked this as RFC as I am suspicious that I have seen no other reports, and whether fixing up the libelf.h include issue is the proper approach.

tools/perf/arch/arm64/util/sym-handling.c +2 -1
index 5df788985130..8dfa3e5229f1 100644
@@ -6,9 +6,10 @@
 
 #include "symbol.h" // for the elf__needs_adjust_symbols() prototype
 #include <stdbool.h>
-#include <gelf.h>
 
 #ifdef HAVE_LIBELF_SUPPORT
+#include <gelf.h>
+
 bool elf__needs_adjust_symbols(GElf_Ehdr ehdr)
 {
 	return ehdr.e_type == ET_EXEC ||
Signature · 1 line
2.17.1
#1 lore raw

Em Wed, Nov 06, 2019 at 09:00:54PM +0800, John Garry escreveu:

Quoted 9 lines

Currently when cross compiling perf tool for ARM64 on my x86 machine I get this error:
arch/arm64/util/sym-handling.c:9:10: fatal error: gelf.h: No such file or directory
#include <gelf.h>

For the build, libelf is reported off:
Auto-detecting system features:
...
... libelf: [ OFF ]

Thanks, applied.

- Arnaldo

Quoted 36 lines

Indeed, test-libelf is not built successfully:
more ./build/feature/test-libelf.make.output test-libelf.c:2:10: fatal error: libelf.h: No such file or directory

 #include <libelf.h>
          ^~~~~~~~~~

compilation terminated.

I have no such problems natively compiling on ARM64, and I did not previously have this issue for cross compiling. Fix by relocating the gelf.h include.

---

I marked this as RFC as I am suspicious that I have seen no other reports, and whether fixing up the libelf.h include issue is the proper approach.

tools/perf/arch/arm64/util/sym-handling.c +2 -1
index 5df788985130..8dfa3e5229f1 100644
@@ -6,9 +6,10 @@
 
 #include "symbol.h" // for the elf__needs_adjust_symbols() prototype
 #include <stdbool.h>
-#include <gelf.h>
 
 #ifdef HAVE_LIBELF_SUPPORT
+#include <gelf.h>
+
 bool elf__needs_adjust_symbols(GElf_Ehdr ehdr)
 {
 	return ehdr.e_type == ET_EXEC ||
Signature · 1 line
2.17.1
Signature · 1 line
- Arnaldo
author
#2 lore raw

perf top for arm64?

Hi all,

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top
Couldn't read the cpuid for this machine: No such file or directory

That's v5.5-rc1 release.

It seems that we are just missing an arm64 version of get_cpuid() - with the patch below, I now get as hoped:

PerfTop: 32857 irqs/sec kernel:85.0% exact: 0.0% lost: 0/0
drop: 0/0 [4000Hz cycles], (all, 64 CPUs)
-------------------------------------------------------------------------------

     8.99%  [kernel]          [k] arm_smmu_cmdq_issue_cmdlist
     5.80%  [kernel]          [k] __softirqentry_text_start
     4.49%  [kernel]          [k] _raw_spin_unlock_irqrestore
     3.48%  [kernel]          [k] el0_svc_common.constprop.2
     3.37%  [kernel]          [k] _raw_write_lock_irqsave
     3.28%  [kernel]          [k] __local_bh_enable_ip
     3.05%  [kernel]          [k] __blk_complete_request
     2.07%  [kernel]          [k] queued_spin_lock_slowpath
     1.93%  [vdso]            [.] 0x0000000000000484

Was this just missed? Or is there a good reason to omit?

Thanks,
John

--->8---

Subject: [PATCH] perf: Add perf top support for arm64

Copied from get_cpuid_str() essentially...

tools/perf/arch/arm64/util/header.c +54 -0
index a32e4b72a98f..ecd1f86e29cc 100644
@@ -1,10 +1,12 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <perf/cpumap.h>
+#include <util/cpumap.h>
 #include <internal/cpumap.h>
 #include <api/fs/fs.h>
 #include "debug.h"
 #include "header.h"
+#include <errno.h>
 
 #define MIDR "/regs/identification/midr_el1"
 #define MIDR_SIZE 19
@@ -12,6 +14,59 @@
 #define MIDR_VARIANT_SHIFT      20
 #define MIDR_VARIANT_MASK       (0xf << MIDR_VARIANT_SHIFT)
 
+int
+get_cpuid(char *buffer, size_t sz)
+{
+	char *buf = NULL;
+	char path[PATH_MAX];
+	const char *sysfs = sysfs__mountpoint();
+	int cpu;
+	u64 midr = 0;
+	FILE *file;
+
+	if (!sysfs)
+		return EINVAL;
+
+	buf = malloc(MIDR_SIZE);
+	if (!buf)
+		return EINVAL;
+
+	/* read midr from list of cpus mapped to this pmu */
+	for (cpu = 0; cpu < cpu__max_present_cpu(); cpu++) {
+		scnprintf(path, sz, "%s/devices/system/cpu/cpu%d"MIDR,
+				sysfs, cpu);
+
+		file = fopen(path, "r");
+		if (!file) {
+			pr_debug("fopen failed for file %s\n", path);
+			continue;
+		}
+
+		if (!fgets(buf, MIDR_SIZE, file)) {
+			fclose(file);
+			continue;
+		}
+		fclose(file);
+
+		/* Ignore/clear Variant[23:20] and
+		 * Revision[3:0] of MIDR
+		 */
+		midr = strtoul(buf, NULL, 16);
+		midr &= (~(MIDR_VARIANT_MASK | MIDR_REVISION_MASK));
+		scnprintf(buffer, MIDR_SIZE, "0x%016lx", midr);
+		/* got midr break loop */
+		break;
+	}
+
+	if (!midr) {
+		pr_err("failed to get cpuid string\n");
+		free(buf);
+		return EINVAL;
+	}
+	return 0;
+}
+
#3 lore raw

Re: perf top for arm64?

Quoted 6 lines · John Garry

Hi all,

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top
Couldn't read the cpuid for this machine: No such file or directory

there was recent change that check on cpuid and quits:
608127f73779 perf top: Initialize perf_env->cpuid, needed by the per arch annotation init routine

Arnaldo,
maybe this should be just a warning/info, because it seems to be related to annotations only..?

get_cpuid is defined only for s390/x86/powerpc, so I guess it won't work on the rest as well

jirka

Quoted 109 lines

That's v5.5-rc1 release.

It seems that we are just missing an arm64 version of get_cpuid() - with the patch below, I now get as hoped:

   PerfTop:   32857 irqs/sec  kernel:85.0%  exact:  0.0% lost: 0/0 drop: 0/0

[4000Hz cycles], (all, 64 CPUs)
-------------------------------------------------------------------------------

     8.99%  [kernel]          [k] arm_smmu_cmdq_issue_cmdlist
     5.80%  [kernel]          [k] __softirqentry_text_start
     4.49%  [kernel]          [k] _raw_spin_unlock_irqrestore
     3.48%  [kernel]          [k] el0_svc_common.constprop.2
     3.37%  [kernel]          [k] _raw_write_lock_irqsave
     3.28%  [kernel]          [k] __local_bh_enable_ip
     3.05%  [kernel]          [k] __blk_complete_request
     2.07%  [kernel]          [k] queued_spin_lock_slowpath
     1.93%  [vdso]            [.] 0x0000000000000484

Was this just missed? Or is there a good reason to omit?

Thanks,
John

--->8---

Subject: [PATCH] perf: Add perf top support for arm64

Copied from get_cpuid_str() essentially...

tools/perf/arch/arm64/util/header.c +54 -0
index a32e4b72a98f..ecd1f86e29cc 100644
@@ -1,10 +1,12 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <perf/cpumap.h>
+#include <util/cpumap.h>
 #include <internal/cpumap.h>
 #include <api/fs/fs.h>
 #include "debug.h"
 #include "header.h"
+#include <errno.h>
 
 #define MIDR "/regs/identification/midr_el1"
 #define MIDR_SIZE 19
@@ -12,6 +14,59 @@
 #define MIDR_VARIANT_SHIFT      20
 #define MIDR_VARIANT_MASK       (0xf << MIDR_VARIANT_SHIFT)
 
+int
+get_cpuid(char *buffer, size_t sz)
+{
+	char *buf = NULL;
+	char path[PATH_MAX];
+	const char *sysfs = sysfs__mountpoint();
+	int cpu;
+	u64 midr = 0;
+	FILE *file;
+
+	if (!sysfs)
+		return EINVAL;
+
+	buf = malloc(MIDR_SIZE);
+	if (!buf)
+		return EINVAL;
+
+	/* read midr from list of cpus mapped to this pmu */
+	for (cpu = 0; cpu < cpu__max_present_cpu(); cpu++) {
+		scnprintf(path, sz, "%s/devices/system/cpu/cpu%d"MIDR,
+				sysfs, cpu);
+
+		file = fopen(path, "r");
+		if (!file) {
+			pr_debug("fopen failed for file %s\n", path);
+			continue;
+		}
+
+		if (!fgets(buf, MIDR_SIZE, file)) {
+			fclose(file);
+			continue;
+		}
+		fclose(file);
+
+		/* Ignore/clear Variant[23:20] and
+		 * Revision[3:0] of MIDR
+		 */
+		midr = strtoul(buf, NULL, 16);
+		midr &= (~(MIDR_VARIANT_MASK | MIDR_REVISION_MASK));
+		scnprintf(buffer, MIDR_SIZE, "0x%016lx", midr);
+		/* got midr break loop */
+		break;
+	}
+
+	if (!midr) {
+		pr_err("failed to get cpuid string\n");
+		free(buf);
+		return EINVAL;
+	}
+	return 0;
+}
+
 
author
#4 lore raw

Re: perf top for arm64?

Quoted 8 lines · Jiri Olsa
Quoted 6 lines · John Garry

Hi all,

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top
Couldn't read the cpuid for this machine: No such file or directory

Hi Jirka,

Quoted 3 lines

there was recent change that check on cpuid and quits:
608127f73779 perf top: Initialize perf_env->cpuid, needed by the per arch annotation init routine

ok, this is new code. I obviously didn't check the git history...

But, apart from this, there are many other places where get_cpuid() is called. I wonder what else we're missing out on, and whether we should still add it.

Thanks,
John

Quoted 121 lines

Arnaldo,
maybe this should be just a warning/info, because it seems to be related to annotations only..?

get_cpuid is defined only for s390/x86/powerpc, so I guess it won't work on the rest as well

jirka

Quoted 99 lines

That's v5.5-rc1 release.

It seems that we are just missing an arm64 version of get_cpuid() - with the patch below, I now get as hoped:

    PerfTop:   32857 irqs/sec  kernel:85.0%  exact:  0.0% lost: 0/0 drop: 0/0

[4000Hz cycles], (all, 64 CPUs)
-------------------------------------------------------------------------------

      8.99%  [kernel]          [k] arm_smmu_cmdq_issue_cmdlist
      5.80%  [kernel]          [k] __softirqentry_text_start
      4.49%  [kernel]          [k] _raw_spin_unlock_irqrestore
      3.48%  [kernel]          [k] el0_svc_common.constprop.2
      3.37%  [kernel]          [k] _raw_write_lock_irqsave
      3.28%  [kernel]          [k] __local_bh_enable_ip
      3.05%  [kernel]          [k] __blk_complete_request
      2.07%  [kernel]          [k] queued_spin_lock_slowpath
      1.93%  [vdso]            [.] 0x0000000000000484

Was this just missed? Or is there a good reason to omit?

Thanks,
John

--->8---

Subject: [PATCH] perf: Add perf top support for arm64

Copied from get_cpuid_str() essentially...

tools/perf/arch/arm64/util/header.c +45 -0
index a32e4b72a98f..ecd1f86e29cc 100644
@@ -1,10 +1,12 @@
  #include <stdio.h>
  #include <stdlib.h>
  #include <perf/cpumap.h>
+#include <util/cpumap.h>
  #include <internal/cpumap.h>
  #include <api/fs/fs.h>
  #include "debug.h"
  #include "header.h"
+#include <errno.h>
 
  #define MIDR "/regs/identification/midr_el1"
  #define MIDR_SIZE 19
@@ -12,6 +14,59 @@
  #define MIDR_VARIANT_SHIFT      20
  #define MIDR_VARIANT_MASK       (0xf << MIDR_VARIANT_SHIFT)
 
+int
+get_cpuid(char *buffer, size_t sz)
+{
+	char *buf = NULL;
+	char path[PATH_MAX];
+	const char *sysfs = sysfs__mountpoint();
+	int cpu;
+	u64 midr = 0;
+	FILE *file;
+
+	if (!sysfs)
+		return EINVAL;
+
+	buf = malloc(MIDR_SIZE);
+	if (!buf)
+		return EINVAL;
+
+	/* read midr from list of cpus mapped to this pmu */
+	for (cpu = 0; cpu < cpu__max_present_cpu(); cpu++) {
+		scnprintf(path, sz, "%s/devices/system/cpu/cpu%d"MIDR,
+				sysfs, cpu);
+
+		file = fopen(path, "r");
+		if (!file) {
+			pr_debug("fopen failed for file %s\n", path);
+			continue;
+		}
+
+		if (!fgets(buf, MIDR_SIZE, file)) {
+			fclose(file);
+			continue;
+		}
+		fclose(file);
+
+		/* Ignore/clear Variant[23:20] and
+		 * Revision[3:0] of MIDR
+		 */
+		midr = strtoul(buf, NULL, 16);
+		midr &= (~(MIDR_VARIANT_MASK | MIDR_REVISION_MASK));
+		scnprintf(buffer, MIDR_SIZE, "0x%016lx", midr);
+		/* got midr break loop */
+		break;
+	}

tools/perf/arch/arm64/util/header.c line 69

Quoted 10 lines
tools/perf/arch/arm64/util/header.c +9 -0
+
+	if (!midr) {
+		pr_err("failed to get cpuid string\n");
+		free(buf);
+		return EINVAL;
+	}
+	return 0;
+}
+
 

.

#5 lore raw

Re: perf top for arm64?

Quoted 16 lines · John Garry
Quoted 8 lines · Jiri Olsa
Quoted 6 lines · John Garry

Hi all,

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top
Couldn't read the cpuid for this machine: No such file or directory

Hi Jirka,

Quoted 3 lines

there was recent change that check on cpuid and quits:
608127f73779 perf top: Initialize perf_env->cpuid, needed by the per arch annotation init routine

Quoted 5 lines

ok, this is new code. I obviously didn't check the git history...

But, apart from this, there are many other places where get_cpuid() is called. I wonder what else we're missing out on, and whether we should still add it.

right, I was just wondering how come vendor events are working for you, but realized we have get_cpuid_str being called in there ;-)

I think we should add it as you have it prepared already, could you post it with bigger changelog that would explain where it's being used for arm?

jirka

author
#6 lore raw

Re: perf top for arm64?

Quoted 23 lines · Jiri Olsa
Quoted 16 lines · John Garry
Quoted 8 lines · Jiri Olsa
Quoted 6 lines · John Garry

Hi all,

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top
Couldn't read the cpuid for this machine: No such file or directory

Hi Jirka,

Quoted 3 lines

there was recent change that check on cpuid and quits:

    608127f73779 perf top: Initialize perf_env->cpuid, needed by the per arch annotation init routine
Quoted 5 lines

ok, this is new code. I obviously didn't check the git history...

But, apart from this, there are many other places where get_cpuid() is called. I wonder what else we're missing out on, and whether we should still add it.

Quoted 6 lines

right, I was just wondering how come vendor events are working for you, but realized we have get_cpuid_str being called in there ;-)

I think we should add it as you have it prepared already, could you post it with bigger changelog that would explain where it's being used for arm?

ok, I can look to do that.

But, as you know, we still need to fix perf top for other architectures affected.

Thanks,
John

#7 lore raw

Re: perf top for arm64?

Em Tue, Dec 10, 2019 at 05:17:56PM +0000, John Garry escreveu:

Quoted 27 lines · Jiri Olsa
Quoted 26 lines · John Garry
Quoted 16 lines · Jiri Olsa
Quoted 8 lines · John Garry
Quoted 6 lines

Hi all,

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top
Couldn't read the cpuid for this machine: No such file or directory

Hi Jirka,

Quoted 3 lines

there was recent change that check on cpuid and quits:

    608127f73779 perf top: Initialize perf_env->cpuid, needed by the per arch annotation init routine
Quoted 5 lines

ok, this is new code. I obviously didn't check the git history...

But, apart from this, there are many other places where get_cpuid() is called. I wonder what else we're missing out on, and whether we should still add it.

right, I was just wondering how come vendor events are working for you, but realized we have get_cpuid_str being called in there ;-)

Quoted 8 lines
Quoted 3 lines

I think we should add it as you have it prepared already, could you post it with bigger changelog that would explain where it's being used for arm?

ok, I can look to do that.

But, as you know, we still need to fix perf top for other architectures affected.

Right, I need to make that just a pr_debug() message and then check in the annotation code when that is needed to see if it is set, if not, then show a popup error message and refuse to do whatever annotation feature requires that.

Anyway, your patch should make sense and provide info that the ARM64 annotation may use now or in the future.

- Arnaldo

#8 lore raw

PATCH 1/1 Fix 'perf top' breakage on architectures not providing get_cpuid() Re: perf top for arm64?

Em Tue, Dec 10, 2019 at 04:51:13PM -0300, Arnaldo Carvalho de Melo escreveu:

Quoted 31 lines

Em Tue, Dec 10, 2019 at 05:17:56PM +0000, John Garry escreveu:

Quoted 21 lines · Jiri Olsa
Quoted 19 lines · John Garry
Quoted 7 lines · Jiri Olsa
Quoted 5 lines · John Garry
Quoted 4 lines

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top
Couldn't read the cpuid for this machine: No such file or directory
Quoted 8 lines
Quoted 2 lines

there was recent change that check on cpuid and quits:

    608127f73779 perf top: Initialize perf_env->cpuid, needed by the per arch annotation init routine

ok, this is new code. I obviously didn't check the git history...

But, apart from this, there are many other places where get_cpuid() is called. I wonder what else we're missing out on, and whether we should still add it.

right, I was just wondering how come vendor events are working for you, but realized we have get_cpuid_str being called in there ;-)

Quoted 8 lines
Quoted 3 lines

I think we should add it as you have it prepared already, could you post it with bigger changelog that would explain where it's being used for arm?

ok, I can look to do that.

But, as you know, we still need to fix perf top for other architectures affected.

Quoted 7 lines

Right, I need to make that just a pr_debug() message and then check in the annotation code when that is needed to see if it is set, if not, then show a popup error message and refuse to do whatever annotation feature requires that.

Anyway, your patch should make sense and provide info that the ARM64 annotation may use now or in the future.

So can you take a look at the two patches below and provide me Acked-by and/or Reviewed-by and/or Tested-by?

From 53c6cde6a71a1a9283763bd2e938b229b50c2cd5 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <[email protected]>
Date: Wed, 11 Dec 2019 10:09:24 -0300
Subject: [PATCH 1/2] perf arch: Make the default get_cpuid() return compatible
error

Some of the functions calling get_cpuid() propagate back the error it returns, and all are using errno (positive) values, make the weak default get_cpuid() function return ENOSYS to be consistent and to allow checking if this is an arch not providing this function or if a provided one is having trouble getting the cpuid, to decide if the warning should be provided to the user or just a debug message should be emitted.

---
 tools/perf/util/header.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/util/header.c +1 -1
index becc2d109423..4d39a75551a0 100644
@@ -850,7 +850,7 @@ int __weak strcmp_cpuid_str(const char *mapcpuid, const char *cpuid)
  */
 int __weak get_cpuid(char *buffer __maybe_unused, size_t sz __maybe_unused)
 {
-	return -1;
+	return ENOSYS; /* Not implemented */
 }
 
 static int write_cpuid(struct feat_fd *ff,
Signature · 48 lines
2.21.0

From c6c6a3e2eb6982e37294abcac389effd298cf730 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <[email protected]>
Date: Wed, 11 Dec 2019 10:21:59 -0300
Subject: [PATCH 2/2] perf top: Do not bail out when perf_env__read_cpuid()
 returns ENOSYS

'perf top' stopped working on hw architectures that do not provide a
get_cpuid() implementation and thus fallback to the weak get_cpuid()
default function.

This is done because at annotation time we may need it in the arch
specific annotation init routine, but that is only being used by arches
that do provide a get_cpuid() implementation:

  $ find tools/  -name "*.[ch]" | xargs grep 'evlist->env'
  tools/perf/builtin-top.c:	top.evlist->env = &perf_env;
  tools/perf/util/evsel.c:		return evsel->evlist->env;
  tools/perf/util/s390-cpumsf.c:	sf->machine_type = s390_cpumsf_get_type(session->evlist->env->cpuid);
  tools/perf/util/header.c:	session->evlist->env = &header->env;
  tools/perf/util/sample-raw.c:	const char *arch_pf = perf_env__arch(evlist->env);
  $

  $ find tools/perf/arch  -name "*.[ch]" | xargs grep -w get_cpuid
  tools/perf/arch/x86/util/auxtrace.c:	ret = get_cpuid(buffer, sizeof(buffer));
  tools/perf/arch/x86/util/header.c:get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/powerpc/util/header.c:get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/s390/util/header.c: * Implementation of get_cpuid().
  tools/perf/arch/s390/util/header.c:int get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/s390/util/header.c:	if (buf && get_cpuid(buf, 128))
  $

For 'report' or 'script', i.e. tools working on perf.data files, that is
setup while reading the header, its just top that needs to explicitely
read it at tool start.

Reported-by: John Garry <[email protected]>
Analysed-by: Jiri Olsa <[email protected]>
Cc: Adrian Hunter <[email protected]>
Cc: Mark Rutland <[email protected]>
Cc: Namhyung Kim <[email protected]>
Cc: Will Deacon <[email protected]>
Link: https://lkml.kernel.org/n/[email protected]
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
---
 tools/perf/builtin-top.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)
tools/perf/builtin-top.c +7 -3
index dc80044bc46f..795e353de095 100644
@@ -1568,9 +1568,13 @@ int cmd_top(int argc, const char **argv)
 	 */
 	status = perf_env__read_cpuid(&perf_env);
 	if (status) {
-		pr_err("Couldn't read the cpuid for this machine: %s\n",
-		       str_error_r(errno, errbuf, sizeof(errbuf)));
-		goto out_delete_evlist;
+		/*
+		 * Some arches do not provide a get_cpuid(), so just use pr_debug, otherwise
+		 * warn the user explicitely.
+		 */
+		eprintf(status == ENOSYS ? 1 : 0, verbose,
+			"Couldn't read the cpuid for this machine: %s\n",
+			str_error_r(errno, errbuf, sizeof(errbuf)));
 	}
 	top.evlist->env = &perf_env;
 
Signature · 1 line
2.21.0
reviewer Tested-by Reviewed-by on v1 1/1
#9 lore raw

Re: [PATCHES] Fix 'perf top' breakage on architectures not providing get_cpuid() Re: perf top for arm64?

Quoted 2 lines · Arnaldo Carvalho de Melo

So can you take a look at the two patches below and provide me Acked-by and/or Reviewed-by and/or Tested-by?

I just gave this a spin. With vanilla v5.5-rc1 perf top behaves as John reported, and with these applied atop perf works as expected:

From scaning the code, the only other user of get_cpuid() that ends up giving up is perf kvm, but AFAICT that never supported arm64, so that looks sound to me:

Thanks,
Mark.

Quoted 112 lines

From 53c6cde6a71a1a9283763bd2e938b229b50c2cd5 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <[email protected]>
Date: Wed, 11 Dec 2019 10:09:24 -0300
Subject: [PATCH 1/2] perf arch: Make the default get_cpuid() return compatible
error

Some of the functions calling get_cpuid() propagate back the error it returns, and all are using errno (positive) values, make the weak default get_cpuid() function return ENOSYS to be consistent and to allow checking if this is an arch not providing this function or if a provided one is having trouble getting the cpuid, to decide if the warning should be provided to the user or just a debug message should be emitted.

---
 tools/perf/util/header.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/util/header.c +1 -1
index becc2d109423..4d39a75551a0 100644
@@ -850,7 +850,7 @@ int __weak strcmp_cpuid_str(const char *mapcpuid, const char *cpuid)
  */
 int __weak get_cpuid(char *buffer __maybe_unused, size_t sz __maybe_unused)
 {
-	return -1;
+	return ENOSYS; /* Not implemented */
 }
 
 static int write_cpuid(struct feat_fd *ff,
Signature · 48 lines
2.21.0

From c6c6a3e2eb6982e37294abcac389effd298cf730 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <[email protected]>
Date: Wed, 11 Dec 2019 10:21:59 -0300
Subject: [PATCH 2/2] perf top: Do not bail out when perf_env__read_cpuid()
 returns ENOSYS

'perf top' stopped working on hw architectures that do not provide a
get_cpuid() implementation and thus fallback to the weak get_cpuid()
default function.

This is done because at annotation time we may need it in the arch
specific annotation init routine, but that is only being used by arches
that do provide a get_cpuid() implementation:

  $ find tools/  -name "*.[ch]" | xargs grep 'evlist->env'
  tools/perf/builtin-top.c:	top.evlist->env = &perf_env;
  tools/perf/util/evsel.c:		return evsel->evlist->env;
  tools/perf/util/s390-cpumsf.c:	sf->machine_type = s390_cpumsf_get_type(session->evlist->env->cpuid);
  tools/perf/util/header.c:	session->evlist->env = &header->env;
  tools/perf/util/sample-raw.c:	const char *arch_pf = perf_env__arch(evlist->env);
  $

  $ find tools/perf/arch  -name "*.[ch]" | xargs grep -w get_cpuid
  tools/perf/arch/x86/util/auxtrace.c:	ret = get_cpuid(buffer, sizeof(buffer));
  tools/perf/arch/x86/util/header.c:get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/powerpc/util/header.c:get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/s390/util/header.c: * Implementation of get_cpuid().
  tools/perf/arch/s390/util/header.c:int get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/s390/util/header.c:	if (buf && get_cpuid(buf, 128))
  $

For 'report' or 'script', i.e. tools working on perf.data files, that is
setup while reading the header, its just top that needs to explicitely
read it at tool start.

Reported-by: John Garry <[email protected]>
Analysed-by: Jiri Olsa <[email protected]>
Cc: Adrian Hunter <[email protected]>
Cc: Mark Rutland <[email protected]>
Cc: Namhyung Kim <[email protected]>
Cc: Will Deacon <[email protected]>
Link: https://lkml.kernel.org/n/[email protected]
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
---
 tools/perf/builtin-top.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)
tools/perf/builtin-top.c +7 -3
index dc80044bc46f..795e353de095 100644
@@ -1568,9 +1568,13 @@ int cmd_top(int argc, const char **argv)
 	 */
 	status = perf_env__read_cpuid(&perf_env);
 	if (status) {
-		pr_err("Couldn't read the cpuid for this machine: %s\n",
-		       str_error_r(errno, errbuf, sizeof(errbuf)));
-		goto out_delete_evlist;
+		/*
+		 * Some arches do not provide a get_cpuid(), so just use pr_debug, otherwise
+		 * warn the user explicitely.
+		 */
+		eprintf(status == ENOSYS ? 1 : 0, verbose,
+			"Couldn't read the cpuid for this machine: %s\n",
+			str_error_r(errno, errbuf, sizeof(errbuf)));
 	}
 	top.evlist->env = &perf_env;
 
Signature · 1 line
2.21.0
#10 lore raw

Re: [PATCHES] Fix 'perf top' breakage on architectures not providing get_cpuid() Re: perf top for arm64?

Em Wed, Dec 11, 2019 at 02:46:43PM +0000, Mark Rutland escreveu:

Quoted 7 lines · Arnaldo Carvalho de Melo
Quoted 2 lines

So can you take a look at the two patches below and provide me Acked-by and/or Reviewed-by and/or Tested-by?

I just gave this a spin. With vanilla v5.5-rc1 perf top behaves as John reported, and with these applied atop perf works as expected:

Quoted 7 lines
Quoted 1 line

From scaning the code, the only other user of get_cpuid() that ends up

giving up is perf kvm, but AFAICT that never supported arm64, so that looks sound to me:

Thanks guys, added all the tags, appreciated, I also added this:

Fixes 608127f73779 ("perf top: Initialize perf_env->cpuid, needed by the per arch annotation init routine")

Thanks Jiri for pointing to it.

- Arnaldo

Quoted 115 lines

Thanks,
Mark.

Quoted 112 lines

From 53c6cde6a71a1a9283763bd2e938b229b50c2cd5 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <[email protected]>
Date: Wed, 11 Dec 2019 10:09:24 -0300
Subject: [PATCH 1/2] perf arch: Make the default get_cpuid() return compatible
error

Some of the functions calling get_cpuid() propagate back the error it returns, and all are using errno (positive) values, make the weak default get_cpuid() function return ENOSYS to be consistent and to allow checking if this is an arch not providing this function or if a provided one is having trouble getting the cpuid, to decide if the warning should be provided to the user or just a debug message should be emitted.

---
 tools/perf/util/header.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/util/header.c +1 -1
index becc2d109423..4d39a75551a0 100644
@@ -850,7 +850,7 @@ int __weak strcmp_cpuid_str(const char *mapcpuid, const char *cpuid)
  */
 int __weak get_cpuid(char *buffer __maybe_unused, size_t sz __maybe_unused)
 {
-	return -1;
+	return ENOSYS; /* Not implemented */
 }
 
 static int write_cpuid(struct feat_fd *ff,
Signature · 48 lines
2.21.0

From c6c6a3e2eb6982e37294abcac389effd298cf730 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <[email protected]>
Date: Wed, 11 Dec 2019 10:21:59 -0300
Subject: [PATCH 2/2] perf top: Do not bail out when perf_env__read_cpuid()
 returns ENOSYS

'perf top' stopped working on hw architectures that do not provide a
get_cpuid() implementation and thus fallback to the weak get_cpuid()
default function.

This is done because at annotation time we may need it in the arch
specific annotation init routine, but that is only being used by arches
that do provide a get_cpuid() implementation:

  $ find tools/  -name "*.[ch]" | xargs grep 'evlist->env'
  tools/perf/builtin-top.c:	top.evlist->env = &perf_env;
  tools/perf/util/evsel.c:		return evsel->evlist->env;
  tools/perf/util/s390-cpumsf.c:	sf->machine_type = s390_cpumsf_get_type(session->evlist->env->cpuid);
  tools/perf/util/header.c:	session->evlist->env = &header->env;
  tools/perf/util/sample-raw.c:	const char *arch_pf = perf_env__arch(evlist->env);
  $

  $ find tools/perf/arch  -name "*.[ch]" | xargs grep -w get_cpuid
  tools/perf/arch/x86/util/auxtrace.c:	ret = get_cpuid(buffer, sizeof(buffer));
  tools/perf/arch/x86/util/header.c:get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/powerpc/util/header.c:get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/s390/util/header.c: * Implementation of get_cpuid().
  tools/perf/arch/s390/util/header.c:int get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/s390/util/header.c:	if (buf && get_cpuid(buf, 128))
  $

For 'report' or 'script', i.e. tools working on perf.data files, that is
setup while reading the header, its just top that needs to explicitely
read it at tool start.

Reported-by: John Garry <[email protected]>
Analysed-by: Jiri Olsa <[email protected]>
Cc: Adrian Hunter <[email protected]>
Cc: Mark Rutland <[email protected]>
Cc: Namhyung Kim <[email protected]>
Cc: Will Deacon <[email protected]>
Link: https://lkml.kernel.org/n/[email protected]
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
---
 tools/perf/builtin-top.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)
tools/perf/builtin-top.c +7 -3
index dc80044bc46f..795e353de095 100644
@@ -1568,9 +1568,13 @@ int cmd_top(int argc, const char **argv)
 	 */
 	status = perf_env__read_cpuid(&perf_env);
 	if (status) {
-		pr_err("Couldn't read the cpuid for this machine: %s\n",
-		       str_error_r(errno, errbuf, sizeof(errbuf)));
-		goto out_delete_evlist;
+		/*
+		 * Some arches do not provide a get_cpuid(), so just use pr_debug, otherwise
+		 * warn the user explicitely.
+		 */
+		eprintf(status == ENOSYS ? 1 : 0, verbose,
+			"Couldn't read the cpuid for this machine: %s\n",
+			str_error_r(errno, errbuf, sizeof(errbuf)));
 	}
 	top.evlist->env = &perf_env;
 
Signature · 1 line
2.21.0
Signature · 1 line
- Arnaldo
author Tested-by on v1 1/1
#11 lore raw

Re: [PATCHES] Fix 'perf top' breakage on architectures not providing get_cpuid() Re: perf top for arm64?

In reply to Arnaldo Carvalho de Melo
Quoted 28 lines
Quoted 9 lines

Right, I need to make that just a pr_debug() message and then check in the annotation code when that is needed to see if it is set, if not, then show a popup error message and refuse to do whatever annotation feature requires that.

Anyway, your patch should make sense and provide info that the ARM64 annotation may use now or in the future.

So can you take a look at the two patches below and provide me Acked-by and/or Reviewed-by and/or Tested-by?

Quoted 1 line

From 53c6cde6a71a1a9283763bd2e938b229b50c2cd5 Mon Sep 17 00:00:00 2001

From: Arnaldo Carvalho de Melo <[email protected]>
Date: Wed, 11 Dec 2019 10:09:24 -0300
Subject: [PATCH 1/2] perf arch: Make the default get_cpuid() return compatible
error

Some of the functions calling get_cpuid() propagate back the error it returns, and all are using errno (positive) values, make the weak default get_cpuid() function return ENOSYS to be consistent and to allow checking if this is an arch not providing this function or if a provided one is having trouble getting the cpuid, to decide if the warning should be provided to the user or just a debug message should be emitted.

Quoted 3 lines
Cc Adrian Hunter <[email protected]>
Cc Jiri Olsa <[email protected]>
Cc John Garry <[email protected]>

For both patches:

I'll still look to get the arm64 version of get_cpuid() patch sent.

Quoted 23 lines
---
  tools/perf/util/header.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/util/header.c +1 -1
index becc2d109423..4d39a75551a0 100644
@@ -850,7 +850,7 @@ int __weak strcmp_cpuid_str(const char *mapcpuid, const char *cpuid)
   */
  int __weak get_cpuid(char *buffer __maybe_unused, size_t sz __maybe_unused)
  {
-	return -1;
+	return ENOSYS; /* Not implemented */
  }
 
  static int write_cpuid(struct feat_fd *ff,
Acked-by on v1 1/1
#12 lore raw

Re: [PATCHES] Fix 'perf top' breakage on architectures not providing get_cpuid() Re: perf top for arm64?

In reply to Arnaldo Carvalho de Melo
Quoted 37 lines · Arnaldo Carvalho de Melo

Em Tue, Dec 10, 2019 at 04:51:13PM -0300, Arnaldo Carvalho de Melo escreveu:

Quoted 35 lines

Em Tue, Dec 10, 2019 at 05:17:56PM +0000, John Garry escreveu:

Quoted 21 lines · Jiri Olsa
Quoted 19 lines · John Garry
Quoted 7 lines · Jiri Olsa
Quoted 5 lines · John Garry
Quoted 4 lines

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top
Couldn't read the cpuid for this machine: No such file or directory
Quoted 8 lines
Quoted 2 lines

there was recent change that check on cpuid and quits:

    608127f73779 perf top: Initialize perf_env->cpuid, needed by the per arch annotation init routine

ok, this is new code. I obviously didn't check the git history...

But, apart from this, there are many other places where get_cpuid() is called. I wonder what else we're missing out on, and whether we should still add it.

right, I was just wondering how come vendor events are working for you, but realized we have get_cpuid_str being called in there ;-)

Quoted 8 lines
Quoted 3 lines

I think we should add it as you have it prepared already, could you post it with bigger changelog that would explain where it's being used for arm?

ok, I can look to do that.

But, as you know, we still need to fix perf top for other architectures affected.

Right, I need to make that just a pr_debug() message and then check in the annotation code when that is needed to see if it is set, if not, then show a popup error message and refuse to do whatever annotation feature requires that.

Quoted 5 lines
Quoted 2 lines

Anyway, your patch should make sense and provide info that the ARM64 annotation may use now or in the future.

So can you take a look at the two patches below and provide me Acked-by and/or Reviewed-by and/or Tested-by?

looks good to me, but don't have arm server at the moment..

thanks,
jirka

Quoted 114 lines

From 53c6cde6a71a1a9283763bd2e938b229b50c2cd5 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <[email protected]>
Date: Wed, 11 Dec 2019 10:09:24 -0300
Subject: [PATCH 1/2] perf arch: Make the default get_cpuid() return compatible
error

Some of the functions calling get_cpuid() propagate back the error it returns, and all are using errno (positive) values, make the weak default get_cpuid() function return ENOSYS to be consistent and to allow checking if this is an arch not providing this function or if a provided one is having trouble getting the cpuid, to decide if the warning should be provided to the user or just a debug message should be emitted.

---
 tools/perf/util/header.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/util/header.c +1 -1
index becc2d109423..4d39a75551a0 100644
@@ -850,7 +850,7 @@ int __weak strcmp_cpuid_str(const char *mapcpuid, const char *cpuid)
  */
 int __weak get_cpuid(char *buffer __maybe_unused, size_t sz __maybe_unused)
 {
-	return -1;
+	return ENOSYS; /* Not implemented */
 }
 
 static int write_cpuid(struct feat_fd *ff,
Signature · 48 lines
2.21.0

From c6c6a3e2eb6982e37294abcac389effd298cf730 Mon Sep 17 00:00:00 2001
From: Arnaldo Carvalho de Melo <[email protected]>
Date: Wed, 11 Dec 2019 10:21:59 -0300
Subject: [PATCH 2/2] perf top: Do not bail out when perf_env__read_cpuid()
 returns ENOSYS

'perf top' stopped working on hw architectures that do not provide a
get_cpuid() implementation and thus fallback to the weak get_cpuid()
default function.

This is done because at annotation time we may need it in the arch
specific annotation init routine, but that is only being used by arches
that do provide a get_cpuid() implementation:

  $ find tools/  -name "*.[ch]" | xargs grep 'evlist->env'
  tools/perf/builtin-top.c:	top.evlist->env = &perf_env;
  tools/perf/util/evsel.c:		return evsel->evlist->env;
  tools/perf/util/s390-cpumsf.c:	sf->machine_type = s390_cpumsf_get_type(session->evlist->env->cpuid);
  tools/perf/util/header.c:	session->evlist->env = &header->env;
  tools/perf/util/sample-raw.c:	const char *arch_pf = perf_env__arch(evlist->env);
  $

  $ find tools/perf/arch  -name "*.[ch]" | xargs grep -w get_cpuid
  tools/perf/arch/x86/util/auxtrace.c:	ret = get_cpuid(buffer, sizeof(buffer));
  tools/perf/arch/x86/util/header.c:get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/powerpc/util/header.c:get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/s390/util/header.c: * Implementation of get_cpuid().
  tools/perf/arch/s390/util/header.c:int get_cpuid(char *buffer, size_t sz)
  tools/perf/arch/s390/util/header.c:	if (buf && get_cpuid(buf, 128))
  $

For 'report' or 'script', i.e. tools working on perf.data files, that is
setup while reading the header, its just top that needs to explicitely
read it at tool start.

Reported-by: John Garry <[email protected]>
Analysed-by: Jiri Olsa <[email protected]>
Cc: Adrian Hunter <[email protected]>
Cc: Mark Rutland <[email protected]>
Cc: Namhyung Kim <[email protected]>
Cc: Will Deacon <[email protected]>
Link: https://lkml.kernel.org/n/[email protected]
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
---
 tools/perf/builtin-top.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)
tools/perf/builtin-top.c +7 -3
index dc80044bc46f..795e353de095 100644
@@ -1568,9 +1568,13 @@ int cmd_top(int argc, const char **argv)
 	 */
 	status = perf_env__read_cpuid(&perf_env);
 	if (status) {
-		pr_err("Couldn't read the cpuid for this machine: %s\n",
-		       str_error_r(errno, errbuf, sizeof(errbuf)));
-		goto out_delete_evlist;
+		/*
+		 * Some arches do not provide a get_cpuid(), so just use pr_debug, otherwise
+		 * warn the user explicitely.
+		 */
+		eprintf(status == ENOSYS ? 1 : 0, verbose,
+			"Couldn't read the cpuid for this machine: %s\n",
+			str_error_r(errno, errbuf, sizeof(errbuf)));
 	}
 	top.evlist->env = &perf_env;
 
Signature · 1 line
2.21.0
#13 lore raw

RE: perf top for arm64?

In reply to Jiri Olsa
Quoted 38 lines

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Jiri Olsa
Sent: 2019年12月11日 1:09
To: John Garry <[email protected]>
Cc: Arnaldo Carvalho de Melo <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Linuxarm <[email protected]>; [email protected]
Subject: Re: perf top for arm64?

Quoted 17 lines · John Garry
Quoted 8 lines · Jiri Olsa
Quoted 6 lines · John Garry

Hi all,

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top Couldn't read
the cpuid for this machine: No such file or directory

Hi Jirka,

Quoted 4 lines

there was recent change that check on cpuid and quits:
608127f73779 perf top: Initialize perf_env->cpuid, needed by the
per arch annotation init routine

Quoted 5 lines

ok, this is new code. I obviously didn't check the git history...

But, apart from this, there are many other places where get_cpuid() is called. I wonder what else we're missing out on, and whether we should still add it.

Quoted 5 lines

right, I was just wondering how come vendor events are working for you, but realized we have get_cpuid_str being called in there ;-)

I think we should add it as you have it prepared already, could you post it with bigger changelog that would explain where it's being used for arm?

Hi Jirka,

I reported metricgroup cannot work on ARM64 before, however, no one can come up with a solution, could you take a look how to fix it? Thanks a lot!

You can refer to below link for more info:

	[1] https://www.spinics.net/lists/linux-perf-users/msg09190.html (NACK by Will Deason)
	[2] https://www.spinics.net/lists/linux-perf-users/msg09324.html

Best Regards,
Joakim Zhang

Quoted 1 line

jirka

#14 lore raw

Re: perf top for arm64?

In reply to Jiri Olsa

Hi John,

Quoted 23 lines · Jiri Olsa
Quoted 16 lines · John Garry
Quoted 8 lines · Jiri Olsa
Quoted 6 lines · John Garry

Hi all,

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top
Couldn't read the cpuid for this machine: No such file or directory

Hi Jirka,

Quoted 3 lines

there was recent change that check on cpuid and quits:

    608127f73779 perf top: Initialize perf_env->cpuid, needed by the per arch annotation init routine
Quoted 5 lines

ok, this is new code. I obviously didn't check the git history...

But, apart from this, there are many other places where get_cpuid() is called. I wonder what else we're missing out on, and whether we should still add it.

Quoted 6 lines

right, I was just wondering how come vendor events are working for you, but realized we have get_cpuid_str being called in there ;-)

I think we should add it as you have it prepared already, could you post it with bigger changelog that would explain where it's being used for arm?

I've also seen the similar problem when I was looking to add support for 'perf kvm stat' on arm64 [1] (which though got stuck due to some other reasons for a very long time :(

It would be great if your patch can address this issue!

[1] https://lore.kernel.org/patchwork/patch/1087531/

Thanks,
Zenghui

#15 lore raw

Re: perf top for arm64?

In reply to Jiri Olsa

Em Tue, Dec 10, 2019 at 05:36:55PM +0100, Jiri Olsa escreveu:

Quoted 8 lines · John Garry
Quoted 6 lines

Hi all,

I find to my surprise that "perf top" does not work for arm64:

root@ubuntu:/home/john/linux# tools/perf/perf top
Couldn't read the cpuid for this machine: No such file or directory
Quoted 6 lines

there was recent change that check on cpuid and quits:
608127f73779 perf top: Initialize perf_env->cpuid, needed by the per arch annotation init routine

Arnaldo,
maybe this should be just a warning/info, because it seems to be related to annotations only..?

Right, my bad, I'll look into making this just a debug message and then check in the annotation code when this is really needed to show an error/popup window :-\

- Arnaldo

Quoted 114 lines

get_cpuid is defined only for s390/x86/powerpc, so I guess it won't work on the rest as well

jirka

Quoted 109 lines

That's v5.5-rc1 release.

It seems that we are just missing an arm64 version of get_cpuid() - with the patch below, I now get as hoped:

   PerfTop:   32857 irqs/sec  kernel:85.0%  exact:  0.0% lost: 0/0 drop: 0/0

[4000Hz cycles], (all, 64 CPUs)
-------------------------------------------------------------------------------

     8.99%  [kernel]          [k] arm_smmu_cmdq_issue_cmdlist
     5.80%  [kernel]          [k] __softirqentry_text_start
     4.49%  [kernel]          [k] _raw_spin_unlock_irqrestore
     3.48%  [kernel]          [k] el0_svc_common.constprop.2
     3.37%  [kernel]          [k] _raw_write_lock_irqsave
     3.28%  [kernel]          [k] __local_bh_enable_ip
     3.05%  [kernel]          [k] __blk_complete_request
     2.07%  [kernel]          [k] queued_spin_lock_slowpath
     1.93%  [vdso]            [.] 0x0000000000000484

Was this just missed? Or is there a good reason to omit?

Thanks,
John

--->8---

Subject: [PATCH] perf: Add perf top support for arm64

Copied from get_cpuid_str() essentially...

tools/perf/arch/arm64/util/header.c +54 -0
index a32e4b72a98f..ecd1f86e29cc 100644
@@ -1,10 +1,12 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <perf/cpumap.h>
+#include <util/cpumap.h>
 #include <internal/cpumap.h>
 #include <api/fs/fs.h>
 #include "debug.h"
 #include "header.h"
+#include <errno.h>
 
 #define MIDR "/regs/identification/midr_el1"
 #define MIDR_SIZE 19
@@ -12,6 +14,59 @@
 #define MIDR_VARIANT_SHIFT      20
 #define MIDR_VARIANT_MASK       (0xf << MIDR_VARIANT_SHIFT)
 
+int
+get_cpuid(char *buffer, size_t sz)
+{
+	char *buf = NULL;
+	char path[PATH_MAX];
+	const char *sysfs = sysfs__mountpoint();
+	int cpu;
+	u64 midr = 0;
+	FILE *file;
+
+	if (!sysfs)
+		return EINVAL;
+
+	buf = malloc(MIDR_SIZE);
+	if (!buf)
+		return EINVAL;
+
+	/* read midr from list of cpus mapped to this pmu */
+	for (cpu = 0; cpu < cpu__max_present_cpu(); cpu++) {
+		scnprintf(path, sz, "%s/devices/system/cpu/cpu%d"MIDR,
+				sysfs, cpu);
+
+		file = fopen(path, "r");
+		if (!file) {
+			pr_debug("fopen failed for file %s\n", path);
+			continue;
+		}
+
+		if (!fgets(buf, MIDR_SIZE, file)) {
+			fclose(file);
+			continue;
+		}
+		fclose(file);
+
+		/* Ignore/clear Variant[23:20] and
+		 * Revision[3:0] of MIDR
+		 */
+		midr = strtoul(buf, NULL, 16);
+		midr &= (~(MIDR_VARIANT_MASK | MIDR_REVISION_MASK));
+		scnprintf(buffer, MIDR_SIZE, "0x%016lx", midr);
+		/* got midr break loop */
+		break;
+	}
+
+	if (!midr) {
+		pr_err("failed to get cpuid string\n");
+		free(buf);
+		return EINVAL;
+	}
+	return 0;
+}
+
 
Signature · 1 line
- Arnaldo
merge bot Applied
#16 lore raw
In reply to John Garry
The following commit has been merged into the perf/core branch of tip:

The following commit has been merged into the perf/core branch of tip:

Commit-ID:     71f699078b154fcb1c9162fd0208ada9ce532ffc
Gitweb:        https://git.kernel.org/tip/71f699078b154fcb1c9162fd0208ada9ce532ffc
Author:        John Garry <[email protected]>
AuthorDate:    Wed, 06 Nov 2019 21:00:54 +08:00
Committer:     Arnaldo Carvalho de Melo <[email protected]>
CommitterDate: Wed, 06 Nov 2019 15:49:39 -03:00

perf tools: Fix cross compile for ARM64

Currently when cross compiling perf tool for ARM64 on my x86 machine I get this error:

arch/arm64/util/sym-handling.c:9:10: fatal error: gelf.h: No such file or directory
#include <gelf.h>

For the build, libelf is reported off:

Auto-detecting system features:
...
... libelf: [ OFF ]

Indeed, test-libelf is not built successfully:

more ./build/feature/test-libelf.make.output test-libelf.c:2:10: fatal error: libelf.h: No such file or directory

   #include <libelf.h>
          ^~~~~~~~~~

compilation terminated.

I have no such problems natively compiling on ARM64, and I did not previously have this issue for cross compiling. Fix by relocating the gelf.h include.

---
 tools/perf/arch/arm64/util/sym-handling.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
tools/perf/arch/arm64/util/sym-handling.c +2 -1
index 5df7889..8dfa3e5 100644
@@ -6,9 +6,10 @@
 
 #include "symbol.h" // for the elf__needs_adjust_symbols() prototype
 #include <stdbool.h>
-#include <gelf.h>
 
 #ifdef HAVE_LIBELF_SUPPORT
+#include <gelf.h>
+
 bool elf__needs_adjust_symbols(GElf_Ehdr ehdr)
 {
 	return ehdr.e_type == ET_EXEC ||