perf/urgent fixes

Hi Ingo,

Please consider pulling,

- Arnaldo

Build stats at the end of this message.

The following changes since commit 3cace81ea5bb0b3f2b97cab8e2c8e1fae2feb7ed:

Merge tag 'perf-urgent-for-mingo-20160815' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/urgent (2016-08-16 20:08:56 +0200)

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git tags/perf-urgent-for-mingo-20160818

for you to fetch changes up to c53412ee8c7ec31373a4176ff7f3a6b79296c05c:

perf evsel: Do not access outside hw cache name arrays (2016-08-18 16:39:46 -0300)

----------------------------------------------------------------
perf/urgent fixes:

- Do not access outside hw cache name arrays (Arnaldo Carvalho de Melo)

- Use addr_location::addr instead of ip for entries when unwinding using DWARF CFI, fixing the "srcline" information for userspace application callchains (Milian Wolff)

- Reinstate strlcpy() header guard with __UCLIBC__, fixing the build with uclibc, detected when building for the ARC architecture (Vineet Gupta)

----------------------------------------------------------------
Arnaldo Carvalho de Melo (1):
perf evsel: Do not access outside hw cache name arrays

Milian Wolff (1):
perf unwind: Use addr_location::addr instead of ip for entries

Vineet Gupta (1):
tools lib: Reinstate strlcpy() header guard with __UCLIBC__

 tools/include/linux/string.h             | 6 +++++-
 tools/perf/util/evsel.c                  | 6 +++---
 tools/perf/util/unwind-libdw.c           | 2 +-
 tools/perf/util/unwind-libunwind-local.c | 2 +-
 4 files changed, 10 insertions(+), 6 deletions(-)

Build stats, now only android-ndk doesn't build with NO_LIBELF=1, as it doesn't even have libelf to cross build with, and the non-cross builds build objtool, uClibc and an ARC architecture cross build image are in the works:

  # time dm
   1 77.199181517 alpine:3.4: Ok
   2 30.215940343 android-ndk:r12b-arm: Ok
   3 80.593965325 archlinux:latest: Ok
   4 45.102186377 centos:5: Ok
   5 67.078635441 centos:6: Ok
   6 73.582617061 centos:7: Ok
   7 65.462644128 debian:7: Ok
   8 69.919139691 debian:8: Ok
   9 77.192563446 debian:experimental: Ok
  10 80.651363745 fedora:20: Ok
  11 85.566954259 fedora:21: Ok
  12 87.547437162 fedora:22: Ok
  13 94.866327046 fedora:23: Ok
  14 93.829320269 fedora:24: Ok
  15 94.860933711 fedora:rawhide: Ok
  16 88.890456556 mageia:5: Ok
  17 85.622890293 opensuse:13.2: Ok
  18 87.232034972 opensuse:42.1: Ok
  19 94.807172866 opensuse:tumbleweed: Ok
  20 72.189153053 ubuntu:12.04.5: Ok
  21 86.275708891 ubuntu:14.04.4: Ok
  22 90.743742216 ubuntu:15.10: Ok
  23 86.931085419 ubuntu:16.04: Ok
  24 67.823595709 ubuntu:16.04-x-arm: Ok
  25 62.985390761 ubuntu:16.04-x-arm64: Ok
  26 64.727635984 ubuntu:16.04-x-powerpc64: Ok
  27 63.208264990 ubuntu:16.04-x-powerpc64el: Ok
  28 64.685772233 ubuntu:16.04-x-s390: Ok
  29 85.089091168 ubuntu:16.10: Ok
   2224.88s

real 37m6.200s
user 0m1.830s
sys 0m2.653s
#

From: Milian Wolff <[email protected]>

This fixes the srcline translation for call chains of user space applications.

Before we got:

    perf report --stdio --no-children -s sym,srcline -g address
     8.92%  [.] main                                 mandelbrot.h:41
            |
            |--3.70%--main +8390240
            |          __libc_start_main +139950056726769
            |          _start +8388650
            |
            |--2.74%--main +8390189
            |
             --2.08%--main +8390296
                       __libc_start_main +139950056726769
                       _start +8388650

     7.59%  [.] main                                 complex:1326
            |
            |--4.79%--main +8390203
            |          __libc_start_main +139950056726769
            |          _start +8388650
            |
             --2.80%--main +8390219

     7.12%  [.] __muldc3                             libgcc2.c:1945
            |
            |--3.76%--__muldc3 +139950060519490
            |          main +8390224
            |          __libc_start_main +139950056726769
            |          _start +8388650
            |
             --3.32%--__muldc3 +139950060519512
                       main +8390224

With this patch applied, we instead get:

    perf report --stdio --no-children -s sym,srcline -g address
     8.92%  [.] main                                 mandelbrot.h:41
            |
            |--3.70%--main mandelbrot.h:41
            |          __libc_start_main +241
            |          _start +4194346
            |
            |--2.74%--main mandelbrot.h:41
            |
             --2.08%--main mandelbrot.h:41
                       __libc_start_main +241
                       _start +4194346

     7.59%  [.] main                                 complex:1326
            |
            |--4.79%--main complex:1326
            |          __libc_start_main +241
            |          _start +4194346
            |
             --2.80%--main complex:1326

     7.12%  [.] __muldc3                             libgcc2.c:1945
            |
            |--3.76%--__muldc3 libgcc2.c:1945
            |          main mandelbrot.h:39
            |          __libc_start_main +241
            |          _start +4194346
            |
             --3.32%--__muldc3 libgcc2.c:1945
                       main mandelbrot.h:39

LPU-Reference: [email protected]

---
 tools/perf/util/unwind-libdw.c           | 2 +-
 tools/perf/util/unwind-libunwind-local.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

index cf5e250bc78e..783a53fb7a4e 100644
@@ -66,7 +66,7 @@ static int entry(u64 ip, struct unwind_info *ui)
 	if (__report_module(&al, ip, ui))
 		return -1;
 
-	e->ip  = ip;
+	e->ip  = al.addr;
 	e->map = al.map;
 	e->sym = al.sym;
 

index 97c0f8fc5561..20c2e5743903 100644
@@ -542,7 +542,7 @@ static int entry(u64 ip, struct thread *thread,
 	thread__find_addr_location(thread, PERF_RECORD_MISC_USER,
 				   MAP__FUNCTION, ip, &al);
 
-	e.ip = ip;
+	e.ip = al.addr;
 	e.map = al.map;
 	e.sym = al.sym;
 

2.7.4

Pulled, thanks a lot Arnaldo!

Ingo

From: Vineet Gupta <[email protected]>

perf tools build in recent kernels spews splat when cross compiling with uClibc:

|   CC       util/alias.o
| In file included from tools/perf/util/../ui/../util/cache.h:8:0,
|                 from tools/perf/util/../ui/helpline.h:7,
|                 from tools/perf/util/debug.h:8,
|                 from arch/../util/cpumap.h:9,
|                 from arch/../util/env.h:5,
|                 from arch/common.h:4,
|                 from arch/common.c:3:
| tools/include/linux/string.h:12:15: warning: redundant redeclaration of ?strlcpy? [-Wredundant-decls]
|  extern size_t strlcpy(char *dest, const char *src, size_t size);
               ^

This is after commit 61a6445e463a31 ("tools lib: Guard the strlcpy() header with __GLIBC__").

The problem is uClibc also defines __GLIBC__ for exported headers for applications. So add that specific check to not trip for uClibc.

---
 tools/include/linux/string.h | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

index b96879477311..f436d2420a18 100644
@@ -8,7 +8,11 @@ void *memdup(const void *src, size_t len);
 
 int strtobool(const char *s, bool *res);
 
-#ifdef __GLIBC__
+/*
+ * glibc based builds needs the extern while uClibc doesn't.
+ * However uClibc headers also define __GLIBC__ hence the hack below
+ */
+#if defined(__GLIBC__) && !defined(__UCLIBC__)
 extern size_t strlcpy(char *dest, const char *src, size_t size);
 #endif
 

2.7.4

From: Arnaldo Carvalho de Melo <[email protected]>

We have to check if the values are >= *_MAX, not just >, fix it.

''In file /tools/perf/util/evsel.c function __perf_evsel__hw_cache_name it appears that there is a bug that reads beyond the end of the buffer. The statement "if (type > PERF_COUNT_HW_CACHE_MAX)" allows type to be equal to the maximum value. Later, when statement "if (!perf_evsel__is_cache_op_valid(type, op))" is executed, the function can access array perf_evsel__hw_cache_stat[type] beyond the end of the buffer.

It appears to me that the statement "if (type > PERF_COUNT_HW_CACHE_MAX)" should be "if (type >= PERF_COUNT_HW_CACHE_MAX)"

Bug found with Coverity and manual code review. No attempts were made to execute the code with a maximum type value.''

Committer note:

Testing it:

  $ perf record -e $(echo $(perf list cache | cut -d \[ -f1) | sed 's/ /,/g') usleep 1
  [ perf record: Woken up 16 times to write data ]
  [ perf record: Captured and wrote 0.023 MB perf.data (34 samples) ]
  $ perf evlist
  L1-dcache-load-misses
  L1-dcache-loads
  L1-dcache-stores
  L1-icache-load-misses
  LLC-load-misses
  LLC-loads
  LLC-store-misses
  LLC-stores
  branch-load-misses
  branch-loads
  dTLB-load-misses
  dTLB-loads
  dTLB-store-misses
  dTLB-stores
  iTLB-load-misses
  iTLB-loads
  node-load-misses
  node-loads
  node-store-misses
  node-stores
  $ perf list cache

List of pre-defined events (to be used in -e):

    L1-dcache-load-misses        [Hardware cache event]
    L1-dcache-loads              [Hardware cache event]
    L1-dcache-stores             [Hardware cache event]
    L1-icache-load-misses        [Hardware cache event]
    LLC-load-misses              [Hardware cache event]
    LLC-loads                    [Hardware cache event]
    LLC-store-misses             [Hardware cache event]
    LLC-stores                   [Hardware cache event]
    branch-load-misses           [Hardware cache event]
    branch-loads                 [Hardware cache event]
    dTLB-load-misses             [Hardware cache event]
    dTLB-loads                   [Hardware cache event]
    dTLB-store-misses            [Hardware cache event]
    dTLB-stores                  [Hardware cache event]
    iTLB-load-misses             [Hardware cache event]
    iTLB-loads                   [Hardware cache event]
    node-load-misses             [Hardware cache event]
    node-loads                   [Hardware cache event]
    node-store-misses            [Hardware cache event]
    node-stores                  [Hardware cache event]

$

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=153351

---
 tools/perf/util/evsel.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

index d9b80ef881cd..21fd573106ed 100644
@@ -507,17 +507,17 @@ static int __perf_evsel__hw_cache_name(u64 config, char *bf, size_t size)
 	u8 op, result, type = (config >>  0) & 0xff;
 	const char *err = "unknown-ext-hardware-cache-type";
 
-	if (type > PERF_COUNT_HW_CACHE_MAX)
+	if (type >= PERF_COUNT_HW_CACHE_MAX)
 		goto out_err;
 
 	op = (config >>  8) & 0xff;
 	err = "unknown-ext-hardware-cache-op";
-	if (op > PERF_COUNT_HW_CACHE_OP_MAX)
+	if (op >= PERF_COUNT_HW_CACHE_OP_MAX)
 		goto out_err;
 
 	result = (config >> 16) & 0xff;
 	err = "unknown-ext-hardware-cache-result";
-	if (result > PERF_COUNT_HW_CACHE_RESULT_MAX)
+	if (result >= PERF_COUNT_HW_CACHE_RESULT_MAX)
 		goto out_err;
 
 	err = "invalid-cache";

2.7.4