[PATCH v3] perf: add support for the xtensa architecture

[PATCH v3] perf: add support for the xtensa architecture

[Baruch Siach]

Tested using kernel tracepoints on a QEMU simulated environment. Kernel support
for perf depends on the patch "xtensa: enable HAVE_PERF_EVENTS", which is
scheduled for v3.14. Hardware performance counters are not supported under
xtensa yet.

Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
---
v2: use the "core ID" field from /proc/cpuinfo (Max Filippov)

v3: add testing and kernel dependency information (Ingo Molnar)
---
 tools/perf/perf.h | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/tools/perf/perf.h b/tools/perf/perf.h
index b079304bd53d..e441864dbe03 100644
--- a/tools/perf/perf.h
+++ b/tools/perf/perf.h
@@ -132,6 +132,13 @@
 #define CPUINFO_PROC	"CPU"
 #endif
 
+#ifdef __xtensa__
+#define mb()		asm volatile("memw" ::: "memory")
+#define wmb()		asm volatile("memw" ::: "memory")
+#define rmb()		asm volatile("" ::: "memory")
+#define CPUINFO_PROC	"core ID"
+#endif
+
 #define barrier() asm volatile ("" ::: "memory")
 
 #ifndef cpu_relax
-- 
1.8.5.2

Re: [PATCH v3] perf: add support for the xtensa architecture

[Ingo Molnar]

* Baruch Siach <baruch@tkos.co.il> wrote:

> Tested using kernel tracepoints on a QEMU simulated environment. Kernel support
> for perf depends on the patch "xtensa: enable HAVE_PERF_EVENTS", which is
> scheduled for v3.14. Hardware performance counters are not supported under
> xtensa yet.
> 
> Acked-by: Max Filippov <jcmvbkbc@gmail.com>
> Signed-off-by: Baruch Siach <baruch@tkos.co.il>

Acked-by: Ingo Molnar <mingo@kernel.org>

Thanks,

	Ingo

Re: [PATCH v3] perf: add support for the xtensa architecture

[Baruch Siach]

Hi Ingo,

On Mon, Jan 13, 2014 at 11:46:47AM +0100, Ingo Molnar wrote:
> * Baruch Siach <baruch@tkos.co.il> wrote:
> > Tested using kernel tracepoints on a QEMU simulated environment. Kernel 
> > support
> > for perf depends on the patch "xtensa: enable HAVE_PERF_EVENTS", which is
> > scheduled for v3.14. Hardware performance counters are not supported under
> > xtensa yet.
> > 
> > Acked-by: Max Filippov <jcmvbkbc@gmail.com>
> > Signed-off-by: Baruch Siach <baruch@tkos.co.il>
> 
> Acked-by: Ingo Molnar <mingo@kernel.org>

Thanks. So is it going into v3.14 by any chance?

baruch

-- 
     http://baruch.siach.name/blog/                  ~. .~   Tk Open Systems
=}------------------------------------------------ooO--U--Ooo------------{=
   - baruch@tkos.co.il - tel: +972.2.679.5364, http://www.tkos.co.il -

[tip:perf/urgent] perf tools: Add support for the xtensa architecture

[tip-bot for Baruch Siach]

Commit-ID:  3a46817f674389d5fb2f4dce19fbc73ebd1976bb
Gitweb:     http://git.kernel.org/tip/3a46817f674389d5fb2f4dce19fbc73ebd1976bb
Author:     Baruch Siach <baruch@tkos.co.il>
AuthorDate: Mon, 13 Jan 2014 12:27:35 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Mon, 20 Jan 2014 16:19:08 -0300

perf tools: Add support for the xtensa architecture

Tested using kernel tracepoints on a QEMU simulated environment.

Kernel support for perf depends on the patch "xtensa: enable
HAVE_PERF_EVENTS", which is scheduled for v3.14.

Hardware performance counters are not supported under xtensa yet.

Acked-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: linux-xtensa@linux-xtensa.org
Link: http://lkml.kernel.org/r/aafcdb22f04e2d3188d2938528939481be56b649.1389608855.git.baruch@tkos.co.il
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/perf.h | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/tools/perf/perf.h b/tools/perf/perf.h
index 3c2f213..7daa806 100644
--- a/tools/perf/perf.h
+++ b/tools/perf/perf.h
@@ -132,6 +132,13 @@
 #define CPUINFO_PROC	"CPU"
 #endif
 
+#ifdef __xtensa__
+#define mb()		asm volatile("memw" ::: "memory")
+#define wmb()		asm volatile("memw" ::: "memory")
+#define rmb()		asm volatile("" ::: "memory")
+#define CPUINFO_PROC	"core ID"
+#endif
+
 #define barrier() asm volatile ("" ::: "memory")
 
 #ifndef cpu_relax