From: Maynard Johnson <maynardj@us.ibm.com>
To: linuxppc-dev@ozlabs.org
Subject: [PATCH] Fix oprofile sampling of marked events on POWER7
Date: Fri, 01 May 2009 12:17:57 -0500 [thread overview]
Message-ID: <49FB2EC5.4040400@us.ibm.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 1296 bytes --]
Description
-----------
Change ppc64 oprofile kernel driver to use the SLOT bits (MMCRA[37:39]only on
older processors where those bits are defined.
Background
----------
The performance monitor unit of the 64-bit POWER processor family has the
ability to collect accurate instruction-level samples when profiling on marked
events (i.e., "PM_MRK_<event-name>"). In processors prior to POWER6, the MMCRA
register contained "slot information" that the oprofile kernel driver used to
adjust the value latched in the SIAR at the time of a PMU interrupt. But as of
POWER6, these slot bits in MMCRA are no longer necessary for oprofile to use,
since the SIAR itself holds the accurate sampled instruction address. With
POWER6, these MMCRA slot bits were zero'ed out by hardware so oprofile's use of
these slot bits was, in effect, a NOP. But with POWER7, these bits are no
longer zero'ed out; however, they serve some other purpose rather than slot
information. Thus, using these bits on POWER7 to adjust the SIAR value results
in samples being attributed to the wrong instructions. The attached patch
changes the oprofile kernel driver to ignore these slot bits on all newer
processors starting with POWER6.
Thanks.
-Maynard
Signed-off-by: Maynard Johnson <maynardj@us.ibm.com>
[-- Attachment #2: oprof-p7.patch --]
[-- Type: text/plain, Size: 1222 bytes --]
diff -paur linux/arch/powerpc/oprofile/op_model_power4.c linux-p7-oprofile-patch//arch/powerpc/oprofile/op_model_power4.c
--- linux/arch/powerpc/oprofile/op_model_power4.c 2009-05-01 08:20:21.000000000 -0500
+++ linux-p7-oprofile-patch//arch/powerpc/oprofile/op_model_power4.c 2009-05-01 08:20:05.000000000 -0500
@@ -26,6 +26,7 @@
static unsigned long reset_value[OP_MAX_COUNTER];
static int oprofile_running;
+static int use_slot_nums;
/* mmcr values are set in power4_reg_setup, used in power4_cpu_setup */
static u32 mmcr0_val;
@@ -61,6 +62,12 @@ static int power4_reg_setup(struct op_co
else
mmcr0_val |= MMCR0_PROBLEM_DISABLE;
+ if (__is_processor(PV_POWER4) || __is_processor(PV_POWER4p) ||
+ __is_processor(PV_970) || __is_processor(PV_970FX) ||
+ __is_processor(PV_970MP) || __is_processor(PV_970GX) ||
+ __is_processor(PV_POWER5) || __is_processor(PV_POWER5p))
+ use_slot_nums = 1;
+
return 0;
}
@@ -206,7 +213,7 @@ static unsigned long get_pc(struct pt_re
mmcra = mfspr(SPRN_MMCRA);
- if (mmcra & MMCRA_SAMPLE_ENABLE) {
+ if (use_slot_nums && (mmcra & MMCRA_SAMPLE_ENABLE)) {
slot = ((mmcra & MMCRA_SLOT) >> MMCRA_SLOT_SHIFT);
if (slot > 1)
pc += 4 * (slot - 1);
next reply other threads:[~2009-05-01 17:18 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-05-01 17:17 Maynard Johnson [this message]
2009-06-15 13:40 ` [PATCH] Fix oprofile sampling of marked events on POWER7 Maynard Johnson
2009-06-16 4:12 ` Benjamin Herrenschmidt
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=49FB2EC5.4040400@us.ibm.com \
--to=maynardj@us.ibm.com \
--cc=linuxppc-dev@ozlabs.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).