From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Anton Blanchard <anton@samba.org>
Cc: linuxppc-dev@ozlabs.org, mingo@elte.hu, paulus@samba.org,
a.p.zijlstra@chello.nl, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] powerpc/perf_counter: Add alignment-faults and emulation-faults sw events
Date: Mon, 27 Jul 2009 11:47:52 +1000 [thread overview]
Message-ID: <1248659272.28464.29.camel@pasglop> (raw)
In-Reply-To: <20090706121845.GD4391@kryten>
On Mon, 2009-07-06 at 22:18 +1000, Anton Blanchard wrote:
> Hook up the alignment-faults and emulation-faults events for powerpc.
>
> Signed-off-by: Anton Blanchard <anton@samba.org>
> ---
>
> Lots of duplication between PPC_WARN_EMULATED() and perf_swcounter_event()
> here. Maybe we need to create PPC_WARN_ALIGNMENT(), use it and hide all
> calls to perf_swcounter_event in the macros.
Sounds good.
BTW. The patch relies on some perctr changes I don't have in my tree
(PERF_COUNT_SW_ALIGNMENT_FAULTS isn't defined for me)
Ingo, how do you want to handle that ? Should I wait til round 2 of the
merge window before putting Anton's patch in or can I merge some of your
stuff in powerpc-next as pre-reqs ? Or you can just take Anton's patch
with my:
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
And stick it in your own queue (though pls, give me a pointer to the git
tree in question so I can verify that I don't get new stuff that
conflicts before we get to the merge window).
Cheers,
Ben.
> Index: linux.trees.git/arch/powerpc/kernel/align.c
> ===================================================================
> --- linux.trees.git.orig/arch/powerpc/kernel/align.c 2009-07-06 21:50:53.000000000 +1000
> +++ linux.trees.git/arch/powerpc/kernel/align.c 2009-07-06 22:10:41.000000000 +1000
> @@ -25,6 +25,7 @@
> #include <asm/cache.h>
> #include <asm/cputable.h>
> #include <asm/emulated_ops.h>
> +#include <linux/perf_counter.h>
>
> struct aligninfo {
> unsigned char len;
> @@ -707,6 +708,9 @@
> */
> CHECK_FULL_REGS(regs);
>
> + perf_swcounter_event(PERF_COUNT_SW_ALIGNMENT_FAULTS, 1, 0, regs,
> + regs->nip);
> +
> dsisr = regs->dsisr;
>
> /* Some processors don't provide us with a DSISR we can use here,
> Index: linux.trees.git/arch/powerpc/kernel/traps.c
> ===================================================================
> --- linux.trees.git.orig/arch/powerpc/kernel/traps.c 2009-07-06 21:50:53.000000000 +1000
> +++ linux.trees.git/arch/powerpc/kernel/traps.c 2009-07-06 22:11:52.000000000 +1000
> @@ -34,6 +34,7 @@
> #include <linux/bug.h>
> #include <linux/kdebug.h>
> #include <linux/debugfs.h>
> +#include <linux/perf_counter.h>
>
> #include <asm/emulated_ops.h>
> #include <asm/pgtable.h>
> @@ -759,6 +760,8 @@
>
> /* Emulate the mfspr rD, PVR. */
> if ((instword & PPC_INST_MFSPR_PVR_MASK) == PPC_INST_MFSPR_PVR) {
> + perf_swcounter_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, 0,
> + regs, regs->nip);
> PPC_WARN_EMULATED(mfpvr);
> rd = (instword >> 21) & 0x1f;
> regs->gpr[rd] = mfspr(SPRN_PVR);
> @@ -767,6 +770,8 @@
>
> /* Emulating the dcba insn is just a no-op. */
> if ((instword & PPC_INST_DCBA_MASK) == PPC_INST_DCBA) {
> + perf_swcounter_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, 0,
> + regs, regs->nip);
> PPC_WARN_EMULATED(dcba);
> return 0;
> }
> @@ -776,6 +781,8 @@
> int shift = (instword >> 21) & 0x1c;
> unsigned long msk = 0xf0000000UL >> shift;
>
> + perf_swcounter_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, 0,
> + regs, regs->nip);
> PPC_WARN_EMULATED(mcrxr);
> regs->ccr = (regs->ccr & ~msk) | ((regs->xer >> shift) & msk);
> regs->xer &= ~0xf0000000UL;
> @@ -784,18 +791,24 @@
>
> /* Emulate load/store string insn. */
> if ((instword & PPC_INST_STRING_GEN_MASK) == PPC_INST_STRING) {
> + perf_swcounter_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, 0,
> + regs, regs->nip);
> PPC_WARN_EMULATED(string);
> return emulate_string_inst(regs, instword);
> }
>
> /* Emulate the popcntb (Population Count Bytes) instruction. */
> if ((instword & PPC_INST_POPCNTB_MASK) == PPC_INST_POPCNTB) {
> + perf_swcounter_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, 0,
> + regs, regs->nip);
> PPC_WARN_EMULATED(popcntb);
> return emulate_popcntb_inst(regs, instword);
> }
>
> /* Emulate isel (Integer Select) instruction */
> if ((instword & PPC_INST_ISEL_MASK) == PPC_INST_ISEL) {
> + perf_swcounter_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, 0,
> + regs, regs->nip);
> PPC_WARN_EMULATED(isel);
> return emulate_isel(regs, instword);
> }
> @@ -994,8 +1007,11 @@
>
> #ifdef CONFIG_MATH_EMULATION
> errcode = do_mathemu(regs);
> - if (errcode >= 0)
> + if (errcode >= 0) {
> + perf_swcounter_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, 0,
> + regs, regs->nip);
> PPC_WARN_EMULATED(math);
> + }
>
> switch (errcode) {
> case 0:
> @@ -1017,8 +1033,11 @@
>
> #elif defined(CONFIG_8XX_MINIMAL_FPEMU)
> errcode = Soft_emulate_8xx(regs);
> - if (errcode >= 0)
> + if (errcode >= 0) {
> + perf_swcounter_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, 0,
> + regs, regs->nip);
> PPC_WARN_EMULATED(8xx);
> + }
>
> switch (errcode) {
> case 0:
> @@ -1129,6 +1148,8 @@
>
> flush_altivec_to_thread(current);
>
> + perf_swcounter_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, 0, regs,
> + regs->nip);
> PPC_WARN_EMULATED(altivec);
> err = emulate_altivec(regs);
> if (err == 0) {
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
next prev parent reply other threads:[~2009-07-27 1:48 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-07-06 12:18 [PATCH] powerpc/perf_counter: Add alignment-faults and emulation-faults sw events Anton Blanchard
2009-07-27 1:47 ` Benjamin Herrenschmidt [this message]
2009-07-27 1:51 ` Anton Blanchard
2009-07-27 4:57 ` Benjamin Herrenschmidt
2009-07-27 6:26 ` Peter Zijlstra
2009-07-27 6:27 ` Peter Zijlstra
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=1248659272.28464.29.camel@pasglop \
--to=benh@kernel.crashing.org \
--cc=a.p.zijlstra@chello.nl \
--cc=anton@samba.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@ozlabs.org \
--cc=mingo@elte.hu \
--cc=paulus@samba.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).