linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Namhyung Kim <namhyung@kernel.org>
To: Ian Rogers <irogers@google.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@redhat.com>,
	Arnaldo Carvalho de Melo <acme@kernel.org>,
	Mark Rutland <mark.rutland@arm.com>,
	Alexander Shishkin <alexander.shishkin@linux.intel.com>,
	Jiri Olsa <jolsa@kernel.org>,
	Adrian Hunter <adrian.hunter@intel.com>,
	Kan Liang <kan.liang@linux.intel.com>,
	John Garry <john.g.garry@oracle.com>,
	Will Deacon <will@kernel.org>,
	James Clark <james.clark@linaro.org>,
	Mike Leach <mike.leach@linaro.org>, Leo Yan <leo.yan@linux.dev>,
	Guo Ren <guoren@kernel.org>,
	Paul Walmsley <paul.walmsley@sifive.com>,
	Palmer Dabbelt <palmer@dabbelt.com>,
	Albert Ou <aou@eecs.berkeley.edu>, Nick Terrell <terrelln@fb.com>,
	"Steven Rostedt (Google)" <rostedt@goodmis.org>,
	Guilherme Amadio <amadio@gentoo.org>,
	Changbin Du <changbin.du@huawei.com>,
	Daniel Bristot de Oliveira <bristot@kernel.org>,
	Daniel Wagner <dwagner@suse.de>,
	Aditya Gupta <adityag@linux.ibm.com>,
	Athira Rajeev <atrajeev@linux.vnet.ibm.com>,
	Masahiro Yamada <masahiroy@kernel.org>,
	Kajol Jain <kjain@linux.ibm.com>,
	Huacai Chen <chenhuacai@kernel.org>,
	Bibo Mao <maobibo@loongson.cn>, Anup Patel <anup@brainfault.org>,
	Atish Patra <atishp@rivosinc.com>,
	Shenlin Liang <liangshenlin@eswincomputing.com>,
	Oliver Upton <oliver.upton@linux.dev>,
	"Steinar H. Gunderson" <sesse@google.com>,
	"Dr. David Alan Gilbert" <linux@treblig.org>,
	Chen Pei <cp0613@linux.alibaba.com>,
	Dima Kogan <dima@secretsauce.net>,
	Yury Norov <yury.norov@gmail.com>,
	Alexander Lobakin <aleksander.lobakin@intel.com>,
	linux-kernel@vger.kernel.org, linux-perf-users@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org, linux-csky@vger.kernel.org,
	linux-riscv@lists.infradead.org
Subject: Re: [PATCH v2 16/31] perf dwarf-regs: Pass accurate disassembly machine to get_dwarf_regnum
Date: Thu, 7 Nov 2024 11:18:50 -0800	[thread overview]
Message-ID: <Zy0SmlDrmksra_OD@google.com> (raw)
In-Reply-To: <CAP-5=fU4hTL1hfB7-FpMnFopJJriZAOXY_8iakW5yHC_gfhTWg@mail.gmail.com>

Hello,

Sorry for the late reply.

On Mon, Oct 07, 2024 at 08:46:59AM -0700, Ian Rogers wrote:
> On Mon, Oct 7, 2024 at 1:08 AM Masami Hiramatsu <mhiramat@kernel.org> wrote:
> >
> > On Sat,  5 Oct 2024 12:55:26 -0700
> > Ian Rogers <irogers@google.com> wrote:
> >
> > > Rather than pass 0/EM_NONE, use the value computed in the disasm
> > > struct arch. Switch the EM_NONE case to EM_HOST, rewriting EM_NONE if
> > > it were passed to get_dwarf_regnum. Pass a flags value as
> > > architectures like csky need the flags to determine the ABI variant.
> > >
> >
> > Does this change the command output when we use it for cross-build
> > environment? E.g. remote arch is different from host arch? If so,
> > please add output examples with/without this change.
> 
> The cases where this would apply are small as get_arch_regnum is only
> implemented for x86. get_dwarf_regnum likewise only works for x86 and
> it is only called by annotate.

Yep, it's used only in the data type profiling.  I don't expect it to
work in the cross environment yet.  This change moves it closer to the
working state though. :)


> In this code without this patch the behavior is to return -ENOTSUP, ie
> the code is set up to fail and this code just makes it not fail for
> the x86 case (when not on x86) with code that is well tested on x86.
> The code exists as x86 registers may be the same dwarf number but have
> different names: e.g. rax, eax, ax, al. I'm not sure this reaches a
> high complexity level for extensive testing. I'll see if I can grab an
> x86 perf.data file to analyze on ARM, but I don't think doing this
> should gate the series.

Yep, as I said it's not supported yet and I don't think that's the goal
of this patch series.  So probably it's ok to merge it without the
extensive testing.

Thanks,
Namhyung

> 
> > > Signed-off-by: Ian Rogers <irogers@google.com>
> > > ---
> > >  tools/perf/util/annotate.c           | 6 +++---
> > >  tools/perf/util/dwarf-regs.c         | 8 ++++++--
> > >  tools/perf/util/include/dwarf-regs.h | 5 +++--
> > >  3 files changed, 12 insertions(+), 7 deletions(-)
> > >
> > > diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
> > > index 37ce43c4eb8f..b1d98da79be8 100644
> > > --- a/tools/perf/util/annotate.c
> > > +++ b/tools/perf/util/annotate.c
> > > @@ -2292,7 +2292,7 @@ static int extract_reg_offset(struct arch *arch, const char *str,
> > >       if (regname == NULL)
> > >               return -1;
> > >
> > > -     op_loc->reg1 = get_dwarf_regnum(regname, 0);
> > > +     op_loc->reg1 = get_dwarf_regnum(regname, arch->e_machine, arch->e_flags);
> > >       free(regname);
> > >
> > >       /* Get the second register */
> > > @@ -2305,7 +2305,7 @@ static int extract_reg_offset(struct arch *arch, const char *str,
> > >               if (regname == NULL)
> > >                       return -1;
> > >
> > > -             op_loc->reg2 = get_dwarf_regnum(regname, 0);
> > > +             op_loc->reg2 = get_dwarf_regnum(regname, arch->e_machine, arch->e_flags);
> > >               free(regname);
> > >       }
> > >       return 0;
> > > @@ -2405,7 +2405,7 @@ int annotate_get_insn_location(struct arch *arch, struct disasm_line *dl,
> > >                               return -1;
> > >
> > >                       if (*s == arch->objdump.register_char)
> > > -                             op_loc->reg1 = get_dwarf_regnum(s, 0);
> > > +                             op_loc->reg1 = get_dwarf_regnum(s, arch->e_machine, arch->e_flags);
> > >                       else if (*s == arch->objdump.imm_char) {
> > >                               op_loc->offset = strtol(s + 1, &p, 0);
> > >                               if (p && p != s + 1)
> > > diff --git a/tools/perf/util/dwarf-regs.c b/tools/perf/util/dwarf-regs.c
> > > index 7c01bc4d7e5b..1321387f6948 100644
> > > --- a/tools/perf/util/dwarf-regs.c
> > > +++ b/tools/perf/util/dwarf-regs.c
> > > @@ -70,7 +70,7 @@ __weak int get_arch_regnum(const char *name __maybe_unused)
> > >  }
> > >
> > >  /* Return DWARF register number from architecture register name */
> > > -int get_dwarf_regnum(const char *name, unsigned int machine)
> > > +int get_dwarf_regnum(const char *name, unsigned int machine, unsigned int flags __maybe_unused)
> > >  {
> > >       char *regname = strdup(name);
> > >       int reg = -1;
> > > @@ -84,8 +84,12 @@ int get_dwarf_regnum(const char *name, unsigned int machine)
> > >       if (p)
> > >               *p = '\0';
> > >
> > > +     if (machine == EM_NONE) {
> > > +             /* Generic arch - use host arch */
> > > +             machine = EM_HOST;
> > > +     }
> > >       switch (machine) {
> > > -     case EM_NONE:   /* Generic arch - use host arch */
> > > +     case EM_HOST:
> > >               reg = get_arch_regnum(regname);
> > >               break;
> > >       default:
> > > diff --git a/tools/perf/util/include/dwarf-regs.h b/tools/perf/util/include/dwarf-regs.h
> > > index f4f87ded5e3d..ee0a734564c7 100644
> > > --- a/tools/perf/util/include/dwarf-regs.h
> > > +++ b/tools/perf/util/include/dwarf-regs.h
> > > @@ -93,12 +93,13 @@ int get_arch_regnum(const char *name);
> > >   * name: architecture register name
> > >   * machine: ELF machine signature (EM_*)
> > >   */
> > > -int get_dwarf_regnum(const char *name, unsigned int machine);
> > > +int get_dwarf_regnum(const char *name, unsigned int machine, unsigned int flags);
> > >
> > >  #else /* HAVE_LIBDW_SUPPORT */
> > >
> > >  static inline int get_dwarf_regnum(const char *name __maybe_unused,
> > > -                                unsigned int machine __maybe_unused)
> > > +                                unsigned int machine __maybe_unused,
> > > +                                unsigned int flags __maybe_unused)
> > >  {
> > >       return -1;
> > >  }
> > > --
> > > 2.47.0.rc0.187.ge670bccf7e-goog
> > >
> >
> >
> > --
> > Masami Hiramatsu (Google) <mhiramat@kernel.org>

  reply	other threads:[~2024-11-07 19:18 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-10-05 19:55 [PATCH v2 00/31] Libdw/dwarf build clean up Ian Rogers
2024-10-05 19:55 ` [PATCH v2 01/31] perf build: Fix !HAVE_DWARF_GETLOCATIONS_SUPPORT Ian Rogers
2024-10-05 19:55 ` [PATCH v2 02/31] perf build: Rename NO_DWARF to NO_LIBDW Ian Rogers
2024-10-05 19:55 ` [PATCH v2 03/31] perf build: Remove defined but never used variable Ian Rogers
2024-10-05 19:55 ` [PATCH v2 04/31] perf build: Rename test-dwarf to test-libdw Ian Rogers
2024-10-05 19:55 ` [PATCH v2 05/31] perf build: Combine libdw-dwarf-unwind into libdw feature tests Ian Rogers
2024-10-05 19:55 ` [PATCH v2 06/31] perf build: Combine test-dwarf-getlocations into test-libdw Ian Rogers
2024-10-05 19:55 ` [PATCH v2 07/31] perf build: Combine test-dwarf-getcfi " Ian Rogers
2024-10-05 19:55 ` [PATCH v2 08/31] perf probe: Move elfutils support check to libdw check Ian Rogers
2024-10-05 19:55 ` [PATCH v2 09/31] perf libdw: Remove unnecessary defines Ian Rogers
2024-10-05 19:55 ` [PATCH v2 10/31] perf build: Rename HAVE_DWARF_SUPPORT to HAVE_LIBDW_SUPPORT Ian Rogers
2024-10-05 19:55 ` [PATCH v2 11/31] perf build: Rename CONFIG_DWARF to CONFIG_LIBDW Ian Rogers
2024-10-05 19:55 ` [PATCH v2 12/31] perf bpf-prologue: Remove unused file Ian Rogers
2024-10-05 19:55 ` [PATCH v2 13/31] perf dwarf-regs: Remove PERF_HAVE_ARCH_REGS_QUERY_REGISTER_OFFSET Ian Rogers
2024-10-05 19:55 ` [PATCH v2 14/31] perf dwarf-regs: Add EM_HOST and EF_HOST defines Ian Rogers
2024-10-05 19:55 ` [PATCH v2 15/31] perf disasm: Add e_machine/e_flags to struct arch Ian Rogers
2024-10-05 19:55 ` [PATCH v2 16/31] perf dwarf-regs: Pass accurate disassembly machine to get_dwarf_regnum Ian Rogers
2024-10-07  8:07   ` Masami Hiramatsu
2024-10-07 15:46     ` Ian Rogers
2024-11-07 19:18       ` Namhyung Kim [this message]
2024-10-05 19:55 ` [PATCH v2 17/31] perf dwarf-regs: Pass ELF flags to get_dwarf_regstr Ian Rogers
2024-10-07  8:09   ` Masami Hiramatsu
2024-10-07 15:05     ` Ian Rogers
2024-11-07 19:24       ` Namhyung Kim
2024-10-05 19:55 ` [PATCH v2 18/31] perf dwarf-regs: Move x86 dwarf-regs out of arch Ian Rogers
2024-10-07  8:34   ` Masami Hiramatsu
2024-10-07 15:32     ` Ian Rogers
2024-10-05 19:55 ` [PATCH v2 19/31] perf arm64: Remove dwarf-regs.c Ian Rogers
2024-10-07  8:19   ` Masami Hiramatsu
2024-10-05 19:55 ` [PATCH v2 20/31] perf arm: " Ian Rogers
2024-10-05 19:55 ` [PATCH v2 21/31] perf dwarf-regs: Move csky dwarf-regs out of arch Ian Rogers
2024-10-05 19:55 ` [PATCH v2 22/31] perf loongarch: Remove dwarf-regs.c Ian Rogers
2024-10-05 19:55 ` [PATCH v2 23/31] perf mips: " Ian Rogers
2024-10-05 19:55 ` [PATCH v2 24/31] perf dwarf-regs: Move powerpc dwarf-regs out of arch Ian Rogers
2024-10-05 19:55 ` [PATCH v2 25/31] perf riscv: Remove dwarf-regs.c and add dwarf-regs-table.h Ian Rogers
2024-10-05 19:55 ` [PATCH v2 26/31] perf s390: Remove dwarf-regs.c Ian Rogers
2024-10-05 19:55 ` [PATCH v2 27/31] perf sh: " Ian Rogers
2024-10-05 19:55 ` [PATCH v2 28/31] perf sparc: " Ian Rogers
2024-10-05 19:55 ` [PATCH v2 29/31] perf xtensa: " Ian Rogers
2024-10-05 19:55 ` [PATCH v2 30/31] perf dwarf-regs: Remove get_arch_regstr code Ian Rogers
2024-10-05 19:55 ` [PATCH v2 31/31] perf build: Remove PERF_HAVE_DWARF_REGS Ian Rogers
2024-10-16 17:21 ` [PATCH v2 00/31] Libdw/dwarf build clean up Ian Rogers

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=Zy0SmlDrmksra_OD@google.com \
    --to=namhyung@kernel.org \
    --cc=acme@kernel.org \
    --cc=adityag@linux.ibm.com \
    --cc=adrian.hunter@intel.com \
    --cc=aleksander.lobakin@intel.com \
    --cc=alexander.shishkin@linux.intel.com \
    --cc=amadio@gentoo.org \
    --cc=anup@brainfault.org \
    --cc=aou@eecs.berkeley.edu \
    --cc=atishp@rivosinc.com \
    --cc=atrajeev@linux.vnet.ibm.com \
    --cc=bristot@kernel.org \
    --cc=changbin.du@huawei.com \
    --cc=chenhuacai@kernel.org \
    --cc=cp0613@linux.alibaba.com \
    --cc=dima@secretsauce.net \
    --cc=dwagner@suse.de \
    --cc=guoren@kernel.org \
    --cc=irogers@google.com \
    --cc=james.clark@linaro.org \
    --cc=john.g.garry@oracle.com \
    --cc=jolsa@kernel.org \
    --cc=kan.liang@linux.intel.com \
    --cc=kjain@linux.ibm.com \
    --cc=leo.yan@linux.dev \
    --cc=liangshenlin@eswincomputing.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-csky@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=linux-riscv@lists.infradead.org \
    --cc=linux@treblig.org \
    --cc=maobibo@loongson.cn \
    --cc=mark.rutland@arm.com \
    --cc=masahiroy@kernel.org \
    --cc=mhiramat@kernel.org \
    --cc=mike.leach@linaro.org \
    --cc=mingo@redhat.com \
    --cc=oliver.upton@linux.dev \
    --cc=palmer@dabbelt.com \
    --cc=paul.walmsley@sifive.com \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.org \
    --cc=sesse@google.com \
    --cc=terrelln@fb.com \
    --cc=will@kernel.org \
    --cc=yury.norov@gmail.com \
    /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).