From: Rong Xu <xur@google.com>
To: Masahiro Yamada <masahiroy@kernel.org>
Cc: Alice Ryhl <aliceryhl@google.com>,
Andrew Morton <akpm@linux-foundation.org>,
Arnd Bergmann <arnd@arndb.de>, Bill Wendling <morbo@google.com>,
Borislav Petkov <bp@alien8.de>, Breno Leitao <leitao@debian.org>,
Brian Gerst <brgerst@gmail.com>,
Dave Hansen <dave.hansen@linux.intel.com>,
David Li <davidxl@google.com>, Han Shen <shenhan@google.com>,
Heiko Carstens <hca@linux.ibm.com>,
"H. Peter Anvin" <hpa@zytor.com>, Ingo Molnar <mingo@redhat.com>,
Jann Horn <jannh@google.com>, Jonathan Corbet <corbet@lwn.net>,
Josh Poimboeuf <jpoimboe@kernel.org>,
Juergen Gross <jgross@suse.com>,
Justin Stitt <justinstitt@google.com>,
Kees Cook <kees@kernel.org>,
"Mike Rapoport (IBM)" <rppt@kernel.org>,
Nathan Chancellor <nathan@kernel.org>,
Nick Desaulniers <ndesaulniers@google.com>,
Nicolas Schier <nicolas@fjasle.eu>,
"Paul E. McKenney" <paulmck@kernel.org>,
Peter Zijlstra <peterz@infradead.org>,
Sami Tolvanen <samitolvanen@google.com>,
Thomas Gleixner <tglx@linutronix.de>,
Wei Yang <richard.weiyang@gmail.com>,
workflows@vger.kernel.org,
Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>,
Maksim Panchenko <max4bolt@gmail.com>,
x86@kernel.org, linux-arch@vger.kernel.org,
linux-doc@vger.kernel.org, linux-kbuild@vger.kernel.org,
linux-kernel@vger.kernel.org, llvm@lists.linux.dev,
Sriraman Tallam <tmsriram@google.com>
Subject: Re: [PATCH v4 4/6] AutoFDO: Enable -ffunction-sections for the AutoFDO build
Date: Mon, 21 Oct 2024 16:32:00 -0700 [thread overview]
Message-ID: <CAF1bQ=Tp8Dc=jpNNq+B+qj90bowod8dQ1JYRsM4q5ARdf=Jd_Q@mail.gmail.com> (raw)
In-Reply-To: <CAK7LNAQpFdHxAGk1SSRrJwyKA1XjfJLbyAeka7-YemJ1zEevnQ@mail.gmail.com>
The answers are the same as the reply in [PATCH v4 5/6]
On Sun, Oct 20, 2024 at 7:26 PM Masahiro Yamada <masahiroy@kernel.org> wrote:
>
> On Tue, Oct 15, 2024 at 6:33 AM Rong Xu <xur@google.com> wrote:
> >
> > Enable -ffunction-sections by default for the AutoFDO build.
> >
> > With -ffunction-sections, the compiler places each function in its own
> > section named .text.function_name instead of placing all functions in
> > the .text section. In the AutoFDO build, this allows the linker to
> > utilize profile information to reorganize functions for improved
> > utilization of iCache and iTLB.
> >
> > Co-developed-by: Han Shen <shenhan@google.com>
> > Signed-off-by: Han Shen <shenhan@google.com>
> > Signed-off-by: Rong Xu <xur@google.com>
> > Suggested-by: Sriraman Tallam <tmsriram@google.com>
> > ---
> > include/asm-generic/vmlinux.lds.h | 37 ++++++++++++++++++++++++-------
> > scripts/Makefile.autofdo | 2 +-
> > 2 files changed, 30 insertions(+), 9 deletions(-)
> >
> > diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
> > index 5df589c60401..ace617d1af9b 100644
> > --- a/include/asm-generic/vmlinux.lds.h
> > +++ b/include/asm-generic/vmlinux.lds.h
> > @@ -95,18 +95,25 @@
> > * With LTO_CLANG, the linker also splits sections by default, so we need
> > * these macros to combine the sections during the final link.
> > *
> > + * With LTO_CLANG, the linker also splits sections by default, so we need
> > + * these macros to combine the sections during the final link.
> > + *
> > * RODATA_MAIN is not used because existing code already defines .rodata.x
> > * sections to be brought in with rodata.
> > */
> > -#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG)
> > +#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG) || \
> > +defined(CONFIG_AUTOFDO_CLANG)
> > #define TEXT_MAIN .text .text.[0-9a-zA-Z_]*
> > +#else
> > +#define TEXT_MAIN .text
> > +#endif
> > +#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG)
> > #define DATA_MAIN .data .data.[0-9a-zA-Z_]* .data..L* .data..compoundliteral* .data.$__unnamed_* .data.$L*
> > #define SDATA_MAIN .sdata .sdata.[0-9a-zA-Z_]*
> > #define RODATA_MAIN .rodata .rodata.[0-9a-zA-Z_]* .rodata..L*
> > #define BSS_MAIN .bss .bss.[0-9a-zA-Z_]* .bss..L* .bss..compoundliteral*
> > #define SBSS_MAIN .sbss .sbss.[0-9a-zA-Z_]*
> > #else
> > -#define TEXT_MAIN .text
> > #define DATA_MAIN .data
> > #define SDATA_MAIN .sdata
> > #define RODATA_MAIN .rodata
> > @@ -549,6 +556,20 @@
> > __cpuidle_text_end = .; \
> > __noinstr_text_end = .;
> >
> > +#ifdef CONFIG_AUTOFDO_CLANG
> > +#define TEXT_HOT \
> > + __hot_text_start = .; \
> > + *(.text.hot .text.hot.*) \
> > + __hot_text_end = .;
> > +#define TEXT_UNLIKELY \
> > + __unlikely_text_start = .; \
> > + *(.text.unlikely .text.unlikely.*) \
> > + __unlikely_text_end = .;
> > +#else
> > +#define TEXT_HOT *(.text.hot .text.hot.*)
> > +#define TEXT_UNLIKELY *(.text.unlikely .text.unlikely.*)
> > +#endif
>
>
>
> Again, why is this conditional?
The condition is to ensure that we don't change the default kernel
build by any means. The new code will introduce a few new symbols.
>
>
> The only difference is *_start and *_end symbols are defined
> when CONFIG_AUTOFDO_CLANG=y.
>
> And, where are these symbols used?
These new symbols are currently unreferenced within the kernel source tree.
However, they provide a valuable means of identifying hot and cold
sections of text, and how large they are. I think they are useful information.
>
>
>
>
>
>
>
>
>
>
>
> > +
> > /*
> > * .text section. Map to function alignment to avoid address changes
> > * during second ld run in second ld pass when generating System.map
> > @@ -557,30 +578,30 @@
> > * code elimination or function-section is enabled. Match these symbols
> > * first when in these builds.
> > */
> > -#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG)
> > +#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG) || \
> > +defined(CONFIG_AUTOFDO_CLANG)
> > #define TEXT_TEXT \
> > ALIGN_FUNCTION(); \
> > *(.text.asan.* .text.tsan.*) \
> > *(.text.unknown .text.unknown.*) \
> > - *(.text.unlikely .text.unlikely.*) \
> > + TEXT_UNLIKELY \
> > . = ALIGN(PAGE_SIZE); \
> > - *(.text.hot .text.hot.*) \
> > + TEXT_HOT \
> > *(TEXT_MAIN .text.fixup) \
> > NOINSTR_TEXT \
> > *(.ref.text)
> > #else
> > #define TEXT_TEXT \
> > ALIGN_FUNCTION(); \
> > - *(.text.hot .text.hot.*) \
> > + TEXT_HOT \
> > *(TEXT_MAIN .text.fixup) \
> > - *(.text.unlikely .text.unlikely.*) \
> > + TEXT_UNLIKELY \
> > *(.text.unknown .text.unknown.*) \
> > NOINSTR_TEXT \
> > *(.ref.text) \
> > *(.text.asan.* .text.tsan.*)
> > #endif
> >
> > -
> > /* sched.text is aling to function alignment to secure we have same
> > * address even at second ld pass when generating System.map */
> > #define SCHED_TEXT \
> > diff --git a/scripts/Makefile.autofdo b/scripts/Makefile.autofdo
> > index 1c9f224bc221..9c9a530ef090 100644
> > --- a/scripts/Makefile.autofdo
> > +++ b/scripts/Makefile.autofdo
> > @@ -10,7 +10,7 @@ ifndef CONFIG_DEBUG_INFO
> > endif
> >
> > ifdef CLANG_AUTOFDO_PROFILE
> > - CFLAGS_AUTOFDO_CLANG += -fprofile-sample-use=$(CLANG_AUTOFDO_PROFILE)
> > + CFLAGS_AUTOFDO_CLANG += -fprofile-sample-use=$(CLANG_AUTOFDO_PROFILE) -ffunction-sections
> > endif
> >
> > ifdef CONFIG_LTO_CLANG_THIN
> > --
> > 2.47.0.rc1.288.g06298d1525-goog
> >
> >
>
>
> --
> Best Regards
> Masahiro Yamada
next prev parent reply other threads:[~2024-10-21 23:32 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-14 21:33 [PATCH v4 0/6] Add AutoFDO and Propeller support for Clang build Rong Xu
2024-10-14 21:33 ` [PATCH v4 1/6] Add AutoFDO " Rong Xu
2024-10-20 16:32 ` Masahiro Yamada
2024-10-21 22:43 ` Rong Xu
2024-10-23 6:44 ` Masahiro Yamada
2024-10-14 21:33 ` [PATCH v4 2/6] objtool: Fix unreachable instruction warnings for weak funcitons Rong Xu
2024-10-14 21:33 ` [PATCH v4 3/6] Change the symbols order when --ffuntion-sections is enabled Rong Xu
2024-10-21 2:15 ` Masahiro Yamada
2024-10-21 23:43 ` Rong Xu
2024-10-23 6:48 ` Masahiro Yamada
2024-10-14 21:33 ` [PATCH v4 4/6] AutoFDO: Enable -ffunction-sections for the AutoFDO build Rong Xu
2024-10-21 2:25 ` Masahiro Yamada
2024-10-21 23:32 ` Rong Xu [this message]
2024-10-23 6:49 ` Masahiro Yamada
2024-10-14 21:33 ` [PATCH v4 5/6] AutoFDO: Enable machine function split optimization for AutoFDO Rong Xu
2024-10-21 3:18 ` Masahiro Yamada
2024-10-21 23:28 ` Rong Xu
2024-10-23 6:49 ` Masahiro Yamada
2024-10-23 16:24 ` Rong Xu
2024-10-14 21:33 ` [PATCH v4 6/6] Add Propeller configuration for kernel build Rong Xu
2024-10-20 17:48 ` Masahiro Yamada
2024-10-22 0:00 ` Rong Xu
2024-10-23 7:06 ` Masahiro Yamada
2024-10-23 7:25 ` Arnd Bergmann
2024-10-23 7:28 ` Masahiro Yamada
2024-10-23 16:23 ` Rong Xu
2024-10-19 6:20 ` [PATCH v4 0/6] Add AutoFDO and Propeller support for Clang build Rong Xu
2024-10-20 3:20 ` Yonghong Song
2024-10-20 3:25 ` Nathan Chancellor
2024-10-21 21:12 ` Rong Xu
2024-10-20 3:31 ` Nathan Chancellor
2024-10-20 15:45 ` Masahiro Yamada
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='CAF1bQ=Tp8Dc=jpNNq+B+qj90bowod8dQ1JYRsM4q5ARdf=Jd_Q@mail.gmail.com' \
--to=xur@google.com \
--cc=akpm@linux-foundation.org \
--cc=aliceryhl@google.com \
--cc=arnd@arndb.de \
--cc=bp@alien8.de \
--cc=brgerst@gmail.com \
--cc=corbet@lwn.net \
--cc=dave.hansen@linux.intel.com \
--cc=davidxl@google.com \
--cc=hca@linux.ibm.com \
--cc=hpa@zytor.com \
--cc=jannh@google.com \
--cc=jgross@suse.com \
--cc=jpoimboe@kernel.org \
--cc=justinstitt@google.com \
--cc=kees@kernel.org \
--cc=leitao@debian.org \
--cc=linux-arch@vger.kernel.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kbuild@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=llvm@lists.linux.dev \
--cc=masahiroy@kernel.org \
--cc=max4bolt@gmail.com \
--cc=miguel.ojeda.sandonis@gmail.com \
--cc=mingo@redhat.com \
--cc=morbo@google.com \
--cc=nathan@kernel.org \
--cc=ndesaulniers@google.com \
--cc=nicolas@fjasle.eu \
--cc=paulmck@kernel.org \
--cc=peterz@infradead.org \
--cc=richard.weiyang@gmail.com \
--cc=rppt@kernel.org \
--cc=samitolvanen@google.com \
--cc=shenhan@google.com \
--cc=tglx@linutronix.de \
--cc=tmsriram@google.com \
--cc=workflows@vger.kernel.org \
--cc=x86@kernel.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