[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAFP8O3J2pUddEfL+W2j=tpjdj1ufsQMJqjcARbYvY3+khtBPig@mail.gmail.com>
Date: Tue, 22 Aug 2023 00:05:56 -0700
From: Fangrui Song <maskray@...gle.com>
To: Denis Nikitin <denik@...omium.org>
Cc: linux-kbuild@...r.kernel.org,
Masahiro Yamada <masahiroy@...nel.org>,
Nathan Chancellor <nathan@...nel.org>,
Nick Desaulniers <ndesaulniers@...gle.com>,
Nicolas Schier <nicolas@...sle.eu>, Tom Rix <trix@...hat.com>,
linux-kernel@...r.kernel.org, llvm@...ts.linux.dev
Subject: Re: [PATCH] modpost: Skip .llvm.call-graph-profile section check
On Mon, Aug 21, 2023 at 11:54 PM Denis Nikitin <denik@...omium.org> wrote:
>
> .llvm.call-graph-profile section is added when the kernel is built with
> profiles (e.g. -fprofile-sample-use=<llvm.profile>).
Right. .llvm.call-graph-profile may also be added when the kernel is
built with clang -fprofile-use= (though instrumentation-based PGO
support is not upstreamed yet).
> The section holds
> metadata for symbols beloning to other sections and hence doesn't need
> modpost checks.
Typo: belonging.
.llvm.call-graph-profile contains edge information derived from text
sections, so .llvm.call-graph-profile itself doesn't need more
analysis as the text sections have been analyzed.
> This change fixes the kernel build with sample profiles which fails
> with:
> "FATAL: modpost: Please add code to calculate addend for this architecture"
>
> Signed-off-by: Denis Nikitin <denik@...omium.org>
> ---
> scripts/mod/modpost.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/scripts/mod/modpost.c b/scripts/mod/modpost.c
> index b29b29707f10..64bd13f7199c 100644
> --- a/scripts/mod/modpost.c
> +++ b/scripts/mod/modpost.c
> @@ -761,6 +761,7 @@ static const char *const section_white_list[] =
> ".fmt_slot*", /* EZchip */
> ".gnu.lto*",
> ".discard.*",
> + ".llvm.call-graph-profile", /* call graph */
> NULL
> };
>
> --
> 2.42.0.rc1.204.g551eb34607-goog
>
>
Otherwise the change looks good to me.
--
宋方睿
Powered by blists - more mailing lists