public inbox for linux-kbuild@vger.kernel.org
 help / color / mirror / Atom feed
From: Nathan Chancellor <nathan@kernel.org>
To: Masahiro Yamada <masahiroy@kernel.org>
Cc: linux-kbuild@vger.kernel.org, linux-kernel@vger.kernel.org,
	Nick Desaulniers <ndesaulniers@google.com>,
	Tom Rix <trix@redhat.com>,
	llvm@lists.linux.dev
Subject: Re: [PATCH 2/3] Kconfig.debug: add toolchain checks for DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT
Date: Mon, 3 Oct 2022 13:32:18 -0700	[thread overview]
Message-ID: <YztG0rb+cgf0IfPo@dev-arch.thelio-3990X> (raw)
In-Reply-To: <CAK7LNASnp4DA9AADx4Nj9b_h6r5wfPFUaonfmOfaHaSizpyfgg@mail.gmail.com>

On Tue, Oct 04, 2022 at 05:18:49AM +0900, Masahiro Yamada wrote:
> On Tue, Oct 4, 2022 at 2:10 AM Nathan Chancellor <nathan@kernel.org> wrote:
> >
> > On Mon, Oct 03, 2022 at 03:11:06AM +0900, Masahiro Yamada wrote:
> > > CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT passes the -g option to the
> > > command line. The actual DWARF version is up to the toolchain.
> > >
> > > The combination of GCC and GAS works fine, and Clang with the integrated
> > > assembler is good too.
> > >
> > > The combination of Clang and GAS is a bit tricky, but at least, the
> > > default -g flag worked until LLVM 14 was released because Clang <=13
> > > defaults to DWARF v4.
> > >
> > > Clang 14 switched to DWARF v5 by default.
> > >
> > > Now, CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT has the same issue as
> > > addressed by commit 98cd6f521f10 ("Kconfig: allow explicit opt in to
> > > DWARF v5").
> > >
> > > CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y for Clang >= 14 and
> > > GAS < 2.35 produces a ton of errors like follows:
> > >
> > >   /tmp/main-c2741c.s: Assembler messages:
> > >   /tmp/main-c2741c.s:109: Error: junk at end of line, first unrecognized character is `"'
> > >   /tmp/main-c2741c.s:109: Error: file number less than one
> > >
> > > Add 'depends on' to check toolchains.
> > >
> > > Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
> > > ---
> > >
> > >  lib/Kconfig.debug | 1 +
> > >  1 file changed, 1 insertion(+)
> > >
> > > diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
> > > index f4b2165f24db..cc90414d492e 100644
> > > --- a/lib/Kconfig.debug
> > > +++ b/lib/Kconfig.debug
> > > @@ -253,6 +253,7 @@ config DEBUG_INFO_NONE
> > >  config DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT
> > >       bool "Rely on the toolchain's implicit default DWARF version"
> > >       select DEBUG_INFO
> > > +     depends on !CC_IS_CLANG || AS_IS_LLVM || CLANG_VERSION < 140000 || (AS_IS_GNU && AS_VERSION >= 23502)
> > >       help
> > >         The implicit default version of DWARF debug info produced by a
> > >         toolchain changes over time.
> > > --
> > > 2.34.1
> > >
> > >
> >
> > Could we do something like this instead? Bikeshed the configuration
> > names as much as you want but we are duplicating most of this
> > conditional in a few different places so it would be good to consolidate
> > those in my opinion. Even though the DWARF4 and DWARF5 conditions are
> > the same, they might diverge so I think it makes sense to keep them
> > separate.
> >
> > diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
> > index cc90414d492e..8e34a639cd40 100644
> > --- a/lib/Kconfig.debug
> > +++ b/lib/Kconfig.debug
> > @@ -231,6 +231,17 @@ config DEBUG_INFO
> >           in the "Debug information" choice below, indicating that debug
> >           information will be generated for build targets.
> >
> > +config CC_AND_AS_SUPPORT_DWARF4
> > +       def_bool !CC_IS_CLANG || AS_IS_LLVM || (AS_IS_GNU && AS_VERSION >= 23502)
> > +
> > +config CC_AND_AS_SUPPORT_DWARF5
> > +       def_bool !CC_IS_CLANG || AS_IS_LLVM || (AS_IS_GNU && AS_VERSION >= 23502)
> > +
> > +config CC_AND_AS_SUPPORT_DWARF_DEFAULT
> > +       bool
> > +       default y if CC_AND_AS_SUPPORT_DWARF5 && (CLANG_VERSION >= 140000 || GCC_VERSION >= 110000)
> > +       default y if CC_AND_AS_SUPPORT_DWARF4 && (CLANG_VERSION < 140000 || GCC_VERSION < 110000)
> 
> 
> No, this is an overkill.
> 
> In my patch, you can enable DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT
> for the combination of:
>     CC_IS_CLANG=y
>     CLANG_VERSION < 140000
>     AS_IS_GNU=y
>     AS_VERSION < 23502
> 
> But, your code disallows it.
> 
> You are requiring binutils 2.35.2 for Clang+GAS, but
> it is not necessary.

Ah, I figured that part was necessary because of the condition added to
CONFIG_DEBUG_INFO_DWARF4 by commit 32ef9e5054ec ("Makefile.debug:
re-enable debug info for .S files") but I see now that was because of
some binutils behavior that you pointed out on Nick's v2 of that change,
which I failed to pay attention to because I had already reviewed the
change and did not read subsequent reviews/revisions as I should have,
so my apologies.

https://lore.kernel.org/CAK7LNATWDH01=ZKLnsxc0vcib1zGDbEq8jLQwhWP7HkkmSb_Mw@mail.gmail.com/

Regardless, I still think the duplication in these conditions is
unfortunate and I would like to see it cleaned up in some way but I
don't have a strong opinion on the matter. To the original change:

Reviewed-by: Nathan Chancellor <nathan@kernel.org>

Once these are applied to linux-kbuild, I will send a v2 of my RISC-V
fix.

> >  choice
> >         prompt "Debug information"
> >         depends on DEBUG_KERNEL
> > @@ -253,7 +264,7 @@ config DEBUG_INFO_NONE
> >  config DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT
> >         bool "Rely on the toolchain's implicit default DWARF version"
> >         select DEBUG_INFO
> > -       depends on !CC_IS_CLANG || AS_IS_LLVM || CLANG_VERSION < 140000 || (AS_IS_GNU && AS_VERSION >= 23502)
> > +       depends on CC_AND_AS_SUPPORT_DWARF_DEFAULT
> >         help
> >           The implicit default version of DWARF debug info produced by a
> >           toolchain changes over time.
> > @@ -265,7 +276,7 @@ config DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT
> >  config DEBUG_INFO_DWARF4
> >         bool "Generate DWARF Version 4 debuginfo"
> >         select DEBUG_INFO
> > -       depends on !CC_IS_CLANG || AS_IS_LLVM || (AS_IS_GNU && AS_VERSION >= 23502)
> > +       depends on CC_AND_AS_SUPPORT_DWARF4
> >         help
> >           Generate DWARF v4 debug info. This requires gcc 4.5+, binutils 2.35.2
> >           if using clang without clang's integrated assembler, and gdb 7.0+.
> > @@ -277,7 +288,7 @@ config DEBUG_INFO_DWARF4
> >  config DEBUG_INFO_DWARF5
> >         bool "Generate DWARF Version 5 debuginfo"
> >         select DEBUG_INFO
> > -       depends on !CC_IS_CLANG || AS_IS_LLVM || (AS_IS_GNU && AS_VERSION >= 23502)
> > +       depends on CC_AND_AS_SUPPORT_DWARF5
> >         help
> >           Generate DWARF v5 debug info. Requires binutils 2.35.2, gcc 5.0+ (gcc
> >           5.0+ accepts the -gdwarf-5 flag but only had partial support for some
> 
> 
> 
> -- 
> Best Regards
> Masahiro Yamada

  reply	other threads:[~2022-10-03 20:32 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-02 18:11 [PATCH 0/3] Kconfig.debug: make more efforts to fix CONFIG_DEBUG_INFO for Clang+GAS Masahiro Yamada
2022-10-02 18:11 ` [PATCH 1/3] Kconfig.debug: simplify the dependency of DEBUG_INFO_DWARF4/5 Masahiro Yamada
2022-10-03 16:47   ` Nathan Chancellor
2022-10-03 16:53   ` Nick Desaulniers
2022-10-03 19:56     ` Masahiro Yamada
2022-10-02 18:11 ` [PATCH 2/3] Kconfig.debug: add toolchain checks for DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT Masahiro Yamada
2022-10-03 17:10   ` Nathan Chancellor
2022-10-03 20:18     ` Masahiro Yamada
2022-10-03 20:32       ` Nathan Chancellor [this message]
2022-10-02 18:11 ` [PATCH 3/3] Kconfig.debug: split debug-level and DWARF-version into separate choices Masahiro Yamada
2022-10-03 17:20   ` Nathan Chancellor
2022-10-03 20:58     ` 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=YztG0rb+cgf0IfPo@dev-arch.thelio-3990X \
    --to=nathan@kernel.org \
    --cc=linux-kbuild@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=llvm@lists.linux.dev \
    --cc=masahiroy@kernel.org \
    --cc=ndesaulniers@google.com \
    --cc=trix@redhat.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