linux-kernel - Re: [PATCH] Makefile: Improve compressed debug info support detection

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20200611214413.GA3924125@rani.riverdale.lan>
Date:   Thu, 11 Jun 2020 17:44:13 -0400
From:   Arvind Sankar <nivedita@...m.mit.edu>
To:     Nick Desaulniers <ndesaulniers@...gle.com>
Cc:     Arvind Sankar <nivedita@...m.mit.edu>,
        Masahiro Yamada <masahiroy@...nel.org>,
        Fangrui Song <maskray@...gle.com>,
        Rong Chen <rong.a.chen@...el.com>,
        kernel test robot <lkp@...el.com>, kbuild-all@...ts.01.org,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] Makefile: Improve compressed debug info support detection

On Thu, Jun 11, 2020 at 01:44:53PM -0700, Nick Desaulniers wrote:
> On Wed, Jun 10, 2020 at 4:39 PM Arvind Sankar <nivedita@...m.mit.edu> wrote:
> >
> > On Wed, Jun 10, 2020 at 07:30:46PM -0400, Arvind Sankar wrote:
> > > On Wed, Jun 10, 2020 at 02:27:55PM -0700, Nick Desaulniers wrote:
> > >
> > > No, as-option does invoke the assembler. The problem here is that with
> > > -Wa, the option is only seen by the assembler, not the gcc driver. So it
> > > will succeed because the assembler supports it, but it will not test
> > > whether the gcc driver also supports it.
> > >
> >
> > I think in theory another way to fix it is to just use -Wa even for
> > CFLAGS, assuming the compiler itself doesn't have to do anything with
> > the option and it's purely an assembler thing. Then you'd just do the as
> > and ld tests.
> 
> I don't think the CFLAGS are used for .S source files though.  We may
> drive assembler via the compiler, but AFLAGS are used in place of
> CFLAGS IIRC.  (eh, maybe not for ARCH=arc, ARCH=sh, ARCH=csky,
> ARCH=mips)
> 
> -- 
> Thanks,
> ~Nick Desaulniers

I meant just replace
	KBUILD_CFLAGS += -gz=zlib
with
	KBUILD_CFLAGS += -Wa,--compress-debug-sections=zlib
leaving KBUILD_AFLAGS the same as before
	KBUILD_AFLAGS += -Wa,--compress-debug-sections=zlib

This would avoid depending on the subtleties in the gcc driver's support
for this option. But -gz=zlib is passed to cc1 as well as the assembler,
so I'm not sure if there would be some loss from not passing it to the
compiler.