linux-kernel - Re: [PATCH v2 1/3] perf report: Support LLVM for addr2line()

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] [day] [month] [year] [list]

Date: Sun, 19 May 2024 16:00:27 -0700
From: Ian Rogers <irogers@...gle.com>
To: "Steinar H. Gunderson" <sesse@...gle.com>
Cc: acme@...nel.org, linux-perf-users@...r.kernel.org, 
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 1/3] perf report: Support LLVM for addr2line()

On Sun, May 19, 2024 at 2:01 PM Steinar H. Gunderson <sesse@...gle.com> wrote:
>
> On Sun, May 19, 2024 at 11:23:26AM -0700, Ian Rogers wrote:
> > I think we might want to display this in the feature list during the build:
> > https://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools-next.git/tree/tools/build/Makefile.feature?h=perf-tools-next#n123
>
> Ack.
>
> > s/HAVE_LIBLLVM_SUPPORT/HAVE_LLVM_SUPPORT/
>
> Ack.
>
> >> +#pragma GCC diagnostic push
> >> +#pragma GCC diagnostic ignored "-Wunused-parameter"  /* Needed for LLVM 14. */
> > nit: pehaps disabling this should be conditional:
> > #if LLVM_VERSION_MAJOR == 14
>
> It doesn't seem LLVM_VERSION_MAJOR is defined yet at this point, so I
> don't think that would work. (I just checked; 15 has the same issue.)
> In any case, I think it would be just more clutter for dubious gain.
>
> >> +       if (file) {
> >> +               if (line_info.FileName == "<invalid>") {
> >> +                       /* Match the convention of libbfd. */
> >> +                       *file = nullptr;
> > Should "*file" be cleared if "!file" so the caller can reliably free it?
>
> I don't understand. If “!file”, then file == NULL and surely accessing
> *file would mean a crash?

You're right, sorry. Thanks,
Ian

> /* Steinar */