linux-kernel - Re: [PATCH v7 4/4] perf report: LLVM-based symbol listing

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: <ZlOomG5NacWlV_RY@google.com>
Date: Sun, 26 May 2024 23:24:40 +0200
From: "Steinar H. Gunderson" <sesse@...gle.com>
To: acme@...nel.org
Cc: linux-perf-users@...r.kernel.org, linux-kernel@...r.kernel.org,
	irogers@...gle.com
Subject: Re: [PATCH v7 4/4] perf report: LLVM-based symbol listing

On Sun, May 26, 2024 at 08:22:12PM +0200, Steinar H. Gunderson wrote:
> +
> +	symbols__fixup_end(dso__symbols(dso), false);
> +	symbols__fixup_duplicate(dso__symbols(dso));
> +	dso__set_adjust_symbols(dso, true);
> +	ret = 0;

Urgh, evidently I forgot to copy over dso->text_offset and dso->text_end
here, which I guess needs to happen.

Even after doing that, I still cannot annotate Windows symbols, though;
I get

  Couldn't annotate RtlLeaveCriticalSection:
  Internal error: Invalid -1 error code

This is true even with libbfd in a non-distro build (without any of my
LLVM patches), so there must be something else at play here.

I _can_ run a Windows binary in WINE and get the right symbols out with
perf report, though. So this leaves only demangling as the final piece
of the puzzle to make distro builds essentially equal to nondistro builds.

/* Steinar */