linux-kernel - Re: [PATCH v1] perf symbol: Add rust v0 demangling support

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]

Message-ID: <Z6PSyvPJ09b3pfPE@google.com>
Date: Wed, 5 Feb 2025 13:06:18 -0800
From: Namhyung Kim <namhyung@...nel.org>
To: Ian Rogers <irogers@...gle.com>
Cc: Daniel Xu <dxu@...uu.xyz>, Peter Zijlstra <peterz@...radead.org>,
	Ingo Molnar <mingo@...hat.com>,
	Arnaldo Carvalho de Melo <acme@...nel.org>,
	Mark Rutland <mark.rutland@....com>,
	Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
	Jiri Olsa <jolsa@...nel.org>,
	Adrian Hunter <adrian.hunter@...el.com>,
	Kan Liang <kan.liang@...ux.intel.com>,
	Miguel Ojeda <ojeda@...nel.org>,
	Alex Gaynor <alex.gaynor@...il.com>,
	Boqun Feng <boqun.feng@...il.com>, Gary Guo <gary@...yguo.net>,
	Björn Roy Baron <bjorn3_gh@...tonmail.com>,
	Benno Lossin <benno.lossin@...ton.me>,
	Andreas Hindborg <a.hindborg@...nel.org>,
	Alice Ryhl <aliceryhl@...gle.com>, Trevor Gross <tmgross@...ch.edu>,
	James Clark <james.clark@...aro.org>,
	Howard Chu <howardchu95@...il.com>,
	Ravi Bangoria <ravi.bangoria@....com>,
	Masami Hiramatsu <mhiramat@...nel.org>,
	linux-kernel@...r.kernel.org, linux-perf-users@...r.kernel.org,
	rust-for-linux@...r.kernel.org
Subject: Re: [PATCH v1] perf symbol: Add rust v0 demangling support

Hello,

On Mon, Feb 03, 2025 at 09:20:30AM -0800, Ian Rogers wrote:
> On Thu, Jan 30, 2025 at 11:35 AM Daniel Xu <dxu@...uu.xyz> wrote:
> >
> > Hi Ian,
> >
> > Thanks so much for working on this!
> >
> > Only did a quick skim. Seems like it'd be nice to reuse the official
> > demangler now that we know about it. Will be interesting to
> > see how that gets packaged for distros.
> 
> I think the best solution is to use the official demangler but in
> general we'd like fewer rather than more library dependencies. For
> example your initial comments were to work to make the coupling with
> capstone and LLVM disassemblers more loose via dlopen:
> https://lore.kernel.org/lkml/20250122174308.350350-1-irogers@google.com/
> 
> Thinking through the options (note I'm not doing Rust things and so my
> level of skin in the game is minimal):
> 1) stay as we are - well this is a mess as there is no v0 support and
> the legacy support comes from libbfd/libiberty.
> 2) option (1) plus this change cleaned up - there's still the mess
> with the legacy support, which is probably more important given v0
> isn't widely adopted.
> 3) wait for the official demangler to be packaged as a library and
> depend upon it, add testing from here - solves legacy and v0 issues,
> but brings in an extra library dependency to the build. Timeline not
> clear.
> 4) "contribute"/copy-paste the official demangler into perf tree and
> do like (3). Perhaps remove the forked file when a dependable library
> exists (there's some precedent for this with how libtraceevent was in
> and then removed from the linux tree) - solves the timeline, not great
> to fork the official demangler, so far not heard from the author Ariel
> Ben-Yehuda and I would strongly prefer they contribute rather than I
> copy-paste it.

Agreed.  I'm hesitant to add this change to perf code base without
commitment of the Rust community.

Thanks,
Namhyung

> 
> For (3) and (4) it is probably a good idea to engage the Rust folks in:
> https://github.com/rust-lang/rust/issues/60705
> 
> Having broken Rust demangling seems like a high priority issue for the
> perf tool.