From: Luis Machado <luis.machado@arm.com>
To: Joel Brobecker <brobecker@adacore.com>,
Torbjorn SVENSSON <torbjorn.svensson@foss.st.com>,
simon.marchi@polymtl.ca, tom@tromey.com
Cc: gdb-patches@sourceware.org
Subject: Re: GDB 13 release -- 2023-01-21 Update
Date: Fri, 27 Jan 2023 06:38:52 +0000 [thread overview]
Message-ID: <47dff98e-b714-2d72-bc3b-c14c4e0313d0@arm.com> (raw)
In-Reply-To: <Y9NvgBUvtCBfp0Qi@adacore.com>
On 1/27/23 06:30, Joel Brobecker wrote:
> Hello,
>
>>> * [Torbjorn] tdep/29738
>>> Arm M-profile dwarf2 unwinder performance suffers from exponential growth
>>> https://sourceware.org/bugzilla/show_bug.cgi?id=29738
>>>
>>> patch v3, 2023-01-19, reviewed 2023-01-20:
>>> https://sourceware.org/pipermail/gdb-patches/2023-January/195915.html
>>
>> I just pushed this for master.
>> Is it okay to also push the 2 patches to gdb-13-branch?
>
> For the avoidance of doubt, my understanding is that we are talking
> about the following two patches:
>
> commit d72ba177c85f2ad18d0dcabdd8844532c9acb819
> Author: Torbj�rn SVENSSON <torbjorn.svensson@foss.st.com>
> Date: Thu Nov 17 12:17:53 2022 +0100
> Subject: gdb: dwarf2 generic implementation for caching function data
>
> ... and ...
>
> commit 5cf11483141a58314834653003e49709b47822d5
> Author: Torbj�rn SVENSSON <torbjorn.svensson@foss.st.com>
> Date: Thu Nov 17 12:18:20 2022 +0100
> Subject: gdb/arm: Use new dwarf2 function cache
>
> I hope I having missed any other patch!
>
> The first one adds, as the subject indicates, a framework for
> caching frame-related information, and the second patch takes
> advantage of that framework,
>
> Luis marked the corresponding PR as "important to fix", so
> I'm assuming the impact if we do not backport is significant
> (exponentional performance degradation). I was confused into
> thinking that this would "only" impact Cortex-m without security
> extensions, but maybe it's the opposite actually.
It affects m-profiles, but it affects the targets with the security extensions more, as they have more SP registers.
>
> So the next question is, what is the potential impact if we
> backport the patch and there is a bug in it:
>
> - Well, it touches the generic Arm unwinding code, so
> worse case scenario, DWARF-based unwinding is broken?
The change is restricted to m-profiles. Unless there is something really off with the caching code, it should break all Arm unwinding.
>
> - With that said, the patch appears to simply add a cache,
> so the logic of it all doesn't appear to be extremely
> complicated. So I would rate the risk to be low.
That's what it seems to me.
I think another round of testing would be a good step to make sure there are no hidden bugs.
>
> If Tom and/or Simon agree, my assessment is that it is fine
> to backport those two patches onto the gdb-13 branch.
>
next prev parent reply other threads:[~2023-01-27 6:39 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-01-21 6:08 Joel Brobecker
2023-01-25 20:18 ` Torbjorn SVENSSON
2023-01-27 6:30 ` Joel Brobecker
2023-01-27 6:38 ` Luis Machado [this message]
2023-01-27 17:17 ` Simon Marchi
2023-01-28 8:39 ` Luis Machado
2023-01-29 11:52 ` Joel Brobecker
2023-01-30 11:23 ` Luis Machado
2023-01-31 7:03 ` Joel Brobecker
2023-01-31 13:57 ` Luis Machado
2023-01-31 14:33 ` Luis Machado
2023-02-02 3:44 ` Joel Brobecker
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=47dff98e-b714-2d72-bc3b-c14c4e0313d0@arm.com \
--to=luis.machado@arm.com \
--cc=brobecker@adacore.com \
--cc=gdb-patches@sourceware.org \
--cc=simon.marchi@polymtl.ca \
--cc=tom@tromey.com \
--cc=torbjorn.svensson@foss.st.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).