From: Florian Weimer <fweimer@redhat.com>
To: Jan Kratochvil via Libc-alpha <libc-alpha@sourceware.org>
Cc: Jan Kratochvil <jkratochvil@azul.com>, Anton Kozlov <akozlov@azul.com>
Subject: Re: [PATCH] RFC: Provide a function to reset IFUNC PLTs
Date: Tue, 07 Mar 2023 09:40:46 +0100 [thread overview]
Message-ID: <87v8jdq7ht.fsf@oldenburg.str.redhat.com> (raw)
In-Reply-To: <DM6PR11MB40736D2E35582FCFDA780C5FC6B69@DM6PR11MB4073.namprd11.prod.outlook.com> (Jan Kratochvil via Libc-alpha's message of "Mon, 6 Mar 2023 09:04:34 +0100")
* Jan Kratochvil via Libc-alpha:
> Some projects snapshot+restore process images to migrate them to
> a different machine. The target machine may have different (particularly
> lower) set of CPU features. Restored process does crash in glibc IFUNC
> functions which have been already set in PLTs due to the former more
> rich CPU features on the snapshotting machine.
>
> Providing a glibc function which can be called during the restore.
>
> I understand the code may need more adjustments before its upstreaming
> but is this an acceptable approach?
Do you have a high-level overview how this is supposed to work? Is CRIU
expected to call the new _dl_reset_ifunc symbol after restore?
Use of <sys/platform/x86.h> is somewhat rare. Not even GCC uses it
AFAIK. It has its own cached CPU data used for target clones and
similar features. Re-running those IFUNC resolvers will just give the
same results.
Distributions have started to deploy alternate builds in
glibc-hwcaps/x86-64-v3 directories:
openSUSE Tumbleweed gains optional x86-64-v3 optimization
<https://news.opensuse.org/2023/03/02/tw-gains-optional-optimizations/>
This might be a common situation fairly soon. This is not IFUNC-based
at all, so any IFUNC-based approach is not going to help with that.
Practically speaking, I think cluster heterogeneity needs to be hidden
at the cluster level.
Thanks,
Florian
next prev parent reply other threads:[~2023-03-07 8:40 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-03-06 8:04 Jan Kratochvil
2023-03-07 8:40 ` Florian Weimer [this message]
2023-03-07 13:07 ` Adhemerval Zanella Netto
2023-03-08 10:21 ` Jan Kratochvil
2023-03-08 13:04 ` Adhemerval Zanella Netto
2023-03-09 11:32 ` Jan Kratochvil
2023-03-09 15:47 ` Adhemerval Zanella Netto
2023-03-09 17:43 ` Adhemerval Zanella Netto
2023-03-16 14:38 ` Jan Kratochvil
2023-03-20 16:47 ` Adhemerval Zanella Netto
2023-03-29 12:12 ` Jan Kratochvil
2023-03-29 13:14 ` Adhemerval Zanella Netto
2023-03-13 13:59 ` Florian Weimer
2023-03-14 12:55 ` Jan Kratochvil
2023-03-14 14:49 ` Florian Weimer
2023-03-14 15:06 ` Jan Kratochvil
2023-03-08 10:23 ` Jan Kratochvil
2023-03-08 10:44 ` Florian Weimer
2023-03-08 11:03 ` Jan Kratochvil
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=87v8jdq7ht.fsf@oldenburg.str.redhat.com \
--to=fweimer@redhat.com \
--cc=akozlov@azul.com \
--cc=jkratochvil@azul.com \
--cc=libc-alpha@sourceware.org \
/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).