From: Mahmoud Adam <mngyadam@amazon.com>
To: <peterz@infradead.org>
Cc: <tglx@linutronix.de>, <bp@alien8.de>,
<gregkh@linuxfoundation.org>, <regressions@lists.linux.dev>
Subject: Regression caused by 'x86/returnthunk: Allow different return thunks'
Date: Tue, 12 Mar 2024 20:58:39 +0100 [thread overview]
Message-ID: <lrkyq5xxrz0gg.fsf@dev-dsk-mngyadam-1c-a2602c62.eu-west-1.amazon.com> (raw)
Hello,
We've seen ~35% regression on fio benchamrk on AMD caused by these
commits:
`
545a94ffc29a4 * x86/alternative: Make custom return thunk unconditional
b012dcf39d9e7 * Revert "x86/alternative: Make custom return thunk unconditional"
1221b8ea25cc8 * x86/returnthunk: Allow different return thunks
`
the finding and bisection was done on v6.1, but the regression is
already seen on the new rebases of other stable kernel 5.10 & 5.15 that
have the same patches.
the following exact fio command that sees this regression:
fio --time_based --name=benchmark --size=50G --runtime=60
--filename=$device --ioengine=psync --randrepeat=0 --iodepth=1
--fsync=64 --invalidate=1 --verify=0 --verify_fatal=0 --blocksize=4k
--group_reporting --rw=randwrite --numjobs=1
note: the regression is seen even with retbleed=off and also with spectre_v2=off
-MNAdam
next reply other threads:[~2024-03-12 19:58 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-03-12 19:58 Mahmoud Adam [this message]
2024-03-12 20:02 ` Regression caused by 'x86/returnthunk: Allow different return thunks' Borislav Petkov
2024-03-12 21:00 ` Mahmoud Adam
2024-03-12 21:04 ` Borislav Petkov
2024-03-12 21:33 ` Mahmoud Adam
2024-03-14 14:38 ` Borislav Petkov
2024-03-14 16:43 ` Mahmoud Adam
2024-03-27 17:58 ` Borislav Petkov
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=lrkyq5xxrz0gg.fsf@dev-dsk-mngyadam-1c-a2602c62.eu-west-1.amazon.com \
--to=mngyadam@amazon.com \
--cc=bp@alien8.de \
--cc=gregkh@linuxfoundation.org \
--cc=peterz@infradead.org \
--cc=regressions@lists.linux.dev \
--cc=tglx@linutronix.de \
/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).