From: Richard Weinberger <richard@nod.at>
To: Jeff Dike <jdike@addtoit.com>, Eli Cooper <elicooper@gmx.com>
Cc: user-mode-linux-devel@lists.sourceforge.net
Subject: Re: [uml-devel] [PATCH] um: fix FPU register double-restore after sigreturn
Date: Fri, 18 Mar 2016 21:12:07 +0100 [thread overview]
Message-ID: <56EC6117.90806@nod.at> (raw)
In-Reply-To: <20160318164243.GA979@parts.user-mode-linux.org>
Jeff, Eli,
Am 18.03.2016 um 17:42 schrieb Jeff Dike:
> On Sat, Mar 19, 2016 at 12:13:47AM +0800, Eli Cooper wrote:
>> I agree that we should keep userspace() minimal. So what in the first
>> place is the FPU state save/restore code for? Or, what (except for
>> sigreturn) could possibly mess the FPU state of the ptrace'd process
>> without it?
>
> My (vague) recollection was that with libc in there, I couldn't be
> confident that it wouldn't unexpectedly use FP for something. So,
> saving and restoring FP state was a hedge against that happening.
>
> Jeff
>
git blame points to:
commit fbfe9c847edf57ac8232aeafb290f272289893a3
Author: Ingo van Lil <inguin@gmx.de>
Date: Wed Sep 14 16:21:23 2011 -0700
um: Save FPU registers between task switches
Some time ago Jeff prepared 42daba316557 ("uml: stop saving process FP
state") for UML to stop saving the process FP state between task
switches. The assumption was that since with SKAS0 every guest process
runs inside a host process context the host OS will take care of keeping
the proper FP state.
Unfortunately this is not true for multi-threaded applications, where
all guest threads share a single host process context yet all may use
the FPU on their own. Although I haven't verified it I suspect things
to be even worse in SKAS3 mode where all guest processes run inside a
single host process.
The patch reintroduces the saving and restoring of the FP context
between task switches.
[richard@nod.at: Ingo posted this patch in 2009, sadly it was never applied
and got lost. Now in 2011 the problem was reported by Gunnar.]
Signed-off-by: Ingo van Lil <inguin@gmx.de>
Signed-off-by: Richard Weinberger <richard@nod.at>
Reported-by: <gunnarlindroth@hotmail.com>
Tested-by: <gunnarlindroth@hotmail.com>
Cc: Stanislav Meduna <stano@meduna.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
So, it is not that easy. :=)
Thanks,
//richard
------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785231&iu=/4140
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel
prev parent reply other threads:[~2016-03-18 20:12 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-03-16 1:25 [uml-devel] [PATCH] um: fix FPU register double-restore after sigreturn Eli Cooper
2016-03-17 22:21 ` Richard Weinberger
2016-03-18 1:41 ` Eli Cooper
2016-03-18 8:20 ` Richard Weinberger
2016-03-18 16:13 ` Eli Cooper
2016-03-18 16:42 ` Jeff Dike
2016-03-18 20:12 ` Richard Weinberger [this message]
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=56EC6117.90806@nod.at \
--to=richard@nod.at \
--cc=elicooper@gmx.com \
--cc=jdike@addtoit.com \
--cc=user-mode-linux-devel@lists.sourceforge.net \
/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 NNTP newsgroup(s).