[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAHk-=wjyp3Y_vXJwvoieBJpmmTrs46kc4GKbq5x_nvonHvPJBw@mail.gmail.com>
Date: Fri, 23 Oct 2020 10:50:55 -0700
From: Linus Torvalds <torvalds@...ux-foundation.org>
To: Naresh Kamboju <naresh.kamboju@...aro.org>
Cc: Daniel Díaz <daniel.diaz@...aro.org>,
Stephen Rothwell <sfr@...b.auug.org.au>,
"Matthew Wilcox (Oracle)" <willy@...radead.org>,
"Peter Zijlstra (Intel)" <peterz@...radead.org>,
Viresh Kumar <viresh.kumar@...aro.org>,
X86 ML <x86@...nel.org>,
open list <linux-kernel@...r.kernel.org>,
lkft-triage@...ts.linaro.org,
"Eric W. Biederman" <ebiederm@...ssion.com>,
linux-mm <linux-mm@...ck.org>,
linux-m68k <linux-m68k@...ts.linux-m68k.org>,
Linux-Next Mailing List <linux-next@...r.kernel.org>,
Thomas Gleixner <tglx@...utronix.de>,
kasan-dev <kasan-dev@...glegroups.com>,
Dmitry Vyukov <dvyukov@...gle.com>,
Geert Uytterhoeven <geert@...ux-m68k.org>,
Christian Brauner <christian.brauner@...ntu.com>,
Ingo Molnar <mingo@...hat.com>, LTP List <ltp@...ts.linux.it>,
Al Viro <viro@...iv.linux.org.uk>
Subject: Re: [LTP] mmstress[1309]: segfault at 7f3d71a36ee8 ip
00007f3d77132bdf sp 00007f3d71a36ee8 error 4 in libc-2.27.so[7f3d77058000+1aa000]
On Fri, Oct 23, 2020 at 10:00 AM Naresh Kamboju
<naresh.kamboju@...aro.org> wrote:
>
> [Old patch from yesterday]
>
> After applying your patch on top on linux next tag 20201015
> there are two observations,
> 1) i386 build failed. please find build error build
Yes, this was expected. That patch explicitly only works on x86-64,
because 32-bit needs the double register handling for 64-bit values
(mainly loff_t).
> 2) x86_64 kasan test PASS and the reported error not found.
Ok, good. That confirms that the problem you reported is indeed the
register allocation.
The patch I sent an hour ago (the one based on Rasmus' one from
yesterday) should fix things too, and - unlike yesterday's - work on
32-bit.
But I'll wait for confirmation (and hopefully a sign-off from Rasmus
so that I can give him authorship) before actually committing it.
Linus
Powered by blists - more mailing lists