[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <35a07d38d6d59b79d28543bb4a862f14af1fa713.camel@physik.fu-berlin.de>
Date: Sun, 07 Sep 2025 19:57:37 +0200
From: John Paul Adrian Glaubitz <glaubitz@...sik.fu-berlin.de>
To: René Rebe <rene@...ctco.de>
Cc: Michael Karcher <kernel@...rcher.dialup.fu-berlin.de>, Andreas Larsson
<andreas@...sler.com>, sparclinux@...r.kernel.org,
linux-kernel@...r.kernel.org, Anthony Yznaga <anthony.yznaga@...cle.com>
Subject: Re: [PATCH v4 2/5] sparc: fix accurate exception reporting in
copy_{from_to}_user for UltraSPARC III
Hi Rene,
On Sun, 2025-09-07 at 19:32 +0200, René Rebe wrote:
> It is probably a good time to mention that there are likely some other major
> TLB (or so) bug on U3. For example, I could never boot any Linux kernel
> (probably ever) with 8GB installed in my Sun Blade 1000 - it would NULL ptr
> deref very early:
Have a look at arch/sparc/kernel/head_64.S:
/* Patch copy/page operations to cheetah optimized versions. */
call cheetah_patch_copyops
nop
call cheetah_patch_copy_page
nop
call cheetah_patch_cachetlbops
nop
These patch in UltraSPARC-III-optimized versions of copy_{to,from}_user
and copy_page and TLB operations. Replacing these calls with "nop" might
tell us whether any of those is broken.
I have already replaced cheetah_patch_copyops with generic_patch_copyops
and was able to verify that the recently discovered OOPS is not related
to copy_{to,from}_user, so Michael's UltraSPARC III may still be correct
and complete.
I am testing the other two now as I think that these are good candidates.
Adrian
--
.''`. John Paul Adrian Glaubitz
: :' : Debian Developer
`. `' Physicist
`- GPG: 62FF 8A75 84E0 2956 9546 0006 7426 3B37 F5B5 F913
Powered by blists - more mailing lists