[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAADnVQJLhKPA1MGtpqDwAnSDrCXBd1dxoCor2RwD22a-VLpiKA@mail.gmail.com>
Date: Fri, 8 Mar 2024 07:49:14 -0800
From: Alexei Starovoitov <alexei.starovoitov@...il.com>
To: Naresh Kamboju <naresh.kamboju@...aro.org>, Christoph Hellwig <hch@...radead.org>,
Linus Torvalds <torvalds@...ux-foundation.org>
Cc: Arnd Bergmann <arnd@...db.de>, linux-mm <linux-mm@...ck.org>,
linux-next <linux-next@...r.kernel.org>, open list <linux-kernel@...r.kernel.org>,
Linux Regressions <regressions@...ts.linux.dev>, lkft-triage@...ts.linaro.org,
Andrew Morton <akpm@...ux-foundation.org>, Ard Biesheuvel <ardb@...nel.org>,
Alexei Starovoitov <ast@...nel.org>, Andrii Nakryiko <andrii@...nel.org>,
Dan Carpenter <dan.carpenter@...aro.org>
Subject: Re: arm64: WARNING: at mm/vmalloc.c:315 ioremap_page_range
On Fri, Mar 8, 2024 at 6:24 AM Naresh Kamboju <naresh.kamboju@...aro.org> wrote:
>
> On Fri, 8 Mar 2024 at 17:34, Arnd Bergmann <arnd@...db.de> wrote:
> >
> > On Fri, Mar 8, 2024, at 12:53, Naresh Kamboju wrote:
> > > The following warning was noticed while boot arm and arm64 devices.
> > > The below log is extracted from arm64 ROCK Pi 4B while booting and
> > > another issue is kernel oops noticed while testing selftests: kvm:
> > > memslot_perf_test.
> > >
> > > 1) WARNING: CPU: 1 PID: 201 at mm/vmalloc.c:315 ioremap_page_range
> > > 2) Unable to handle kernel NULL pointer dereference at virtual address
> > > Internal error: Oops: 0000000096000006 [#1] PREEMPT SMP
> > > Kernel panic - not syncing: Oops: Fatal exception in interrupt
> > >
> > > Please find steps to reproduce, logs and config file links below.
> > >
> > > Reported-by: Linux Kernel Functional Testing <lkft@...aro.org>
>
> <trim>
>
> > The warning was added by commit 3e49a866c9dc ("mm: Enforce
> > VM_IOREMAP flag and range in ioremap_page_range.").
> >
> > This indeed uses a special memory area at a fixed physical address,
> > but I don't see where the vm_struct comes from.
>
> Thanks for your explanation.
>
> > No idea here, but it seems unrelated to the first one.
>
> I will reproduce this issue a few more times and get back to you.
See the proposed fix:
https://lore.kernel.org/bpf/CAADnVQLP=dxBb+RiMGXoaCEuRrbK387J6B+pfzWKF_F=aRgCPQ@mail.gmail.com/
but Christoph believes a different approach is necessary:
https://lore.kernel.org/bpf/Zeso7eNj1sBhH5zY@infradead.org/
Powered by blists - more mailing lists