Re: mm: Unable to handle kernel NULL pointer dereference at virtual address - mmap_region (include/linux/fs.h:580 mm/mmap.c:2946)

[Dan Carpenter]

On Mon, Oct 16, 2023 at 05:32:00PM +0100, Lorenzo Stoakes wrote:
> On Mon, Oct 16, 2023 at 12:05:37PM +0100, Lorenzo Stoakes wrote:
> > On Mon, Oct 16, 2023 at 03:52:07PM +0530, Naresh Kamboju wrote:
> > > Following kernel crash noticed while running LTP hugetlb and selftests on
> > > qemu-x86_64 and qemu-arm64 running with Linux next 6.6.0-rc6-next-20231016.
> > >
> > > Reported-by: Linux Kernel Functional Testing <lkft@xxxxxxxxxx>
> > > Reported-by: Naresh Kamboju <naresh.kamboju@xxxxxxxxxx>
> > >
> > > Test Logs:
> > > -----
> >
> > [snip]
> >
> > > <4>[   97.499871] Call trace:
> > > <4>[ 97.500288] mmap_region (include/linux/fs.h:580 mm/mmap.c:2946)
> >
> > OK this is from a patch of mine, and an easy fix (incorrect assumption about
> > vm->vm_file == file).
> >
> > I will put a fix forward tonight.
> >
> > > <4>[ 97.500814] do_mmap (mm/mmap.c:1379)
> > > <4>[ 97.501243] vm_mmap_pgoff (mm/util.c:546)
> > > <4>[ 97.501711] ksys_mmap_pgoff (mm/mmap.c:1425)
> > > <4>[ 97.502166] __arm64_sys_mmap (arch/arm64/kernel/sys.c:21)
> > > <4>[ 97.502634] invoke_syscall (arch/arm64/include/asm/current.h:19
> > > arch/arm64/kernel/syscall.c:56)
> > > <4>[ 97.503175] el0_svc_common.constprop.0
> > > (include/linux/thread_info.h:127 (discriminator 2)
> > > arch/arm64/kernel/syscall.c:144 (discriminator 2))
> > > <4>[ 97.503763] do_el0_svc (arch/arm64/kernel/syscall.c:156)
> > > <4>[ 97.504191] el0_svc (arch/arm64/include/asm/daifflags.h:28
> > > arch/arm64/kernel/entry-common.c:133
> > > arch/arm64/kernel/entry-common.c:144
> > > arch/arm64/kernel/entry-common.c:679)
> >
> > [snip]
> 
> Have cc-d people in this thread on it, but for the record, -fix patch is at
> https://lore.kernel.org/all/c9eb4cc6-7db4-4c2b-838d-43a0b319a4f0@lucifer.local/

Smatch also caught this bug.  Your patch silences the warning.

mm/mmap.c:2946 mmap_region() error: we previously assumed 'file' could be null (see line 2849)

It's amazing that Naresh was able to hit this after it had only been in
linux-next for less than a day.

regards,
dan carpenter