Re: [PATCH] Fix oops in NFS

From: Dominique Martinet
Date: Fri Jan 05 2024 - 12:21:19 EST


Matthew Wilcox wrote on Fri, Jan 05, 2024 at 01:17:36PM +0000:
> host on /host type 9p (rw,relatime,access=client,trans=virtio)

David Howells wrote on Fri, Jan 05, 2024 at 02:33:32PM +0000:
> > This commit (100ccd18bb41 in linux-next 20240104) is bad for me. After
> > it, running xfstests gives me first a bunch of errors along these lines:
>
> This may be related to a patch that is in linux-next 20240105, but not
> 20240104 ("9p: Fix initialisation of netfs_inode for 9p").

Yes, you'd be reading zeroes without that patch because the netfs code
thinks the file has 0 size and doesn't bother reading, that'd explain
the exec format error loading other modules...

One thing that surprised me is that this also affects cache=none, I
thought we had different file ops going straight to p9_client_read in
this case?
But turning my brain on this would be the read-only mmap case that we
need to support for execs, which module loading also uses, so this came
biting there alright.

--
Dominique Martinet | Asmadeus