Re: OOPS: 2.6.24-rc5-mm1 -- EIP is at r_show+0x2a/0x70 -- (triggered by "cat /proc/iomem")

[Miles Lane]

On Dec 20, 2007 6:37 AM, David Howells <dhowells@xxxxxxxxxx> wrote:
> Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote:
>
> > I would be suspecting iget-stop-procfs-from-using-iget-and-read_inode.patch.
>
> I think your suspicions are very unlikely.  The patch only affects
> proc_get_inode() - and looking at the patch backtrace, it looks like the
> system is successfully past that already (it's unlikely that
> proc_reg_read+0x60/0x74 would have been reached otherwise).
>
> If my patch to procfs is wrong, it would affect all proc files and ought be
> immediately detectable.

I tested the patch Andrew sent.  I ran "cat /proc/iomem" before trying
a suspend-to-disk.  It worked fine.  Then I suspended and resumed.
This time, "cat /proc/iomem" caused the stack trace to be generated.
So, you are right, you patch is not the problem.

Thanks,
              Miles
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/