[lkml]   [2010]   [Oct]   [28]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [next-20101038] Call trace in ext4
On Thu, Oct 28, 2010 at 09:54:23PM +0200, Sedat Dilek wrote:
> Hm, unfortunately NO (see logs).
> I have compiled via M=fs/ext4 in an already compiled build-tree with
> these 3 patches.

Ok, stupid question. You did make sure the new ext4 module was
loaded, right?

> [ 100.884524] ------------[ cut here ]------------
> [ 100.884718] kernel BUG at /home/sd/src/linux-2.6/linux-2.6.36/debian/build/source_i386_none/fs/ext4/inode.c:2721!

OK, so after the patch, line 2721: changed from page_buffers() to:

if (!page_has_buffers(page))
page_has_buffers() expands to:

#define page_has_buffers(page) PagePrivate(page)

which expands to test_bit() call to see if PG_private1 is set in
page->flags. There is no BUG_ON anywhere there as far as I can tell.

Line 2721 in the older kernel was page_has_buffers() which does have a
BUG_ON check.

- Ted

 \ /
  Last update: 2010-10-28 22:09    [from the cache]
©2003-2011 Jasper Spaans