On Thu, Jul 21, 2011 at 11:55:04PM +0300, T?r?k Edwin wrote: > Not sure what the BUG means: did XFS ran out of memory when trying to write the 4GB coredump, as it wanted to allocate an extra 4G temporarely to hold it? > It means we dereferenced an invalid pointer in xfs_is_delayed_page, which is called from the writeback code, which is pretty odd. We haven't really touched that particular piece of code for a long time, and it really doesn't derefence many pointers, basically just page->mapping, page->private and the flags and b_this_page fields in the buffer head. _______________________________________________ xfs mailing list xfs@xxxxxxxxxxx http://oss.sgi.com/mailman/listinfo/xfs
