Messages in this thread | | | Date | Sat, 21 Oct 2000 19:19:06 -0400 | From | Chris Mason <> | Subject | Re: reproducable oops with 2.4.0-test9/reiserfs-3.6.17 |
| |
--On 10/21/00 13:08:26 -0700 "Brian J. Murrell" <9d0d5eb514ea07016f41967b8dc37026@interlinx.bc.ca> wrote:
[ BUG with highmem on reiserfs ] > > Unfortunately, that BUG might have been a red herring.
I wish ;-) That was a real bug, you just found a second one.
For highmem pages, set_bh_page() sets bh->b_data to (char *)0 + offset. Based on this, I don't think it is legal for highmem pages to be used here.
Guys, what am I missing? reiserfs is using GFP_BUFFER on kmallocs except for the bitmaps, and during journal replay (I'm fixing both of those spots). It looks like this BUG is coming from accessing bh->b_data from a page in the page cache.
-chris
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/
| |