Hello, On Nov 16, 2008, at 2:34 PM, Joerg Sonnenberger wrote:
On Sun, Nov 16, 2008 at 12:40:37PM -0500, Michael wrote:I just found out that the transaction size was negative.Misassumptions on my part. It seems the vfs_bio.c change was the culprit. WAPBL is trying to trace the size of buffers and that the badoptimisation failed when the buffer was changed slightly, but still hadthe same buffer cache size. Can you try again after an update please?
Seems to work again, thanks! Any idea why (apparently) it bit only on LP64? have fun Michael