Skip to contentSkip to conversation

2.5.32-mm1

DISCUSSION [email protected] 1 participant 1 message Aug 28, 2002

lore
Share & tools
lore
#0 lore raw

2.5.32-mm1 fixes some page allocation failures on 3.75 GB quad xeon. 2.5.32 vanilla had a lot of these:

page allocation failure. order:0, mode:0x50

The page alloc failures began showing up in the first few minutes of testing (dbench). 2.5.32 stopped responding after about 4 hours of testing.

2.5.32-mm1 has been testing for over 14 hours with no page allocation failures.

2.5.32 didn't have page alloc failures on uniprocessor 384 MB machine during 22 hours of testing.

Signature · 2 lines
Randy Hron
http://home.earthlink.net/~rwhron/kernel/bigbox.html