o reverse some vmscan 2.1.130 patch
(s/free_page/free_page_and_swap_cache). This rejecting plus my own
mm changes give me a _perfect_ mm tuning (at least everything
is fine here so far). Here an example:
andrea@dragon:~$ find /usr/src/linux/ -type f -exec grep -l sdfsdfa {} \;
&
[1] 695
andrea@dragon:~$ cp /dev/zero /tmp/p
cp: /tmp/p: No space left on device
andrea@dragon:~$ rm /tmp/p
andrea@dragon:~$ free
total used free shared buffers cached
Mem: 63564 62040 1524 36252 3276 35756
-/+ buffers/cache: 23008 40556
Swap: 52412 0 52412
I was also just running emacs gnome and X without any problem in
the meantime.
o very very experimental scsi code. This will change _only_ the
resetting case. I have not yet tested it... I should hack
ppa to generate a scsi reset... If somebody has a machine that
generate a scsi reset frequently it would be nice if he could try
my code. The code should not generate fs corruption at all but
could hang the machine if I missed something... I have not hear
comments about the snapshot from scsi maintainers...
o Little s/time_before/time_after/ in the scsi code (patch sent to
Linus). Also this influence only the scsi reset path.
Andrea Arcangeli
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/