ATA FLUSHCACHE

[Ollivier Robert]

According to Ollivier Robert:
> Allowing retries on ATA_FLUSHCACHE operations (by default they are not --
> see ata-disk.c around line 280) gets me further but it still panics.

Trying it differently by putting a larger timeout (the same as above in
ata-disk.c that is, 5s instead of 1s), made the system survive an hg clone
of the src HEAD source and seems to survive a clone of /usr/ports.

Don't know yet what Søren think of that but I'll keep it in my tree.
-- 
Ollivier ROBERT -=- FreeBSD: The Power to Serve! -=- roberto at keltia.freenix.fr
Darwin sidhe.keltia.net Kernel Version 8.8.2: Thu Sep 28 20:43:26 PDT 2006 i386