On Mon, 11 Aug 2003 09:21:58 -0400 (EDT) tallison@tacocat.net wrote: > In my case it was shitty EIDE code in the kernel screwing up my VIA > chipset (I think that's the name) (which is also shitty or even > shittier I find out) and causing massive problems with data > corruptions and really poor performance (7MB/s with a tail wind). > > downgrade the kernel to 2.4.18 or lower and this problem may go away. > > Alternatively, at least in theory, upgrade to 2.6.0-testX. > > But the entire 2.4.2x series has been horribly problematic for these > motherboards and has been a real dissappointment. Kernel 2.4.21 with a VIA VT8366: jkoenig@note:~$ sudo hdparm -Tt /dev/hda Password: /dev/hda: Timing buffer-cache reads: 268 MB in 2.02 seconds = 132.67 MB/sec Timing buffered disk reads: 100 MB in 3.02 seconds = 33.11 MB/sec jkoenig@note:~$ It's a little low right now, I've gotten close to 200 buffer-cache and 45 buffered disk (I have a lot of disk activity right now) hdparm was not used to tweak the drive. My .config: CONFIG_IDEDMA_PCI_AUTO=y CONFIG_BLK_DEV_IDEDMA=y CONFIG_BLK_DEV_VIA82CXXX=y -- -johann koenig Now Playing: Rancid - Rats In The Hallway : Punk-O-Rama (Vol. 3) Today is Pungenday, the 4th day of Bureaucracy in the YOLD 3169 My public pgp key: http://mental-graffiti.com/pgp/johannkoenig.pgp
Attachment:
pgpRSiSB2c8Zx.pgp
Description: PGP signature