> currently experiencing a heavy performance regression in 2.6.26. Until > 2.6.26, my external firewire disk (Maxtor OneTouch 4 Plus / 1TB) was > reading with 38MB/s and writing with 26MB/s. When I do writing tests wi= th > dd from /dev/zero to disk, I get 26MBs but when I read with KDE, hdparm= , > dd or cp I'm stuck at 24MB/s. This speed is very stable, only varies > 0.5MB/s among successive tests. >=20 > I also upgraded my BIOS to latest version after the upgrade but I think= > this is not the cause of the regression because I also have a SoundBlas= ter > audigy2ZS and it has a firewire port too and using the port on it rathe= r > than the motherboard's one also doesn't changes the speed in a reasonab= le > manner. >=20 > Is anybody other than me having the same problem or have some ideas? >=20 > Thanks in advance, >=20 > Hakan >=20 > My basic system configuration is as follows: > CPU: Intel Core2Quad Q6600 > RAM: 4GB OCz Flex XLC CL3 High performance DDR2 > M/B: MSI P35 Platinum / BIOS 1.9 >=20 >=20 Hi all, I've filed a bug report about the performance regression and a message ex= planing the current situation arrived shortly. According to message and the linked wikipedia page, the new firewire stac= k that enabled high performance (via gap count optimization) is not currently supporting all features of the old stack hence, lenny will be shipped wit= h the old stack. Since lenny is nearing freeze and will be released, packagers started to = pack the old stack, not the new to support more hardware and operations over f= irewire. Bug report and the reply can be seen at: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=3D497713 Cheers, Hakan
Attachment:
signature.asc
Description: OpenPGP digital signature