Paul Barton-Davis wrote:
>
> >I've refreshed the patch for 2.4.0-test8 and put that up at
> >http://www.uow.edu.au/~andrewm/linux/schedlat.html#downloads
> >
> >Some of the recent changes to ext2 truncation have broken this patch, so
> >there will probably again be scheduling stalls with deleting large
> >files. I'll take a look at what they've done and mend it later this
> >week.
>
> quick latency patch report: with a "real" load (running a multitrack
> hard disk recorder system), 2.4.0-test7(SMP,all SCSI) + your patch
> does not prevent audio dropouts. I still get system stalls on the
> order of 30msec from time to time. I am running the recorder more
> these days, and will try to figure out more about this as it happens.
mm.. There was a scheduler problem which only bites with the
low-latency patches. Ingo fixed it and that patch went into the main
kernel. I _think_ it was in test8-pre. So a retest with test8 may be
worthwhile.
Don't go near test8-pre2, 3, 4, 5 or 6 - they'll chew your files.
