> However, since version 2.1.33, it's been down-hill. Many of the new
> "improvements" may be causing problems that "look" as though they are
> SMP related but I don't think so. I recompiled the 2.1.37 without
> SMP, I needed to change the location of a couple #defines because
> it would not compile "out-of-the-box" with SMP disabled.
there is no point in fixing half-done cleanups. But now it really looks
very nifty :)
btw, my IDE-disk SMP system is not flaky at all (under 2.1.36, ... pre 37
has slight booting problems ;), so i really suspect that there is some
unresolved SCSI problem haunting you, so if you are looking at problem
sources you should look into that area IMHO.
-- mingo