Hello Linux-kernel,
If I start an SMP kernel with the "nosmp" option, it locks within 5
minutes with very little load.
Others are way more stable: a UP kernel seems just really stable. An SMP
kernel makes it unstable but workable (and it *is* a BP6 board so
instability may be inherent). An SMP kernel with the "nosmp" option will
lock the machine by doing something like starting X, then start netscape
or so. The strange fact is that it starts with hanging keyboard and
mouse, but will let you log in remotely and/or kill X; but the keyboard
remains dead (tried "kbdrate" and unplugging, to no avail). Then the
eth0 hang too, another time I had just enough time to do a remote
reboot.
I would guess the "nosmp" option should work like a UP kernel, not like
this, should it?
As said: a UP kernel is *stable*, whatever I try.
A normal SMP kernel is unstable but it takes more effort to show this:
something like cp -av /mnt/hd0 /mnt/hd1 will work - mostly. It stops
with an DMA timeout and a few seconds later, it freezes. "noapic"
doesn't help.
I also saw a couple of messages from people that had *SCSI* drives.
Those started resetting. Could these two facts be related? It seems to
look like some type of bug in the SMP code...
Now I *do* have a BP6 board, but this still does not explain the easy
lock with "nosmp".
Details:
2.2.15pre14 with ide.2.2.15pre13
BP6, 2x Celeron 466, 128Mb. 15G with HPT366. Matrox, Epic100 and
inactive SB16. No shared PCI interrupts (9, 10, 11, 12).
Best regards,
Valentijn
(tried to subscribe to linux-kernel but got no response... is that
normal?)
--- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sat Mar 18 2000 - 05:51:32 EST