>>>>> Markus W Kilbinger writes:

YAMAMOTO> can you try sys/malloc.h rev.1.81?

Markus> I did, and it seems to solve _all_ problems I had (beside the
Markus> crashed disklabel of my tibook ;-))! Great and thanks!

Hmmm, interesting. Did you update _only_ malloc.h?

I cvs updated me entire source tree, and saw that NEW_BUFQ_STRATEGY
started to work. Great I thought, but then backed down malloc.h to
1.80 again, just to test. And it still worked... Seems like some of
the other checkins yesterday may have fixed something too.

In any case NEW_BUFQ_STRATEGY works for me now, so I'm happy now. I
noticed however that QUEUEDEBUG doesn't seem to work at all:

fatal page fault in supervisor mode
trap type 6 code 0 eip c01897f7 cs 8 eflags 10282 cr2 5 imask 50
panic: trap
Begin traceback...
trap() at trap+0x20f
--- trap (number 6) ---
uhci_check_intr(c07c0000,1,c018977b,c07c0000,c07b5e80) at uhci_check_intr+0xb
uhci_softintr(c07c0000,c0434000,c0189634,c07b5c40) at uhci_softintr+0x26
softintr_dispatch(1) at softintr_dispatch+0x10a
Xsoftnet() at Xsoftnet+0x40
--- interrupt ---
mpidle(c03d77a0,0,c0437f80,c020bc5b) at mpidle
bpendtsleep(c03da474,20,c0340cce,0,0) at bpendtsleep
main(0,0,0,0,0) at main+0x52c
End traceback...
syncing disks... done
rebooting...

Kind regards,

        -- Urban