Re: SMP HT + USB2.0 crash

["davor emard" <davoremard@gmail.com>]

HI

I downloaded memtest86 iso image and let it running. I passed
1st battery of tests and continues. I'm pretty sure it will keep
pass those test without anyerrors, after all memories there
are 2x512MB samsung and for this machine I was choosing
ram's recommended by the manufacturer this motherboard.

Regarding nvidia, I don't like their binary policy but I may
tell you that both intel 925XE machines can run indefinitely
(for days even months until I reinstall new kernel)
each with a lot of simultaneous openGL with nvidia on 2 monitors
in dual head, dvb receiving over usb (here I mean USB 1.1 only)
firewire traffic, SPDIF output, ethernet traffic on both interfaces
and all that
through enabled SMP, PREEMPT and PREEMPT BIG_KERNEL_LOCK
and it is not going to crash.

As soon as I enable EHCI there it comes.
Otherwise I have option to use EHCI and don't use SMP
but not them both.



On 6/5/06, davor emard <davoremard@gmail.com> wrote:
> HI
>
> Due to popular demand I'm mailing the recent crash
> captured with serial cable on 2.6.16.19 on intel 925X
> triggered with SMP + USB2.0 combination.
>
> Here we got it, it's pretty random and useless for
> tracking the real problem so please don't blame
> software demux in DVB core just becaise it randomly
> crashed there - only thing that matters is that a800
> terrestroal receiver was running in USB 2.0 mode in
> order to generate enough usb2.0 traffic
> to trigger SMP+EHCI bug.
>
> I contributed myself to the dvb software demux code
> and it runs stable with it's demux core part almost
> unchanged for 2 years.
>
> If I remove EHCI and run a800 in USB 1.1 mode everything
> can run stable for any amount of time.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/