[lkml]   [2004]   [May]   [30]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: Strange DMA-errors and system hang with Promise 20268
On Sat, 2004-05-29 at 16:42, Mario 'BitKoenig' Holbe wrote:
> Hi,
> On Sun, Mar 07, 2004 at 02:05:46AM +0100, Mario 'BitKoenig' Holbe wrote:
> > Mar 4 01:01:06 darkside kernel: hde: dma_timer_expiry: dma status == 0x21
> hmmm, it seems I solved the issue in my case.
> I did just connect the two disks on the Promise to a second separate
> power supply and everything works rock-stable and survives all things
> that went wrong before (parallel fsck, parallel hdparm -t as well as
> 'normal operation' over days).
> Things remain stable even with the WDC disk connected back to the
> Promise, which was heavy unstable before.
> Moreover, when I connect the disks back to the machines internal power
> supply, the problems arise again - immediately.
> IMHO, this does also explain, why the problems happen while heavy I/O
> (parallel over all disks) and/or while S.M.A.R.T selftests are running
> (also: parallel over all disks).
> Well, I thought, a 300W power supply would be enough for 1GHz P-III
> and 4 disks. And it definitely was enough for a long time.
> Most likely, the power supply just aged and its capacity decreased.
> Perhaps, the disks consume more power with newer kernels, but while
> crawling the lkm archives I found similar reports also for 2.4.19.
> I have no idea, why always the disks connected to the Promise
> controller did fail. Perhaps, the Promise is more sensitive regarding
> signal quality on the IDE wire.
> I have no idea, why my WDC disk failed when connected to the Promise,
> while others did work far more stable. Perhaps, the WDC disks signal
> quality under low-power is more bad than the one of other disks.
> I have no idea, why the WDC disk did work well when it was connected
> to the onboard controller. Perhaps, the lower signal quality of the
> WDC (if so) and the sensitivity of the Promise (if so) added together
> was too much at all.
> Henrik, Sebastian: if you still have problems, this is probably
> something to test for you.

Well. I don't have those problems anymore but..I have a question..
Should the box freeze just because there is some powerfailures? :/

Henrik Persson <>

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
More majordomo info at
Please read the FAQ at

 \ /
  Last update: 2005-03-22 14:03    [W:0.152 / U:0.308 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site