Re: ad WRITE_DMA timing out frequently

From: Paul Mather <paul_at_gromit.dlib.vt.edu>
Date: Wed, 02 Mar 2005 13:55:50 -0500
On Sat, 2005-02-19 at 15:56 +0000, Ian Dowse wrote:
> In message <1108752445.1105.34.camel_at_zappa.Chelsea-Ct.Org>, Paul Mather writes:
> >The "TIMEOUT - WRITE_DMA" issue has been a recurring problem for me
> >since somewhere in the 5.2.1--5.3 release range.  (It's been so long now
> >that I don't remember whether it first started plaguing me in 5.2.1 or
> >5.3.  I do know for definite I never got this problem in 5.1 and it only
> >crept in during an "upgrade.")
> 
> On a recent -CURRENT you could try the following patch. It attempts
> to clean up the handling of timeouts in the ATA code by using the
> new callout_init_mtx() function, and appeared to cure fairly frequent
> WRITE_DMA timeout messages for me.

[[Patch omitted.]]


I've been using the patch Ian posted for a week or so now and have not
had any of the serious problems I had before I applied the patch.  I
have had two incidences of "TIMEOUT - WRITE_DMA" in the time I've been
running with the patch, but they have not resulted in the "FAILURE -
WRITE_DMA timed out" and associated fatal provider disconnection I
experienced previously.  It may be completely unrelated, but since
running with this patch, my system also hasn't experienced any of the
occasional spontaneous reboots I would sometimes get before, and seems
overall more robust.

It would be great if this patch could be applied to -CURRENT and MFC5'd,
too.

Cheers,

Paul.
-- 
e-mail: paul_at_gromit.dlib.vt.edu

"Without music to decorate it, time is just a bunch of boring production
 deadlines or dates by which bills must be paid."
        --- Frank Vincent Zappa
Received on Wed Mar 02 2005 - 17:55:59 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:29 UTC