Re: ATAPI DMA timeouts showing up in logs

Re: ATAPI DMA timeouts showing up in logs

[Dennis Grant]

>On Thu, 1970-01-01 at 05:59, Dennis Grant wrote:

>> Now that I've got the proper IDE driver in place (2.4.20rc4)
>> and the master drive on the primary interface is running at
>> a full ATA133, these have started showing up in the logs - 
>> 2 or 3 a day:

> My guess is its the IDE command/DMA sequence bug that Khalid
> fixed in -ac. Some drives also take a very long time on 
> retrying blocks and that might cause a timeout/reset too.

OK.

Interestingly enough, the drive wasn't being used at all during the timeframe
when the messages show up - at least, not explictly. Perhaps some random process
was polling the drive....

Alan, I tried 2.4.20rc4-ac1 but while the patch took, it wouldn't compile. I'll
try again later and fwd where the error was. (I don't have access to that machine
at the moment)

>> This last one is the only indication that something might be >> amiss - the
two instances of "invalid argument" Other than
>> that, the drive appears to work just fine.

>Those are ones CD-ROM's dont support

Ah, OK. Perhaps hdparm shouldn't try them then. :)

Thanks.

DG

Re: IDE documentation bitrot, (Was: ATAPI DMA timeouts showing up in logs)

[Alan Cox]

> Is the current maintainer Mark Lord or Marcin Dalecki?  In the 2.5.x

Neither

> tree, it's had obsolete info removed, but it hasn't been brought up to
> date.

In theory its Andre & I - it does want updating badly

IDE documentation bitrot, (Was: ATAPI DMA timeouts showing up in logs)

[John Bradford]

>>> Now that I've got the proper IDE driver in place (2.4.20rc4)
>>> and the master drive on the primary interface is running at
>>> a full ATA133, these have started showing up in the logs - 
>>> 2 or 3 a day:
>> My guess is its the IDE command/DMA sequence bug that Khalid
>> fixed in -ac. Some drives also take a very long time on 
>> retrying blocks and that might cause a timeout/reset too.
> OK.

>>> This last one is the only indication that something might be amiss - the
>>> two instances of "invalid argument" Other than that, the drive
>>> appears to work just fine.
>> Those are ones CD-ROM's dont support
> Ah, OK. Perhaps hdparm shouldn't try them then. :)

There have been quite a few posts to linux-kernel about IDE diagnostic
messages that people are interpreting as fatal errors when they
aren't, and I've noticed that Documentation/ide.txt hasn't been
updated much for a long time - maybe it would be worth bringing it up
to date?

Is the current maintainer Mark Lord or Marcin Dalecki?  In the 2.5.x
tree, it's had obsolete info removed, but it hasn't been brought up to
date.

John.