[PATCH] Fix for 3ware driver
Jason Thomson
jason.thomson at mintel.com
Wed May 5 01:13:24 PDT 2004
Well, it didn't fix my problem.
I'm guessing that the problem I have might possibly be something to do
with bad hardware. (We have about 10 similarly configured boxes and
this is the only one we're having problems with - although the workload
is not the same on any two boxes).
I'm going to start replacing hardware until the problem goes away.
(Fingers crossed).
Jason Thomson wrote:
> Sorry to reply to my own message (should have checked first). I've just
> looked at the code, and from a cursory glance, it seems that the code
> that has been changed is the same in the old driver and the new driver.
>
> As I recall from a previous investigation in December, the differences
> between the two versions were small, anyway. The new (3ware approved)
> driver looked like a tidier version of the old one.
>
> So I guess it's likely that the problem did exist before. I'll cvsup
> and apply the patch.
>
>
>
>
>
> Jason Thomson wrote:
>
>> Is it possible that this problem is also exhibited by a 4.9-STABLE
>> Kernel from January 22nd? i.e. might this same (or a similar) problem
>> have existed in the previous driver?
>>
>> I am seeing i/o hangs (swap_pager: indefinite wait buffer messages) on
>> one of our boxes here. (4 x Maxtor 300GB drives / 3ware 7506-4 /
>> 2.4GHz Xeon Dell 1600SC).
>>
>> The machine is a backup server running BackupPC - it does experience
>> heavy load and i/o stress.
>>
>>
>> Scott Long wrote:
>>
>>> Mike Tancsa wrote:
>>>
>>>>
>>>> What cards has this shown up with and what versions of the BIOS ? I
>>>> have quite a few 3ware boxes deployed and have not seen any
>>>> problems. When was the bug introduced ?
>>
>>
>>
>>>
>>> The bugs were introduced into 4.x with the vendor update that happened
>>> on April 7. So far the only cards that we have tested against have been
>>> 6xxx series cards. The bugs deal with resource shortage handling, so
>>> it's likely that the slower 6xxx cards are more likely than the newer
>>> cards to experience the problems. However, on a fast machine under
>>> heavy I/O load, I would expect it to happen on any card.
>>>
>>> Scott
>>
>>
>>
>> _______________________________________________
>> freebsd-stable at freebsd.org mailing list
>> http://lists.freebsd.org/mailman/listinfo/freebsd-stable
>> To unsubscribe, send any mail to "freebsd-stable-unsubscribe at freebsd.org"
>>
>>
>
>
More information about the freebsd-stable
mailing list