Re: [PATCH] sym53c8xx_2: slave_alloc/destroy safety (2.6.27.5)

[Tony Battersby <tonyb@cybernetics.com>]

James Bottomley wrote:
> On Tue, 2008-12-30 at 12:10 +0200, Aaro Koskinen wrote:
>   
>> Thanks. The updated patch is below.
>>     
>
> Tony,
>
> Since we're lacking an active maintainer on this one and you seem
> interested, could you run these patches through your tests to make sure
> they're reasonably OK and then respond with an ack?
>
> Thanks,
>
> James
>
>
>
>   
This patch survived an overnight testing of "modprobe sym53c8xx;
modprobe -r sym53c8xx" in a loop, using a SCSI cable of marginal quality
that sometimes fails DV, with a mix of single-lun and multi-lun devices
of various SCSI revision levels, with many kernel debug options
enabled.  The only kernel warning message I see is the pre-existing and
already-reported problem about dma_free_coherent() being called with
interrupts disabled during rmmod, but that is not caused by this patch.

>From the description of this patch, I assume that it is a cleanup
instead of a fix for an actual problem.  Correspondingly, I don't have a
specific testcase that is improved by this patch.  I also have not
analyzed the source to see what the patch is trying to do.  However, the
current form of the patch passes testing as described above.

I will reply to the other patches in separate messages later.

Tested-by: Tony Battersby <tonyb@cybernetics.com>

Tony