Re: [PATCH v8 0/10] More device removal fixes

[Bart Van Assche <bvanassche@acm.org>]

On 02/06/13 23:31, Joe Lawrence wrote:
> On Tue, 5 Feb 2013, Bart Van Assche wrote:
>> Fix a few race conditions that can be triggered by removing a device:
>> - Fix a race between starved list processing and device removal.
>> - Avoid that a SCSI LLD callback can get invoked after
>>    scsi_remove_host() finished.
>> - Speed up device removal by stopping error handling as soon as
>>    the SHOST_DEL or SHOST_DEL_RECOVERY state has been reached.
>> - Save and restore the host_scribble field during error handling
>>
>> These patches have been tested on top of kernel v3.8-rc6 and are also
>> available here: http://github.com/bvanassche/linux/tree/device-removal-fixes.
>
> Hi Bart,
>
> I gave your patchset a spin on our ftServer, which supports hotplug of PCI
> devices.  Before attempting a surprise HW removal of an LSI Fusion SAS-2
> SCSI controller, I issued an orderly removal through our hotplug driver.
> The device removal appeared to hang (this does not occur with vanilla
> 3.8.0-rc6).

Hello Joe,

Thanks for running this test. I have had a quick look at the mpt2sas 
_scsih_remove() implementation but I have not yet had the time to 
analyze that function entirely. My proposal is to defer further analysis 
of the interaction between the two patches that make scsi_remove_host() 
wait and the mpt2sas driver until a later time. It would be appreciated 
if you could run the following test:
* Revert the two patches that make scsi_remove_host() wait (these
   two patches should revert cleanly).
* Repeat your device removal test.
If that test runs fine, I will retest and repost this patch series 
without the two patches that make scsi_remove_host() wait.

Thanks,

Bart.