On Mon, Jun 05 2017, Shaohua Li wrote:

> On Mon, Jun 05, 2017 at 04:49:39PM +1000, Neil Brown wrote:
>> 
>> If mddev_suspend() races with md_write_start() we can deadlock
>> with mddev_suspend() waiting for the request that is currently
>> in md_write_start() to complete the ->make_request() call,
>> and md_write_start() waiting for the metadata to be updated
>> to mark the array as 'dirty'.
>> As metadata updates done by md_check_recovery() only happen then
>> the mddev_lock() can be claimed, and as mddev_suspend() is often
>> called with the lock held, these threads wait indefinitely for each
>> other.
>> 
>> We fix this by having md_write_start() abort if mddev_suspend()
>> is happening, and ->make_request() aborts if md_write_start()
>> aborted.
>> md_make_request() can detect this abort, decrease the ->active_io
>> count, and wait for mddev_suspend().
>
> Applied the two patches, thanks! For this one, I added md_start_write symbol back. Will
> push these to Linus in the 4.12 cycle.
>


Thanks for fixing the md_start_write export!  At one stage in
development I made it "static", but then forgot the export when I put it
back :-(

Thanks,
NeilBrown