Re: [dm-devel] dm-io async WRITE_SAME results in iSCSI NULL pointer [was: Re: Write same support]

["Martin K. Petersen" <martin petersen oracle com>]

>>>>> "Vivek" == Vivek Goyal <vgoyal redhat com> writes:

Vivek> First we returned "false" from bio_has_data() and skipped above
Vivek> assignment, in blk_rq_bio_prep() and now we are trying to make up
Vivek> for it.

Vivek> Maybe returning false from bio_has_data() for WRITE_SAME is not
Vivek> such a good idea because this bio has one logical block of data.

bio_has_data() is used for two different things. In my current tree I
split it up into bio_has_data() and bio_is_rw().

But given Mike's merge observation I think I'll take a step back and see
if I can come up with a better way to handle these
REQ_I_CANT_BELIEVE_ITS_NOT_FS requests in general. I really don't want
to have all the special cases scattered all over the place.

-- 
Martin K. Petersen	Oracle Linux Engineering