On Tue, 2011-10-18 at 11:08 +0200, Christoph Hellwig wrote: > This needs a patch like the one below folded in. In addition to not > unessecarily offload new command mapping to the fabric thread this > also fixes the currenly incorrect (outdated) handling of the return > values from transport_generic_allocate_tasks. I don't even have > a ppc compiler at hand right now, so handle it with care. > > --- > From: Christoph Hellwig <hch@xxxxxx> > Subject: ibmvscsis: do not offload new command processing to the fabric thread > > tcm_queuecommand is always called from process context, so prepare the command > in-process instead of offloading it. > > Signed-off-by: Christoph Hellwig <hch@xxxxxx> Hi Christoph, Fixed some minor compile breakage with this patch. Aside from that, I think it looks like a reasonably safe conversion to perform the necessary setup in tcm_queuecommand() process context instead of the extra context switch into transport_generic_handle_cdb_map(). I've not heard Tomo-san or BR-King feedback on this for a v3.2 merge yet, but will respin with this change included for them to review again for next week. Thanks, --nab -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
