| Message ID | 20210622091153.29231-1-hare@suse.de |
|---|---|
| State | New |
| Headers | show |
| Series | [PATCHv2] virtio_scsi: do not overwrite SCSI status | expand |
Hannes, > When a sense code is present we should not override the scsi status; > the driver already sets it based on the response from the hypervisor. Applied to 5.14/scsi-staging, thanks! -- Martin K. Petersen Oracle Linux Engineering
> Applied to 5.14/scsi-staging, thanks! This fixes virtio-scsi for me. Thanks. Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
On Tue, 22 Jun 2021 11:11:53 +0200, Hannes Reinecke wrote: > When a sense code is present we should not override the scsi status; > the driver already sets it based on the response from the hypervisor. Applied to 5.14/scsi-queue, thanks! [1/1] virtio_scsi: do not overwrite SCSI status https://git.kernel.org/mkp/scsi/c/c43ddbf97f46 -- Martin K. Petersen Oracle Linux Engineering
Hi, On Tue, Jun 22, 2021 at 11:11:53AM +0200, Hannes Reinecke wrote: > When a sense code is present we should not override the scsi status; > the driver already sets it based on the response from the hypervisor. > > Fixes: 464a00c9e0ad ("scsi: core: Kill DRIVER_SENSE") > Signed-off-by: Hannes Reinecke <hare@suse.de> > Tested-by: Guenter Roeck <linux@roeck-us.net> > Tested-by: Jiri Slaby <jirislaby@kernel.org> > Tested-by: Christian Borntraeger <borntraeger@de.ibm.com> This patch didn't make it into mainline, but commit 464a00c9e0ad is there. This means that virtio_scsi is now broken in mainline (as of v5.13-9356-g4b820e167bf6). Did it get lost, or is it still queued somewhere ? Thanks, Guenter > --- > drivers/scsi/virtio_scsi.c | 3 +-- > 1 file changed, 1 insertion(+), 2 deletions(-) > > diff --git a/drivers/scsi/virtio_scsi.c b/drivers/scsi/virtio_scsi.c > index fd69a03d6137..ad78bf631900 100644 > --- a/drivers/scsi/virtio_scsi.c > +++ b/drivers/scsi/virtio_scsi.c > @@ -156,12 +156,11 @@ static void virtscsi_complete_cmd(struct virtio_scsi *vscsi, void *buf) > > WARN_ON(virtio32_to_cpu(vscsi->vdev, resp->sense_len) > > VIRTIO_SCSI_SENSE_SIZE); > - if (sc->sense_buffer) { > + if (resp->sense_len) { > memcpy(sc->sense_buffer, resp->sense, > min_t(u32, > virtio32_to_cpu(vscsi->vdev, resp->sense_len), > VIRTIO_SCSI_SENSE_SIZE)); > - set_status_byte(sc, SAM_STAT_CHECK_CONDITION); > } > > sc->scsi_done(sc);
Guenter, > This patch didn't make it into mainline, but commit 464a00c9e0ad is > there. This means that virtio_scsi is now broken in mainline (as of > v5.13-9356-g4b820e167bf6). Did it get lost, or is it still queued > somewhere ? It's still in 5.14/scsi-queue: https://git.kernel.org/pub/scm/linux/kernel/git/mkp/scsi.git/log/?h=5.14/scsi-queue -- Martin K. Petersen Oracle Linux Engineering
diff --git a/drivers/scsi/virtio_scsi.c b/drivers/scsi/virtio_scsi.c index fd69a03d6137..ad78bf631900 100644 --- a/drivers/scsi/virtio_scsi.c +++ b/drivers/scsi/virtio_scsi.c @@ -156,12 +156,11 @@ static void virtscsi_complete_cmd(struct virtio_scsi *vscsi, void *buf) WARN_ON(virtio32_to_cpu(vscsi->vdev, resp->sense_len) > VIRTIO_SCSI_SENSE_SIZE); - if (sc->sense_buffer) { + if (resp->sense_len) { memcpy(sc->sense_buffer, resp->sense, min_t(u32, virtio32_to_cpu(vscsi->vdev, resp->sense_len), VIRTIO_SCSI_SENSE_SIZE)); - set_status_byte(sc, SAM_STAT_CHECK_CONDITION); } sc->scsi_done(sc);
When a sense code is present we should not override the scsi status; the driver already sets it based on the response from the hypervisor. Fixes: 464a00c9e0ad ("scsi: core: Kill DRIVER_SENSE") Signed-off-by: Hannes Reinecke <hare@suse.de> Tested-by: Guenter Roeck <linux@roeck-us.net> Tested-by: Jiri Slaby <jirislaby@kernel.org> --- drivers/scsi/virtio_scsi.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)