Hi Sagi,
On Thu, 19 Mar 2015 12:36:41 +0200, Sagi Grimberg wrote :
> On 3/18/2015 4:45 PM, Christophe Vu-Brugier wrote:
> > A check that rejects a CDB with FUA bit set if no write cache is
> > emulated was added by the following commit:
> >
> > fde9f50 target: Add sanity checks for DPO/FUA bit usage
> >
> > However, this additional check is wrong: if FUA is supported or
> > emulated, then CDBs with FUA bit set should be accepted.
> >
> > Signed-off-by: Christophe Vu-Brugier <cvubrugier@xxxxxxxxxxx>
> > ---
> > drivers/target/target_core_sbc.c | 3 +--
> > 1 file changed, 1 insertion(+), 2 deletions(-)
> >
> > diff --git a/drivers/target/target_core_sbc.c b/drivers/target/target_core_sbc.c
> > index 856e800..a4431b4 100644
> > --- a/drivers/target/target_core_sbc.c
> > +++ b/drivers/target/target_core_sbc.c
> > @@ -703,8 +703,7 @@ sbc_check_dpofua(struct se_device *dev, struct se_cmd *cmd, unsigned char *cdb)
> > }
> > }
> > if (cdb[1] & 0x8) {
> > - if (!dev->dev_attrib.emulate_fua_write ||
> > - !dev->dev_attrib.emulate_write_cache) {
> > + if (!dev->dev_attrib.emulate_fua_write) {
> > pr_err("Got CDB: 0x%02x with FUA bit set, but device"
> > " does not advertise support for FUA write\n",
> > cdb[0]);
> >
>
>
> Why did the initiator thinks the device supports FUA if the
> modsense response depends on both FUA and WCE?
>
> static sense_reason_t spc_emulate_modesense(struct se_cmd *cmd)
> {
> ....
> if ((spc_check_dev_wce(dev)) &&
> (dev->dev_attrib.emulate_fua_write > 0))
> spc_modesense_dpofua(&buf[length], type);
> ....
> }
>
> I think maybe the device supports WCE but wasn't explicitly set with
> emulate_write_cache. Perhaps the check should be
> if (!dev->dev_attrib.emulate_fua_write || !spc_check_dev_wce(dev))
>
> Thoughts?
The backstore device I use is a disk partition.
The "emulate_write_cache" attribute is set to 0 and cannot be set to 1.
When I try to enable "emulate_write_cache", I get the following error
in `dmesg`:
emulate_write_cache not supported for this device
However, the write cache is enabled on the disk itself:
# hdparm -I /dev/sda | grep "Write cache"
* Write cache
So yes, the device supports WCE but it is not explicitly set with
the "emulate_write_cache" attribute.
Also, I tested what you suggest and confirm that it works. I will send
a patch.
Thank you Sagi!
--
Christophe Vu-Brugier
--
To unsubscribe from this list: send the line "unsubscribe target-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
