Re: [2.6 patch] scsi/qla4xxx/ql4_isr.c: remove dead code

[Andrew Vasquez <andrew.vasquez@qlogic.com>]

On Tue, 19 Feb 2008, James Bottomley wrote:

> On Tue, 2008-02-19 at 18:35 -0800, Andrew Vasquez wrote:
> > On Tue, 19 Feb 2008, James Bottomley wrote:
> > 
> > > On Tue, 2008-02-19 at 21:29 +0200, Adrian Bunk wrote:
> > > > This patch removes dead code spotted by the Coverity checker.
> > > > 
> > > > Signed-off-by: Adrian Bunk <bunk@kernel.org>
> > > > 
> > > > ---
> > > > 
> > > >  drivers/scsi/qla4xxx/ql4_isr.c |   18 +-----------------
> > > >  1 file changed, 1 insertion(+), 17 deletions(-)
> > > > 
> > > > --- linux-2.6/drivers/scsi/qla4xxx/ql4_isr.c.old	2008-02-19 20:29:16.000000000 +0200
> > > > +++ linux-2.6/drivers/scsi/qla4xxx/ql4_isr.c	2008-02-19 20:30:37.000000000 +0200
> > > > @@ -91,38 +91,22 @@ static void qla4xxx_status_entry(struct 
> > > >  		if (scsi_status == 0) {
> > > >  			cmd->result = DID_OK << 16;
> > > >  			break;
> > > >  		}
> > > >  
> > > >  		if (sts_entry->iscsiFlags & ISCSI_FLAG_RESIDUAL_OVER) {
> > > >  			cmd->result = DID_ERROR << 16;
> > > >  			break;
> > > >  		}
> > > >  
> > > > -		if (sts_entry->iscsiFlags &ISCSI_FLAG_RESIDUAL_UNDER) {
> > > > +		if (sts_entry->iscsiFlags &ISCSI_FLAG_RESIDUAL_UNDER)
> > > >  			scsi_set_resid(cmd, residual);
> > > > -			if (!scsi_status && ((scsi_bufflen(cmd) - residual) <
> > > > -				cmd->underflow)) {
> > > > -
> > > > -				cmd->result = DID_ERROR << 16;
> > > > -
> > > > -				DEBUG2(printk("scsi%ld:%d:%d:%d: %s: "
> > > > -					"Mid-layer Data underrun0, "
> > > > -					"xferlen = 0x%x, "
> > > > -					"residual = 0x%x\n", ha->host_no,
> > > > -					cmd->device->channel,
> > > > -					cmd->device->id,
> > > > -					cmd->device->lun, __func__,
> > > > -					scsi_bufflen(cmd), residual));
> > > > -				break;
> > > > -			}
> > > > -		}
> > > 
> > > This code doesn't look dead to me, it looks to be enforcing
> > > cmd->underrun if set ... what makes the coverity checker think it can
> > > never be executed?
> > 
> > Hmm, guess it's the earlier 'if (scsi_status == 0)' check a few lines
> > up...  Dave S., can you take a look at this...  Thanks, av
> 
> Ah, so the !scsi_status is wrong it was supposed to be scsi_status !=
> 0 ... and even then it can just be dropped.

My guess is that the check should have been written as:

	...
	if (sts_entry->iscsiFlags &ISCSI_FLAG_RESIDUAL_UNDER)
		scsi_set_resid(cmd, residual);
		if ((scsi_bufflen(cmd) - residual) < cmd->underflow) {
			...

It looks to be a logic-error while porting from qla2xxx, where
scsi_status during CS_COMPLETE is the full 16-bit status (high-byte is
transport, low-byte SCSI status) from from the FCP_RSP frame (not so
in iSCSI, where it's just the SCSI-status) and the residual check
in qla_isr.c::qla2x00_status_entry() looks like:

	if (!lscsi_status &&
	    ((unsigned)(scsi_bufflen(cp) - resid) <
	     cp->underflow)) {
		...

I'll defer to Dave S. for verification.