[PATCH] scsi: Simplify scsi_cdl_check_cmd()

[PATCH] scsi: Simplify scsi_cdl_check_cmd()

[Damien Le Moal]

Reading the 800+ pages of SPC often leads to a brain shutdown and to
less than ideal code... This resulted in the checks of the rwcdlp and
cdlp fields in scsi_cdl_check_cmd() to have identical if-else branches.

Replace this with a comment describing the cases we are interested in
and replace the if-else code block with a simple test of the cdlp field
that is used as the function return value.

Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Julia Lawall <julia.lawall@inria.fr>
Closes: https://lore.kernel.org/r/202306221657.BJHEADkz-lkp@intel.com/
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
---
 drivers/scsi/scsi.c | 37 ++++++++++++++-----------------------
 1 file changed, 14 insertions(+), 23 deletions(-)

diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c
index c4bf99a842f3..cea10616e630 100644
--- a/drivers/scsi/scsi.c
+++ b/drivers/scsi/scsi.c
@@ -586,31 +586,22 @@ static bool scsi_cdl_check_cmd(struct scsi_device *sdev, u8 opcode, u16 sa,
 	if ((buf[1] & 0x03) != 0x03)
 		return false;
 
-	/* See SPC-6, one command format of REPORT SUPPORTED OPERATION CODES */
+	/*
+	 * See SPC-6, one command format of REPORT SUPPORTED OPERATION CODES.
+	 * We have the following cases depending on rwcdlp (buf[0] & 0x01)
+	 * value:
+	 *  - rwcdlp == 0: then cdlp indicates support for the A mode page when
+	 *		   it is equal to 1 and for the B mode page when it is
+	 *		   equal to 2.
+	 *  - rwcdlp == 1: then cdlp indicates support for the T2A mode page
+	 *		   when it is equal to 1 and for the T2B mode page when
+	 *		   it is equal to 2.
+	 * Overall, to detect support for command duration limits, we only need
+	 * to check that cdlp is 1 or 2.
+	 */
 	cdlp = (buf[1] & 0x18) >> 3;
-	if (buf[0] & 0x01) {
-		/* rwcdlp == 1 */
-		switch (cdlp) {
-		case 0x01:
-			/* T2A page */
-			return true;
-		case 0x02:
-			/* T2B page */
-			return true;
-		}
-	} else {
-		/* rwcdlp == 0 */
-		switch (cdlp) {
-		case 0x01:
-			/* A page */
-			return true;
-		case 0x02:
-			/* B page */
-			return true;
-		}
-	}
 
-	return false;
+	return cdlp == 0x01 || cdlp == 0x02;
 }
 
 /**
-- 
2.40.1

Re: [PATCH] scsi: Simplify scsi_cdl_check_cmd()

[Christoph Hellwig]

> +	 * See SPC-6, one command format of REPORT SUPPORTED OPERATION CODES.

/one/on/ or even on the?

Re: [PATCH] scsi: Simplify scsi_cdl_check_cmd()

[Damien Le Moal]

On 6/23/23 14:40, Christoph Hellwig wrote:
>> +	 * See SPC-6, one command format of REPORT SUPPORTED OPERATION CODES.
> 
> /one/on/ or even on the?

The spec has that as "One_command parameter data format", as opposed to the
other format which is "All_command parameter data format". Will re-phrase it
exactly like that.

-- 
Damien Le Moal
Western Digital Research

Re: [PATCH] scsi: Simplify scsi_cdl_check_cmd()

[Christoph Hellwig]

On Fri, Jun 23, 2023 at 04:26:09PM +0900, Damien Le Moal wrote:
> On 6/23/23 14:40, Christoph Hellwig wrote:
> >> +	 * See SPC-6, one command format of REPORT SUPPORTED OPERATION CODES.
> > 
> > /one/on/ or even on the?
> 
> The spec has that as "One_command parameter data format", as opposed to the
> other format which is "All_command parameter data format". Will re-phrase it
> exactly like that.

Oh, ok.