On Sun, 9 Jul 2006, Justin Piszcz wrote:

On Sun, 9 Jul 2006, Mark Lord wrote:

quoted

Mmm.. there are two main paths into those messages,
and my current patch only caught one of them.

Here's a reworked version that catches the ata_op on both paths.
Maybe this will dump out the info we need to diagnose Justin's system.

Compiles & links fine on 2.6.17, but not tested.

Cheers

--- linux/drivers/scsi/libata-scsi.c.orig	2006-06-23 13:38:37.000000000 
-0400
+++ linux/drivers/scsi/libata-scsi.c	2006-07-09 12:19:52.000000000 -0400

@@ -542,6 +542,7 @@

	struct ata_taskfile *tf = &qc->tf;
	unsigned char *sb = cmd->sense_buffer;
	unsigned char *desc = sb + 8;
+	unsigned char ata_op = tf->command;

	memset(sb, 0, SCSI_SENSE_BUFFERSIZE);

@@ -558,6 +559,7 @@

	 * onto sense key, asc & ascq.
	 */
	if (tf->command & (ATA_BUSY | ATA_DF | ATA_ERR | ATA_DRQ)) {
+		printk(KERN_WARNING "ata_gen_ata_desc_sense: failed 
ata_op=0x%02x\n", ata_op);
		ata_to_sense_error(qc->ap->id, tf->command, tf->feature,
				   &sb[1], &sb[2], &sb[3]);
		sb[1] &= 0x0f;

@@ -617,6 +619,7 @@

	struct scsi_cmnd *cmd = qc->scsicmd;
	struct ata_taskfile *tf = &qc->tf;
	unsigned char *sb = cmd->sense_buffer;
+	unsigned char ata_op = tf->command;

	memset(sb, 0, SCSI_SENSE_BUFFERSIZE);

@@ -633,6 +636,7 @@

	 * onto sense key, asc & ascq.
	 */
	if (tf->command & (ATA_BUSY | ATA_DF | ATA_ERR | ATA_DRQ)) {
+		printk(KERN_WARNING "ata_gen_fixed_sense: failed 
ata_op=0x%02x\n", ata_op);
		ata_to_sense_error(qc->ap->id, tf->command, tf->feature,
				   &sb[2], &sb[12], &sb[13]);
		sb[2] &= 0x0f;

Thanks Mark!

Applying now.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Mark,

Check line 519, this is where it is printing the error (I believe) and 
the patch does not print the ata_op here.

It is in the ata_to_sense_error() function.

I've already patched, as you can see, recompiled, etc..

# patch -p0 < /tmp/b
patching file linux/drivers/scsi/libata-scsi.c
Reversed (or previously applied) patch detected!  Assume -R? [n]
#

Jul  9 15:22:57 p34 kernel: [4300704.724000] ata3: translated ATA stat/err 
0x51/04 to SCSI SK/ASC/ASCQ 0xb/00/00
Jul  9 15:22:57 p34 kernel: [4300704.724000] ata3: status=0x51 { 
DriveReady SeekComplete Error }
Jul  9 15:22:57 p34 kernel: [4300704.724000] ata3: error=0x04 { 
DriveStatusError }

This part needs the ata_op:

     519  translate_done:
     520         printk(KERN_ERR "ata%u: translated ATA stat/err 0x%02x/%02x to "
     521                "SCSI SK/ASC/ASCQ 0x%x/%02x/%02x\n", id, drv_stat, drv_err,
     522                *sk, *asc, *ascq);


Justin.