Skip to contentSkip to conversation

libata-core.c conversion to new debugging scheme, part 1 (25% done)

PATCH Applied Borislav Petkov 4 participants 9 messages May 31, 2006 – Jun 12, 2006

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 libata-core.c conversion to new debugging scheme, part 1 (25% done)

This patch converts the first 25% of libata-core.c to the new debugging scheme.

Signed-off-by <[email protected]>
drivers/scsi/libata-core.c renamed from drivers/scsi/libata-core.c.orig +36 -21
@@ -411,7 +411,7 @@ static const char *sata_spd_string(unsig
 
 void ata_dev_disable(struct ata_device *dev)
 {
-	if (ata_dev_enabled(dev)) {
+	if (ata_dev_enabled(dev) && ata_msg_drv(dev->ap)) {
 		ata_dev_printk(dev, KERN_WARNING, "disabled\n");
 		dev->class++;
 	}
@@ -770,8 +770,12 @@ void ata_std_dev_select (struct ata_port
 void ata_dev_select(struct ata_port *ap, unsigned int device,
 			   unsigned int wait, unsigned int can_sleep)
 {
-	VPRINTK("ENTER, ata%u: device %u, wait %u\n",
-		ap->id, device, wait);
+	if (ata_msg_probe(ap)) {
+		struct ata_device *dev = container_of(&ap, struct ata_device, ap);
+		ata_dev_printk(dev, KERN_INFO, "ata_dev_select: ENTER, ata%u: "
+				"device %u, wait %u\n",
+				ap->id, device, wait);
+	}
 
 	if (wait)
 		ata_wait_idle(ap);
@@ -1297,18 +1301,21 @@ static int ata_dev_configure(struct ata_
 	unsigned int xfer_mask;
 	int i, rc;
 
-	if (!ata_dev_enabled(dev)) {
-		DPRINTK("ENTER/EXIT (host %u, dev %u) -- nodev\n",
-			ap->id, dev->devno);
+	if (!ata_dev_enabled(dev) && ata_msg_info(ap)) {
+		ata_dev_printk(dev, KERN_INFO, "%s: ENTER/EXIT (host %u, dev %u) -- nodev\n",
+			__FUNCTION__, ap->id, dev->devno);
 		return 0;
 	}
 
-	DPRINTK("ENTER, host %u, dev %u\n", ap->id, dev->devno);
+	if (ata_msg_probe(ap))
+		ata_dev_printk(dev, KERN_DEBUG, "%s: ENTER, host %u, dev %u\n", 
+			__FUNCTION__, ap->id, dev->devno);
 
 	/* print device capabilities */
-	if (print_info)
-		ata_dev_printk(dev, KERN_DEBUG, "cfg 49:%04x 82:%04x 83:%04x "
+	if (ata_msg_probe(ap))
+		ata_dev_printk(dev, KERN_DEBUG, "%s: cfg 49:%04x 82:%04x 83:%04x "
 			       "84:%04x 85:%04x 86:%04x 87:%04x 88:%04x\n",
+			       __FUNCTION__,
 			       id[49], id[82], id[83], id[84],
 			       id[85], id[86], id[87], id[88]);
 
@@ -1328,7 +1335,8 @@ static int ata_dev_configure(struct ata_
 	/* find max transfer mode; for printk only */
 	xfer_mask = ata_id_xfermask(id);
 
-	ata_dump_id(id);
+	if (ata_msg_probe(ap))
+		ata_dump_id(id);
 
 	/* ATA-specific feature tests */
 	if (dev->class == ATA_DEV_ATA) {
@@ -1349,7 +1357,7 @@ static int ata_dev_configure(struct ata_
 			ata_dev_config_ncq(dev, ncq_desc, sizeof(ncq_desc));
 
 			/* print device info to dmesg */
-			if (print_info)
+			if (ata_msg_info(ap))
 				ata_dev_printk(dev, KERN_INFO, "ATA-%d, "
 					"max %s, %Lu sectors: %s %s\n",
 					ata_id_major_version(id),
@@ -1372,7 +1380,7 @@ static int ata_dev_configure(struct ata_
 			}
 
 			/* print device info to dmesg */
-			if (print_info)
+			if (ata_msg_info(ap))
 				ata_dev_printk(dev, KERN_INFO, "ATA-%d, "
 					"max %s, %Lu sectors: CHS %u/%u/%u\n",
 					ata_id_major_version(id),
@@ -1383,7 +1391,8 @@ static int ata_dev_configure(struct ata_
 
 		if (dev->id[59] & 0x100) {
 			dev->multi_count = dev->id[59] & 0xff;
-			DPRINTK("ata%u: dev %u multi count %u\n",
+			if (ata_msg_info(ap))
+				ata_dev_printk(dev, KERN_INFO, "ata%u: dev %u multi count %u\n",
 				ap->id, dev->devno, dev->multi_count);
 		}
 
@@ -1396,8 +1405,9 @@ static int ata_dev_configure(struct ata_
 
 		rc = atapi_cdb_len(id);
 		if ((rc < 12) || (rc > ATAPI_CDB_LEN)) {
-			ata_dev_printk(dev, KERN_WARNING,
-				       "unsupported CDB len\n");
+			if (ata_msg_warn(ap))
+				ata_dev_printk(dev, KERN_WARNING, 
+					"unsupported CDB len\n");
 			rc = -EINVAL;
 			goto err_out_nosup;
 		}
@@ -1409,7 +1419,7 @@ static int ata_dev_configure(struct ata_
 		}
 
 		/* print device info to dmesg */
-		if (print_info)
+		if (ata_msg_info(ap))
 			ata_dev_printk(dev, KERN_INFO, "ATAPI, max %s%s\n",
 				       ata_mode_string(xfer_mask),
 				       cdb_intr_string);
@@ -1423,7 +1433,7 @@ static int ata_dev_configure(struct ata_
 
 	/* limit bridge transfers to udma5, 200 sectors */
 	if (ata_dev_knobble(dev)) {
-		if (print_info)
+		if (ata_msg_info(ap))
 			ata_dev_printk(dev, KERN_INFO,
 				       "applying bridge limits\n");
 		dev->udma_mask &= ATA_UDMA5;
@@ -1433,11 +1443,15 @@ static int ata_dev_configure(struct ata_
 	if (ap->ops->dev_config)
 		ap->ops->dev_config(ap, dev);
 
-	DPRINTK("EXIT, drv_stat = 0x%x\n", ata_chk_status(ap));
+	if (ata_msg_probe(ap))
+		ata_dev_printk(dev, KERN_DEBUG, "%s: EXIT, drv_stat = 0x%x\n",
+			__FUNCTION__, ata_chk_status(ap));
 	return 0;
 
 err_out_nosup:
-	DPRINTK("EXIT, err\n");
+	if (ata_msg_probe(ap))
+		ata_dev_printk(dev, KERN_DEBUG, 
+				"%s: EXIT, err\n", __FUNCTION__);
 	return rc;
 }
 
@@ -1478,8 +1492,9 @@ static int ata_bus_probe(struct ata_port
 	if (ap->ops->probe_reset) {
 		rc = ap->ops->probe_reset(ap, classes);
 		if (rc) {
-			ata_port_printk(ap, KERN_ERR,
-					"reset failed (errno=%d)\n", rc);
+			if (ata_msg_err(ap))
+				ata_port_printk(ap, KERN_ERR, 
+						"reset failed (errno=%d)\n", rc);
 			return rc;
 		}
 	} else {

___________________________________________________________
Gesendet von Yahoo! Mail - Jetzt mit 1GB Speicher kostenlos - Hier anmelden: http://mail.yahoo.de

#1 lore raw
Quoted 3 lines · Borislav Petkov

This patch converts the first 25% of libata-core.c to the new debugging scheme.

Signed-off-by <[email protected]>

Looks good at first glance, though I'll hold off on apply until I return from the Red Hat Summit in Nashville.

It made me remember another point, though: in order to avoid regressions, after applying your patch, I would think that you would want to create a patch which did something like

#ifndef ATA_VERBOSE_DEBUG
	ap->msg_enable = xxx
#else ATA_DEBUG
	ap->msg_enable = yyy
#else
	ap->msg_enable = ...

#endif

Thus, we are assured that applying your patch will not change the behavior(much?) when ATA_VERBOSE_DEBUG is enabled.

Regards,

Jeff

#2 lore raw
Quoted 12 lines · Jeff Garzik
Quoted 3 lines · Borislav Petkov

This patch converts the first 25% of libata-core.c to the new debugging scheme.

Signed-off-by <[email protected]>

Looks good at first glance, though I'll hold off on apply until I return from the Red Hat Summit in Nashville.

It made me remember another point, though: in order to avoid regressions, after applying your patch, I would think that you would want to create a patch which did something like

Quoted 7 lines
#ifndef ATA_VERBOSE_DEBUG
	ap->msg_enable = xxx
#else ATA_DEBUG
	ap->msg_enable = yyy
#else
	ap->msg_enable = ...

#endif

Ok, i was thinking something along the lines of the following; for better granularity we might add more ATA_XXX defines or let the user subsequently adjust dbg level through sysfs...? (The patch applies on top of the one i sent you before.)

Adjust initial debugging levels through preprocessor defines.

Signed-off-by <[email protected]>
drivers/scsi/libata-core.c renamed from drivers/scsi/libata-core.c.orig1 +9 -1
@@ -5200,7 +5200,15 @@ static void ata_host_init(struct ata_por
 	ap->sata_spd_limit = UINT_MAX;
 	ap->active_tag = ATA_TAG_POISON;
 	ap->last_ctl = 0xFF;
-	ap->msg_enable = ATA_MSG_DRV;
+
+#if defined(ATA_VERBOSE_DEBUG)
+	/* turn on all debugging levels */
+	ap->msg_enable = 0x00FF;
+#elif defined(ATA_DEBUG)
+	ap->msg_enable = ATA_MSG_DRV | ATA_MSG_INFO | ATA_MSG_CTL | ATA_MSG_WARN | ATA_MSG_ERR;
+#else 
+	ap->msg_enable = ATA_MSG_DRV | ATA_MSG_ERR;
+#endif
 
 	INIT_WORK(&ap->port_task, NULL, NULL);
 	INIT_LIST_HEAD(&ap->eh_done_q);

___________________________________________________________
Telefonate ohne weitere Kosten vom PC zum PC: http://messenger.yahoo.de

#3 lore raw
Quoted 39 lines · Borislav Petkov
Quoted 11 lines · Jeff Garzik
Quoted 3 lines · Borislav Petkov

This patch converts the first 25% of libata-core.c to the new debugging scheme.

Signed-off-by <[email protected]>

Looks good at first glance, though I'll hold off on apply until I return from the Red Hat Summit in Nashville.

It made me remember another point, though: in order to avoid regressions, after applying your patch, I would think that you would want to create a patch which did something like

Quoted 7 lines
#ifndef ATA_VERBOSE_DEBUG
	ap->msg_enable = xxx
#else ATA_DEBUG
	ap->msg_enable = yyy
#else
	ap->msg_enable = ...

#endif

Ok, i was thinking something along the lines of the following; for better granularity we might add more ATA_XXX defines or let the user subsequently adjust dbg level through sysfs...? (The patch applies on top of the one i sent you before.)

Adjust initial debugging levels through preprocessor defines.

Signed-off-by <[email protected]>
drivers/scsi/libata-core.c renamed from drivers/scsi/libata-core.c.orig1 +2 -1
@@ -5200,7 +5200,15 @@ static void ata_host_init(struct ata_por
 	ap->sata_spd_limit = UINT_MAX;
 	ap->active_tag = ATA_TAG_POISON;
 	ap->last_ctl = 0xFF;
-	ap->msg_enable = ATA_MSG_DRV;
+
+#if defined(ATA_VERBOSE_DEBUG)

drivers/scsi/libata-core.c static void ata_host_init(struct ata_por · line 5214

Quoted 10 lines
drivers/scsi/libata-core.c renamed from drivers/scsi/libata-core.c.orig1 +7 -0
+	/* turn on all debugging levels */
+	ap->msg_enable = 0x00FF;
+#elif defined(ATA_DEBUG)
+	ap->msg_enable = ATA_MSG_DRV | ATA_MSG_INFO | ATA_MSG_CTL | ATA_MSG_WARN | ATA_MSG_ERR;
+#else 
+	ap->msg_enable = ATA_MSG_DRV | ATA_MSG_ERR;
+#endif
 
 	INIT_WORK(&ap->port_task, NULL, NULL);
 	INIT_LIST_HEAD(&ap->eh_done_q);

Hello,

Those ATA_MSG_* constants designates two things..

* message level (debug, info, warning...)
* message origin (probe, intr...)

although above distinction isn't clear for some constants. libata now uses ata_port/dev_printk() macros to print messages and the second argument is message level (KERN_INFO, KERN_WARNING...), which carries duplicate information as above ATA_* constants. IMHO, it would be better to fold the two into one. e.g.

  ata_port_printk(ap, ATA_MSG_INFO, "blah blah\n");

instead of

if (ata_msg_info(ap))

	ata_port_printk(ap, KERN_INFO, "blah blah\n");

Some constants probably need to be adjusted a bit though.

Thanks.

Signature · 1 line
tejun
#4 lore raw
Quoted 11 lines · Tejun Heo

Those ATA_MSG_* constants designates two things..

* message level (debug, info, warning...)
* message origin (probe, intr...)

although above distinction isn't clear for some constants. libata now uses ata_port/dev_printk() macros to print messages and the second argument is message level (KERN_INFO, KERN_WARNING...), which carries duplicate information as above ATA_* constants. IMHO, it would be better to fold the two into one. e.g.

Quoted 8 lines
  ata_port_printk(ap, ATA_MSG_INFO, "blah blah\n");

instead of

if (ata_msg_info(ap))

	ata_port_printk(ap, KERN_INFO, "blah blah\n");

Some constants probably need to be adjusted a bit though.

Although I agree, it is best to integrate the msg_enable support in "waves", not doing too much at once, so that we have time to reflect better on the next step -- just like what we're doing now.

The ata_msg_xxx level and KERN_xxx level are not perfectly mapped, nor should they be, so the proposed scheme may not work out once we have all the message levels integrated.

The most important goal to achieve is (a) eliminating the need for users to rebuild libata to get verbose messages and (b) allowing fine-grained per-port verbose debugging and tracing. We can address cleanups after those goals.

Jeff

#5 lore raw
Quoted 3 lines · Jeff Garzik

The ata_msg_xxx level and KERN_xxx level are not perfectly mapped, nor should they be, so the proposed scheme may not work out once we have all the message levels integrated.

How about the folowing mapping for ATA-to-printk debugging levels:

ATA_MSG_DRV ----> KERN_INFO
ATA_MSG_INFO

ATA_MSG_PROBE ----> KERN_DEBUG
ATA_MSG_MALLOC
ATA_MSG_CTL
ATA_MSG_INTR

since those are in their nature debugging infos containing ENTER/EXIT function_name calls

ATA_MSG_WARN ----> KERN_WARNING

ATA_MSG_ERR ----> KERN_ERR

but this I could do after the whole conversion thing is done so we have the new dbg infrastructure first.

Regards,
Boris.

___________________________________________________________
Gesendet von Yahoo! Mail - Jetzt mit 1GB Speicher kostenlos - Hier anmelden: http://mail.yahoo.de

#6 lore raw
In reply to Tejun Heo
Quoted 1 line

Some constants probably need to be adjusted a bit though.

I couldn't agree more, I'll think of some mapping later on and post it here.

Regards,
Boris.

___________________________________________________________
Telefonate ohne weitere Kosten vom PC zum PC: http://messenger.yahoo.de

Applied
#7 lore raw
In reply to Borislav Petkov
Quoted 36 lines · Borislav Petkov
Quoted 11 lines · Jeff Garzik
Quoted 3 lines · Borislav Petkov

This patch converts the first 25% of libata-core.c to the new debugging scheme.

Signed-off-by <[email protected]>

Looks good at first glance, though I'll hold off on apply until I return from the Red Hat Summit in Nashville.

It made me remember another point, though: in order to avoid regressions, after applying your patch, I would think that you would want to create a patch which did something like

Quoted 7 lines
#ifndef ATA_VERBOSE_DEBUG
	ap->msg_enable = xxx
#else ATA_DEBUG
	ap->msg_enable = yyy
#else
	ap->msg_enable = ...

#endif

Ok, i was thinking something along the lines of the following; for better granularity we might add more ATA_XXX defines or let the user subsequently adjust dbg level through sysfs...? (The patch applies on top of the one i sent you before.)

Adjust initial debugging levels through preprocessor defines.

Signed-off-by <[email protected]>
drivers/scsi/libata-core.c renamed from drivers/scsi/libata-core.c.orig1 +0 -0
@@ -5200,7 +5200,15 @@ static void ata_host_init(struct ata_por
 	ap->sata_spd_limit = UINT_MAX;
 	ap->active_tag = ATA_TAG_POISON;
 	ap->last_ctl = 0xFF;

drivers/scsi/libata-core.c static void ata_host_init(struct ata_por · line 5211

Quoted 10 lines
drivers/scsi/libata-core.c renamed from drivers/scsi/libata-core.c.orig1 +9 -1
-	ap->msg_enable = ATA_MSG_DRV;
+
+#if defined(ATA_VERBOSE_DEBUG)
+	/* turn on all debugging levels */
+	ap->msg_enable = 0x00FF;
+#elif defined(ATA_DEBUG)
+	ap->msg_enable = ATA_MSG_DRV | ATA_MSG_INFO | ATA_MSG_CTL | ATA_MSG_WARN | ATA_MSG_ERR;
+#else 
+	ap->msg_enable = ATA_MSG_DRV | ATA_MSG_ERR;
+#endif

applied

#8 lore raw
In reply to Borislav Petkov
Quoted 3 lines · Borislav Petkov

This patch converts the first 25% of libata-core.c to the new debugging scheme.

Signed-off-by <[email protected]>

After I hand-applied a lot of this, and looked at the result, I would request further updates:

1) The message level order is a bit off. Following the original Becker-designed strategy (netif.txt), the ATA_MSG_xxx bits should be ordered in order of verbosity, where that order is passed to ata_msg_init() from the user.

This implies that that ATA_MSG_ERR (the most important) should be 0x1, _WARN should be 0x2, PROBE(?), DRV(?), and so on.

2) rediff and resend this "part 1" patch against the latest libata-dev.git#upstream, which looked mostly OK

Jeff