Skip to contentSkip to conversation

ieee1394/ohci1394 : CycleTooLong interrupt management

PATCH In review Jean-Baptiste MUR 2 participants 3 messages Mar 17, 2006 – Mar 23, 2006

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ieee1394/ohci1394 : CycleTooLong interrupt management

This patch modifies the ohci1394.c file to enable and manage the "cycle too long" interrupt.
If this interrupt occurs, the "LinkControl.CycleMaster" bit of the host controller is reseted. This implies, that the host controller does not send "cycle start" packet anymore freezing then the isochronous communication. The management of the interrupt added by the patch is that when the interrupt occurs, the OHCI irq handler prints a kernel log warning and then sets the "LinkControl.CycleMaster" bit again resuming the isochronous communication.

---

Kernel version : 2.6.16-rc4

#############Patch begin

drivers/ieee1394/ohci1394.c +1 -0
index ab01a54..33850eb 100644
@@ -580,6 +580,7 @@ static void ohci_initialize(struct ti_oh
                  OHCI1394_isochRx |
                  OHCI1394_isochTx |
                  OHCI1394_postedWriteErr |
+                  OHCI1394_cycleTooLong |
                  OHCI1394_cycleInconsistent);
 
        /* Enable link */
@@ -2386,6 +2387,15 @@ static irqreturn_t ohci_irq_handler(int
                PRINT(KERN_ERR, "physical posted write error");
                /* no recovery strategy yet, had to involve protocol drivers 

*/

        }

+ if (event & OHCI1394_cycleTooLong) {
+ if(printk_ratelimit())
+ PRINT(KERN_WARNING, "isochronous cycle too long");
+ else
+ DBGMSG("OHCI1394_cycleTooLong");
+ /* If this event occurs, we try to reactivate the "cycle
master" bit. */
+ reg_write(ohci, OHCI1394_LinkControlSet,
OHCI1394_LinkControl_CycleMaster);
+ event &= ~OHCI1394_cycleTooLong;
+ }

        if (event & OHCI1394_cycleInconsistent) {
                /* We subscribe to the cycleInconsistent event only to
                 * clear the corresponding event bit... otherwise,

#############Patch end

#1 lore raw

Jean-Baptiste MUR wrote:
...

Quoted 1 line

Kernel version : 2.6.16-rc4

(actually plus an ohci1394 patch from the linux1394 development tree)

...

Quoted 4 lines

+ /* If this event occurs, we try to reactivate the "cycle
master" bit. */
+ reg_write(ohci, OHCI1394_LinkControlSet,
OHCI1394_LinkControl_CycleMaster);

...

Three minor nits:

There are tabs replaced by spaces and superfluous line breaks inserted in your posting. Take care that the mailer does not mangle whitespace. If you cannot prevent this nor switch the mailer, send the patch as attachment without recoding (i.e. 7bit encoding, not base64 or the like).

Take care of a maximum line length of 80 characters. For example:

		/* Try to reactivate the "cycle master" bit. */
		reg_write(ohci, OHCI1394_LinkControlSet,
			  OHCI1394_LinkControl_CycleMaster);

The comment is unnecessary since the actual code is clear enough, at least IMO.

Anyway, thanks again for identifying this problem.

Signature · 3 lines
Stefan Richter
-=====-=-==- --== =---=
http://arcgraph.de/sr/
#2 lore raw

Sorry for pollution. I have updated the patch accordingly to your comments. It seems my mailer don't like tab very much so I send it as an attachment. I hope this will be ok.

Jean-Baptiste

Quoted 23 lines · Stefan Richter

Jean-Baptiste MUR wrote:
...

Quoted 1 line

Kernel version : 2.6.16-rc4

(actually plus an ohci1394 patch from the linux1394 development tree)

...

Quoted 4 lines

+ /* If this event occurs, we try to reactivate the "cycle
master" bit. */
+ reg_write(ohci, OHCI1394_LinkControlSet,
OHCI1394_LinkControl_CycleMaster);

...

Three minor nits:

There are tabs replaced by spaces and superfluous line breaks inserted in your posting. Take care that the mailer does not mangle whitespace. If you cannot prevent this nor switch the mailer, send the patch as attachment without recoding (i.e. 7bit encoding, not base64 or the like).

Take care of a maximum line length of 80 characters. For example:

Quoted 13 lines
		/* Try to reactivate the "cycle master" bit. */
		reg_write(ohci, OHCI1394_LinkControlSet,
			  OHCI1394_LinkControl_CycleMaster);

The comment is unnecessary since the actual code is clear enough, at least IMO.

Anyway, thanks again for identifying this problem.

Signature · 3 lines
Stefan Richter
-=====-=-==- --== =---=
http://arcgraph.de/sr/
drivers/ieee1394/ohci1394.c +10 -0
index ab01a54..a240b2a 100644
@@ -580,6 +580,7 @@ static void ohci_initialize(struct ti_oh
 		  OHCI1394_isochRx |
 		  OHCI1394_isochTx |
 		  OHCI1394_postedWriteErr |
+		  OHCI1394_cycleTooLong |
 		  OHCI1394_cycleInconsistent);
 
 	/* Enable link */
@@ -2386,6 +2387,15 @@ static irqreturn_t ohci_irq_handler(int 
 		PRINT(KERN_ERR, "physical posted write error");
 		/* no recovery strategy yet, had to involve protocol drivers */
 	}
+	if (event & OHCI1394_cycleTooLong) {
+		if(printk_ratelimit())
+			PRINT(KERN_WARNING, "isochronous cycle too long");
+		else
+			DBGMSG("OHCI1394_cycleTooLong");
+		reg_write(ohci, OHCI1394_LinkControlSet, 
+			  OHCI1394_LinkControl_CycleMaster);
+		event &= ~OHCI1394_cycleTooLong;
+	}
 	if (event & OHCI1394_cycleInconsistent) {
 		/* We subscribe to the cycleInconsistent event only to
 		 * clear the corresponding event bit... otherwise,