[Xenomai-core] [PATCH] bug in xnintr_irq_proc()

[Xenomai-core] [PATCH] bug in xnintr_irq_proc()

[Wolfgang Grandegger]

[-- Attachment #1: Type: text/plain, Size: 171 bytes --]

Hello,

attached is a fix for a nice bug in xnintr_irq_proc() of Xenomai v2.3.x.
Check for tabs. I have not checked if it's present in recent versions as
well.

Wolfgang.

[-- Attachment #2: xenomai-2.3.x-irq-proc.patch --]
[-- Type: text/x-patch, Size: 837 bytes --]

Index: ksrc/nucleus/intr.c
===================================================================
--- ksrc/nucleus/intr.c	(revision 3441)
+++ ksrc/nucleus/intr.c	(working copy)
@@ -836,7 +836,7 @@ int xnintr_irq_proc(unsigned int irq, ch
 #if defined(CONFIG_XENO_OPT_SHIRQ_LEVEL) || defined(CONFIG_XENO_OPT_SHIRQ_EDGE)
 	intr = xnirqs[irq].handlers;
 	if (intr) {
-		strcpy(p, "	   "); p += 8;
+		strcpy(p, "	   "); p += 4;
 
 		do {
 			*p = ' '; p += 1;
@@ -848,7 +848,7 @@ int xnintr_irq_proc(unsigned int irq, ch
 #else /* !CONFIG_XENO_OPT_SHIRQ_LEVEL && !CONFIG_XENO_OPT_SHIRQ_EDGE */
 	intr = rthal_irq_cookie(&rthal_domain, irq);
 	if (intr) {
-		strcpy(p, "	    "); p += 9;
+		strcpy(p, "	    "); p += 5;
 		strcpy(p, intr->name); p += strlen(intr->name);
 	}
 #endif /* CONFIG_XENO_OPT_SHIRQ_LEVEL || CONFIG_XENO_OPT_SHIRQ_EDGE */

Re: [Xenomai-core] [PATCH] bug in xnintr_irq_proc()

[Philippe Gerum]

Wolfgang Grandegger wrote:
> Hello,
> 
> attached is a fix for a nice bug in xnintr_irq_proc() of Xenomai v2.3.x.

Mmf... Thanks.

> Check for tabs. I have not checked if it's present in recent versions as
> well.

Even if a bit overkill, I would even go for something like
p += sprintf(p, "...
just to be sure we won't get hit again.

(or even snprintf() for the paranoïds among us).

> 
> Wolfgang.
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Xenomai-core mailing list
> Xenomai-core@domain.hid
> https://mail.gna.org/listinfo/xenomai-core


-- 
Philippe.

Re: [Xenomai-core] [PATCH] bug in xnintr_irq_proc()

[Wolfgang Grandegger]

Philippe Gerum wrote:
> Wolfgang Grandegger wrote:
>> Hello,
>>
>> attached is a fix for a nice bug in xnintr_irq_proc() of Xenomai v2.3.x.
> 
> Mmf... Thanks.
> 
>> Check for tabs. I have not checked if it's present in recent versions as
>> well.
> 
> Even if a bit overkill, I would even go for something like
> p += sprintf(p, "...
> just to be sure we won't get hit again.

Yes, and replacing the tabs with "\t" in case we need them. They seem to
be gone in the trunk.

> (or even snprintf() for the paranoïds among us).

Definitely, especially because the name field of xnintr_t is a pointer.

Wolfgang.