Linus, Alan,
I don't know if you were following the 64-bit jiffies thread, but it came
out that initializing jiffies to a pre-wrap state causes system instability,
partly caused by failure to handle jiffies wrap properly.
I started checking a bunch of code for improper timer wrap handling and
thought I would add this patch to clarify the usage of the time_defore()
and time_after() macros so that it is clearer how to use them. I always
have to sit down and think about how to use these macros, so we may
as well make it clear in the comment. I expanded the time_before{_eq}
definitions as it makes it easier to see which one to use when existing
code to use one of the macros.
I also put a pointer to these macros at the jiffies definition, so people
are more likely to use them.
I will be sending patches for fixing up jiffies comparisons separately.
As the comment at the time_{before,after} macros says, I'm changing all
comparisons of jiffies values to use these macros (even if they are
correctly done) so that (a) it insulates driver code if we change to a
64-bit jiffies value, and (b) makes people more aware that they should
always use these macros when dealing with jiffies.
Cheers, Andreas
=========================================================================
--- linux/include/linux/timer.h.orig Thu Oct 25 10:02:34 2001
+++ linux/include/linux/timer.h Sat Nov 3 12:38:01 2001
@@ -53,20 +53,23 @@
}
/*
- * These inlines deal with timer wrapping correctly. You are
- * strongly encouraged to use them
+ * These inlines deal with timer wrapping correctly. You are
+ * strongly encouraged to use them:
* 1. Because people otherwise forget
* 2. Because if the timer wrap changes in future you wont have to
* alter your driver code.
*
+ * The macros are most commonly used in the form (unknown) after (known),
+ * or (unknown) before (known), where "unknown" is usually "jiffies".
+ *
* Do this with "<0" and ">=0" to only test the sign of the result. A
* good compiler would generate better code (and a really good compiler
* wouldn't care). Gcc is currently neither.
*/
-#define time_after(a,b) ((long)(b) - (long)(a) < 0)
-#define time_before(a,b) time_after(b,a)
+#define time_after(unknown,known) ((long)(known) - (long)(unknown) < 0)
+#define time_before(unknown,known) ((long)(unknown) - (long)(known) < 0)
-#define time_after_eq(a,b) ((long)(a) - (long)(b) >= 0)
-#define time_before_eq(a,b) time_after_eq(b,a)
+#define time_after_eq(unknown,known) ((long)(unknown) - (long)(known) >= 0)
+#define time_before_eq(unknown,known) ((long)(known) - (long)(unknown) >= 0)
#endif
--- linux/kernel/timer.c.orig Thu Oct 25 02:09:35 2001
+++ linux/kernel/timer.c Sat Nov 3 12:50:41 2001
@@ -65,6 +65,10 @@
extern int do_setitimer(int, struct itimerval *, struct itimerval *);
+/*
+ * Use the time_before{_eq}() and time_after{_eq}() macros when comparing
+ * two jiffies values to correctly handle timer wrap. See timer.h.
+ */
unsigned long volatile jiffies;
unsigned int * prof_buffer;
-- Andreas Dilger http://sourceforge.net/projects/ext2resize/ http://www-mddsp.enel.ucalgary.ca/People/adilger/- 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/
This archive was generated by hypermail 2b29 : Wed Nov 07 2001 - 21:00:26 EST