2.4.20-aa1 questions.
what do the following patches actually *fix*?
00_backout-gcc-3_0-patch-1
00_gcc-30-volatile-xtime-1
I'm trying to get 2.4.20 patched up by using the -aa split patches for 2.4.20 and I'm incorporating only the things I want, but I use gcc 3.2 for compiling, and these confused me a bit.
Signature · 2 lines
Austin Gonyou <[email protected]> Coremetrics, Inc.
Quoted 8 lines · Austin Gonyou
what do the following patches actually *fix*?
00_backout-gcc-3_0-patch-1
00_gcc-30-volatile-xtime-1I'm trying to get 2.4.20 patched up by using the -aa split patches for 2.4.20 and I'm incorporating only the things I want, but I use gcc 3.2 for compiling, and these confused me a bit.
Oooh, I had lengthy discussion with andrea on those two. These patches are a) grossly misnamed and b) should be one. They change xtime to a volatile because andrea thinks that's safer.
The background on the silly naming is that earlier 2.4 kernels had xtime not volatile but the prototype (or vice versa) and gcc3 didn't like that.
So the best idea would be to merge them into 00_xtime_volatile-1 if you want to keep them.
Quoted 10 lines · Christoph Hellwig
Quoted 8 lines · Austin Gonyou
what do the following patches actually *fix*?
00_backout-gcc-3_0-patch-1
00_gcc-30-volatile-xtime-1I'm trying to get 2.4.20 patched up by using the -aa split patches for 2.4.20 and I'm incorporating only the things I want, but I use gcc 3.2 for compiling, and these confused me a bit.
Quoted 6 lines
Oooh, I had lengthy discussion with andrea on those two. These patches are a) grossly misnamed and b) should be one. They change xtime to a volatile because andrea thinks that's safer.
The background on the silly naming is that earlier 2.4 kernels had xtime not volatile but the prototype (or vice versa) and gcc3 didn't like that.
Ahh...kewl. thanks much. I was thinking that, but I was very confused by the naming versus the code. (not that I was *that* sure of the code anyway, but you know.) :)
Quoted 2 lines
So the best idea would be to merge them into 00_xtime_volatile-1 if you want to keep them.
Ok...that sounds like a plan. Thanks so much.
Signature · 1 line
GrandMasterLee <[email protected]>
PATCH 1/1 00_xtime_volatile-1
On Tue, 2002-12-03 at 15:01, GrandMasterLee wrote:
[...]
Quoted 9 lines
Quoted 2 lines
The background on the silly naming is that earlier 2.4 kernels had xtime not volatile but the prototype (or vice versa) and gcc3 didn't like that.
Ahh...kewl. thanks much. I was thinking that, but I was very confused by the naming versus the code. (not that I was *that* sure of the code anyway, but you know.) :)
Quoted 4 lines
Quoted 2 lines
So the best idea would be to merge them into if
you want to keep them.Ok...that sounds like a plan. Thanks so much.
Here is the patch. I think it should be OK. I just took the two questioned patches and applied them to 2.4.20, then diffed against a pristine 2.4.20.
--The GrandMaster
<[email protected]>
include/linux/sched.h +1 -1
@@ -582,7 +582,7 @@extern unsigned long volatile jiffies;extern unsigned long itimer_ticks;extern unsigned long itimer_next; -extern struct timeval xtime; +extern volatile struct timeval xtime;extern void do_timer (struct pt_regs *);extern unsigned int * prof_buffer;
include/linux/sched.h.orig +959 -0
959 lines not on this page · raw message
kernel/timer.c +1 -1
@@ -32,7 +32,7 @@long tick = (1000000 +HZ /2 ) /HZ ;/* timer interrupt period */ /* The current time */ -struct timeval xtime __attribute__ ((aligned (16))); +volatile struct timeval xtime __attribute__ ((aligned (16)));/* Don't completely fail for HZ > 500. */ int tickadj =500 /HZ ? :1 ;/* microsecs */