From: William Lee Irwin III <wli@holomorphy.com>
To: Andrew Morton <akpm@osdl.org>
Cc: Nick Sanders <sandersn@btinternet.com>, linux-kernel@vger.kernel.org
Subject: Re: 2.6.9-rc3-mm3
Date: Thu, 7 Oct 2004 04:40:40 -0700 [thread overview]
Message-ID: <20041007114040.GV9106@holomorphy.com> (raw)
In-Reply-To: <20041007025007.77ec1a44.akpm@osdl.org>
Nick Sanders <sandersn@btinternet.com> wrote:
>> I get the following oops when booting and it also stops kde
>> (artswrapper) from starting with the same call trace. USB seems to
>> be working which is good.
On Thu, Oct 07, 2004 at 02:50:07AM -0700, Andrew Morton wrote:
> Could you please do
> cd /usr/src/linux
> wget ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.9-rc3/2.6.9-rc3-mm3/broken-out/optimize-profile-path-slightly.patch
> patch -R -p1 < optimize-profile-path-slightly.patch
> and retest?
Here is a more likely correct patch for what that was trying to do,
however misguided that may be. Untested, uncompiled, vs. 2.6.9-rc3-mm3
without the bad patch:
Expose some variables from kernel/profile.c to the world to short-
circuit some checks in some inline wrappers for the purportedly
``heavier'' profiling functions. I took the liberty of declaring
some variables extern in the bodies of the inlines, which is rare
in the kernel, but protects against the use of the now-exposed
variables by things that shouldn't touch them.
Signed-off-by: William Irwin <wli@holomorphy.com>
Index: mm3-2.6.9-rc3/kernel/profile.c
===================================================================
--- mm3-2.6.9-rc3.orig/kernel/profile.c 2004-10-07 04:07:13.019223245 -0700
+++ mm3-2.6.9-rc3/kernel/profile.c 2004-10-07 04:38:08.299177821 -0700
@@ -34,10 +34,10 @@
#define NR_PROFILE_HIT (PAGE_SIZE/sizeof(struct profile_hit))
#define NR_PROFILE_GRP (NR_PROFILE_HIT/PROFILE_GRPSZ)
-static atomic_t *prof_buffer;
+atomic_t *prof_buffer;
static unsigned long prof_len, prof_shift;
-static int prof_on;
-static cpumask_t prof_cpu_mask = CPU_MASK_ALL;
+int prof_on;
+cpumask_t prof_cpu_mask = CPU_MASK_ALL;
#ifdef CONFIG_SMP
static DEFINE_PER_CPU(struct profile_hit *[2], cpu_profile_hits);
static DEFINE_PER_CPU(int, cpu_profile_flip);
@@ -284,14 +284,12 @@
up(&profile_flip_mutex);
}
-void profile_hit(int type, void *__pc)
+void fastcall __profile_hit(void *__pc)
{
unsigned long primary, secondary, flags, pc = (unsigned long)__pc;
int i, j, cpu;
struct profile_hit *hits;
- if (prof_on != type || !prof_buffer)
- return;
pc = min((pc - (unsigned long)_stext) >> prof_shift, prof_len - 1);
i = primary = (pc & (NR_PROFILE_GRP - 1)) << PROFILE_GRPSHIFT;
secondary = (~(pc << 1) & (NR_PROFILE_GRP - 1)) << PROFILE_GRPSHIFT;
@@ -392,14 +390,6 @@
}
#endif /* !CONFIG_SMP */
-void profile_tick(int type, struct pt_regs *regs)
-{
- if (type == CPU_PROFILING)
- profile_hook(regs);
- if (!user_mode(regs) && cpu_isset(smp_processor_id(), prof_cpu_mask))
- profile_hit(type, (void *)profile_pc(regs));
-}
-
#ifdef CONFIG_PROC_FS
#include <linux/proc_fs.h>
#include <asm/uaccess.h>
Index: mm3-2.6.9-rc3/include/linux/profile.h
===================================================================
--- mm3-2.6.9-rc3.orig/include/linux/profile.h 2004-10-07 04:03:39.960613077 -0700
+++ mm3-2.6.9-rc3/include/linux/profile.h 2004-10-07 04:28:46.333609573 -0700
@@ -8,6 +8,7 @@
#include <linux/init.h>
#include <linux/cpumask.h>
#include <asm/errno.h>
+#include <asm/atomic.h>
#define CPU_PROFILING 1
#define SCHED_PROFILING 2
@@ -17,8 +18,8 @@
/* init basic kernel profiler */
void __init profile_init(void);
-void profile_tick(int, struct pt_regs *);
-void profile_hit(int, void *);
+void FASTCALL(__profile_hit(void *));
+
#ifdef CONFIG_PROC_FS
void create_prof_cpu_mask(struct proc_dir_entry *);
#else
@@ -101,6 +102,26 @@
#endif /* CONFIG_PROFILING */
+static inline void profile_hit(int type, void *pc)
+{
+ extern int prof_on;
+ extern atomic_t *prof_buffer;
+
+ if (prof_on == type && prof_buffer)
+ __profile_hit(pc);
+}
+
+static inline void profile_tick(int type, struct pt_regs *regs)
+{
+ extern cpumask_t prof_cpu_mask;
+
+ if (type != CPU_PROFILING)
+ return;
+ profile_hook(regs);
+ if (!user_mode(regs) && cpu_isset(smp_processor_id(), prof_cpu_mask))
+ profile_hit(type, (void *)profile_pc(regs));
+}
+
#endif /* __KERNEL__ */
#endif /* _LINUX_PROFILE_H */
next prev parent reply other threads:[~2004-10-07 11:41 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-10-07 8:51 2.6.9-rc3-mm3 Andrew Morton
2004-10-07 9:30 ` 2.6.9-rc3-mm3 Martin Zwickel
2004-10-07 9:41 ` 2.6.9-rc3-mm3 Nick Sanders
2004-10-07 9:50 ` 2.6.9-rc3-mm3 Andrew Morton
2004-10-07 10:10 ` 2.6.9-rc3-mm3 Martin Zwickel
2004-10-07 10:44 ` 2.6.9-rc3-mm3 Hanno Meyer-Thurow
[not found] ` <20041007125610.0019eb3d@mango.fruits.de>
2004-10-07 10:58 ` 2.6.9-rc3-mm3 Hanno Meyer-Thurow
2004-10-07 10:16 ` 2.6.9-rc3-mm3 Nick Sanders
2004-10-07 11:40 ` William Lee Irwin III [this message]
2004-10-07 21:25 ` 2.6.9-rc3-mm3 J.A. Magallon
2004-10-07 21:46 ` 2.6.9-rc3-mm3 J.A. Magallon
2004-10-07 22:07 ` 2.6.9-rc3-mm3 Andrew Morton
2004-10-07 22:41 ` 2.6.9-rc3-mm3 J.A. Magallon
2004-10-07 22:54 ` 2.6.9-rc3-mm3 Andrew Morton
2004-10-07 23:06 ` 2.6.9-rc3-mm3 J.A. Magallon
2004-10-07 23:40 ` 2.6.9-rc3-mm3 Greg KH
2004-10-08 1:05 ` 2.6.9-rc3-mm3 Paul Fulghum
2004-10-08 11:23 ` 2.6.9-rc3-mm3 William Lee Irwin III
2004-10-08 11:34 ` 2.6.9-rc3-mm3 Ryan Cumming
2004-10-08 7:00 ` 2.6.9-rc3-mm3 Ingo Molnar
2004-10-07 22:03 ` 2.6.9-rc3-mm3 William Lee Irwin III
2004-10-07 9:59 ` 2.6.9-rc3-mm3 Martin Zwickel
2004-10-07 10:04 ` 2.6.9-rc3-mm3 Jesper Juhl
2004-10-07 10:15 ` 2.6.9-rc3-mm3 Russell King
2004-10-07 16:58 ` 2.6.9-rc3-mm3: `risc_code_addr01' multiple definition Adrian Bunk
2004-10-07 17:29 ` Andrew Vasquez
2004-10-07 17:50 ` James Bottomley
2004-10-07 18:13 ` Andrew Vasquez
2004-10-07 18:15 ` Andrew Vasquez
2004-10-07 18:30 ` Andrew Vasquez
2004-10-21 13:02 ` Jes Sorensen
2004-10-07 17:08 ` 2.6.9-rc3-mm3 (compile stats) John Cherry
2004-10-07 20:56 ` 2.6.9-rc3-mm3 Joel Schopp
2004-10-07 21:33 ` 2.6.9-rc3-mm3 Andrew Morton
2004-10-07 21:01 ` 2.6.9-rc3-mm3: build problem on dual-Opteron w/ NUMA Rafael J. Wysocki
2004-10-07 21:40 ` Andrew Morton
2004-10-07 21:46 ` Andi Kleen
2004-10-08 10:33 ` 2.6.9-rc3-mm3: vm-thrashing-control-tuning Hideo AOKI
2004-10-08 20:49 ` Andrew Morton
-- strict thread matches above, loose matches on Subject: below --
2004-10-07 11:10 2.6.9-rc3-mm3 Oleg Nesterov
2004-10-07 18:18 ` 2.6.9-rc3-mm3 Andrew Morton
[not found] <20041007015139.6f5b833b.akpm@osdl.org.suse.lists.linux.kernel>
[not found] ` <200410071041.20723.sandersn@btinternet.com.suse.lists.linux.kernel>
[not found] ` <20041007025007.77ec1a44.akpm@osdl.org.suse.lists.linux.kernel>
2004-10-07 11:23 ` 2.6.9-rc3-mm3 Andi Kleen
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20041007114040.GV9106@holomorphy.com \
--to=wli@holomorphy.com \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.org \
--cc=sandersn@btinternet.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.