2009-07-06 11:33:18

by Jaswinder Singh Rajput

[permalink] [raw]
Subject: [PATCH -tip] x86: thread_info.h moving comment to where it should be


By mistake commit 2052e8d40ad58 moved following comment to wrong place,
where it does not make any sense :

"preempt_count needs to be 1 initially, until the scheduler is functional."

Moving comment back to right place where preempt_count is setting to 1

Signed-off-by: Jaswinder Singh Rajput <[email protected]>
---
arch/x86/include/asm/thread_info.h | 8 ++++----
1 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h
index b078352..2dda0f8 100644
--- a/arch/x86/include/asm/thread_info.h
+++ b/arch/x86/include/asm/thread_info.h
@@ -43,6 +43,9 @@ struct thread_info {
int uaccess_err;
};

+/*
+ * preempt_count needs to be 1 initially, until the scheduler is functional.
+ */
#define INIT_THREAD_INFO(tsk) \
{ \
.task = &tsk, \
@@ -167,14 +170,12 @@ struct thread_info {
#ifdef CONFIG_X86_32

#define STACK_WARN (THREAD_SIZE/8)
+
/*
* macros/functions for gaining access to the thread information structure
- *
- * preempt_count needs to be 1 initially, until the scheduler is functional.
*/
#ifndef __ASSEMBLY__

-
/* how to get the current stack pointer from C */
register unsigned long current_stack_pointer asm("esp") __used;

@@ -205,7 +206,6 @@ static inline struct thread_info *current_thread_info(void)

/*
* macros/functions for gaining access to the thread information structure
- * preempt_count needs to be 1 initially, until the scheduler is functional.
*/
#ifndef __ASSEMBLY__
DECLARE_PER_CPU(unsigned long, kernel_stack);
--
1.6.0.6



2009-07-06 18:56:22

by Thomas Gleixner

[permalink] [raw]
Subject: Re: [PATCH -tip] x86: thread_info.h moving comment to where it should be

On Mon, 6 Jul 2009, Jaswinder Singh Rajput wrote:
>
> By mistake commit 2052e8d40ad58 moved following comment to wrong place,
> where it does not make any sense :

Wrong. commit 2052e8d4 had the comments in the right place. commit
3351cc03 replaced the two identical INIT_THREAD_INFO macros and did
not update the comments.

> #define STACK_WARN (THREAD_SIZE/8)
> +

How is that related ?

> /*
> * macros/functions for gaining access to the thread information structure
> - *
> - * preempt_count needs to be 1 initially, until the scheduler is functional.
> */
> #ifndef __ASSEMBLY__
>
> -

Ditto.

Thanks,

tglx

2009-07-06 21:06:18

by Jaswinder Singh Rajput

[permalink] [raw]
Subject: [PATCH -tip V2] x86: thread_info.h moving comment to where it should be

On Mon, 2009-07-06 at 20:54 +0200, Thomas Gleixner wrote:
> On Mon, 6 Jul 2009, Jaswinder Singh Rajput wrote:
> >
> > By mistake commit 2052e8d40ad58 moved following comment to wrong place,
> > where it does not make any sense :
>
> Wrong. commit 2052e8d4 had the comments in the right place. commit
> 3351cc03 replaced the two identical INIT_THREAD_INFO macros and did
> not update the comments.
>

Ahh so you was also part of it, here is updated patch:

[PATCH -tip] x86: thread_info.h moving comment to where it should be

By mistake commit 3351cc03 forget to move following comment along
with INIT_THREAD_INFO :

"preempt_count needs to be 1 initially, until the scheduler is functional."

Moving comment back to right place where preempt_count is setting to 1

Also by mistake commit 2052e8d4 forget to fix extra line which is not required.
Removed that extra line.

Signed-off-by: Jaswinder Singh Rajput <[email protected]>
---
arch/x86/include/asm/thread_info.h | 7 +++----
1 files changed, 3 insertions(+), 4 deletions(-)

diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h
index b078352..1131636 100644
--- a/arch/x86/include/asm/thread_info.h
+++ b/arch/x86/include/asm/thread_info.h
@@ -43,6 +43,9 @@ struct thread_info {
int uaccess_err;
};

+/*
+ * preempt_count needs to be 1 initially, until the scheduler is functional
+ */
#define INIT_THREAD_INFO(tsk) \
{ \
.task = &tsk, \
@@ -169,12 +172,9 @@ struct thread_info {
#define STACK_WARN (THREAD_SIZE/8)
/*
* macros/functions for gaining access to the thread information structure
- *
- * preempt_count needs to be 1 initially, until the scheduler is functional.
*/
#ifndef __ASSEMBLY__

-
/* how to get the current stack pointer from C */
register unsigned long current_stack_pointer asm("esp") __used;

@@ -205,7 +205,6 @@ static inline struct thread_info *current_thread_info(void)

/*
* macros/functions for gaining access to the thread information structure
- * preempt_count needs to be 1 initially, until the scheduler is functional.
*/
#ifndef __ASSEMBLY__
DECLARE_PER_CPU(unsigned long, kernel_stack);
--
1.6.0.6

2009-07-06 22:46:14

by Thomas Gleixner

[permalink] [raw]
Subject: Re: [PATCH -tip V2] x86: thread_info.h moving comment to where it should be

Jaswinder,

On Tue, 7 Jul 2009, Jaswinder Singh Rajput wrote:
> On Mon, 2009-07-06 at 20:54 +0200, Thomas Gleixner wrote:
> > On Mon, 6 Jul 2009, Jaswinder Singh Rajput wrote:
> > >
> > > By mistake commit 2052e8d40ad58 moved following comment to wrong place,
> > > where it does not make any sense :
> >
> > Wrong. commit 2052e8d4 had the comments in the right place. commit
> > 3351cc03 replaced the two identical INIT_THREAD_INFO macros and did
> > not update the comments.
> >
>
> Ahh so you was also part of it, here is updated patch:

if you expect that you earn more trust by such snotty comments, then
you are really on the wrong track. To keep the record straight:

_You_ sent a patch with a bogus patch description.

Instead of silently sending a fixed up patch after I pointed out to
you that your commit log is incorrect you come back and make pompous
comments about who is doing wrong and who has taken part of it.

> [PATCH -tip] x86: thread_info.h moving comment to where it should be
>
> By mistake commit 3351cc03 forget to move following comment along
> with INIT_THREAD_INFO :
>
> "preempt_count needs to be 1 initially, until the scheduler is functional."
>
> Moving comment back to right place where preempt_count is setting to 1
>
> Also by mistake commit 2052e8d4 forget to fix extra line which is not required.
> Removed that extra line.

That changelog is just another proof of your attempts to make a
moutain out of a molehill:

It does not matter at all which commit did not move a comment and it
does even matter less which commit did not remove or added a blank
line. For this patch a commit message consisting of a single subject
line:

x86: thread_info.h: move comment and remove stray newline

is sufficient and more useful than your "by mistake ..."
fingerpointing.

We care about those details with _real_ code bugs, but not for the
sake of pointing at the person who made a mistake. We simply care
because we want a reference to the context where and why it happened.

Thanks,

tglx

2009-07-07 05:23:00

by Jaswinder Singh Rajput

[permalink] [raw]
Subject: Re: [PATCH -tip V2] x86: thread_info.h moving comment to where it should be

On Tue, 2009-07-07 at 00:44 +0200, Thomas Gleixner wrote:
> Jaswinder,
>
> On Tue, 7 Jul 2009, Jaswinder Singh Rajput wrote:
> > On Mon, 2009-07-06 at 20:54 +0200, Thomas Gleixner wrote:
> > > On Mon, 6 Jul 2009, Jaswinder Singh Rajput wrote:
> > > >
> > > > By mistake commit 2052e8d40ad58 moved following comment to wrong place,
> > > > where it does not make any sense :
> > >
> > > Wrong. commit 2052e8d4 had the comments in the right place. commit
> > > 3351cc03 replaced the two identical INIT_THREAD_INFO macros and did
> > > not update the comments.
> > >
> >
> > Ahh so you was also part of it, here is updated patch:
>
> if you expect that you earn more trust by such snotty comments, then
> you are really on the wrong track. To keep the record straight:
>

You started it and when it comes to you why you start jumping.

> _You_ sent a patch with a bogus patch description.
>

Have you applied it or you only shouts like always.

--
JSR

2009-07-07 16:16:42

by Christoph Lameter

[permalink] [raw]
Subject: Re: [PATCH -tip V2] x86: thread_info.h moving comment to where it should be

On Tue, 7 Jul 2009, Jaswinder Singh Rajput wrote:

> Have you applied it or you only shouts like always.

Could we reduce the noise please (that applies bto both of you)?

If it helps: Blame me (since I am less involved now) but please get it
fixed.