Skip to contentSkip to conversation

x86: fix arch/x86/lib/clear_page_64.S::clear_page_c annotation

PATCH In review Mike Galbraith 3 participants 5 messages Jun 30, 2009 x86

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 x86: fix arch/x86/lib/clear_page_64.S::clear_page_c annotation

x86: fix arch/x86/lib/clear_page_64.S::clear_page_c annotation.

Cc Ingo Molnar <[email protected]>
Cc Alexander van Heukelum <[email protected]>
Cc Cyrill Gorcunov <[email protected]>

LKML-Reference: <new-submission>

---
 arch/x86/lib/clear_page_64.S |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/lib/clear_page_64.S +2 -2
Index: linux-2.6/arch/x86/lib/clear_page_64.S
===================================================================
@@ -6,14 +6,14 @@
  * rdi	page
  */			
 	ALIGN
-clear_page_c:
+ENTRY(clear_page_c)
 	CFI_STARTPROC
 	movl $4096/8,%ecx
 	xorl %eax,%eax
 	rep stosq
 	ret
 	CFI_ENDPROC
-ENDPROC(clear_page)
+ENDPROC(clear_page_c)
 
 ENTRY(clear_page)
 	CFI_STARTPROC
Acked-by on v1 1/1
#1 lore raw

Hi Mike,

I agree with the code, but I do have a remark...

The ENTRY() macro already includes the ALIGN, so you could get rid of the preceding ALIGN. But it's just a neatness thing, because two consecutive ALIGNS don't change the resulting binary.

At any rate:

(maybe the x86-committer feels comfortable enough to remove the ALIGN ;) )

Quoted 35 lines · Mike Galbraith

x86: fix arch/x86/lib/clear_page_64.S::clear_page_c annotation.

Cc Ingo Molnar <[email protected]>
Cc Alexander van Heukelum <[email protected]>
Cc Cyrill Gorcunov <[email protected]>

LKML-Reference: <new-submission>

---
 arch/x86/lib/clear_page_64.S |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Index: linux-2.6/arch/x86/lib/clear_page_64.S
===================================================================

arch/x86/lib/clear_page_64.S +2 -2
@@ -6,14 +6,14 @@
  * rdi	page
  */			
 	ALIGN
-clear_page_c:
+ENTRY(clear_page_c)
 	CFI_STARTPROC
 	movl $4096/8,%ecx
 	xorl %eax,%eax
 	rep stosq
 	ret
 	CFI_ENDPROC
-ENDPROC(clear_page)
+ENDPROC(clear_page_c)
 
 ENTRY(clear_page)
 	CFI_STARTPROC
author Acked-by Alexander van Heukelum
#2 lore raw
Quoted 8 lines · Alexander van Heukelum

Hi Mike,

I agree with the code, but I do have a remark...

The ENTRY() macro already includes the ALIGN, so you could get rid of the preceding ALIGN. But it's just a neatness thing, because two consecutive ALIGNS don't change the resulting binary.

Quoted 6 lines

At any rate:

(maybe the x86-committer feels comfortable enough to remove the ALIGN ;) )

No doubt, but I can save him the trouble ;-)

x86: fix incorrect arch/x86/lib/clear_page_64.S::clear_page_c annotation.

LKML-Reference: <new-submission>

---
 arch/x86/lib/clear_page_64.S |    5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
arch/x86/lib/clear_page_64.S +2 -3
Index: linux-2.6/arch/x86/lib/clear_page_64.S
===================================================================
@@ -5,15 +5,14 @@
  * Zero a page. 	
  * rdi	page
  */			
-	ALIGN
-clear_page_c:
+ENTRY(clear_page_c)
 	CFI_STARTPROC
 	movl $4096/8,%ecx
 	xorl %eax,%eax
 	rep stosq
 	ret
 	CFI_ENDPROC
-ENDPROC(clear_page)
+ENDPROC(clear_page_c)
 
 ENTRY(clear_page)
 	CFI_STARTPROC
#3 lore raw

[Mike Galbraith - Tue, Jun 30, 2009 at 03:00:20PM +0200]

| On Tue, 2009-06-30 at 13:42 +0200, Alexander van Heukelum wrote:
| > Hi Mike,
| > 
| > I agree with the code, but I do have a remark...
| > 
| > The ENTRY() macro already includes the ALIGN, so you could get
| > rid of the preceding ALIGN. But it's just a neatness thing, because
| > two consecutive ALIGNS don't change the resulting binary.
| > 
| > At any rate:
| > 
| > Acked-by: Alexander van Heukelum <[email protected]>
| > 
| > (maybe the x86-committer feels comfortable enough to remove the ALIGN ;)
| > )
| 
| No doubt, but I can save him the trouble ;-)
| 
| x86: fix incorrect arch/x86/lib/clear_page_64.S::clear_page_c annotation.
| 
| Signed-off-by: Mike Galbraith <[email protected]>
| Acked-by: Alexander van Heukelum <[email protected]>
| Cc: Ingo Molnar <[email protected]>
| Cc: Cyrill Gorcunov <[email protected]>
| LKML-Reference: <new-submission>
| 
| ---
|  arch/x86/lib/clear_page_64.S |    5 ++---
|  1 file changed, 2 insertions(+), 3 deletions(-)
|

...

Thanks Mike, here is my Ack (if needed)

-- Cyrill

#4 lore raw

[tip:x86/urgent] x86: Fix symbol annotation for arch/x86/lib/clear_page_64.S::clear_page_c

In reply to Mike Galbraith
Commit-ID:  9e314996e3dc5189b9b36dce67088e882e989897
Gitweb:     http://git.kernel.org/tip/9e314996e3dc5189b9b36dce67088e882e989897
Author:     Mike Galbraith <[email protected]>
AuthorDate: Tue, 30 Jun 2009 15:00:20 +0200
Committer:  Ingo Molnar <[email protected]>
CommitDate: Tue, 30 Jun 2009 23:43:15 +0200

x86: Fix symbol annotation for arch/x86/lib/clear_page_64.S::clear_page_c

Noticed the zero-sized function symbol while looking at 'perf' profiles, it causes the profiler to display those addresses in hexa.

Turns out that this was wrong/bogus for an eternity.

LKML-Reference: <[email protected]>

---
 arch/x86/lib/clear_page_64.S |    5 ++---
 1 files changed, 2 insertions(+), 3 deletions(-)
arch/x86/lib/clear_page_64.S +2 -3
index 9a10a78..ebeafcc 100644
@@ -5,15 +5,14 @@
  * Zero a page. 	
  * rdi	page
  */			
-	ALIGN
-clear_page_c:
+ENTRY(clear_page_c)
 	CFI_STARTPROC
 	movl $4096/8,%ecx
 	xorl %eax,%eax
 	rep stosq
 	ret
 	CFI_ENDPROC
-ENDPROC(clear_page)
+ENDPROC(clear_page_c)
 
 ENTRY(clear_page)
 	CFI_STARTPROC