[GIT PULL] clean header untangling

[GIT PULL] clean header untangling

[Jeremy Fitzhardinge]

I cleaned up the untangle series.

The following changes since commit 9049a11de73d3ecc623f1903100d099f82ede56c:
  Jeremy Fitzhardinge (1):
        Merge commit 'remotes/tip/x86/paravirt' into x86/untangle2

are available in the git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git x86/untangle2

Jeremy Fitzhardinge (10):
      x86: rename *-defs.h to *-_types.h for consistency
      Split pgtable.h into pgtable_types.h and pgtable.h
      x86: Split pgtable_32.h into pgtable_32.h and pgtable_32_types.h
      x86: Split pgtable_64.h into pgtable_64_types.h and pgtable_64.h
      x86: Include pgtable_32|64_types.h in pgtable_types.h
      x86: create _types.h counterparts for page*.h
      x86: move 2 and 3 level asm-generic defs into page-defs
      x86: move defs around to allow paravirt.h to just include page_types.h
      x86: define pud_flags and pud_large properly to allow non-PAE builds
      x86: move pte types into pgtable*.h

 arch/x86/include/asm/page.h                        |  168 +----------
 arch/x86/include/asm/page_32.h                     |   89 +------
 arch/x86/include/asm/page_32_types.h               |   62 ++++
 arch/x86/include/asm/page_64.h                     |  101 +------
 arch/x86/include/asm/page_64.h.rej                 |  114 +++++++
 arch/x86/include/asm/page_64_types.h               |   91 ++++++
 arch/x86/include/asm/page_types.h                  |   64 ++++
 arch/x86/include/asm/paravirt.h                    |    2 +-
 arch/x86/include/asm/pgtable-2level-defs.h         |   20 --
 arch/x86/include/asm/pgtable-2level_types.h        |   35 +++
 ...gtable-3level-defs.h => pgtable-3level_types.h} |   18 ++
 arch/x86/include/asm/pgtable.h                     |  322 ++++----------------
 arch/x86/include/asm/pgtable_32.h                  |   42 +---
 arch/x86/include/asm/pgtable_32_types.h            |   46 +++
 arch/x86/include/asm/pgtable_64.h                  |   48 +---
 arch/x86/include/asm/pgtable_64_types.h            |   62 ++++
 arch/x86/include/asm/pgtable_types.h               |  321 +++++++++++++++++++
 arch/x86/include/asm/processor.h                   |    1 +
 18 files changed, 891 insertions(+), 715 deletions(-)
 create mode 100644 arch/x86/include/asm/page_32_types.h
 create mode 100644 arch/x86/include/asm/page_64.h.rej
 create mode 100644 arch/x86/include/asm/page_64_types.h
 create mode 100644 arch/x86/include/asm/page_types.h
 delete mode 100644 arch/x86/include/asm/pgtable-2level-defs.h
 create mode 100644 arch/x86/include/asm/pgtable-2level_types.h
 rename arch/x86/include/asm/{pgtable-3level-defs.h => pgtable-3level_types.h} (67%)
 create mode 100644 arch/x86/include/asm/pgtable_32_types.h
 create mode 100644 arch/x86/include/asm/pgtable_64_types.h
 create mode 100644 arch/x86/include/asm/pgtable_types.h

Re: [GIT PULL] clean header untangling

[Ingo Molnar]

* Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> I cleaned up the untangle series.
>
> The following changes since commit 9049a11de73d3ecc623f1903100d099f82ede56c:
>  Jeremy Fitzhardinge (1):
>        Merge commit 'remotes/tip/x86/paravirt' into x86/untangle2
>
> are available in the git repository at:
>
>  git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git x86/untangle2
>
> Jeremy Fitzhardinge (10):
>      x86: rename *-defs.h to *-_types.h for consistency
>      Split pgtable.h into pgtable_types.h and pgtable.h
>      x86: Split pgtable_32.h into pgtable_32.h and pgtable_32_types.h
>      x86: Split pgtable_64.h into pgtable_64_types.h and pgtable_64.h
>      x86: Include pgtable_32|64_types.h in pgtable_types.h
>      x86: create _types.h counterparts for page*.h
>      x86: move 2 and 3 level asm-generic defs into page-defs
>      x86: move defs around to allow paravirt.h to just include page_types.h
>      x86: define pud_flags and pud_large properly to allow non-PAE builds
>      x86: move pte types into pgtable*.h
>
> arch/x86/include/asm/page.h                        |  168 +----------
> arch/x86/include/asm/page_32.h                     |   89 +------
> arch/x86/include/asm/page_32_types.h               |   62 ++++
> arch/x86/include/asm/page_64.h                     |  101 +------
> arch/x86/include/asm/page_64.h.rej                 |  114 +++++++
> arch/x86/include/asm/page_64_types.h               |   91 ++++++
> arch/x86/include/asm/page_types.h                  |   64 ++++
> arch/x86/include/asm/paravirt.h                    |    2 +-
> arch/x86/include/asm/pgtable-2level-defs.h         |   20 --
> arch/x86/include/asm/pgtable-2level_types.h        |   35 +++
> ...gtable-3level-defs.h => pgtable-3level_types.h} |   18 ++
> arch/x86/include/asm/pgtable.h                     |  322 ++++----------------
> arch/x86/include/asm/pgtable_32.h                  |   42 +---
> arch/x86/include/asm/pgtable_32_types.h            |   46 +++
> arch/x86/include/asm/pgtable_64.h                  |   48 +---
> arch/x86/include/asm/pgtable_64_types.h            |   62 ++++
> arch/x86/include/asm/pgtable_types.h               |  321 +++++++++++++++++++
> arch/x86/include/asm/processor.h                   |    1 +
> 18 files changed, 891 insertions(+), 715 deletions(-)
> create mode 100644 arch/x86/include/asm/page_32_types.h
> create mode 100644 arch/x86/include/asm/page_64.h.rej
> create mode 100644 arch/x86/include/asm/page_64_types.h
> create mode 100644 arch/x86/include/asm/page_types.h
> delete mode 100644 arch/x86/include/asm/pgtable-2level-defs.h
> create mode 100644 arch/x86/include/asm/pgtable-2level_types.h
> rename arch/x86/include/asm/{pgtable-3level-defs.h => pgtable-3level_types.h} (67%)
> create mode 100644 arch/x86/include/asm/pgtable_32_types.h
> create mode 100644 arch/x86/include/asm/pgtable_64_types.h
> create mode 100644 arch/x86/include/asm/pgtable_types.h

Pulled into tip:x86/headers, thanks Jeremy!

	Ingo

[PATCH] x86 headers: remove duplicate pud_large() definition

[Ingo Molnar]

* Ingo Molnar <mingo@elte.hu> wrote:

> 
> * Jeremy Fitzhardinge <jeremy@goop.org> wrote:
> 
> > I cleaned up the untangle series.
> >
> > The following changes since commit 9049a11de73d3ecc623f1903100d099f82ede56c:
> >  Jeremy Fitzhardinge (1):
> >        Merge commit 'remotes/tip/x86/paravirt' into x86/untangle2
> >
> > are available in the git repository at:
> >
> >  git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git x86/untangle2
> >
> > Jeremy Fitzhardinge (10):
> >      x86: rename *-defs.h to *-_types.h for consistency
> >      Split pgtable.h into pgtable_types.h and pgtable.h
> >      x86: Split pgtable_32.h into pgtable_32.h and pgtable_32_types.h
> >      x86: Split pgtable_64.h into pgtable_64_types.h and pgtable_64.h
> >      x86: Include pgtable_32|64_types.h in pgtable_types.h
> >      x86: create _types.h counterparts for page*.h
> >      x86: move 2 and 3 level asm-generic defs into page-defs
> >      x86: move defs around to allow paravirt.h to just include page_types.h
> >      x86: define pud_flags and pud_large properly to allow non-PAE builds
> >      x86: move pte types into pgtable*.h
> >
> > arch/x86/include/asm/page.h                        |  168 +----------
> > arch/x86/include/asm/page_32.h                     |   89 +------
> > arch/x86/include/asm/page_32_types.h               |   62 ++++
> > arch/x86/include/asm/page_64.h                     |  101 +------
> > arch/x86/include/asm/page_64.h.rej                 |  114 +++++++
> > arch/x86/include/asm/page_64_types.h               |   91 ++++++
> > arch/x86/include/asm/page_types.h                  |   64 ++++
> > arch/x86/include/asm/paravirt.h                    |    2 +-
> > arch/x86/include/asm/pgtable-2level-defs.h         |   20 --
> > arch/x86/include/asm/pgtable-2level_types.h        |   35 +++
> > ...gtable-3level-defs.h => pgtable-3level_types.h} |   18 ++
> > arch/x86/include/asm/pgtable.h                     |  322 ++++----------------
> > arch/x86/include/asm/pgtable_32.h                  |   42 +---
> > arch/x86/include/asm/pgtable_32_types.h            |   46 +++
> > arch/x86/include/asm/pgtable_64.h                  |   48 +---
> > arch/x86/include/asm/pgtable_64_types.h            |   62 ++++
> > arch/x86/include/asm/pgtable_types.h               |  321 +++++++++++++++++++
> > arch/x86/include/asm/processor.h                   |    1 +
> > 18 files changed, 891 insertions(+), 715 deletions(-)
> > create mode 100644 arch/x86/include/asm/page_32_types.h
> > create mode 100644 arch/x86/include/asm/page_64.h.rej
> > create mode 100644 arch/x86/include/asm/page_64_types.h
> > create mode 100644 arch/x86/include/asm/page_types.h
> > delete mode 100644 arch/x86/include/asm/pgtable-2level-defs.h
> > create mode 100644 arch/x86/include/asm/pgtable-2level_types.h
> > rename arch/x86/include/asm/{pgtable-3level-defs.h => pgtable-3level_types.h} (67%)
> > create mode 100644 arch/x86/include/asm/pgtable_32_types.h
> > create mode 100644 arch/x86/include/asm/pgtable_64_types.h
> > create mode 100644 arch/x86/include/asm/pgtable_types.h
> 
> Pulled into tip:x86/headers, thanks Jeremy!

btw., i had to do these conflict resolutions:

    Conflicts:
        arch/x86/include/asm/page.h
        arch/x86/include/asm/pgtable.h
        arch/x86/mach-voyager/voyager_smp.c
        arch/x86/mm/fault.c

and this in kmemcheck:

    Merge branch 'x86/headers' into kmemcheck
    
    Conflicts:
        arch/x86/include/asm/pgtable.h
    
    Merged into:
    
        arch/x86/include/asm/pgtable_types.h
    
Plus the fixlet below.

	Ingo

------------------------->
>From 999c7880cc8eeb0cbe6610b8c6d0ab0ec51cd848 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Fri, 13 Feb 2009 13:15:55 +0100
Subject: [PATCH] x86 headers: remove duplicate pud_large() definition

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/include/asm/pgtable.h |    7 -------
 1 files changed, 0 insertions(+), 7 deletions(-)

diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h
index 62024ff..1c097a3 100644
--- a/arch/x86/include/asm/pgtable.h
+++ b/arch/x86/include/asm/pgtable.h
@@ -387,13 +387,6 @@ static inline unsigned long pages_to_mb(unsigned long npg)
 #define io_remap_pfn_range(vma, vaddr, pfn, size, prot)	\
 	remap_pfn_range(vma, vaddr, pfn, size, prot)
 
-#if PAGETABLE_LEVELS == 2
-static inline int pud_large(pud_t pud)
-{
-	return 0;
-}
-#endif
-
 #if PAGETABLE_LEVELS > 2
 static inline int pud_none(pud_t pud)
 {

[PATCH] x86 headers: include linux/types.h

[Ingo Molnar]

>From 56cefcea7c8769ffc04a5609e6ac849e36685468 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Fri, 13 Feb 2009 13:23:02 +0100
Subject: [PATCH] x86 headers: include linux/types.h

To properly pick up types relied on by prototypes like 'bool'.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/include/asm/page.h |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/arch/x86/include/asm/page.h b/arch/x86/include/asm/page.h
index 467ce69..89ed9d7 100644
--- a/arch/x86/include/asm/page.h
+++ b/arch/x86/include/asm/page.h
@@ -1,6 +1,8 @@
 #ifndef _ASM_X86_PAGE_H
 #define _ASM_X86_PAGE_H
 
+#include <linux/types.h>
+
 #ifdef __KERNEL__
 
 #include <asm/page_types.h>

Re: [PATCH] x86 headers: include linux/types.h

[Ingo Molnar]

Regarding the fixlet below - wouldnt PTE_FLAGS_MASK be better placed in
pgtable_types.h ? It's specifically related to the pte format. OTOH we
have a lot of PTE details in the page_*types.h files already.

The boundary between page_types.h and pgtable_types.h does not seem to
be clear to me - what is the principle you followed?

	Ingo

----------------------->
>From e43623b4ed1d0a1a86f0e05a2df8b9637b90ddd7 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Fri, 13 Feb 2009 13:24:19 +0100
Subject: [PATCH] x86 headers: include page_types.h in pgtable_types.h

To properly pick up details like PTE_FLAGS_MASK.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/include/asm/pgtable_types.h |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/arch/x86/include/asm/pgtable_types.h b/arch/x86/include/asm/pgtable_types.h
index a7452f1..9dafe87 100644
--- a/arch/x86/include/asm/pgtable_types.h
+++ b/arch/x86/include/asm/pgtable_types.h
@@ -2,6 +2,7 @@
 #define _ASM_X86_PGTABLE_DEFS_H
 
 #include <linux/const.h>
+#include <asm/page_types.h>
 
 #define FIRST_USER_ADDRESS	0
 

[PATCH] x86 headers: protect page_32.h via __ASSEMBLY__

[Ingo Molnar]

>From beb6943d8df7ce9278282101af4e0f6f7b648451 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Fri, 13 Feb 2009 13:36:47 +0100
Subject: [PATCH] x86 headers: protect page_32.h via __ASSEMBLY__

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/include/asm/page_32.h |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/x86/include/asm/page_32.h b/arch/x86/include/asm/page_32.h
index b3f0bf7..da4e762 100644
--- a/arch/x86/include/asm/page_32.h
+++ b/arch/x86/include/asm/page_32.h
@@ -3,6 +3,8 @@
 
 #include <asm/page_32_types.h>
 
+#ifndef __ASSEMBLY__
+
 #ifdef CONFIG_HUGETLB_PAGE
 #define HAVE_ARCH_HUGETLB_UNMAPPED_AREA
 #endif
@@ -19,8 +21,6 @@ extern unsigned long __phys_addr(unsigned long);
 #define pfn_valid(pfn)		((pfn) < max_mapnr)
 #endif /* CONFIG_FLATMEM */
 
-#ifndef __ASSEMBLY__
-
 #ifdef CONFIG_X86_USE_3DNOW
 #include <asm/mmx.h>
 

Re: [PATCH] x86 headers: include linux/types.h

[Jeremy Fitzhardinge]

Ingo Molnar wrote:
> Regarding the fixlet below - wouldnt PTE_FLAGS_MASK be better placed in
> pgtable_types.h ? It's specifically related to the pte format. OTOH we
> have a lot of PTE details in the page_*types.h files already.
>   

No, you're right.

> The boundary between page_types.h and pgtable_types.h does not seem to
> be clear to me - what is the principle you followed?
>   

Well, basically anything to do with the internals of a pte*_t go in 
pgtable.h.  PTE_*_MASK should go there too, and is an oversight (though 
its on the margin because of its dependency on the page size/mask 
definitions in page*.h).

    J

Re: [PATCH] x86 headers: remove duplicate pud_large() definition

[Jeremy Fitzhardinge]

Ingo Molnar wrote:
>         arch/x86/mach-voyager/voyager_smp.c
>         arch/x86/mm/fault.c
>   

I'm not sure where these two came from.  I certainly didn't touch those 
files; they must have diverged somewhere else along their history.

> and this in kmemcheck:
>
>     Merge branch 'x86/headers' into kmemcheck
>     
>     Conflicts:
>         arch/x86/include/asm/pgtable.h
>     
>     Merged into:
>     
>         arch/x86/include/asm/pgtable_types.h
>     
> Plus the fixlet below.
>
> 	Ingo
>
> ------------------------->
> From 999c7880cc8eeb0cbe6610b8c6d0ab0ec51cd848 Mon Sep 17 00:00:00 2001
> From: Ingo Molnar <mingo@elte.hu>
> Date: Fri, 13 Feb 2009 13:15:55 +0100
> Subject: [PATCH] x86 headers: remove duplicate pud_large() definition
>
> Signed-off-by: Ingo Molnar <mingo@elte.hu>
> ---
>  arch/x86/include/asm/pgtable.h |    7 -------
>  1 files changed, 0 insertions(+), 7 deletions(-)
>
> diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h
> index 62024ff..1c097a3 100644
> --- a/arch/x86/include/asm/pgtable.h
> +++ b/arch/x86/include/asm/pgtable.h
> @@ -387,13 +387,6 @@ static inline unsigned long pages_to_mb(unsigned long npg)
>  #define io_remap_pfn_range(vma, vaddr, pfn, size, prot)	\
>  	remap_pfn_range(vma, vaddr, pfn, size, prot)
>  
> -#if PAGETABLE_LEVELS == 2
> -static inline int pud_large(pud_t pud)
> -{
> -	return 0;
> -}
> -#endif
>   

OK.

    J

Re: [PATCH] x86 headers: remove duplicate pud_large() definition

[Ingo Molnar]

* Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> Ingo Molnar wrote:
>>         arch/x86/mach-voyager/voyager_smp.c
>>         arch/x86/mm/fault.c
>>   
>
> I'm not sure where these two came from.  I certainly didn't touch those  
> files; they must have diverged somewhere else along their history.

yep, it's drive-by conflicts. Not particularly interesting either as
those files are dead right now.

	Ingo

Re: [PATCH] x86 headers: include linux/types.h

[Ingo Molnar]

* Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> Ingo Molnar wrote:
>> Regarding the fixlet below - wouldnt PTE_FLAGS_MASK be better placed in
>> pgtable_types.h ? It's specifically related to the pte format. OTOH we
>> have a lot of PTE details in the page_*types.h files already.
>>   
>
> No, you're right.
>
>> The boundary between page_types.h and pgtable_types.h does not seem to
>> be clear to me - what is the principle you followed?
>>   
>
> Well, basically anything to do with the internals of a pte*_t go in  
> pgtable.h.  PTE_*_MASK should go there too, and is an oversight (though  
> its on the margin because of its dependency on the page size/mask  
> definitions in page*.h).

likewise, PUD_PAGE_* and PAGETABLE_LEVELS should move to pgtable_types.h
too i guess. Same goes for *MASK_SHIFT*.

	Ingo

Re: [PATCH] x86 headers: protect page_32.h via __ASSEMBLY__

[Jeremy Fitzhardinge]

Ingo Molnar wrote:
> From beb6943d8df7ce9278282101af4e0f6f7b648451 Mon Sep 17 00:00:00 2001
> From: Ingo Molnar <mingo@elte.hu>
> Date: Fri, 13 Feb 2009 13:36:47 +0100
> Subject: [PATCH] x86 headers: protect page_32.h via __ASSEMBLY__
>   

I think this is wrong in principle.  There should never be a reason to 
include a non-_types.h into asm.

(Fine as a build fix, of course.)

    J

> Signed-off-by: Ingo Molnar <mingo@elte.hu>
> ---
>  arch/x86/include/asm/page_32.h |    4 ++--
>  1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/arch/x86/include/asm/page_32.h b/arch/x86/include/asm/page_32.h
> index b3f0bf7..da4e762 100644
> --- a/arch/x86/include/asm/page_32.h
> +++ b/arch/x86/include/asm/page_32.h
> @@ -3,6 +3,8 @@
>  
>  #include <asm/page_32_types.h>
>  
> +#ifndef __ASSEMBLY__
> +
>  #ifdef CONFIG_HUGETLB_PAGE
>  #define HAVE_ARCH_HUGETLB_UNMAPPED_AREA
>  #endif
> @@ -19,8 +21,6 @@ extern unsigned long __phys_addr(unsigned long);
>  #define pfn_valid(pfn)		((pfn) < max_mapnr)
>  #endif /* CONFIG_FLATMEM */
>  
> -#ifndef __ASSEMBLY__
> -
>  #ifdef CONFIG_X86_USE_3DNOW
>  #include <asm/mmx.h>
>  
>
>   

Re: [PATCH] x86 headers: protect page_32.h via __ASSEMBLY__

[Ingo Molnar]

* Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> Ingo Molnar wrote:
>> From beb6943d8df7ce9278282101af4e0f6f7b648451 Mon Sep 17 00:00:00 2001
>> From: Ingo Molnar <mingo@elte.hu>
>> Date: Fri, 13 Feb 2009 13:36:47 +0100
>> Subject: [PATCH] x86 headers: protect page_32.h via __ASSEMBLY__
>>   
>
> I think this is wrong in principle.  There should never be a reason to  
> include a non-_types.h into asm.
>
> (Fine as a build fix, of course.)

Agreed - but it needs fixes to the .S files, these assembly files
include pgtable.h or page.h:

arch/x86/boot/compressed/head_32.S
arch/x86/boot/compressed/head_64.S
arch/x86/boot/header.S
arch/x86/kernel/acpi/realmode/wakeup.S
arch/x86/kernel/acpi/wakeup_32.S
arch/x86/kernel/acpi/wakeup_64.S
arch/x86/kernel/efi_stub_32.S
arch/x86/kernel/entry_32.S
arch/x86/kernel/entry_64.S
arch/x86/kernel/head_32.S
arch/x86/kernel/head_64.S
arch/x86/kernel/relocate_kernel_32.S
arch/x86/kernel/relocate_kernel_64.S
arch/x86/kernel/trampoline_32.S
arch/x86/kernel/trampoline_64.S
arch/x86/kernel/vmlinux_32.lds.S
arch/x86/kernel/vmlinux_64.lds.S
arch/x86/lib/getuser.S
arch/x86/power/hibernate_asm_32.S
arch/x86/power/hibernate_asm_64.S
arch/x86/xen/xen-head.S

	Ingo

Re: [PATCH] x86 headers: include linux/types.h

[Jeremy Fitzhardinge]

Ingo Molnar wrote:
> From 56cefcea7c8769ffc04a5609e6ac849e36685468 Mon Sep 17 00:00:00 2001
> From: Ingo Molnar <mingo@elte.hu>
> Date: Fri, 13 Feb 2009 13:23:02 +0100
> Subject: [PATCH] x86 headers: include linux/types.h
>
> To properly pick up types relied on by prototypes like 'bool'.
>   

linux/types.h is not assembler-safe.

    J

Re: [PATCH] x86 headers: protect page_32.h via __ASSEMBLY__

[Jeremy Fitzhardinge]

Ingo Molnar wrote:
> * Jeremy Fitzhardinge <jeremy@goop.org> wrote:
>
>   
>> Ingo Molnar wrote:
>>     
>>> From beb6943d8df7ce9278282101af4e0f6f7b648451 Mon Sep 17 00:00:00 2001
>>> From: Ingo Molnar <mingo@elte.hu>
>>> Date: Fri, 13 Feb 2009 13:36:47 +0100
>>> Subject: [PATCH] x86 headers: protect page_32.h via __ASSEMBLY__
>>>   
>>>       
>> I think this is wrong in principle.  There should never be a reason to  
>> include a non-_types.h into asm.
>>
>> (Fine as a build fix, of course.)
>>     
>
> Agreed - but it needs fixes to the .S files, these assembly files
> include pgtable.h or page.h:
>   

Piecemeal, or one sweep?

    J

Re: [PATCH] x86 headers: include linux/types.h

[Ingo Molnar]

* Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> Ingo Molnar wrote:
>> From 56cefcea7c8769ffc04a5609e6ac849e36685468 Mon Sep 17 00:00:00 2001
>> From: Ingo Molnar <mingo@elte.hu>
>> Date: Fri, 13 Feb 2009 13:23:02 +0100
>> Subject: [PATCH] x86 headers: include linux/types.h
>>
>> To properly pick up types relied on by prototypes like 'bool'.
>
> linux/types.h is not assembler-safe.

it is in tip:master, due to:

 0fb807c: unconditionally include asm/types.h from linux/types.h
 527bdfe: make linux/types.h as assembly safe

	Ingo

Re: [PATCH] x86 headers: protect page_32.h via __ASSEMBLY__

[Ingo Molnar]

* Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> Ingo Molnar wrote:
>> * Jeremy Fitzhardinge <jeremy@goop.org> wrote:
>>
>>   
>>> Ingo Molnar wrote:
>>>     
>>>> From beb6943d8df7ce9278282101af4e0f6f7b648451 Mon Sep 17 00:00:00 2001
>>>> From: Ingo Molnar <mingo@elte.hu>
>>>> Date: Fri, 13 Feb 2009 13:36:47 +0100
>>>> Subject: [PATCH] x86 headers: protect page_32.h via __ASSEMBLY__
>>>>         
>>> I think this is wrong in principle.  There should never be a reason 
>>> to  include a non-_types.h into asm.
>>>
>>> (Fine as a build fix, of course.)
>>>     
>>
>> Agreed - but it needs fixes to the .S files, these assembly files
>> include pgtable.h or page.h:
>>   
>
> Piecemeal, or one sweep?

i think the change itself should be in one commit, it's rather trivial.

	Ingo

Re: [PATCH] x86 headers: include linux/types.h

[Jaswinder Singh Rajput]

On Sat, Feb 14, 2009 at 12:22 AM, Jeremy Fitzhardinge <jeremy@goop.org> wrote:
> Ingo Molnar wrote:
>>
>> From 56cefcea7c8769ffc04a5609e6ac849e36685468 Mon Sep 17 00:00:00 2001
>> From: Ingo Molnar <mingo@elte.hu>
>> Date: Fri, 13 Feb 2009 13:23:02 +0100
>> Subject: [PATCH] x86 headers: include linux/types.h
>>
>> To properly pick up types relied on by prototypes like 'bool'.
>>
>
> linux/types.h is not assembler-safe.
>

AFAIK -tip's linux/types.h is assembly safe ;-)

JSR

Re: [PATCH] x86 headers: include linux/types.h

[Jeremy Fitzhardinge]

Ingo Molnar wrote:
> it is in tip:master, due to:
>
>  0fb807c: unconditionally include asm/types.h from linux/types.h
>  527bdfe: make linux/types.h as assembly safe
>   
OK, but that change isn't in x86/headers, so it won't compile standalone.

    J
> 	Ingo
>   

Re: [PATCH] x86 headers: include linux/types.h

[Ingo Molnar]

* Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> Ingo Molnar wrote:
>> it is in tip:master, due to:
>>
>>  0fb807c: unconditionally include asm/types.h from linux/types.h
>>  527bdfe: make linux/types.h as assembly safe
>>   
> OK, but that change isn't in x86/headers, so it won't compile standalone.

ok - i merged it into x86/headers.

	Ingo

Re: [PATCH] x86 headers: include linux/types.h

[Jeremy Fitzhardinge]

Ingo Molnar wrote:
> likewise, PUD_PAGE_* and PAGETABLE_LEVELS should move to pgtable_types.h
> too i guess. Same goes for *MASK_SHIFT*.

Hm, iffy.  I think those are getting more to the larger-scale virtual 
and physical memory layout for the machine: the various page sizes 
available and the number of P and V bits.

    J