[patch] mm: export cancel_dirty_page()

[patch] mm: export cancel_dirty_page()

[Ingo Molnar]

From: Ingo Molnar <mingo@elte.hu>
Subject: [patch] export cancel_dirty_page()

export cancel_dirty_page() - it's used by hugetlbfs which can be 
modular. (This makes my -git based kernel yum repository build again.)

Signed-off-by: Ingo Molnar <mingo@elte.hu>

Index: linux/mm/truncate.c
===================================================================
--- linux.orig/mm/truncate.c
+++ linux/mm/truncate.c
@@ -65,6 +65,7 @@ void cancel_dirty_page(struct page *page
 		task_io_account_cancelled_write(account_size);
 	}
 }
+EXPORT_SYMBOL(cancel_dirty_page);
 
 /*
  * If truncate cannot remove the fs-private metadata from the page, the page

Re: [patch] mm: export cancel_dirty_page()

[Adrian Bunk]

On Fri, Dec 22, 2006 at 12:13:28AM +0100, Ingo Molnar wrote:
> From: Ingo Molnar <mingo@elte.hu>
> Subject: [patch] export cancel_dirty_page()
> 
> export cancel_dirty_page() - it's used by hugetlbfs which can be 
> modular. (This makes my -git based kernel yum repository build again.)
>...

No, it can't be:

config HUGETLBFS
        bool "HugeTLB file system support"
        ^^^^

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: [patch] mm: export cancel_dirty_page()

[Ingo Molnar]

* Adrian Bunk <bunk@stusta.de> wrote:

> On Fri, Dec 22, 2006 at 12:13:28AM +0100, Ingo Molnar wrote:
> > From: Ingo Molnar <mingo@elte.hu>
> > Subject: [patch] export cancel_dirty_page()
> > 
> > export cancel_dirty_page() - it's used by hugetlbfs which can be 
> > modular. (This makes my -git based kernel yum repository build again.)
> >...
> 
> No, it can't be:
> 
> config HUGETLBFS
>         bool "HugeTLB file system support"
>         ^^^^

ah, indeed - but i dont see a fundamental reason why hugetlbfs is not 
modular. Nevertheless exporting this makes sense. My quick hack below to 
guess to convert reiserfs (just to make the rpm build) also needs it.

	Ingo

Index: linux/fs/reiserfs/stree.c
===================================================================
--- linux.orig/fs/reiserfs/stree.c
+++ linux/fs/reiserfs/stree.c
@@ -1439,6 +1439,8 @@ static void unmap_buffers(struct page *p
 
 	if (page) {
 		if (page_has_buffers(page)) {
+			cancel_dirty_page(page, PAGE_CACHE_SIZE);
+
 			tail_index = pos & (PAGE_CACHE_SIZE - 1);
 			cur_index = 0;
 			head = page_buffers(page);
@@ -1458,9 +1460,6 @@ static void unmap_buffers(struct page *p
 				}
 				bh = next;
 			} while (bh != head);
-			if (PAGE_SIZE == bh->b_size) {
-				clear_page_dirty(page);
-			}
 		}
 	}
 }

Re: [patch] mm: export cancel_dirty_page()

[Linus Torvalds]

On Fri, 22 Dec 2006, Ingo Molnar wrote:
> 
> ah, indeed - but i dont see a fundamental reason why hugetlbfs is not 
> modular. Nevertheless exporting this makes sense. My quick hack below to 
> guess to convert reiserfs (just to make the rpm build) also needs it.

Yes, it should be exported regardless.

Hoiwever, I'm not sure your reiserfs change is valid: why was that old 
code testing "PAGE_SIZE == bh->b_size"?

(Not that I see why the _old_ code would be valid either, and why you'd 
ever care about b_size being PAGE_SIZE, but I'm just wondering..)

		Linus