Re: 2.2.0-final

[ebiederm+eric@ccr.net (Eric W. Biederman)]

>>>>> "AA" == Andrea Arcangeli <andrea@e-mind.com> writes:

bh-> b_count++ with bget(bh) and implementing bget() this way: 

AA> extern inline unsigned int bget(struct buffer_head * bh)
AA> {
AA>         buffer_get(bh);
AA>         return ++bh->b_count;
AA> }

AA> where buffer_get() is this:

AA> extern inline void buffer_get(struct buffer_head *bh)
AA> {
AA>         struct page * page = mem_map + MAP_NR(bh->b_data);

AA>         switch (atomic_read(&page->count))
AA>         {
AA>         case 1:
AA>                 atomic_inc(&page->count);
AA>                 nr_freeable_pages--;

This is bogus.   Consider the case when you have 4 buffers per page (common with ext2fs)
You will way underestimate the number of freeable pages.

AA>                 break;
AA> #if 1 /* PARANOID */
AA>         case 0:
AA>                 printk(KERN_ERR "buffer_get: page was unused!\n");
AA> #endif
AA>         }
AA> }

AA> And for b_count-- exists a bput().

AA> Taking uptodate the file cache instead is been very easier (some line
AA> changed and nothing more). Lukily the only b_count++ or b_count-- are in
AA> buffer.c and in ext2fs, other fs has one or two b_count only.

Eric
--
To unsubscribe, send a message with 'unsubscribe linux-mm my@address'
in the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://humbolt.geo.uu.nl/Linux-MM/