Messages in this thread Patch in this message | | | Subject | Re: [PATCH] mm: fix page_mkclean_one (was: 2.6.19 file content corruption on ext3) | From | Dave Kleikamp <> | Date | Wed, 20 Dec 2006 15:55:35 -0600 |
| |
On Wed, 2006-12-20 at 11:50 -0800, Linus Torvalds wrote:
> NOTE NOTE NOTE! I _only_ did enough to make things compile for my > particular configuration. That means that right now the following > filesystems are broken with this patch (because they use the totally > broken old crap): > > CIFS, FUSE, JFS, ReiserFS, XFS > > and I don't know exactly what they need to be fixed. But most likely their > usage was insane and pointless anyway (looking at the ReiserFS case, for > example, that was DEFINITELY the case. I can't even imagine what the heck > it thinks it is doing).
Here's a patch to get rid of clear_page_dirty() from jfs. I'm not convinced it was totally broken, but I'm not convinced it wasn't. Either way, I don't think that bit of code was particularly beneficial.
Feel free to apply this patch independent of your patch if you really think that jfs's use of clear_page_dirty is crap, or I can push it through -mm first.
This patch removes some questionable code that attempted to make a no-longer-used page easier to reclaim.
Calling metapage_writepage against such a page will not result in any I/O being performed, so removing this code shouldn't be a big deal.
Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
diff -Nurp linux-orig/fs/jfs/jfs_metapage.c linux/fs/jfs/jfs_metapage.c --- linux-orig/fs/jfs/jfs_metapage.c 2006-12-07 17:12:58.000000000 -0600 +++ linux/fs/jfs/jfs_metapage.c 2006-12-20 15:19:48.000000000 -0600 @@ -764,22 +764,9 @@ void release_metapage(struct metapage * } else if (mp->lsn) /* discard_metapage doesn't remove it */ remove_from_logsync(mp); -#if MPS_PER_PAGE == 1 - /* - * If we know this is the only thing in the page, we can throw - * the page out of the page cache. If pages are larger, we - * don't want to do this. - */ - - /* Retest mp->count since we may have released page lock */ - if (test_bit(META_discard, &mp->flag) && !mp->count) { - clear_page_dirty(page); - ClearPageUptodate(page); - } -#else /* Try to keep metapages from using up too much memory */ drop_metapage(page, mp); -#endif + unlock_page(page); page_cache_release(page); }
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |