swap: Add ability to mark swap pages bad
Subsystems (3)
PATCH 2/4 swap: Add ability to mark swap pages bad
Add swap_free_markbad(), equivalent to swap_free but marks the swap page as bad. Update try_to_unuse_entry() and shmem_unuse() to call the new function when working on pages with errors.
These patches are a based on a patch by Nick Piggin and some of my own patches/bugfixes as discussed on LKML.
Signed-off-by Richard Purdie <[email protected]>
--- include/linux/swap.h | 1 + mm/shmem.c | 17 ++++++++++------- mm/swap_state.c | 2 ++ mm/swapfile.c | 38 +++++++++++++++++++++++++++++--------- 4 files changed, 42 insertions(+), 16 deletions(-)
include/linux/swap.h +1 -0
@@ -244,6 +244,7 @@ extern swp_entry_t get_swap_page_of_typeextern int swap_duplicate (swp_entry_t );extern int valid_swaphandles (swp_entry_t ,unsigned long *);extern void swap_free (swp_entry_t ); +extern void swap_free_markbad (swp_entry_t );extern void free_swap_and_cache (swp_entry_t );extern int swap_type_of (dev_t ,sector_t ,struct block_device **);extern unsigned int count_swap_pages (int ,int );
mm/shmem.c +10 -7
@@ -734,7 +734,7 @@ static int shmem_unuse_inode(struct shmestruct page **dir;struct page *subdir;swp_entry_t *ptr; -int offset; +int offset, moved, error; idx =0 ; ptr = info->i_direct;
@@ -792,17 +792,20 @@ lost2: found: idx += offset; inode = &info->vfs_inode; -if (move_from_swap_cache (page, idx, inode->i_mapping) ==0 ) { + error =PageError (page); + moved = (move_from_swap_cache (page, idx, inode->i_mapping) ==0 ); +if (moved) { info->flags |=SHMEM_PAGEIN ;shmem_swp_set (info, ptr + offset,0 ); }shmem_swp_unmap (ptr);spin_unlock (&info->lock); -/* -* Decrement swap count even when the entry is left behind: -* try_to_unuse will skip over mms, then reincrement count. -*/ -swap_free (entry); +if (moved) { +if (!error) +swap_free (entry); +else +swap_free_markbad (entry); + }return 1 ; }
mm/swapfile.c +29 -9
@@ -304,6 +304,23 @@ void swap_free(swp_entry_t entry) } } +void swap_free_markbad (swp_entry_t entry) +{ +struct swap_info_struct * p; + + p =swap_info_get (entry); +if (p) { +unsigned long offset =swp_offset (entry); +if (swap_entry_free (p, offset) ==0 ) { + p->swap_map[offset] =SWAP_MAP_BAD ; + p->pages--; + nr_swap_pages--; + total_swap_pages--; + } +spin_unlock (&swap_lock); + } +} +/* * How many references to page are currently swapped out? */
@@ -764,11 +781,6 @@ retry:* remains (rarer), it will be read from disk into another page. * Splitting into two pages would be incorrect if swap supported * "shared private" pages, but they are handled by tmpfs files. -* -* Note shmem_unuse already deleted a swappage from the swap cache, -* unless the move to filepage failed: in which case it left swappage -* in cache, lowered its swap count to pass quickly through the loops -* above, and now we must reincrement count to try again later. */ if (PageSwapCache (page)) {if ((*swap_map >1 ) &&PageDirty (page)) {
@@ -781,11 +793,19 @@ retry:wait_on_page_writeback (page); goto retry; } +if (!shmem) { +int error =PageError (page); -if (shmem) -swap_duplicate (entry); -else -delete_from_swap_cache (page); +write_lock_irq (&swapper_space.tree_lock); +__delete_from_swap_cache (page); +write_unlock_irq (&swapper_space.tree_lock); +page_cache_release (page);/* the swapcache ref */ + +if (!error) +swap_free (entry); +else +swap_free_markbad (entry); + } }/*
mm/swap_state.c +2 -0
@@ -131,6 +131,8 @@ void __delete_from_swap_cache(struct pagradix_tree_delete (&swapper_space.page_tree,page_private (page));set_page_private (page,0 );ClearPageSwapCache (page); +if (unlikely (PageError (page))) +ClearPageError (page); total_swapcache_pages--;__dec_zone_page_state (page,NR_FILE_PAGES );INC_CACHE_INFO (del_total);