Fix the problem (BUG 4964) with unmapped buffers in transaction's t_sync_data list. The problem is we need to call filesystem's own invalidatepage() from block_write_full_page(). block_write_full_page() must call filesystem's invalidatepage(). Otherwise following nasty race can happen: proc 1 proc 2 ------ ------ - write some new data to 'offset' => bh gets to the transactions data list - starts truncate => i_size set to new size - mpage_writepages() - ext3_ordered_writepage() to 'offset' - block_write_full_page() - page->index > end_index+1 - block_invalidatepage() - discard_buffer() - clear_buffer_mapped() - commit triggers and finds unmapped buffer - BOOM!
Showing the changes to include/linux/buffer_head.h (1 of 3 files). Show all
Unified · Split
| @@ -190,6 +190,7 @@ extern int buffer_heads_over_limit; · 189 unmodified lines | |||
| 190 | */ | 190 | */ |
| 191 | int try_to_release_page(struct page * page, gfp_t gfp_mask); | 191 | int try_to_release_page(struct page * page, gfp_t gfp_mask); |
| 192 | int block_invalidatepage(struct page *page, unsigned long offset); | 192 | int block_invalidatepage(struct page *page, unsigned long offset); |
| 193 | int do_invalidatepage(struct page *page, unsigned long offset); | ||
| 193 | int block_write_full_page(struct page *page, get_block_t *get_block, | 194 | int block_write_full_page(struct page *page, get_block_t *get_block, |
| 194 | struct writeback_control *wbc); | 195 | struct writeback_control *wbc); |
| 195 | int block_read_full_page(struct page*, get_block_t*); | 196 | int block_read_full_page(struct page*, get_block_t*); |