Doug / smarc-fsl-linux-kernel

Download as
Browse Code ยป

Commit b3af9468aebf5fcb573d0a116b31d2be1d43c0e9

Authored by Wu Fengguang
Committed by Jens Axboe
1 parent 9ecc2738ac
Exists in master and in 7 other branches imx_3.0.35_4.1.0, imx_3.10.17_1.0.1_ga, imx_3.10.53_1.1.0_ga, imx_3.14.28_1.0.0_ga, smarc-imx_3.10.53_1.1.0_ga, smarc-imx_3.14.28_1.0.0_ga, smarc-l5.0.0_1.0.0-ga

writeback: don't delay inodes redirtied by a fast dirtier 

Debug traces show that in per-bdi writeback, the inode under writeback
almost always get redirtied by a busy dirtier.  We used to call
redirty_tail() in this case, which could delay inode for up to 30s.

This is unacceptable because it now happens so frequently for plain cp/dd,
that the accumulated delays could make writeback of big files very slow.

So let's distinguish between data redirty and metadata only redirty.
The first one is caused by a busy dirtier, while the latter one could
happen in XFS, NFS, etc. when they are doing delalloc or updating isize.

The inode being busy dirtied will now be requeued for next io, while
the inode being redirtied by fs will continue to be delayed to avoid
repeated IO.

CC: Jan Kara <jack@suse.cz>
CC: Theodore Ts'o <tytso@mit.edu>
CC: Dave Chinner <david@fromorbit.com>
CC: Chris Mason <chris.mason@oracle.com>
CC: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

Showing 1 changed file with 9 additions and 3 deletions Side-by-side Diff

... ... @@ -474,11 +474,16 @@
474 474 spin_lock(&inode_lock);
475 475 inode->i_state &= ~I_SYNC;
476 476 if (!(inode->i_state & (I_FREEING | I_CLEAR))) {
477   - if (inode->i_state & I_DIRTY) {
  477 + if ((inode->i_state & I_DIRTY_PAGES) && wbc->for_kupdate) {
478 478 /*
479   - * Someone redirtied the inode while were writing back
480   - * the pages.
  479 + * More pages get dirtied by a fast dirtier.
481 480 */
  481 + goto select_queue;
  482 + } else if (inode->i_state & I_DIRTY) {
  483 + /*
  484 + * At least XFS will redirty the inode during the
  485 + * writeback (delalloc) and on io completion (isize).
  486 + */
482 487 redirty_tail(inode);
483 488 } else if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) {
484 489 /*
... ... @@ -502,6 +507,7 @@
502 507 * soon as the queue becomes uncongested.
503 508 */
504 509 inode->i_state |= I_DIRTY_PAGES;
  510 +select_queue:
505 511 if (wbc->nr_to_write <= 0) {
506 512 /*
507 513 * slice used up: queue for next turn