linux-master/fs/fs-writeback.c

8  * upon dirty inodes against superblocks, and writing back dirty
9  * pages against inodes.  ie: data writeback.  Writeout of the
64  * seconds), which means most of the time inodes will have their
66  * few inodes might not their timestamps updated for 24 hours.
147  * periodic background write-out of dirty inodes. Since the write-out would
256  * Maximum inodes per isw.  A specific value has been chosen to make
302  * list.  Only inodes attached to cgwb's are kept on this list.
376 	 * Multiple inodes can be switched at once.  The switching procedure
384 	struct inode		*inodes[];
495 	struct backing_dev_info *bdi = inode_to_bdi(isw->inodes[0]);
496 	struct bdi_writeback *old_wb = isw->inodes[0]->i_wb;
525 	for (inodep = isw->inodes; *inodep; inodep++) {
541 	for (inodep = isw->inodes; *inodep; inodep++)
599 	isw = kzalloc(struct_size(isw, inodes, 2), GFP_ATOMIC);
622 	isw->inodes[0] = inode;
650 		isw->inodes[*nr] = inode;
660  * cleanup_offline_cgwb - detach associated inodes
663  * Switch all inodes attached to @wb to a nearest living ancestor's wb in order
664  * to eventually release the dying @wb.  Returns %true if not all inodes were
674 	isw = kzalloc(struct_size(isw, inodes, WB_MAX_INODES_PER_ISW),
693 	 * In addition to the inodes that have completed writeback, also switch
694 	 * cgwbs for those inodes only with dirty timestamps. Otherwise, those
695 	 * inodes won't be written back for a long time when lazytime is
705 	/* no attached inodes? bail out */
949  * relation to the total write bandwidth of all wb's w/ dirty inodes on
978  * have dirty inodes.  If @base_work->nr_page isn't %LONG_MAX, it's
1229  * Add in the number of potentially dirty inodes, because each inode
1398 	 * For inodes being constantly redirtied, dirtied_when can get stuck.
1409  * Move expired (dirtied before dirtied_before) dirty inodes from
1464  * Queue all expired dirty inodes for io, eldest first.
1559  * inodes. This function can be called only by flusher thread - noone else
1560  * processes all inodes in writeback lists and requeueing inodes behind flusher
1582 		 * is odd for clean inodes, it can happen for some
1815 	 * inodes/pages in one big loop. Setting wbc.nr_to_write=LONG_MAX
1840  * Write a portion of b_io inodes which belong to @sb.
1842  * Return the number of pages and/or inodes written.
1864 	long total_wrote = 0;  /* count both pages and inodes */
1875 				 * superblock, move all inodes not belonging
1891 		 * Don't bother with new inodes or inodes being freed, first
1906 			 * other inodes on s_io.
2023 	/* Leave any unwritten inodes on b_io */
2054  * just walks the superblock inode list, writing back any inodes which are
2103 		 * include all inodes that need writing. Livelock avoidance is
2125 		 * Dirty inodes are moved to b_io for writeback in batches.
2128 		 * as made some progress on cleaning pages or inodes.
2136 		 * No more inodes for IO, bail
2367  * Wake up bdi's periodically to make sure dirtytime inodes gets
2371  * inodes on the system.  So instead we define a separate delayed work
2433  * refers to a blockdev.  Unhashed inodes will never be added to the dirty list
2436  * In short, ensure you hash any inodes _before_ you start marking them dirty.
2513 		 * list handling so that we don't move inodes under flush worker's
2531 		 * Only add valid (hashed) inodes to the superblock's
2532 		 * dirty list.  Add blockdev inodes as well.
2607 	 * inodes that have started writeback after this point.
2609 	 * Use rcu_read_lock() to keep the inodes around until we have a
2611 	 * the local list because inodes can be dropped from either by writeback
2701  * writeback_inodes_sb_nr -	writeback dirty inodes from given super_block
2706  * Start writeback on some inodes on this super_block. No guarantees are made
2719  * writeback_inodes_sb	-	writeback dirty inodes from given super_block
2723  * Start writeback on some inodes on this super_block. No guarantees are made
2773 	 * inodes under writeback and I_DIRTY_TIME inodes ignored by