fs/xfs/xfs_discard.c

37  * extremely slow discard implementations (e.g. ceph RBD) and so walking a
51  * state of the free extent is out of our control and we cannot run a discard
53  * extent as busy and undergoing a discard operation whilst we held the AGF
56  * This is exactly how online discard works - free extents are marked busy when
58  * the busy extent record is marked as "undergoing discard" and the discard is
59  * then issued on the free extent. Once the discard completes, the busy extent
62  * In the context of fstrim, if we find a free extent we need to discard, we
63  * don't have to discard it immediately. All we need to do it record that free
64  * extent as being busy and under discard, and all the allocation routines will
66  * the AGF lock, we can safely discard it without holding the AGF lock because
67  * nothing will attempt to allocate that free space until the discard completes.
70  * discard, and so for fast devices fstrim will run much faster as we can have
71  * multiple discard operations in flight at once, as well as pipeline the free
72  * extent search so that it overlaps in flight discard IO.
104  * Walk the discard list and issue discards on all the busy extents in the
129 	 "discard failed for extent [0x%llx,%u], error %d",  in xfs_discard_extents()
187 		/* sub-AG discard request always starts at tcur->start */  in xfs_trim_gather_extents()
273 		 * discard and try again the next time.  in xfs_trim_gather_extents()
301 	 * we aren't going to issue a discard on them any more.  in xfs_trim_gather_extents()
320  * for the repeated iteration of gather/discard loop, so use the longest extent
361 		 * We hand the extent list to the discard function here so the  in xfs_trim_perag_extents()
364 		 * the next round of extents to discard.  in xfs_trim_perag_extents()
457  * Walk the discard list and issue discards on all the busy extents in the
496 	 "discard failed for rtextent [0x%llx,%llu], error %d",  in xfs_discard_rtdev_extents()