From dd05add7b2b6004c5fa3a1f276f56d8668f86b65 Mon Sep 17 00:00:00 2001 From: Paulo Alcantara Date: Sat, 26 Sep 2026 19:20:30 -0300 Subject: netfs: zero the tail of a short DIO/unbuffered read The buffered read collector zero-fills the tail of a short read that stops below the inode's i_size (netfs_clear_unread()), so a read that races an extending write still returns the expected number of bytes. The non-buffered collector path does no such thing: it just records how much was transferred. Add the same zero-fill for the non-buffered case, gated on NETFS_SREQ_CLEAR_TAIL: a subreq's source sets that flag when a short result from it is known to be safe to treat as a hole, as opposed to NETFS_SREQ_HIT_EOF, which means the read genuinely ran off the end of the file and should be reported short as-is. Only CLEAR_TAIL should zero-fill here; a real EOF must stay a real short read. No source currently sets CLEAR_TAIL on an unbuffered/DIO subrequest, so this is inert on its own -- a following change teaches cifs to set it in the one case that needs it. Fixes: e2d46f2ec332 ("netfs: Change the read result collector to only use one work item") Reviewed-by: David Howells Reviewed-by: Namjae Jeon Signed-off-by: Paulo Alcantara Cc: Christian Brauner Cc: Matthew Wilcox Cc: Ronnie Sahlberg Cc: Shyam Prasad N Cc: Tom Talpey Cc: Bharath SM Cc: stable@vger.kernel.org --- Documentation/admin-guide/device-mapper/delay.rst | 54 +++++++++++++++++++++++ 1 file changed, 54 insertions(+) create mode 100644 Documentation/admin-guide/device-mapper/delay.rst (limited to 'Documentation/admin-guide/device-mapper/delay.rst') diff --git a/Documentation/admin-guide/device-mapper/delay.rst b/Documentation/admin-guide/device-mapper/delay.rst new file mode 100644 index 000000000..a1e673c0e --- /dev/null +++ b/Documentation/admin-guide/device-mapper/delay.rst @@ -0,0 +1,54 @@ +======== +dm-delay +======== + +Device-Mapper's "delay" target delays reads and/or writes +and/or flushes and optionally maps them to different devices. + +Arguments:: + + [ + [ ]] + +Table line has to either have 3, 6 or 9 arguments: + +3: apply offset and delay to read, write and flush operations on device + +6: apply offset and delay to device, also apply write_offset and write_delay + to write and flush operations on optionally different write_device with + optionally different sector offset + +9: same as 6 arguments plus define flush_offset and flush_delay explicitly + on/with optionally different flush_device/flush_offset. + +Offsets are specified in sectors. + +Delays are specified in milliseconds. + + +Example scripts +=============== + +:: + #!/bin/sh + # + # Create mapped device named "delayed" delaying read, write and flush operations for 500ms. + # + dmsetup create delayed --table "0 `blockdev --getsz $1` delay $1 0 500" + +:: + #!/bin/sh + # + # Create mapped device delaying write and flush operations for 400ms and + # splitting reads to device $1 but writes and flushes to different device $2 + # to different offsets of 2048 and 4096 sectors respectively. + # + dmsetup create delayed --table "0 `blockdev --getsz $1` delay $1 2048 0 $2 4096 400" + +:: + #!/bin/sh + # + # Create mapped device delaying reads for 50ms, writes for 100ms and flushes for 333ms + # onto the same backing device at offset 0 sectors. + # + dmsetup create delayed --table "0 `blockdev --getsz $1` delay $1 0 50 $2 0 100 $1 0 333" -- cgit v1.3.1