Description of problem:
Since some time iSCSI target is not functioning. After some initial I/O, target throws "fd_do_rw() write returned -22" and initiator throws IO error.
Detailed configuration steps follow.

TARGET (iqn.2003-01.org.linux-iscsi.seurat.x8664, 192.168.22.1) :
Rootfs is btrfs (rw,relatime,seclabel,compress=zlib,ssd,space_cache,autodefrag)
1. truncate -s 800M /iscsi.bs0
2. targetcli:
  cd backstores/fileio
  create file_or_dev=/iscsi.bs0 name=fbs0
  cd /iscsi
  create
  cd iqn.2003-01.org.linux-iscsi.seurat.x8664:sn.513490d29dc7/tpg1 \
  cd luns
  create /backstores/fileio/fbs0
  cd /iscsi/iqn.2003-01.org.linux-iscsi.seurat.x8664:sn.513490d29dc7/tpg1/acls
  create iqn.2003-01.org.linux-iscsi.ginjo.x8664:sn.37648a3d9cbe

  ls /
o- / ............................................................................ [...]
  o- backstores ................................................................. [...]
  | o- block ..................................................... [Storage Objects: 0]
  | o- fileio .................................................... [Storage Objects: 1]
  | | o- fbs0 ............................ [/iscsi.bs0 (800.0MiB) write-back activated]
  | o- pscsi ..................................................... [Storage Objects: 0]
  | o- ramdisk ................................................... [Storage Objects: 0]
  | o- user ...................................................... [Storage Objects: 0]
  o- iscsi ............................................................... [Targets: 1]
  | o- iqn.2003-01.org.linux-iscsi.seurat.x8664:sn.513490d29dc7 ............. [TPGs: 1]
  |   o- tpg1 .................................................. [no-gen-acls, no-auth]
  |     o- acls ............................................................. [ACLs: 1]
  |     | o- iqn.2003-01.org.linux-iscsi.ginjo.x8664:sn.37648a3d9cbe . [Mapped LUNs: 1]
  |     |   o- mapped_lun0 .................................... [lun0 fileio/fbs0 (rw)]
  |     o- luns ............................................................. [LUNs: 1]
  |     | o- lun0 .......................................... [fileio/fbs0 (/iscsi.bs0)]
  |     o- portals ....................................................... [Portals: 1]
  |       o- 0.0.0.0:3260 ........................................................ [OK]




INITIATOR (iqn.2003-01.org.linux-iscsi.ginjo.x8664:sn.37648a3d9cbe, 192.168.22.3) :
1. iscsiadm -m discovery -t st -p 192.168.22.1
   192.168.22.1:3260,1 iqn.2003-01.org.linux-iscsi.seurat.x8664:sn.513490d29dc7
2. iscsiadm -m node -l
[May 6 08:05] scsi host4: iSCSI Initiator over TCP/IP
[  +0.041362] scsi 4:0:0:0: Direct-Access     LIO-ORG  fbs0             4.0  PQ: 0 ANSI: 5
[  +0.028042] sd 4:0:0:0: Attached scsi generic sg2 type 0
[  +0.014333] sd 4:0:0:0: [sdb] 1638400 512-byte logical blocks: (838 MB/800 MiB)
[  +0.010611] sd 4:0:0:0: [sdb] Write Protect is off
[  +0.003731] sd 4:0:0:0: [sdb] Mode Sense: 43 00 10 08
[  +0.013365] sd 4:0:0:0: [sdb] Write cache: enabled, read cache: enabled, supports DPO and FUA
[  +0.083011] sd 4:0:0:0: [sdb] Attached SCSI disk

3. mkfs.btrfs -M -L seurat-iscsi /dev/sdb
4. btrfstune -x /dev/sdb

5. mount /dev/sdb /mnt/tmp -o autodefrag,compress=lzo
[May 6 08:06] BTRFS: device label seurat-iscsi devid 1 transid 3 /dev/sdb
[ +22.736061] BTRFS info (device sdb): disk space caching is enabled
[  +0.003400] BTRFS: has skinny extents
[  +0.010614] BTRFS: creating UUID tree
[May 6 08:07] BTRFS info (device sdb): enabling auto defrag
[  +0.005219] BTRFS info (device sdb): setting 8 feature flag
[  +0.003252] BTRFS info (device sdb): disk space caching is enabled
(there are two mounts in the log snippet above)


NOW START THE I/O. Soon:

INITIATOR:
[May 6 08:08] fd_do_rw() write returned -22
[  +0.107407] fd_do_rw() write returned -22


TARGET:
[May 6 08:08] sd 4:0:0:0: [sdb] FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[  +0.004159] sd 4:0:0:0: [sdb] Sense Key : Not Ready [current] 
[  +0.001032] sd 4:0:0:0: [sdb] Add. Sense: Logical unit communication failure
[  +0.003389] sd 4:0:0:0: [sdb] CDB: Write(10) 2a 00 00 01 6b 20 00 2b e0 00
[  +0.003223] blk_update_request: I/O error, dev sdb, sector 92960
[  +0.001409] BTRFS: bdev /dev/sdb errs: wr 1, rd 0, flush 0, corrupt 0, gen 0
[  +0.002002] BTRFS: bdev /dev/sdb errs: wr 2, rd 0, flush 0, corrupt 0, gen 0
[  +0.000759] BTRFS: bdev /dev/sdb errs: wr 3, rd 0, flush 0, corrupt 0, gen 0
[  +0.001236] BTRFS: bdev /dev/sdb errs: wr 4, rd 0, flush 0, corrupt 0, gen 0
[  +0.000801] BTRFS: bdev /dev/sdb errs: wr 5, rd 0, flush 0, corrupt 0, gen 0
[  +0.000644] BTRFS: bdev /dev/sdb errs: wr 6, rd 0, flush 0, corrupt 0, gen 0
[  +0.088333] sd 4:0:0:0: [sdb] FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[  +0.000003] sd 4:0:0:0: [sdb] Sense Key : Not Ready [current] 
[  +0.000003] sd 4:0:0:0: [sdb] Add. Sense: Logical unit communication failure
[  +0.000003] sd 4:0:0:0: [sdb] CDB: Write(10) 2a 00 00 00 a0 00 00 3b 20 00
[  +0.000003] blk_update_request: I/O error, dev sdb, sector 40960
[  +0.000006] BTRFS: bdev /dev/sdb errs: wr 7, rd 0, flush 0, corrupt 0, gen 0
[  +0.000195] BTRFS: bdev /dev/sdb errs: wr 8, rd 0, flush 0, corrupt 0, gen 0
[  +0.000085] BTRFS: bdev /dev/sdb errs: wr 9, rd 0, flush 0, corrupt 0, gen 0
[  +0.000082] BTRFS: bdev /dev/sdb errs: wr 10, rd 0, flush 0, corrupt 0, gen 0

Application exits with I/O errors, btrfs sometimes switches to read-only.



Version-Release number of selected component (if applicable):
TARGET:
kernel-core-4.0.1-300.fc22.x86_64
targetcli-2.1.fb40-1.fc21.noarch

INITIATOR:
kernel-core-4.1.0-0.rc1.git1.1.fc23.x86_64
iscsi-initiator-utils-6.2.0.873-25.gitc9d830b.fc22.x86_64

How reproducible:
I've reproduced it twice on different hosts/VMs. In both cases initiator/target are directly connected: a) target on physical host, initiator is a VM hosted on this host - KVM is used; b) target and initiator are VMs hosted on the same machine, VirtualBox is used.