Tejun says:
"At least for libata, worrying about suspend/resume failures don't make
whole lot of sense. If suspend failed, just proceed with suspend. If
the device can't be woken up afterwards, that's that. There isn't
anything we could have done differently anyway. The same for resume, if
spinup fails, the device is dud and the following commands will invoke
EH actions and will eventually fail. Again, there really isn't any
*choice* to make. Just making sure the errors are handled gracefully
(ie. don't crash) and the following commands are handled correctly
should be enough."
The only libata user that actually cares about the result from a suspend
operation is libsas. However, it only cares about whether queuing a new
operation collides with an in-flight one. All libsas does with the
error is retry, but we can just let libata wait for the previous
operation before continuing.
Other cleanups include:
1/ Unifying all ata port pm operations on an ata_port_pm_ prefix
2/ Marking all ata port pm helper routines as returning void, only
ata_port_pm_ entry points need to fake a 0 return value.
3/ Killing ata_port_{suspend|resume}_common() in favor of calling
ata_port_request_pm() directly
4/ Killing the wrappers that just do a to_ata_port() conversion
5/ Clearly marking the entry points that do async operations with an
_async suffix.
Reference: http://marc.info/?l=linux-scsi&m=138995409532286&w=2
Cc: Phillip Susi <psusi@ubuntu.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Todd Brandt <todd.e.brandt@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Pull SCSI target updates from Nicholas Bellinger:
"The highlights this round include:
- add support for SCSI Referrals (Hannes)
- add support for T10 DIF into target core (nab + mkp)
- add support for T10 DIF emulation in FILEIO + RAMDISK backends (Sagi + nab)
- add support for T10 DIF -> bio_integrity passthrough in IBLOCK backend (nab)
- prep changes to iser-target for >= v3.15 T10 DIF support (Sagi)
- add support for qla2xxx N_Port ID Virtualization - NPIV (Saurav + Quinn)
- allow percpu_ida_alloc() to receive task state bitmask (Kent)
- fix >= v3.12 iscsi-target session reset hung task regression (nab)
- fix >= v3.13 percpu_ref se_lun->lun_ref_active race (nab)
- fix a long-standing network portal creation race (Andy)"
* 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending: (51 commits)
target: Fix percpu_ref_put race in transport_lun_remove_cmd
target/iscsi: Fix network portal creation race
target: Report bad sector in sense data for DIF errors
iscsi-target: Convert gfp_t parameter to task state bitmask
iscsi-target: Fix connection reset hang with percpu_ida_alloc
percpu_ida: Make percpu_ida_alloc + callers accept task state bitmask
iscsi-target: Pre-allocate more tags to avoid ack starvation
qla2xxx: Configure NPIV fc_vport via tcm_qla2xxx_npiv_make_lport
qla2xxx: Enhancements to enable NPIV support for QLOGIC ISPs with TCM/LIO.
qla2xxx: Fix scsi_host leak on qlt_lport_register callback failure
IB/isert: pass scatterlist instead of cmd to fast_reg_mr routine
IB/isert: Move fastreg descriptor creation to a function
IB/isert: Avoid frwr notation, user fastreg
IB/isert: seperate connection protection domains and dma MRs
tcm_loop: Enable DIF/DIX modes in SCSI host LLD
target/rd: Add DIF protection into rd_execute_rw
target/rd: Add support for protection SGL setup + release
target/rd: Refactor rd_build_device_space + rd_release_device_space
target/file: Add DIF protection support to fd_execute_rw
target/file: Add DIF protection init/format support
...
Pull exofs and ore fixes from Boaz Harrosh:
"The main fix here, the first patch, is also destined for -stable. The
rest is small trivia and cosmetics. The ORE patches effect both exofs
and pnfs-objects very reproducible bugs"
[ ORE is "object raid engine", used by exofs and pnfs - Linus ]
* 'for-linus' of git://git.open-osd.org/linux-open-osd:
exofs: Print less in r4w
exofs: Allow corrupted directory entry to be empty file
exofs: Allow O_DIRECT open
ore: Don't crash on NULL bio in _clear_bio
ore: Fix wrong math in allocation of per device BIO
Pull infiniband updates from Roland Dreier:
"Main batch of InfiniBand/RDMA changes for 3.14:
- Flow steering for InfiniBand UD traffic
- IP-based addressing for IBoE aka RoCE
- Pass SRP submaintainership from Dave to Bart
- SRP transport fixes from Bart
- Add the new Cisco usNIC low-level device driver
- Various other fixes"
* tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (75 commits)
IB/mlx5: Verify reserved fields are cleared
IB/mlx5: Remove old field for create mkey mailbox
IB/mlx5: Abort driver cleanup if teardown hca fails
IB/mlx5: Allow creation of QPs with zero-length work queues
mlx5_core: Fix PowerPC support
mlx5_core: Improve debugfs readability
IB/mlx5: Add support for resize CQ
IB/mlx5: Implement modify CQ
IB/mlx5: Make sure doorbell record is visible before doorbell
mlx5_core: Use mlx5 core style warning
IB/mlx5: Clear out struct before create QP command
mlx5_core: Fix out arg size in access_register command
RDMA/nes: Slight optimization of Ethernet address compare
IB/qib: Fix QP check when looping back to/from QP1
RDMA/cxgb4: Fix gcc warning on 32-bit arch
IB/usnic: Remove unused includes of <linux/version.h>
RDMA/amso1100: Add check if cache memory was allocated before freeing it
IPoIB: Report operstate consistently when brought up without a link
IB/core: Fix unused variable warning
RDMA/cma: Handle global/non-linklocal IPv6 addresses in cma_check_linklocal()
...
At IO preparation we calculate the max pages at each device and
allocate a BIO per device of that size. The calculation was wrong
on some unaligned corner cases offset/length combination and would
make prepare return with -ENOMEM. This would be bad for pnfs-objects
that would in that case IO through MDS. And fatal for exofs were it
would fail writes with EIO.
Fix it by doing the proper math, that will work in all cases. (I
ran a test with all possible offset/length combinations this time
round).
Also when reading we do not need to allocate for the parity units
since we jump over them.
Also lower the max_io_length to take into account the parity pages
so not to allocate BIOs bigger than PAGE_SIZE
CC: Stable Kernel <stable@vger.kernel.org>
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
The following command has been used to verify that the kernel-doc
tool no longer complains about undocumented fields:
scripts/kernel-doc -html drivers/scsi/scsi_transport_srp.c \
include/scsi/scsi_transport_srp.h >srp-transport-doc.html
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Acked-by: Sebastian Riemer <sebastian.riemer@profitbricks.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
The rport timers must be stopped before the SRP initiator destroys the
resources associated with the SCSI host. This is necessary because
otherwise the callback functions invoked from the SRP transport layer
could trigger a use-after-free. Stopping the rport timers before
invoking scsi_remove_host() can trigger long delays in the SCSI error
handler if a transport layer failure occurs while scsi_remove_host()
is in progress. Hence move the code for stopping the rport timers from
srp_rport_release() into a new function and invoke that function after
scsi_remove_host() has finished. This patch fixes the following
sporadic kernel crash:
kernel BUG at include/asm-generic/dma-mapping-common.h:64!
invalid opcode: 0000 [#1] SMP
RIP: 0010:[<ffffffffa03b20b1>] [<ffffffffa03b20b1>] srp_unmap_data+0x121/0x130 [ib_srp]
Call Trace:
[<ffffffffa03b20fc>] srp_free_req+0x3c/0x80 [ib_srp]
[<ffffffffa03b2188>] srp_finish_req+0x48/0x70 [ib_srp]
[<ffffffffa03b21fb>] srp_terminate_io+0x4b/0x60 [ib_srp]
[<ffffffffa03a6fb5>] __rport_fail_io_fast+0x75/0x80 [scsi_transport_srp]
[<ffffffffa03a7438>] rport_fast_io_fail_timedout+0x88/0xc0 [scsi_transport_srp]
[<ffffffff8108b370>] worker_thread+0x170/0x2a0
[<ffffffff81090876>] kthread+0x96/0xa0
[<ffffffff8100c0ca>] child_rip+0xa/0x20
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Add infrastructure for referrals.
v2 changes:
- Fix unsigned long long division in core_alua_state_lba_dependent on
32-bit (Fengguang + Chen + Hannes)
- Fix compile warning in core_alua_state_lba_dependent (nab)
- Convert segment_* + sectors variables in core_alua_state_lba_dependent
to u64 (Hannes)
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
When a command runs into a timeout we need to send an 'ABORT TASK'
TMF. This is typically done by the 'eh_abort_handler' LLDD callback.
Conceptually, however, this function is a normal SCSI command, so
there is no need to enter the error handler.
This patch implements a new scsi_abort_command() function which
invokes an asynchronous function scsi_eh_abort_handler() to
abort the commands via the usual 'eh_abort_handler'.
If abort succeeds the command is either retried or terminated,
depending on the number of allowed retries. However, 'eh_eflags'
records the abort, so if the retry would fail again the
command is pushed onto the error handler without trying to
abort it (again); it'll be cleared up from SCSI EH.
[hare: smatch detected stray switch fixed]
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Commit 18a4d0a22e
(Handle disk devices which can not process medium access commands)
was introduced to offline any device which cannot process medium
access commands.
However, commit 3eef6257de
(Reduce error recovery time by reducing use of TURs) reduced
the number of TURs by sending it only on the first failing
command, which might or might not be a medium access command.
So in combination this results in an erratic device offlining
during EH; if the command where the TUR was sent upon happens
to be a medium access command the device will be set offline,
if not everything proceeds as normal.
This patch moves the check to the final test, eliminating
this problem.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Some host adapters do not pass commands through to the target disk
directly. Instead they provide an emulated target which may or may not
accurately report its capabilities. In some cases the physical device
characteristics are reported even when the host adapter is processing
commands on the device's behalf. This can lead to adapter firmware hangs
or excessive I/O errors.
This patch disables WRITE SAME for devices connected to host adapters
that provide an emulated target. Driver writers can disable WRITE SAME
by setting the no_write_same flag in the host adapter template.
[jejb: fix up rejections due to eh_deadline patch]
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Cc: stable@kernel.org
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Pull infiniband/rdma updates from Roland Dreier:
- Re-enable flow steering verbs with new improved userspace ABI
- Fixes for slow connection due to GID lookup scalability
- IPoIB fixes
- Many fixes to HW drivers including mlx4, mlx5, ocrdma and qib
- Further improvements to SRP error handling
- Add new transport type for Cisco usNIC
* tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (66 commits)
IB/core: Re-enable create_flow/destroy_flow uverbs
IB/core: extended command: an improved infrastructure for uverbs commands
IB/core: Remove ib_uverbs_flow_spec structure from userspace
IB/core: Use a common header for uverbs flow_specs
IB/core: Make uverbs flow structure use names like verbs ones
IB/core: Rename 'flow' structs to match other uverbs structs
IB/core: clarify overflow/underflow checks on ib_create/destroy_flow
IB/ucma: Convert use of typedef ctl_table to struct ctl_table
IB/cm: Convert to using idr_alloc_cyclic()
IB/mlx5: Fix page shift in create CQ for userspace
IB/mlx4: Fix device max capabilities check
IB/mlx5: Fix list_del of empty list
IB/mlx5: Remove dead code
IB/core: Encorce MR access rights rules on kernel consumers
IB/mlx4: Fix endless loop in resize CQ
RDMA/cma: Remove unused argument and minor dead code
RDMA/ucma: Discard events for IDs not yet claimed by user space
IB/core: Add Cisco usNIC rdma node and transport types
RDMA/nes: Remove self-assignment from nes_query_qp()
IB/srp: Report receive errors correctly
...
Add support for periodically reconnecting to an SRP target until
the dev_loss timer expires. After the tenth reconnection attempt,
gradually slow down subsequent reconnect attempts.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Acked-by: David Dillow <dillowda@ornl.gov>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Add the necessary functions in the SRP transport module to allow an
SRP initiator driver to implement transport layer error handling
similar to the functionality already provided by the FC transport
layer. This includes:
- Support for implementing fast_io_fail_tmo, the time that should
elapse after having detected a transport layer problem and
before failing I/O.
- Support for implementing dev_loss_tmo, the time that should
elapse after having detected a transport layer problem and
before removing a remote port.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Acked-by: David Dillow <dillowda@ornl.gov>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Keep the rport data structure around after srp_remove_host() has
finished until cleanup of the IB transport layer has finished
completely. This is necessary because later patches use the rport
pointer inside the queuecommand callback. Without this patch
accessing the rport from inside a queuecommand callback is racy
because srp_remove_host() must be invoked before scsi_remove_host()
and because the queuecommand callback could get invoked after
srp_remove_host() has finished. In other words, without this patch
the queuecommand callback can get invoked after the rport data
structure has been freed.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Acked-by: David Dillow <dillowda@ornl.gov>
Signed-off-by: Roland Dreier <roland@purestorage.com>
This patchs adds an 'eh_deadline' sysfs attribute to the scsi
host which limits the overall runtime of the SCSI EH.
The 'eh_deadline' value is stored in the now obsolete field
'resetting'.
When a command is failed the start time of the EH is stored
in 'last_reset'. If the overall runtime of the SCSI EH is longer
than last_reset + eh_deadline, the EH is short-circuited and
falls through to issue a host reset only.
[jejb: add comments in Scsi_Host about new fields]
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Field is now unused, so this is dead code.
[jejb: remove resetting and last_reset from Scsi_Host]
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
For offload iSCSI like qla4xxx, CHAP entries are stored in adapter's
flash.
This patch adds support to add/update CHAP entries in adapter's flash
using iscsi tools, like Open-iSCSI.
Signed-off-by: Adheer Chandravanshi <adheer.chandravanshi@qlogic.com>
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
fnic doesn't use any of the create/destroy/enable/disable interfaces
either from the (legacy) module paramaters or the (new) fcoe_sysfs
interfaces. When fcoe_sysfs was introduced fnic wasn't changed since
it wasn't using the interfaces. libfcoe incorrectly assumed that that
all of its users were using fcoe_sysfs and when adding and deleting
FCFs would assume the existance of a fcoe_ctlr_device. fnic was not
allocating this structure because it doesn't care about the standard
user interfaces (fnic starts on link only). If/When libfcoe tried to use
the fcoe_ctlr_device's lock for the first time a NULL pointer exception
would be triggered.
Since fnic doesn't care about sysfs or user interfaces, the solution
is to drop libfcoe's assumption that all drivers are using fcoe_sysfs.
This patch accomplishes this by changing some of the structure
relationships.
We need a way to determine when a LLD is using fcoe_sysfs or not and
we can do that by checking for the existance of the fcoe_ctlr_device.
Prior to this patch, it was assumed that the fcoe_ctlr structure was
allocated with the fcoe_ctlr_device and immediately followed it in
memory. To reach the fcoe_ctlr_device we would simply go back in memory
from the fcoe_ctlr to get the fcoe_ctlr_device.
Since fnic doesn't allocate the fcoe_ctlr_device, we cannot keep that
assumption. This patch adds a pointer from the fcoe_ctlr to the
fcoe_ctlr_device. For bnx2fc and fcoe we will continue to allocate the
two structures together, but then we'll set the ctlr->cdev pointer
to point at the fcoe_ctlr_device. fnic will not change and will continue
to allocate the fcoe_ctlr itself, and ctlr->cdev will remain NULL.
When libfcoe adds fcoe_fcf's to the fcoe_ctlr it will check if ctlr->cdev
is set and only if so will it continue to interact with fcoe_sysfs.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Tested-by: Hiral Patel <hiralpat@cisco.com>
Dan Williams
Linus Torvalds
Boaz Harrosh
Bart Van Assche
Hannes Reinecke
Adheer Chandravanshi
Lalit Chandivade
James Bottomley
Harish Zunjarrao
Martin K. Petersen
Robert Love