- 15 Mar, 2014 40 commits
-
-
Stephen M. Cameron authored
Don't wait for *all* commands to complete, only for accelerated mode commands. Signed-off-by:
Stephen M. Cameron <scameron@beardog.cce.hp.com> Signed-off-by:
James Bottomley <JBottomley@Parallels.com>
-
Scott Teel authored
Add controller-based data-at-rest encryption compatibility to ioaccel2 path (HP SSD Smart Path). Encryption feature requires driver to supply additional fields for encryption enable, tweak index, and data encryption key index in the ioaccel2 request structure. Encryption enable flag and data encryption key index come from raid_map data structure from raid offload command. During ioaccel2 submission, check device structure's raid map to see if encryption is enabled for the device. If so, call new function below. Add function set_encrypt_ioaccel2 to set encryption flag, data encryption key index, and calculate tweak value from request's logical block address. Signed-off-by:
Scott Teel <scott.teel@hp.com> Signed-off-by:
-
Scott Teel authored
Signed-off-by:
-
Scott Teel authored
Change the handling of HP SSD Smart Path errors with status: 0x02 CHECK CONDITION 0x08 BUSY 0x18 RESERVATION CONFLICT 0x40 TASK ABORTED So that they get retried on the RAID Path. Signed-off-by:
-
Stephen M. Cameron authored
Signed-off-by:
-
Stephen M. Cameron authored
Code was confused and assumed that page zero was not VPD page and all non-zero pages were VPD pages. Signed-off-by:
-
Scott Teel authored
Allow driver to schedule a rescan whenever a request fails on the ioaccel2 path. This eliminates the possibility of driver getting stuck in non-ioaccel mode. IOaccel mode (HP SSD Smart Path) is disabled by driver upon error detection. Driver relied on idea that request would be retried through normal path, and a subsequent error would occur on that path, and be processed by controller firmware. As part of that process, controller disables ioaccel mode and later reinstates it, signalling driver to change modes. In some error cases, the error will not duplicate on the standard path, so the driver could get stuck in non-ioaccel mode. To avoid that, we allow driver to request a rescan during the next run of the rescan thread. Signed-off-by:
-
Scott Teel authored
Allow SSD Smart Path for a controller to be disabled by the user, regardless of settings in controller firmware or array configuration. To disable: echo 0 > /sys/class/scsi_host/host<id>/acciopath_status To re-enable: echo 1 > /sys/class/scsi_host/host<id>/acciopath_status To check state: cat /sys/class/scsi_host/host<id>/acciopath_status Signed-off-by:
-
Scott Teel authored
Load balance across members of a N-way mirror set, and handle the meta-RAID levels: R10, R50, R60. Signed-off-by:
-
Stephen M. Cameron authored
Otherwise we could wind up using incorrect raid map data, and then very bad things would likely happen. Signed-off-by:
-
Scott Teel authored
Underlying firmware cannot handle task abort on accelerated path (SSD Smart Path). Change abort requests for accelerated path commands to physical target reset. Send reset request on normal IO path. Signed-off-by:
Mike Miller <michael.miller@canonical.com> Signed-off-by:
-
Scott Teel authored
Signed-off-by:
-
Scott Teel authored
Signed-off-by:
Joe Handzik <Joseph.T.Handzik@hp.com> Signed-off-by:
-
Stephen M. Cameron authored
Signed-off-by:
-
Mike MIller authored
Signed-off-by:
-
Stephen M. Cameron authored
Signed-off-by:
-
Stephen M. Cameron authored
Signed-off-by:
-
Mike Miller authored
Signed-off-by:
-
Scott Teel authored
Signed-off-by:
-
Scott Teel authored
Signed-off-by:
-
Stephen M. Cameron authored
* Do not check event bits on locked up controllers to see if they need to be rescanned. * Do not initiate any device rescans on controllers which are known to be locked up. Signed-off-by:
-
Stephen M. Cameron authored
For shared SAS configurations, hosts need to poll Smart Arrays periodically in order to be able to detect configuration changes such as logical drives being added or removed from remote hosts. A register on the controller indicates when such events have occurred, and the driver polls the register via a workqueue and kicks off a rescan of devices if such an event is detected. Additionally, changes to logical drive raid offload eligibility are autodetected in this way. Signed-off-by:
-
Stephen M. Cameron authored
When rescanning for logical drives, store information about whather raid offload is enabled for each logical drive, and update the driver's internal record of this. Signed-off-by:
-
Stephen M. Cameron authored
This enables sending i/o's destined for RAID logical drives which can be serviced by a single physical disk down a different, faster i/o path directly to physical drives for certain logical volumes on SSDs bypassing the Smart Array RAID stack for a performance improvement. Signed-off-by:
Matt Gates <matthew.gates@hp.com> Signed-off-by:
Don Brace <brace@beardog.cce.hp.com> Signed-off-by:
-
Scott Teel authored
For "mode 1" io accelerated commands, the command tag is in a different location than for commands that go down the normal RAID path, so the abort handler needs to take this into account. Signed-off-by:
-
Stephen M. Cameron authored
When commands sent down the "fast path" fail, they must be re-tried down the normal RAID path. We do this by kicking i/o's back to the scsi mid layer with a DID_SOFT_ERROR status, which causes them to be retried. This won't work for SG_IO's and other non REQ_TYPE_FS i/o's which could get kicked all the way back to the application, which may have no idea that the command needs resubmitting and likely no way to resubmit it in such a way the that driver can recognize it as a resubmit and send it down the normal RAID path. So we just always send non REQ_TYPE_FS i/o's down the normal RAID path, never down the "fast path". Signed-off-by:
-
Matt Gates authored
For certain i/o's to certain devices (unmasked physical disks) we can bypass the RAID stack firmware and do the i/o to the device directly and it will be faster. Signed-off-by:
-
Matt Gates authored
This is normally optional, but for SSD Smart Path support (in subsequent patches) it is required. Signed-off-by:
-
Matt Gates authored
There is an extended report luns command which contains additional information about physical devices. In particular we need to get the physical device handle so we can use an alternate i/o path for fast physical devices like SSDs so we can speed up certain i/o's by bypassing the RAID stack code in the controller firmware. Signed-off-by:
-
Suresh Thiagarajan authored
spin_lock_irqsave for the HBA lock is called in one function where flag is local to that function. Another function is called from the first function where lock has to be released using spin_unlock_irqrestore for calling task_done of libsas. In the second function also flag is declared and used. For calling task_done there is no need to enable the irq. So instead of using spin_lock_irqsave and spin_unlock_irqrestore, spin_lock and spin_unlock is used now. This also avoids passing the flags across all the functions where HBA lock is being used. Also removed redundant code. Reported-by:
Jason Seba <jason.seba42@gmail.com> Signed-off-by:
Oleg Nesterov <oleg@redhat.com> Signed-off-by:
Suresh Thiagarajan <Suresh.Thiagarajan@pmcs.com> Signed-off-by:
Viswas G <viswas.g@pmcs.com> Acked-by:
Jack Wang <xjtuwjp@gmail.com> Signed-off-by:
-
Hannes Reinecke authored
Commit 254f796b updated the driver to use 16 MSI-X vectors, despite the fact that older controllers would provide only 4. This was causing MSI-X registration to drop down to INTx mode. But as the controller support performant mode, the initialisation will become confused and cause the machine to stall during boot. This patch fixes up the MSI-X registration to re-issue the pci_enable_msix() call with the correct number of MSI-X vectors. With that the hpsa driver continues to works on older controllers like the P200. Signed-off-by:
Hannes Reinecke <hare@suse.de> Acked-by:
-
Mahesh Rajashekhara authored
This patch fixes kernel panic issue while booting into the kdump kernel. We have triggered crash and kdump vmcore was successful. No issues seen while booting into the OS. Signed-off-by:
Mahesh Rajashekhara <Mahesh.Rajashekhara@pmcs.com> Reviewed-by:
Tomas Henzl <thenzl@redhat.com> Signed-off-by:
-
Dan Carpenter authored
pthru32->dataxferlen comes from the user so we need to check that it's not too large so we don't overflow the buffer. Reported-by:
Nico Golde <nico@ngolde.de> Reported-by:
Fabian Yamaguchi <fabs@goesec.de> Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Acked-by:
Sumit Saxena <sumit.saxena@lsi.com> Signed-off-by:
-
Vaughan Cao authored
This patch makes reject messages show right value for opcode and itt, which is converse previously. Signed-off-by:
Vaughan Cao <vaughan.cao@oracle.com> Acked-by:
Mike Christie <michaelc@cs.wisc.edu> Signed-off-by:
-
Vijaya Mohan Guvva authored
Fixed following smatch warnings in bfa. drivers/scsi/bfa/bfa_ioc.c:3882 bfa_sfp_show_comp() error: memcpy() 'des' too small (64 vs 248) drivers/scsi/bfa/bfa_ioc.c:6859 bfa_flash_status_read() warn: unsigned 'status' is never less than zero. drivers/scsi/bfa/bfa_ioc.c:6881 bfa_flash_status_read() warn: unsigned 'status' is never less than zero. drivers/scsi/bfa/bfa_ioc.c:6917 bfa_flash_read_start() warn: unsigned 'status' is never less than zero. drivers/scsi/bfa/bfa_ioc.c:7043 bfa_flash_raw_read() warn: unsigned 'status' is never less than zero. Signed-off-by:
Vijaya Mohan Guvva <vmohan@brocade.com> Signed-off-by:
-
Wei Yongjun authored
Add the missing unlock before return from function bfad_iocmd_cfg_trunk() in the error handling case. Signed-off-by:
Wei Yongjun <yongjun_wei@trendmicro.com.cn> Acked-by:
-
Hannes Reinecke authored
When the host template doesn't declare an eh_host_reset_handler the eh_deadline mechanism is pointless and will set the device to offline. So disable eh_deadline if no eh_host_reset_handler is present. Signed-off-by:
-
James Bottomley authored
In the highly unusual case where two threads are running concurrently through the scanning code scanning the same target, we run into the situation where one may allocate the target while the other is still using it. In this case, because the reap checks for STARGET_CREATED and kills the target without reference counting, the second thread will do the wrong thing on reap. Fix this by reference counting even creates and doing the STARGET_CREATED check in the final put. Tested-by:
Sarah Sharp <sarah.a.sharp@linux.intel.com> Cc: stable@vger.kernel.org # delay backport for 2 months for field testing Signed-off-by:
-
James Bottomley authored
This patch eliminates the reap_ref and replaces it with a proper kref. On last put of this kref, the target is removed from visibility in sysfs. The final call to scsi_target_reap() for the device is done from __scsi_remove_device() and only if the device was made visible. This ensures that the target disappears as soon as the last device is gone rather than waiting until final release of the device (which is often too long). Reviewed-by:
Alan Stern <stern@rowland.harvard.edu> Tested-by:
-
Bradley Grove authored
Signed-off-by:
Bradley Grove <bgrove@attotech.com> Signed-off-by:
-
