- 15 Mar, 2014 40 commits
-
-
adam radford authored
The following patch for megaraid_sas will return leaked MPT frames from any polled DCMD's that timeout to the MPT frame pool. Signed-off-by: Adam Radford <aradford@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
adam radford authored
The following patch for megaraid_sas fixes the megasas_ioc_init_fusion function to use a local stack variable for the IOCinit frame physical address instead of clobbering the first request descriptor. Signed-off-by: Adam Radford <aradford@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
adam radford authored
The following patch for megaraid_sas loads the correct raid context timeout value for multpathing and clustering. Signed-off-by: Adam Radford <aradford@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Sumit.Saxena@lsi.com authored
Host lock is added back around queuecommand. Host lock removal can create race conditon between ISR path(when RAID map update interrupt is raised) and IO build path of driver, since IO build path is making use of RAID map, and in case of RAID map update interrupt, old RAID map copy is memset to zero, which some IOs may be referencing in build IO path. Changes done for performance boost- 1) Added code to set SMP IRQ affinity per CPU. 2) Pass MSI-x index, while issuing sysPD IO. Signed-off-by: Kashyap Desai <kashyap.desai@lsi.com> Signed-off-by: Sumit Saxena <sumit.saxena@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Sumit.Saxena@lsi.com authored
If consistent DMA mask is set to 64 bit, fall back to 32bit DMA mask and 32bit consistent DMA mask. 64bit consistent DMA mask may be set on some 64bit DMA slot, which causes DMA offset "10000000000000" and MFI_INIT and IOCTL frames will have high memory addresses, leads to firmware FAULT. Signed-off-by: Kashyap Desai <kashyap.desai@lsi.com> Signed-off-by: Sumit Saxena <sumit.saxena@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Sumit.Saxena@lsi.com authored
Signed-off-by: Sumit Saxena <sumit.saxena@lsi.com> Signed-off-by: Kashyap Desai <kashyap.desai@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Sumit.Saxena@lsi.com authored
Don't wait forever for firmware response for internal DCMDs sent from driver firmware. Such DCMDs will be posted to firmware with timeout. Timeout is also introduced for DCMD sent to abort the commands. DCMD sent via IOCTL path will still be always blocking to keep the IOCTL design intact. Signed-off-by: Sumit Saxena <sumit.saxena@lsi.com> Signed-off-by: Kashyap Desai <kashyap.desai@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Dan Carpenter authored
The original code always set the upper 32 bits to zero because it was doing a shift of the wrong variable. Fixes: 1a4f550a ('[SCSI] arcmsr: 1.20.00.15: add SATA RAID plus other fixes') Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Eiichi Tsukata authored
Currently, scsi error handling in scsi_io_completion() tries to unconditionally requeue scsi command when device keeps some error state. For example, UNIT_ATTENTION causes infinite retry with action == ACTION_RETRY. This is because retryable errors are thought to be temporary and the scsi device will soon recover from those errors. Normally, such retry policy is appropriate because the device will soon recover from temporary error state. But there is no guarantee that device is able to recover from error state immediately. Some hardware error can prevent device from recovering. This patch adds timeout in scsi_io_completion() to avoid infinite command retry in scsi_io_completion(). Once scsi command retry time is longer than this timeout, the command is treated as failure. Signed-off-by: Eiichi Tsukata <eiichi.tsukata.xh@hitachi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Mike Christie authored
Check that the session is setup before accessing its connection. This fixes a oops where userspace tries to get the ip address before the session is bound to a host. Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Ethan Zhao authored
Load and unload iscsi_ibft module will cause kernel memory leak, fix it in scsi/iscsi_boot_sysfs.c iscsi_boot_destroy_kset(). Signed-off-by: Ethan Zhao <ethan.kernel@gmail.com> Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Mike Christie authored
iscsi_queuecommand will only take in commands that can fit in the current window. So, if a command is on the cmdqueue then it can fit in the current window. If a command is on the mgmtqueue, then we are setting the immediate bit so they will also fit in the window. As a result, we never need to to do a iscsi_conn_queue_work when the maxCmdSn is increased. What should happen is that a command will complete the window will be increased, then the scsi layer will send us more commands by running the scsi_device queues. Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Shlomo Pongratz authored
Replace the session lock with two locks, a forward lock and a backwards lock named frwd_lock and back_lock respectively. The forward lock protects resources that change while sending a request to the target, such as cmdsn, queued_cmdsn, and allocating task from the commands' pool with kfifo_out. The backward lock protects resources that change while processing a response or in error path, such as cmdsn_exp, cmdsn_max, and returning tasks to the commands' pool with kfifo_in. Under a steady state fast-path situation, that is when one or more processes/threads submit IO to an iscsi device and a single kernel upcall (e.g softirq) is dealing with processing of responses without errors, this patch eliminates the contention between the queuecommand()/request response/scsi_done() flows associated with iscsi sessions. Between the forward and the backward locks exists a strict locking hierarchy. The mutual exclusion zone protected by the forward lock can enclose the mutual exclusion zone protected by the backward lock but not vice versa. For example, in iscsi_conn_teardown or in iscsi_xmit_data when there is a failure and __iscsi_put_task is called, the backward lock is taken while the forward lock is still taken. On the other hand, if in the RX path a nop is to be sent, for example in iscsi_handle_reject or __iscsi_complete_pdu than the forward lock is released and the backward lock is taken for the duration of iscsi_send_nopout, later the backward lock is released and the forward lock is retaken. libiscsi_tcp uses two kernel fifos the r2t pool and the r2t queue. The insertion and deletion from these queues didn't corespond to the assumption taken by the new forward/backwards session locking paradigm. That is, in iscsi_tcp_clenup_task which belongs to the RX (backwards) path, r2t is taken out from r2t queue and inserted to the r2t pool. In iscsi_tcp_get_curr_r2t which belong to the TX (forward) path, r2t is also inserted to the r2t pool and another r2t is pulled from r2t queue. Only in iscsi_tcp_r2t_rsp which is called in the RX path but can requeue to the TX path, r2t is taken from the r2t pool and inserted to the r2t queue. In order to cope with this situation, two spin locks were added, pool2queue and queue2pool. The former protects extracting from the r2t pool and inserting to the r2t queue, and the later protects the extracing from the r2t queue and inserting to the r2t pool. Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> [minor fix up to apply cleanly and compile fix] Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Shlomo Pongratz authored
Restructure the iscsi_tcp_r2t_rsp routine in order to avoid allocating r2t from r2tpool.queue and returning it back in case the parameters rhdr->data_length and or rhdr->data_offset prohibit the requing. Since the values of these parameters are known prior to the allocation, we can pre-check and thus avoid futile allocations. [jejb: checkpatch fixes] Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Karen Xie authored
Fixed calculates wrong tuple values on T5 adapter: switch to use the exported API cxgb4_select_ntuple() from cxgb4 base driver. Signed-off-by: Karen Xie <kxie@chelsio.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Hannes Reinecke authored
When megasas_get_pd_list() fails we cannot detect any drives, so we should be checking the return value accordingly. [jejb: checkpatch fix] Signed-off-by: Hannes Reinecke <hare@suse.de> Acked-by: Kashyap Desai <kashyap.desai@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Hannes Reinecke authored
When no HBA is found we should be returning '-ENXIO' to be consistent with the other return values. Signed-off-by: Hannes Reinecke <hare@suse.de> Acked-by: Kashyap Desai <kashyap.desai@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Hannes Reinecke authored
The pointer to the queue info structure is potentially a 64-bit value, so we should be using the correct macros to set the values in the init frame. Signed-off-by: Hannes Reinecke <hare@suse.de> Acked-by: Kashyap Desai <kashyap.desai@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Lukasz Dorau authored
The version of isci driver has not been updated for 2 years. It was 83 isci commits ago. Suspend/resume support has been implemented and many bugs have been fixed since 1.1. Now update the version to 1.2. Signed-off-by: Lukasz Dorau <lukasz.dorau@intel.com> Signed-off-by: Dave Jiang <dave.jiang@intel.com> Signed-off-by: Maciej Patelczyk <maciej.patelczyk@intel.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Hannes Reinecke authored
For various error conditions the bfa driver just returns 'DID_ERROR', which carries no information at all about the actual source of error. This patch updates the error handling to return a correct error code, depending on the type of error occurred. Signed-off-by: Hannes Reinecke <hare@suse.de> Acked-by: Vijaya Mohan Guvva <vmohan@brocade.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Fix following sparse warnings:- drivers/scsi/qla4xxx/ql4_os.c:2109:33: warning: cast truncates bits from constant value (ffff7fff becomes 7fff) drivers/scsi/qla4xxx/ql4_os.c:2306:33: warning: cast truncates bits from constant value (ffff7fff becomes 7fff) Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Nilesh Javali authored
Signed-off-by: Nilesh Javali <nilesh.javali@qlogic.com> Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Removing unused code as FW does not need any value in mbox-5. Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Issue: While unloading driver MBOX 0x31 fail as DDB logout (MBOX 0x56) operation is not completed. Fix: Wait for DDB Logout completion before MBOX 0x31 Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Issue: Driver holds rom-lock for too long during reset recovery. During adapter reset testing, it was found that the driver holds the rom-lock for too long, because of which other drivers fail to acquire the rom-lock, leading to reset failures. The primary cause is, in the bootstrap code, while holding the rom-lock, the driver checks if the peg is halted, causing a 2 second contention. Fix: When a reset recovery starts, the driver deduces the cause, and sets appropriate flags in watchdog & recover_adapter routines. This flag should be used to determine if bootstrap is invoked from probe or reset context, reducing the rom-lock footprint of the drivers. Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Tej Parkash authored
Issue: minidump data collection fails as driver reports data mismatch Fix: When the driver encounters a new entry type that it cannot process, it should just skip the entry and adjust the total buffer size by subtracting the skipped bytes from it. This is to ensure that there is no data mismatch because of the new entries. Signed-off-by: Tej Parkash <tej.parkash@qlogic.com> Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Tej Parkash authored
Issue: Pending IO wait does not complete after triggering Graceful reset, causing ack timeout and call traces. Fix: 1. Reducing the IO command wait timeout before triggering reset, as logically also timeout should be less than reset timeout (10sec). 2. Moving the abort IO after chip reset, because only after chip reset, driver owns the IO otherwise it is with firmware and can still revert back with response. Signed-off-by: Tej Parkash <tej.parkash@qlogic.com> Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Tej Parkash authored
Issue: While booting with kdump kernel, driver receive IOCB interrupts for which it is not ready which results in processing them before init_firmware during driver probe Fix: Two steps solution 1. Make driver ready to process the interrupt before interupts handlers is registered. 2. Stop driver processing iocb interrupts if not generated as per firmware protocol i.e R2H bit set Signed-off-by: Tej Parkash <tej.parkash@qlogic.com> Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Issue: Mailbox command 0x102 (IDC Time Extend) failure seen while applying changes to iface using iscsiadm Fix: Added fix to extend IDC timeout only for ISP84xx when IDC ACK needs to be posted and disable ACB mailbox command completion is in intermediate state. Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Nilesh Javali authored
Issue: qla4xxx Unable to clear DDB indices when logout fails due to failure of connection close mbox command. Root cause: If login to session fail, iscsiadm make call to destroy_session. qla4xxx driver does not free ddb index map before free_ddb() Fix: Clear DDB Index map before free_ddb in "destroy_session" in case of connection close mailbox command failure with 4005h. Signed-off-by: Nilesh Javali <nilesh.javali@qlogic.com> Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Tej Parkash authored
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Vikas Chaudhary authored
If waiting for signals was interrupted then the device was put to FAILED state. Use msleep instead of msleep_interruptible to handle this correctly. Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Nilesh Javali authored
Signed-off-by: Nilesh Javali <nilesh.javali@qlogic.com> Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-
Nilesh Javali authored
Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
-