summaryrefslogtreecommitdiffstats
path: root/drivers/scsi
Commit message (Collapse)AuthorAgeFilesLines
* [SCSI] qla4xxx: Modified debug log messages for boot info.Manish Rangankar2012-02-191-14/+12
| | | | | | | | | | | | | In some configurations user may not have boot targets configured. In such cases the debug messages printed out by driver look like some kind of failure happening. However this could be a valid case, so modified the messages to appear as warning messages versus failure messages. Signed-off-by: Manish Rangankar <manish.rangankar@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>
* [SCSI] qla4xxx: Fix verify boot idx correctlyLalit Chandivade2012-02-191-3/+2
| | | | | | | | | | | | | | qla4xxx_verify_boot_idx can falsely report a DDB to be boot target if ha->pri_ddb_idx and ha->sec_ddb_idx are not initialized correctly. What this could cause is if there is DDB entry in FLash at index 0, then qla4xxx_verify_boot_idx would return wrong result as ha->pri_ddb_idx is not set correctly. Fixed the qla4xxx_get_boot_info to set the ha->pri_ddb_idx and ha->sec_ddb_idx correctly. Signed-off-by: Lalit Chandivade <lalit.chandivade@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>
* [SCSI] qla4xxx: Fix un-necessary delay on invalid DDBLalit Chandivade2012-02-192-0/+6
| | | | | | | | | | | | Fix the un-necessary wait for completion of a sendtarget on an invalid DDB entry. The state of an invalid DDB entry is 0 (unassigned) This will also avoid the delays during system boot. Signed-off-by: Lalit Chandivade <lalit.chandivade@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>
* [SCSI] qla4xxx: Remove unused codeVikas Chaudhary2012-02-191-12/+0
| | | | | | | | | This code initially added for FW debugging, we don't need this code now so taking it out. 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>
* [SCSI] libfc: Handle discovery failure during ctlr link downBhanu Prakash Gollapudi2012-02-191-0/+7
| | | | | | | | | | | | | While we wait for GPN_FT response, if the ctlr link goes down, the stack generates a completion for GPN_FT with error FC_EXCH_CLOSED, and reports a discovery error. Discovery is not retried in this case, and rightly so. However, the 'pending' flag stays set, which does not allow subsequent discovery to succeed as GPN_FT will never be issued. Fix it by clearing the pending flag when the discovery fails due to GPN_FT failure. Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com> Signed-off-by: Robert Love <robert.w.love@intel.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] libfc: Fix panic in fc_exch_recvBhanu Prakash Gollapudi2012-02-191-3/+4
| | | | | | | | | | | | | | | | | | | | | | | Adding and removing the host into the zone causes this panic. BUG: unable to handle kernel NULL pointer dereference at 00000000000000a0 IP: [<ffffffffa0491707>] fc_exch_recv+0xc57/0xe70 [libfc] Call Trace: [<ffffffffa050e04b>] bnx2fc_l2_rcv_thread+0x37b/0x430 [bnx2fc] [<ffffffffa050dcd0>] ? bnx2fc_l2_rcv_thread+0x0/0x430 [bnx2fc] [<ffffffff81090886>] kthread+0x96/0xa0 [<ffffffff8100c14a>] child_rip+0xa/0x20 [<ffffffff810907f0>] ? kthread+0x0/0xa0 [<ffffffff8100c140>] ? child_rip+0x0/0x20 During fc_exch_reset, the active exchanges are aborted and the exch is deleted. As part of processing ABTS response, due to 'ep' being NULL, any access to ep in fc_exch_recv_bls() causes this panic. Fixed to access 'ep' only if non-NULL. Reviewed-by: Neerav Parikh <neerav.parikh@intel.com> Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com> Signed-off-by: Robert Love <robert.w.love@intel.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] fcoe: Remove reference counting on 'stuct fcoe_interface'Robert Love2012-02-192-46/+5
| | | | | | | | | | | | The reference counting was necessary on these instances because it was possible for NPIV ports to be destroyed after the N_Port. A previous patch ensures that all NPIV ports are destroyed before the N_Port making the need to track references on the interface unnecessary. Signed-off-by: Robert Love <robert.w.love@intel.com> Tested-by: Ross Brattain <ross.b.brattain@intel.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] fcoe: Do not switch context in vport_delete callbackRobert Love2012-02-191-9/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently all port deletion is routed though the FCoE workqueue (fcoe_wq). When fc_remove_host is called on an N_Port (for example, from fcoe_destroy) the vports are queued into a FC Transport workqueue. fc_remove_host flushes that queue and each vport is passed to fcoe's fcoe_vport_destroy, which simply queues the associated fcoe_ports for later deletion. This queue cannot be flushed within the N_Ports destroy path because of circular locking issues. The result is that the NPIV ports are destroyed after the N_Port, which is reverse of how they are created. This quirk causes fcoe to keep references on the fcoe_interface shared by each of these ports (N_Port and NPIV). Changing the ordering such that NPIV ports are destroyed before the N_Port will allow us to remove reference counting on the fcoe_interface instances. This patch simply allows fcoe_vport_destory to destroy NPIV ports without deferring them to a workqueue context. This ensures that when fc_remove_host is called the NPIV ports will be destroyed first before the N_Port and allows reference counting on the fcoe's fcoe_interface to be remove in a later patch. Signed-off-by: Robert Love <robert.w.love@intel.com> Tested-by: Ross Brattain <ross.b.brattain@intel.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] fcoe: Rename out_nomod label to out_putmodRobert Love2012-02-191-3/+3
| | | | | | | | | | | | | | | | The label implies that it should be called when there is 'nomod.' I read that to mean that the module reference 'get' failed. However, it's only called when the module reference 'get' succeeded. I think it makes more sense to name the label, 'out_putmod' since it should be called when we need to 'put' the module reference taken in the routine before returning. Signed-off-by: Robert Love <robert.w.love@intel.com> Tested-by: Ross Brattain <ross.b.brattain@intel.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] fcoe: Allow exposing FDMI attributes via sysfsNeerav Parikh2012-02-191-0/+8
| | | | | | | | | | Allow FDMI attributes to be exposed via the fc_host class object for the fcoe driver. Signed-off-by: Neerav Parikh <neerav.parikh@intel.com> Tested-by: Ross Brattain <ross.b.brattain@intel.com> Signed-off-by: Robert Love <robert.w.love@intel.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] libfcoe: Don't KERN_ERR on netdev notificationRobert Love2012-02-191-2/+2
| | | | | | | | | | | This is more of a debug statement. As a KERN_ERR we generate log entries anytime any netdev goes up or down, so when booting there are notification log entries for all system interfaces including 'lo'. This is too much. Let's just log when necessary. Signed-off-by: Robert Love <robert.w.love@intel.com> Tested-by: Ross Brattain <ross.b.brattain@intel.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] isci: T10 DIF supportDave Jiang2012-02-193-16/+193
| | | | | | | | | | | | | | | This allows the controller to do WRITE_INSERT and READ_STRIP for SAS disks that support protection information. SAS disks must be formatted with protection information to use this feature via sg_format. sg3_utils-1.32 -- sg_format version 1.19 20110730 sg_format usage: sg_format --format --verbose --pinfo /dev/sda Acked-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Dave Jiang <dave.jiang@intel.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Avoid invalid request queue dereference for bad response ↵Arun Easi2012-02-191-11/+17
| | | | | | | | packets. Signed-off-by: Arun Easi <arun.easi@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Stop iteration after first failure in *_id functions.Arun Easi2012-02-191-0/+6
| | | | | | Signed-off-by: Arun Easi <arun.easi@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Fix incorrect register access in qla2x00_start_iocbs().Arun Easi2012-02-191-1/+1
| | | | | | Signed-off-by: Arun Easi <arun.easi@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Handle device mapping changes due to device logout.Arun Easi2012-02-192-0/+27
| | | | | | | | | | | | A device logout sent in the delete path of a fcport would clear the port handle binding inside the firmware. This could lead to queued work items for the fcport, if any, getting incorrect results. This patch fixes the issue by checking for device name changes after a call to get port database. Signed-off-by: Arun Easi <arun.easi@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Add ha->max_fibre_devices to keep track of the maximum ↵Chad Dupuis2012-02-197-48/+77
| | | | | | | | | | | | number of targets. Add a field to the qla_hw_data struct to allow us to set the maximum number of fabric devices on a per adapter basis based on ISP type. [jejb: fix up missing rval = QLA_SUCCESS to prevent uninit var warning] Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Cache swl during fabric discovery.Andrew Vasquez2012-02-193-7/+8
| | | | | | | | | | Rather than continuously allocating and freeing swl within the discovery process, simply pre-allocate it the first time that it's needed, cache it through the rest of the lifecycle of the driver and free it at module unload. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Remove EDC sysfs interface.Joe Carnuccio2012-02-194-149/+4
| | | | | | Signed-off-by: Joe Carnuccio <joe.carnuccio@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Perform firmware dump procedure on mailbox command timeout.Chad Dupuis2012-02-191-0/+6
| | | | | | Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Change the log message when previous dump is available to ↵Giridhar Malavali2012-02-191-2/+3
| | | | | | | | retrieve for ISP82xx. Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Log messages to use correct vha.Arun Easi2012-02-192-47/+47
| | | | | | Signed-off-by: Arun Easi <arun.easi@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Add new message when a new loopid is assigned.Arun Easi2012-02-192-1/+4
| | | | | | Signed-off-by: Arun Easi <arun.easi@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Fix ql_dbg arguments.Arun Easi2012-02-191-2/+2
| | | | | | Signed-off-by: Arun Easi <arun.easi@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Use ql_log* #define's in ql_log() and ql_log_pci().Chad Dupuis2012-02-191-6/+6
| | | | | | Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Convert remaining printk's to ql_log format.Chad Dupuis2012-02-193-16/+16
| | | | | | Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Print mailbox command opcode and return code when a command ↵Chad Dupuis2012-02-191-7/+9
| | | | | | | | times out. Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Reduce mbx-command timeout for Login/Logout requests.Andrew Vasquez2012-02-191-3/+4
| | | | | | | | | | Don't use default 30 second mailbox-command timeout for these serial requests, instead, limit the TMO to the standard 2*RATOV plus some fudge-factor. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Prep zero-length BSG data-transfer requests.Andrew Vasquez2012-02-191-17/+7
| | | | | | | | | | | During command failure/non-recognition, the upper-layer FC-transport expects the drivers to set job-reply->reply_payload_rcv_len. Do this in a consistent manner to avoid duplication. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Perform implicit logout during rport tear-down.Andrew Vasquez2012-02-195-5/+51
| | | | | | | | | During rport tear-down, make sure we do an implicit LOGO of the fcport in our firmware to try to clear any residual commands associated with that fcport. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Handle failure cases during fabric_loginChad Dupuis2012-02-193-19/+42
| | | | | | | | | Make sure that all calls to ha->isp_ops->fabric_login() check the return value for failure. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Increase speed of flash access in ISP82xx adapters to ↵Chad Dupuis2012-02-191-40/+27
| | | | | | | | improve firmware load speed. Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Return blank sysfs string on initial get thermal failure.Joe Carnuccio2012-02-191-1/+1
| | | | | | | | | | | When thermal temperature initially fails, return a blank string to the sysfs interface. This fixes the initial display of 0.00 followed by subsequent display of blank line; the initial 0.00 should have not displayed for cards that do not support thermal temperature. Signed-off-by: Joe Carnuccio <joe.carnuccio@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Handle change notifications based on switch scan results.Arun Easi2012-02-193-144/+13
| | | | | | | | | Instead of processing each RSCN individually, use only the name server results from the switch to tell the existance of a given fcport. Signed-off-by: Arun Easi <arun.easi@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Correct print format for edc ql_log() calls.Joe Carnuccio2012-02-191-2/+2
| | | | | | Signed-off-by: Joe Carnuccio <joe.carnuccio@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Use consistent DL mask for ELS/CT passthru requests.Andrew Vasquez2012-02-191-4/+4
| | | | | | | | | | | | The driver is logging a slew of 'good' status requests for ELS/CT passthrough commands. Change some log messages from: * ql_log() -> ql_dbg() * ql_log_info -> ql_dbg_user Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Consolidation of SRB processing.Giridhar Malavali2012-02-1911-628/+485
| | | | | | | | | | Rework the structures related to SRB processing to minimize the memory allocations per I/O and manage resources associated with and completions from common routines. Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Use a valid enode-mac if none defined.Andrew Vasquez2012-02-191-3/+3
| | | | | | | | Original 'defaults' were not OUI valid. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Enhancements to support ISP83xx.Giridhar Malavali2012-02-1917-170/+1233
| | | | | | | | | | Signed-off-by: Saurav Kashyap <saurav.kashyap@qlogic.com> Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: Harish Zunjarrao <harish.zunjarrao@qlogic.com> Signed-off-by: Nigel Kirkland <nigel.kirkland@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] qla2xxx: Enhanced the dump routines to capture multiple request and ↵Giridhar Malavali2012-02-193-3/+105
| | | | | | | | response queues. Signed-off-by: Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] aacraid: Added Sync.mode to support series 7/8/9 controllersMahesh Rajashekhara2012-02-198-63/+338
| | | | | | | | | | Added Sync. mode to support Series 7/8/9 controller families: This is a compatibility mode for all these controller families. The Async. (Performance) mode can be changed in the future. First Async. mode version added for Series 7; Controller parameter aac_sync_mode added Signed-off-by: Mahesh Rajashekhara <aacraid@pmc-sierra.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] bnx2i: Fixed the override of the error_mask module paramEddie Wai2012-02-192-6/+10
| | | | | | | | | | | | The error_mask module param overrides has a bug which prevented the new module param values to take effect. Also changed the type attribute of the error_mask1/2 module params from int to uint to allow the MSB to be set. Signed-off-by: Eddie Wai <eddie.wai@broadcom.com> Acked-by: Anil Veerabhadrappa <anilgv@broadcom.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] bnx2i: use kthread_create_on_node()Eric Dumazet2012-02-191-2/+3
| | | | | | | | | bnx2i_percpu_thread_create() create per cpu kthread, and should use proper NUMA aware API. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Acked-by: Eddie Wai <eddie.wai@broadcom.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] bfa: don't leak mem in bfad_im_bsg_els_ct_request()Jesper Juhl2012-02-191-2/+2
| | | | | | | | | | | | | | | If 'drv_fcxp = kzalloc(sizeof(struct bfad_fcxp), GFP_KERNEL);' fails and returns NULL, then we'll leak the memory allocated to 'bsg_fcpt' when we jump to 'out:' and the variable subsequently goes out of scope. Also remove the cast of the kzalloc() return value. kzalloc() returns a void* which is implicitly converted, so the explicit cast is pointless. Signed-off-by: Jesper Juhl <jj@chaosbits.net> Acked-by: Krishna Gudipati <kgudipat@brocade.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] isci: enable clock gatingMarcin Tomczak2012-02-192-2/+34
| | | | | | | | | | | | Enabling clock gating for power savings on entry to controller ready state. Disable SCU clock gating for power savings on exit from the controller ready state. The gating is fully automated by silicon after setting the mode. Signed-off-by: Marcin Tomczak <marcin.tomczak@intel.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] libiscsi: fix cmd timeout/completion raceMike Christie2012-02-191-10/+10
| | | | | | | | | | | | | | | | If the driver/lib has called scsi_done and cleaned up internally but scsi layer has not yet called blk_mark_rq_complete when the command times out we hit a problem if the timeout code calls blk_mark_rq_complete first. When the time out code calls into the driver we were returning BLK_EH_RESET_TIMER and that causes the timeout code to just call us again later. We need to be calling BLK_EH_HANDLED so the timeout code can complete the completion process because it had called blk_mark_rq_complete on the command and now owns its processing. Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] libiscsi_tcp: fix max_r2t manipulationMike Christie2012-02-194-24/+22
| | | | | | | | | | Problem description from Xi Wang: A large max_r2t could lead to integer overflow in subsequent call to iscsi_tcp_r2tpool_alloc(), allocating a smaller buffer than expected and leading to out-of-bounds write. Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] iscsi: fix setting of pid from netlink skbMike Christie2012-02-191-2/+2
| | | | | | | | | | | NETLINK_CREDS's pid now returns 0, so I guess we are supposed to be using NETLINK_CB. This changed while the patch to export the pid was getting merged upstream, so it was not noticed until both the network and iscsi changes were in the same tree. Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Cc: stable@kernel.org Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] iscsi: don't hang in endless loop if no targets presentSasha Levin2012-02-191-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | iscsi_if_send_reply() may return -ESRCH if there were no targets to send data to. Currently we're ignoring this value and looping in attempt to do it over and over, which will usually lead in a hung task like this one: [ 4920.817298] INFO: task trinity:9074 blocked for more than 120 seconds. [ 4920.818527] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 4920.819982] trinity D 0000000000000000 5504 9074 2756 0x00000004 [ 4920.825374] ffff880003961a98 0000000000000086 ffff8800001aa000 ffff8800001aa000 [ 4920.826791] 00000000001d4340 ffff880003961fd8 ffff880003960000 00000000001d4340 [ 4920.828241] 00000000001d4340 00000000001d4340 ffff880003961fd8 00000000001d4340 [ 4920.833231] [ 4920.833519] Call Trace: [ 4920.834010] [<ffffffff826363fa>] schedule+0x3a/0x50 [ 4920.834953] [<ffffffff82634ac9>] __mutex_lock_common+0x209/0x5b0 [ 4920.836226] [<ffffffff81af805d>] ? iscsi_if_rx+0x2d/0x990 [ 4920.837281] [<ffffffff81053943>] ? sched_clock+0x13/0x20 [ 4920.838305] [<ffffffff81af805d>] ? iscsi_if_rx+0x2d/0x990 [ 4920.839336] [<ffffffff82634eb0>] mutex_lock_nested+0x40/0x50 [ 4920.840423] [<ffffffff81af805d>] iscsi_if_rx+0x2d/0x990 [ 4920.841434] [<ffffffff810dffed>] ? sub_preempt_count+0x9d/0xd0 [ 4920.842548] [<ffffffff82637bb0>] ? _raw_read_unlock+0x30/0x60 [ 4920.843666] [<ffffffff821f71de>] netlink_unicast+0x1ae/0x1f0 [ 4920.844751] [<ffffffff821f7997>] netlink_sendmsg+0x227/0x350 [ 4920.845850] [<ffffffff821857bd>] ? sock_update_netprioidx+0xdd/0x1b0 [ 4920.847060] [<ffffffff82185732>] ? sock_update_netprioidx+0x52/0x1b0 [ 4920.848276] [<ffffffff8217f226>] sock_aio_write+0x166/0x180 [ 4920.849348] [<ffffffff810dfe41>] ? get_parent_ip+0x11/0x50 [ 4920.850428] [<ffffffff811d0d9a>] do_sync_write+0xda/0x120 [ 4920.851465] [<ffffffff810dffed>] ? sub_preempt_count+0x9d/0xd0 [ 4920.852579] [<ffffffff810dfe41>] ? get_parent_ip+0x11/0x50 [ 4920.853608] [<ffffffff81791887>] ? security_file_permission+0x27/0xb0 [ 4920.854821] [<ffffffff811d0f4c>] vfs_write+0x16c/0x180 [ 4920.855781] [<ffffffff811d104f>] sys_write+0x4f/0xa0 [ 4920.856798] [<ffffffff82638e79>] system_call_fastpath+0x16/0x1b [ 4920.877487] 1 lock held by trinity/9074: [ 4920.878239] #0: (rx_queue_mutex){+.+...}, at: [<ffffffff81af805d>] iscsi_if_rx+0x2d/0x990 [ 4920.880005] Kernel panic - not syncing: hung_task: blocked tasks Signed-off-by: Sasha Levin <levinsasha928@gmail.com> Acked-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
* [SCSI] fix the new host byte settings (DID_TARGET_FAILURE and DID_NEXUS_FAILURE)Moger, Babu2012-02-192-4/+4
| | | | | | | | | | | | | | | | | | This patch fixes the host byte settings DID_TARGET_FAILURE and DID_NEXUS_FAILURE. The function __scsi_error_from_host_byte, tries to reset the host byte to DID_OK. But that does not happen because of the OR operation. Here is the flow. scsi_softirq_done-> scsi_decide_disposition -> __scsi_error_from_host_byte Let's take an example with DID_NEXUS_FAILURE. In scsi_decide_disposition, result will be set as DID_NEXUS_FAILURE (=0x11). Then in __scsi_error_from_host_byte, when we do OR with DID_OK. Purpose is to reset it back to DID_OK. But that does not happen. This patch fixes this issue. Signed-off-by: Babu Moger <babu.moger@netapp.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
OpenPOWER on IntegriCloud