- 11 Apr, 2010 40 commits
-
-
Avi Kaplan authored
This patch changes the arguments to iscsi_sw_tcp_conn_restore_callbacks, so that it works like the function to set the callbacks and because in upcoming patches we need a iscsi_conn. Signed-off-by:
Avi Kaplan <savik751@gmail.com> Signed-off-by:
Mike Christie <michaelc@cs.wisc.edu> Signed-off-by:
James Bottomley <James.Bottomley@suse.de>
-
Mike Christie authored
blk_get_request sets the cmd_flags, so we should not and do not need to set them. If we did set them to a different value then it can cause a oops in the elevator code. Signed-off-by:
-
Dan Carpenter authored
This change is basically a clean up. datadgst_en is an int which comes from the user. I didn't see anything limiting it to 1 and 0 although obviously that's what it's supposed to be. If the user passed in 2 this test would fail. This same if condition is repeated in another function and it uses && there. Signed-off-by:
Dan Carpenter <error27@gmail.com> Signed-off-by:
-
Chris Leech authored
Print all world wide node names (node, port and fabric) with the same format specifier of "%16.16llx". That makes sure they all print as a 16 character hex string, with lower case letters, no 0x prefix, and without stripping off any leading 0s. Signed-off-by:
Chris Leech <christopher.leech@intel.com> Signed-off-by:
Robert Love <robert.w.love@intel.com> Signed-off-by:
-
Chris Leech authored
Most of the prints of fabric IDs were specified as %6x, which will not print any leading 0s. It's nice to see leading 0s for identifiers like this, which are a fixed length. This patch sets the precision modifier as well, making the specifier %6.6x, which forces the printing of leading 0s. Signed-off-by:
-
Yi Zou authored
fc_fcp_resp is assuming when FCP_SNS_LEN_VAL is set, the FCP_RSP_LEN_VAL is not, which is not true. This leads to not copying the sense data and error out a valid FCP_RSP. Signed-off-by:
Yi Zou <yi.zou@intel.com> Signed-off-by:
-
Vasu Dev authored
No reason to restrict CDB size to 12 bytes in fcoe, so increased to 16 so that 16 bytes SCSI CDB doesn't fail. Uses common define to set max_cmd_len for fcoe and fnic, fnic is already setting max_cmd_len to 16. sg_readcap -l fails without this fix. Signed-off-by:
Vasu Dev <vasu.dev@intel.com> Signed-off-by:
-
Vasu Dev authored
Signed-off-by:
-
Brian Uchino authored
fnic_flush_tx() is used to send frames held while fabric login is in progress. The frames are held in tx_queue, but fnic_flush_tx() was incorrectly flushing from recv_queue which is used for received frames. Signed-off-by:
Brian Uchino <buchino@cisco.com> Signed-off-by:
Abhijeet Joglekar <abjoglek@cisco.com> Signed-off-by:
-
Brian Uchino authored
New fnic version to mark inclusion of tx_flush bugfix. Signed-off-by:
-
Joe Eykholt authored
hton24(p + 3, value) would fail to compile because p + 3[0] is not a valid expression. Went ahead and converted hton24 and ntoh24 to inline functions, which is better because the parameters are evalutated only once. Signed-off-by:
Joe Eykholt <jeykholt@cisco.com> Signed-off-by:
-
Joe Eykholt authored
fc_exch_recv_req has variables eof, sof, and f_ctl, which are set but never used. Delete them. Signed-off-by:
-
Chris Leech authored
Allow for dormant states while link configuration completes. In the default link mode, this is equivalent to the old check. Signed-off-by:
-
Chris Leech authored
The FIP controler state wasn't being reset on a disable. A disable/enable sequence should be treated as a link event. Otherwise, when using disable to mask a time when the link is up but unusable, FCF discovery would attempt to continue and login would jump directly to the non-FIP fallback on enable. Signed-off-by:
-
Robert Love authored
FPMA indicates that the Fabric will provide the host's N_Port's MAC address. When sending a FLOGI/FDISC frame and FPMA was negotiated through FIP discovery we still need to provide the MAC descriptor, as per the specification, but the MAC should be zero'd out since the FCF will be providing it in the FLOGI/FDISC ACC. In FC-BB-5 section 7.8.7.4.2 (Fabric login) it states: The MAC address field in the MAC address descriptor of a FIP FLOGI Request operation or a FIP NPIV FDISC Request operation shall contain: a) the proposed MAC address to use as VN_Port MAC address if the ENode is requesting to use SPMA (see table 27); b) all zeroes to indicate no MAC address is proposed if the ENode is requesting to use FPMA (see table 27); or c) the proposed MAC address to use as VN_Port MAC address if the ENode supports both SPMA and FPMA and leaves the decision of which addressing scheme to use to the FCF (i.e., if both the FP and SP bits are set to one, see table 27). This patch fixes case B. This patch also adds debug statements to illustrate whether a FPMA or SPMA MAC is added to a FLOGI/FDISC frame. Signed-off-by:
-
Kashyap, Desai authored
Upgrade version to 05.100.00.02 Signed-off-by:
Kashyap Desai <kashyap.desai@lsi.com> Signed-off-by:
-
Kashyap, Desai authored
Fix a oops in _scsih_sas_device_remove. The driver was attempting to delete a object from the sas_device link list when the object was not present. Added sanity check for sas_device NULL dereference. before deleting sas_device now driver will search device in list then only it will follow device removal. Signed-off-by:
-
Kashyap, Desai authored
we added support to set the deleted flag prior to device scan, then clear the flag for responding devices, leaving the deleted flag only set for missing devices. The problem is for internal generated host resets, IO queues are not blocked at scsi mid layer level. IO will be continued sent to driver, and driver will return SCSI_MLQUEUE_HOST_BUSY. The problem is the driver checks for the deleted flag before it checks for the controller being in reset, so there is a window where the driver would be returning DID_NO_CONNECT for responding devices. This occurs during the time between calling _scsih_prep_device_scan, and _scsih_mark_responding_sas_device & _scsih_mark_responding_raid_device. Fix the queuecommand entry point so ioc->shost_recovery flag sanity check is given higher presidence then the device "deleted flag" check. Signed-off-by:
-
James Smart authored
Signed-off-by:
Alex Iannicelli <alex.iannicelli@emulex.com> Signed-off-by:
James Smart <james.smart@emulex.com> Signed-off-by:
-
James Smart authored
- Add code to prevent unreg_vpi mailbox command from failing. - Add code to reset the HBA if unreg_vpi mailbox fails with busy status. - Remove code that was clearing the nlp_type stored during rport discovery. Signed-off-by:
-
James Smart authored
- Add memcpy to mailbox completion to transfer reply correctly. - Add support for BSG mailbox commands (dump, update_cfg, and event_log_status). - Add warning message and refuse mailbox command while mgmt is blocked. - Add checks for memory allocation failure. - Add check for sli4 dump zero BDE size. - Only copy data if mailbox commands succeed. - Add support for Read Event Log mailbox command. Signed-off-by:
-
James Smart authored
- Move the code to increase the sg seg count for LP21000 adapters. - Check pcmd on command completion before dereferencing it. - Clear queue memory when creating firmware queues to prevent stale entries. - Replace the use of PAGE_SIZE in many areas that assumed it was always 4k. - Add an else clause to a conditional that needed to unlock the hba_lock. Signed-off-by:
-
James Smart authored
Update driver for change in T10-PI interface on adapter - Remove Profiles. - Add new SLI Opcodes. - Add new PDE structures used for BlockGuard. Signed-off-by:
-
James Smart authored
- Prevent log message 1801 during vport delete. - Enable NPIV by default. - Display correct value for max Vports on SLI4 HBAs. Signed-off-by:
-
James Smart authored
- Add the new Logical Link speed event support. - Add RATOV and EDTOV to the REG_VFI mailbox command. Signed-off-by:
-
Kashyap, Desai authored
Upgraded version string. Signed-off-by:
-
Kashyap, Desai authored
Added check before free_pages just to make sure ioc->scsi_lookup is not NULL. Signed-off-by:
-
Kashyap, Desai authored
Added support in the driver to support EEH and PCIe Advanced Error Recovery. This involves adding new pci_error_handler interface for recovering the controller from PCI Bus errors, such as SERR and PERR. Some tools are available for simulating PCI errors in order to validate this interface: http://www.kernel.org/pub/linux/utils/pci/aer-injectSigned-off-by:
-
Kashyap, Desai authored
RAID_SCSI_IO_PASSTHROUGH: Driver needs to be sending the default descriptor for RAID Passthru, currently its sending SCSI_IO descriptor. Signed-off-by:
-
Kashyap, Desai authored
There are few special cases which needs to be handled deleting old port. CASE1: In topology you need cascaded expanders. Through sysfs just make sure topology is up. Erase the manufacturing image of the cascaded expander and reset the board. In some cases Adapter will receive Exapnder Add event before expander delete. In such a case, driver needs to delete duplicate port before adding new port. CASE2: Enable Device Missing delay of HBA through lsiutils. If expander or end device is hotswapped with different device before DMD timer expires, driver will get device add for new device first and then device deletion event for the original devices will arrive later at DMD timer expires. In this case also driver need to delete duplicate port before adding port for new device. Added new function which will make sure when new port is added, that its not claiming the same phy resources already in use by another port. If it does, then it will delete the other port before adding the new port. Signed-off-by:
-
Douglas Gilbert authored
While testing the midlevel q_at_head and q_at_tail patch for sg and the block SG_IO ioctl I found it useful to reduce the queuing within the scsi_debug driver. The reason is that the midlevel queue only comes into play when the corresponding LLD queue is full. It is also useful when testing to be confident that your program is the only thing issuing commands to the (virtual) scsi_debug device. The no_uld=1 parameter will stop a scsi_debug virtual disk appearing as /dev/sd* . Changelog: - add max_queue parameter to reduce the number of queued commands the driver will accept. This parameter can be changed after the driver is loaded. - add no_uld parameter that restricts scsi_debug's virtual devices to the sg and bsg drivers - correct stale url Signed-off-by:Douglas Gilbert <dgilbert@interlog.com> Signed-off-by:
-
James Bottomley authored
Hazard testing uncovered yet another bug in sd. Under heavy reset activity the retry counter might be exhausted and the command will be returned with sense UNIT_ATTENTION/0x29/00 (POWER ON, RESET, OR BUS DEVICE RESET OCCURRED). In those cases we should just increase the retry counter again, retrying one more to clear up this Unit Attention state. [jejb: update to work with RC16 devices and not to loop endlessly] Signed-off-by:
Hannes Reinecke <hare@suse.de> Signed-off-by:
-
Kashyap, Desai authored
After Host Reset firmware will have new list of device handles for the target. Device handle refresh in driver is part of Rescan topology logic. (See functions like *_search_responding_*). This needs to be done from Host Reset context before making shost_recovery to 0. Currently it is done in Firwmare event context, which may leads IO to a wrong device. Now handler refresh is moved to HBA reset context. Apart from this, Now driver will stop IOs for all device setting deleted flag to 1 at the time of HBA Reset through _scsih_prep_device_scan. It will only unblock devices, if devices has been found as part of RESCAN. This way it will make more safe IO blocking at the time of HBA reset at mpt2sas driver layer. Signed-off-by:
-
Hannes Reinecke authored
sd always tries to submit a READ_CAPACITY(16) CDB, regardless whether the host actually supports it. queuecommand() will then return DID_ABORT, which is not qualified enough to detect the true cause here. So better check in sd_try_rc16 first if the cdblen is supported. Signed-off-by:
-
Anil Veerabhadrappa authored
missing 10G drivers added to description Signed-off-by:
Anil Veerabhadrappa <anilgv@broadcom.com> Reviewed-by:
-
Anil Veerabhadrappa authored
When bnx2/cnic/bnx2i drivers are loaded in certain order, bnx2i will will not initialize the device correctly because 'hba->cnic' will be NULL when bnx2i_start() is called from register_device() context. Under this condition 'ifdown' and 'ifup' of associated network interface is required to bring iscsi adapter state to ready state so that it will accept iscsi connection setup within the chip Initializing 'hba->cnic' before calling register_device() will fix this issue Signed-off-by:
Michael Chan <mchan@broadcom.com> Signed-off-by:
-
Anil Veerabhadrappa authored
instruct firmware to use driver/iscsid provided expected statsn field while constructing login pdu header. Initialize 'flags' to instruct chip to use driver/iscsid provided ExpStatSN value while constructing iSCSI login PDU header Signed-off-by:
Eddie Wai <waie@broadcom.com> Signed-off-by:
-
Christof Schmitt authored
The return code FAST_IO_FAIL from fc_block_scsi_eh indicates that the pending I/O requests have been terminated as a result of the fast_io_fail_tmo. Pass this return code back to the scsi eh to stop the scsi eh in this case. Signed-off-by:
Christof Schmitt <christof.schmitt@de.ibm.com> Signed-off-by:
-
Christof Schmitt authored
If the scsi eh is running and then a FC LLD calls fc_remote_port_delete, the SCSI commands sent from the eh will fail. To prevent this, a FC LLD can call fc_block_scsi_eh from the eh callback, blocking the eh thread until the dev_loss_tmo fires or the remote port is available again. If (e.g. for a multipathing setup) the dev_loss_tmo is set to a very large value, thus preventing the scsi device removal , the scsi eh can block for a long time. For multipathing, the fast_io_fail_tmo is then set to a low value to detect path problems sooner. This patch introduces a new return code FAST_IO_FAIL. The function fc_block_scsi_eh now returns FAST_IO_FAIL when the fast_io_fail_tmo fires. This indicates that the LLD terminated all pending I/O requests and there are no more pending SCSI commands for the scsi eh to wait for. This return code can be passed back to the scsi eh to stop the escalation and finish the recovery process for this device. Signed-off-by:
-
Ferenc Wagner authored
The default is enabled since 7e47e5ca. Signed-off-by:
Ferenc Wagner <wferi@niif.hu> Acked-by:
Giridhar Malavali <giridhar.malavali@qlogic.com> Signed-off-by:
-
