- 19 Jul, 2022 6 commits
-
-
Quinn Tran authored
vref_count took an extra decrement in the task management path. Add an extra ref count to compensate the imbalance. Link: https://lore.kernel.org/r/20220713052045.10683-7-njavali@marvell.com Cc: stable@vger.kernel.org Reviewed-by:
Himanshu Madhani <himanshu.madhani@oracle.com> Signed-off-by:
Quinn Tran <qutran@marvell.com> Signed-off-by:
Nilesh Javali <njavali@marvell.com> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com>
-
Quinn Tran authored
This patch fixes IKE message being dropped due to error in processing Purex IOCB and Continuation IOCBs. Link: https://lore.kernel.org/r/20220713052045.10683-6-njavali@marvell.com Fixes: fac28079 ("scsi: qla2xxx: edif: Add extraction of auth_els from the wire") Cc: stable@vger.kernel.org Reviewed-by:
-
Arun Easi authored
On some platforms, the current logic of relying on finding new packet solely based on signature pattern can lead to driver reading stale packets. Though this is a bug in those platforms, reduce such exposures by limiting reading packets until the IN pointer. Two module parameters are introduced: ql2xrspq_follow_inptr: When set, on newer adapters that has queue pointer shadowing, look for response packets only until response queue in pointer. When reset, response packets are read based on a signature pattern logic (old way). ql2xrspq_follow_inptr_legacy: Like ql2xrspq_follow_inptr, but for those adapters where there is no queue pointer shadowing. Link: https://lore.kernel.org/r/20220713052045.10683-5-njavali@marvell.com Cc: stable@vger.kernel.org Reviewed-by:Arun Easi <aeasi@marvell.com> Signed-off-by:
-
Bikash Hazarika authored
While requesting a new mailbox command, driver does not write any data to unused registers. Initialize the unused register value to zero while requesting a new mailbox command to prevent stale entry access by firmware. Link: https://lore.kernel.org/r/20220713052045.10683-4-njavali@marvell.com Cc: stable@vger.kernel.org Reviewed-by:
Bikash Hazarika <bhazarika@marvell.com> Signed-off-by:
-
Bikash Hazarika authored
Replace display field with the correct field. Link: https://lore.kernel.org/r/20220713052045.10683-3-njavali@marvell.com Fixes: 8777e431 ("scsi: qla2xxx: Migrate NVME N2N handling into state machine") Cc: stable@vger.kernel.org Reviewed-by:
-
Nilesh Javali authored
This fixes the regression of NVMe discovery failure during driver load time. This reverts commit 6a45c8e1. Link: https://lore.kernel.org/r/20220713052045.10683-2-njavali@marvell.com Cc: stable@vger.kernel.org Reviewed-by:
-
- 14 Jul, 2022 20 commits
-
-
Don Brace authored
Link: https://lore.kernel.org/r/165730608687.177165.11815510982277242966.stgit@brunhildaReviewed-by:
Gerry Morong <gerry.morong@microchip.com> Reviewed-by:
Scott Benesh <scott.benesh@microchip.com> Reviewed-by:
Scott Teel <scott.teel@microchip.com> Reviewed-by:
Kevin Barnett <kevin.barnett@microchip.com> Signed-off-by:
Don Brace <don.brace@microchip.com> Signed-off-by:
-
Don Brace authored
Update copyright to current year. Link: https://lore.kernel.org/r/165730608177.177165.13184715486635363193.stgit@brunhildaReviewed-by:
-
Kevin Barnett authored
Allow user to override the default driver timeout for controller ready. There are some rare configurations which require the driver to wait longer than the normal 3 minutes for the controller to complete its bootup sequence and be ready to accept commands from the driver. The module parameter is: ctrl_ready_timeout= { 0 | 30-1800 } and specifies the timeout in seconds for the driver to wait for controller ready. The valid range is 0 or 30-1800. The default value is 0, which causes the driver to use a timeout of 180 seconds (3 minutes). Link: https://lore.kernel.org/r/165730607666.177165.9221211345284471213.stgit@brunhildaReviewed-by:Mike McGowen <mike.mcgowen@microchip.com> Signed-off-by:
-
Kevin Barnett authored
Change removing a LUN using sysfs from an internal driver function pqi_remove_all_scsi_devices() to using the .slave_destroy entry in the scsi_host_template. A LUN can be deleted via sysfs using this syntax: echo 1 > /sys/block/sdX/device/delete Link: https://lore.kernel.org/r/165730607154.177165.9723066932202995774.stgit@brunhildaReviewed-by:
-
Mike McGowen authored
Allow SMP affinity to be changeable by disabling managed interrupts. On distributions where the driver is enabled for multi-queue support the driver utilizes kernel managed interrupts, which automatically distributes interrupts to all available CPUs and assigns SMP affinity. On most distributions, the affinity can not be changed by the user. This change will allow managed interrupts to be disabled by the user via a module parameter while still allowing multi-queue support to function properly. Use the module parameter disable_managed_interrupts=1 Link: https://lore.kernel.org/r/165730606638.177165.12846020942931640329.stgit@brunhildaReviewed-by:
-
Kevin Barnett authored
Correct a rare stale RAID map access when performing AIO during a RAID configuration change. A race condition in the driver could cause it to access a stale RAID map when a logical volume is reconfigured. Modify the driver logic to invalidate a RAID map very early when a RAID configuration change is detected and only switch to a new RAID map after the driver detects that the RAID map has changed. Link: https://lore.kernel.org/r/165730606128.177165.7671413443814750829.stgit@brunhildaReviewed-by:
-
Mahesh Rajashekhara authored
Correct a SOP READ and WRITE DMA flags for some requests. This update corrects DMA direction issues with SCSI commands removed from the controller's internal lookup table. Currently, SCSI READ BLOCK LIMITS (0x5) was removed from the controller lookup table and exposed a DMA direction flag issue. SCSI READ BLOCK LIMITS was recently removed from our controller lookup table so the controller uses the respective IU flag field to set the DMA data direction. Since the DMA direction is incorrect the FW never completes the request causing a hang. Some SCSI commands which use SCSI READ BLOCK LIMITS * sg_map * mt -f /dev/stX status After updating controller firmware, users may notice their tape units failing. This patch resolves the issue. Also, the AIO path DMA direction is correct. The DMA direction flag is a day-one bug with no reported BZ. Fixes: 6c223761 ("smartpqi: initial commit of Microsemi smartpqi driver") Link: https://lore.kernel.org/r/165730605618.177165.9054223644512926624.stgit@brunhildaReviewed-by:Mahesh Rajashekhara <Mahesh.Rajashekhara@microchip.com> Signed-off-by:
-
Kevin Barnett authored
Change method used to detect controller firmware crash during PQI reset. PQI reset can fail with error -6 if firmware takes > 100ms to complete reset. Method used by driver to detect controller firmware crash during PQI was incorrect in some cases. Link: https://lore.kernel.org/r/165730605108.177165.1132931838384767071.stgit@brunhildaReviewed-by:
-
Mike McGowen authored
Add PCI IDs for Lenovo controllers (values in hex): VID / DID / SVID / SDID ---- ---- ---- ---- Lenovo 4350-8i HBA 9005 / 028f / 1d49 / 0220 Lenovo 4350-16i HBA 9005 / 028f / 1d49 / 0221 Lenovo 5350-8i RAID 9005 / 028f / 1d49 / 0520 Lenovo 5350-8i Internal RAID 9005 / 028f / 1d49 / 0522 Lenovo 9350-8i RAID 9005 / 028f / 1d49 / 0620 Lenovo 9350-8i Internal RAID 9005 / 028f / 1d49 / 0621 Lenovo 9350-16i RAID 9005 / 028f / 1d49 / 0622 Lenovo 9350-16i Internal RAID 9005 / 028f / 1d49 / 0623 Link: https://lore.kernel.org/r/165730604598.177165.9910276232981721083.stgit@brunhildaReviewed-by: -
Mike McGowen authored
Add the PCI ID for (values in hex): VID / DID / SVID / SDID ---- ---- ---- ---- Adaptec SmartHBA 2100-8i-o 9005 / 0285 / 9005 / 0659 Link: https://lore.kernel.org/r/165730604089.177165.17257514581321583667.stgit@brunhildaReviewed-by: -
Sagar Biradar authored
Fail all outstanding requests after a PCI linkdown. Block access to device SCSI attributes during the following conditions: "Cable pull" is called PQI_CTRL_SURPRISE_REMOVAL. "PCIe Link Down" is called PQI_CTRL_GRACEFUL_REMOVAL. Block access to device SCSI attributes during and in rare instances when the controller goes offline. Either outstanding requests or the access of SCSI attributes post linkdown can lead to a hang. Post linkdown, driver does not fail the outstanding requests leading to long wait time before all the IOs eventually fail. Also access of the SCSI attributes by host applications can lead to a system hang. Link: https://lore.kernel.org/r/165730603578.177165.4699352086827187263.stgit@brunhildaReviewed-by:
Sagar Biradar <sagar.biradar@microchip.com> Signed-off-by:
-
Kumar Meiyappan authored
Add driver support for up to 256 LUNs per device. Update AIO path to pass the appropriate LUN number for base-code to target the correct LUN. Update RAID IO path to pass the appropriate LUN number for FW to target the correct LUN. Pass the correct LUN number while doing a LUN reset. Count the outstanding commands based on LUN number. While removing a Multi-LUN device, wait for all outstanding commands to complete for all LUNs. Add Feature bit support. Link: https://lore.kernel.org/r/165730603067.177165.14016422176841798336.stgit@brunhildaReviewed-by:
Kumar Meiyappan <Kumar.Meiyappan@microchip.com> Signed-off-by:
-
Mike McGowen authored
Insert a minimum 1 millisecond delay after writing to a register before reading from it. SIS and PQI registers that can be both written to and read from can return stale data if read from too soon after having been written to. There is no read/write ordering or hazard detection on the inbound path to the MSGU from the PCIe bus, therefore reads could pass writes. Link: https://lore.kernel.org/r/165730602555.177165.11181012469428348394.stgit@brunhildaReviewed-by:
-
Murthy Bhat authored
Add the following controllers (values in hex): VID / DID / SVID / SDID ---- / ---- / ---- / ---- Ramaxel FBGF-RAD PM8204 9005 / 028F / 1CC4 / 0101 Ramaxel FBGF-RAD PM8222 9005 / 028F / 1CC4 / 0201 Link: https://lore.kernel.org/r/165730602045.177165.3720208650043407285.stgit@brunhildaReviewed-by:Murthy Bhat <Murthy.Bhat@microchip.com> Signed-off-by:
-
Gilbert Wu authored
Print controller firmware version to OS message log during driver initialization or after OFA. Link: https://lore.kernel.org/r/165730601536.177165.17698744242908911822.stgit@brunhildaReviewed-by:
Gilbert Wu <Gilbert.Wu@microchip.com> Signed-off-by:
-
Mike McGowen authored
Check the response code returned from the LUN reset task management function and if it indicates the LUN is not valid, do not retry. Reduce rescan worker delay to 5 seconds for the event handler only. The removal of a drive from the OS could have been delayed up to 30 seconds after being physically pulled. The driver was retrying a LUN reset 3 times even though the return code indiciated the LUN was no longer valid. There was a 10 second delay between each retry. Additionally, the rescan worker was scheduled to run 10 seconds after the driver received the event. Link: https://lore.kernel.org/r/165730601025.177165.9416869335174437006.stgit@brunhildaReviewed-by:
-
Karan Tilak Kumar authored
Refactor code from fnic probe into a different function so that scsi layer initialization code is grouped together. Also, add log messages for better debugging. Link: https://lore.kernel.org/r/20220707205155.692688-1-kartilak@cisco.comCo-developed-by:
Gian Carlo Boffa <gcboffa@cisco.com> Signed-off-by:
Arulprabhu Ponnusamy <arulponn@cisco.com> Signed-off-by:
Karan Tilak Kumar <kartilak@cisco.com> Signed-off-by:
-
Michael Schmitz authored
DMA masks given in the Zorro ID table don't contain the 2 byte alignment quirk seen in the GVP11_XFER_MASK macro from gvp11.h so no need to account for that. DMA masks passed to dma_set_mask_and_coherent() must be 64 bit, add the missing cast in the TO_DMA_MASK macro used to convert driver DMA masks to DMA API masks. Link: https://lore.kernel.org/r/6d1d88ee-1cf6-c735-1e6d-bafd2096e322@gmail.com Link: https://lore.kernel.org/r/20220713074913.7873-1-schmitzmic@gmail.com Fixes: 158da6bc ("scsi: gvp11: Convert m68k WD33C93 drivers to DMA API") CC: linux-scsi@vger.kernel.org Signed-off-by:
Michael Schmitz <schmitzmic@gmail.com> Signed-off-by:
-
Keoseong Park authored
Commit 4db7a236 ("scsi: ufs: Fix concurrency of error handler and other error recovery paths") removed all callers of UIC_HIBERN8_ENTER_RETRIES. Hence also remove the macro itself. Link: https://lore.kernel.org/r/20220708052006epcms2p2d1129dbf95fd77f46906200ccb0a9ccd@epcms2p2Reviewed-by:
Avri Altman <avri.altman@wdc.com> Signed-off-by:
Keoseong Park <keosung.park@samsung.com> Signed-off-by:
-
Chanho Park authored
Since commit 1599069a ("phy: core: Warn when phy_power_on is called before phy_init"), the following warning has been reported: phy_power_on was called before phy_init To address this, we need to remove phy_power_on from exynos_ufs_phy_init() and move it after phy_init. phy_power_off and phy_exit are also necessary in exynos_ufs_remove(). Link: https://lore.kernel.org/r/20220706020255.151177-4-chanho61.park@samsung.comReviewed-by:
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by:
Chanho Park <chanho61.park@samsung.com> Signed-off-by:
-
- 07 Jul, 2022 14 commits
-
-
Jiang Jian authored
There is an unexpected word "the" in the comments that needs to be dropped. file: ./drivers/s390/scsi/zfcp_diag.h line: 5 * Definitions for handling diagnostics in the the zfcp device driver. changed to * Definitions for handling diagnostics in the zfcp device driver. Link: https://lore.kernel.org/r/20220621114207.106405-1-jiangjian@cdjrlc.com Link: https://lore.kernel.org/r/4c02fefa19ab46f0f163990bde3ce10bd9c7caf1.1657122360.git.bblock@linux.ibm.comReviewed-by:
Benjamin Block <bblock@linux.ibm.com> Signed-off-by:
Jiang Jian <jiangjian@cdjrlc.com> Signed-off-by:
-
Julian Wiedmann authored
These attributes are now only accessed through the zfcp_sysfs_sdev_attr_group. Link: https://lore.kernel.org/r/0791b9149ebfa39e6b8eab093113cd2527dbf3d3.1657122360.git.bblock@linux.ibm.com Fixes: d8d7cf3f ("scsi: zfcp: Switch to attribute groups") Reviewed-by:
Julian Wiedmann <jwi@linux.ibm.com> Signed-off-by:
-
Junwoo Lee authored
Once the host link startup fails 3 times, all host registers are reset to default values except in ufshcd_hba_enable(). The ufs host controller is disabled and enabled in ufshcd_hba_enable(). Consequently we need to skip last hci reset to get valid host register values. e.g. [ 1.898026] [2: kworker/u16:2: 211] ufs: link startup failed 1 [ 1.898133] [2: kworker/u16:2: 211] host_regs: 00000000: 1383ff1f 00000000 00000300 00000000 [ 1.898141] [2: kworker/u16:2: 211] host_regs: 00000010: 00000106 000001ce 00000000 00000000 [ 1.898148] [2: kworker/u16:2: 211] host_regs: 00000020: 00000000 00000470 00000000 00000000 [ 1.898155] [2: kworker/u16:2: 211] host_regs: 00000030: 00000008 00000003 00000000 00000000 [ 1.898163] [2: kworker/u16:2: 211] host_regs: 00000040: 00000000 00000000 00000000 00000000 [ 1.898171] [2: kworker/u16:2: 211] host_regs: 00000050: 00000000 00000000 00000000 00000000 [ 1.898177] [2: kworker/u16:2: 211] host_regs: 00000060: 00000000 00000000 00000000 00000000 [ 1.898186] [2: kworker/u16:2: 211] host_regs: 00000070: 00000000 00000000 00000000 00000000 [ 1.898194] [2: kworker/u16:2: 211] host_regs: 00000080: 00000000 00000000 00000000 00000000 [ 1.898201] [2: kworker/u16:2: 211] host_regs: 00000090: 00000000 00000000 00000000 00000000 Link: https://lore.kernel.org/r/20220705083538.15143-1-sh043.lee@samsung.comReviewed-by:
Junwoo Lee <junwoo80.lee@samsung.com> Signed-off-by:
Seunghui Lee <sh043.lee@samsung.com> Signed-off-by:
-
Daniil Lunev authored
Sets the WriteBooster capability flag when ADL's UFS controller is used. Link: https://lore.kernel.org/r/20220705165316.v2.1.Ib5ebec952d9a59f5c69c89b694777f517d22466d@changeidAcked-by:
Adrian Hunter <adrian.hunter@intel.com> Signed-off-by:
Daniil Lunev <dlunev@chromium.org> Signed-off-by:
-
James Smart authored
Update copyrights to 2022 for files modified in the 14.2.0.5 patch set. Link: https://lore.kernel.org/r/20220701211425.2708-13-jsmart2021@gmail.comCo-developed-by:
Justin Tee <justin.tee@broadcom.com> Signed-off-by:
James Smart <jsmart2021@gmail.com> Signed-off-by:
-
James Smart authored
Update lpfc version to 14.2.0.5 Link: https://lore.kernel.org/r/20220701211425.2708-12-jsmart2021@gmail.comCo-developed-by:
-
James Smart authored
The Menlo/Hornet adapter was never released to the field. As such, driver code specific to the adapter is unnecessary and should be removed. Link: https://lore.kernel.org/r/20220701211425.2708-11-jsmart2021@gmail.comCo-developed-by:
-
James Smart authored
lpfc_nvmet_prep_abort_wqe() has a lot of common code with lpfc_sli_prep_abort_xri(). Delete lpfc_nvmet_prep_abort_wqe() as the wqe can be filled out using the generic lpfc_sli_prep_abort_xri routine(). Add the wqec option to lpfc_sli_prep_abort_xri() for lpfc_nvmet_prep_abort_wqe(). Link: https://lore.kernel.org/r/20220701211425.2708-10-jsmart2021@gmail.comCo-developed-by:
-
James Smart authored
The RSCN_MEMENTO logic was to workaround a target that does not register both FCP and NVMe FC4 types at the same time. This caused the configuration to not produce a second RSCN for the NVMe FC4 type registration in a timely manner. The intention of the RSCN_MEMENTO flag was to always signal to try NVMe PRLI. However, there are other FCP-only target arrays in correctly behaved configurations that reject the NVMe PRLI followed by a LOGO leading to never rediscovering the target after an issue_lip (as LOGO causes a repeat of PLOGI/PRLIs). Revert the RSCN_MEMENTO patch as it is causing correctly behaved configs to fail while it exists only to succeed on a misbehaved config. Link: https://lore.kernel.org/r/20220701211425.2708-9-jsmart2021@gmail.com Fixes: 1045592f ("scsi: lpfc: Introduce FC_RSCN_MEMENTO flag for tracking post RSCN completion") Co-developed-by:
-
James Smart authored
During a target link bounce test, the driver sees a mismatch between the NPortId and the WWPN on the node structures (ndlps) involved. When this occurs, the driver "swaps" the ndlp and new_ndlp node parameters to restore WWPN/DID uniqueness in the fc_nodes list per vport. However, the driver neglected to swap the nlp_fc4_type in the ndlp passed to lpfc_plogi_confirm_nport causing a failure to recover the NVMe PLOGI/PRLI and ultimately the NVMe paths. Correct confirm_nport to preserve the fc4 types from the new-ndlp when the data is moved over ot the ndlp structure. Link: https://lore.kernel.org/r/20220701211425.2708-8-jsmart2021@gmail.comCo-developed-by:
-
James Smart authored
Disabling FA-PWWN should be effective after port reset, but in some cases it was found to be impossible to clear FA-PWWN usage without a driver reload. Clean up FA-PWWN flag management to make enable and disable of the feature more robust. Link: https://lore.kernel.org/r/20220701211425.2708-7-jsmart2021@gmail.comCo-developed-by:
-
James Smart authored
There is no corresponding free routine if lpfc_sli4_issue_wqe fails to issue the CMF WQE in lpfc_issue_cmf_sync_wqe. If ret_val is non-zero, then free the iocbq request structure. Link: https://lore.kernel.org/r/20220701211425.2708-6-jsmart2021@gmail.comCo-developed-by:
-
James Smart authored
VMID introduced an extra increment of cmd_pending, causing double-counting of the I/O. The normal increment ios performed in lpfc_get_scsi_buf. Link: https://lore.kernel.org/r/20220701211425.2708-5-jsmart2021@gmail.com Fixes: 33c79741 ("scsi: lpfc: vmid: Introduce VMID in I/O path") Cc: <stable@vger.kernel.org> # v5.14+ Co-developed-by:
-
James Smart authored
When providing a D_ID in XMIT_ELS_RSP64_CX iocb the PU field should be set to 3 to describe the parameter being passed to firmware. Link: https://lore.kernel.org/r/20220701211425.2708-4-jsmart2021@gmail.comCo-developed-by:
-
