- 26 Nov, 2020 9 commits
-
-
Vaibhav Gupta authored
Drivers should do only device-specific jobs. But in general, drivers using legacy PCI PM framework for .suspend()/.resume() have to manage many PCI PM-related tasks themselves which can be done by PCI Core itself. This brings extra load on the driver and it directly calls PCI helper functions to handle them. Switch to the new generic framework by updating function signatures and define a "struct dev_pm_ops" variable to bind PM callbacks. Also, remove unnecessary calls to the PCI Helper functions along with the legacy .suspend & .resume bindings. Link: https://lore.kernel.org/r/20201102164730.324035-10-vaibhavgupta40@gmail.comSigned-off-by:
Vaibhav Gupta <vaibhavgupta40@gmail.com> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com>
-
Vaibhav Gupta authored
The driver calls pci_enable_wake(...., false) in arcmsr_resume(), and there is no corresponding pci_enable_wake(...., true) in arcmsr_suspend(). Either it should do enable-wake the device in .suspend() or should not invoke pci_enable_wake() at all. Concluding that this driver doesn't support enable-wake and PCI core calls pci_enable_wake(pci_dev, PCI_D0, false) during resume, drop it from arcmsr_resume(). Link: https://lore.kernel.org/r/20201102164730.324035-9-vaibhavgupta40@gmail.comSigned-off-by:
-
Vaibhav Gupta authored
Drivers should do only device-specific jobs. But in general, drivers using legacy PCI PM framework for .suspend()/.resume() have to manage many PCI PM-related tasks themselves which can be done by PCI Core itself. This brings extra load on the driver and it directly calls PCI helper functions to handle them. Switch to the new generic framework by updating function signatures and define a "struct dev_pm_ops" variable to bind PM callbacks. Also, remove unnecessary calls to the PCI Helper functions along with the legacy .suspend & .resume bindings. Link: https://lore.kernel.org/r/20201102164730.324035-8-vaibhavgupta40@gmail.comSigned-off-by:
-
Vaibhav Gupta authored
Drivers should do only device-specific jobs. But in general, drivers using legacy PCI PM framework for .suspend()/.resume() have to manage many PCI PM-related tasks themselves which can be done by PCI Core itself. This brings extra load on the driver and it directly calls PCI helper functions to handle them. Switch to the new generic framework by updating function signatures and define a "struct dev_pm_ops" variable to bind PM callbacks. Also, remove unnecessary calls to the PCI Helper functions along with the legacy .suspend & .resume bindings. Link: https://lore.kernel.org/r/20201102164730.324035-7-vaibhavgupta40@gmail.comSigned-off-by:
-
Vaibhav Gupta authored
Drivers should do only device-specific jobs. But in general, drivers using legacy PCI PM framework for .suspend()/.resume() have to manage many PCI PM-related tasks themselves which can be done by PCI Core itself. This brings extra load on the driver and it directly calls PCI helper functions to handle them. Switch to the new generic framework by updating function signatures and define a "struct dev_pm_ops" variable to bind PM callbacks. Also, remove unnecessary calls to the PCI Helper functions along with the legacy .suspend & .resume bindings. Link: https://lore.kernel.org/r/20201102164730.324035-6-vaibhavgupta40@gmail.comAcked-by:
Balsundar P <balsundar.p@microchip.com> Signed-off-by:
-
Vaibhav Gupta authored
The driver calls pci_enable_wake(...., false) in aac_resume(), and there is no corresponding pci_enable_wake(...., true) in aac_suspend(). Either it should do enable-wake the device in .suspend() or should not invoke pci_enable_wake() at all. Concluding that this is a bug and PCI core calls pci_enable_wake(pci_dev, PCI_D0, false) during resume, drop it from aac_resume(). Link: https://lore.kernel.org/r/20201102164730.324035-5-vaibhavgupta40@gmail.comSigned-off-by:
-
Vaibhav Gupta authored
Function parameter 'pdev 'is described as Generic Device Structure. It is a PCI device structure. Link: https://lore.kernel.org/r/20201102164730.324035-4-vaibhavgupta40@gmail.comSigned-off-by:
-
Vaibhav Gupta authored
Drivers should do only device-specific jobs. But in general, drivers using legacy PCI PM framework for .suspend()/.resume() have to manage many PCI PM-related tasks themselves which can be done by PCI Core itself. This brings extra load on the driver and it directly calls PCI helper functions to handle them. Switch to the new generic framework by updating function signatures and define a "struct dev_pm_ops" variable to bind PM callbacks. Also, remove unnecessary calls to the PCI Helper functions along with the legacy .suspend & .resume bindings. Link: https://lore.kernel.org/r/20201102164730.324035-3-vaibhavgupta40@gmail.comSigned-off-by:
-
Vaibhav Gupta authored
The driver calls pci_enable_wake(...., false) in megasas_resume(), and there is no corresponding pci_enable_wake(...., true) in megasas_suspend(). Either it should do enable-wake the device in .suspend() or should not invoke pci_enable_wake() at all. Concluding that this driver doesn't support enable-wake and PCI core calls pci_enable_wake(pci_dev, PCI_D0, false) during resume, drop it from megasas_resume(). Link: https://lore.kernel.org/r/20201102164730.324035-2-vaibhavgupta40@gmail.comSigned-off-by:
-
- 24 Nov, 2020 11 commits
-
-
Karan Tilak Kumar authored
We need to check for a valid io_req before we check other data. Also, remove redundant checks. Link: https://lore.kernel.org/r/20201121023337.19295-1-kartilak@cisco.comReviewed-by:
Arulprabhu Ponnusamy <arulponn@cisco.com> Co-developed-by:
Satish Kharat <satishkh@cisco.com> Signed-off-by:
Karan Tilak Kumar <kartilak@cisco.com> Signed-off-by:
-
Karan Tilak Kumar authored
Set scsi_set_resid() only if FCPIO_ICMND_CMPL_RESID_UNDER is set. Link: https://lore.kernel.org/r/20201121015134.18872-1-kartilak@cisco.comReviewed-by:
-
Karan Tilak Kumar authored
Replace shost_printk() with FNIC_MAIN_DBG() so that these log messages are controlled by fnic_log_level flag in fnic_handle_link. Link: https://lore.kernel.org/r/20201121013739.18701-1-kartilak@cisco.comReviewed-by:
-
Karan Tilak Kumar authored
Avoid looping in fnic_scsi_abort_io() before sending fw reset when fnic is in TRANS ETH state and when we have not received any link events. Link: https://lore.kernel.org/r/20201121012145.18522-1-kartilak@cisco.comReviewed-by:
-
Karan Tilak Kumar authored
Replacing shost_printk with FNIC_FCS_DBG() so that these log messages are controlled by fnic_log_level flag in fnic_fip_handler_timer. Bumping up version number from 47 to 49 to maintain same level as internal version. Link: https://lore.kernel.org/r/20201120220712.16708-1-kartilak@cisco.comReviewed-by:
-
Joe Perches authored
Make the pm8001_printk() macro take an explicit HBA instead of assuming the existence of an unspecified pm8001_ha argument. Miscellanea: - Add pm8001_ha to the few uses of pm8001_printk() - Add HBA to the pm8001_dbg macro call to pm8001_printk() Link: https://lore.kernel.org/r/0e17a4c845f15e18f98b346ffb9b039584d21cdd.1605914030.git.joe@perches.comAcked-by:
Jack Wang <jinpu.wang@cloud.ionos.com> Signed-off-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
Joe Perches authored
Every PM8001_<FOO>_DBG macro uses an internal call to pm8001_printk. Convert all uses of: PM8001_<FOO>_DBG(hba, pm8001_printk(fmt, ...)) to pm8001_dbg(hba, <FOO>, fmt, ...) so the visual complexity of each macro is reduced. The repetitive macro definitions are converted to a single pm8001_dbg and the level is concatenated using PM8001_##level##_LOGGING for the specific level test. Done with coccinelle, checkpatch and a little typing of the new macro definition. Miscellanea: - Coalesce formats - Realign arguments - Add missing terminating newlines to formats - Remove trailing spaces from formats - Change defective loop with printk(KERN_INFO... to emit a 16 byte hex block to %p16h Link: https://lore.kernel.org/r/49f36a93af7752b613d03c89a87078243567fd9a.1605914030.git.joe@perches.comReported-by:
kernel test robot <lkp@intel.com> Acked-by:
-
Xu Wang authored
Stop typecasting the value returned by kcalloc(). Link: https://lore.kernel.org/r/20201120083648.9319-1-vulab@iscas.ac.cnAcked-by:
Xu Wang <vulab@iscas.ac.cn> Signed-off-by:
-
Finn Thain authored
Refactor to avoid needless calls to NCR5380_maybe_release_dma_irq(). This makes the machine code smaller and the source more readable. Link: https://lore.kernel.org/r/c1317ae8fdcb498460de5d7ea0bd62a42f5eeca8.1605847196.git.fthain@telegraphics.com.auTested-by:
Michael Schmitz <schmitzmic@gmail.com> Reviewed-by:
Finn Thain <fthain@telegraphics.com.au> Signed-off-by:
-
Finn Thain authored
It is possible that bus_reset_cleanup() or .eh_abort_handler could be invoked during NCR5380_queuecommand(). If that takes place before the new command is enqueued and after the ST-DMA "lock" has been acquired, the ST-DMA "lock" will be released again. This will result in a lost DMA interrupt and a command timeout. Fix this by excluding EH and interrupt handlers while the new command is enqueued. Link: https://lore.kernel.org/r/af25163257796b50bb99d4ede4025cea55787b8f.1605847196.git.fthain@telegraphics.com.auTested-by:
-
Bjorn Andersson authored
The introduction of ufshcd_dme_configure_adapt() refactored out duplication from the Mediatek and Qualcomm drivers. Both these implementations had the logic of: gear_tx == UFS_HS_G4 => PA_INITIAL_ADAPT gear_tx != UFS_HS_G4 => PA_NO_ADAPT but now both implementations pass PA_INITIAL_ADAPT as "adapt_val" and if gear_tx is not UFS_HS_G4 that is replaced with PA_INITIAL_ADAPT. In other words, it's PA_INITIAL_ADAPT in both above cases. The result is that e.g. Qualcomm SM8150 has no longer functional UFS, so adjust the logic to match the previous implementation. Link: https://lore.kernel.org/r/20201121044810.507288-1-bjorn.andersson@linaro.org Fixes: fc85a74e ("scsi: ufs: Refactor ADAPT configuration function") Reviewed-by:Can Guo <cang@codeaurora.org> Reviewed-by:
Stanley Chu <stanley.chu@mediatek.com> Signed-off-by:
Bjorn Andersson <bjorn.andersson@linaro.org> Signed-off-by:
-
- 20 Nov, 2020 20 commits
-
-
James Smart authored
Remove vport variable that is assigned but not used in lpfc_sli4_abts_err_handler(). Link: https://lore.kernel.org/r/20201119203407.121913-1-james.smart@broadcom.com Fixes: e7dab164 ("scsi: lpfc: Fix scheduling call while in softirq context in lpfc_unreg_rpi") Reported-by:
James Smart <james.smart@broadcom.com> Signed-off-by:
-
James Smart authored
lpfc_nvmet_prep_abort_wqe() needs to be declared static. Link: https://lore.kernel.org/r/20201119203316.121725-1-james.smart@broadcom.com Fixes: db7531d2 ("scsi: lpfc: Convert abort handling to SLI-3 and SLI-4 handlers") Reported-by:
-
James Smart authored
Remove set but not used variable shost in lpfc_dev_loss_tmo_handler(). Link: https://lore.kernel.org/r/20201119203353.121866-1-james.smart@broadcom.com Fixes: 52edb2ca ("scsi: lpfc: Remove ndlp when a PLOGI/ADISC/PRLI/REG_RPI ultimately fails") Reported-by:
-
James Smart authored
Remove local variables that are set but not used. Link: https://lore.kernel.org/r/20201119203340.121819-1-james.smart@broadcom.com Fixes: c6adba15 ("scsi: lpfc: Rework remote port lock handling") Reported-by:
-
James Smart authored
Function needs to be declared as static. Link: https://lore.kernel.org/r/20201119203328.121772-1-james.smart@broadcom.com Fixes: 8aaa7bcf ("scsi: lpfc: Add FDMI Vendor MIB support") Reported-by:
-
Colin Ian King authored
Currently there is an error return path that neglects to free the allocation for lcb_context. Fix this by adding a new error free exit path that kfree's lcb_context before returning. Use this new kfree exit path in another exit error path that also kfree's the same object, allowing a line of code to be removed. Link: https://lore.kernel.org/r/20201118141314.462471-1-colin.king@canonical.com Fixes: 4430f7fd ("scsi: lpfc: Rework locations of ndlp reference taking") Reviewed-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Colin Ian King authored
Currently there is a null check on the pointer ndlp that exits via error path issue_ct_rsp_exit followed by another null check on the same pointer that is almost identical to the previous null check stanza and yet can never can be reached because the previous check exited via issue_ct_rsp_exit. This is deadcode and can be removed. Link: https://lore.kernel.org/r/20201118133744.461385-1-colin.king@canonical.comReviewed-by:
-
Colin Ian King authored
There is a null check on pointer lpfc_cmd after the pointer has been dereferenced when pointers rdata and ndlp are initialized at the start of the function. Fix this by only assigning rdata and ndlp after the pointer lpfc_cmd has been null checked. Link: https://lore.kernel.org/r/20201118131345.460631-1-colin.king@canonical.com Fixes: 96e209be ("scsi: lpfc: Convert SCSI I/O completions to SLI-3 and SLI-4 handlers") Reviewed-by:
-
Tyrel Datwyler authored
The previous patch added support for the targetWWPN field in version 2 MADs and vfcFrame structures. Set the IBMVFC_CAN_SEND_VF_WWPN bit in our capabailites flag during NPIV Login to inform the VIOS that this client supports the feature. Link: https://lore.kernel.org/r/20201118011104.296999-7-tyreld@linux.ibm.comAcked-by:
Brian King <brking@linux.vnet.ibm.com> Signed-off-by:
Tyrel Datwyler <tyreld@linux.ibm.com> Signed-off-by:
-
Tyrel Datwyler authored
Several version 2 MADs and the version 2 vfcFrame structures introduced a new targetWWPN field for better identification of a target over the scsi_id. Set this field and MAD versioning fields when the VIOS advertises the IBMVFC_HANDLE_VF_WWPN capability. Link: https://lore.kernel.org/r/20201118011104.296999-6-tyreld@linux.ibm.comAcked-by:
-
Tyrel Datwyler authored
The FC iu and response payloads are located at different offsets depending on the ibmvfc_cmd version. This is a result of the version 2 vfcFrame definition adding an extra 64bytes of reserved space to the structure prior to the payloads. Add helper routines to determine the current vfcFrame version and return a pointer to the proper iu or response structure within that ibmvfc_cmd. Link: https://lore.kernel.org/r/20201118011104.296999-5-tyreld@linux.ibm.comAcked-by:
-
Tyrel Datwyler authored
Testing the NPIV Login response capabilities is a long winded process of dereferencing the vhost->login_buf->resp.capabilities field, then byte swapping that value to host endian, and performing the bitwise test. Currently we only ever check this in ibmvfc_cancel_all(), but follow-up patches will need to regularly check for targetWWPN and channelization support. Add a helper to simplify checking various VIOS capabilities, namely ibmvfc_check_caps(). Link: https://lore.kernel.org/r/20201118011104.296999-4-tyreld@linux.ibm.comAcked-by:
-
Tyrel Datwyler authored
Introduce a target_wwpn field to several MADs. Its possible that a SCSI ID of a target can change due to some fabric changes. The WWPN of the SCSI target provides a better way to identify the target. Also, add flags for receiving MAD versioning information and advertising client support for targetWWPN with the VIOS. This latter capability flag will be required for future clients capable of requesting multiple hardware queues from the host adapter. Link: https://lore.kernel.org/r/20201118011104.296999-3-tyreld@linux.ibm.comAcked-by:
-
Tyrel Datwyler authored
The virtual FC frame command exchanged with the VIOS is used for device reset and command abort TMF as well as normally queued commands. When initializing the ibmvfc_cmd there are several elements of the command that are set the same way regardless of the command type. Deduplicate code by moving these commonally set fields into a initialization helper routine, namely ibmvfc_init_vfc_cmd(). Link: https://lore.kernel.org/r/20201118011104.296999-2-tyreld@linux.ibm.comAcked-by:
-
Tyrel Datwyler authored
The vfcFrame correlation field is a 64bit handle that is intended to trace I/O operations through both the client stack and VIOS stack when the underlying physical FC adapter supports tagging. Tag vfcFrames with the associated ibmvfc_event pointer handle. Link: https://lore.kernel.org/r/20201117185031.129939-3-tyreld@linux.ibm.comSigned-off-by:
-
Tyrel Datwyler authored
Remove a superfluous semicolon following a closing function block bracket. Link: https://lore.kernel.org/r/20201117185031.129939-2-tyreld@linux.ibm.comSigned-off-by:
-
Tyrel Datwyler authored
Both ibmvfc_show_host_(capabilities|npiv_version) functions retrieve values from vhost->login_buf.resp buffer. This is the MAD response buffer from the VIOS and as such any multi-byte non-string values are in big endian format. Byte swap these values to host CPU endian format for better human readability. Link: https://lore.kernel.org/r/20201117185031.129939-1-tyreld@linux.ibm.comSigned-off-by:
-
Leo Liou authored
We have LBA and length for unmap commands. Link: https://lore.kernel.org/r/20201117165839.1643377-8-jaegeuk@kernel.orgReviewed-by:
Leo Liou <leoliou@google.com> Signed-off-by:
Jaegeuk Kim <jaegeuk@kernel.org> Signed-off-by:
-
Jaegeuk Kim authored
The following call stack prevents clk_gating at every I/O completion. We can remove the condition, ufshcd_any_tag_in_use(), since clkgating_work will check it again. ufshcd_complete_requests(struct ufs_hba *hba) ufshcd_transfer_req_compl() __ufshcd_transfer_req_compl() __ufshcd_release(hba) if (ufshcd_any_tag_in_use() == 1) return; ufshcd_tmc_handler(hba); blk_mq_tagset_busy_iter(); Note that this still requires work to deal with a potential race condition when user sets clkgating.delay_ms to very small value. That can cause preventing clkgating by the check of ufshcd_any_tag_in_use() in gate_work. Link: https://lore.kernel.org/r/20201117165839.1643377-7-jaegeuk@kernel.org Fixes: 7252a360 ("scsi: ufs: Avoid busy-waiting by eliminating tag conflicts") Reviewed-by:Asutosh Das <asutoshd@codeaurora.org> Reviewed-by:
-
Jaegeuk Kim authored
This adds user-friendly tracepoints with group id. Link: https://lore.kernel.org/r/20201117165839.1643377-6-jaegeuk@kernel.orgReviewed-by:
Jaegeuk Kim <jaegeuk@google.com> Signed-off-by:
-
