- Jul 08, 2022
-
-
James Smart authored
Update lpfc version to 14.2.0.5 Link: https://lore.kernel.org/r/20220701211425.2708-12-jsmart2021@gmail.com Co-developed-by:
Justin Tee <justin.tee@broadcom.com> Signed-off-by:
James Smart <jsmart2021@gmail.com> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com>
-
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.com Co-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.com Co-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.com Co-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.com Co-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.com Co-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.com Co-developed-by:
-
James Smart authored
Malformed user input to debugfs results in buffer overflow crashes. Adapt input string lengths to fit within internal buffers, leaving space for NULL terminators. Link: https://lore.kernel.org/r/20220701211425.2708-3-jsmart2021@gmail.com Co-developed-by:
-
James Smart authored
In lpfc_nvme_cancel_iocb(), a cqe is created locally from stack storage. The code didn't initialize the total_data_placed word, inheriting stack content. Initialize the total_data_placed word. Link: https://lore.kernel.org/r/20220701211425.2708-2-jsmart2021@gmail.com Co-developed-by:
-
Martin K. Petersen authored
Bring in fixes to resolve a merge conflict in the lpfc driver update. Signed-off-by: -
Bart Van Assche authored
For some technologies, e.g. an ATA bus, resuming can take multiple seconds. Waiting for resume to finish can cause a very noticeable delay. Hence this commit that restores the behavior from before "scsi: core: pm: Rely on the device driver core for async power management" for most SCSI devices. This commit introduces a behavior change: if the START command fails, do not consider this as a SCSI disk resume failure. Link: https://bugzilla.kernel.org/show_bug.cgi?id=215880 Link: https://lore.kernel.org/r/20220630195703.10155-3-bvanassche@acm.org Fixes: a19a93e4 ("scsi: core: pm: Rely on the device driver core for async power management") Cc: Ming Lei <ming.lei@redhat.com> Cc: Hannes Reinecke <hare@suse.de> Cc: John Garry <john.garry@huawei.com> Cc: ericspero@icloud.com Cc: jason600.groome@gmail.com Tested-by:
<jason600.groome@gmail.com> Signed-off-by:
Bart Van Assche <bvanassche@acm.org> Signed-off-by:
-
Bart Van Assche authored
Move the definition of SCSI_QUEUE_DELAY to just above the function that uses it. Link: https://lore.kernel.org/r/20220630195703.10155-2-bvanassche@acm.org Cc: Ming Lei <ming.lei@redhat.com> Cc: Hannes Reinecke <hare@suse.de> Cc: John Garry <john.garry@huawei.com> Signed-off-by:
-
Jiapeng Chong authored
This was found by coccicheck: drivers/scsi/megaraid/megaraid_sas_base.c:3950 process_fw_state_change_wq() warn: inconsistent indenting. Link: https://lore.kernel.org/r/20220630074152.29171-1-jiapeng.chong@linux.alibaba.com Signed-off-by:
Jiapeng Chong <jiapeng.chong@linux.alibaba.com> Signed-off-by:
-
Michael Schmitz authored
Use dma_map_single() for gvp11 driver (leave bounce buffer logic unchanged). Use dma_set_mask_and_coherent() to avoid explicit cache flushes. Compile-tested only. CC: linux-scsi@vger.kernel.org Link: https://lore.kernel.org/r/6d1d88ee-1cf6-c735-1e6d-bafd2096e322@gmail.com Link: https://lore.kernel.org/r/20220630033302.3183-4-schmitzmic@gmail.com Reviewed-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
Michael Schmitz <schmitzmic@gmail.com> Signed-off-by:
-
Michael Schmitz authored
Use dma_map_single() for a2091 driver (leave bounce buffer logic unchanged). Use dma_set_mask_and_coherent() to avoid explicit cache flushes. Compile-tested only. CC: linux-scsi@vger.kernel.org Link: https://lore.kernel.org/r/6d1d88ee-1cf6-c735-1e6d-bafd2096e322@gmail.com Link: https://lore.kernel.org/r/20220630033302.3183-3-schmitzmic@gmail.com Reviewed-by:
-
Michael Schmitz authored
Use dma_map_single() for a3000 driver (leave bounce buffer logic unchanged). Use dma_set_mask_and_coherent() to avoid explicit cache flushes. Compile-tested only. CC: linux-scsi@vger.kernel.org Link: https://lore.kernel.org/r/6d1d88ee-1cf6-c735-1e6d-bafd2096e322@gmail.com Link: https://lore.kernel.org/r/20220630033302.3183-2-schmitzmic@gmail.com Reviewed-by:
-
Li Zhijian authored
sdev_printk() will only accept messages up to 128 bytes. Shorten strings exceeding 128 bytes avoid printing an incomplete sentence like: [ 475.156955] sd 9:0:0:0: Warning! Received an indication that the LUN assignments on this target have changed. The Linux SCSI layer does not automatical Link: https://lore.kernel.org/r/20220630024516.1571209-1-lizhijian@fujitsu.com Suggested-by:
Finn Thain <fthain@linux-m68k.org> Reviewed-by:
Li Zhijian <lizhijian@fujitsu.com> Signed-off-by:
-
Mike Christie authored
On our backend we can do something similar to LIO where we can enable and disable UNMAP support on the fly. In the SCSI/block layer we can detect this by just doing a rescan. However, LIO cannot detect this change because we only check during the initial configuration. This patch allows UNMAP detection to also happen when the user tries to turn it on. Link: https://lore.kernel.org/r/20220628200230.15052-6-michael.christie@oracle.com Reviewed-by:
Christoph Hellwig <hch@lst.de> Signed-off-by:
Mike Christie <michael.christie@oracle.com> Signed-off-by:
-
Mike Christie authored
Move file's UNMAP setup code to a configure_unmap callout. Link: https://lore.kernel.org/r/20220628200230.15052-5-michael.christie@oracle.com Reviewed-by:
-
Mike Christie authored
Move iblock's UNMAP setup code to a configure_unmap callout. Link: https://lore.kernel.org/r/20220628200230.15052-4-michael.christie@oracle.com Reviewed-by:
-
Mike Christie authored
Add a callout to configure a backend's UNMAP settings. This will be used to allow userspace to configure UNMAP after the initial device setup, similar to how we can set up the other attributes post device configuration. Link: https://lore.kernel.org/r/20220628200230.15052-3-michael.christie@oracle.com Reviewed-by:
-
Mike Christie authored
We use WSNZ=1 so if we get a WRITE_SAME with zero logical blocks we are supposed to fail it. We do this check and failure in target_core_sbc.c before calling into the backend, so we can remove the incorrect check in target_core_file. Link: https://lore.kernel.org/r/20220628200230.15052-2-michael.christie@oracle.com Reviewed-by:
-
Sreekanth Reddy authored
Increase cmd_per_lun to 128. Link: https://lore.kernel.org/r/20220628074848.5036-3-sreekanth.reddy@broadcom.com Signed-off-by:
Sreekanth Reddy <sreekanth.reddy@broadcom.com> Signed-off-by:
-
Sreekanth Reddy authored
Enable shared host tagset to make sure that total outstanding I/O count can not exceed controller's can_queue setting. Link: https://lore.kernel.org/r/20220628074848.5036-2-sreekanth.reddy@broadcom.com Signed-off-by:
-
Dan Carpenter authored
There is a copy and paste bug here. It should check ".rsp" instead of ".req". The error message is copy and pasted as well so update that too. Link: https://lore.kernel.org/r/YrK1A/t3L6HKnswO@kili Fixes: 9c40c36e ("scsi: qla2xxx: edif: Reduce Initiator-Initiator thrashing") Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
-
- Jun 28, 2022
-
-
ChanWoo Lee authored
Checks information about tx_lanes, but is not used. Since commit 1e1e465c ("scsi/ufs: qcom: Remove ufs_qcom_phy_*() calls from host"), tx_lanes is deprecated. As a result, ufs_qcom_link_startup_notify -> POST_CHANGE action does nothing. If it is not going to be updated, it can be removed. Link: https://lore.kernel.org/r/20220627235545.16943-1-cw9316.lee@samsung.com Reviewed-by:
ChanWoo Lee <cw9316.lee@samsung.com> Signed-off-by:
-
Alice Chao authored
NULL pointer access issue was found for the regulator released by ufs_mtk_vreg_fix_vccq(). Simply fix this issue by clearing the released vreg pointer in ufs_hba struct. Link: https://lore.kernel.org/r/20220623035052.18802-9-stanley.chu@mediatek.com Reviewed-by:
Stanley Chu <stanley.chu@mediatek.com> Reviewed-by:
Alice Chao <alice.chao@mediatek.com> Signed-off-by:
-
Peter Wang authored
Add pm-qos request to support performance boosting in MediaTek UFS platforms. At the same time, adjust the order of function calls to be symmetric during the low-power control flow. Link: https://lore.kernel.org/r/20220623035052.18802-8-stanley.chu@mediatek.com Reviewed-by:
Peter Wang <peter.wang@mediatek.com> Signed-off-by:
-
Po-Wen Kao authored
Add interfaces for controlling the host power to optimize the power consumption in MediaTek UFS platforms. Link: https://lore.kernel.org/r/20220623035052.18802-7-stanley.chu@mediatek.com Reviewed-by:
Po-Wen Kao <powen.kao@mediatek.com> Signed-off-by:
-
Po-Wen Kao authored
In MediaTek UFS platforms, UniPro will not return reset confirm if it is in POWERDOWN state thus hang issue may happen while disabling UFSHCI. Simply disable this feature before UniPro leaves POWERDOWN state. Link: https://lore.kernel.org/r/20220623035052.18802-6-stanley.chu@mediatek.com Reviewed-by:
-
Peter Wang authored
Add "PRE_CHANGE" and "POST_CHANGE" information for ref-clk control to precisely configure the low-power state of the parent of ref-clk. Link: https://lore.kernel.org/r/20220623035052.18802-5-stanley.chu@mediatek.com Reviewed-by:
-
Peter Wang authored
Some UFSHCI hosts in MediaTek UFS platform need workaround to prevent host hang issue by setting CLK_CG bit before host is enabled. This operation shall have no side effect on those platforms which do not support this bit. Link: https://lore.kernel.org/r/20220623035052.18802-4-stanley.chu@mediatek.com Reviewed-by:
-
Peter Wang authored
MediaTek decides to always add delays before and after VCC is turned-off. Link: https://lore.kernel.org/r/20220623035052.18802-3-stanley.chu@mediatek.com Reviewed-by:
-
Stanley Chu authored
Fix build warnings: 1. ../drivers/ufs/host/ufs-mediatek.c:1375:5: error: no previous prototype for function 'ufs_mtk_system_suspend' [-Werror,-Wmissing-prototypes] int ufs_mtk_system_suspend(struct device *dev) ^ ../drivers/ufs/host/ufs-mediatek.c:1375:1: note: declare 'static' if the function is not intended to be used outside of this translation unit int ufs_mtk_system_suspend(struct device *dev) ^ static 2. ../drivers/ufs/host/ufs-mediatek.c:702:50: error: format specifies type 'unsigned long' but the argument has type 'int' [-Werror,-Wformat] snprintf(vcc_name, MAX_VCC_NAME, "vcc-ufs%lu", ver); ~~~ ^~~ %d Link: https://lore.kernel.org/r/20220623035052.18802-2-stanley.chu@mediatek.com Reviewed-by: -
Arnd Bergmann authored
The dpt_i2o driver was fixed to stop using virt_to_bus() in 2008, but it still has a stale reference in an error handling code path that could never work. I submitted a patch to fix this reference earlier, but Hannes Reinecke suggested that removing the driver may be just as good here. The i2o driver layer was removed in 2015 with commit 4a72a7af ("staging: remove i2o subsystem"), but the even older dpt_i2o scsi driver stayed around. The last non-cleanup patches I could find were from Miquel van Smoorenburg and Mark Salyzyn back in 2008, they might know if there is any chance of the hardware still being used anywhere. Link: https://lore.kernel.org/linux-scsi/CAK8P3a1XfwkTOV7qOs1fTxf4vthNBRXKNu8A5V7TWnHT081NGA@mail.gmail.com/T/ Link: https://lore.kernel.org/r/20220624155226.2889613-3-arnd@kernel.org Cc: Miquel van Smoorenburg <mikevs@xs4all.net> Cc: Mark Salyzyn <salyzyn@android.com> Cc: Hannes Reinecke <hare@suse.de> Signed-off-by:
-
Arnd Bergmann authored
The BusLogic driver is the last remaining driver that relies on the deprecated bus_to_virt() function, which in turn only works on a few architectures, and is incompatible with both swiotlb and iommu support. Before commit 391e2f25 ("[SCSI] BusLogic: Port driver to 64-bit."), the driver had a dependency on x86-32, presumably because of this problem. However, the change introduced another bug that made it still impossible to use the driver on any 64-bit machine. This was in turn fixed in commit 56f39614 ("scsi: BusLogic: Fix 64-bit system enumeration error for Buslogic"), 8 years later, which shows that there are not a lot of users. Maciej is still using the driver on 32-bit hardware, and Khalid mentioned that the driver works with the device emulation used in VirtualBox and VMware. Both of those only emulate it for Windows 2000 and older operating systems that did not ship with the better LSI logic driver. Do a minimum fix that searches through the list of descriptors to find one that matches the bus address. This is clearly as inefficient as was indicated in the code comment about the lack of a bus_to_virt() replacement. A better fix would likely involve changing out the entire descriptor allocation for a simpler one, but that would be much more invasive. Link: https://lore.kernel.org/r/20220624155226.2889613-2-arnd@kernel.org Cc: Maciej W. Rozycki <macro@orcam.me.uk> Cc: Matt Wang <wwentao@vmware.com> Tested-by:
Khalid Aziz <khalid@gonehiking.org> Reviewed-by:
Robin Murphy <robin.murphy@arm.com> Reviewed-by:
Hannes Reinecke <hare@suse.de> Acked-by:
-
Colin Ian King authored
Variable wlen is being assigned a value that is never read, it is being re-assigned with a different value later on. The assignment is redundant and can be removed. Cleans up clang scan build warning: drivers/scsi/fcoe/fcoe.c:1491:2: warning: Value stored to 'wlen' is never read [deadcode.DeadStores] Link: https://lore.kernel.org/r/20220623164710.76831-1-colin.i.king@gmail.com Signed-off-by:
Colin Ian King <colin.i.king@gmail.com> Signed-off-by:
-
John Garry authored
If the controller is behind an IOMMU then the IOMMU IOVA caching range can affect performance, as discussed in [0]. Limit the max HW sectors to not exceed this limit. We need to hardcode the value until a proper DMA mapping API is available. [0] https://lore.kernel.org/linux-iommu/20210129092120.1482-1-thunder.leizhen@huawei.com/ Link: https://lore.kernel.org/r/1655988119-223714-1-git-send-email-john.garry@huawei.com Signed-off-by:
John Garry <john.garry@huawei.com> Signed-off-by:
-
