- 01 Aug, 2021 16 commits
-
-
Avri Altman authored
In host mode, the host is expected to send HPB WRITE BUFFER with buffer-id = 0x1 when it inactivates a region. Use the map-requests pool as there is no point in assigning a designated cache for umap-requests. [mkp: REQ_OP_DRV_*] Link: https://lore.kernel.org/r/20210712095039.8093-7-avri.altman@wdc.comReviewed-by:
Daejun Park <daejun7.park@samsung.com> Signed-off-by:
Avri Altman <avri.altman@wdc.com> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com>
-
Avri Altman authored
In host mode, eviction is considered an extreme measure. Verify that the entering region has enough reads, and the exiting region has fewer reads. Link: https://lore.kernel.org/r/20210712095039.8093-6-avri.altman@wdc.comReviewed-by:
-
Avri Altman authored
In host control mode, reads are the major source of activation trials. Keep track of those reads counters, for both active as well inactive regions. We reset the read counter upon write - we are only interested in "clean" reads. Keep those counters normalized, as we are using those reads as a comparative score, to make various decisions. If during consecutive normalizations an active region has exhaust its reads - inactivate it. While at it, protect the {active,inactive}_count stats by adding them into the applicable handler. Link: https://lore.kernel.org/r/20210712095039.8093-5-avri.altman@wdc.comReviewed-by: -
Avri Altman authored
Given a transfer length, set_dirty meticulously iterates over all the entries, across subregions and regions if needed. Currently its only use is to mark dirty blocks, but HCM may benefit from it as well to manage its read counters. Link: https://lore.kernel.org/r/20210712095039.8093-4-avri.altman@wdc.comReviewed-by:
-
Avri Altman authored
In device control mode, the device may recommend the host to either activate or inactivate a region, and the host should follow. Meaning those are not actually recommendations, but more of instructions. Conversely, in host control mode, the recommendation protocol is slightly changed: a) The device may only recommend the host to update a subregion of an already-active region. And, b) The device may *not* recommend to inactivate a region. Furthermore, in host control mode, the host may choose not to follow any of the device's recommendations. However, in case of a recommendation to update an active and clean subregion, it is better to follow those recommendation because otherwise the host has no other way to know that some internal relocation took place. Link: https://lore.kernel.org/r/20210712095039.8093-3-avri.altman@wdc.comReviewed-by: -
Avri Altman authored
We will use control_mode later when we need to differentiate between device and host control modes. Link: https://lore.kernel.org/r/20210712095039.8093-2-avri.altman@wdc.comReviewed-by:
-
Daejun Park authored
Version 2.0 of HBP supports reads of varying sizes from 4KB to 1MB. A read operation <= 32KB is supported as single HPB read. A read between 36KB and 1MB is supported by a combination of write buffer command and HPB read command to deliver more PPN. The write buffer commands may not be issued immediately due to busy tags. To use HPB read more aggressively, the driver can requeue the write buffer command. The requeue threshold is implemented as timeout and can be modified with requeue_timeout_ms entry in sysfs. [mkp: REQ_OP_DRV_* and blk_rq_is_passthrough()] Link: https://lore.kernel.org/r/20210712090025epcms2p3b3d94f6f1b2cfa394e3d9ba130ca0fa7@epcms2p3Tested-by:
Can Guo <cang@codeaurora.org> Tested-by:
Stanley Chu <stanley.chu@mediatek.com> Reviewed-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Reviewed-by:
Bean Huo <beanhuo@micron.com> Reviewed-by:
-
Daejun Park authored
If the logical address of a read I/O belongs to an active sub-region, the HPB driver modifies the read I/O command to an HPB read. The driver modifies the UFS UPIU instead of modifying the existing SCSI command. In HPB version 1.0, the maximum read I/O size that can be converted to HPB read is 4KB. The dirty map of the active sub-region prevents an incorrect HPB read that has stale physical page number which is updated by previous write I/O. [mkp: REQ_OP_DRV_* and blk_rq_is_passthrough()] Link: https://lore.kernel.org/r/20210712085936epcms2p4b0ec5c8cecdeea6cc043d684363842b6@epcms2p4Tested-by:
Bart Van Assche <bvanassche@acm.org> Reviewed-by:
-
Daejun Park authored
Implement L2P map management in HPB. The HPB divides logical addresses into several regions. A region consists of several sub-regions. The sub-region is a basic unit where L2P mapping is managed. The driver loads L2P mapping data of each sub-region. The loaded sub-region is called active-state. The HPB driver unloads L2P mapping data as region unit. The unloaded region is called inactive-state. Sub-region/region candidates to be loaded and unloaded are delivered from the UFS device. The UFS device delivers the recommended active sub-region and inactivate region to the driver using sense data. The HPB module performs L2P mapping management on the host through the delivered information. A pinned region is a preset region on the UFS device that is always in activate-state. The data structures for map data requests and L2P mappings use the mempool API, minimizing allocation overhead while avoiding static allocation. The mininum size of the memory pool used in the HPB is implemented as a module parameter so that it can be configurable by the user. To guarantee a minimum memory pool size of 4MB: ufshpb_host_map_kbytes=4096. The map_work manages active/inactive via 2 "to-do" lists: - hpb->lh_inact_rgn: regions to be inactivated - hpb->lh_act_srgn: subregions to be activated These lists are maintained on I/O completion. [mkp: switch to REQ_OP_DRV_*] Link: https://lore.kernel.org/r/20210712085859epcms2p36e420f19564f6cd0c4a45d54949619eb@epcms2p3Tested-by:
-
Daejun Park authored
Implement Host Performance Buffer (HPB) initialization and add function calls to UFS core driver. NAND flash-based storage devices, including UFS, have mechanisms to translate logical addresses of I/O requests to the corresponding physical addresses of the flash storage. In UFS, logical-to-physical-address (L2P) map data, which is required to identify the physical address for the requested I/Os, can only be partially stored in SRAM from NAND flash. Due to this partial loading, accessing the flash address area, where the L2P information for that address is not loaded in the SRAM, can result in serious performance degradation. The basic concept of HPB is to cache L2P mapping entries in host system memory so that both physical block address (PBA) and logical block address (LBA) can be delivered in HPB read command. The HPB read command allows to read data faster than a regular read command in UFS since it provides the physical address (HPB Entry) of the desired logical block in addition to its logical address. The UFS device can access the physical block in NAND directly without searching and uploading L2P mapping table. This improves read performance because the NAND read operation for uploading L2P mapping table is removed. In HPB initialization, the host checks if the UFS device supports HPB feature and retrieves related device capabilities. Then, HPB parameters are configured in the device. Total start-up time of popular applications was measured and the difference observed between HPB being enabled and disabled. Popular applications are 12 game apps and 24 non-game apps. Each test cycle consists of running 36 applications in sequence. We repeated the cycle for observing performance improvement by L2P mapping cache hit in HPB. The following is the test environment: - kernel version: 4.4.0 - RAM: 8GB - UFS 2.1 (64GB) Results: +-------+----------+----------+-------+ | cycle | baseline | with HPB | diff | +-------+----------+----------+-------+ | 1 | 272.4 | 264.9 | -7.5 | | 2 | 250.4 | 248.2 | -2.2 | | 3 | 226.2 | 215.6 | -10.6 | | 4 | 230.6 | 214.8 | -15.8 | | 5 | 232.0 | 218.1 | -13.9 | | 6 | 231.9 | 212.6 | -19.3 | +-------+----------+----------+-------+ We also measured HPB performance using iozone: $ iozone -r 4k -+n -i2 -ecI -t 16 -l 16 -u 16 -s $IO_RANGE/16 -F \ mnt/tmp_1 mnt/tmp_2 mnt/tmp_3 mnt/tmp_4 mnt/tmp_5 mnt/tmp_6 mnt/tmp_7 \ mnt/tmp_8 mnt/tmp_9 mnt/tmp_10 mnt/tmp_11 mnt/tmp_12 mnt/tmp_13 \ mnt/tmp_14 mnt/tmp_15 mnt/tmp_16 Results: +----------+--------+---------+ | IO range | HPB on | HPB off | +----------+--------+---------+ | 1 GB | 294.8 | 300.87 | | 4 GB | 293.51 | 179.35 | | 8 GB | 294.85 | 162.52 | | 16 GB | 293.45 | 156.26 | | 32 GB | 277.4 | 153.25 | +----------+--------+---------+ Link: https://lore.kernel.org/r/20210712085830epcms2p8c1288b7f7a81b044158a18232617b572@epcms2p8Reported-by:
kernel test robot <lkp@intel.com> Tested-by:
-
Dwaipayan Ray authored
The macros cpu_to_le16() and cpu_to_le32() have special cases for constants. Their __constant_<foo> versions are not required. On little endian systems, both cpu_to_le16() and __constant_cpu_to_le16() expand to the same expression. Same is the case with cpu_to_le32(). On big endian systems, cpu_to_le16() expands to __swab16() which has a __builtin_constant_p check. Similarly, cpu_to_le32() expands to __swab32(). Consequently these macros can be safely used with constants, and hence all those uses are converted. This was discovered as a part of a checkpatch evaluation, looking at all reports of WARNING:CONSTANT_CONVERSION error type. Link: https://lore.kernel.org/r/20210716112852.24598-1-dwaipayanray1@gmail.comSigned-off-by:
Dwaipayan Ray <dwaipayanray1@gmail.com> Signed-off-by:
-
Colin Ian King authored
An earlier fix changed the print format specifier for adapter->bios_addr to use %lX. However, the integer is a u32 so the fix was wrong. Fix this by using the correct %X format specifier. Link: https://lore.kernel.org/r/20210730095031.26981-1-colin.king@canonical.com Fixes: 43622697 ("scsi: BusLogic: use %lX for unsigned long rather than %X") Acked-by:
Khalid Aziz <khalid@gonehiking.org> Signed-off-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Maciej W. Rozycki authored
Existing blogic_msg() invocations do not appear to overrun its internal buffer of a fixed length of 100, which would cause stack corruption, but it's easy to miss with possible further updates and a fix is cheap in performance terms, so limit the output produced into the buffer by using vscnprintf() rather than vsprintf(). Link: https://lore.kernel.org/r/alpine.DEB.2.21.2104201939390.44318@angie.orcam.me.ukAcked-by:
Maciej W. Rozycki <macro@orcam.me.uk> Signed-off-by:
-
Maciej W. Rozycki authored
Update BusLogic driver's messaging system to use pr_cont() for continuation lines, bringing messy output: pci 0000:00:13.0: PCI->APIC IRQ transform: INT A -> IRQ 17 scsi: ***** BusLogic SCSI Driver Version 2.1.17 of 12 September 2013 ***** scsi: Copyright 1995-1998 by Leonard N. Zubkoff <lnz@dandelion.com> scsi0: Configuring BusLogic Model BT-958 PCI Wide Ultra SCSI Host Adapter scsi0: Firmware Version: 5.07B, I/O Address: 0x7000, IRQ Channel: 17/Level scsi0: PCI Bus: 0, Device: 19, Address: 0xE0012000, Host Adapter SCSI ID: 7 scsi0: Parity Checking: Enabled, Extended Translation: Enabled scsi0: Synchronous Negotiation: Ultra, Wide Negotiation: Enabled scsi0: Disconnect/Reconnect: Enabled, Tagged Queuing: Enabled scsi0: Scatter/Gather Limit: 128 of 8192 segments, Mailboxes: 211 scsi0: Driver Queue Depth: 211, Host Adapter Queue Depth: 192 scsi0: Tagged Queue Depth: Automatic , Untagged Queue Depth: 3 scsi0: SCSI Bus Termination: Both Enabled , SCAM: Disabled scsi0: *** BusLogic BT-958 Initialized Successfully *** scsi host0: BusLogic BT-958 back to order: pci 0000:00:13.0: PCI->APIC IRQ transform: INT A -> IRQ 17 scsi: ***** BusLogic SCSI Driver Version 2.1.17 of 12 September 2013 ***** scsi: Copyright 1995-1998 by Leonard N. Zubkoff <lnz@dandelion.com> scsi0: Configuring BusLogic Model BT-958 PCI Wide Ultra SCSI Host Adapter scsi0: Firmware Version: 5.07B, I/O Address: 0x7000, IRQ Channel: 17/Level scsi0: PCI Bus: 0, Device: 19, Address: 0xE0012000, Host Adapter SCSI ID: 7 scsi0: Parity Checking: Enabled, Extended Translation: Enabled scsi0: Synchronous Negotiation: Ultra, Wide Negotiation: Enabled scsi0: Disconnect/Reconnect: Enabled, Tagged Queuing: Enabled scsi0: Scatter/Gather Limit: 128 of 8192 segments, Mailboxes: 211 scsi0: Driver Queue Depth: 211, Host Adapter Queue Depth: 192 scsi0: Tagged Queue Depth: Automatic, Untagged Queue Depth: 3 scsi0: SCSI Bus Termination: Both Enabled, SCAM: Disabled scsi0: *** BusLogic BT-958 Initialized Successfully *** scsi host0: BusLogic BT-958 Also diagnostic output such as with the BusLogic=TraceConfiguration parameter is affected and becomes vertical and therefore hard to read. This has now been corrected, e.g.: pci 0000:00:13.0: PCI->APIC IRQ transform: INT A -> IRQ 17 blogic_cmd(86) Status = 30: 4 ==> 4: FF 05 93 00 blogic_cmd(95) Status = 28: (Modify I/O Address) blogic_cmd(91) Status = 30: 1 ==> 1: 01 blogic_cmd(04) Status = 30: 4 ==> 4: 41 41 35 30 blogic_cmd(8D) Status = 30: 14 ==> 14: 45 DC 00 20 00 00 00 00 00 40 30 37 42 1D scsi: ***** BusLogic SCSI Driver Version 2.1.17 of 12 September 2013 ***** scsi: Copyright 1995-1998 by Leonard N. Zubkoff <lnz@dandelion.com> blogic_cmd(04) Status = 30: 4 ==> 4: 41 41 35 30 blogic_cmd(0B) Status = 30: 3 ==> 3: 00 08 07 blogic_cmd(0D) Status = 30: 34 ==> 34: 03 01 07 04 00 00 00 00 00 00 00 00 00 00 00 00 FF 42 44 46 FF 00 00 00 00 00 00 00 00 00 FF 00 FF 00 blogic_cmd(8D) Status = 30: 14 ==> 14: 45 DC 00 20 00 00 00 00 00 40 30 37 42 1D blogic_cmd(84) Status = 30: 1 ==> 1: 37 blogic_cmd(8B) Status = 30: 5 ==> 5: 39 35 38 20 20 blogic_cmd(85) Status = 30: 1 ==> 1: 42 blogic_cmd(86) Status = 30: 4 ==> 4: FF 05 93 00 blogic_cmd(91) Status = 30: 64 ==> 64: 41 46 3E 20 39 35 38 20 20 00 C4 00 04 01 07 2F 07 04 35 FF FF FF FF FF FF FF FF FF FF 01 00 FE FF 08 FF FF 00 00 00 00 00 00 00 01 00 01 00 00 FF FF 00 00 00 00 00 00 00 00 00 00 00 00 00 FC scsi0: Configuring BusLogic Model BT-958 PCI Wide Ultra SCSI Host Adapter etc. Link: https://lore.kernel.org/r/alpine.DEB.2.21.2104201940430.44318@angie.orcam.me.uk Fixes: 4bcc595c ("printk: reinstate KERN_CONT for printing continuation lines") Cc: stable@vger.kernel.org # v4.9+ Acked-by:
-
Christoph Hellwig authored
Set ret to 0 after the initial permission checks to avoid leaking -EPERM for commands without data transfer. Link: https://lore.kernel.org/r/20210731074027.1185545-3-hch@lst.de Fixes: 75ca5640 ("scsi: bsg: Move the whole request execution into the SCSI/transport handlers") Signed-off-by:
Christoph Hellwig <hch@lst.de> Signed-off-by:
-
Christoph Hellwig authored
Set ret to 0 after the initial permission checks to avoid leaking -EPERM for commands without data transfer. Link: https://lore.kernel.org/r/20210731074027.1185545-2-hch@lst.de Fixes: 75ca5640 ("scsi: bsg: Move the whole request execution into the SCSI/transport handlers") Signed-off-by:
-
- 31 Jul, 2021 8 commits
-
-
Vincent Palomares authored
Allow UFS suspend/resume callbacks to run in parallel with other suspend/resume callbacks. This can recoup dozens of milliseconds on the resume path if UFS hardware needs to be powered back on. Suspending and resuming asynchronously is safe to do so long as the driver callbacks only depend on resources made available by either a) parent devices or b) devices explicitly marked as suppliers with device_link_add. Link: https://lore.kernel.org/r/20210728012743.1063928-1-paillon@google.com Cc: Bjorn Helgaas <bhelgaas@google.com> Cc: Jaegeuk Kim <jaegeuk@kernel.org> Cc: Bart Van Assche <bvanassche@acm.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Stanley Chu <stanley.chu@mediatek.com> Cc: Can Guo <cang@codeaurora.org> Cc: Asutosh Das <asutoshd@codeaurora.org> Cc: Avri Altman <avri.altman@wdc.com> Cc: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by:
Vincent Palomares <paillon@google.com> Signed-off-by:
-
James Smart authored
The lpfc_sli4_nvmet_xri_aborted() routine takes out the abts_buf_list_lock and traverses the buffer contexts to match the xri. Upon match, it then takes the context lock before potentially removing the context from the associated buffer list. This violates the lock hierarchy used elsewhere in the driver of locking context, then the abts_buf_list_lock - thus a possible deadlock. Resolve by: after matching, release the abts_buf_list_lock, then take the context lock, and if to be deleted from the list, retake the abts_buf_list_lock, maintaining lock hierarchy. This matches same list lock hierarchy as elsewhere in the driver Link: https://lore.kernel.org/r/20210730163309.25809-1-jsmart2021@gmail.comReported-by:
Jia-Ju Bai <baijiaju1990@gmail.com> Signed-off-by:
James Smart <jsmart2021@gmail.com> Signed-off-by:
-
Colin Ian King authored
There are two spelling mistakes with the same triple l in alloc, one in a comment, the other in a ql_dbg() debug message. Fix them. Link: https://lore.kernel.org/r/20210729082413.4761-1-colin.king@canonical.comSigned-off-by:
-
Christoph Hellwig authored
Remove the amount of indirect calls by making the handler responsible for the entire execution of the request. Link: https://lore.kernel.org/r/20210729064845.1044147-5-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Move the sg_timeout and sg_reserved_size fields into the bsg_device and scsi_device structures as they have nothing to do with generic block I/O. Note that these values are now separate for bsg vs. SCSI device node access, but that just matches how /dev/sg vs the other nodes has always behaved. Link: https://lore.kernel.org/r/20210729064845.1044147-4-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
This was used for the table based SCSI passthough permission checking that is gone now. Link: https://lore.kernel.org/r/20210729064845.1044147-3-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Use the per-device cdev_device_interface to store the bsg data in the char device inode, and thus remove the need to embedd the bsg_class_device structure in the request_queue. Link: https://lore.kernel.org/r/20210729064845.1044147-2-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
cdrom_read_cdda_bpc() relies on sending SCSI command to the low level driver using a REQ_OP_SCSI_IN request. This isn't generic block layer functionality, so move the actual low-level code into the sr driver and call it through a new read_cdda_bpc method in the cdrom_device_ops structure. With this the CDROM code does not have to pull in scsi_normalize_sense() and depend on CONFIG_SCSI_COMMON. Link: https://lore.kernel.org/r/20210730072752.GB23847%40lst.deTested-by:
Anders Roxell <anders.roxell@linaro.org> Signed-off-by:
-
- 29 Jul, 2021 16 commits
-
-
Christoph Hellwig authored
Just call scsi_ioctl() in sg as that has the same effect. Link: https://lore.kernel.org/r/20210724072033.1284840-25-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Split the SG_IO handler from the main scsi_ioctl() routine. Link: https://lore.kernel.org/r/20210724072033.1284840-24-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Split the SCSI_IOCTL_GET_IDLUN handler from the main scsi_ioctl() routine. Link: https://lore.kernel.org/r/20210724072033.1284840-23-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Factor out a helper for the various flavors of START STOP UNIT command ioctls. Link: https://lore.kernel.org/r/20210724072033.1284840-22-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Remove the comment above ioctl_internal_command() which doesn't document this function at all. Link: https://lore.kernel.org/r/20210724072033.1284840-21-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
CONFIG_BLK_SCSI_REQUEST is rather misnamed as it enables building a small amount of code shared by the SCSI initiator, target, and consumers of the scsi_request passthrough API. Rename it and also allow building it as a module. [mkp: add module license] Link: https://lore.kernel.org/r/20210724072033.1284840-20-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Merge the ioctl handling in block/scsi_ioctl.c into its only caller in drivers/scsi/scsi_ioctl.c. Link: https://lore.kernel.org/r/20210724072033.1284840-19-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Remove the separate command filter structure and just use a switch statement (which also cought two duplicate commands), return a bool and give the function a sensible name. Link: https://lore.kernel.org/r/20210724072033.1284840-18-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Move the SCSI command size table to common SCSI code. Link: https://lore.kernel.org/r/20210724072033.1284840-17-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Merge scsi_req_init() into its only caller. Link: https://lore.kernel.org/r/20210724072033.1284840-16-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Move the SCSI-specific bsg code in the SCSI midlayer instead of in the common bsg code. This just keeps the common bsg code block/ and also allows building it as a module. Link: https://lore.kernel.org/r/20210724072033.1284840-15-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Decouple bsg from scsi_cmd_ioctl(). This requires a small amount of code duplication, but will allow moving all SCSI ioctl handling into SCSI midlayer. Link: https://lore.kernel.org/r/20210724072033.1284840-14-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Return the max_sectors value in bytes. Lifted from scsi_ioctl.c. Link: https://lore.kernel.org/r/20210724072033.1284840-13-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Ensure SCSI ULD only has to call a single ioctl helper. This also adds a bunch of missing ioctls to the ch driver, and removes the need for a duplicate implementation of SCSI_IOCTL_SEND_COMMAND command. Link: https://lore.kernel.org/r/20210724072033.1284840-12-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Manually verify that the device is not a partition and the caller has admin privіleges at the beginning of the sr ioctl method and open code the trivial check for sd as well. Link: https://lore.kernel.org/r/20210724072033.1284840-11-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Open code scsi_cmd_blk_ioctl() in its two callers. Link: https://lore.kernel.org/r/20210724072033.1284840-10-hch@lst.deSigned-off-by:
-
