- 23 Jun, 2021 6 commits
-
-
Hannes Reinecke authored
When a sense code is present we should not override the SAM status; the driver already sets it based on the response from the hypervisor. In addition we should only copy the sense buffer if one is actually provided by the hypervisor. Link: https://lore.kernel.org/r/20210622091153.29231-1-hare@suse.de Fixes: 464a00c9 ("scsi: core: Kill DRIVER_SENSE") Tested-by:
Guenter Roeck <linux@roeck-us.net> Tested-by:
Jiri Slaby <jirislaby@kernel.org> Signed-off-by:
Hannes Reinecke <hare@suse.de> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com>
-
Yufen Yu authored
Offlining a SATA device connected to a hisi SAS controller and then scanning the host will result in detecting 255 non-existent devices: # lsscsi [2:0:0:0] disk ATA Samsung SSD 860 2B6Q /dev/sda [2:0:1:0] disk ATA WDC WD2003FYYS-3 1D01 /dev/sdb [2:0:2:0] disk SEAGATE ST600MM0006 B001 /dev/sdc # echo "offline" > /sys/block/sdb/device/state # echo "- - -" > /sys/class/scsi_host/host2/scan # lsscsi [2:0:0:0] disk ATA Samsung SSD 860 2B6Q /dev/sda [2:0:1:0] disk ATA WDC WD2003FYYS-3 1D01 /dev/sdb [2:0:1:1] disk ATA WDC WD2003FYYS-3 1D01 /dev/sdh ... [2:0:1:255] disk ATA WDC WD2003FYYS-3 1D01 /dev/sdjb After a REPORT LUN command issued to the offline device fails, the SCSI midlayer tries to do a sequential scan of all devices whose LUN number is not 0. However, SATA does not support LUN numbers at all. Introduce a generic sas_slave_alloc() handler which will return -ENXIO for SATA devices if the requested LUN number is larger than 0 and make libsas drivers use this function as their .slave_alloc callback. Link: https://lore.kernel.org/r/20210622034037.1467088-1-yuyufen@huawei.comReported-by:
Wu Bo <wubo40@huawei.com> Suggested-by:
John Garry <john.garry@huawei.com> Reviewed-by:
Jason Yan <yanaijie@huawei.com> Signed-off-by:
Yufen Yu <yuyufen@huawei.com> Signed-off-by:
-
Bart Van Assche authored
Since scsi_mq_alloc_queue() only has one caller, inline it. This change was suggested by Christoph Hellwig. Link: https://lore.kernel.org/r/20210622024654.12543-1-bvanassche@acm.org Cc: Christoph Hellwig <hch@lst.de> Cc: Ming Lei <ming.lei@redhat.com> Cc: Ed Tsai <ed.tsai@mediatek.com> Reviewed-by:
Christoph Hellwig <hch@lst.de> Reviewed-by:
Ming Lei <ming.lei@redhat.com> Signed-off-by:
Bart Van Assche <bvanassche@acm.org> Signed-off-by:
-
Colin Ian King authored
The shifting of the u8 integer returned fom ahc_inb(ahc, port+3) by 24 bits to the left will be promoted to a 32 bit signed int and then sign-extended to a u64. In the event that the top bit of the u8 is set then all then all the upper 32 bits of the u64 end up as also being set because of the sign-extension. Fix this by casting the u8 values to a u64 before the 24 bit left shift. [ This dates back to 2002, I found the offending commit from the git history git://git.kernel.org/pub/scm/linux/kernel/git/tglx/history.git, commit f58eb66c ("Update aic7xxx driver to 6.2.10...") ] Link: https://lore.kernel.org/r/20210621151727.20667-1-colin.king@canonical.comSigned-off-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
James Smart authored
vport is linked onto the driver's vport list at allocation, but failure path fails to remove it from the list. Change location of linkage until after complete vport completion. Link: https://lore.kernel.org/r/20210619155729.20049-1-jsmart2021@gmail.com Fixes: 692e5d73 ("scsi: elx: efct: LIO backend interface routines") Reported-by:
Dan Carpenter <dan.carpenter@oracle.com> Co-developed-by:
Ram Vegesna <ram.vegesna@broadcom.com> Signed-off-by:
James Smart <james.smart@broadcom.com> Signed-off-by:
-
James Smart authored
Flags value is being set to a constant and ANDed with 0 which always results in 0. Remove the assignment line. Link: https://lore.kernel.org/r/20210619155641.19942-1-jsmart2021@gmail.com Fixes: 1628f5b4 ("scsi: elx: libefc_sli: Populate and post different WQEs") Reported-by:
-
- 19 Jun, 2021 24 commits
-
-
Kees Cook authored
In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memcpy(), memmove(), and memset(), avoid intentionally reading across neighboring array fields. SCtmp->sense_buffer is 96 bytes, but ecbptr->sense is 14 bytes. Instead of over-reading ecbptr->sense, copy only the actual contents and zero pad the remaining bytes, avoiding potential over-reads. Link: https://lore.kernel.org/r/20210616212437.1727088-1-keescook@chromium.orgSigned-off-by:
Kees Cook <keescook@chromium.org> Signed-off-by:
-
Kees Cook authored
In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memcpy(), memmove(), and memset(), avoid intentionally reading across neighboring array fields. pcmd->sense_buffer is 96 bytes, and was being manually zero-filled. However, struct SENSE_DATA is 18 bytes, with ccb->arcmsr_cdb.SenseData only being 15 bytes, resulting in a 3 byte over-read. Copy only the contents of ccb->arcmsr_cdb.SenseData and zero fill the remainder, avoiding potential over-reads. Link: https://lore.kernel.org/r/20210616212428.1726958-1-keescook@chromium.orgSigned-off-by:
-
Kees Cook authored
In preparation for FORTIFY_SOURCE performing compile-time and run-time field bounds checking for memcpy() avoid intentionally reading across neighboring array fields. scb->scsi_cmd->sense_buffer is 96 bytes: #define SCSI_SENSE_BUFFERSIZE 96 tapeDCDB->sense_info is 56 bytes: typedef struct { ... uint8_t sense_info[56]; } IPS_DCDB_TABLE_TAPE, ... scb->dcdb.sense_info is 64 bytes: typedef struct { ... uint8_t sense_info[64]; ... } IPS_DCDB_TABLE, ... Copying 96 bytes from either was copying beyond the end of the respective buffers, leading to potential memory content exposures. Correctly copy the actual buffer contents and zero pad the remaining bytes. Link: https://lore.kernel.org/r/20210616212408.1726812-1-keescook@chromium.orgSigned-off-by: -
Zou Wei authored
The function is missing a of_node_put() on node. Fix this by adding the call before returning. Link: https://lore.kernel.org/r/1623929522-4389-1-git-send-email-zou_wei@huawei.comReported-by:
Hulk Robot <hulkci@huawei.com> Reviewed-by:
Stanley Chu <stanley.chu@mediatek.com> Signed-off-by:
Zou Wei <zou_wei@huawei.com> Signed-off-by:
-
Dan Carpenter authored
Calling a nested spin_lock_irqsave() will overwrite the original "flags" so that they can not be enabled again at the end. Link: https://lore.kernel.org/r/YMyjH16k4M1yEmmU@mwanda Fixes: 3146240f ("scsi: elx: libefc: FC Domain state machine interfaces") Signed-off-by:
-
Colin Ian King authored
The comparison of the u32 variable rc to less than zero always false because it is unsigned. Fix this by making it an int. Link: https://lore.kernel.org/r/20210616170401.15831-1-colin.king@canonical.com Fixes: 202bfdff ("scsi: elx: libefc: FC node ELS and state handling") Signed-off-by:
-
James Smart authored
debugfs_create_xxx routines, which return pointers, are being checked for error by looking for NULL values. The routines may return pointer-munged -Exxx codes, so they should be using IS_ERR() to adapt. There are two cases: - The first case is on initial directory creation, which actually doesn't need to be checked. So remove the check. - Creation of the sessions subdirectory. Modify this creation to create under the initial directory created, and fix failure check. Link: https://lore.kernel.org/r/20210618233004.83769-1-jsmart2021@gmail.com Fixes: 4df84e84 ("scsi: elx: efct: Driver initialization routines") Reported-by:
James Smart <jsmart2021@gmail.com> Signed-off-by:
-
James Smart authored
efct_hw_iotype_is_originator() is returning a negative (-EIO) status which doesn't make sense for a u8 function type. Reviewing the code, the function only needs to return true/false, thus a bool status is most appropriate. Change the function return type and patch up the one callee as the bool inverses the if check. Link: https://lore.kernel.org/r/20210618231524.83179-1-jsmart2021@gmail.com Fixes: 4df84e84 ("scsi: elx: efct: Driver initialization routines") Reported-by:
-
James Smart authored
cmpxchg is being used on a bool type, which is requiring architecture support that isn't compatible with a bool. Convert variable abort_in_progress from bool to int. Link: https://lore.kernel.org/r/20210618174050.80302-1-jsmart2021@gmail.com Fixes: ebc076b3 ("scsi: elx: efct: Tie into kernel Kconfig and build process") Reported-by:
kernel test robot <lkp@intel.com> Co-developed-by:
-
Nathan Chancellor authored
clang warns: drivers/scsi/elx/efct/efct_hw.c:1523:17: warning: address of array 'ctx->buf' will always evaluate to 'true' [-Wpointer-bool-conversion] (!ctx->buf ? U32_MAX : *((u32 *)ctx->buf))); ~~~~~~^~~ buf is an array in the middle of a struct so deferencing it is not a problem as long as ctx is not NULL. Eliminate the check, which fixes the warning. Link: https://github.com/ClangBuiltLinux/linux/issues/1398 Link: https://lore.kernel.org/r/20210617063123.21239-1-nathan@kernel.org Fixes: 580c0255 ("scsi: elx: efct: RQ buffer, memory pool allocation and deallocation APIs") Reviewed-by:Nathan Chancellor <nathan@kernel.org> Signed-off-by:
-
Nathan Chancellor authored
clang warns: drivers/scsi/elx/efct/efct_lio.c:1216:24: warning: variable 'id' is uninitialized when used here [-Wuninitialized] se_sess, node, id); ^~ Shuffle the debug print after id's initialization so that the actual value is printed. Link: https://github.com/ClangBuiltLinux/linux/issues/1397 Link: https://lore.kernel.org/r/20210617061721.2405511-1-nathan@kernel.org Fixes: 692e5d73 ("scsi: elx: efct: LIO backend interface routines") Reviewed-by: -
Wei Yongjun authored
Fix to return negative error code -ENOMEM from the error handling case instead of 0. Also fix typo in error message. Link: https://lore.kernel.org/r/20210617024837.1023069-1-weiyongjun1@huawei.com Fixes: 4df84e84 ("scsi: elx: efct: Driver initialization routines") Reported-by:
Wei Yongjun <weiyongjun1@huawei.com> Signed-off-by:
-
Colin Ian King authored
The variable "lun" is being initialized with a value that is never read, it is being updated later on. The assignment is redundant and can be removed. Link: https://lore.kernel.org/r/20210616171621.16176-1-colin.king@canonical.comReviewed-by:
-
Colin Ian King authored
There is a spelling mistake in a efc_log_info message. Fix it. Link: https://lore.kernel.org/r/20210616142637.12706-1-colin.king@canonical.comSigned-off-by:
-
James Smart authored
Integration with VMID patches resulted in a build error when CONFIG_DEBUG_FS is disabled and driver option CONFIG_SCSI_LPFC_DEBUG_FS is disabled. It results in an undefined variable: lpfc_scsi:5595:3: error: 'uuid' undeclared (first use in this function); did you mean 'upid'? Link: https://lore.kernel.org/r/20210618171842.79710-1-jsmart2021@gmail.com Fixes: 33c79741 ("scsi: lpfc: vmid: Introduce VMID in I/O path") Reported-by:
-
Colin Ian King authored
The continue statement at the end of a loop has no effect, remove it. Link: https://lore.kernel.org/r/20210617114347.10247-1-colin.king@canonical.comSigned-off-by:
-
Colin Ian King authored
The continue statement at the end of a for-loop has no effect, remove it. Link: https://lore.kernel.org/r/20210617073743.151008-1-colin.king@canonical.comSigned-off-by:
-
Andy Shevchenko authored
Switch to use module_parport_driver() to reduce boilerplate code. Link: https://lore.kernel.org/r/20210616142540.45676-1-andriy.shevchenko@linux.intel.comSigned-off-by:
Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by:
-
Andy Shevchenko authored
Switch to use module_parport_driver() to reduce boilerplate code. Link: https://lore.kernel.org/r/20210616142429.45373-1-andriy.shevchenko@linux.intel.comSigned-off-by:
-
Zhen Lei authored
When an expander does not contain any 'phys', an appropriate error code -1 should be returned, as done elsewhere in this function. However, we currently do not explicitly assign this error code to 'rc'. As a result, 0 was incorrectly returned. Link: https://lore.kernel.org/r/20210514081300.6650-1-thunder.leizhen@huawei.com Fixes: f92363d1 ("[SCSI] mpt3sas: add new driver supporting 12GB SAS") Reported-by:
Zhen Lei <thunder.leizhen@huawei.com> Signed-off-by:
-
Suganath Prabu S authored
Fix the structurally dead code (UNREACHABLE) type of error reported by Coverity. Link: https://lore.kernel.org/r/20210618155506.2609112-1-suganath-prabu.subramani@broadcom.comReported-by:
Suganath Prabu S <suganath-prabu.subramani@broadcom.com> Signed-off-by:
-
Yang Yingliang authored
Fix the following warnings: drivers/scsi/mpi3mr/mpi3mr_os.c:24:5: warning: symbol 'prot_mask' was not declared. Should it be static? drivers/scsi/mpi3mr/mpi3mr_os.c:28:5: warning: symbol 'prot_guard_mask' was not declared. Should it be static? drivers/scsi/mpi3mr/mpi3mr_os.c:31:5: warning: symbol 'logging_level' was not declared. Should it be static? Link: https://lore.kernel.org/r/20210604071407.1360742-1-yangyingliang@huawei.comReported-by:
Kashyap Desai <kashyap.desai@broadcom.com> Signed-off-by:
Yang Yingliang <yangyingliang@huawei.com> Signed-off-by:
-
Yang Yingliang authored
Fix to return a negative error code from the error handling case instead of 0 as done elsewhere in this function. Link: https://lore.kernel.org/r/20210603151653.711020-1-yangyingliang@huawei.com Fixes: fb9b0457 ("scsi: mpi3mr: Add support for recovering controller") Fixes: 824a1566 ("scsi: mpi3mr: Base driver code") Reported-by:
-
Yang Yingliang authored
Goto unlock path before return from function in the error handling case. Link: https://lore.kernel.org/r/20210603152803.717505-1-yangyingliang@huawei.com Fixes: c9566231 ("scsi: mpi3mr: Create operational request and reply queue pair") Reported-by:
-
- 16 Jun, 2021 10 commits
-
-
James Smart authored
This final patch ties the efct driver into the kernel Kconfig and build linkages in the drivers/scsi directory. Link: https://lore.kernel.org/r/20210601235512.20104-32-jsmart2021@gmail.comReported-by:
Daniel Wagner <dwagner@suse.de> Co-developed-by:
-
James Smart authored
Add efct driver Kconfig and Makefiles. Link: https://lore.kernel.org/r/20210601235512.20104-31-jsmart2021@gmail.comReviewed-by:
-
James Smart authored
Integration with the scsi_fc_transport host interfaces. Link: https://lore.kernel.org/r/20210601235512.20104-30-jsmart2021@gmail.comReviewed-by:
-
James Smart authored
Implement routines to detach transport and hardware objects. Link: https://lore.kernel.org/r/20210601235512.20104-29-jsmart2021@gmail.comReviewed-by:
-
James Smart authored
Add routines to retrieve link stats and host stats, add firmware update helper routines. Link: https://lore.kernel.org/r/20210601235512.20104-28-jsmart2021@gmail.comReviewed-by:
-
James Smart authored
Routines that write I/O to work queue, send SRRs and raw frames. Link: https://lore.kernel.org/r/20210601235512.20104-27-jsmart2021@gmail.comReviewed-by:
-
James Smart authored
Add LIO backend template registration and template functions. Link: https://lore.kernel.org/r/20210601235512.20104-26-jsmart2021@gmail.comReviewed-by:
-
James Smart authored
Routines for SCSI transport IO alloc, build and send I/O. Link: https://lore.kernel.org/r/20210601235512.20104-25-jsmart2021@gmail.comReviewed-by:
-
James Smart authored
Add routines to handle unsolicited FC frames. Link: https://lore.kernel.org/r/20210601235512.20104-24-jsmart2021@gmail.comReviewed-by:
-
James Smart authored
Add driver definitions for: - Routines for EQ, CQ, WQ and RQ processing. - Routines for I/O object pool allocation and deallocation. Link: https://lore.kernel.org/r/20210601235512.20104-23-jsmart2021@gmail.comReviewed-by:
-
