- Aug 20, 2019
-
-
Martin Wilck authored
Avoid code duplication between qla2x00_alloc_offload_mem() and qla2x00_alloc_fw_dump() by moving the FCE and EFT buffer allocation and initialization to separate functions. Cleanly track failure and success by making sure that the ha->eft, ha->fce and respective eft_dma, fce_dma members are set if and only if the buffers are properly allocated and initialized. Avoid pointless buffer reallocation. Eliminate some goto statements. Make sure the fce_enabled flag is cleared when the FCE buffer is freed. Fixes: ad0a0b01 ("scsi: qla2xxx: Fix Firmware dump size for Extended login and Exchange Offload") Fixes: a28d9e4e ("scsi: qla2xxx: Add support for multiple fwdump templates/segments") Cc: Joe Carnuccio <joe.carnuccio@cavium.com> Cc: Quinn Tran <qutran@marvell.com> Cc: Himanshu Madhani <hmadhani@marvell.com> Cc: Bart Van Assche <bvanassche@acm.org> Signed-off-by:
Martin Wilck <mwilck@suse.com> Tested-by:
Himanshu Madhani <hmadhani@marvell.com> Reviewed-by:
Martin K. Petersen <martin.petersen@oracle.com>
-
Martin Wilck authored
In qla2x00_alloc_fw_dump(), an existing EFT buffer (e.g. from previous invocation of qla2x00_alloc_offload_mem()) is freed. The buffer is then re-allocated, but without setting the eft and eft_dma fields to the new values. Fixes: a28d9e4e ("scsi: qla2xxx: Add support for multiple fwdump templates/segments") Cc: Joe Carnuccio <joe.carnuccio@cavium.com> Cc: Quinn Tran <qutran@marvell.com> Cc: Himanshu Madhani <hmadhani@marvell.com> Cc: Bart Van Assche <bvanassche@acm.org> Signed-off-by:
-
- Aug 13, 2019
-
-
Masahiro Yamada authored
When CONFIG_UAPI_HEADER_TEST=y, exported headers are compile-tested to make sure they can be included from user-space. Currently, scsi_bsg_fc.h, scsi_netlink.h, and scsi_netlink_fc.h are excluded from the test coverage. To make them join the compile-test, we need to fix the build errors attached below. For a case like this, we decided to use __u{8,16,32,64} variable types in this discussion: https://lkml.org/lkml/2019/6/5/18 Build log: CC usr/include/scsi/scsi_netlink_fc.h.s CC usr/include/scsi/scsi_netlink.h.s CC usr/include/scsi/scsi_bsg_fc.h.s In file included from ./usr/include/scsi/scsi_netlink_fc.h:10:0, from <command-line>:32: ./usr/include/scsi/scsi_netlink.h:29:2: error: unknown type name uint8_t uint8_t version; ^~~~~~~ ./usr/include/scsi/scsi_netlink.h:30:2: error: unknown type name uint8_t uint8_t transport; ^~~~~~~ ./usr/include/scsi/scsi_netlink.h:31:2: error: unknown type name uint16_t uint16_t magic; ^~~~~~~~ ./usr/include/scsi/scsi_netlink.h:32:2: error: unknown type name uint16_t uint16_t msgtype; ^~~~~~~~ CC usr/include/rdma/vmw_pvrdma-abi.h.s ./usr/include/scsi/scsi_netlink.h:33:2: error: unknown type name uint16_t uint16_t msglen; ^~~~~~~~ ./usr/include/scsi/scsi_netlink.h:34:33: error: uint64_t undeclared here (not in a function); did you mean __uint128_t ? } __attribute__((aligned(sizeof(uint64_t)))); ^~~~~~~~ __uint128_t ./usr/include/scsi/scsi_netlink.h:78:2: error: expected specifier-qualifier-list before uint64_t uint64_t vendor_id; ^~~~~~~~ In file included from <command-line>:32:0: ./usr/include/scsi/scsi_netlink_fc.h:46:2: error: expected specifier-qualifier-list before uint64_t uint64_t seconds; ^~~~~~~~ make[2]: *** [scripts/Makefile.build;302: usr/include/scsi/scsi_netlink_fc.h.s] Error 1 make[2]: *** Waiting for unfinished jobs.... In file included from <command-line>:32:0: ./usr/include/scsi/scsi_netlink.h:29:2: error: unknown type name uint8_t uint8_t version; ^~~~~~~ ./usr/include/scsi/scsi_netlink.h:30:2: error: unknown type name uint8_t uint8_t transport; ^~~~~~~ ./usr/include/scsi/scsi_netlink.h:31:2: error: unknown type name uint16_t uint16_t magic; ^~~~~~~~ ./usr/include/scsi/scsi_netlink.h:32:2: error: unknown type name uint16_t uint16_t msgtype; ^~~~~~~~ ./usr/include/scsi/scsi_netlink.h:33:2: error: unknown type name uint16_t uint16_t msglen; ^~~~~~~~ ./usr/include/scsi/scsi_netlink.h:34:33: error: uint64_t undeclared here (not in a function); did you mean __uint128_t ? } __attribute__((aligned(sizeof(uint64_t)))); ^~~~~~~~ __uint128_t ./usr/include/scsi/scsi_netlink.h:78:2: error: expected specifier-qualifier-list before uint64_t uint64_t vendor_id; ^~~~~~~~ make[2]: *** [scripts/Makefile.build;302: usr/include/scsi/scsi_netlink.h.s] Error 1 In file included from <command-line>:32:0: ./usr/include/scsi/scsi_bsg_fc.h:69:2: error: unknown type name uint8_t uint8_t reserved; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:72:2: error: unknown type name uint8_t uint8_t port_id[3]; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:90:2: error: unknown type name uint8_t uint8_t reserved; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:93:2: error: unknown type name uint8_t uint8_t port_id[3]; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:114:2: error: unknown type name uint8_t uint8_t command_code; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:117:2: error: unknown type name uint8_t uint8_t port_id[3]; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:154:2: error: unknown type name uint32_t uint32_t status; /* See FC_CTELS_STATUS_xxx */ ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:158:3: error: unknown type name uint8_t uint8_t action; /* fragment_id for CT REJECT */ ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:159:3: error: unknown type name uint8_t uint8_t reason_code; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:160:3: error: unknown type name uint8_t uint8_t reason_explanation; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:161:3: error: unknown type name uint8_t uint8_t vendor_unique; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:177:2: error: unknown type name uint8_t uint8_t reserved; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:180:2: error: unknown type name uint8_t uint8_t port_id[3]; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:185:2: error: unknown type name uint32_t uint32_t preamble_word0; /* revision & IN_ID */ ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:186:2: error: unknown type name uint32_t uint32_t preamble_word1; /* GS_Type, GS_SubType, Options, Rsvd */ ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:187:2: error: unknown type name uint32_t uint32_t preamble_word2; /* Cmd Code, Max Size */ ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:207:2: error: unknown type name uint64_t uint64_t vendor_id; ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:210:2: error: unknown type name uint32_t uint32_t vendor_cmd[0]; ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:217:2: error: unknown type name uint32_t uint32_t vendor_rsp[0]; ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:236:2: error: unknown type name uint8_t uint8_t els_code; ^~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:254:2: error: unknown type name uint32_t uint32_t preamble_word0; /* revision & IN_ID */ ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:255:2: error: unknown type name uint32_t uint32_t preamble_word1; /* GS_Type, GS_SubType, Options, Rsvd */ ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:256:2: error: unknown type name uint32_t uint32_t preamble_word2; /* Cmd Code, Max Size */ ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:268:2: error: unknown type name uint32_t uint32_t msgcode; ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:292:2: error: unknown type name uint32_t uint32_t result; ^~~~~~~~ ./usr/include/scsi/scsi_bsg_fc.h:295:2: error: unknown type name uint32_t uint32_t reply_payload_rcv_len; ^~~~~~~~ Signed-off-by:Masahiro Yamada <yamada.masahiro@socionext.com> Signed-off-by:
-
Frederick Lawler authored
Commit 8c0d3a02 ("PCI: Add accessors for PCI Express Capability") added accessors for the PCI Express Capability so that drivers didn't need to be aware of differences between v1 and v2 of the PCI Express Capability. Replace pci_read_config_word() and pci_write_config_word() calls with pcie_capability_read_word() and pcie_capability_write_word(). Signed-off-by:
Frederick Lawler <fred@fredlawl.com> Signed-off-by:
-
Frederick Lawler authored
Commit 8c0d3a02 ("PCI: Add accessors for PCI Express Capability") added accessors for the PCI Express Capability so that drivers didn't need to be aware of differences between v1 and v2 of the PCI Express Capability. Replace pci_read_config_word() and pci_write_config_word() calls with pcie_capability_read_word() and pcie_capability_write_word(). Signed-off-by:
-
Colin Ian King authored
Variable retv is initialized to a value that is never read and it is re-assigned later. The initialization is redundant and can be removed. Addresses-Coverity: ("Unused value") Signed-off-by:Colin Ian King <colin.king@canonical.com> Acked-by:
Matthew Wilcox (Oracle) <willy@infradead.org> Signed-off-by:
-
Christophe JAILLET authored
This should be 'pmcraid', not 'pcmraid' Signed-off-by:
Christophe JAILLET <christophe.jaillet@wanadoo.fr> Signed-off-by:
-
Gustavo A. R. Silva authored
Mark switch cases where we are expecting to fall through. Fix the following warnings (Building: rpc_defconfig arm): drivers/scsi/arm/fas216.c: In function fas216_disconnect_intr : drivers/scsi/arm/fas216.c:913:6: warning: this statement may fall through [-Wimplicit-fallthrough=] if (fas216_get_last_msg(info, info->scsi.msgin_fifo) == ABORT) { ^ drivers/scsi/arm/fas216.c:919:2: note: here default: /* huh? */ ^~~~~~~ drivers/scsi/arm/fas216.c: In function fas216_kick : drivers/scsi/arm/fas216.c:1959:3: warning: this statement may fall through [-Wimplicit-fallthrough=] fas216_allocate_tag(info, SCpnt); ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/scsi/arm/fas216.c:1960:2: note: here case TYPE_OTHER: ^~~~ drivers/scsi/arm/fas216.c: In function fas216_busservice_intr : drivers/scsi/arm/fas216.c:1413:3: warning: this statement may fall through [-Wimplicit-fallthrough=] fas216_stoptransfer(info); ^~~~~~~~~~~~~~~~~~~~~~~~~ drivers/scsi/arm/fas216.c:1414:2: note: here case STATE(STAT_STATUS, PHASE_SELSTEPS):/* Sel w/ steps -> Status */ ^~~~ drivers/scsi/arm/fas216.c:1424:3: warning: this statement may fall through [-Wimplicit-fallthrough=] fas216_stoptransfer(info); ^~~~~~~~~~~~~~~~~~~~~~~~~ drivers/scsi/arm/fas216.c:1425:2: note: here case STATE(STAT_MESGIN, PHASE_COMMAND): /* Command -> Message In */ ^~~~ drivers/scsi/arm/fas216.c: In function fas216_funcdone_intr : drivers/scsi/arm/fas216.c:1573:6: warning: this statement may fall through [-Wimplicit-fallthrough=] if ((stat & STAT_BUSMASK) == STAT_MESGIN) { ^ drivers/scsi/arm/fas216.c:1579:2: note: here default: ^~~~~~~ drivers/scsi/arm/fas216.c: In function fas216_handlesync : drivers/scsi/arm/fas216.c:605:20: warning: this statement may fall through [-Wimplicit-fallthrough=] info->scsi.phase = PHASE_MSGOUT_EXPECT; ~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~ drivers/scsi/arm/fas216.c:607:2: note: here case async: ^~~~ Signed-off-by:Gustavo A. R. Silva <gustavo@embeddedor.com> Signed-off-by:
-
Chandrakanth Patil authored
This patch provides a module parameter and sysfs interface to select whether the queue depth for each device should be based on the value suggested by firmware (the default) or the maximum supported by the controller (can_queue). Although we have a sysfs interface per sdev to change the queue depth of individual scsi devices, this implementation provides a single sysfs entry per shost to switch between the controller max and the value reported by firmware. The module parameter can provide an interface for one time grub settings and provides persistent settings across the boot. [mkp: tweaked commit desc] Signed-off-by:
Kashyap Desai <kashyap.desai@broadcom.com> Signed-off-by:
Chandrakanth Patil <chandrakanth.patil@broadcom.com> Signed-off-by:
-
Bart Van Assche authored
BUG: KASAN: null-ptr-deref in qla24xx_handle_plogi_done_event+0x134/0x9f0 [qla2xxx] Read of size 4 at addr 00000000000000a0 by task swapper/2/0 CPU: 2 PID: 0 Comm: swapper/2 Not tainted 5.2.0-dbg+ #1 Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 Call Trace: <IRQ> dump_stack+0x8a/0xd6 __kasan_report.cold+0x5/0x41 kasan_report+0x16/0x20 __asan_load4+0x7e/0x80 qla24xx_handle_plogi_done_event+0x134/0x9f0 [qla2xxx] qla2x00_els_dcmd2_sp_done+0x15f/0x230 [qla2xxx] qla24xx_els_ct_entry+0x3b3/0x610 [qla2xxx] qla24xx_process_response_queue+0x514/0x10e0 [qla2xxx] qla24xx_msix_rsp_q+0x80/0x100 [qla2xxx] __handle_irq_event_percpu+0x72/0x450 handle_irq_event_percpu+0x74/0xf0 handle_irq_event+0x5e/0x8f handle_edge_irq+0x13a/0x320 handle_irq+0x30/0x40 do_IRQ+0x91/0x190 common_interrupt+0xf/0xf </IRQ> RIP: 0010:default_idle+0x31/0x230 Fixes: 8777e431 ("scsi: qla2xxx: Migrate NVME N2N handling into state machine") # v4.19. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
Bart Van Assche <bvanassche@acm.org> Tested-by:
-
Bart Van Assche authored
Make the code easier to read by converting 'goto' statements into 'return' statements. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
This patch fixes the following Coverity complaint: Null-checking sp->u.iocb_cmd.u.ctarg.rsp suggests that it may be null, but it has already been dereferenced on all paths leading to the check. See also commit e374f9f5 ("scsi: qla2xxx: Migrate switch registration commands away from mailbox interface") # v4.16. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
This patch reduces code duplication. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Instead of calling qla2x00_fcport_event_handler() and letting the switch statement inside that function decide which other function to call, call the latter function directly. Remove the event member from the event_arg structure because it is no longer needed. Remove the qla_handle_els_plogi_done() function because it is never called. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
It is easy to mix up the QLA_* and the MBS_* status codes. Complain loudly if that happens. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
The C language supports implicit casting of void pointers to non-void pointers. Remove explicit sts_entry_* casts that are not necessary. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Split srb_cmd.ctx into two pointers such that the compiler can check the type of that pointer. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Not calling sp->done() from the command completion path is a severe bug. Hence complain loudly if that happens. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
The LIO core requires that the target driver callback functions .queue_data_in() and .queue_status() call target_put_sess_cmd() or transport_generic_free_cmd(). These calls may happen synchronously or asynchronously. Make sure that one of these LIO functions is called in case a command has been aborted. This patch avoids that the code for removing a session hangs due to commands that do not make progress. Cc: Himanshu Madhani <hmadhani@marvell.com> Fixes: 694833ee ("scsi: tcm_qla2xxx: Do not allow aborted cmd to advance.") # v4.13. Fixes: a07100e0 ("qla2xxx: Fix TMR ABORT interaction issue between qla2xxx and TCM") # v4.5. Signed-off-by:
-
Bart Van Assche authored
This patch reduces code duplication. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Since struct sg_table is used in nvme-fc-driver.h, include <linux/scatterlist.h> from that header file. Since no definitions or declarations from <linux/blk-mq.h> are used in the qla_nvme.h header file, do not include <linux/blk-mq.h> from that header file. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Avoid that this function crashes if mcmd == NULL. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Instead of allocating a struct srb dynamically from inside .queuecommand(), set qla2xxx_driver_template.cmd_size such that struct scsi_cmnd and struct srb are contiguous. Do not call QLA_QPAIR_MARK_BUSY() / QLA_QPAIR_MARK_NOT_BUSY() for SRBs associated with SCSI commands. That is safe because scsi_remove_host() is called before queue pairs are deleted and scsi_remove_host() waits for all outstanding SCSI commands to finish. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
This patch does not change any functionality but makes the next patch easier to read. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Since all pointers passed to the srb_t.done() and srb_t.free() functions have type srb_t, change the type of the first argument of these functions from void * into struct srb *. This allows the compiler to verify the argument types for these functions. This patch does not change any functionality. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Use the same approach for encoding the destination ID as the approach used by qlt_update_vp_map(). Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
According to the firmware documentation responder mode must be set for ELS pass-through IOCBs if a response is expected. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
According to the firmware documentation the firmware expects all ELS pass-through IOCB parameters in little endian format. Make this explicit. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
This patch fixes several Coverity complaints about reading data that has not been initialized. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
This change makes it easier to detect qla24xx_read_flash_data() failures and also to handle such failures. This change does not modify the behavior of the driver since all callers ignore the qla24xx_read_flash_data() return value. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Introduce the be_id_t and le_id_t data types for Fibre Channel source and destination ID formats supported by the firmware instead of using an uint8_t[3] array. Introduce functions for converting from and to the port_id_t data types. This patch does not change the behavior of the qla2xxx driver but improves source code readability and also allows the compiler to verify the endianness of Fibre Channel IDs. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Failure of a soft reset is a severe failure. Hence report such failures. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
This patch makes the string manipulation code easier to verify. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Pass the output buffer size to the code that generates a PCI info string and check the output buffer size while generating a PCI info string. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Such a timeout is a severe issue. Hence complain if waiting for pending commands times out. This patch fixes a small bug: it modifies qla82xx_chip_reset_cleanup() such that the "Done waiting" message is reported if qla82xx_chip_reset_cleanup() succeeded instead of if that function failed. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
Make it clear to humans and also to the compiler that the string passed as fourth argument is not modified. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
This patch fixes several Coverity complaints about not always checking the qla2x00_wait_for_hba_online() return value. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
This patch does not change any functionality. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
This patch fixes the following Coverity complaint: Unintentional integer overflow (OVERFLOW_BEFORE_WIDEN) overflow_before_widen: Potentially overflowing expression dma_eng_num * 65536U with type unsigned int (32 bits, unsigned) is evaluated using 32-bit arithmetic, and then used in a context that expects an expression of type uint64_t (64 bits, unsigned). Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
Bart Van Assche authored
This was detected by Coverity. Cc: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by:
-
