GIT 631c228cd09bd5b93090fa60bd9803ec14aa0586 git+ssh://master.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-rc-fixes-2.6.git commit 631c228cd09bd5b93090fa60bd9803ec14aa0586 Author: Christoph Hellwig Date: Sat Jul 8 20:42:15 2006 +0200 [SCSI] hide EH backup data outside the scsi_cmnd Currently struct scsi_cmnd has various fields that are used to backup original data after the corresponding fields have been overridden for EH commands. This means drivers can easily get at it and misuse it. Due to the old_ naming this doesn't happen for most of them, but two that have different names have been used wrong a lot (see previous patch). Another downside is that they unessecarily bloat the scsi_cmnd size. This patch moves them onstack in scsi_send_eh_cmnd to fix those two issues aswell as allowing future EH fixes like moving the EH command submissions to use SG lists like everything else. Signed-off-by: Christoph Hellwig Signed-off-by: James Bottomley commit ae0fda0cdf009731a33f031f86df8566f1977ffe Author: Dave C Boutcher Date: Thu Jul 6 22:08:49 2006 -0500 [SCSI] ibmvscsi: handle inactive SCSI target during probe Without this patch we register an interrupt with request_irq, but then return a bad return code from the module probe. Signed-off-by: James Bottomley commit 6c51fe10475eda9f5c535f337411e1ed0e6915b9 Author: Dave C Boutcher Date: Thu Jul 6 22:07:47 2006 -0500 [SCSI] ibmvscsi: allocate lpevents for ibmvscsi on iseries Allocate the correct number of lp events when running ibmvscsi on legacy iseries Signed-off-by: Dave Boutcher Signed-off-by: James Bottomley commit d14164316d48a4319d8da9dc24be8663b2472bae Author: Matthew Wilcox Date: Sat Jul 8 14:02:26 2006 -0600 [SCSI] aic7[9x]xx: Remove last vestiges of reverse_scan Remove last vestiges of the reverse_scan paramater from aic7xxx and aic79xx. Signed-off-by: Matthew Wilcox Signed-off-by: James Bottomley commit 5e13cdfa5ba94724d6ab11de26b3ec3c94b88b00 Author: Christoph Hellwig Date: Sat Jul 8 20:39:30 2006 +0200 [SCSI] aha152x: stop poking at saved scsi_cmnd members Stop poking into the old_ & co scsi_cmnd fields that should only be used in the EH code. Untested, but this is required to move ahead with the EH fixes. Signed-off-by: Christoph Hellwig Signed-off-by: James Bottomley commit 4e73ea7b02a8d2759b20ab52f13f40d5f5421044 Author: Luben Tuikov Date: Fri Jul 7 00:02:18 2006 -0700 [SCSI] st.c: Improve sense output Convert this: st0: Error with sense data: <6>st: Current: sense key: Illegal Request Additional sense: Invalid field in cdb To this: st0: Current: sense key: Illegal Request Additional sense: Invalid field in cdb Signed-off-by: Luben Tuikov Acked-by: Kai Makisara Signed-off-by: Andrew Morton Signed-off-by: James Bottomley commit 035bff20bff2b6779ed3ad0a24a51a1a96828bc8 Author: James Smart Date: Thu Jul 6 15:50:58 2006 -0400 [SCSI] lpfc 8.1.7: Change version number to 8.1.7 Change version number to 8.1.7 Signed-off-by: James Smart Signed-off-by: James Bottomley commit 65a29c166fe331574880a375559405ac802b027a Author: James Smart Date: Thu Jul 6 15:50:50 2006 -0400 [SCSI] lpfc 8.1.7: Misc Fixes Misc Fixes: - Fix some sparse warnings - casts of address space - Fix handling of the adapter registration string. Each invocation was byteswapping, so every other adapter init attempt failed. - Correct comments and default value for the lpfc_max_luns parameter Signed-off-by: James Smart Signed-off-by: James Bottomley commit b4c026520ff0a4cb838a941bb0ed8996075e3d8c Author: James Smart Date: Thu Jul 6 15:50:43 2006 -0400 [SCSI] lpfc 8.1.7: Add lpfc_sli_flush_mbox_queue() function Add lpfc_sli_flush_mbox_queue() function and use it in lpfc_offline() call to avoid deadlock on thread block. Signed-off-by: James Smart Signed-off-by: James Bottomley commit ce8b3ce55bb0bd9318954efab0f714b088178978 Author: James Smart Date: Thu Jul 6 15:50:36 2006 -0400 [SCSI] lpfc 8.1.7: Correct the wait in attachment that delays for topology discovery Correct the wait in attachment that delays for topology discovery Signed-off-by: James Smart Signed-off-by: James Bottomley commit 5e0b43385559bfe78c1f80d71ec203effe244d9a Author: James Smart Date: Thu Jul 6 15:50:29 2006 -0400 [SCSI] lpfc 8.1.7: Remove depricated sysfs attribute board_online Remove depricated sysfs attribute board_online, as it's replaced by the new issue_reset attribute Signed-off-by: James Smart Signed-off-by: James Bottomley commit 40496f073ffe75c59414a7236f95ecbc49562596 Author: James Smart Date: Thu Jul 6 15:50:22 2006 -0400 [SCSI] lpfc 8.1.7: Adding new issue_reset sysfs attribute Adding new issue_reset sysfs attribute Signed-off-by: James Smart Signed-off-by: James Bottomley commit 420b630d6e9ff41dc3e2a2a2808b67907951e094 Author: James Smart Date: Thu Jul 6 15:50:16 2006 -0400 [SCSI] lpfc 8.1.7: Fix panic in lpfc_sli_validate_fcp_iocb Fix panic in lpfc_sli_validate_fcp_iocb due to access of scsi_cmnd after returning it to the midlayer Signed-off-by: James Smart Signed-off-by: James Bottomley commit bcf4dbfaf38502d3919e0601c06afac7d5e42685 Author: James Smart Date: Thu Jul 6 15:50:08 2006 -0400 [SCSI] lpfc 8.1.7: Consolidate dma buf cleanup into a separate function Consolidate dma buf cleanup into a separate function Signed-off-by: James Smart Signed-off-by: James Bottomley commit 9279565046f39f3ab338818c9923a23b9f25be5f Author: James Smart Date: Thu Jul 6 15:50:02 2006 -0400 [SCSI] lpfc 8.1.7: Correct bogus nodev_tmo message on NPort that changes its NPort Id Correct bogus nodev_tmo message on NPort that changes its NPort Id Signed-off-by: James Smart Signed-off-by: James Bottomley commit 688a88635f9d0d9251d35198e931eaac8816abef Author: James Smart Date: Thu Jul 6 15:49:56 2006 -0400 [SCSI] lpfc 8.1.7: Fix txcmplq related panics on heavy IO while downloading firmware Fix txcmplq related panics on heavy IO while downloading firmware Signed-off-by: James Smart Signed-off-by: James Bottomley commit 4db621e0f3817e14c79513b42c0479d44d8467de Author: James Smart Date: Thu Jul 6 15:49:49 2006 -0400 [SCSI] lpfc 8.1.7: Issue DOWN_LINK prior to INIT_LINK to work around link failure issue Issue DOWN_LINK prior to INIT_LINK to work around link failure issue Signed-off-by: James Smart Signed-off-by: James Bottomley commit d0e56dad5c10e2a34a830e02dfd3aca20e63e387 Author: James Smart Date: Thu Jul 6 15:49:42 2006 -0400 [SCSI] lpfc 8.1.7: Fixed infinite retry of REG_LOGIN mailbox failed due to MBXERR_RPI_FULL Fixed infinite retry of REG_LOGIN mailbox failed due to MBXERR_RPI_FULL Signed-off-by: James Smart Signed-off-by: James Bottomley commit 9f49d3b05f092b89e774be20c654ceb0c64a7d19 Author: James Smart Date: Thu Jul 6 15:49:34 2006 -0400 [SCSI] lpfc 8.1.7: Fix memory leak and cleanup code related to per ring lookup array Fix memory leak and cleanup code related to per ring lookup array. Signed-off-by: James Smart Signed-off-by: James Bottomley commit e17da18e2f78845899467d6e580146ef673f5330 Author: James Smart Date: Thu Jul 6 15:49:25 2006 -0400 [SCSI] lpfc 8.1.7: Standardize the driver on a single define for the maximum supported targets Standardize the driver on a single define for the maximum supported targets. Signed-off-by: James Smart Signed-off-by: James Bottomley commit 5a0e326dfa6d87a1ac6929a7d6d327e8803b7aef Author: James Smart Date: Thu Jul 6 15:49:16 2006 -0400 [SCSI] lpfc 8.1.7: Use mod_timer instead of add_timer in lpfc_els_timeout_handler Use mod_timer instead of add_timer in lpfc_els_timeout_handler This patch was formerly posted by Mark Haverkamp. http://marc.theaimsgroup.com/?l=linux-scsi&m=114246089015681&w=2 Signed-off-by: Mark Haverkamp Signed-off-by: James Smart Signed-off-by: James Bottomley commit dc22f16d4e46cee699c3835cc6e2a8005023517f Author: Eric Moore Date: Thu Jul 6 11:23:14 2006 -0600 [SCSI] mptsas: use unnumbered port API and remove driver porttracking This allows us to be rid of the machinery in mptsas for creating and tracking port numbers. Since mptsas is merely inventing the numbers, the SAS transport class may as well do it instead. Signed-off-by: Eric Moore Signed-off-by: James Bottomley commit a507c6a2d91c52615bd3bb5a8ff439d9e8530939 Author: Douglas Gilbert Date: Sun Jul 2 13:48:10 2006 -0400 [SCSI] update additional sense codes and some opcode names Following on from my post titled: "additional sense codes need update" see the attachment against lk 2.6.17 . ChangeLog: - update additional sense codes table to agree with SPC-4 revision 5a (14 June 2006) - adjust some of the opcode names Signed-off-by: Douglas Gilbert Signed-off-by: James Bottomley commit c9fefeb26457b87f4a767faefcf77321bb90db52 Author: James Bottomley Date: Sun Jul 2 11:10:18 2006 -0500 [SCSI] scsi_transport_sas: add unindexed ports Some SAS HBAs don't want to go to the trouble of tracking port numbers, so they'd simply like to say "add this port and give it a number". This is especially beneficial from the hotplug point of view, since tracking ports and the available number space can be a real pain. The current implementation uses an incrementing number per expander to add the port on. However, since there can never be more ports than there are phys, a later implementation will try to be more intelligent about this. Signed-off-by: James Bottomley commit 24f6d2fd314f8580fcfd96391ce9689727d55572 Author: Adrian Bunk Date: Fri Jun 30 02:33:07 2006 -0700 [SCSI] aic79xx: make ahd_done_with_status() static This patch makes a needlessly global function static. Signed-off-by: Adrian Bunk Acked-by: Hannes Reinecke Signed-off-by: Andrew Morton Signed-off-by: James Bottomley --- diff --git a/drivers/message/fusion/mptbase.h b/drivers/message/fusion/mptbase.h index a5ce10b..0b555e2 100644 --- a/drivers/message/fusion/mptbase.h +++ b/drivers/message/fusion/mptbase.h @@ -644,7 +644,6 @@ #endif struct work_struct fc_rescan_work; char fc_rescan_work_q_name[KOBJ_NAME_LEN]; struct workqueue_struct *fc_rescan_work_q; - u8 port_serial_number; } MPT_ADAPTER; /* diff --git a/drivers/message/fusion/mptsas.c b/drivers/message/fusion/mptsas.c index f7bd8b1..0fb7d58 100644 --- a/drivers/message/fusion/mptsas.c +++ b/drivers/message/fusion/mptsas.c @@ -144,7 +144,6 @@ struct mptsas_devinfo { * Specific details on ports, wide/narrow */ struct mptsas_portinfo_details{ - u8 port_id; /* port number provided to transport */ u16 num_phys; /* number of phys belong to this port */ u64 phy_bitmask; /* TODO, extend support for 255 phys */ struct sas_rphy *rphy; /* transport layer rphy object */ @@ -350,10 +349,10 @@ mptsas_port_delete(struct mptsas_portinf port_info = port_details->port_info; phy_info = port_info->phy_info; - dsaswideprintk((KERN_DEBUG "%s: [%p]: port=%02d num_phys=%02d " + dsaswideprintk((KERN_DEBUG "%s: [%p]: num_phys=%02d " "bitmask=0x%016llX\n", - __FUNCTION__, port_details, port_details->port_id, - port_details->num_phys, port_details->phy_bitmask)); + __FUNCTION__, port_details, port_details->num_phys, + port_details->phy_bitmask)); for (i = 0; i < port_info->num_phys; i++, phy_info++) { if(phy_info->port_details != port_details) @@ -462,9 +461,8 @@ mptsas_setup_wide_ports(MPT_ADAPTER *ioc * phy be removed by firmware events. */ dsaswideprintk((KERN_DEBUG - "%s: [%p]: port=%d deleting phy = %d\n", - __FUNCTION__, port_details, - port_details->port_id, i)); + "%s: [%p]: deleting phy = %d\n", + __FUNCTION__, port_details, i)); port_details->num_phys--; port_details->phy_bitmask &= ~ (1 << phy_info->phy_id); memset(&phy_info->attached, 0, sizeof(struct mptsas_devinfo)); @@ -493,7 +491,6 @@ mptsas_setup_wide_ports(MPT_ADAPTER *ioc goto out; port_details->num_phys = 1; port_details->port_info = port_info; - port_details->port_id = ioc->port_serial_number++; if (phy_info->phy_id < 64 ) port_details->phy_bitmask |= (1 << phy_info->phy_id); @@ -525,12 +522,8 @@ mptsas_setup_wide_ports(MPT_ADAPTER *ioc mptsas_get_port(phy_info_cmp); port_details->starget = mptsas_get_starget(phy_info_cmp); - port_details->port_id = - phy_info_cmp->port_details->port_id; port_details->num_phys = phy_info_cmp->port_details->num_phys; -// port_info->port_serial_number--; - ioc->port_serial_number--; if (!phy_info_cmp->port_details->num_phys) kfree(phy_info_cmp->port_details); } else @@ -554,11 +547,11 @@ #ifdef MPT_DEBUG_SAS_WIDE if (!port_details) continue; dsaswideprintk((KERN_DEBUG - "%s: [%p]: phy_id=%02d port_id=%02d num_phys=%02d " + "%s: [%p]: phy_id=%02d num_phys=%02d " "bitmask=0x%016llX\n", __FUNCTION__, - port_details, i, port_details->port_id, - port_details->num_phys, port_details->phy_bitmask)); + port_details, i, port_details->num_phys, + port_details->phy_bitmask)); dsaswideprintk((KERN_DEBUG"\t\tport = %p rphy=%p\n", port_details->port, port_details->rphy)); } @@ -1608,11 +1601,7 @@ static int mptsas_probe_one_phy(struct d if (phy_info->sas_port_add_phy) { if (!port) { - port = sas_port_alloc(dev, - phy_info->port_details->port_id); - dsaswideprintk((KERN_DEBUG - "sas_port_alloc: port=%p dev=%p port_id=%d\n", - port, dev, phy_info->port_details->port_id)); + port = sas_port_alloc_num(dev); if (!port) { error = -ENOMEM; goto out; @@ -1625,6 +1614,9 @@ static int mptsas_probe_one_phy(struct d goto out; } mptsas_set_port(phy_info, port); + dsaswideprintk((KERN_DEBUG + "sas_port_alloc: port=%p dev=%p port_id=%d\n", + port, dev, port->port_identifier)); } dsaswideprintk((KERN_DEBUG "sas_port_add_phy: phy_id=%d\n", phy_info->phy_id)); @@ -1939,7 +1931,8 @@ mptsas_delete_expander_phys(MPT_ADAPTER expander_sas_address) continue; #ifdef MPT_DEBUG_SAS_WIDE - dev_printk(KERN_DEBUG, &port->dev, "delete\n"); + dev_printk(KERN_DEBUG, &port->dev, + "delete port (%d)\n", port->port_identifier); #endif sas_port_delete(port); mptsas_port_delete(phy_info->port_details); @@ -2185,7 +2178,8 @@ mptsas_hotplug_work(void *arg) ioc->name, ds, ev->channel, ev->id, phy_info->phy_id); #ifdef MPT_DEBUG_SAS_WIDE - dev_printk(KERN_DEBUG, &port->dev, "delete\n"); + dev_printk(KERN_DEBUG, &port->dev, + "delete port (%d)\n", port->port_identifier); #endif sas_port_delete(port); mptsas_port_delete(phy_info->port_details); diff --git a/drivers/scsi/aha152x.c b/drivers/scsi/aha152x.c index 36e63f8..f974869 100644 --- a/drivers/scsi/aha152x.c +++ b/drivers/scsi/aha152x.c @@ -551,6 +551,11 @@ #endif struct aha152x_scdata { Scsi_Cmnd *next; /* next sc in queue */ struct semaphore *sem; /* semaphore to block on */ + unsigned char cmd_len; + unsigned char cmnd[MAX_COMMAND_SIZE]; + unsigned short use_sg; + unsigned request_bufflen; + void *request_buffer; }; @@ -1006,11 +1011,20 @@ #endif return FAILED; } } else { + struct aha152x_scdata *sc; + SCpnt->host_scribble = kmalloc(sizeof(struct aha152x_scdata), GFP_ATOMIC); if(SCpnt->host_scribble==0) { printk(ERR_LEAD "allocation failed\n", CMDINFO(SCpnt)); return FAILED; } + + sc = SCDATA(SCpnt); + memcpy(sc->cmnd, SCpnt->cmnd, sizeof(sc->cmnd)); + sc->request_buffer = SCpnt->request_buffer; + sc->request_bufflen = SCpnt->request_bufflen; + sc->use_sg = SCpnt->use_sg; + sc->cmd_len = SCpnt->cmd_len; } SCNEXT(SCpnt) = NULL; @@ -1165,6 +1179,10 @@ static int aha152x_device_reset(Scsi_Cmn DECLARE_MUTEX_LOCKED(sem); struct timer_list timer; int ret, issued, disconnected; + unsigned char old_cmd_len = SCpnt->cmd_len; + unsigned short old_use_sg = SCpnt->use_sg; + void *old_buffer = SCpnt->request_buffer; + unsigned old_bufflen = SCpnt->request_bufflen; unsigned long flags; #if defined(AHA152X_DEBUG) @@ -1198,11 +1216,11 @@ #endif add_timer(&timer); down(&sem); del_timer(&timer); - - SCpnt->cmd_len = SCpnt->old_cmd_len; - SCpnt->use_sg = SCpnt->old_use_sg; - SCpnt->request_buffer = SCpnt->buffer; - SCpnt->request_bufflen = SCpnt->bufflen; + + SCpnt->cmd_len = old_cmd_len; + SCpnt->use_sg = old_use_sg; + SCpnt->request_buffer = old_buffer; + SCpnt->request_bufflen = old_bufflen; DO_LOCK(flags); @@ -1565,6 +1583,9 @@ #if defined(AHA152X_STAT) #endif if(DONE_SC->SCp.phase & check_condition) { + struct scsi_cmnd *cmd = HOSTDATA(shpnt)->done_SC; + struct aha152x_scdata *sc = SCDATA(cmd); + #if 0 if(HOSTDATA(shpnt)->debug & debug_eh) { printk(ERR_LEAD "received sense: ", CMDINFO(DONE_SC)); @@ -1573,13 +1594,13 @@ #if 0 #endif /* restore old command */ - memcpy((void *) DONE_SC->cmnd, (void *) DONE_SC->data_cmnd, sizeof(DONE_SC->data_cmnd)); - DONE_SC->request_buffer = DONE_SC->buffer; - DONE_SC->request_bufflen = DONE_SC->bufflen; - DONE_SC->use_sg = DONE_SC->old_use_sg; - DONE_SC->cmd_len = DONE_SC->old_cmd_len; + memcpy(cmd->cmnd, sc->cmnd, sizeof(sc->cmnd)); + cmd->request_buffer = sc->request_buffer; + cmd->request_bufflen = sc->request_bufflen; + cmd->use_sg = sc->use_sg; + cmd->cmd_len = sc->cmd_len; - DONE_SC->SCp.Status = 0x02; + cmd->SCp.Status = 0x02; HOSTDATA(shpnt)->commands--; if (!HOSTDATA(shpnt)->commands) diff --git a/drivers/scsi/aic7xxx/aic79xx_core.c b/drivers/scsi/aic7xxx/aic79xx_core.c index a1e8ca7..653818d 100644 --- a/drivers/scsi/aic7xxx/aic79xx_core.c +++ b/drivers/scsi/aic7xxx/aic79xx_core.c @@ -7289,7 +7289,7 @@ ahd_reset_cmds_pending(struct ahd_softc ahd->flags &= ~AHD_UPDATE_PEND_CMDS; } -void +static void ahd_done_with_status(struct ahd_softc *ahd, struct scb *scb, uint32_t status) { cam_status ostat; diff --git a/drivers/scsi/aic7xxx/aic79xx_osm.c b/drivers/scsi/aic7xxx/aic79xx_osm.c index b244c71..998999c 100644 --- a/drivers/scsi/aic7xxx/aic79xx_osm.c +++ b/drivers/scsi/aic7xxx/aic79xx_osm.c @@ -243,25 +243,6 @@ ahd_print_path(struct ahd_softc *ahd, st static uint32_t aic79xx_no_reset; /* - * Certain PCI motherboards will scan PCI devices from highest to lowest, - * others scan from lowest to highest, and they tend to do all kinds of - * strange things when they come into contact with PCI bridge chips. The - * net result of all this is that the PCI card that is actually used to boot - * the machine is very hard to detect. Most motherboards go from lowest - * PCI slot number to highest, and the first SCSI controller found is the - * one you boot from. The only exceptions to this are when a controller - * has its BIOS disabled. So, we by default sort all of our SCSI controllers - * from lowest PCI slot number to highest PCI slot number. We also force - * all controllers with their BIOS disabled to the end of the list. This - * works on *almost* all computers. Where it doesn't work, we have this - * option. Setting this option to non-0 will reverse the order of the sort - * to highest first, then lowest, but will still leave cards with their BIOS - * disabled at the very end. That should fix everyone up unless there are - * really strange cirumstances. - */ -static uint32_t aic79xx_reverse_scan; - -/* * Should we force EXTENDED translation on a controller. * 0 == Use whatever is in the SEEPROM or default to off * 1 == Use whatever is in the SEEPROM or default to on @@ -350,7 +331,6 @@ MODULE_PARM_DESC(aic79xx, " periodically to prevent tag starvation.\n" " This may be required by some older disk\n" " or drives/RAID arrays.\n" -" reverse_scan Sort PCI devices highest Bus/Slot to lowest\n" " tag_info: Set per-target tag depth\n" " global_tag_depth: Global tag depth for all targets on all buses\n" " slewrate:Set the signal slew rate (0-15).\n" @@ -1031,7 +1011,6 @@ aic79xx_setup(char *s) #ifdef AHD_DEBUG { "debug", &ahd_debug }, #endif - { "reverse_scan", &aic79xx_reverse_scan }, { "periodic_otag", &aic79xx_periodic_otag }, { "pci_parity", &aic79xx_pci_parity }, { "seltime", &aic79xx_seltime }, diff --git a/drivers/scsi/aic7xxx/aic7xxx_osm.c b/drivers/scsi/aic7xxx/aic7xxx_osm.c index debf3e2..aa4be8a 100644 --- a/drivers/scsi/aic7xxx/aic7xxx_osm.c +++ b/drivers/scsi/aic7xxx/aic7xxx_osm.c @@ -353,7 +353,6 @@ MODULE_PARM_DESC(aic7xxx, " periodically to prevent tag starvation.\n" " This may be required by some older disk\n" " drives or RAID arrays.\n" -" reverse_scan Sort PCI devices highest Bus/Slot to lowest\n" " tag_info: Set per-target tag depth\n" " global_tag_depth: Global tag depth for every target\n" " on every bus\n" diff --git a/drivers/scsi/constants.c b/drivers/scsi/constants.c index dddd2ac..61f6024 100644 --- a/drivers/scsi/constants.c +++ b/drivers/scsi/constants.c @@ -5,6 +5,7 @@ * Additions for SCSI 3+ (SPC-3 T10/1416-D Rev 07 3 May 2002) * by D. Gilbert and aeb (20020609) * Additions for SPC-3 T10/1416-D Rev 21 22 Sept 2004, D. Gilbert 20041025 + * Update to SPC-4 T10/1713-D Rev 5a, 14 June 2006, D. Gilbert 20060702 */ #include @@ -36,55 +37,56 @@ static const char * cdb_byte0_names[] = /* 00-03 */ "Test Unit Ready", "Rezero Unit/Rewind", NULL, "Request Sense", /* 04-07 */ "Format Unit/Medium", "Read Block Limits", NULL, "Reasssign Blocks", -/* 08-0d */ "Read (6)", NULL, "Write (6)", "Seek (6)", NULL, NULL, +/* 08-0d */ "Read(6)", NULL, "Write(6)", "Seek(6)", NULL, NULL, /* 0e-12 */ NULL, "Read Reverse", "Write Filemarks", "Space", "Inquiry", -/* 13-16 */ "Verify (6)", "Recover Buffered Data", "Mode Select (6)", - "Reserve (6)", -/* 17-1a */ "Release (6)", "Copy", "Erase", "Mode Sense (6)", +/* 13-16 */ "Verify(6)", "Recover Buffered Data", "Mode Select(6)", + "Reserve(6)", +/* 17-1a */ "Release(6)", "Copy", "Erase", "Mode Sense(6)", /* 1b-1d */ "Start/Stop Unit", "Receive Diagnostic", "Send Diagnostic", /* 1e-1f */ "Prevent/Allow Medium Removal", NULL, /* 20-22 */ NULL, NULL, NULL, /* 23-28 */ "Read Format Capacities", "Set Window", - "Read Capacity (10)", NULL, NULL, "Read (10)", -/* 29-2d */ "Read Generation", "Write (10)", "Seek (10)", "Erase (10)", - "Read updated block", -/* 2e-31 */ "Write Verify (10)", "Verify (10)", "Search High", "Search Equal", + "Read Capacity(10)", NULL, NULL, "Read(10)", +/* 29-2d */ "Read Generation", "Write(10)", "Seek(10)", "Erase(10)", + "Read updated block", +/* 2e-31 */ "Write Verify(10)", "Verify(10)", "Search High", "Search Equal", /* 32-34 */ "Search Low", "Set Limits", "Prefetch/Read Position", -/* 35-37 */ "Synchronize Cache (10)", "Lock/Unlock Cache (10)", +/* 35-37 */ "Synchronize Cache(10)", "Lock/Unlock Cache(10)", "Read Defect Data(10)", /* 38-3c */ "Medium Scan", "Compare", "Copy Verify", "Write Buffer", "Read Buffer", -/* 3d-3f */ "Update Block", "Read Long (10)", "Write Long (10)", -/* 40-41 */ "Change Definition", "Write Same (10)", +/* 3d-3f */ "Update Block", "Read Long(10)", "Write Long(10)", +/* 40-41 */ "Change Definition", "Write Same(10)", /* 42-48 */ "Read sub-channel", "Read TOC/PMA/ATIP", "Read density support", - "Play audio (10)", "Get configuration", "Play audio msf", + "Play audio(10)", "Get configuration", "Play audio msf", "Play audio track/index", -/* 49-4f */ "Play track relative (10)", "Get event status notification", +/* 49-4f */ "Play track relative(10)", "Get event status notification", "Pause/resume", "Log Select", "Log Sense", "Stop play/scan", NULL, /* 50-55 */ "Xdwrite", "Xpwrite, Read disk info", "Xdread, Read track info", - "Reserve track", "Send OPC info", "Mode Select (10)", -/* 56-5b */ "Reserve (10)", "Release (10)", "Repair track", "Read master cue", - "Mode Sense (10)", "Close track/session", + "Reserve track", "Send OPC info", "Mode Select(10)", +/* 56-5b */ "Reserve(10)", "Release(10)", "Repair track", "Read master cue", + "Mode Sense(10)", "Close track/session", /* 5c-5f */ "Read buffer capacity", "Send cue sheet", "Persistent reserve in", "Persistent reserve out", /* 60-67 */ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, /* 68-6f */ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, /* 70-77 */ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, /* 78-7f */ NULL, NULL, NULL, NULL, NULL, NULL, NULL, "Variable length", -/* 80-84 */ "Xdwrite (16)", "Rebuild (16)", "Regenerate (16)", "Extended copy", +/* 80-84 */ "Xdwrite(16)", "Rebuild(16)", "Regenerate(16)", "Extended copy", "Receive copy results", -/* 85-89 */ "Memory Export In (16)", "Access control in", "Access control out", - "Read (16)", "Memory Export Out (16)", -/* 8a-8f */ "Write (16)", NULL, "Read attributes", "Write attributes", - "Write and verify (16)", "Verify (16)", -/* 90-94 */ "Pre-fetch (16)", "Synchronize cache (16)", - "Lock/unlock cache (16)", "Write same (16)", NULL, +/* 85-89 */ "ATA command pass through(16)", "Access control in", + "Access control out", "Read(16)", "Memory Export Out(16)", +/* 8a-8f */ "Write(16)", NULL, "Read attributes", "Write attributes", + "Write and verify(16)", "Verify(16)", +/* 90-94 */ "Pre-fetch(16)", "Synchronize cache(16)", + "Lock/unlock cache(16)", "Write same(16)", NULL, /* 95-99 */ NULL, NULL, NULL, NULL, NULL, -/* 9a-9f */ NULL, NULL, NULL, NULL, "Service action in (16)", - "Service action out (16)", -/* a0-a5 */ "Report luns", "Blank", "Send event", "Maintenance in", - "Maintenance out", "Move medium/play audio(12)", +/* 9a-9f */ NULL, NULL, NULL, NULL, "Service action in(16)", + "Service action out(16)", +/* a0-a5 */ "Report luns", "ATA command pass through(12)/Blank", + "Security protocol in", "Maintenance in", "Maintenance out", + "Move medium/play audio(12)", /* a6-a9 */ "Exchange medium", "Move medium attached", "Read(12)", "Play track relative(12)", /* aa-ae */ "Write(12)", NULL, "Erase(12), Get Performance", @@ -92,12 +94,12 @@ static const char * cdb_byte0_names[] = /* af-b1 */ "Verify(12)", "Search data high(12)", "Search data equal(12)", /* b2-b4 */ "Search data low(12)", "Set limits(12)", "Read element status attached", -/* b5-b6 */ "Request volume element address", "Send volume tag, set streaming", +/* b5-b6 */ "Security protocol out", "Send volume tag, set streaming", /* b7-b9 */ "Read defect data(12)", "Read element status", "Read CD msf", /* ba-bc */ "Redundancy group (in), Scan", - "Redundancy group (out), Set cd-rom speed", "Spare in, Play cd", -/* bd-bf */ "Spare out, Mechanism status", "Volume set in, Read cd", - "Volume set out, Send DVD structure", + "Redundancy group (out), Set cd-rom speed", "Spare (in), Play cd", +/* bd-bf */ "Spare (out), Mechanism status", "Volume set (in), Read cd", + "Volume set (out), Send DVD structure", }; struct value_name_pair { @@ -112,6 +114,7 @@ static const struct value_name_pair main {0xc, "Report supported operation codes"}, {0xd, "Report supported task management functions"}, {0xe, "Report priority"}, + {0xf, "Report timestamp"}, }; #define MAINT_IN_SZ ARRAY_SIZE(maint_in_arr) @@ -120,6 +123,7 @@ static const struct value_name_pair main {0xa, "Set target port groups"}, {0xb, "Change aliases"}, {0xe, "Set priority"}, + {0xe, "Set timestamp"}, }; #define MAINT_OUT_SZ ARRAY_SIZE(maint_out_arr) @@ -427,6 +431,7 @@ static struct error_info additional[] = {0x001A, "Rewind operation in progress"}, {0x001B, "Set capacity operation in progress"}, {0x001C, "Verify operation in progress"}, + {0x001D, "ATA pass through information available"}, {0x0100, "No index/sector signal"}, @@ -438,7 +443,7 @@ static struct error_info additional[] = {0x0400, "Logical unit not ready, cause not reportable"}, {0x0401, "Logical unit is in process of becoming ready"}, - {0x0402, "Logical unit not ready, initializing cmd. required"}, + {0x0402, "Logical unit not ready, initializing command required"}, {0x0403, "Logical unit not ready, manual intervention required"}, {0x0404, "Logical unit not ready, format in progress"}, {0x0405, "Logical unit not ready, rebuild in progress"}, @@ -478,6 +483,9 @@ static struct error_info additional[] = {0x0B00, "Warning"}, {0x0B01, "Warning - specified temperature exceeded"}, {0x0B02, "Warning - enclosure degraded"}, + {0x0B03, "Warning - background self-test failed"}, + {0x0B04, "Warning - background pre-scan detected medium error"}, + {0x0B05, "Warning - background medium scan detected medium error"}, {0x0C00, "Write error"}, {0x0C01, "Write error - recovered with auto reallocation"}, @@ -493,6 +501,7 @@ static struct error_info additional[] = {0x0C0B, "Auxiliary memory write error"}, {0x0C0C, "Write error - unexpected unsolicited data"}, {0x0C0D, "Write error - not enough unsolicited data"}, + {0x0C0F, "Defects in error window"}, {0x0D00, "Error detected by third party temporary initiator"}, {0x0D01, "Third party device failure"}, @@ -504,11 +513,12 @@ static struct error_info additional[] = {0x0E00, "Invalid information unit"}, {0x0E01, "Information unit too short"}, {0x0E02, "Information unit too long"}, + {0x0E03, "Invalid field in command information unit"}, {0x1000, "Id CRC or ECC error"}, - {0x1001, "Data block guard check failed"}, - {0x1002, "Data block application tag check failed"}, - {0x1003, "Data block reference tag check failed"}, + {0x1001, "Logical block guard check failed"}, + {0x1002, "Logical block application tag check failed"}, + {0x1003, "Logical block reference tag check failed"}, {0x1100, "Unrecovered read error"}, {0x1101, "Read retries exhausted"}, @@ -530,6 +540,7 @@ static struct error_info additional[] = {0x1111, "Read error - loss of streaming"}, {0x1112, "Auxiliary memory read error"}, {0x1113, "Read error - failed retransmission request"}, + {0x1114, "Read error - lba marked bad by application client"}, {0x1200, "Address mark not found for id field"}, @@ -610,11 +621,14 @@ static struct error_info additional[] = {0x2100, "Logical block address out of range"}, {0x2101, "Invalid element address"}, {0x2102, "Invalid address for write"}, + {0x2103, "Invalid write crossing layer jump"}, {0x2200, "Illegal function (use 20 00, 24 00, or 26 00)"}, {0x2400, "Invalid field in cdb"}, {0x2401, "CDB decryption error"}, + {0x2402, "Obsolete"}, + {0x2403, "Obsolete"}, {0x2404, "Security audit value frozen"}, {0x2405, "Security working key frozen"}, {0x2406, "Nonce not unique"}, @@ -637,7 +651,10 @@ static struct error_info additional[] = {0x260C, "Invalid operation for copy source or destination"}, {0x260D, "Copy segment granularity violation"}, {0x260E, "Invalid parameter while port is enabled"}, - {0x260F, "Invalid data-out buffer integrity"}, + {0x260F, "Invalid data-out buffer integrity check value"}, + {0x2610, "Data decryption key fail limit reached"}, + {0x2611, "Incomplete key-associated data set"}, + {0x2612, "Vendor specific key reference not found"}, {0x2700, "Write protected"}, {0x2701, "Hardware write protected"}, @@ -649,6 +666,7 @@ static struct error_info additional[] = {0x2800, "Not ready to ready change, medium may have changed"}, {0x2801, "Import or export element accessed"}, + {0x2802, "Format-layer may have changed"}, {0x2900, "Power on, reset, or bus device reset occurred"}, {0x2901, "Power on occurred"}, @@ -669,6 +687,11 @@ static struct error_info additional[] = {0x2A07, "Implicit asymmetric access state transition failed"}, {0x2A08, "Priority changed"}, {0x2A09, "Capacity data has changed"}, + {0x2A10, "Timestamp changed"}, + {0x2A11, "Data encryption parameters changed by another i_t nexus"}, + {0x2A12, "Data encryption parameters changed by vendor specific " + "event"}, + {0x2A13, "Data encryption key instance counter has changed"}, {0x2B00, "Copy cannot execute since host cannot disconnect"}, @@ -690,6 +713,7 @@ static struct error_info additional[] = {0x2E00, "Insufficient time for operation"}, {0x2F00, "Commands cleared by another initiator"}, + {0x2F01, "Commands cleared by power loss notification"}, {0x3000, "Incompatible medium installed"}, {0x3001, "Cannot read medium - unknown format"}, @@ -702,7 +726,8 @@ static struct error_info additional[] = {0x3008, "Cannot write - application code mismatch"}, {0x3009, "Current session not fixated for append"}, {0x300A, "Cleaning request rejected"}, - {0x300C, "WORM medium, overwrite attempted"}, + {0x300C, "WORM medium - overwrite attempted"}, + {0x300D, "WORM medium - integrity check"}, {0x3010, "Medium not formatted"}, {0x3100, "Medium format corrupted"}, @@ -790,6 +815,9 @@ static struct error_info additional[] = {0x3F0F, "Echo buffer overwritten"}, {0x3F10, "Medium loadable"}, {0x3F11, "Medium auxiliary memory accessible"}, + {0x3F12, "iSCSI IP address added"}, + {0x3F13, "iSCSI IP address removed"}, + {0x3F14, "iSCSI IP address changed"}, /* * {0x40NN, "Ram failure"}, * {0x40NN, "Diagnostic failure on component nn"}, @@ -799,6 +827,7 @@ static struct error_info additional[] = {0x4300, "Message error"}, {0x4400, "Internal target failure"}, + {0x4471, "ATA device failed set features"}, {0x4500, "Select or reselect failure"}, @@ -807,9 +836,10 @@ static struct error_info additional[] = {0x4700, "Scsi parity error"}, {0x4701, "Data phase CRC error detected"}, {0x4702, "Scsi parity error detected during st data phase"}, - {0x4703, "Information unit CRC error detected"}, + {0x4703, "Information unit iuCRC error detected"}, {0x4704, "Asynchronous information protection error detected"}, {0x4705, "Protocol service CRC error"}, + {0x4706, "Phy test function in progress"}, {0x477f, "Some commands cleared by iSCSI Protocol event"}, {0x4800, "Initiator detected error message received"}, @@ -844,6 +874,8 @@ static struct error_info additional[] = {0x5300, "Media load or eject failed"}, {0x5301, "Unload tape failure"}, {0x5302, "Medium removal prevented"}, + {0x5303, "Medium removal prevented by data transfer element"}, + {0x5304, "Medium thread or unthread failure"}, {0x5400, "Scsi to host system interface failure"}, @@ -855,6 +887,7 @@ static struct error_info additional[] = {0x5505, "Insufficient access control resources"}, {0x5506, "Auxiliary memory out of space"}, {0x5507, "Quota error"}, + {0x5508, "Maximum number of supplemental decryption keys exceeded"}, {0x5700, "Unable to recover table-of-contents"}, @@ -1004,6 +1037,7 @@ static struct error_info additional[] = {0x6708, "Assign failure occurred"}, {0x6709, "Multiply assigned logical unit"}, {0x670A, "Set target port groups command failed"}, + {0x670B, "ATA device feature not enabled"}, {0x6800, "Logical unit not configured"}, @@ -1030,6 +1064,8 @@ static struct error_info additional[] = {0x6F03, "Read of scrambled sector without authentication"}, {0x6F04, "Media region code is mismatched to logical unit region"}, {0x6F05, "Drive region must be permanent/region reset count error"}, + {0x6F06, "Insufficient block count for binding nonce recording"}, + {0x6F07, "Conflict in binding nonce recording"}, /* * {0x70NN, "Decompression exception short algorithm id of nn"}, */ @@ -1041,6 +1077,8 @@ static struct error_info additional[] = {0x7203, "Session fixation error - incomplete track in session"}, {0x7204, "Empty or partially written reserved track"}, {0x7205, "No more track reservations allowed"}, + {0x7206, "RMZ extension is not allowed"}, + {0x7207, "No more test zone extensions are allowed"}, {0x7300, "Cd control error"}, {0x7301, "Power calibration area almost full"}, @@ -1049,6 +1087,18 @@ static struct error_info additional[] = {0x7304, "Program memory area update failure"}, {0x7305, "Program memory area is full"}, {0x7306, "RMA/PMA is almost full"}, + {0x7310, "Current power calibration area almost full"}, + {0x7311, "Current power calibration area is full"}, + {0x7317, "RDZ is full"}, + + {0x7400, "Security error"}, + {0x7401, "Unable to decrypt data"}, + {0x7402, "Unencrypted data encountered while decrypting"}, + {0x7403, "Incorrect data encryption key"}, + {0x7404, "Cryptographic integrity validation failed"}, + {0x7405, "Error decrypting data"}, + {0x7471, "Logical unit access not authorized"}, + {0, NULL} }; diff --git a/drivers/scsi/ibmvscsi/iseries_vscsi.c b/drivers/scsi/ibmvscsi/iseries_vscsi.c index 7eed0b0..6aeb5f0 100644 --- a/drivers/scsi/ibmvscsi/iseries_vscsi.c +++ b/drivers/scsi/ibmvscsi/iseries_vscsi.c @@ -81,7 +81,7 @@ int ibmvscsi_init_crq_queue(struct crq_q int rc; single_host_data = hostdata; - rc = viopath_open(viopath_hostLp, viomajorsubtype_scsi, 0); + rc = viopath_open(viopath_hostLp, viomajorsubtype_scsi, max_requests); if (rc < 0) { printk("viopath_open failed with rc %d in open_event_path\n", rc); diff --git a/drivers/scsi/ibmvscsi/rpa_vscsi.c b/drivers/scsi/ibmvscsi/rpa_vscsi.c index 242b887..ed22b96 100644 --- a/drivers/scsi/ibmvscsi/rpa_vscsi.c +++ b/drivers/scsi/ibmvscsi/rpa_vscsi.c @@ -238,6 +238,7 @@ int ibmvscsi_init_crq_queue(struct crq_q if (rc == 2) { /* Adapter is good, but other end is not ready */ printk(KERN_WARNING "ibmvscsi: Partner adapter not ready\n"); + retrc = 0; } else if (rc != 0) { printk(KERN_WARNING "ibmvscsi: Error %d opening adapter\n", rc); goto reg_crq_failed; diff --git a/drivers/scsi/lpfc/lpfc.h b/drivers/scsi/lpfc/lpfc.h index f81691f..d44f9aa 100644 --- a/drivers/scsi/lpfc/lpfc.h +++ b/drivers/scsi/lpfc/lpfc.h @@ -21,10 +21,12 @@ struct lpfc_sli2_slim; -#define LPFC_MAX_TARGET 256 /* max targets supported */ -#define LPFC_MAX_DISC_THREADS 64 /* max outstanding discovery els req */ -#define LPFC_MAX_NS_RETRY 3 /* max NameServer retries */ +#define LPFC_MAX_TARGET 256 /* max number of targets supported */ +#define LPFC_MAX_DISC_THREADS 64 /* max outstanding discovery els + requests */ +#define LPFC_MAX_NS_RETRY 3 /* Number of retry attempts to contact + the NameServer before giving up. */ #define LPFC_DFT_HBA_Q_DEPTH 2048 /* max cmds per hba */ #define LPFC_LC_HBA_Q_DEPTH 1024 /* max cmds per low cost hba */ #define LPFC_LP101_HBA_Q_DEPTH 128 /* max cmds per low cost hba */ @@ -41,7 +43,6 @@ #define getPaddr(high, low) ((dma_addr_ (( (u64)(high)<<16 ) << 16)|( (u64)(low)))) /* Provide maximum configuration definitions. */ #define LPFC_DRVR_TIMEOUT 16 /* driver iocb timeout value in sec */ -#define MAX_FCP_TARGET 256 /* max num of FCP targets supported */ #define FC_MAX_ADPTMSG 64 #define MAX_HBAEVT 32 diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c index b62a72d..5c68cdd 100644 --- a/drivers/scsi/lpfc/lpfc_attr.c +++ b/drivers/scsi/lpfc/lpfc_attr.c @@ -219,9 +219,19 @@ lpfc_issue_lip(struct Scsi_Host *host) return -ENOMEM; memset((void *)pmboxq, 0, sizeof (LPFC_MBOXQ_t)); - lpfc_init_link(phba, pmboxq, phba->cfg_topology, phba->cfg_link_speed); + pmboxq->mb.mbxCommand = MBX_DOWN_LINK; + pmboxq->mb.mbxOwner = OWN_HOST; + mbxstatus = lpfc_sli_issue_mbox_wait(phba, pmboxq, phba->fc_ratov * 2); + if ((mbxstatus == MBX_SUCCESS) && (pmboxq->mb.mbxStatus == 0)) { + memset((void *)pmboxq, 0, sizeof (LPFC_MBOXQ_t)); + lpfc_init_link(phba, pmboxq, phba->cfg_topology, + phba->cfg_link_speed); + mbxstatus = lpfc_sli_issue_mbox_wait(phba, pmboxq, + phba->fc_ratov * 2); + } + if (mbxstatus == MBX_TIMEOUT) pmboxq->mbox_cmpl = lpfc_sli_def_mbox_cmpl; else @@ -233,51 +243,53 @@ lpfc_issue_lip(struct Scsi_Host *host) return 0; } -static ssize_t -lpfc_nport_evt_cnt_show(struct class_device *cdev, char *buf) +static int +lpfc_selective_reset(struct lpfc_hba *phba) { - struct Scsi_Host *host = class_to_shost(cdev); - struct lpfc_hba *phba = (struct lpfc_hba*)host->hostdata; - return snprintf(buf, PAGE_SIZE, "%d\n", phba->nport_event_cnt); + struct completion online_compl; + int status = 0; + + init_completion(&online_compl); + lpfc_workq_post_event(phba, &status, &online_compl, + LPFC_EVT_OFFLINE); + wait_for_completion(&online_compl); + + if (status != 0) + return -EIO; + + init_completion(&online_compl); + lpfc_workq_post_event(phba, &status, &online_compl, + LPFC_EVT_ONLINE); + wait_for_completion(&online_compl); + + if (status != 0) + return -EIO; + + return 0; } static ssize_t -lpfc_board_online_show(struct class_device *cdev, char *buf) +lpfc_issue_reset(struct class_device *cdev, const char *buf, size_t count) { struct Scsi_Host *host = class_to_shost(cdev); struct lpfc_hba *phba = (struct lpfc_hba*)host->hostdata; + int status = -EINVAL; - if (phba->fc_flag & FC_OFFLINE_MODE) - return snprintf(buf, PAGE_SIZE, "0\n"); + if (strncmp(buf, "selective", sizeof("selective") - 1) == 0) + status = lpfc_selective_reset(phba); + + if (status == 0) + return strlen(buf); else - return snprintf(buf, PAGE_SIZE, "1\n"); + return status; } static ssize_t -lpfc_board_online_store(struct class_device *cdev, const char *buf, - size_t count) +lpfc_nport_evt_cnt_show(struct class_device *cdev, char *buf) { struct Scsi_Host *host = class_to_shost(cdev); struct lpfc_hba *phba = (struct lpfc_hba*)host->hostdata; - struct completion online_compl; - int val=0, status=0; - - if (sscanf(buf, "%d", &val) != 1) - return -EINVAL; - - init_completion(&online_compl); - - if (val) - lpfc_workq_post_event(phba, &status, &online_compl, - LPFC_EVT_ONLINE); - else - lpfc_workq_post_event(phba, &status, &online_compl, - LPFC_EVT_OFFLINE); - wait_for_completion(&online_compl); - if (!status) - return strlen(buf); - else - return -EIO; + return snprintf(buf, PAGE_SIZE, "%d\n", phba->nport_event_cnt); } static ssize_t @@ -532,10 +544,9 @@ static CLASS_DEVICE_ATTR(lpfc_drvr_versi NULL); static CLASS_DEVICE_ATTR(management_version, S_IRUGO, management_version_show, NULL); -static CLASS_DEVICE_ATTR(board_online, S_IRUGO | S_IWUSR, - lpfc_board_online_show, lpfc_board_online_store); static CLASS_DEVICE_ATTR(board_mode, S_IRUGO | S_IWUSR, lpfc_board_mode_show, lpfc_board_mode_store); +static CLASS_DEVICE_ATTR(issue_reset, S_IWUSR, NULL, lpfc_issue_reset); static int lpfc_poll = 0; module_param(lpfc_poll, int, 0); @@ -695,12 +706,12 @@ LPFC_ATTR(discovery_threads, 32, 1, 64, "during discovery"); /* -# lpfc_max_luns: maximum number of LUNs per target driver will support -# Value range is [1,32768]. Default value is 256. -# NOTE: The SCSI layer will scan each target for this many luns +# lpfc_max_luns: maximum allowed LUN. +# Value range is [0,65535]. Default value is 255. +# NOTE: The SCSI layer might probe all allowed LUN on some old targets. */ -LPFC_ATTR_R(max_luns, 256, 1, 32768, - "Maximum number of LUNs per target driver will support"); +LPFC_ATTR_R(max_luns, 255, 0, 65535, + "Maximum allowed LUN"); /* # lpfc_poll_tmo: .Milliseconds driver will wait between polling FCP ring. @@ -739,8 +750,8 @@ struct class_device_attribute *lpfc_host &class_device_attr_lpfc_max_luns, &class_device_attr_nport_evt_cnt, &class_device_attr_management_version, - &class_device_attr_board_online, &class_device_attr_board_mode, + &class_device_attr_issue_reset, &class_device_attr_lpfc_poll, &class_device_attr_lpfc_poll_tmo, NULL, diff --git a/drivers/scsi/lpfc/lpfc_crtn.h b/drivers/scsi/lpfc/lpfc_crtn.h index ee22173..517e9e4 100644 --- a/drivers/scsi/lpfc/lpfc_crtn.h +++ b/drivers/scsi/lpfc/lpfc_crtn.h @@ -147,6 +147,7 @@ int lpfc_sli_hba_setup(struct lpfc_hba * int lpfc_sli_hba_down(struct lpfc_hba *); int lpfc_sli_issue_mbox(struct lpfc_hba *, LPFC_MBOXQ_t *, uint32_t); int lpfc_sli_handle_mb_event(struct lpfc_hba *); +int lpfc_sli_flush_mbox_queue(struct lpfc_hba *); int lpfc_sli_handle_slow_ring_event(struct lpfc_hba *, struct lpfc_sli_ring *, uint32_t); void lpfc_sli_def_mbox_cmpl(struct lpfc_hba *, LPFC_MBOXQ_t *); diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c index 4126fd8..b89f6cb 100644 --- a/drivers/scsi/lpfc/lpfc_els.c +++ b/drivers/scsi/lpfc/lpfc_els.c @@ -648,33 +648,32 @@ lpfc_more_plogi(struct lpfc_hba * phba) } static struct lpfc_nodelist * -lpfc_plogi_confirm_nport(struct lpfc_hba * phba, struct lpfc_iocbq * cmdiocb, +lpfc_plogi_confirm_nport(struct lpfc_hba * phba, struct lpfc_dmabuf *prsp, struct lpfc_nodelist *ndlp) { struct lpfc_nodelist *new_ndlp; - struct lpfc_dmabuf *pcmd, *prsp; uint32_t *lp; struct serv_parm *sp; uint8_t name[sizeof (struct lpfc_name)]; uint32_t rc; - pcmd = (struct lpfc_dmabuf *) cmdiocb->context2; - prsp = (struct lpfc_dmabuf *) pcmd->list.next; lp = (uint32_t *) prsp->virt; sp = (struct serv_parm *) ((uint8_t *) lp + sizeof (uint32_t)); + memset(name, 0, sizeof (struct lpfc_name)); /* Now we to find out if the NPort we are logging into, matches the WWPN * we have for that ndlp. If not, we have some work to do. */ new_ndlp = lpfc_findnode_wwpn(phba, NLP_SEARCH_ALL, &sp->portName); - memset(name, 0, sizeof (struct lpfc_name)); - rc = memcmp(&ndlp->nlp_portname, name, sizeof(struct lpfc_name)); - if (!rc || (new_ndlp == ndlp)) { + if (new_ndlp == ndlp) return ndlp; - } if (!new_ndlp) { + rc = + memcmp(&ndlp->nlp_portname, name, sizeof(struct lpfc_name)); + if (!rc) + return ndlp; new_ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_ATOMIC); if (!new_ndlp) return ndlp; @@ -683,17 +682,21 @@ lpfc_plogi_confirm_nport(struct lpfc_hba } lpfc_unreg_rpi(phba, new_ndlp); - new_ndlp->nlp_prev_state = ndlp->nlp_state; new_ndlp->nlp_DID = ndlp->nlp_DID; - new_ndlp->nlp_state = NLP_STE_PLOGI_ISSUE; - lpfc_nlp_list(phba, new_ndlp, NLP_PLOGI_LIST); + new_ndlp->nlp_prev_state = ndlp->nlp_prev_state; + new_ndlp->nlp_state = ndlp->nlp_state; + lpfc_nlp_list(phba, new_ndlp, ndlp->nlp_flag & NLP_LIST_MASK); /* Move this back to NPR list */ - lpfc_unreg_rpi(phba, ndlp); - ndlp->nlp_DID = 0; /* Two ndlps cannot have the same did */ - ndlp->nlp_state = NLP_STE_NPR_NODE; - lpfc_nlp_list(phba, ndlp, NLP_NPR_LIST); - + if (memcmp(&ndlp->nlp_portname, name, sizeof(struct lpfc_name)) == 0) { + lpfc_nlp_list(phba, ndlp, NLP_NO_LIST); + } + else { + lpfc_unreg_rpi(phba, ndlp); + ndlp->nlp_DID = 0; /* Two ndlps cannot have the same did */ + ndlp->nlp_state = NLP_STE_NPR_NODE; + lpfc_nlp_list(phba, ndlp, NLP_NPR_LIST); + } return new_ndlp; } @@ -703,6 +706,7 @@ lpfc_cmpl_els_plogi(struct lpfc_hba * ph { IOCB_t *irsp; struct lpfc_nodelist *ndlp; + struct lpfc_dmabuf *prsp; int disc, rc, did, type; @@ -769,7 +773,10 @@ lpfc_cmpl_els_plogi(struct lpfc_hba * ph } } else { /* Good status, call state machine */ - ndlp = lpfc_plogi_confirm_nport(phba, cmdiocb, ndlp); + prsp = list_entry(((struct lpfc_dmabuf *) + cmdiocb->context2)->list.next, + struct lpfc_dmabuf, list); + ndlp = lpfc_plogi_confirm_nport(phba, prsp, ndlp); rc = lpfc_disc_state_machine(phba, ndlp, cmdiocb, NLP_EVT_CMPL_PLOGI); } @@ -3282,10 +3289,9 @@ lpfc_els_timeout_handler(struct lpfc_hba } else lpfc_sli_release_iocbq(phba, piocb); } - if (phba->sli.ring[LPFC_ELS_RING].txcmplq_cnt) { - phba->els_tmofunc.expires = jiffies + HZ * timeout; - add_timer(&phba->els_tmofunc); - } + if (phba->sli.ring[LPFC_ELS_RING].txcmplq_cnt) + mod_timer(&phba->els_tmofunc, jiffies + HZ * timeout); + spin_unlock_irq(phba->host->host_lock); } @@ -3442,6 +3448,8 @@ lpfc_els_unsol_event(struct lpfc_hba * p if ((did & Fabric_DID_MASK) == Fabric_DID_MASK) { ndlp->nlp_type |= NLP_FABRIC; } + ndlp->nlp_state = NLP_STE_UNUSED_NODE; + lpfc_nlp_list(phba, ndlp, NLP_UNUSED_LIST); } phba->fc_stat.elsRcvFrame++; @@ -3463,13 +3471,14 @@ lpfc_els_unsol_event(struct lpfc_hba * p rjt_err = 1; break; } + ndlp = lpfc_plogi_confirm_nport(phba, mp, ndlp); lpfc_disc_state_machine(phba, ndlp, elsiocb, NLP_EVT_RCV_PLOGI); break; case ELS_CMD_FLOGI: phba->fc_stat.elsRcvFLOGI++; lpfc_els_rcv_flogi(phba, elsiocb, ndlp, newnode); if (newnode) { - mempool_free( ndlp, phba->nlp_mem_pool); + lpfc_nlp_list(phba, ndlp, NLP_NO_LIST); } break; case ELS_CMD_LOGO: @@ -3492,7 +3501,7 @@ lpfc_els_unsol_event(struct lpfc_hba * p phba->fc_stat.elsRcvRSCN++; lpfc_els_rcv_rscn(phba, elsiocb, ndlp, newnode); if (newnode) { - mempool_free( ndlp, phba->nlp_mem_pool); + lpfc_nlp_list(phba, ndlp, NLP_NO_LIST); } break; case ELS_CMD_ADISC: @@ -3535,28 +3544,28 @@ lpfc_els_unsol_event(struct lpfc_hba * p phba->fc_stat.elsRcvLIRR++; lpfc_els_rcv_lirr(phba, elsiocb, ndlp); if (newnode) { - mempool_free( ndlp, phba->nlp_mem_pool); + lpfc_nlp_list(phba, ndlp, NLP_NO_LIST); } break; case ELS_CMD_RPS: phba->fc_stat.elsRcvRPS++; lpfc_els_rcv_rps(phba, elsiocb, ndlp); if (newnode) { - mempool_free( ndlp, phba->nlp_mem_pool); + lpfc_nlp_list(phba, ndlp, NLP_NO_LIST); } break; case ELS_CMD_RPL: phba->fc_stat.elsRcvRPL++; lpfc_els_rcv_rpl(phba, elsiocb, ndlp); if (newnode) { - mempool_free( ndlp, phba->nlp_mem_pool); + lpfc_nlp_list(phba, ndlp, NLP_NO_LIST); } break; case ELS_CMD_RNID: phba->fc_stat.elsRcvRNID++; lpfc_els_rcv_rnid(phba, elsiocb, ndlp); if (newnode) { - mempool_free( ndlp, phba->nlp_mem_pool); + lpfc_nlp_list(phba, ndlp, NLP_NO_LIST); } break; default: @@ -3568,7 +3577,7 @@ lpfc_els_unsol_event(struct lpfc_hba * p "%d:0115 Unknown ELS command x%x received from " "NPORT x%x\n", phba->brd_no, cmd, did); if (newnode) { - mempool_free( ndlp, phba->nlp_mem_pool); + lpfc_nlp_list(phba, ndlp, NLP_NO_LIST); } break; } diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c index adb0860..4d6cf99 100644 --- a/drivers/scsi/lpfc/lpfc_hbadisc.c +++ b/drivers/scsi/lpfc/lpfc_hbadisc.c @@ -1084,7 +1084,7 @@ lpfc_register_remote_port(struct lpfc_hb fc_remote_port_rolechg(rport, rport_ids.roles); if ((rport->scsi_target_id != -1) && - (rport->scsi_target_id < MAX_FCP_TARGET)) { + (rport->scsi_target_id < LPFC_MAX_TARGET)) { ndlp->nlp_sid = rport->scsi_target_id; } @@ -1313,7 +1313,7 @@ lpfc_nlp_list(struct lpfc_hba * phba, st if ((rport_add == mapped) && ((!nlp->rport) || (nlp->rport->scsi_target_id == -1) || - (nlp->rport->scsi_target_id >= MAX_FCP_TARGET))) { + (nlp->rport->scsi_target_id >= LPFC_MAX_TARGET))) { nlp->nlp_state = NLP_STE_UNMAPPED_NODE; spin_lock_irq(phba->host->host_lock); nlp->nlp_flag |= NLP_TGT_NO_SCSIID; diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c index 81755a3..ef47b82 100644 --- a/drivers/scsi/lpfc/lpfc_init.c +++ b/drivers/scsi/lpfc/lpfc_init.c @@ -71,6 +71,7 @@ lpfc_config_port_prep(struct lpfc_hba * uint16_t offset = 0; static char licensed[56] = "key unlock for use with gnu public licensed code only\0"; + static int init_key = 1; pmb = mempool_alloc(phba->mbox_mem_pool, GFP_KERNEL); if (!pmb) { @@ -82,10 +83,13 @@ lpfc_config_port_prep(struct lpfc_hba * phba->hba_state = LPFC_INIT_MBX_CMDS; if (lpfc_is_LC_HBA(phba->pcidev->device)) { - uint32_t *ptext = (uint32_t *) licensed; + if (init_key) { + uint32_t *ptext = (uint32_t *) licensed; - for (i = 0; i < 56; i += sizeof (uint32_t), ptext++) - *ptext = cpu_to_be32(*ptext); + for (i = 0; i < 56; i += sizeof (uint32_t), ptext++) + *ptext = cpu_to_be32(*ptext); + init_key = 0; + } lpfc_read_nv(phba, pmb); memset((char*)mb->un.varRDnvp.rsvd3, 0, @@ -405,19 +409,26 @@ lpfc_config_port_post(struct lpfc_hba * } /* MBOX buffer will be freed in mbox compl */ - i = 0; + return (0); +} + +static int +lpfc_discovery_wait(struct lpfc_hba *phba) +{ + int i = 0; + while ((phba->hba_state != LPFC_HBA_READY) || (phba->num_disc_nodes) || (phba->fc_prli_sent) || ((phba->fc_map_cnt == 0) && (i<2)) || - (psli->sli_flag & LPFC_SLI_MBOX_ACTIVE)) { + (phba->sli.sli_flag & LPFC_SLI_MBOX_ACTIVE)) { /* Check every second for 30 retries. */ i++; if (i > 30) { - break; + return -ETIMEDOUT; } if ((i >= 15) && (phba->hba_state <= LPFC_LINK_DOWN)) { /* The link is down. Set linkdown timeout */ - break; + return -ETIMEDOUT; } /* Delay for 1 second to give discovery time to complete. */ @@ -425,12 +436,7 @@ lpfc_config_port_post(struct lpfc_hba * } - /* Since num_disc_nodes keys off of PLOGI, delay a bit to let - * any potential PRLIs to flush thru the SLI sub-system. - */ - msleep(50); - - return (0); + return 0; } /************************************************************************/ @@ -1339,7 +1345,8 @@ lpfc_offline(struct lpfc_hba * phba) struct lpfc_sli_ring *pring; struct lpfc_sli *psli; unsigned long iflag; - int i = 0; + int i; + int cnt = 0; if (!phba) return 0; @@ -1348,17 +1355,27 @@ lpfc_offline(struct lpfc_hba * phba) return 0; psli = &phba->sli; - pring = &psli->ring[psli->fcp_ring]; lpfc_linkdown(phba); + lpfc_sli_flush_mbox_queue(phba); - /* The linkdown event takes 30 seconds to timeout. */ - while (pring->txcmplq_cnt) { - mdelay(10); - if (i++ > 3000) - break; + for (i = 0; i < psli->num_rings; i++) { + pring = &psli->ring[i]; + /* The linkdown event takes 30 seconds to timeout. */ + while (pring->txcmplq_cnt) { + mdelay(10); + if (cnt++ > 3000) { + lpfc_printf_log(phba, + KERN_WARNING, LOG_INIT, + "%d:0466 Outstanding IO when " + "bringing Adapter offline\n", + phba->brd_no); + break; + } + } } + /* stop all timers associated with this hba */ lpfc_stop_timer(phba); phba->work_hba_events = 0; @@ -1639,6 +1656,8 @@ lpfc_pci_probe_one(struct pci_dev *pdev, goto out_free_irq; } + lpfc_discovery_wait(phba); + if (phba->cfg_poll & DISABLE_FCP_RING_INT) { spin_lock_irq(phba->host->host_lock); lpfc_poll_start_timer(phba); diff --git a/drivers/scsi/lpfc/lpfc_mem.c b/drivers/scsi/lpfc/lpfc_mem.c index 0701765..066292d 100644 --- a/drivers/scsi/lpfc/lpfc_mem.c +++ b/drivers/scsi/lpfc/lpfc_mem.c @@ -133,6 +133,11 @@ lpfc_mem_free(struct lpfc_hba * phba) pci_pool_destroy(phba->lpfc_scsi_dma_buf_pool); pci_pool_destroy(phba->lpfc_mbuf_pool); + + /* Free the iocb lookup array */ + kfree(psli->iocbq_lookup); + psli->iocbq_lookup = NULL; + } void * diff --git a/drivers/scsi/lpfc/lpfc_nportdisc.c b/drivers/scsi/lpfc/lpfc_nportdisc.c index 27d60ad..bd0b0e2 100644 --- a/drivers/scsi/lpfc/lpfc_nportdisc.c +++ b/drivers/scsi/lpfc/lpfc_nportdisc.c @@ -1110,6 +1110,17 @@ lpfc_cmpl_reglogin_reglogin_issue(struct phba->brd_no, did, mb->mbxStatus, phba->hba_state); + /* + * If RegLogin failed due to lack of HBA resources do not + * retry discovery. + */ + if (mb->mbxStatus == MBXERR_RPI_FULL) { + ndlp->nlp_prev_state = NLP_STE_UNUSED_NODE; + ndlp->nlp_state = NLP_STE_UNUSED_NODE; + lpfc_nlp_list(phba, ndlp, NLP_UNUSED_LIST); + return ndlp->nlp_state; + } + /* Put ndlp in npr list set plogi timer for 1 sec */ mod_timer(&ndlp->nlp_delayfunc, jiffies + HZ * 1); spin_lock_irq(phba->host->host_lock); diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c index aea1ee4..a760a44 100644 --- a/drivers/scsi/lpfc/lpfc_scsi.c +++ b/drivers/scsi/lpfc/lpfc_scsi.c @@ -153,22 +153,6 @@ static void lpfc_release_scsi_buf(struct lpfc_hba * phba, struct lpfc_scsi_buf * psb) { unsigned long iflag = 0; - /* - * There are only two special cases to consider. (1) the scsi command - * requested scatter-gather usage or (2) the scsi command allocated - * a request buffer, but did not request use_sg. There is a third - * case, but it does not require resource deallocation. - */ - if ((psb->seg_cnt > 0) && (psb->pCmd->use_sg)) { - dma_unmap_sg(&phba->pcidev->dev, psb->pCmd->request_buffer, - psb->seg_cnt, psb->pCmd->sc_data_direction); - } else { - if ((psb->nonsg_phys) && (psb->pCmd->request_bufflen)) { - dma_unmap_single(&phba->pcidev->dev, psb->nonsg_phys, - psb->pCmd->request_bufflen, - psb->pCmd->sc_data_direction); - } - } spin_lock_irqsave(&phba->scsi_buf_list_lock, iflag); psb->pCmd = NULL; @@ -282,6 +266,27 @@ lpfc_scsi_prep_dma_buf(struct lpfc_hba * } static void +lpfc_scsi_unprep_dma_buf(struct lpfc_hba * phba, struct lpfc_scsi_buf * psb) +{ + /* + * There are only two special cases to consider. (1) the scsi command + * requested scatter-gather usage or (2) the scsi command allocated + * a request buffer, but did not request use_sg. There is a third + * case, but it does not require resource deallocation. + */ + if ((psb->seg_cnt > 0) && (psb->pCmd->use_sg)) { + dma_unmap_sg(&phba->pcidev->dev, psb->pCmd->request_buffer, + psb->seg_cnt, psb->pCmd->sc_data_direction); + } else { + if ((psb->nonsg_phys) && (psb->pCmd->request_bufflen)) { + dma_unmap_single(&phba->pcidev->dev, psb->nonsg_phys, + psb->pCmd->request_bufflen, + psb->pCmd->sc_data_direction); + } + } +} + +static void lpfc_handle_fcp_err(struct lpfc_scsi_buf *lpfc_cmd) { struct scsi_cmnd *cmnd = lpfc_cmd->pCmd; @@ -454,6 +459,7 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba cmd->scsi_done(cmd); if (phba->cfg_poll & ENABLE_FCP_RING_POLLING) { + lpfc_scsi_unprep_dma_buf(phba, lpfc_cmd); lpfc_release_scsi_buf(phba, lpfc_cmd); return; } @@ -511,6 +517,7 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba } } + lpfc_scsi_unprep_dma_buf(phba, lpfc_cmd); lpfc_release_scsi_buf(phba, lpfc_cmd); } @@ -609,6 +616,7 @@ lpfc_scsi_prep_cmnd(struct lpfc_hba * ph static int lpfc_scsi_prep_task_mgmt_cmd(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd, + unsigned int lun, uint8_t task_mgmt_cmd) { struct lpfc_sli *psli; @@ -627,8 +635,7 @@ lpfc_scsi_prep_task_mgmt_cmd(struct lpfc piocb = &piocbq->iocb; fcp_cmnd = lpfc_cmd->fcp_cmnd; - int_to_scsilun(lpfc_cmd->pCmd->device->lun, - &lpfc_cmd->fcp_cmnd->fcp_lun); + int_to_scsilun(lun, &lpfc_cmd->fcp_cmnd->fcp_lun); fcp_cmnd->fcpCntl2 = task_mgmt_cmd; piocb->ulpCommand = CMD_FCP_ICMND64_CR; @@ -655,14 +662,16 @@ lpfc_scsi_prep_task_mgmt_cmd(struct lpfc static int lpfc_scsi_tgt_reset(struct lpfc_scsi_buf * lpfc_cmd, struct lpfc_hba * phba, - unsigned tgt_id, struct lpfc_rport_data *rdata) + unsigned tgt_id, unsigned int lun, + struct lpfc_rport_data *rdata) { struct lpfc_iocbq *iocbq; struct lpfc_iocbq *iocbqrsp; int ret; lpfc_cmd->rdata = rdata; - ret = lpfc_scsi_prep_task_mgmt_cmd(phba, lpfc_cmd, FCP_TARGET_RESET); + ret = lpfc_scsi_prep_task_mgmt_cmd(phba, lpfc_cmd, lun, + FCP_TARGET_RESET); if (!ret) return FAILED; @@ -822,6 +831,7 @@ lpfc_queuecommand(struct scsi_cmnd *cmnd return 0; out_host_busy_free_buf: + lpfc_scsi_unprep_dma_buf(phba, lpfc_cmd); lpfc_release_scsi_buf(phba, lpfc_cmd); out_host_busy: return SCSI_MLQUEUE_HOST_BUSY; @@ -969,12 +979,12 @@ lpfc_reset_lun_handler(struct scsi_cmnd if (lpfc_cmd == NULL) goto out; - lpfc_cmd->pCmd = cmnd; lpfc_cmd->timeout = 60; lpfc_cmd->scsi_hba = phba; lpfc_cmd->rdata = rdata; - ret = lpfc_scsi_prep_task_mgmt_cmd(phba, lpfc_cmd, FCP_LUN_RESET); + ret = lpfc_scsi_prep_task_mgmt_cmd(phba, lpfc_cmd, cmnd->device->lun, + FCP_LUN_RESET); if (!ret) goto out_free_scsi_buf; @@ -1001,7 +1011,6 @@ lpfc_reset_lun_handler(struct scsi_cmnd cmd_status = iocbqrsp->iocb.ulpStatus; lpfc_sli_release_iocbq(phba, iocbqrsp); - lpfc_release_scsi_buf(phba, lpfc_cmd); /* * All outstanding txcmplq I/Os should have been aborted by the device. @@ -1040,6 +1049,8 @@ lpfc_reset_lun_handler(struct scsi_cmnd } out_free_scsi_buf: + lpfc_release_scsi_buf(phba, lpfc_cmd); + lpfc_printf_log(phba, KERN_ERR, LOG_FCP, "%d:0713 SCSI layer issued LUN reset (%d, %d) " "Data: x%x x%x x%x\n", @@ -1070,7 +1081,6 @@ lpfc_reset_bus_handler(struct scsi_cmnd /* The lpfc_cmd storage is reused. Set all loop invariants. */ lpfc_cmd->timeout = 60; - lpfc_cmd->pCmd = cmnd; lpfc_cmd->scsi_hba = phba; /* @@ -1078,7 +1088,7 @@ lpfc_reset_bus_handler(struct scsi_cmnd * targets known to the driver. Should any target reset * fail, this routine returns failure to the midlayer. */ - for (i = 0; i < MAX_FCP_TARGET; i++) { + for (i = 0; i < LPFC_MAX_TARGET; i++) { /* Search the mapped list for this target ID */ match = 0; list_for_each_entry(ndlp, &phba->fc_nlpmap_list, nlp_listp) { @@ -1090,8 +1100,8 @@ lpfc_reset_bus_handler(struct scsi_cmnd if (!match) continue; - ret = lpfc_scsi_tgt_reset(lpfc_cmd, phba, - i, ndlp->rport->dd_data); + ret = lpfc_scsi_tgt_reset(lpfc_cmd, phba, i, cmnd->device->lun, + ndlp->rport->dd_data); if (ret != SUCCESS) { lpfc_printf_log(phba, KERN_ERR, LOG_FCP, "%d:0713 Bus Reset on target %d failed\n", diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c index bb69a7a..350a625 100644 --- a/drivers/scsi/lpfc/lpfc_sli.c +++ b/drivers/scsi/lpfc/lpfc_sli.c @@ -191,35 +191,12 @@ static int lpfc_sli_ringtxcmpl_put(struct lpfc_hba * phba, struct lpfc_sli_ring * pring, struct lpfc_iocbq * piocb) { - uint16_t iotag; - list_add_tail(&piocb->list, &pring->txcmplq); pring->txcmplq_cnt++; if (unlikely(pring->ringno == LPFC_ELS_RING)) mod_timer(&phba->els_tmofunc, jiffies + HZ * (phba->fc_ratov << 1)); - if (pring->fast_lookup) { - /* Setup fast lookup based on iotag for completion */ - iotag = piocb->iocb.ulpIoTag; - if (iotag && (iotag < pring->fast_iotag)) - *(pring->fast_lookup + iotag) = piocb; - else { - - /* Cmd ring put: iotag greater then - configured max wd0 */ - lpfc_printf_log(phba, - KERN_ERR, - LOG_SLI, - "%d:0316 Cmd ring %d put: iotag x%x " - "greater then configured max x%x " - "wd0 x%x\n", - phba->brd_no, - pring->ringno, iotag, - pring->fast_iotag, - *(((uint32_t *)(&piocb->iocb)) + 7)); - } - } return (0); } @@ -601,7 +578,7 @@ lpfc_sli_handle_mb_event(struct lpfc_hba /* Stray Mailbox Interrupt, mbxCommand mbxStatus */ lpfc_printf_log(phba, - KERN_ERR, + KERN_WARNING, LOG_MBOX | LOG_SLI, "%d:0304 Stray Mailbox Interrupt " "mbxCommand x%x mbxStatus x%x\n", @@ -1570,8 +1547,8 @@ #define BARRIER_TEST_PATTERN (0xdeadbeef void lpfc_reset_barrier(struct lpfc_hba * phba) { - uint32_t * resp_buf; - uint32_t * mbox_buf; + uint32_t __iomem *resp_buf; + uint32_t __iomem *mbox_buf; volatile uint32_t mbox; uint32_t hc_copy; int i; @@ -1587,7 +1564,7 @@ void lpfc_reset_barrier(struct lpfc_hba * Tell the other part of the chip to suspend temporarily all * its DMA activity. */ - resp_buf = (uint32_t *)phba->MBslimaddr; + resp_buf = phba->MBslimaddr; /* Disable the error attention */ hc_copy = readl(phba->HCregaddr); @@ -1605,7 +1582,7 @@ void lpfc_reset_barrier(struct lpfc_hba ((MAILBOX_t *)&mbox)->mbxOwner = OWN_CHIP; writel(BARRIER_TEST_PATTERN, (resp_buf + 1)); - mbox_buf = (uint32_t *)phba->MBslimaddr; + mbox_buf = phba->MBslimaddr; writel(mbox, mbox_buf); for (i = 0; @@ -1805,7 +1782,7 @@ lpfc_sli_brdrestart(struct lpfc_hba * ph skip_post = 0; word0 = 0; /* This is really setting up word1 */ } - to_slim = (uint8_t *) phba->MBslimaddr + sizeof (uint32_t); + to_slim = phba->MBslimaddr + sizeof (uint32_t); writel(*(uint32_t *) mb, to_slim); readl(to_slim); /* flush */ @@ -2659,8 +2636,6 @@ lpfc_sli_hba_down(struct lpfc_hba * phba INIT_LIST_HEAD(&(pring->txq)); - kfree(pring->fast_lookup); - pring->fast_lookup = NULL; } spin_unlock_irqrestore(phba->host->host_lock, flags); @@ -3110,6 +3085,24 @@ lpfc_sli_issue_mbox_wait(struct lpfc_hba return retval; } +int +lpfc_sli_flush_mbox_queue(struct lpfc_hba * phba) +{ + int i = 0; + + while (phba->sli.sli_flag & LPFC_SLI_MBOX_ACTIVE && !phba->stopped) { + if (i++ > LPFC_MBOX_TMO * 1000) + return 1; + + if (lpfc_sli_handle_mb_event(phba) == 0) + i = 0; + + msleep(1); + } + + return (phba->sli.sli_flag & LPFC_SLI_MBOX_ACTIVE) ? 1 : 0; +} + irqreturn_t lpfc_intr_handler(int irq, void *dev_id, struct pt_regs * regs) { diff --git a/drivers/scsi/lpfc/lpfc_sli.h b/drivers/scsi/lpfc/lpfc_sli.h index a52d6c6..d8ef0d2 100644 --- a/drivers/scsi/lpfc/lpfc_sli.h +++ b/drivers/scsi/lpfc/lpfc_sli.h @@ -135,8 +135,6 @@ #define LPFC_STOP_IOCB_MASK 0x030 / uint32_t fast_iotag; /* max fastlookup based iotag */ uint32_t iotag_ctr; /* keeps track of the next iotag to use */ uint32_t iotag_max; /* max iotag value to use */ - struct lpfc_iocbq ** fast_lookup; /* array of IOCB ptrs indexed by - iotag */ struct list_head txq; uint16_t txq_cnt; /* current length of queue */ uint16_t txq_max; /* max length */ diff --git a/drivers/scsi/lpfc/lpfc_version.h b/drivers/scsi/lpfc/lpfc_version.h index 6b73756..10e89c6 100644 --- a/drivers/scsi/lpfc/lpfc_version.h +++ b/drivers/scsi/lpfc/lpfc_version.h @@ -18,7 +18,7 @@ * included with this package. * *******************************************************************/ -#define LPFC_DRIVER_VERSION "8.1.6" +#define LPFC_DRIVER_VERSION "8.1.7" #define LPFC_DRIVER_NAME "lpfc" diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c index 2ab7df0..b332cad 100644 --- a/drivers/scsi/scsi.c +++ b/drivers/scsi/scsi.c @@ -346,7 +346,7 @@ void scsi_log_send(struct scsi_cmnd *cmd if (level > 3) { printk(KERN_INFO "buffer = 0x%p, bufflen = %d," " done = 0x%p, queuecommand 0x%p\n", - cmd->buffer, cmd->bufflen, + cmd->request_buffer, cmd->request_bufflen, cmd->done, sdev->host->hostt->queuecommand); @@ -661,11 +661,6 @@ void __scsi_done(struct scsi_cmnd *cmd) */ int scsi_retry_command(struct scsi_cmnd *cmd) { - /* - * Restore the SCSI command state. - */ - scsi_setup_cmd_retry(cmd); - /* * Zero the sense information from the last time we tried * this command. @@ -711,10 +706,6 @@ void scsi_finish_command(struct scsi_cmn "Notifying upper driver of completion " "(result %x)\n", cmd->result)); - /* - * We can get here with use_sg=0, causing a panic in the upper level - */ - cmd->use_sg = cmd->old_use_sg; cmd->done(cmd); } EXPORT_SYMBOL(scsi_finish_command); diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c index 6683d59..6a5b731 100644 --- a/drivers/scsi/scsi_error.c +++ b/drivers/scsi/scsi_error.c @@ -460,19 +460,67 @@ static void scsi_eh_done(struct scsi_cmn * Return value: * SUCCESS or FAILED or NEEDS_RETRY **/ -static int scsi_send_eh_cmnd(struct scsi_cmnd *scmd, int timeout) +static int scsi_send_eh_cmnd(struct scsi_cmnd *scmd, int timeout, int copy_sense) { struct scsi_device *sdev = scmd->device; struct Scsi_Host *shost = sdev->host; + int old_result = scmd->result; DECLARE_COMPLETION(done); unsigned long timeleft; unsigned long flags; + unsigned char old_cmnd[MAX_COMMAND_SIZE]; + enum dma_data_direction old_data_direction; + unsigned short old_use_sg; + unsigned char old_cmd_len; + unsigned old_bufflen; + void *old_buffer; int rtn; + /* + * We need saved copies of a number of fields - this is because + * error handling may need to overwrite these with different values + * to run different commands, and once error handling is complete, + * we will need to restore these values prior to running the actual + * command. + */ + old_buffer = scmd->request_buffer; + old_bufflen = scmd->request_bufflen; + memcpy(old_cmnd, scmd->cmnd, sizeof(scmd->cmnd)); + old_data_direction = scmd->sc_data_direction; + old_cmd_len = scmd->cmd_len; + old_use_sg = scmd->use_sg; + + if (copy_sense) { + int gfp_mask = GFP_ATOMIC; + + if (shost->hostt->unchecked_isa_dma) + gfp_mask |= __GFP_DMA; + + scmd->sc_data_direction = DMA_FROM_DEVICE; + scmd->request_bufflen = 252; + scmd->request_buffer = kzalloc(scmd->request_bufflen, gfp_mask); + if (!scmd->request_buffer) + return FAILED; + } else { + scmd->request_buffer = NULL; + scmd->request_bufflen = 0; + scmd->sc_data_direction = DMA_NONE; + } + + scmd->underflow = 0; + scmd->use_sg = 0; + scmd->cmd_len = COMMAND_SIZE(scmd->cmnd[0]); + if (sdev->scsi_level <= SCSI_2) scmd->cmnd[1] = (scmd->cmnd[1] & 0x1f) | (sdev->lun << 5 & 0xe0); + /* + * Zero the sense buffer. The scsi spec mandates that any + * untransferred sense data should be interpreted as being zero. + */ + memset(scmd->sense_buffer, 0, sizeof(scmd->sense_buffer)); + shost->eh_action = &done; spin_lock_irqsave(shost->host_lock, flags); @@ -522,6 +570,29 @@ static int scsi_send_eh_cmnd(struct scsi rtn = FAILED; } + + /* + * Last chance to have valid sense data. + */ + if (copy_sense) { + if (!SCSI_SENSE_VALID(scmd)) { + memcpy(scmd->sense_buffer, scmd->request_buffer, + sizeof(scmd->sense_buffer)); + } + kfree(scmd->request_buffer); + } + + + /* + * Restore original data + */ + scmd->request_buffer = old_buffer; + scmd->request_bufflen = old_bufflen; + memcpy(scmd->cmnd, old_cmnd, sizeof(scmd->cmnd)); + scmd->sc_data_direction = old_data_direction; + scmd->cmd_len = old_cmd_len; + scmd->use_sg = old_use_sg; + scmd->result = old_result; return rtn; } @@ -537,56 +608,10 @@ static int scsi_send_eh_cmnd(struct scsi static int scsi_request_sense(struct scsi_cmnd *scmd) { static unsigned char generic_sense[6] = - {REQUEST_SENSE, 0, 0, 0, 252, 0}; - unsigned char *scsi_result; - int saved_result; - int rtn; + {REQUEST_SENSE, 0, 0, 0, 252, 0}; memcpy(scmd->cmnd, generic_sense, sizeof(generic_sense)); - - scsi_result = kmalloc(252, GFP_ATOMIC | ((scmd->device->host->hostt->unchecked_isa_dma) ? __GFP_DMA : 0)); - - - if (unlikely(!scsi_result)) { - printk(KERN_ERR "%s: cannot allocate scsi_result.\n", - __FUNCTION__); - return FAILED; - } - - /* - * zero the sense buffer. some host adapters automatically always - * request sense, so it is not a good idea that - * scmd->request_buffer and scmd->sense_buffer point to the same - * address (db). 0 is not a valid sense code. - */ - memset(scmd->sense_buffer, 0, sizeof(scmd->sense_buffer)); - memset(scsi_result, 0, 252); - - saved_result = scmd->result; - scmd->request_buffer = scsi_result; - scmd->request_bufflen = 252; - scmd->use_sg = 0; - scmd->cmd_len = COMMAND_SIZE(scmd->cmnd[0]); - scmd->sc_data_direction = DMA_FROM_DEVICE; - scmd->underflow = 0; - - rtn = scsi_send_eh_cmnd(scmd, SENSE_TIMEOUT); - - /* last chance to have valid sense data */ - if(!SCSI_SENSE_VALID(scmd)) { - memcpy(scmd->sense_buffer, scmd->request_buffer, - sizeof(scmd->sense_buffer)); - } - - kfree(scsi_result); - - /* - * when we eventually call scsi_finish, we really wish to complete - * the original request, so let's restore the original data. (db) - */ - scsi_setup_cmd_retry(scmd); - scmd->result = saved_result; - return rtn; + return scsi_send_eh_cmnd(scmd, SENSE_TIMEOUT, 1); } /** @@ -605,12 +630,6 @@ void scsi_eh_finish_cmd(struct scsi_cmnd { scmd->device->host->host_failed--; scmd->eh_eflags = 0; - - /* - * set this back so that the upper level can correctly free up - * things. - */ - scsi_setup_cmd_retry(scmd); list_move_tail(&scmd->eh_entry, done_q); } EXPORT_SYMBOL(scsi_eh_finish_cmd); @@ -715,47 +734,26 @@ static int scsi_eh_tur(struct scsi_cmnd { static unsigned char tur_command[6] = {TEST_UNIT_READY, 0, 0, 0, 0, 0}; int retry_cnt = 1, rtn; - int saved_result; retry_tur: memcpy(scmd->cmnd, tur_command, sizeof(tur_command)); - /* - * zero the sense buffer. the scsi spec mandates that any - * untransferred sense data should be interpreted as being zero. - */ - memset(scmd->sense_buffer, 0, sizeof(scmd->sense_buffer)); - - saved_result = scmd->result; - scmd->request_buffer = NULL; - scmd->request_bufflen = 0; - scmd->use_sg = 0; - scmd->cmd_len = COMMAND_SIZE(scmd->cmnd[0]); - scmd->underflow = 0; - scmd->sc_data_direction = DMA_NONE; - rtn = scsi_send_eh_cmnd(scmd, SENSE_TIMEOUT); + rtn = scsi_send_eh_cmnd(scmd, SENSE_TIMEOUT, 0); - /* - * when we eventually call scsi_finish, we really wish to complete - * the original request, so let's restore the original data. (db) - */ - scsi_setup_cmd_retry(scmd); - scmd->result = saved_result; - - /* - * hey, we are done. let's look to see what happened. - */ SCSI_LOG_ERROR_RECOVERY(3, printk("%s: scmd %p rtn %x\n", __FUNCTION__, scmd, rtn)); - if (rtn == SUCCESS) - return 0; - else if (rtn == NEEDS_RETRY) { + + switch (rtn) { + case NEEDS_RETRY: if (retry_cnt--) goto retry_tur; + /*FALLTHRU*/ + case SUCCESS: return 0; + default: + return 1; } - return 1; } /** @@ -837,44 +835,16 @@ static int scsi_try_bus_device_reset(str static int scsi_eh_try_stu(struct scsi_cmnd *scmd) { static unsigned char stu_command[6] = {START_STOP, 0, 0, 0, 1, 0}; - int rtn; - int saved_result; - if (!scmd->device->allow_restart) - return 1; - - memcpy(scmd->cmnd, stu_command, sizeof(stu_command)); - - /* - * zero the sense buffer. the scsi spec mandates that any - * untransferred sense data should be interpreted as being zero. - */ - memset(scmd->sense_buffer, 0, sizeof(scmd->sense_buffer)); - - saved_result = scmd->result; - scmd->request_buffer = NULL; - scmd->request_bufflen = 0; - scmd->use_sg = 0; - scmd->cmd_len = COMMAND_SIZE(scmd->cmnd[0]); - scmd->underflow = 0; - scmd->sc_data_direction = DMA_NONE; + if (scmd->device->allow_restart) { + int rtn; - rtn = scsi_send_eh_cmnd(scmd, START_UNIT_TIMEOUT); - - /* - * when we eventually call scsi_finish, we really wish to complete - * the original request, so let's restore the original data. (db) - */ - scsi_setup_cmd_retry(scmd); - scmd->result = saved_result; + memcpy(scmd->cmnd, stu_command, sizeof(stu_command)); + rtn = scsi_send_eh_cmnd(scmd, START_UNIT_TIMEOUT, 0); + if (rtn == SUCCESS) + return 0; + } - /* - * hey, we are done. let's look to see what happened. - */ - SCSI_LOG_ERROR_RECOVERY(3, printk("%s: scmd %p rtn %x\n", - __FUNCTION__, scmd, rtn)); - if (rtn == SUCCESS) - return 0; return 1; } @@ -1684,8 +1654,6 @@ scsi_reset_provider(struct scsi_device * scmd->scsi_done = scsi_reset_provider_done_command; scmd->done = NULL; - scmd->buffer = NULL; - scmd->bufflen = 0; scmd->request_buffer = NULL; scmd->request_bufflen = 0; diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 08af9aa..077c1c6 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -436,60 +436,16 @@ EXPORT_SYMBOL_GPL(scsi_execute_async); * * Arguments: cmd - command that is ready to be queued. * - * Returns: Nothing - * * Notes: This function has the job of initializing a number of * fields related to error handling. Typically this will * be called once for each command, as required. */ -static int scsi_init_cmd_errh(struct scsi_cmnd *cmd) +static void scsi_init_cmd_errh(struct scsi_cmnd *cmd) { cmd->serial_number = 0; - memset(cmd->sense_buffer, 0, sizeof cmd->sense_buffer); - if (cmd->cmd_len == 0) cmd->cmd_len = COMMAND_SIZE(cmd->cmnd[0]); - - /* - * We need saved copies of a number of fields - this is because - * error handling may need to overwrite these with different values - * to run different commands, and once error handling is complete, - * we will need to restore these values prior to running the actual - * command. - */ - cmd->old_use_sg = cmd->use_sg; - cmd->old_cmd_len = cmd->cmd_len; - cmd->sc_old_data_direction = cmd->sc_data_direction; - cmd->old_underflow = cmd->underflow; - memcpy(cmd->data_cmnd, cmd->cmnd, sizeof(cmd->cmnd)); - cmd->buffer = cmd->request_buffer; - cmd->bufflen = cmd->request_bufflen; - - return 1; -} - -/* - * Function: scsi_setup_cmd_retry() - * - * Purpose: Restore the command state for a retry - * - * Arguments: cmd - command to be restored - * - * Returns: Nothing - * - * Notes: Immediately prior to retrying a command, we need - * to restore certain fields that we saved above. - */ -void scsi_setup_cmd_retry(struct scsi_cmnd *cmd) -{ - memcpy(cmd->cmnd, cmd->data_cmnd, sizeof(cmd->data_cmnd)); - cmd->request_buffer = cmd->buffer; - cmd->request_bufflen = cmd->bufflen; - cmd->use_sg = cmd->old_use_sg; - cmd->cmd_len = cmd->old_cmd_len; - cmd->sc_data_direction = cmd->sc_old_data_direction; - cmd->underflow = cmd->old_underflow; } void scsi_device_unbusy(struct scsi_device *sdev) @@ -807,22 +763,13 @@ static void scsi_free_sgtable(struct sca */ static void scsi_release_buffers(struct scsi_cmnd *cmd) { - struct request *req = cmd->request; - - /* - * Free up any indirection buffers we allocated for DMA purposes. - */ if (cmd->use_sg) scsi_free_sgtable(cmd->request_buffer, cmd->sglist_len); - else if (cmd->request_buffer != req->buffer) - kfree(cmd->request_buffer); /* * Zero these out. They now point to freed memory, and it is * dangerous to hang onto the pointers. */ - cmd->buffer = NULL; - cmd->bufflen = 0; cmd->request_buffer = NULL; cmd->request_bufflen = 0; } @@ -858,7 +805,7 @@ static void scsi_release_buffers(struct void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes) { int result = cmd->result; - int this_count = cmd->bufflen; + int this_count = cmd->request_bufflen; request_queue_t *q = cmd->device->request_queue; struct request *req = cmd->request; int clear_errors = 1; @@ -866,28 +813,14 @@ void scsi_io_completion(struct scsi_cmnd int sense_valid = 0; int sense_deferred = 0; - /* - * Free up any indirection buffers we allocated for DMA purposes. - * For the case of a READ, we need to copy the data out of the - * bounce buffer and into the real buffer. - */ - if (cmd->use_sg) - scsi_free_sgtable(cmd->buffer, cmd->sglist_len); - else if (cmd->buffer != req->buffer) { - if (rq_data_dir(req) == READ) { - unsigned long flags; - char *to = bio_kmap_irq(req->bio, &flags); - memcpy(to, cmd->buffer, cmd->bufflen); - bio_kunmap_irq(to, &flags); - } - kfree(cmd->buffer); - } + scsi_release_buffers(cmd); if (result) { sense_valid = scsi_command_normalize_sense(cmd, &sshdr); if (sense_valid) sense_deferred = scsi_sense_is_deferred(&sshdr); } + if (blk_pc_request(req)) { /* SG_IO ioctl from block level */ req->errors = result; if (result) { @@ -908,15 +841,6 @@ void scsi_io_completion(struct scsi_cmnd } /* - * Zero these out. They now point to freed memory, and it is - * dangerous to hang onto the pointers. - */ - cmd->buffer = NULL; - cmd->bufflen = 0; - cmd->request_buffer = NULL; - cmd->request_bufflen = 0; - - /* * Next deal with any sectors which we were able to correctly * handle. */ @@ -1012,7 +936,7 @@ void scsi_io_completion(struct scsi_cmnd if (!(req->flags & REQ_QUIET)) { scmd_printk(KERN_INFO, cmd, "Volume overflow, CDB: "); - __scsi_print_command(cmd->data_cmnd); + __scsi_print_command(cmd->cmnd); scsi_print_sense("", cmd); } /* See SSC3rXX or current. */ @@ -1143,7 +1067,7 @@ static void scsi_blk_pc_done(struct scsi * successfully. Since this is a REQ_BLOCK_PC command the * caller should check the request's errors value */ - scsi_io_completion(cmd, cmd->bufflen); + scsi_io_completion(cmd, cmd->request_bufflen); } static void scsi_setup_blk_pc_cmnd(struct scsi_cmnd *cmd) diff --git a/drivers/scsi/scsi_priv.h b/drivers/scsi/scsi_priv.h index e2fbe9a..ae24c85 100644 --- a/drivers/scsi/scsi_priv.h +++ b/drivers/scsi/scsi_priv.h @@ -57,7 +57,6 @@ extern int scsi_eh_scmd_add(struct scsi_ /* scsi_lib.c */ extern int scsi_maybe_unblock_host(struct scsi_device *sdev); -extern void scsi_setup_cmd_retry(struct scsi_cmnd *cmd); extern void scsi_device_unbusy(struct scsi_device *sdev); extern int scsi_queue_insert(struct scsi_cmnd *cmd, int reason); extern void scsi_next_command(struct scsi_cmnd *cmd); diff --git a/drivers/scsi/scsi_transport_sas.c b/drivers/scsi/scsi_transport_sas.c index dd07562..9bb4e37 100644 --- a/drivers/scsi/scsi_transport_sas.c +++ b/drivers/scsi/scsi_transport_sas.c @@ -41,6 +41,7 @@ struct sas_host_attrs { struct mutex lock; u32 next_target_id; u32 next_expander_id; + int next_port_id; }; #define to_sas_host_attrs(host) ((struct sas_host_attrs *)(host)->shost_data) @@ -146,6 +147,7 @@ static int sas_host_setup(struct transpo mutex_init(&sas_host->lock); sas_host->next_target_id = 0; sas_host->next_expander_id = 0; + sas_host->next_port_id = 0; return 0; } @@ -327,7 +329,7 @@ sas_phy_protocol_attr(identify.target_po sas_phy_simple_attr(identify.sas_address, sas_address, "0x%016llx\n", unsigned long long); sas_phy_simple_attr(identify.phy_identifier, phy_identifier, "%d\n", u8); -//sas_phy_simple_attr(port_identifier, port_identifier, "%d\n", u8); +//sas_phy_simple_attr(port_identifier, port_identifier, "%d\n", int); sas_phy_linkspeed_attr(negotiated_linkrate); sas_phy_linkspeed_attr(minimum_linkrate_hw); sas_phy_linkspeed_attr(minimum_linkrate); @@ -590,6 +592,38 @@ struct sas_port *sas_port_alloc(struct d } EXPORT_SYMBOL(sas_port_alloc); +/** sas_port_alloc_num - allocate and initialize a SAS port structure + * + * @parent: parent device + * + * Allocates a SAS port structure and a number to go with it. This + * interface is really for adapters where the port number has no + * meansing, so the sas class should manage them. It will be added to + * the device tree below the device specified by @parent which must be + * either a Scsi_Host or a sas_expander_device. + * + * Returns %NULL on error + */ +struct sas_port *sas_port_alloc_num(struct device *parent) +{ + int index; + struct Scsi_Host *shost = dev_to_shost(parent); + struct sas_host_attrs *sas_host = to_sas_host_attrs(shost); + + /* FIXME: use idr for this eventually */ + mutex_lock(&sas_host->lock); + if (scsi_is_sas_expander_device(parent)) { + struct sas_rphy *rphy = dev_to_rphy(parent); + struct sas_expander_device *exp = rphy_to_expander_device(rphy); + + index = exp->next_port_id++; + } else + index = sas_host->next_port_id++; + mutex_unlock(&sas_host->lock); + return sas_port_alloc(parent, index); +} +EXPORT_SYMBOL(sas_port_alloc_num); + /** * sas_port_add - add a SAS port to the device hierarchy * diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index 3225d31..98bd3aa 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -502,8 +502,7 @@ static int sd_init_command(struct scsi_c SCpnt->cmnd[4] = (unsigned char) this_count; SCpnt->cmnd[5] = 0; } - SCpnt->request_bufflen = SCpnt->bufflen = - this_count * sdp->sector_size; + SCpnt->request_bufflen = this_count * sdp->sector_size; /* * We shouldn't disconnect in the middle of a sector, so with a dumb diff --git a/drivers/scsi/sr.c b/drivers/scsi/sr.c index fd94408..fae6e95 100644 --- a/drivers/scsi/sr.c +++ b/drivers/scsi/sr.c @@ -360,7 +360,7 @@ static int sr_init_command(struct scsi_c "mismatch count %d, bytes %d\n", size, SCpnt->request_bufflen); if (SCpnt->request_bufflen > size) - SCpnt->request_bufflen = SCpnt->bufflen = size; + SCpnt->request_bufflen = size; } } @@ -387,8 +387,7 @@ static int sr_init_command(struct scsi_c if (this_count > 0xffff) { this_count = 0xffff; - SCpnt->request_bufflen = SCpnt->bufflen = - this_count * s_size; + SCpnt->request_bufflen = this_count * s_size; } SCpnt->cmnd[2] = (unsigned char) (block >> 24) & 0xff; diff --git a/drivers/scsi/st.c b/drivers/scsi/st.c index 756ceb9..7f669b6 100644 --- a/drivers/scsi/st.c +++ b/drivers/scsi/st.c @@ -368,7 +368,7 @@ static int st_chk_result(struct scsi_tap SRpnt->cmd[0], SRpnt->cmd[1], SRpnt->cmd[2], SRpnt->cmd[3], SRpnt->cmd[4], SRpnt->cmd[5]); if (cmdstatp->have_sense) - __scsi_print_sense("st", SRpnt->sense, SCSI_SENSE_BUFFERSIZE); + __scsi_print_sense(name, SRpnt->sense, SCSI_SENSE_BUFFERSIZE); } ) /* end DEB */ if (!debugging) { /* Abnormal conditions for tape */ if (!cmdstatp->have_sense) @@ -384,9 +384,8 @@ static int st_chk_result(struct scsi_tap scode != VOLUME_OVERFLOW && SRpnt->cmd[0] != MODE_SENSE && SRpnt->cmd[0] != TEST_UNIT_READY) { - printk(KERN_WARNING "%s: Error with sense data: ", name); - __scsi_print_sense("st", SRpnt->sense, - SCSI_SENSE_BUFFERSIZE); + + __scsi_print_sense(name, SRpnt->sense, SCSI_SENSE_BUFFERSIZE); } } diff --git a/include/scsi/scsi_cmnd.h b/include/scsi/scsi_cmnd.h index 371f70d..58e6444 100644 --- a/include/scsi/scsi_cmnd.h +++ b/include/scsi/scsi_cmnd.h @@ -58,9 +58,7 @@ struct scsi_cmnd { int timeout_per_command; unsigned char cmd_len; - unsigned char old_cmd_len; enum dma_data_direction sc_data_direction; - enum dma_data_direction sc_old_data_direction; /* These elements define the operation we are about to perform */ #define MAX_COMMAND_SIZE 16 @@ -71,18 +69,11 @@ #define MAX_COMMAND_SIZE 16 void *request_buffer; /* Actual requested buffer */ /* These elements define the operation we ultimately want to perform */ - unsigned char data_cmnd[MAX_COMMAND_SIZE]; - unsigned short old_use_sg; /* We save use_sg here when requesting - * sense info */ unsigned short use_sg; /* Number of pieces of scatter-gather */ unsigned short sglist_len; /* size of malloc'd scatter-gather list */ - unsigned bufflen; /* Size of data buffer */ - void *buffer; /* Data buffer */ unsigned underflow; /* Return error if less than this amount is transferred */ - unsigned old_underflow; /* save underflow here when reusing the - * command for error handling */ unsigned transfersize; /* How much we are guaranteed to transfer with each SCSI transfer diff --git a/include/scsi/scsi_transport_sas.h b/include/scsi/scsi_transport_sas.h index e3c503c..f5772ff 100644 --- a/include/scsi/scsi_transport_sas.h +++ b/include/scsi/scsi_transport_sas.h @@ -106,6 +106,7 @@ #define rphy_to_end_device(r) \ struct sas_expander_device { int level; + int next_port_id; #define SAS_EXPANDER_VENDOR_ID_LEN 8 char vendor_id[SAS_EXPANDER_VENDOR_ID_LEN+1]; @@ -127,7 +128,7 @@ #define rphy_to_expander_device(r) \ struct sas_port { struct device dev; - u8 port_identifier; + int port_identifier; int num_phys; /* the other end of the link */ @@ -168,6 +169,7 @@ extern void sas_rphy_delete(struct sas_r extern int scsi_is_sas_rphy(const struct device *); struct sas_port *sas_port_alloc(struct device *, int); +struct sas_port *sas_port_alloc_num(struct device *); int sas_port_add(struct sas_port *); void sas_port_free(struct sas_port *); void sas_port_delete(struct sas_port *);