OpenCloudOS-Kernel/drivers/scsi/lpfc/lpfc_mbox.c

2660 lines
87 KiB
C
Raw Normal View History

/*******************************************************************
* This file is part of the Emulex Linux Device Driver for *
* Fibre Channel Host Bus Adapters. *
* Copyright (C) 2017-2018 Broadcom. All Rights Reserved. The term *
* Broadcom refers to Broadcom Inc. and/or its subsidiaries. *
* Copyright (C) 2004-2016 Emulex. All rights reserved. *
* EMULEX and SLI are trademarks of Emulex. *
* www.broadcom.com *
* Portions Copyright (C) 2004-2005 Christoph Hellwig *
* *
* This program is free software; you can redistribute it and/or *
* modify it under the terms of version 2 of the GNU General *
* Public License as published by the Free Software Foundation. *
* This program is distributed in the hope that it will be useful. *
* ALL EXPRESS OR IMPLIED CONDITIONS, REPRESENTATIONS AND *
* WARRANTIES, INCLUDING ANY IMPLIED WARRANTY OF MERCHANTABILITY, *
* FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT, ARE *
* DISCLAIMED, EXCEPT TO THE EXTENT THAT SUCH DISCLAIMERS ARE HELD *
* TO BE LEGALLY INVALID. See the GNU General Public License for *
* more details, a copy of which can be found in the file COPYING *
* included with this package. *
*******************************************************************/
#include <linux/blkdev.h>
#include <linux/pci.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/slab.h>
#include <linux/interrupt.h>
#include <scsi/scsi_device.h>
#include <scsi/scsi_transport_fc.h>
#include <scsi/scsi.h>
#include <scsi/fc/fc_fs.h>
#include "lpfc_hw4.h"
#include "lpfc_hw.h"
#include "lpfc_sli.h"
#include "lpfc_sli4.h"
#include "lpfc_nl.h"
#include "lpfc_disc.h"
#include "lpfc_scsi.h"
#include "lpfc.h"
#include "lpfc_logmsg.h"
#include "lpfc_crtn.h"
#include "lpfc_compat.h"
/**
* lpfc_dump_static_vport - Dump HBA's static vport information.
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
* @offset: offset for dumping vport info.
*
* The dump mailbox command provides a method for the device driver to obtain
* various types of information from the HBA device.
*
* This routine prepares the mailbox command for dumping list of static
* vports to be created.
**/
int
lpfc_dump_static_vport(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb,
uint16_t offset)
{
MAILBOX_t *mb;
struct lpfc_dmabuf *mp;
mb = &pmb->u.mb;
/* Setup to dump vport info region */
memset(pmb, 0, sizeof(LPFC_MBOXQ_t));
mb->mbxCommand = MBX_DUMP_MEMORY;
mb->un.varDmp.type = DMP_NV_PARAMS;
mb->un.varDmp.entry_index = offset;
mb->un.varDmp.region_id = DMP_REGION_VPORT;
mb->mbxOwner = OWN_HOST;
/* For SLI3 HBAs data is embedded in mailbox */
if (phba->sli_rev != LPFC_SLI_REV4) {
mb->un.varDmp.cv = 1;
mb->un.varDmp.word_cnt = DMP_RSP_SIZE/sizeof(uint32_t);
return 0;
}
/* For SLI4 HBAs driver need to allocate memory */
mp = kmalloc(sizeof(struct lpfc_dmabuf), GFP_KERNEL);
if (mp)
mp->virt = lpfc_mbuf_alloc(phba, 0, &mp->phys);
if (!mp || !mp->virt) {
kfree(mp);
lpfc_printf_log(phba, KERN_ERR, LOG_MBOX,
"2605 lpfc_dump_static_vport: memory"
" allocation failed\n");
return 1;
}
memset(mp->virt, 0, LPFC_BPL_SIZE);
INIT_LIST_HEAD(&mp->list);
/* save address for completion */
pmb->ctx_buf = (uint8_t *)mp;
mb->un.varWords[3] = putPaddrLow(mp->phys);
mb->un.varWords[4] = putPaddrHigh(mp->phys);
mb->un.varDmp.sli4_length = sizeof(struct static_vport_info);
return 0;
}
/**
* lpfc_down_link - Bring down HBAs link.
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* This routine prepares a mailbox command to bring down HBA link.
**/
void
lpfc_down_link(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
{
MAILBOX_t *mb;
memset(pmb, 0, sizeof(LPFC_MBOXQ_t));
mb = &pmb->u.mb;
mb->mbxCommand = MBX_DOWN_LINK;
mb->mbxOwner = OWN_HOST;
}
/**
* lpfc_dump_mem - Prepare a mailbox command for reading a region.
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
* @offset: offset into the region.
* @region_id: config region id.
*
* The dump mailbox command provides a method for the device driver to obtain
* various types of information from the HBA device.
*
* This routine prepares the mailbox command for dumping HBA's config region.
**/
void
lpfc_dump_mem(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb, uint16_t offset,
uint16_t region_id)
{
MAILBOX_t *mb;
void *ctx;
mb = &pmb->u.mb;
ctx = pmb->ctx_buf;
/* Setup to dump VPD region */
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->mbxCommand = MBX_DUMP_MEMORY;
mb->un.varDmp.cv = 1;
mb->un.varDmp.type = DMP_NV_PARAMS;
mb->un.varDmp.entry_index = offset;
mb->un.varDmp.region_id = region_id;
mb->un.varDmp.word_cnt = (DMP_RSP_SIZE / sizeof (uint32_t));
mb->un.varDmp.co = 0;
mb->un.varDmp.resp_offset = 0;
pmb->ctx_buf = ctx;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_dump_wakeup_param - Prepare mailbox command for retrieving wakeup params
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* This function create a dump memory mailbox command to dump wake up
* parameters.
*/
void
lpfc_dump_wakeup_param(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
{
MAILBOX_t *mb;
void *ctx;
mb = &pmb->u.mb;
/* Save context so that we can restore after memset */
ctx = pmb->ctx_buf;
/* Setup to dump VPD region */
memset(pmb, 0, sizeof(LPFC_MBOXQ_t));
mb->mbxCommand = MBX_DUMP_MEMORY;
mb->mbxOwner = OWN_HOST;
mb->un.varDmp.cv = 1;
mb->un.varDmp.type = DMP_NV_PARAMS;
if (phba->sli_rev < LPFC_SLI_REV4)
mb->un.varDmp.entry_index = 0;
mb->un.varDmp.region_id = WAKE_UP_PARMS_REGION_ID;
mb->un.varDmp.word_cnt = WAKE_UP_PARMS_WORD_SIZE;
mb->un.varDmp.co = 0;
mb->un.varDmp.resp_offset = 0;
pmb->ctx_buf = ctx;
return;
}
/**
* lpfc_read_nv - Prepare a mailbox command for reading HBA's NVRAM param
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The read NVRAM mailbox command returns the HBA's non-volatile parameters
* that are used as defaults when the Fibre Channel link is brought on-line.
*
* This routine prepares the mailbox command for reading information stored
* in the HBA's NVRAM. Specifically, the HBA's WWNN and WWPN.
**/
void
lpfc_read_nv(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
{
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->mbxCommand = MBX_READ_NV;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_config_async - Prepare a mailbox command for enabling HBA async event
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
* @ring: ring number for the asynchronous event to be configured.
*
* The asynchronous event enable mailbox command is used to enable the
* asynchronous event posting via the ASYNC_STATUS_CN IOCB response and
* specifies the default ring to which events are posted.
*
* This routine prepares the mailbox command for enabling HBA asynchronous
* event support on a IOCB ring.
**/
void
lpfc_config_async(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb,
uint32_t ring)
{
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->mbxCommand = MBX_ASYNCEVT_ENABLE;
mb->un.varCfgAsyncEvent.ring = ring;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_heart_beat - Prepare a mailbox command for heart beat
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The heart beat mailbox command is used to detect an unresponsive HBA, which
* is defined as any device where no error attention is sent and both mailbox
* and rings are not processed.
*
* This routine prepares the mailbox command for issuing a heart beat in the
* form of mailbox command to the HBA. The timely completion of the heart
* beat mailbox command indicates the health of the HBA.
**/
void
lpfc_heart_beat(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
{
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->mbxCommand = MBX_HEARTBEAT;
mb->mbxOwner = OWN_HOST;
return;
}
/**
2010-11-21 12:11:48 +08:00
* lpfc_read_topology - Prepare a mailbox command for reading HBA topology
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
* @mp: DMA buffer memory for reading the link attention information into.
*
2010-11-21 12:11:48 +08:00
* The read topology mailbox command is issued to read the link topology
* information indicated by the HBA port when the Link Event bit of the Host
* Attention (HSTATT) register is set to 1 (For SLI-3) or when an FC Link
* Attention ACQE is received from the port (For SLI-4). A Link Event
* Attention occurs based on an exception detected at the Fibre Channel link
* interface.
*
2010-11-21 12:11:48 +08:00
* This routine prepares the mailbox command for reading HBA link topology
* information. A DMA memory has been set aside and address passed to the
* HBA through @mp for the HBA to DMA link attention information into the
* memory as part of the execution of the mailbox command.
*
* Return codes
* 0 - Success (currently always return 0)
**/
int
2010-11-21 12:11:48 +08:00
lpfc_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb,
struct lpfc_dmabuf *mp)
{
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
INIT_LIST_HEAD(&mp->list);
2010-11-21 12:11:48 +08:00
mb->mbxCommand = MBX_READ_TOPOLOGY;
mb->un.varReadTop.lilpBde64.tus.f.bdeSize = LPFC_ALPA_MAP_SIZE;
mb->un.varReadTop.lilpBde64.addrHigh = putPaddrHigh(mp->phys);
mb->un.varReadTop.lilpBde64.addrLow = putPaddrLow(mp->phys);
/* Save address for later completion and set the owner to host so that
* the FW knows this mailbox is available for processing.
*/
pmb->ctx_buf = (uint8_t *)mp;
mb->mbxOwner = OWN_HOST;
return (0);
}
/**
* lpfc_clear_la - Prepare a mailbox command for clearing HBA link attention
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The clear link attention mailbox command is issued to clear the link event
* attention condition indicated by the Link Event bit of the Host Attention
* (HSTATT) register. The link event attention condition is cleared only if
* the event tag specified matches that of the current link event counter.
* The current event tag is read using the read link attention event mailbox
* command.
*
* This routine prepares the mailbox command for clearing HBA link attention
* information.
**/
void
lpfc_clear_la(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
{
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->un.varClearLA.eventTag = phba->fc_eventTag;
mb->mbxCommand = MBX_CLEAR_LA;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_config_link - Prepare a mailbox command for configuring link on a HBA
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The configure link mailbox command is used before the initialize link
* mailbox command to override default value and to configure link-oriented
* parameters such as DID address and various timers. Typically, this
* command would be used after an F_Port login to set the returned DID address
* and the fabric timeout values. This command is not valid before a configure
* port command has configured the HBA port.
*
* This routine prepares the mailbox command for configuring link on a HBA.
**/
void
lpfc_config_link(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
{
struct lpfc_vport *vport = phba->pport;
MAILBOX_t *mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
/* NEW_FEATURE
* SLI-2, Coalescing Response Feature.
*/
if (phba->cfg_cr_delay && (phba->sli_rev < LPFC_SLI_REV4)) {
mb->un.varCfgLnk.cr = 1;
mb->un.varCfgLnk.ci = 1;
mb->un.varCfgLnk.cr_delay = phba->cfg_cr_delay;
mb->un.varCfgLnk.cr_count = phba->cfg_cr_count;
}
mb->un.varCfgLnk.myId = vport->fc_myDID;
mb->un.varCfgLnk.edtov = phba->fc_edtov;
mb->un.varCfgLnk.arbtov = phba->fc_arbtov;
mb->un.varCfgLnk.ratov = phba->fc_ratov;
mb->un.varCfgLnk.rttov = phba->fc_rttov;
mb->un.varCfgLnk.altov = phba->fc_altov;
mb->un.varCfgLnk.crtov = phba->fc_crtov;
mb->un.varCfgLnk.cscn = 0;
if (phba->bbcredit_support && phba->cfg_enable_bbcr) {
mb->un.varCfgLnk.cscn = 1;
mb->un.varCfgLnk.bbscn = bf_get(lpfc_bbscn_def,
&phba->sli4_hba.bbscn_params);
}
if (phba->cfg_ack0 && (phba->sli_rev < LPFC_SLI_REV4))
mb->un.varCfgLnk.ack0_enable = 1;
mb->mbxCommand = MBX_CONFIG_LINK;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_config_msi - Prepare a mailbox command for configuring msi-x
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The configure MSI-X mailbox command is used to configure the HBA's SLI-3
* MSI-X multi-message interrupt vector association to interrupt attention
* conditions.
*
* Return codes
* 0 - Success
* -EINVAL - Failure
**/
int
lpfc_config_msi(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
{
MAILBOX_t *mb = &pmb->u.mb;
uint32_t attentionConditions[2];
/* Sanity check */
if (phba->cfg_use_msi != 2) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
"0475 Not configured for supporting MSI-X "
"cfg_use_msi: 0x%x\n", phba->cfg_use_msi);
return -EINVAL;
}
if (phba->sli_rev < 3) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
"0476 HBA not supporting SLI-3 or later "
"SLI Revision: 0x%x\n", phba->sli_rev);
return -EINVAL;
}
/* Clear mailbox command fields */
memset(pmb, 0, sizeof(LPFC_MBOXQ_t));
/*
* SLI-3, Message Signaled Interrupt Fearure.
*/
/* Multi-message attention configuration */
attentionConditions[0] = (HA_R0ATT | HA_R1ATT | HA_R2ATT | HA_ERATT |
HA_LATT | HA_MBATT);
attentionConditions[1] = 0;
mb->un.varCfgMSI.attentionConditions[0] = attentionConditions[0];
mb->un.varCfgMSI.attentionConditions[1] = attentionConditions[1];
/*
* Set up message number to HA bit association
*/
#ifdef __BIG_ENDIAN_BITFIELD
/* RA0 (FCP Ring) */
mb->un.varCfgMSI.messageNumberByHA[HA_R0_POS] = 1;
/* RA1 (Other Protocol Extra Ring) */
mb->un.varCfgMSI.messageNumberByHA[HA_R1_POS] = 1;
#else /* __LITTLE_ENDIAN_BITFIELD */
/* RA0 (FCP Ring) */
mb->un.varCfgMSI.messageNumberByHA[HA_R0_POS^3] = 1;
/* RA1 (Other Protocol Extra Ring) */
mb->un.varCfgMSI.messageNumberByHA[HA_R1_POS^3] = 1;
#endif
/* Multi-message interrupt autoclear configuration*/
mb->un.varCfgMSI.autoClearHA[0] = attentionConditions[0];
mb->un.varCfgMSI.autoClearHA[1] = attentionConditions[1];
/* For now, HBA autoclear does not work reliably, disable it */
mb->un.varCfgMSI.autoClearHA[0] = 0;
mb->un.varCfgMSI.autoClearHA[1] = 0;
/* Set command and owner bit */
mb->mbxCommand = MBX_CONFIG_MSI;
mb->mbxOwner = OWN_HOST;
return 0;
}
/**
* lpfc_init_link - Prepare a mailbox command for initialize link on a HBA
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
* @topology: the link topology for the link to be initialized to.
* @linkspeed: the link speed for the link to be initialized to.
*
* The initialize link mailbox command is used to initialize the Fibre
* Channel link. This command must follow a configure port command that
* establishes the mode of operation.
*
* This routine prepares the mailbox command for initializing link on a HBA
* with the specified link topology and speed.
**/
void
lpfc_init_link(struct lpfc_hba * phba,
LPFC_MBOXQ_t * pmb, uint32_t topology, uint32_t linkspeed)
{
lpfc_vpd_t *vpd;
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
switch (topology) {
case FLAGS_TOPOLOGY_MODE_LOOP_PT:
mb->un.varInitLnk.link_flags = FLAGS_TOPOLOGY_MODE_LOOP;
mb->un.varInitLnk.link_flags |= FLAGS_TOPOLOGY_FAILOVER;
break;
case FLAGS_TOPOLOGY_MODE_PT_PT:
mb->un.varInitLnk.link_flags = FLAGS_TOPOLOGY_MODE_PT_PT;
break;
case FLAGS_TOPOLOGY_MODE_LOOP:
mb->un.varInitLnk.link_flags = FLAGS_TOPOLOGY_MODE_LOOP;
break;
case FLAGS_TOPOLOGY_MODE_PT_LOOP:
mb->un.varInitLnk.link_flags = FLAGS_TOPOLOGY_MODE_PT_PT;
mb->un.varInitLnk.link_flags |= FLAGS_TOPOLOGY_FAILOVER;
break;
case FLAGS_LOCAL_LB:
mb->un.varInitLnk.link_flags = FLAGS_LOCAL_LB;
break;
}
if ((phba->pcidev->device == PCI_DEVICE_ID_LANCER_G6_FC ||
phba->pcidev->device == PCI_DEVICE_ID_LANCER_G7_FC) &&
!(phba->sli4_hba.pc_sli4_params.pls) &&
mb->un.varInitLnk.link_flags & FLAGS_TOPOLOGY_MODE_LOOP) {
mb->un.varInitLnk.link_flags = FLAGS_TOPOLOGY_MODE_PT_PT;
phba->cfg_topology = FLAGS_TOPOLOGY_MODE_PT_PT;
}
/* Enable asynchronous ABTS responses from firmware */
mb->un.varInitLnk.link_flags |= FLAGS_IMED_ABORT;
/* NEW_FEATURE
* Setting up the link speed
*/
vpd = &phba->vpd;
if (vpd->rev.feaLevelHigh >= 0x02){
switch(linkspeed){
2010-11-21 12:11:48 +08:00
case LPFC_USER_LINK_SPEED_1G:
mb->un.varInitLnk.link_flags |= FLAGS_LINK_SPEED;
mb->un.varInitLnk.link_speed = LINK_SPEED_1G;
break;
case LPFC_USER_LINK_SPEED_2G:
mb->un.varInitLnk.link_flags |= FLAGS_LINK_SPEED;
mb->un.varInitLnk.link_speed = LINK_SPEED_2G;
break;
case LPFC_USER_LINK_SPEED_4G:
mb->un.varInitLnk.link_flags |= FLAGS_LINK_SPEED;
mb->un.varInitLnk.link_speed = LINK_SPEED_4G;
break;
case LPFC_USER_LINK_SPEED_8G:
mb->un.varInitLnk.link_flags |= FLAGS_LINK_SPEED;
mb->un.varInitLnk.link_speed = LINK_SPEED_8G;
break;
case LPFC_USER_LINK_SPEED_10G:
mb->un.varInitLnk.link_flags |= FLAGS_LINK_SPEED;
mb->un.varInitLnk.link_speed = LINK_SPEED_10G;
break;
case LPFC_USER_LINK_SPEED_16G:
mb->un.varInitLnk.link_flags |= FLAGS_LINK_SPEED;
mb->un.varInitLnk.link_speed = LINK_SPEED_16G;
break;
case LPFC_USER_LINK_SPEED_32G:
mb->un.varInitLnk.link_flags |= FLAGS_LINK_SPEED;
mb->un.varInitLnk.link_speed = LINK_SPEED_32G;
break;
case LPFC_USER_LINK_SPEED_64G:
mb->un.varInitLnk.link_flags |= FLAGS_LINK_SPEED;
mb->un.varInitLnk.link_speed = LINK_SPEED_64G;
break;
2010-11-21 12:11:48 +08:00
case LPFC_USER_LINK_SPEED_AUTO:
default:
mb->un.varInitLnk.link_speed = LINK_SPEED_AUTO;
break;
}
}
else
mb->un.varInitLnk.link_speed = LINK_SPEED_AUTO;
mb->mbxCommand = (volatile uint8_t)MBX_INIT_LINK;
mb->mbxOwner = OWN_HOST;
mb->un.varInitLnk.fabric_AL_PA = phba->fc_pref_ALPA;
return;
}
/**
* lpfc_read_sparam - Prepare a mailbox command for reading HBA parameters
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
* @vpi: virtual N_Port identifier.
*
* The read service parameter mailbox command is used to read the HBA port
* service parameters. The service parameters are read into the buffer
* specified directly by a BDE in the mailbox command. These service
* parameters may then be used to build the payload of an N_Port/F_POrt
* login request and reply (LOGI/ACC).
*
* This routine prepares the mailbox command for reading HBA port service
* parameters. The DMA memory is allocated in this function and the addresses
* are populated into the mailbox command for the HBA to DMA the service
* parameters into.
*
* Return codes
* 0 - Success
* 1 - DMA memory allocation failed
**/
int
lpfc_read_sparam(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb, int vpi)
{
struct lpfc_dmabuf *mp;
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->mbxOwner = OWN_HOST;
/* Get a buffer to hold the HBAs Service Parameters */
mp = kmalloc(sizeof (struct lpfc_dmabuf), GFP_KERNEL);
if (mp)
mp->virt = lpfc_mbuf_alloc(phba, 0, &mp->phys);
if (!mp || !mp->virt) {
kfree(mp);
mb->mbxCommand = MBX_READ_SPARM64;
/* READ_SPARAM: no buffers */
lpfc_printf_log(phba, KERN_WARNING, LOG_MBOX,
"0301 READ_SPARAM: no buffers\n");
return (1);
}
INIT_LIST_HEAD(&mp->list);
mb->mbxCommand = MBX_READ_SPARM64;
mb->un.varRdSparm.un.sp64.tus.f.bdeSize = sizeof (struct serv_parm);
mb->un.varRdSparm.un.sp64.addrHigh = putPaddrHigh(mp->phys);
mb->un.varRdSparm.un.sp64.addrLow = putPaddrLow(mp->phys);
if (phba->sli_rev >= LPFC_SLI_REV3)
mb->un.varRdSparm.vpi = phba->vpi_ids[vpi];
/* save address for completion */
pmb->ctx_buf = mp;
return (0);
}
/**
* lpfc_unreg_did - Prepare a mailbox command for unregistering DID
* @phba: pointer to lpfc hba data structure.
* @vpi: virtual N_Port identifier.
* @did: remote port identifier.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The unregister DID mailbox command is used to unregister an N_Port/F_Port
* login for an unknown RPI by specifying the DID of a remote port. This
* command frees an RPI context in the HBA port. This has the effect of
* performing an implicit N_Port/F_Port logout.
*
* This routine prepares the mailbox command for unregistering a remote
* N_Port/F_Port (DID) login.
**/
void
lpfc_unreg_did(struct lpfc_hba * phba, uint16_t vpi, uint32_t did,
LPFC_MBOXQ_t * pmb)
{
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->un.varUnregDID.did = did;
mb->un.varUnregDID.vpi = vpi;
if ((vpi != 0xffff) &&
(phba->sli_rev == LPFC_SLI_REV4))
mb->un.varUnregDID.vpi = phba->vpi_ids[vpi];
mb->mbxCommand = MBX_UNREG_D_ID;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_read_config - Prepare a mailbox command for reading HBA configuration
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The read configuration mailbox command is used to read the HBA port
* configuration parameters. This mailbox command provides a method for
* seeing any parameters that may have changed via various configuration
* mailbox commands.
*
* This routine prepares the mailbox command for reading out HBA configuration
* parameters.
**/
void
lpfc_read_config(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
{
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->mbxCommand = MBX_READ_CONFIG;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_read_lnk_stat - Prepare a mailbox command for reading HBA link stats
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The read link status mailbox command is used to read the link status from
* the HBA. Link status includes all link-related error counters. These
* counters are maintained by the HBA and originated in the link hardware
* unit. Note that all of these counters wrap.
*
* This routine prepares the mailbox command for reading out HBA link status.
**/
void
lpfc_read_lnk_stat(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
{
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->mbxCommand = MBX_READ_LNK_STAT;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_reg_rpi - Prepare a mailbox command for registering remote login
* @phba: pointer to lpfc hba data structure.
* @vpi: virtual N_Port identifier.
* @did: remote port identifier.
* @param: pointer to memory holding the server parameters.
* @pmb: pointer to the driver internal queue element for mailbox command.
* @rpi: the rpi to use in the registration (usually only used for SLI4.
*
* The registration login mailbox command is used to register an N_Port or
* F_Port login. This registration allows the HBA to cache the remote N_Port
* service parameters internally and thereby make the appropriate FC-2
* decisions. The remote port service parameters are handed off by the driver
* to the HBA using a descriptor entry that directly identifies a buffer in
* host memory. In exchange, the HBA returns an RPI identifier.
*
* This routine prepares the mailbox command for registering remote port login.
* The function allocates DMA buffer for passing the service parameters to the
* HBA with the mailbox command.
*
* Return codes
* 0 - Success
* 1 - DMA memory allocation failed
**/
int
lpfc_reg_rpi(struct lpfc_hba *phba, uint16_t vpi, uint32_t did,
uint8_t *param, LPFC_MBOXQ_t *pmb, uint16_t rpi)
{
MAILBOX_t *mb = &pmb->u.mb;
uint8_t *sparam;
struct lpfc_dmabuf *mp;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->un.varRegLogin.rpi = 0;
if (phba->sli_rev == LPFC_SLI_REV4)
mb->un.varRegLogin.rpi = phba->sli4_hba.rpi_ids[rpi];
if (phba->sli_rev >= LPFC_SLI_REV3)
mb->un.varRegLogin.vpi = phba->vpi_ids[vpi];
mb->un.varRegLogin.did = did;
mb->mbxOwner = OWN_HOST;
/* Get a buffer to hold NPorts Service Parameters */
mp = kmalloc(sizeof (struct lpfc_dmabuf), GFP_KERNEL);
if (mp)
mp->virt = lpfc_mbuf_alloc(phba, 0, &mp->phys);
if (!mp || !mp->virt) {
kfree(mp);
mb->mbxCommand = MBX_REG_LOGIN64;
/* REG_LOGIN: no buffers */
lpfc_printf_log(phba, KERN_WARNING, LOG_MBOX,
"0302 REG_LOGIN: no buffers, VPI:%d DID:x%x, "
"rpi x%x\n", vpi, did, rpi);
return 1;
}
INIT_LIST_HEAD(&mp->list);
sparam = mp->virt;
/* Copy param's into a new buffer */
memcpy(sparam, param, sizeof (struct serv_parm));
/* save address for completion */
pmb->ctx_buf = (uint8_t *)mp;
mb->mbxCommand = MBX_REG_LOGIN64;
mb->un.varRegLogin.un.sp64.tus.f.bdeSize = sizeof (struct serv_parm);
mb->un.varRegLogin.un.sp64.addrHigh = putPaddrHigh(mp->phys);
mb->un.varRegLogin.un.sp64.addrLow = putPaddrLow(mp->phys);
return 0;
}
/**
* lpfc_unreg_login - Prepare a mailbox command for unregistering remote login
* @phba: pointer to lpfc hba data structure.
* @vpi: virtual N_Port identifier.
* @rpi: remote port identifier
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The unregistration login mailbox command is used to unregister an N_Port
* or F_Port login. This command frees an RPI context in the HBA. It has the
* effect of performing an implicit N_Port/F_Port logout.
*
* This routine prepares the mailbox command for unregistering remote port
* login.
*
* For SLI4 ports, the rpi passed to this function must be the physical
* rpi value, not the logical index.
**/
void
lpfc_unreg_login(struct lpfc_hba *phba, uint16_t vpi, uint32_t rpi,
LPFC_MBOXQ_t * pmb)
{
MAILBOX_t *mb;
mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->un.varUnregLogin.rpi = rpi;
mb->un.varUnregLogin.rsvd1 = 0;
if (phba->sli_rev >= LPFC_SLI_REV3)
mb->un.varUnregLogin.vpi = phba->vpi_ids[vpi];
mb->mbxCommand = MBX_UNREG_LOGIN;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_sli4_unreg_all_rpis - unregister all RPIs for a vport on SLI4 HBA.
* @vport: pointer to a vport object.
*
* This routine sends mailbox command to unregister all active RPIs for
* a vport.
**/
void
lpfc_sli4_unreg_all_rpis(struct lpfc_vport *vport)
{
struct lpfc_hba *phba = vport->phba;
LPFC_MBOXQ_t *mbox;
int rc;
mbox = mempool_alloc(phba->mbox_mem_pool, GFP_KERNEL);
if (mbox) {
/*
* For SLI4 functions, the rpi field is overloaded for
* the vport context unreg all. This routine passes
* 0 for the rpi field in lpfc_unreg_login for compatibility
* with SLI3 and then overrides the rpi field with the
* expected value for SLI4.
*/
lpfc_unreg_login(phba, vport->vpi, phba->vpi_ids[vport->vpi],
mbox);
mbox->u.mb.un.varUnregLogin.rsvd1 = 0x4000;
mbox->vport = vport;
mbox->mbox_cmpl = lpfc_sli_def_mbox_cmpl;
mbox->ctx_ndlp = NULL;
rc = lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT);
if (rc == MBX_NOT_FINISHED)
mempool_free(mbox, phba->mbox_mem_pool);
}
}
/**
* lpfc_reg_vpi - Prepare a mailbox command for registering vport identifier
* @phba: pointer to lpfc hba data structure.
* @vpi: virtual N_Port identifier.
* @sid: Fibre Channel S_ID (N_Port_ID assigned to a virtual N_Port).
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The registration vport identifier mailbox command is used to activate a
* virtual N_Port after it has acquired an N_Port_ID. The HBA validates the
* N_Port_ID against the information in the selected virtual N_Port context
* block and marks it active to allow normal processing of IOCB commands and
* received unsolicited exchanges.
*
* This routine prepares the mailbox command for registering a virtual N_Port.
**/
void
lpfc_reg_vpi(struct lpfc_vport *vport, LPFC_MBOXQ_t *pmb)
{
MAILBOX_t *mb = &pmb->u.mb;
struct lpfc_hba *phba = vport->phba;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
/*
* Set the re-reg VPI bit for f/w to update the MAC address.
*/
if ((phba->sli_rev == LPFC_SLI_REV4) &&
!(vport->fc_flag & FC_VPORT_NEEDS_REG_VPI))
mb->un.varRegVpi.upd = 1;
mb->un.varRegVpi.vpi = phba->vpi_ids[vport->vpi];
mb->un.varRegVpi.sid = vport->fc_myDID;
if (phba->sli_rev == LPFC_SLI_REV4)
mb->un.varRegVpi.vfi = phba->sli4_hba.vfi_ids[vport->vfi];
else
mb->un.varRegVpi.vfi = vport->vfi + vport->phba->vfi_base;
memcpy(mb->un.varRegVpi.wwn, &vport->fc_portname,
sizeof(struct lpfc_name));
mb->un.varRegVpi.wwn[0] = cpu_to_le32(mb->un.varRegVpi.wwn[0]);
mb->un.varRegVpi.wwn[1] = cpu_to_le32(mb->un.varRegVpi.wwn[1]);
mb->mbxCommand = MBX_REG_VPI;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_unreg_vpi - Prepare a mailbox command for unregistering vport id
* @phba: pointer to lpfc hba data structure.
* @vpi: virtual N_Port identifier.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The unregistration vport identifier mailbox command is used to inactivate
* a virtual N_Port. The driver must have logged out and unregistered all
* remote N_Ports to abort any activity on the virtual N_Port. The HBA will
* unregisters any default RPIs associated with the specified vpi, aborting
* any active exchanges. The HBA will post the mailbox response after making
* the virtual N_Port inactive.
*
* This routine prepares the mailbox command for unregistering a virtual
* N_Port.
**/
void
lpfc_unreg_vpi(struct lpfc_hba *phba, uint16_t vpi, LPFC_MBOXQ_t *pmb)
{
MAILBOX_t *mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
if (phba->sli_rev == LPFC_SLI_REV3)
mb->un.varUnregVpi.vpi = phba->vpi_ids[vpi];
else if (phba->sli_rev >= LPFC_SLI_REV4)
mb->un.varUnregVpi.sli4_vpi = phba->vpi_ids[vpi];
mb->mbxCommand = MBX_UNREG_VPI;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_config_pcb_setup - Set up IOCB rings in the Port Control Block (PCB)
* @phba: pointer to lpfc hba data structure.
*
* This routine sets up and initializes the IOCB rings in the Port Control
* Block (PCB).
**/
static void
lpfc_config_pcb_setup(struct lpfc_hba * phba)
{
struct lpfc_sli *psli = &phba->sli;
struct lpfc_sli_ring *pring;
PCB_t *pcbp = phba->pcb;
dma_addr_t pdma_addr;
uint32_t offset;
uint32_t iocbCnt = 0;
int i;
pcbp->maxRing = (psli->num_rings - 1);
for (i = 0; i < psli->num_rings; i++) {
scsi: lpfc: NVME Initiator: Base modifications NVME Initiator: Base modifications This patch adds base modifications for NVME initiator support. The base modifications consist of: - Formal split of SLI3 rings from SLI-4 WQs (sometimes referred to as rings as well) as implementation now widely varies between the two. - Addition of configuration modes: SCSI initiator only; NVME initiator only; NVME target only; and SCSI and NVME initiator. The configuration mode drives overall adapter configuration, offloads enabled, and resource splits. NVME support is only available on SLI-4 devices and newer fw. - Implements the following based on configuration mode: - Exchange resources are split by protocol; Obviously, if only 1 mode, then no split occurs. Default is 50/50. module attribute allows tuning. - Pools and config parameters are separated per-protocol - Each protocol has it's own set of queues, but share interrupt vectors. SCSI: SLI3 devices have few queues and the original style of queue allocation remains. SLI4 devices piggy back on an "io-channel" concept that eventually needs to merge with scsi-mq/blk-mq support (it is underway). For now, the paradigm continues as it existed prior. io channel allocates N msix and N WQs (N=4 default) and either round robins or uses cpu # modulo N for scheduling. A bunch of module parameters allow the configuration to be tuned. NVME (initiator): Allocates an msix per cpu (or whatever pci_alloc_irq_vectors gets) Allocates a WQ per cpu, and maps the WQs to msix on a WQ # modulo msix vector count basis. Module parameters exist to cap/control the config if desired. - Each protocol has its own buffer and dma pools. I apologize for the size of the patch. Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <james.smart@broadcom.com> ---- Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2017-02-13 05:52:30 +08:00
pring = &psli->sli3_ring[i];
pring->sli.sli3.sizeCiocb =
phba->sli_rev == 3 ? SLI3_IOCB_CMD_SIZE :
SLI2_IOCB_CMD_SIZE;
pring->sli.sli3.sizeRiocb =
phba->sli_rev == 3 ? SLI3_IOCB_RSP_SIZE :
SLI2_IOCB_RSP_SIZE;
/* A ring MUST have both cmd and rsp entries defined to be
valid */
if ((pring->sli.sli3.numCiocb == 0) ||
(pring->sli.sli3.numRiocb == 0)) {
pcbp->rdsc[i].cmdEntries = 0;
pcbp->rdsc[i].rspEntries = 0;
pcbp->rdsc[i].cmdAddrHigh = 0;
pcbp->rdsc[i].rspAddrHigh = 0;
pcbp->rdsc[i].cmdAddrLow = 0;
pcbp->rdsc[i].rspAddrLow = 0;
pring->sli.sli3.cmdringaddr = NULL;
pring->sli.sli3.rspringaddr = NULL;
continue;
}
/* Command ring setup for ring */
pring->sli.sli3.cmdringaddr = (void *)&phba->IOCBs[iocbCnt];
pcbp->rdsc[i].cmdEntries = pring->sli.sli3.numCiocb;
offset = (uint8_t *) &phba->IOCBs[iocbCnt] -
(uint8_t *) phba->slim2p.virt;
pdma_addr = phba->slim2p.phys + offset;
pcbp->rdsc[i].cmdAddrHigh = putPaddrHigh(pdma_addr);
pcbp->rdsc[i].cmdAddrLow = putPaddrLow(pdma_addr);
iocbCnt += pring->sli.sli3.numCiocb;
/* Response ring setup for ring */
pring->sli.sli3.rspringaddr = (void *) &phba->IOCBs[iocbCnt];
pcbp->rdsc[i].rspEntries = pring->sli.sli3.numRiocb;
offset = (uint8_t *)&phba->IOCBs[iocbCnt] -
(uint8_t *)phba->slim2p.virt;
pdma_addr = phba->slim2p.phys + offset;
pcbp->rdsc[i].rspAddrHigh = putPaddrHigh(pdma_addr);
pcbp->rdsc[i].rspAddrLow = putPaddrLow(pdma_addr);
iocbCnt += pring->sli.sli3.numRiocb;
}
}
/**
* lpfc_read_rev - Prepare a mailbox command for reading HBA revision
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The read revision mailbox command is used to read the revision levels of
* the HBA components. These components include hardware units, resident
* firmware, and available firmware. HBAs that supports SLI-3 mode of
* operation provide different response information depending on the version
* requested by the driver.
*
* This routine prepares the mailbox command for reading HBA revision
* information.
**/
void
lpfc_read_rev(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
{
MAILBOX_t *mb = &pmb->u.mb;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->un.varRdRev.cv = 1;
mb->un.varRdRev.v3req = 1; /* Request SLI3 info */
mb->mbxCommand = MBX_READ_REV;
mb->mbxOwner = OWN_HOST;
return;
}
void
lpfc_sli4_swap_str(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
{
MAILBOX_t *mb = &pmb->u.mb;
struct lpfc_mqe *mqe;
switch (mb->mbxCommand) {
case MBX_READ_REV:
mqe = &pmb->u.mqe;
lpfc_sli_pcimem_bcopy(mqe->un.read_rev.fw_name,
mqe->un.read_rev.fw_name, 16);
lpfc_sli_pcimem_bcopy(mqe->un.read_rev.ulp_fw_name,
mqe->un.read_rev.ulp_fw_name, 16);
break;
default:
break;
}
return;
}
/**
* lpfc_build_hbq_profile2 - Set up the HBQ Selection Profile 2
* @hbqmb: pointer to the HBQ configuration data structure in mailbox command.
* @hbq_desc: pointer to the HBQ selection profile descriptor.
*
* The Host Buffer Queue (HBQ) Selection Profile 2 specifies that the HBA
* tests the incoming frames' R_CTL/TYPE fields with works 10:15 and performs
* the Sequence Length Test using the fields in the Selection Profile 2
* extension in words 20:31.
**/
static void
lpfc_build_hbq_profile2(struct config_hbq_var *hbqmb,
struct lpfc_hbq_init *hbq_desc)
{
hbqmb->profiles.profile2.seqlenbcnt = hbq_desc->seqlenbcnt;
hbqmb->profiles.profile2.maxlen = hbq_desc->maxlen;
hbqmb->profiles.profile2.seqlenoff = hbq_desc->seqlenoff;
}
/**
* lpfc_build_hbq_profile3 - Set up the HBQ Selection Profile 3
* @hbqmb: pointer to the HBQ configuration data structure in mailbox command.
* @hbq_desc: pointer to the HBQ selection profile descriptor.
*
* The Host Buffer Queue (HBQ) Selection Profile 3 specifies that the HBA
* tests the incoming frame's R_CTL/TYPE fields with words 10:15 and performs
* the Sequence Length Test and Byte Field Test using the fields in the
* Selection Profile 3 extension in words 20:31.
**/
static void
lpfc_build_hbq_profile3(struct config_hbq_var *hbqmb,
struct lpfc_hbq_init *hbq_desc)
{
hbqmb->profiles.profile3.seqlenbcnt = hbq_desc->seqlenbcnt;
hbqmb->profiles.profile3.maxlen = hbq_desc->maxlen;
hbqmb->profiles.profile3.cmdcodeoff = hbq_desc->cmdcodeoff;
hbqmb->profiles.profile3.seqlenoff = hbq_desc->seqlenoff;
memcpy(&hbqmb->profiles.profile3.cmdmatch, hbq_desc->cmdmatch,
sizeof(hbqmb->profiles.profile3.cmdmatch));
}
/**
* lpfc_build_hbq_profile5 - Set up the HBQ Selection Profile 5
* @hbqmb: pointer to the HBQ configuration data structure in mailbox command.
* @hbq_desc: pointer to the HBQ selection profile descriptor.
*
* The Host Buffer Queue (HBQ) Selection Profile 5 specifies a header HBQ. The
* HBA tests the initial frame of an incoming sequence using the frame's
* R_CTL/TYPE fields with words 10:15 and performs the Sequence Length Test
* and Byte Field Test using the fields in the Selection Profile 5 extension
* words 20:31.
**/
static void
lpfc_build_hbq_profile5(struct config_hbq_var *hbqmb,
struct lpfc_hbq_init *hbq_desc)
{
hbqmb->profiles.profile5.seqlenbcnt = hbq_desc->seqlenbcnt;
hbqmb->profiles.profile5.maxlen = hbq_desc->maxlen;
hbqmb->profiles.profile5.cmdcodeoff = hbq_desc->cmdcodeoff;
hbqmb->profiles.profile5.seqlenoff = hbq_desc->seqlenoff;
memcpy(&hbqmb->profiles.profile5.cmdmatch, hbq_desc->cmdmatch,
sizeof(hbqmb->profiles.profile5.cmdmatch));
}
/**
* lpfc_config_hbq - Prepare a mailbox command for configuring an HBQ
* @phba: pointer to lpfc hba data structure.
* @id: HBQ identifier.
* @hbq_desc: pointer to the HBA descriptor data structure.
* @hbq_entry_index: index of the HBQ entry data structures.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The configure HBQ (Host Buffer Queue) mailbox command is used to configure
* an HBQ. The configuration binds events that require buffers to a particular
* ring and HBQ based on a selection profile.
*
* This routine prepares the mailbox command for configuring an HBQ.
**/
void
lpfc_config_hbq(struct lpfc_hba *phba, uint32_t id,
struct lpfc_hbq_init *hbq_desc,
uint32_t hbq_entry_index, LPFC_MBOXQ_t *pmb)
{
int i;
MAILBOX_t *mb = &pmb->u.mb;
struct config_hbq_var *hbqmb = &mb->un.varCfgHbq;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
hbqmb->hbqId = id;
hbqmb->entry_count = hbq_desc->entry_count; /* # entries in HBQ */
hbqmb->recvNotify = hbq_desc->rn; /* Receive
* Notification */
hbqmb->numMask = hbq_desc->mask_count; /* # R_CTL/TYPE masks
* # in words 0-19 */
hbqmb->profile = hbq_desc->profile; /* Selection profile:
* 0 = all,
* 7 = logentry */
hbqmb->ringMask = hbq_desc->ring_mask; /* Binds HBQ to a ring
* e.g. Ring0=b0001,
* ring2=b0100 */
hbqmb->headerLen = hbq_desc->headerLen; /* 0 if not profile 4
* or 5 */
hbqmb->logEntry = hbq_desc->logEntry; /* Set to 1 if this
* HBQ will be used
* for LogEntry
* buffers */
hbqmb->hbqaddrLow = putPaddrLow(phba->hbqslimp.phys) +
hbq_entry_index * sizeof(struct lpfc_hbq_entry);
hbqmb->hbqaddrHigh = putPaddrHigh(phba->hbqslimp.phys);
mb->mbxCommand = MBX_CONFIG_HBQ;
mb->mbxOwner = OWN_HOST;
/* Copy info for profiles 2,3,5. Other
* profiles this area is reserved
*/
if (hbq_desc->profile == 2)
lpfc_build_hbq_profile2(hbqmb, hbq_desc);
else if (hbq_desc->profile == 3)
lpfc_build_hbq_profile3(hbqmb, hbq_desc);
else if (hbq_desc->profile == 5)
lpfc_build_hbq_profile5(hbqmb, hbq_desc);
/* Return if no rctl / type masks for this HBQ */
if (!hbq_desc->mask_count)
return;
/* Otherwise we setup specific rctl / type masks for this HBQ */
for (i = 0; i < hbq_desc->mask_count; i++) {
hbqmb->hbqMasks[i].tmatch = hbq_desc->hbqMasks[i].tmatch;
hbqmb->hbqMasks[i].tmask = hbq_desc->hbqMasks[i].tmask;
hbqmb->hbqMasks[i].rctlmatch = hbq_desc->hbqMasks[i].rctlmatch;
hbqmb->hbqMasks[i].rctlmask = hbq_desc->hbqMasks[i].rctlmask;
}
return;
}
/**
* lpfc_config_ring - Prepare a mailbox command for configuring an IOCB ring
* @phba: pointer to lpfc hba data structure.
* @ring:
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The configure ring mailbox command is used to configure an IOCB ring. This
* configuration binds from one to six of HBA RC_CTL/TYPE mask entries to the
* ring. This is used to map incoming sequences to a particular ring whose
* RC_CTL/TYPE mask entry matches that of the sequence. The driver should not
* attempt to configure a ring whose number is greater than the number
* specified in the Port Control Block (PCB). It is an error to issue the
* configure ring command more than once with the same ring number. The HBA
* returns an error if the driver attempts this.
*
* This routine prepares the mailbox command for configuring IOCB ring.
**/
void
lpfc_config_ring(struct lpfc_hba * phba, int ring, LPFC_MBOXQ_t * pmb)
{
int i;
MAILBOX_t *mb = &pmb->u.mb;
struct lpfc_sli *psli;
struct lpfc_sli_ring *pring;
memset(pmb, 0, sizeof (LPFC_MBOXQ_t));
mb->un.varCfgRing.ring = ring;
mb->un.varCfgRing.maxOrigXchg = 0;
mb->un.varCfgRing.maxRespXchg = 0;
mb->un.varCfgRing.recvNotify = 1;
psli = &phba->sli;
scsi: lpfc: NVME Initiator: Base modifications NVME Initiator: Base modifications This patch adds base modifications for NVME initiator support. The base modifications consist of: - Formal split of SLI3 rings from SLI-4 WQs (sometimes referred to as rings as well) as implementation now widely varies between the two. - Addition of configuration modes: SCSI initiator only; NVME initiator only; NVME target only; and SCSI and NVME initiator. The configuration mode drives overall adapter configuration, offloads enabled, and resource splits. NVME support is only available on SLI-4 devices and newer fw. - Implements the following based on configuration mode: - Exchange resources are split by protocol; Obviously, if only 1 mode, then no split occurs. Default is 50/50. module attribute allows tuning. - Pools and config parameters are separated per-protocol - Each protocol has it's own set of queues, but share interrupt vectors. SCSI: SLI3 devices have few queues and the original style of queue allocation remains. SLI4 devices piggy back on an "io-channel" concept that eventually needs to merge with scsi-mq/blk-mq support (it is underway). For now, the paradigm continues as it existed prior. io channel allocates N msix and N WQs (N=4 default) and either round robins or uses cpu # modulo N for scheduling. A bunch of module parameters allow the configuration to be tuned. NVME (initiator): Allocates an msix per cpu (or whatever pci_alloc_irq_vectors gets) Allocates a WQ per cpu, and maps the WQs to msix on a WQ # modulo msix vector count basis. Module parameters exist to cap/control the config if desired. - Each protocol has its own buffer and dma pools. I apologize for the size of the patch. Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <james.smart@broadcom.com> ---- Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2017-02-13 05:52:30 +08:00
pring = &psli->sli3_ring[ring];
mb->un.varCfgRing.numMask = pring->num_mask;
mb->mbxCommand = MBX_CONFIG_RING;
mb->mbxOwner = OWN_HOST;
/* Is this ring configured for a specific profile */
if (pring->prt[0].profile) {
mb->un.varCfgRing.profile = pring->prt[0].profile;
return;
}
/* Otherwise we setup specific rctl / type masks for this ring */
for (i = 0; i < pring->num_mask; i++) {
mb->un.varCfgRing.rrRegs[i].rval = pring->prt[i].rctl;
if (mb->un.varCfgRing.rrRegs[i].rval != FC_RCTL_ELS_REQ)
mb->un.varCfgRing.rrRegs[i].rmask = 0xff;
else
mb->un.varCfgRing.rrRegs[i].rmask = 0xfe;
mb->un.varCfgRing.rrRegs[i].tval = pring->prt[i].type;
mb->un.varCfgRing.rrRegs[i].tmask = 0xff;
}
return;
}
/**
* lpfc_config_port - Prepare a mailbox command for configuring port
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The configure port mailbox command is used to identify the Port Control
* Block (PCB) in the driver memory. After this command is issued, the
* driver must not access the mailbox in the HBA without first resetting
* the HBA. The HBA may copy the PCB information to internal storage for
* subsequent use; the driver can not change the PCB information unless it
* resets the HBA.
*
* This routine prepares the mailbox command for configuring port.
**/
void
lpfc_config_port(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
{
MAILBOX_t __iomem *mb_slim = (MAILBOX_t __iomem *) phba->MBslimaddr;
MAILBOX_t *mb = &pmb->u.mb;
dma_addr_t pdma_addr;
uint32_t bar_low, bar_high;
size_t offset;
struct lpfc_hgp hgp;
int i;
uint32_t pgp_offset;
memset(pmb, 0, sizeof(LPFC_MBOXQ_t));
mb->mbxCommand = MBX_CONFIG_PORT;
mb->mbxOwner = OWN_HOST;
mb->un.varCfgPort.pcbLen = sizeof(PCB_t);
offset = (uint8_t *)phba->pcb - (uint8_t *)phba->slim2p.virt;
pdma_addr = phba->slim2p.phys + offset;
mb->un.varCfgPort.pcbLow = putPaddrLow(pdma_addr);
mb->un.varCfgPort.pcbHigh = putPaddrHigh(pdma_addr);
/* Always Host Group Pointer is in SLIM */
mb->un.varCfgPort.hps = 1;
/* If HBA supports SLI=3 ask for it */
if (phba->sli_rev == LPFC_SLI_REV3 && phba->vpd.sli3Feat.cerbm) {
if (phba->cfg_enable_bg)
mb->un.varCfgPort.cbg = 1; /* configure BlockGuard */
mb->un.varCfgPort.cerbm = 1; /* Request HBQs */
mb->un.varCfgPort.ccrp = 1; /* Command Ring Polling */
mb->un.varCfgPort.max_hbq = lpfc_sli_hbq_count();
if (phba->max_vpi && phba->cfg_enable_npiv &&
phba->vpd.sli3Feat.cmv) {
mb->un.varCfgPort.max_vpi = LPFC_MAX_VPI;
mb->un.varCfgPort.cmv = 1;
} else
mb->un.varCfgPort.max_vpi = phba->max_vpi = 0;
} else
phba->sli_rev = LPFC_SLI_REV2;
mb->un.varCfgPort.sli_mode = phba->sli_rev;
/* If this is an SLI3 port, configure async status notification. */
if (phba->sli_rev == LPFC_SLI_REV3)
mb->un.varCfgPort.casabt = 1;
/* Now setup pcb */
phba->pcb->type = TYPE_NATIVE_SLI2;
phba->pcb->feature = FEATURE_INITIAL_SLI2;
/* Setup Mailbox pointers */
phba->pcb->mailBoxSize = sizeof(MAILBOX_t) + MAILBOX_EXT_SIZE;
offset = (uint8_t *)phba->mbox - (uint8_t *)phba->slim2p.virt;
pdma_addr = phba->slim2p.phys + offset;
phba->pcb->mbAddrHigh = putPaddrHigh(pdma_addr);
phba->pcb->mbAddrLow = putPaddrLow(pdma_addr);
/*
* Setup Host Group ring pointer.
*
* For efficiency reasons, the ring get/put pointers can be
* placed in adapter memory (SLIM) rather than in host memory.
* This allows firmware to avoid PCI reads/writes when updating
* and checking pointers.
*
* The firmware recognizes the use of SLIM memory by comparing
* the address of the get/put pointers structure with that of
* the SLIM BAR (BAR0).
*
* Caution: be sure to use the PCI config space value of BAR0/BAR1
* (the hardware's view of the base address), not the OS's
* value of pci_resource_start() as the OS value may be a cookie
* for ioremap/iomap.
*/
pci_read_config_dword(phba->pcidev, PCI_BASE_ADDRESS_0, &bar_low);
pci_read_config_dword(phba->pcidev, PCI_BASE_ADDRESS_1, &bar_high);
/*
* Set up HGP - Port Memory
*
* The port expects the host get/put pointers to reside in memory
* following the "non-diagnostic" mode mailbox (32 words, 0x80 bytes)
* area of SLIM. In SLI-2 mode, there's an additional 16 reserved
* words (0x40 bytes). This area is not reserved if HBQs are
* configured in SLI-3.
*
* CR0Put - SLI2(no HBQs) = 0xc0, With HBQs = 0x80
* RR0Get 0xc4 0x84
* CR1Put 0xc8 0x88
* RR1Get 0xcc 0x8c
* CR2Put 0xd0 0x90
* RR2Get 0xd4 0x94
* CR3Put 0xd8 0x98
* RR3Get 0xdc 0x9c
*
* Reserved 0xa0-0xbf
* If HBQs configured:
* HBQ 0 Put ptr 0xc0
* HBQ 1 Put ptr 0xc4
* HBQ 2 Put ptr 0xc8
* ......
* HBQ(M-1)Put Pointer 0xc0+(M-1)*4
*
*/
if (phba->cfg_hostmem_hgp && phba->sli_rev != 3) {
phba->host_gp = &phba->mbox->us.s2.host[0];
phba->hbq_put = NULL;
offset = (uint8_t *)&phba->mbox->us.s2.host -
(uint8_t *)phba->slim2p.virt;
pdma_addr = phba->slim2p.phys + offset;
phba->pcb->hgpAddrHigh = putPaddrHigh(pdma_addr);
phba->pcb->hgpAddrLow = putPaddrLow(pdma_addr);
} else {
/* Always Host Group Pointer is in SLIM */
mb->un.varCfgPort.hps = 1;
if (phba->sli_rev == 3) {
phba->host_gp = &mb_slim->us.s3.host[0];
phba->hbq_put = &mb_slim->us.s3.hbq_put[0];
} else {
phba->host_gp = &mb_slim->us.s2.host[0];
phba->hbq_put = NULL;
}
/* mask off BAR0's flag bits 0 - 3 */
phba->pcb->hgpAddrLow = (bar_low & PCI_BASE_ADDRESS_MEM_MASK) +
(void __iomem *)phba->host_gp -
(void __iomem *)phba->MBslimaddr;
if (bar_low & PCI_BASE_ADDRESS_MEM_TYPE_64)
phba->pcb->hgpAddrHigh = bar_high;
else
phba->pcb->hgpAddrHigh = 0;
/* write HGP data to SLIM at the required longword offset */
memset(&hgp, 0, sizeof(struct lpfc_hgp));
for (i = 0; i < phba->sli.num_rings; i++) {
lpfc_memcpy_to_slim(phba->host_gp + i, &hgp,
sizeof(*phba->host_gp));
}
}
/* Setup Port Group offset */
if (phba->sli_rev == 3)
pgp_offset = offsetof(struct lpfc_sli2_slim,
mbx.us.s3_pgp.port);
else
pgp_offset = offsetof(struct lpfc_sli2_slim, mbx.us.s2.port);
pdma_addr = phba->slim2p.phys + pgp_offset;
phba->pcb->pgpAddrHigh = putPaddrHigh(pdma_addr);
phba->pcb->pgpAddrLow = putPaddrLow(pdma_addr);
/* Use callback routine to setp rings in the pcb */
lpfc_config_pcb_setup(phba);
/* special handling for LC HBAs */
if (lpfc_is_LC_HBA(phba->pcidev->device)) {
uint32_t hbainit[5];
lpfc_hba_init(phba, hbainit);
memcpy(&mb->un.varCfgPort.hbainit, hbainit, 20);
}
/* Swap PCB if needed */
lpfc_sli_pcimem_bcopy(phba->pcb, phba->pcb, sizeof(PCB_t));
}
/**
* lpfc_kill_board - Prepare a mailbox command for killing board
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* The kill board mailbox command is used to tell firmware to perform a
* graceful shutdown of a channel on a specified board to prepare for reset.
* When the kill board mailbox command is received, the ER3 bit is set to 1
* in the Host Status register and the ER Attention bit is set to 1 in the
* Host Attention register of the HBA function that received the kill board
* command.
*
* This routine prepares the mailbox command for killing the board in
* preparation for a graceful shutdown.
**/
void
lpfc_kill_board(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
{
MAILBOX_t *mb = &pmb->u.mb;
memset(pmb, 0, sizeof(LPFC_MBOXQ_t));
mb->mbxCommand = MBX_KILL_BOARD;
mb->mbxOwner = OWN_HOST;
return;
}
/**
* lpfc_mbox_put - Put a mailbox cmd into the tail of driver's mailbox queue
* @phba: pointer to lpfc hba data structure.
* @mbq: pointer to the driver internal queue element for mailbox command.
*
* Driver maintains a internal mailbox command queue implemented as a linked
* list. When a mailbox command is issued, it shall be put into the mailbox
* command queue such that they shall be processed orderly as HBA can process
* one mailbox command at a time.
**/
void
lpfc_mbox_put(struct lpfc_hba * phba, LPFC_MBOXQ_t * mbq)
{
struct lpfc_sli *psli;
psli = &phba->sli;
list_add_tail(&mbq->list, &psli->mboxq);
psli->mboxq_cnt++;
return;
}
/**
* lpfc_mbox_get - Remove a mailbox cmd from the head of driver's mailbox queue
* @phba: pointer to lpfc hba data structure.
*
* Driver maintains a internal mailbox command queue implemented as a linked
* list. When a mailbox command is issued, it shall be put into the mailbox
* command queue such that they shall be processed orderly as HBA can process
* one mailbox command at a time. After HBA finished processing a mailbox
* command, the driver will remove a pending mailbox command from the head of
* the mailbox command queue and send to the HBA for processing.
*
* Return codes
* pointer to the driver internal queue element for mailbox command.
**/
LPFC_MBOXQ_t *
lpfc_mbox_get(struct lpfc_hba * phba)
{
LPFC_MBOXQ_t *mbq = NULL;
struct lpfc_sli *psli = &phba->sli;
list_remove_head((&psli->mboxq), mbq, LPFC_MBOXQ_t, list);
if (mbq)
psli->mboxq_cnt--;
return mbq;
}
/**
* __lpfc_mbox_cmpl_put - Put mailbox cmd into mailbox cmd complete list
* @phba: pointer to lpfc hba data structure.
* @mbq: pointer to the driver internal queue element for mailbox command.
*
* This routine put the completed mailbox command into the mailbox command
* complete list. This is the unlocked version of the routine. The mailbox
* complete list is used by the driver worker thread to process mailbox
* complete callback functions outside the driver interrupt handler.
**/
void
__lpfc_mbox_cmpl_put(struct lpfc_hba *phba, LPFC_MBOXQ_t *mbq)
{
list_add_tail(&mbq->list, &phba->sli.mboxq_cmpl);
}
/**
* lpfc_mbox_cmpl_put - Put mailbox command into mailbox command complete list
* @phba: pointer to lpfc hba data structure.
* @mbq: pointer to the driver internal queue element for mailbox command.
*
* This routine put the completed mailbox command into the mailbox command
* complete list. This is the locked version of the routine. The mailbox
* complete list is used by the driver worker thread to process mailbox
* complete callback functions outside the driver interrupt handler.
**/
void
lpfc_mbox_cmpl_put(struct lpfc_hba *phba, LPFC_MBOXQ_t *mbq)
{
unsigned long iflag;
/* This function expects to be called from interrupt context */
spin_lock_irqsave(&phba->hbalock, iflag);
__lpfc_mbox_cmpl_put(phba, mbq);
spin_unlock_irqrestore(&phba->hbalock, iflag);
return;
}
/**
* lpfc_mbox_cmd_check - Check the validality of a mailbox command
* @phba: pointer to lpfc hba data structure.
* @mboxq: pointer to the driver internal queue element for mailbox command.
*
* This routine is to check whether a mailbox command is valid to be issued.
* This check will be performed by both the mailbox issue API when a client
* is to issue a mailbox command to the mailbox transport.
*
* Return 0 - pass the check, -ENODEV - fail the check
**/
int
lpfc_mbox_cmd_check(struct lpfc_hba *phba, LPFC_MBOXQ_t *mboxq)
{
/* Mailbox command that have a completion handler must also have a
* vport specified.
*/
if (mboxq->mbox_cmpl && mboxq->mbox_cmpl != lpfc_sli_def_mbox_cmpl &&
mboxq->mbox_cmpl != lpfc_sli_wake_mbox_wait) {
if (!mboxq->vport) {
lpfc_printf_log(phba, KERN_ERR, LOG_MBOX | LOG_VPORT,
"1814 Mbox x%x failed, no vport\n",
mboxq->u.mb.mbxCommand);
dump_stack();
return -ENODEV;
}
}
return 0;
}
/**
* lpfc_mbox_dev_check - Check the device state for issuing a mailbox command
* @phba: pointer to lpfc hba data structure.
*
* This routine is to check whether the HBA device is ready for posting a
* mailbox command. It is used by the mailbox transport API at the time the
* to post a mailbox command to the device.
*
* Return 0 - pass the check, -ENODEV - fail the check
**/
int
lpfc_mbox_dev_check(struct lpfc_hba *phba)
{
/* If the PCI channel is in offline state, do not issue mbox */
if (unlikely(pci_channel_offline(phba->pcidev)))
return -ENODEV;
/* If the HBA is in error state, do not issue mbox */
if (phba->link_state == LPFC_HBA_ERROR)
return -ENODEV;
return 0;
}
/**
* lpfc_mbox_tmo_val - Retrieve mailbox command timeout value
* @phba: pointer to lpfc hba data structure.
* @cmd: mailbox command code.
*
* This routine retrieves the proper timeout value according to the mailbox
* command code.
*
* Return codes
* Timeout value to be used for the given mailbox command
**/
int
lpfc_mbox_tmo_val(struct lpfc_hba *phba, LPFC_MBOXQ_t *mboxq)
{
MAILBOX_t *mbox = &mboxq->u.mb;
uint8_t subsys, opcode;
switch (mbox->mbxCommand) {
case MBX_WRITE_NV: /* 0x03 */
case MBX_DUMP_MEMORY: /* 0x17 */
case MBX_UPDATE_CFG: /* 0x1B */
case MBX_DOWN_LOAD: /* 0x1C */
case MBX_DEL_LD_ENTRY: /* 0x1D */
case MBX_WRITE_VPARMS: /* 0x32 */
case MBX_LOAD_AREA: /* 0x81 */
[SCSI] lpfc 8.2.4 : Miscellaneous Fixes Miscellaneous Fixes: - Fix a couple of sparse complaints - Reset the FCP recovery flag when the node is not a FCP2 device. - Speed up offline prep delays - Fixed a memory leak in lpfc_mem_alloc failure path - Fixed external loopback test. - Fixed error code returned from the driver when HBA is over heated. - Correct Max NPIV vport to limits read from adapter - Add missing locks around fc_flag and FC_NEEDS_REG_VPI - Add missing hba ids for device identification - Added support for SET_VARIABLE and MBX_WRITE_WWN mailbox commands - Changed all temperature event messages from warning to error - Fix reporting of link speed when link is down - Added support for MBX_WRITE_WWN mailbox command - Change del_timer_sync() in ISR to del_timer() in interrupt handler - Correct instances of beXX_to_cpu() that should be cpu_to_beXX() - Perform target flush before releasing node references on module unload - Avoid bogus devloss_tmo messages when driver unloads - Fix panic when HBA generates ERATT interupt - Fix mbox race condition and a workaround on back-to-back mailbox commands - Force NPIV off for pt2pt mode between 2 NPorts - Stop worker thread before removing fc_host. - Fix up discovery timeout error case due to missing clear_la - Tighten mailbox polling code to speed up detection of fast completions - Only allow DUMP_MEMORY if adapter offline due to overtemp errors - Added extended error information to the log messages in chip init. Signed-off-by: James Smart <James.Smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-01-11 14:52:54 +08:00
case MBX_WRITE_WWN: /* 0x98 */
case MBX_LOAD_EXP_ROM: /* 0x9C */
case MBX_ACCESS_VDATA: /* 0xA5 */
return LPFC_MBOX_TMO_FLASH_CMD;
case MBX_SLI4_CONFIG: /* 0x9b */
subsys = lpfc_sli_config_mbox_subsys_get(phba, mboxq);
opcode = lpfc_sli_config_mbox_opcode_get(phba, mboxq);
if (subsys == LPFC_MBOX_SUBSYSTEM_COMMON) {
switch (opcode) {
case LPFC_MBOX_OPCODE_READ_OBJECT:
case LPFC_MBOX_OPCODE_WRITE_OBJECT:
case LPFC_MBOX_OPCODE_READ_OBJECT_LIST:
case LPFC_MBOX_OPCODE_DELETE_OBJECT:
case LPFC_MBOX_OPCODE_GET_PROFILE_LIST:
case LPFC_MBOX_OPCODE_SET_ACT_PROFILE:
case LPFC_MBOX_OPCODE_GET_PROFILE_CONFIG:
case LPFC_MBOX_OPCODE_SET_PROFILE_CONFIG:
case LPFC_MBOX_OPCODE_GET_FACTORY_PROFILE_CONFIG:
case LPFC_MBOX_OPCODE_GET_PROFILE_CAPACITIES:
case LPFC_MBOX_OPCODE_SEND_ACTIVATION:
case LPFC_MBOX_OPCODE_RESET_LICENSES:
case LPFC_MBOX_OPCODE_SET_BOOT_CONFIG:
case LPFC_MBOX_OPCODE_GET_VPD_DATA:
case LPFC_MBOX_OPCODE_SET_PHYSICAL_LINK_CONFIG:
return LPFC_MBOX_SLI4_CONFIG_EXTENDED_TMO;
}
}
if (subsys == LPFC_MBOX_SUBSYSTEM_FCOE) {
switch (opcode) {
case LPFC_MBOX_OPCODE_FCOE_SET_FCLINK_SETTINGS:
return LPFC_MBOX_SLI4_CONFIG_EXTENDED_TMO;
}
}
return LPFC_MBOX_SLI4_CONFIG_TMO;
}
return LPFC_MBOX_TMO;
}
/**
* lpfc_sli4_mbx_sge_set - Set a sge entry in non-embedded mailbox command
* @mbox: pointer to lpfc mbox command.
* @sgentry: sge entry index.
* @phyaddr: physical address for the sge
* @length: Length of the sge.
*
* This routine sets up an entry in the non-embedded mailbox command at the sge
* index location.
**/
void
lpfc_sli4_mbx_sge_set(struct lpfcMboxq *mbox, uint32_t sgentry,
dma_addr_t phyaddr, uint32_t length)
{
struct lpfc_mbx_nembed_cmd *nembed_sge;
nembed_sge = (struct lpfc_mbx_nembed_cmd *)
&mbox->u.mqe.un.nembed_cmd;
nembed_sge->sge[sgentry].pa_lo = putPaddrLow(phyaddr);
nembed_sge->sge[sgentry].pa_hi = putPaddrHigh(phyaddr);
nembed_sge->sge[sgentry].length = length;
}
/**
* lpfc_sli4_mbx_sge_get - Get a sge entry from non-embedded mailbox command
* @mbox: pointer to lpfc mbox command.
* @sgentry: sge entry index.
*
* This routine gets an entry from the non-embedded mailbox command at the sge
* index location.
**/
void
lpfc_sli4_mbx_sge_get(struct lpfcMboxq *mbox, uint32_t sgentry,
struct lpfc_mbx_sge *sge)
{
struct lpfc_mbx_nembed_cmd *nembed_sge;
nembed_sge = (struct lpfc_mbx_nembed_cmd *)
&mbox->u.mqe.un.nembed_cmd;
sge->pa_lo = nembed_sge->sge[sgentry].pa_lo;
sge->pa_hi = nembed_sge->sge[sgentry].pa_hi;
sge->length = nembed_sge->sge[sgentry].length;
}
/**
* lpfc_sli4_mbox_cmd_free - Free a sli4 mailbox command
* @phba: pointer to lpfc hba data structure.
* @mbox: pointer to lpfc mbox command.
*
* This routine frees SLI4 specific mailbox command for sending IOCTL command.
**/
void
lpfc_sli4_mbox_cmd_free(struct lpfc_hba *phba, struct lpfcMboxq *mbox)
{
struct lpfc_mbx_sli4_config *sli4_cfg;
struct lpfc_mbx_sge sge;
dma_addr_t phyaddr;
uint32_t sgecount, sgentry;
sli4_cfg = &mbox->u.mqe.un.sli4_config;
/* For embedded mbox command, just free the mbox command */
if (bf_get(lpfc_mbox_hdr_emb, &sli4_cfg->header.cfg_mhdr)) {
mempool_free(mbox, phba->mbox_mem_pool);
return;
}
/* For non-embedded mbox command, we need to free the pages first */
sgecount = bf_get(lpfc_mbox_hdr_sge_cnt, &sli4_cfg->header.cfg_mhdr);
/* There is nothing we can do if there is no sge address array */
if (unlikely(!mbox->sge_array)) {
mempool_free(mbox, phba->mbox_mem_pool);
return;
}
/* Each non-embedded DMA memory was allocated in the length of a page */
for (sgentry = 0; sgentry < sgecount; sgentry++) {
lpfc_sli4_mbx_sge_get(mbox, sgentry, &sge);
phyaddr = getPaddr(sge.pa_hi, sge.pa_lo);
dma_free_coherent(&phba->pcidev->dev, SLI4_PAGE_SIZE,
mbox->sge_array->addr[sgentry], phyaddr);
}
/* Free the sge address array memory */
kfree(mbox->sge_array);
/* Finally, free the mailbox command itself */
mempool_free(mbox, phba->mbox_mem_pool);
}
/**
* lpfc_sli4_config - Initialize the SLI4 Config Mailbox command
* @phba: pointer to lpfc hba data structure.
* @mbox: pointer to lpfc mbox command.
* @subsystem: The sli4 config sub mailbox subsystem.
* @opcode: The sli4 config sub mailbox command opcode.
* @length: Length of the sli4 config mailbox command (including sub-header).
*
* This routine sets up the header fields of SLI4 specific mailbox command
* for sending IOCTL command.
*
* Return: the actual length of the mbox command allocated (mostly useful
* for none embedded mailbox command).
**/
int
lpfc_sli4_config(struct lpfc_hba *phba, struct lpfcMboxq *mbox,
uint8_t subsystem, uint8_t opcode, uint32_t length, bool emb)
{
struct lpfc_mbx_sli4_config *sli4_config;
union lpfc_sli4_cfg_shdr *cfg_shdr = NULL;
uint32_t alloc_len;
uint32_t resid_len;
uint32_t pagen, pcount;
void *viraddr;
dma_addr_t phyaddr;
/* Set up SLI4 mailbox command header fields */
memset(mbox, 0, sizeof(*mbox));
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_SLI4_CONFIG);
/* Set up SLI4 ioctl command header fields */
sli4_config = &mbox->u.mqe.un.sli4_config;
/* Setup for the embedded mbox command */
if (emb) {
/* Set up main header fields */
bf_set(lpfc_mbox_hdr_emb, &sli4_config->header.cfg_mhdr, 1);
sli4_config->header.cfg_mhdr.payload_length = length;
/* Set up sub-header fields following main header */
bf_set(lpfc_mbox_hdr_opcode,
&sli4_config->header.cfg_shdr.request, opcode);
bf_set(lpfc_mbox_hdr_subsystem,
&sli4_config->header.cfg_shdr.request, subsystem);
sli4_config->header.cfg_shdr.request.request_length =
length - LPFC_MBX_CMD_HDR_LENGTH;
return length;
}
/* Setup for the non-embedded mbox command */
pcount = (SLI4_PAGE_ALIGN(length))/SLI4_PAGE_SIZE;
pcount = (pcount > LPFC_SLI4_MBX_SGE_MAX_PAGES) ?
LPFC_SLI4_MBX_SGE_MAX_PAGES : pcount;
/* Allocate record for keeping SGE virtual addresses */
mbox->sge_array = kzalloc(sizeof(struct lpfc_mbx_nembed_sge_virt),
GFP_KERNEL);
if (!mbox->sge_array) {
lpfc_printf_log(phba, KERN_ERR, LOG_MBOX,
"2527 Failed to allocate non-embedded SGE "
"array.\n");
return 0;
}
for (pagen = 0, alloc_len = 0; pagen < pcount; pagen++) {
/* The DMA memory is always allocated in the length of a
* page even though the last SGE might not fill up to a
* page, this is used as a priori size of SLI4_PAGE_SIZE for
* the later DMA memory free.
*/
viraddr = dma_alloc_coherent(&phba->pcidev->dev,
SLI4_PAGE_SIZE, &phyaddr,
GFP_KERNEL);
/* In case of malloc fails, proceed with whatever we have */
if (!viraddr)
break;
mbox->sge_array->addr[pagen] = viraddr;
/* Keep the first page for later sub-header construction */
if (pagen == 0)
cfg_shdr = (union lpfc_sli4_cfg_shdr *)viraddr;
resid_len = length - alloc_len;
if (resid_len > SLI4_PAGE_SIZE) {
lpfc_sli4_mbx_sge_set(mbox, pagen, phyaddr,
SLI4_PAGE_SIZE);
alloc_len += SLI4_PAGE_SIZE;
} else {
lpfc_sli4_mbx_sge_set(mbox, pagen, phyaddr,
resid_len);
alloc_len = length;
}
}
/* Set up main header fields in mailbox command */
sli4_config->header.cfg_mhdr.payload_length = alloc_len;
bf_set(lpfc_mbox_hdr_sge_cnt, &sli4_config->header.cfg_mhdr, pagen);
/* Set up sub-header fields into the first page */
if (pagen > 0) {
bf_set(lpfc_mbox_hdr_opcode, &cfg_shdr->request, opcode);
bf_set(lpfc_mbox_hdr_subsystem, &cfg_shdr->request, subsystem);
cfg_shdr->request.request_length =
alloc_len - sizeof(union lpfc_sli4_cfg_shdr);
}
/* The sub-header is in DMA memory, which needs endian converstion */
if (cfg_shdr)
lpfc_sli_pcimem_bcopy(cfg_shdr, cfg_shdr,
sizeof(union lpfc_sli4_cfg_shdr));
return alloc_len;
}
/**
* lpfc_sli4_mbox_rsrc_extent - Initialize the opcode resource extent.
* @phba: pointer to lpfc hba data structure.
* @mbox: pointer to an allocated lpfc mbox resource.
* @exts_count: the number of extents, if required, to allocate.
* @rsrc_type: the resource extent type.
* @emb: true if LPFC_SLI4_MBX_EMBED. false if LPFC_SLI4_MBX_NEMBED.
*
* This routine completes the subcommand header for SLI4 resource extent
* mailbox commands. It is called after lpfc_sli4_config. The caller must
* pass an allocated mailbox and the attributes required to initialize the
* mailbox correctly.
*
* Return: the actual length of the mbox command allocated.
**/
int
lpfc_sli4_mbox_rsrc_extent(struct lpfc_hba *phba, struct lpfcMboxq *mbox,
uint16_t exts_count, uint16_t rsrc_type, bool emb)
{
uint8_t opcode = 0;
struct lpfc_mbx_nembed_rsrc_extent *n_rsrc_extnt = NULL;
void *virtaddr = NULL;
/* Set up SLI4 ioctl command header fields */
if (emb == LPFC_SLI4_MBX_NEMBED) {
/* Get the first SGE entry from the non-embedded DMA memory */
virtaddr = mbox->sge_array->addr[0];
if (virtaddr == NULL)
return 1;
n_rsrc_extnt = (struct lpfc_mbx_nembed_rsrc_extent *) virtaddr;
}
/*
* The resource type is common to all extent Opcodes and resides in the
* same position.
*/
if (emb == LPFC_SLI4_MBX_EMBED)
bf_set(lpfc_mbx_alloc_rsrc_extents_type,
&mbox->u.mqe.un.alloc_rsrc_extents.u.req,
rsrc_type);
else {
/* This is DMA data. Byteswap is required. */
bf_set(lpfc_mbx_alloc_rsrc_extents_type,
n_rsrc_extnt, rsrc_type);
lpfc_sli_pcimem_bcopy(&n_rsrc_extnt->word4,
&n_rsrc_extnt->word4,
sizeof(uint32_t));
}
/* Complete the initialization for the particular Opcode. */
opcode = lpfc_sli_config_mbox_opcode_get(phba, mbox);
switch (opcode) {
case LPFC_MBOX_OPCODE_ALLOC_RSRC_EXTENT:
if (emb == LPFC_SLI4_MBX_EMBED)
bf_set(lpfc_mbx_alloc_rsrc_extents_cnt,
&mbox->u.mqe.un.alloc_rsrc_extents.u.req,
exts_count);
else
bf_set(lpfc_mbx_alloc_rsrc_extents_cnt,
n_rsrc_extnt, exts_count);
break;
case LPFC_MBOX_OPCODE_GET_ALLOC_RSRC_EXTENT:
case LPFC_MBOX_OPCODE_GET_RSRC_EXTENT_INFO:
case LPFC_MBOX_OPCODE_DEALLOC_RSRC_EXTENT:
/* Initialization is complete.*/
break;
default:
lpfc_printf_log(phba, KERN_ERR, LOG_MBOX,
"2929 Resource Extent Opcode x%x is "
"unsupported\n", opcode);
return 1;
}
return 0;
}
/**
* lpfc_sli_config_mbox_subsys_get - Get subsystem from a sli_config mbox cmd
* @phba: pointer to lpfc hba data structure.
* @mbox: pointer to lpfc mbox command queue entry.
*
* This routine gets the subsystem from a SLI4 specific SLI_CONFIG mailbox
* command. If the mailbox command is not MBX_SLI4_CONFIG (0x9B) or if the
* sub-header is not present, subsystem LPFC_MBOX_SUBSYSTEM_NA (0x0) shall
* be returned.
**/
uint8_t
lpfc_sli_config_mbox_subsys_get(struct lpfc_hba *phba, LPFC_MBOXQ_t *mbox)
{
struct lpfc_mbx_sli4_config *sli4_cfg;
union lpfc_sli4_cfg_shdr *cfg_shdr;
if (mbox->u.mb.mbxCommand != MBX_SLI4_CONFIG)
return LPFC_MBOX_SUBSYSTEM_NA;
sli4_cfg = &mbox->u.mqe.un.sli4_config;
/* For embedded mbox command, get opcode from embedded sub-header*/
if (bf_get(lpfc_mbox_hdr_emb, &sli4_cfg->header.cfg_mhdr)) {
cfg_shdr = &mbox->u.mqe.un.sli4_config.header.cfg_shdr;
return bf_get(lpfc_mbox_hdr_subsystem, &cfg_shdr->request);
}
/* For non-embedded mbox command, get opcode from first dma page */
if (unlikely(!mbox->sge_array))
return LPFC_MBOX_SUBSYSTEM_NA;
cfg_shdr = (union lpfc_sli4_cfg_shdr *)mbox->sge_array->addr[0];
return bf_get(lpfc_mbox_hdr_subsystem, &cfg_shdr->request);
}
/**
* lpfc_sli_config_mbox_opcode_get - Get opcode from a sli_config mbox cmd
* @phba: pointer to lpfc hba data structure.
* @mbox: pointer to lpfc mbox command queue entry.
*
* This routine gets the opcode from a SLI4 specific SLI_CONFIG mailbox
* command. If the mailbox command is not MBX_SLI4_CONFIG (0x9B) or if
* the sub-header is not present, opcode LPFC_MBOX_OPCODE_NA (0x0) be
* returned.
**/
uint8_t
lpfc_sli_config_mbox_opcode_get(struct lpfc_hba *phba, LPFC_MBOXQ_t *mbox)
{
struct lpfc_mbx_sli4_config *sli4_cfg;
union lpfc_sli4_cfg_shdr *cfg_shdr;
if (mbox->u.mb.mbxCommand != MBX_SLI4_CONFIG)
return LPFC_MBOX_OPCODE_NA;
sli4_cfg = &mbox->u.mqe.un.sli4_config;
/* For embedded mbox command, get opcode from embedded sub-header*/
if (bf_get(lpfc_mbox_hdr_emb, &sli4_cfg->header.cfg_mhdr)) {
cfg_shdr = &mbox->u.mqe.un.sli4_config.header.cfg_shdr;
return bf_get(lpfc_mbox_hdr_opcode, &cfg_shdr->request);
}
/* For non-embedded mbox command, get opcode from first dma page */
if (unlikely(!mbox->sge_array))
return LPFC_MBOX_OPCODE_NA;
cfg_shdr = (union lpfc_sli4_cfg_shdr *)mbox->sge_array->addr[0];
return bf_get(lpfc_mbox_hdr_opcode, &cfg_shdr->request);
}
/**
* lpfc_sli4_mbx_read_fcf_rec - Allocate and construct read fcf mbox cmd
* @phba: pointer to lpfc hba data structure.
* @fcf_index: index to fcf table.
*
* This routine routine allocates and constructs non-embedded mailbox command
* for reading a FCF table entry referred by @fcf_index.
*
* Return: pointer to the mailbox command constructed if successful, otherwise
* NULL.
**/
int
lpfc_sli4_mbx_read_fcf_rec(struct lpfc_hba *phba,
struct lpfcMboxq *mboxq,
uint16_t fcf_index)
{
void *virt_addr;
uint8_t *bytep;
struct lpfc_mbx_sge sge;
uint32_t alloc_len, req_len;
struct lpfc_mbx_read_fcf_tbl *read_fcf;
if (!mboxq)
return -ENOMEM;
req_len = sizeof(struct fcf_record) +
sizeof(union lpfc_sli4_cfg_shdr) + 2 * sizeof(uint32_t);
/* Set up READ_FCF SLI4_CONFIG mailbox-ioctl command */
alloc_len = lpfc_sli4_config(phba, mboxq, LPFC_MBOX_SUBSYSTEM_FCOE,
LPFC_MBOX_OPCODE_FCOE_READ_FCF_TABLE, req_len,
LPFC_SLI4_MBX_NEMBED);
if (alloc_len < req_len) {
lpfc_printf_log(phba, KERN_ERR, LOG_MBOX,
"0291 Allocated DMA memory size (x%x) is "
"less than the requested DMA memory "
"size (x%x)\n", alloc_len, req_len);
return -ENOMEM;
}
/* Get the first SGE entry from the non-embedded DMA memory. This
* routine only uses a single SGE.
*/
lpfc_sli4_mbx_sge_get(mboxq, 0, &sge);
virt_addr = mboxq->sge_array->addr[0];
read_fcf = (struct lpfc_mbx_read_fcf_tbl *)virt_addr;
/* Set up command fields */
bf_set(lpfc_mbx_read_fcf_tbl_indx, &read_fcf->u.request, fcf_index);
/* Perform necessary endian conversion */
bytep = virt_addr + sizeof(union lpfc_sli4_cfg_shdr);
lpfc_sli_pcimem_bcopy(bytep, bytep, sizeof(uint32_t));
return 0;
}
/**
* lpfc_request_features: Configure SLI4 REQUEST_FEATURES mailbox
* @mboxq: pointer to lpfc mbox command.
*
* This routine sets up the mailbox for an SLI4 REQUEST_FEATURES
* mailbox command.
**/
void
lpfc_request_features(struct lpfc_hba *phba, struct lpfcMboxq *mboxq)
{
/* Set up SLI4 mailbox command header fields */
memset(mboxq, 0, sizeof(LPFC_MBOXQ_t));
bf_set(lpfc_mqe_command, &mboxq->u.mqe, MBX_SLI4_REQ_FTRS);
/* Set up host requested features. */
bf_set(lpfc_mbx_rq_ftr_rq_fcpi, &mboxq->u.mqe.un.req_ftrs, 1);
bf_set(lpfc_mbx_rq_ftr_rq_perfh, &mboxq->u.mqe.un.req_ftrs, 1);
/* Enable DIF (block guard) only if configured to do so. */
if (phba->cfg_enable_bg)
bf_set(lpfc_mbx_rq_ftr_rq_dif, &mboxq->u.mqe.un.req_ftrs, 1);
/* Enable NPIV only if configured to do so. */
if (phba->max_vpi && phba->cfg_enable_npiv)
bf_set(lpfc_mbx_rq_ftr_rq_npiv, &mboxq->u.mqe.un.req_ftrs, 1);
Update ABORT processing for NVMET. The driver with nvme had this routine stubbed. Right now XRI_ABORTED_CQE is not handled and the FC NVMET Transport has a new API for the driver. Missing code path, new NVME abort API Update ABORT processing for NVMET There are 3 new FC NVMET Transport API/ template routines for NVMET: lpfc_nvmet_xmt_fcp_release This NVMET template callback routine called to release context associated with an IO This routine is ALWAYS called last, even if the IO was aborted or completed in error. lpfc_nvmet_xmt_fcp_abort This NVMET template callback routine called to abort an exchange that has an IO in progress nvmet_fc_rcv_fcp_req When the lpfc driver receives an ABTS, this NVME FC transport layer callback routine is called. For this case there are 2 paths thru the driver: the driver either has an outstanding exchange / context for the XRI to be aborted or not. If not, a BA_RJT is issued otherwise a BA_ACC NVMET Driver abort paths: There are 2 paths for aborting an IO. The first one is we receive an IO and decide not to process it because of lack of resources. An unsolicated ABTS is immediately sent back to the initiator as a response. lpfc_nvmet_unsol_fcp_buffer lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE) The second one is we sent the IO up to the NVMET transport layer to process, and for some reason the NVME Transport layer decided to abort the IO before it completes all its phases. For this case there are 2 paths thru the driver: the driver either has an outstanding TSEND/TRECEIVE/TRSP WQE or no outstanding WQEs are present for the exchange / context. lpfc_nvmet_xmt_fcp_abort if (LPFC_NVMET_IO_INP) lpfc_nvmet_sol_fcp_issue_abort (ABORT_WQE) lpfc_nvmet_sol_fcp_abort_cmp else lpfc_nvmet_unsol_fcp_issue_abort lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE) lpfc_nvmet_unsol_fcp_abort_cmp Context flags: LPFC_NVMET_IOP - his flag signifies an IO is in progress on the exchange. LPFC_NVMET_XBUSY - this flag indicates the IO completed but the firmware is still busy with the corresponding exchange. The exchange should not be reused until after a XRI_ABORTED_CQE is received for that exchange. LPFC_NVMET_ABORT_OP - this flag signifies an ABORT_WQE was issued on the exchange. LPFC_NVMET_CTX_RLS - this flag signifies a context free was requested, but we are deferring it due to an XBUSY or ABORT in progress. A ctxlock is added to the context structure that is used whenever these flags are set/read within the context of an IO. The LPFC_NVMET_CTX_RLS flag is only set in the defer_relase routine when the transport has resolved all IO associated with the buffer. The flag is cleared when the CTX is associated with a new IO. An exchange can has both an LPFC_NVMET_XBUSY and a LPFC_NVMET_ABORT_OP condition active simultaneously. Both conditions must complete before the exchange is freed. When the abort callback (lpfc_nvmet_xmt_fcp_abort) is envoked: If there is an outstanding IO, the driver will issue an ABORT_WQE. This should result in 3 completions for the exchange: 1) IO cmpl with XB bit set 2) Abort WQE cmpl 3) XRI_ABORTED_CQE cmpl For this scenerio, after completion #1, the NVMET Transport IO rsp callback is called. After completion #2, no action is taken with respect to the exchange / context. After completion #3, the exchange context is free for re-use on another IO. If there is no outstanding activity on the exchange, the driver will send a ABTS to the Initiator. Upon completion of this WQE, the exchange / context is freed for re-use on another IO. Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <james.smart@broadcom.com> Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
2017-04-22 07:05:04 +08:00
if (phba->nvmet_support) {
bf_set(lpfc_mbx_rq_ftr_rq_mrqp, &mboxq->u.mqe.un.req_ftrs, 1);
Update ABORT processing for NVMET. The driver with nvme had this routine stubbed. Right now XRI_ABORTED_CQE is not handled and the FC NVMET Transport has a new API for the driver. Missing code path, new NVME abort API Update ABORT processing for NVMET There are 3 new FC NVMET Transport API/ template routines for NVMET: lpfc_nvmet_xmt_fcp_release This NVMET template callback routine called to release context associated with an IO This routine is ALWAYS called last, even if the IO was aborted or completed in error. lpfc_nvmet_xmt_fcp_abort This NVMET template callback routine called to abort an exchange that has an IO in progress nvmet_fc_rcv_fcp_req When the lpfc driver receives an ABTS, this NVME FC transport layer callback routine is called. For this case there are 2 paths thru the driver: the driver either has an outstanding exchange / context for the XRI to be aborted or not. If not, a BA_RJT is issued otherwise a BA_ACC NVMET Driver abort paths: There are 2 paths for aborting an IO. The first one is we receive an IO and decide not to process it because of lack of resources. An unsolicated ABTS is immediately sent back to the initiator as a response. lpfc_nvmet_unsol_fcp_buffer lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE) The second one is we sent the IO up to the NVMET transport layer to process, and for some reason the NVME Transport layer decided to abort the IO before it completes all its phases. For this case there are 2 paths thru the driver: the driver either has an outstanding TSEND/TRECEIVE/TRSP WQE or no outstanding WQEs are present for the exchange / context. lpfc_nvmet_xmt_fcp_abort if (LPFC_NVMET_IO_INP) lpfc_nvmet_sol_fcp_issue_abort (ABORT_WQE) lpfc_nvmet_sol_fcp_abort_cmp else lpfc_nvmet_unsol_fcp_issue_abort lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE) lpfc_nvmet_unsol_fcp_abort_cmp Context flags: LPFC_NVMET_IOP - his flag signifies an IO is in progress on the exchange. LPFC_NVMET_XBUSY - this flag indicates the IO completed but the firmware is still busy with the corresponding exchange. The exchange should not be reused until after a XRI_ABORTED_CQE is received for that exchange. LPFC_NVMET_ABORT_OP - this flag signifies an ABORT_WQE was issued on the exchange. LPFC_NVMET_CTX_RLS - this flag signifies a context free was requested, but we are deferring it due to an XBUSY or ABORT in progress. A ctxlock is added to the context structure that is used whenever these flags are set/read within the context of an IO. The LPFC_NVMET_CTX_RLS flag is only set in the defer_relase routine when the transport has resolved all IO associated with the buffer. The flag is cleared when the CTX is associated with a new IO. An exchange can has both an LPFC_NVMET_XBUSY and a LPFC_NVMET_ABORT_OP condition active simultaneously. Both conditions must complete before the exchange is freed. When the abort callback (lpfc_nvmet_xmt_fcp_abort) is envoked: If there is an outstanding IO, the driver will issue an ABORT_WQE. This should result in 3 completions for the exchange: 1) IO cmpl with XB bit set 2) Abort WQE cmpl 3) XRI_ABORTED_CQE cmpl For this scenerio, after completion #1, the NVMET Transport IO rsp callback is called. After completion #2, no action is taken with respect to the exchange / context. After completion #3, the exchange context is free for re-use on another IO. If there is no outstanding activity on the exchange, the driver will send a ABTS to the Initiator. Upon completion of this WQE, the exchange / context is freed for re-use on another IO. Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <james.smart@broadcom.com> Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
2017-04-22 07:05:04 +08:00
/* iaab/iaar NOT set for now */
bf_set(lpfc_mbx_rq_ftr_rq_iaab, &mboxq->u.mqe.un.req_ftrs, 0);
bf_set(lpfc_mbx_rq_ftr_rq_iaar, &mboxq->u.mqe.un.req_ftrs, 0);
Update ABORT processing for NVMET. The driver with nvme had this routine stubbed. Right now XRI_ABORTED_CQE is not handled and the FC NVMET Transport has a new API for the driver. Missing code path, new NVME abort API Update ABORT processing for NVMET There are 3 new FC NVMET Transport API/ template routines for NVMET: lpfc_nvmet_xmt_fcp_release This NVMET template callback routine called to release context associated with an IO This routine is ALWAYS called last, even if the IO was aborted or completed in error. lpfc_nvmet_xmt_fcp_abort This NVMET template callback routine called to abort an exchange that has an IO in progress nvmet_fc_rcv_fcp_req When the lpfc driver receives an ABTS, this NVME FC transport layer callback routine is called. For this case there are 2 paths thru the driver: the driver either has an outstanding exchange / context for the XRI to be aborted or not. If not, a BA_RJT is issued otherwise a BA_ACC NVMET Driver abort paths: There are 2 paths for aborting an IO. The first one is we receive an IO and decide not to process it because of lack of resources. An unsolicated ABTS is immediately sent back to the initiator as a response. lpfc_nvmet_unsol_fcp_buffer lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE) The second one is we sent the IO up to the NVMET transport layer to process, and for some reason the NVME Transport layer decided to abort the IO before it completes all its phases. For this case there are 2 paths thru the driver: the driver either has an outstanding TSEND/TRECEIVE/TRSP WQE or no outstanding WQEs are present for the exchange / context. lpfc_nvmet_xmt_fcp_abort if (LPFC_NVMET_IO_INP) lpfc_nvmet_sol_fcp_issue_abort (ABORT_WQE) lpfc_nvmet_sol_fcp_abort_cmp else lpfc_nvmet_unsol_fcp_issue_abort lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE) lpfc_nvmet_unsol_fcp_abort_cmp Context flags: LPFC_NVMET_IOP - his flag signifies an IO is in progress on the exchange. LPFC_NVMET_XBUSY - this flag indicates the IO completed but the firmware is still busy with the corresponding exchange. The exchange should not be reused until after a XRI_ABORTED_CQE is received for that exchange. LPFC_NVMET_ABORT_OP - this flag signifies an ABORT_WQE was issued on the exchange. LPFC_NVMET_CTX_RLS - this flag signifies a context free was requested, but we are deferring it due to an XBUSY or ABORT in progress. A ctxlock is added to the context structure that is used whenever these flags are set/read within the context of an IO. The LPFC_NVMET_CTX_RLS flag is only set in the defer_relase routine when the transport has resolved all IO associated with the buffer. The flag is cleared when the CTX is associated with a new IO. An exchange can has both an LPFC_NVMET_XBUSY and a LPFC_NVMET_ABORT_OP condition active simultaneously. Both conditions must complete before the exchange is freed. When the abort callback (lpfc_nvmet_xmt_fcp_abort) is envoked: If there is an outstanding IO, the driver will issue an ABORT_WQE. This should result in 3 completions for the exchange: 1) IO cmpl with XB bit set 2) Abort WQE cmpl 3) XRI_ABORTED_CQE cmpl For this scenerio, after completion #1, the NVMET Transport IO rsp callback is called. After completion #2, no action is taken with respect to the exchange / context. After completion #3, the exchange context is free for re-use on another IO. If there is no outstanding activity on the exchange, the driver will send a ABTS to the Initiator. Upon completion of this WQE, the exchange / context is freed for re-use on another IO. Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <james.smart@broadcom.com> Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
2017-04-22 07:05:04 +08:00
}
return;
}
/**
* lpfc_init_vfi - Initialize the INIT_VFI mailbox command
* @mbox: pointer to lpfc mbox command to initialize.
* @vport: Vport associated with the VF.
*
* This routine initializes @mbox to all zeros and then fills in the mailbox
* fields from @vport. INIT_VFI configures virtual fabrics identified by VFI
* in the context of an FCF. The driver issues this command to setup a VFI
* before issuing a FLOGI to login to the VSAN. The driver should also issue a
* REG_VFI after a successful VSAN login.
**/
void
lpfc_init_vfi(struct lpfcMboxq *mbox, struct lpfc_vport *vport)
{
struct lpfc_mbx_init_vfi *init_vfi;
memset(mbox, 0, sizeof(*mbox));
2010-11-21 12:11:48 +08:00
mbox->vport = vport;
init_vfi = &mbox->u.mqe.un.init_vfi;
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_INIT_VFI);
bf_set(lpfc_init_vfi_vr, init_vfi, 1);
bf_set(lpfc_init_vfi_vt, init_vfi, 1);
2010-11-21 12:11:48 +08:00
bf_set(lpfc_init_vfi_vp, init_vfi, 1);
bf_set(lpfc_init_vfi_vfi, init_vfi,
vport->phba->sli4_hba.vfi_ids[vport->vfi]);
bf_set(lpfc_init_vfi_vpi, init_vfi,
vport->phba->vpi_ids[vport->vpi]);
bf_set(lpfc_init_vfi_fcfi, init_vfi,
vport->phba->fcf.fcfi);
}
/**
* lpfc_reg_vfi - Initialize the REG_VFI mailbox command
* @mbox: pointer to lpfc mbox command to initialize.
* @vport: vport associated with the VF.
* @phys: BDE DMA bus address used to send the service parameters to the HBA.
*
* This routine initializes @mbox to all zeros and then fills in the mailbox
* fields from @vport, and uses @buf as a DMAable buffer to send the vport's
* fc service parameters to the HBA for this VFI. REG_VFI configures virtual
* fabrics identified by VFI in the context of an FCF.
**/
void
lpfc_reg_vfi(struct lpfcMboxq *mbox, struct lpfc_vport *vport, dma_addr_t phys)
{
struct lpfc_mbx_reg_vfi *reg_vfi;
struct lpfc_hba *phba = vport->phba;
uint8_t bbscn_fabric = 0, bbscn_max = 0, bbscn_def = 0;
memset(mbox, 0, sizeof(*mbox));
reg_vfi = &mbox->u.mqe.un.reg_vfi;
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_REG_VFI);
bf_set(lpfc_reg_vfi_vp, reg_vfi, 1);
bf_set(lpfc_reg_vfi_vfi, reg_vfi,
phba->sli4_hba.vfi_ids[vport->vfi]);
bf_set(lpfc_reg_vfi_fcfi, reg_vfi, phba->fcf.fcfi);
bf_set(lpfc_reg_vfi_vpi, reg_vfi, phba->vpi_ids[vport->vpi]);
memcpy(reg_vfi->wwn, &vport->fc_portname, sizeof(struct lpfc_name));
reg_vfi->wwn[0] = cpu_to_le32(reg_vfi->wwn[0]);
reg_vfi->wwn[1] = cpu_to_le32(reg_vfi->wwn[1]);
reg_vfi->e_d_tov = phba->fc_edtov;
reg_vfi->r_a_tov = phba->fc_ratov;
if (phys) {
reg_vfi->bde.addrHigh = putPaddrHigh(phys);
reg_vfi->bde.addrLow = putPaddrLow(phys);
reg_vfi->bde.tus.f.bdeSize = sizeof(vport->fc_sparam);
reg_vfi->bde.tus.f.bdeFlags = BUFF_TYPE_BDE_64;
}
bf_set(lpfc_reg_vfi_nport_id, reg_vfi, vport->fc_myDID);
/* Only FC supports upd bit */
if ((phba->sli4_hba.lnk_info.lnk_tp == LPFC_LNK_TYPE_FC) &&
(vport->fc_flag & FC_VFI_REGISTERED) &&
(!phba->fc_topology_changed))
bf_set(lpfc_reg_vfi_upd, reg_vfi, 1);
bf_set(lpfc_reg_vfi_bbcr, reg_vfi, 0);
bf_set(lpfc_reg_vfi_bbscn, reg_vfi, 0);
bbscn_fabric = (phba->fc_fabparam.cmn.bbRcvSizeMsb >> 4) & 0xF;
if (phba->bbcredit_support && phba->cfg_enable_bbcr &&
bbscn_fabric != 0) {
bbscn_max = bf_get(lpfc_bbscn_max,
&phba->sli4_hba.bbscn_params);
if (bbscn_fabric <= bbscn_max) {
bbscn_def = bf_get(lpfc_bbscn_def,
&phba->sli4_hba.bbscn_params);
if (bbscn_fabric > bbscn_def)
bf_set(lpfc_reg_vfi_bbscn, reg_vfi,
bbscn_fabric);
else
bf_set(lpfc_reg_vfi_bbscn, reg_vfi, bbscn_def);
bf_set(lpfc_reg_vfi_bbcr, reg_vfi, 1);
}
}
lpfc_printf_vlog(vport, KERN_INFO, LOG_MBOX,
"3134 Register VFI, mydid:x%x, fcfi:%d, "
" vfi:%d, vpi:%d, fc_pname:%x%x fc_flag:x%x"
" port_state:x%x topology chg:%d bbscn_fabric :%d\n",
vport->fc_myDID,
phba->fcf.fcfi,
phba->sli4_hba.vfi_ids[vport->vfi],
phba->vpi_ids[vport->vpi],
reg_vfi->wwn[0], reg_vfi->wwn[1], vport->fc_flag,
vport->port_state, phba->fc_topology_changed,
bbscn_fabric);
}
/**
* lpfc_init_vpi - Initialize the INIT_VPI mailbox command
* @phba: pointer to the hba structure to init the VPI for.
* @mbox: pointer to lpfc mbox command to initialize.
* @vpi: VPI to be initialized.
*
* The INIT_VPI mailbox command supports virtual N_Ports. The driver uses the
* command to activate a virtual N_Port. The HBA assigns a MAC address to use
* with the virtual N Port. The SLI Host issues this command before issuing a
* FDISC to connect to the Fabric. The SLI Host should issue a REG_VPI after a
* successful virtual NPort login.
**/
void
lpfc_init_vpi(struct lpfc_hba *phba, struct lpfcMboxq *mbox, uint16_t vpi)
{
memset(mbox, 0, sizeof(*mbox));
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_INIT_VPI);
bf_set(lpfc_init_vpi_vpi, &mbox->u.mqe.un.init_vpi,
phba->vpi_ids[vpi]);
bf_set(lpfc_init_vpi_vfi, &mbox->u.mqe.un.init_vpi,
phba->sli4_hba.vfi_ids[phba->pport->vfi]);
}
/**
* lpfc_unreg_vfi - Initialize the UNREG_VFI mailbox command
* @mbox: pointer to lpfc mbox command to initialize.
* @vport: vport associated with the VF.
*
* The UNREG_VFI mailbox command causes the SLI Host to put a virtual fabric
* (logical NPort) into the inactive state. The SLI Host must have logged out
* and unregistered all remote N_Ports to abort any activity on the virtual
* fabric. The SLI Port posts the mailbox response after marking the virtual
* fabric inactive.
**/
void
lpfc_unreg_vfi(struct lpfcMboxq *mbox, struct lpfc_vport *vport)
{
memset(mbox, 0, sizeof(*mbox));
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_UNREG_VFI);
bf_set(lpfc_unreg_vfi_vfi, &mbox->u.mqe.un.unreg_vfi,
vport->phba->sli4_hba.vfi_ids[vport->vfi]);
}
/**
* lpfc_sli4_dump_cfg_rg23 - Dump sli4 port config region 23
* @phba: pointer to the hba structure containing.
* @mbox: pointer to lpfc mbox command to initialize.
*
* This function create a SLI4 dump mailbox command to dump configure
* region 23.
**/
int
lpfc_sli4_dump_cfg_rg23(struct lpfc_hba *phba, struct lpfcMboxq *mbox)
{
struct lpfc_dmabuf *mp = NULL;
MAILBOX_t *mb;
memset(mbox, 0, sizeof(*mbox));
mb = &mbox->u.mb;
mp = kmalloc(sizeof(struct lpfc_dmabuf), GFP_KERNEL);
if (mp)
mp->virt = lpfc_mbuf_alloc(phba, 0, &mp->phys);
if (!mp || !mp->virt) {
kfree(mp);
/* dump config region 23 failed to allocate memory */
lpfc_printf_log(phba, KERN_WARNING, LOG_MBOX,
"2569 lpfc dump config region 23: memory"
" allocation failed\n");
return 1;
}
memset(mp->virt, 0, LPFC_BPL_SIZE);
INIT_LIST_HEAD(&mp->list);
/* save address for completion */
mbox->ctx_buf = (uint8_t *)mp;
mb->mbxCommand = MBX_DUMP_MEMORY;
mb->un.varDmp.type = DMP_NV_PARAMS;
mb->un.varDmp.region_id = DMP_REGION_23;
mb->un.varDmp.sli4_length = DMP_RGN23_SIZE;
mb->un.varWords[3] = putPaddrLow(mp->phys);
mb->un.varWords[4] = putPaddrHigh(mp->phys);
return 0;
}
scsi: lpfc: Mark symbols static where possible We get a few warnings when building kernel with W=1: drivers/scsi/lpfc/lpfc_sli.c:5693:1: warning: no previous prototype for 'lpfc_set_features' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_sli.c:8972:1: warning: no previous prototype for 'lpfc_sli_calc_ring' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4621:1: warning: no previous prototype for 'lpfc_rdp_res_link_service' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4633:1: warning: no previous prototype for 'lpfc_rdp_res_sfp_desc' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4698:1: warning: no previous prototype for 'lpfc_rdp_res_link_error' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4727:1: warning: no previous prototype for 'lpfc_rdp_res_bbc_desc' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4752:1: warning: no previous prototype for 'lpfc_rdp_res_oed_temp_desc' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4780:1: warning: no previous prototype for 'lpfc_rdp_res_oed_voltage_desc' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4809:1: warning: no previous prototype for 'lpfc_rdp_res_oed_txbias_desc' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4838:1: warning: no previous prototype for 'lpfc_rdp_res_oed_txpower_desc' [-Wmissing-prototypes] .... In fact, these functions are only used in the file in which they are declared and don't need a declaration, but can be made static. So this patch marks these functions with 'static'. Signed-off-by: Baoyou Xie <baoyou.xie@linaro.org> Acked-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2016-09-25 13:44:55 +08:00
static void
lpfc_mbx_cmpl_rdp_link_stat(struct lpfc_hba *phba, LPFC_MBOXQ_t *mboxq)
{
MAILBOX_t *mb;
int rc = FAILURE;
struct lpfc_rdp_context *rdp_context =
(struct lpfc_rdp_context *)(mboxq->ctx_ndlp);
mb = &mboxq->u.mb;
if (mb->mbxStatus)
goto mbx_failed;
memcpy(&rdp_context->link_stat, &mb->un.varRdLnk, sizeof(READ_LNK_VAR));
rc = SUCCESS;
mbx_failed:
lpfc_sli4_mbox_cmd_free(phba, mboxq);
rdp_context->cmpl(phba, rdp_context, rc);
}
scsi: lpfc: Mark symbols static where possible We get a few warnings when building kernel with W=1: drivers/scsi/lpfc/lpfc_sli.c:5693:1: warning: no previous prototype for 'lpfc_set_features' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_sli.c:8972:1: warning: no previous prototype for 'lpfc_sli_calc_ring' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4621:1: warning: no previous prototype for 'lpfc_rdp_res_link_service' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4633:1: warning: no previous prototype for 'lpfc_rdp_res_sfp_desc' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4698:1: warning: no previous prototype for 'lpfc_rdp_res_link_error' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4727:1: warning: no previous prototype for 'lpfc_rdp_res_bbc_desc' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4752:1: warning: no previous prototype for 'lpfc_rdp_res_oed_temp_desc' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4780:1: warning: no previous prototype for 'lpfc_rdp_res_oed_voltage_desc' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4809:1: warning: no previous prototype for 'lpfc_rdp_res_oed_txbias_desc' [-Wmissing-prototypes] drivers/scsi/lpfc/lpfc_els.c:4838:1: warning: no previous prototype for 'lpfc_rdp_res_oed_txpower_desc' [-Wmissing-prototypes] .... In fact, these functions are only used in the file in which they are declared and don't need a declaration, but can be made static. So this patch marks these functions with 'static'. Signed-off-by: Baoyou Xie <baoyou.xie@linaro.org> Acked-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2016-09-25 13:44:55 +08:00
static void
lpfc_mbx_cmpl_rdp_page_a2(struct lpfc_hba *phba, LPFC_MBOXQ_t *mbox)
{
struct lpfc_dmabuf *mp = (struct lpfc_dmabuf *)mbox->ctx_buf;
struct lpfc_rdp_context *rdp_context =
(struct lpfc_rdp_context *)(mbox->ctx_ndlp);
if (bf_get(lpfc_mqe_status, &mbox->u.mqe))
goto error_mbuf_free;
lpfc_sli_bemem_bcopy(mp->virt, &rdp_context->page_a2,
DMP_SFF_PAGE_A2_SIZE);
/* We don't need dma buffer for link stat. */
lpfc_mbuf_free(phba, mp->virt, mp->phys);
kfree(mp);
memset(mbox, 0, sizeof(*mbox));
lpfc_read_lnk_stat(phba, mbox);
mbox->vport = rdp_context->ndlp->vport;
mbox->mbox_cmpl = lpfc_mbx_cmpl_rdp_link_stat;
mbox->ctx_ndlp = (struct lpfc_rdp_context *)rdp_context;
if (lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT) == MBX_NOT_FINISHED)
goto error_cmd_free;
return;
error_mbuf_free:
lpfc_mbuf_free(phba, mp->virt, mp->phys);
kfree(mp);
error_cmd_free:
lpfc_sli4_mbox_cmd_free(phba, mbox);
rdp_context->cmpl(phba, rdp_context, FAILURE);
}
void
lpfc_mbx_cmpl_rdp_page_a0(struct lpfc_hba *phba, LPFC_MBOXQ_t *mbox)
{
int rc;
struct lpfc_dmabuf *mp = (struct lpfc_dmabuf *)(mbox->ctx_buf);
struct lpfc_rdp_context *rdp_context =
(struct lpfc_rdp_context *)(mbox->ctx_ndlp);
if (bf_get(lpfc_mqe_status, &mbox->u.mqe))
goto error;
lpfc_sli_bemem_bcopy(mp->virt, &rdp_context->page_a0,
DMP_SFF_PAGE_A0_SIZE);
memset(mbox, 0, sizeof(*mbox));
memset(mp->virt, 0, DMP_SFF_PAGE_A2_SIZE);
INIT_LIST_HEAD(&mp->list);
/* save address for completion */
mbox->ctx_buf = mp;
mbox->vport = rdp_context->ndlp->vport;
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_DUMP_MEMORY);
bf_set(lpfc_mbx_memory_dump_type3_type,
&mbox->u.mqe.un.mem_dump_type3, DMP_LMSD);
bf_set(lpfc_mbx_memory_dump_type3_link,
&mbox->u.mqe.un.mem_dump_type3, phba->sli4_hba.physical_port);
bf_set(lpfc_mbx_memory_dump_type3_page_no,
&mbox->u.mqe.un.mem_dump_type3, DMP_PAGE_A2);
bf_set(lpfc_mbx_memory_dump_type3_length,
&mbox->u.mqe.un.mem_dump_type3, DMP_SFF_PAGE_A2_SIZE);
mbox->u.mqe.un.mem_dump_type3.addr_lo = putPaddrLow(mp->phys);
mbox->u.mqe.un.mem_dump_type3.addr_hi = putPaddrHigh(mp->phys);
mbox->mbox_cmpl = lpfc_mbx_cmpl_rdp_page_a2;
mbox->ctx_ndlp = (struct lpfc_rdp_context *)rdp_context;
rc = lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT);
if (rc == MBX_NOT_FINISHED)
goto error;
return;
error:
lpfc_mbuf_free(phba, mp->virt, mp->phys);
kfree(mp);
lpfc_sli4_mbox_cmd_free(phba, mbox);
rdp_context->cmpl(phba, rdp_context, FAILURE);
}
/*
* lpfc_sli4_dump_sfp_pagea0 - Dump sli4 read SFP Diagnostic.
* @phba: pointer to the hba structure containing.
* @mbox: pointer to lpfc mbox command to initialize.
*
* This function create a SLI4 dump mailbox command to dump configure
* type 3 page 0xA0.
*/
int
lpfc_sli4_dump_page_a0(struct lpfc_hba *phba, struct lpfcMboxq *mbox)
{
struct lpfc_dmabuf *mp = NULL;
memset(mbox, 0, sizeof(*mbox));
mp = kmalloc(sizeof(struct lpfc_dmabuf), GFP_KERNEL);
if (mp)
mp->virt = lpfc_mbuf_alloc(phba, 0, &mp->phys);
if (!mp || !mp->virt) {
kfree(mp);
lpfc_printf_log(phba, KERN_WARNING, LOG_MBOX,
"3569 dump type 3 page 0xA0 allocation failed\n");
return 1;
}
memset(mp->virt, 0, LPFC_BPL_SIZE);
INIT_LIST_HEAD(&mp->list);
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_DUMP_MEMORY);
/* save address for completion */
mbox->ctx_buf = mp;
bf_set(lpfc_mbx_memory_dump_type3_type,
&mbox->u.mqe.un.mem_dump_type3, DMP_LMSD);
bf_set(lpfc_mbx_memory_dump_type3_link,
&mbox->u.mqe.un.mem_dump_type3, phba->sli4_hba.physical_port);
bf_set(lpfc_mbx_memory_dump_type3_page_no,
&mbox->u.mqe.un.mem_dump_type3, DMP_PAGE_A0);
bf_set(lpfc_mbx_memory_dump_type3_length,
&mbox->u.mqe.un.mem_dump_type3, DMP_SFF_PAGE_A0_SIZE);
mbox->u.mqe.un.mem_dump_type3.addr_lo = putPaddrLow(mp->phys);
mbox->u.mqe.un.mem_dump_type3.addr_hi = putPaddrHigh(mp->phys);
return 0;
}
/**
* lpfc_reg_fcfi - Initialize the REG_FCFI mailbox command
* @phba: pointer to the hba structure containing the FCF index and RQ ID.
* @mbox: pointer to lpfc mbox command to initialize.
*
* The REG_FCFI mailbox command supports Fibre Channel Forwarders (FCFs). The
* SLI Host uses the command to activate an FCF after it has acquired FCF
* information via a READ_FCF mailbox command. This mailbox command also is used
* to indicate where received unsolicited frames from this FCF will be sent. By
* default this routine will set up the FCF to forward all unsolicited frames
* the the RQ ID passed in the @phba. This can be overridden by the caller for
* more complicated setups.
**/
void
lpfc_reg_fcfi(struct lpfc_hba *phba, struct lpfcMboxq *mbox)
{
struct lpfc_mbx_reg_fcfi *reg_fcfi;
memset(mbox, 0, sizeof(*mbox));
reg_fcfi = &mbox->u.mqe.un.reg_fcfi;
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_REG_FCFI);
scsi: lpfc: NVME Initiator: Base modifications NVME Initiator: Base modifications This patch adds base modifications for NVME initiator support. The base modifications consist of: - Formal split of SLI3 rings from SLI-4 WQs (sometimes referred to as rings as well) as implementation now widely varies between the two. - Addition of configuration modes: SCSI initiator only; NVME initiator only; NVME target only; and SCSI and NVME initiator. The configuration mode drives overall adapter configuration, offloads enabled, and resource splits. NVME support is only available on SLI-4 devices and newer fw. - Implements the following based on configuration mode: - Exchange resources are split by protocol; Obviously, if only 1 mode, then no split occurs. Default is 50/50. module attribute allows tuning. - Pools and config parameters are separated per-protocol - Each protocol has it's own set of queues, but share interrupt vectors. SCSI: SLI3 devices have few queues and the original style of queue allocation remains. SLI4 devices piggy back on an "io-channel" concept that eventually needs to merge with scsi-mq/blk-mq support (it is underway). For now, the paradigm continues as it existed prior. io channel allocates N msix and N WQs (N=4 default) and either round robins or uses cpu # modulo N for scheduling. A bunch of module parameters allow the configuration to be tuned. NVME (initiator): Allocates an msix per cpu (or whatever pci_alloc_irq_vectors gets) Allocates a WQ per cpu, and maps the WQs to msix on a WQ # modulo msix vector count basis. Module parameters exist to cap/control the config if desired. - Each protocol has its own buffer and dma pools. I apologize for the size of the patch. Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <james.smart@broadcom.com> ---- Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2017-02-13 05:52:30 +08:00
if (phba->nvmet_support == 0) {
bf_set(lpfc_reg_fcfi_rq_id0, reg_fcfi,
phba->sli4_hba.hdr_rq->queue_id);
/* Match everything - rq_id0 */
bf_set(lpfc_reg_fcfi_type_match0, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_type_mask0, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_rctl_match0, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_rctl_mask0, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_rq_id1, reg_fcfi, REG_FCF_INVALID_QID);
/* addr mode is bit wise inverted value of fcf addr_mode */
bf_set(lpfc_reg_fcfi_mam, reg_fcfi,
(~phba->fcf.addr_mode) & 0x3);
} else {
/* This is ONLY for NVMET MRQ == 1 */
if (phba->cfg_nvmet_mrq != 1)
return;
bf_set(lpfc_reg_fcfi_rq_id0, reg_fcfi,
phba->sli4_hba.nvmet_mrq_hdr[0]->queue_id);
/* Match type FCP - rq_id0 */
bf_set(lpfc_reg_fcfi_type_match0, reg_fcfi, FC_TYPE_FCP);
bf_set(lpfc_reg_fcfi_type_mask0, reg_fcfi, 0xff);
bf_set(lpfc_reg_fcfi_rctl_match0, reg_fcfi,
FC_RCTL_DD_UNSOL_CMD);
bf_set(lpfc_reg_fcfi_rq_id1, reg_fcfi,
phba->sli4_hba.hdr_rq->queue_id);
/* Match everything else - rq_id1 */
bf_set(lpfc_reg_fcfi_type_match1, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_type_mask1, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_rctl_match1, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_rctl_mask1, reg_fcfi, 0);
scsi: lpfc: NVME Initiator: Base modifications NVME Initiator: Base modifications This patch adds base modifications for NVME initiator support. The base modifications consist of: - Formal split of SLI3 rings from SLI-4 WQs (sometimes referred to as rings as well) as implementation now widely varies between the two. - Addition of configuration modes: SCSI initiator only; NVME initiator only; NVME target only; and SCSI and NVME initiator. The configuration mode drives overall adapter configuration, offloads enabled, and resource splits. NVME support is only available on SLI-4 devices and newer fw. - Implements the following based on configuration mode: - Exchange resources are split by protocol; Obviously, if only 1 mode, then no split occurs. Default is 50/50. module attribute allows tuning. - Pools and config parameters are separated per-protocol - Each protocol has it's own set of queues, but share interrupt vectors. SCSI: SLI3 devices have few queues and the original style of queue allocation remains. SLI4 devices piggy back on an "io-channel" concept that eventually needs to merge with scsi-mq/blk-mq support (it is underway). For now, the paradigm continues as it existed prior. io channel allocates N msix and N WQs (N=4 default) and either round robins or uses cpu # modulo N for scheduling. A bunch of module parameters allow the configuration to be tuned. NVME (initiator): Allocates an msix per cpu (or whatever pci_alloc_irq_vectors gets) Allocates a WQ per cpu, and maps the WQs to msix on a WQ # modulo msix vector count basis. Module parameters exist to cap/control the config if desired. - Each protocol has its own buffer and dma pools. I apologize for the size of the patch. Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <james.smart@broadcom.com> ---- Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2017-02-13 05:52:30 +08:00
}
bf_set(lpfc_reg_fcfi_rq_id2, reg_fcfi, REG_FCF_INVALID_QID);
bf_set(lpfc_reg_fcfi_rq_id3, reg_fcfi, REG_FCF_INVALID_QID);
bf_set(lpfc_reg_fcfi_info_index, reg_fcfi,
phba->fcf.current_rec.fcf_indx);
if (phba->fcf.current_rec.vlan_id != LPFC_FCOE_NULL_VID) {
bf_set(lpfc_reg_fcfi_vv, reg_fcfi, 1);
bf_set(lpfc_reg_fcfi_vlan_tag, reg_fcfi,
phba->fcf.current_rec.vlan_id);
}
}
/**
* lpfc_reg_fcfi_mrq - Initialize the REG_FCFI_MRQ mailbox command
* @phba: pointer to the hba structure containing the FCF index and RQ ID.
* @mbox: pointer to lpfc mbox command to initialize.
* @mode: 0 to register FCFI, 1 to register MRQs
*
* The REG_FCFI_MRQ mailbox command supports Fibre Channel Forwarders (FCFs).
* The SLI Host uses the command to activate an FCF after it has acquired FCF
* information via a READ_FCF mailbox command. This mailbox command also is used
* to indicate where received unsolicited frames from this FCF will be sent. By
* default this routine will set up the FCF to forward all unsolicited frames
* the the RQ ID passed in the @phba. This can be overridden by the caller for
* more complicated setups.
**/
void
lpfc_reg_fcfi_mrq(struct lpfc_hba *phba, struct lpfcMboxq *mbox, int mode)
{
struct lpfc_mbx_reg_fcfi_mrq *reg_fcfi;
/* This is ONLY for MRQ */
if (phba->cfg_nvmet_mrq <= 1)
return;
memset(mbox, 0, sizeof(*mbox));
reg_fcfi = &mbox->u.mqe.un.reg_fcfi_mrq;
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_REG_FCFI_MRQ);
if (mode == 0) {
bf_set(lpfc_reg_fcfi_mrq_info_index, reg_fcfi,
phba->fcf.current_rec.fcf_indx);
if (phba->fcf.current_rec.vlan_id != LPFC_FCOE_NULL_VID) {
bf_set(lpfc_reg_fcfi_mrq_vv, reg_fcfi, 1);
bf_set(lpfc_reg_fcfi_mrq_vlan_tag, reg_fcfi,
phba->fcf.current_rec.vlan_id);
}
return;
}
bf_set(lpfc_reg_fcfi_mrq_rq_id0, reg_fcfi,
phba->sli4_hba.nvmet_mrq_hdr[0]->queue_id);
/* Match NVME frames of type FCP (protocol NVME) - rq_id0 */
bf_set(lpfc_reg_fcfi_mrq_type_match0, reg_fcfi, FC_TYPE_FCP);
bf_set(lpfc_reg_fcfi_mrq_type_mask0, reg_fcfi, 0xff);
bf_set(lpfc_reg_fcfi_mrq_rctl_match0, reg_fcfi, FC_RCTL_DD_UNSOL_CMD);
bf_set(lpfc_reg_fcfi_mrq_rctl_mask0, reg_fcfi, 0xff);
bf_set(lpfc_reg_fcfi_mrq_ptc0, reg_fcfi, 1);
bf_set(lpfc_reg_fcfi_mrq_pt0, reg_fcfi, 1);
bf_set(lpfc_reg_fcfi_mrq_policy, reg_fcfi, 3); /* NVME connection id */
bf_set(lpfc_reg_fcfi_mrq_mode, reg_fcfi, 1);
bf_set(lpfc_reg_fcfi_mrq_filter, reg_fcfi, 1); /* rq_id0 */
bf_set(lpfc_reg_fcfi_mrq_npairs, reg_fcfi, phba->cfg_nvmet_mrq);
bf_set(lpfc_reg_fcfi_mrq_rq_id1, reg_fcfi,
phba->sli4_hba.hdr_rq->queue_id);
/* Match everything - rq_id1 */
bf_set(lpfc_reg_fcfi_mrq_type_match1, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_mrq_type_mask1, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_mrq_rctl_match1, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_mrq_rctl_mask1, reg_fcfi, 0);
bf_set(lpfc_reg_fcfi_mrq_rq_id2, reg_fcfi, REG_FCF_INVALID_QID);
bf_set(lpfc_reg_fcfi_mrq_rq_id3, reg_fcfi, REG_FCF_INVALID_QID);
}
/**
* lpfc_unreg_fcfi - Initialize the UNREG_FCFI mailbox command
* @mbox: pointer to lpfc mbox command to initialize.
* @fcfi: FCFI to be unregistered.
*
* The UNREG_FCFI mailbox command supports Fibre Channel Forwarders (FCFs).
* The SLI Host uses the command to inactivate an FCFI.
**/
void
lpfc_unreg_fcfi(struct lpfcMboxq *mbox, uint16_t fcfi)
{
memset(mbox, 0, sizeof(*mbox));
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_UNREG_FCFI);
bf_set(lpfc_unreg_fcfi, &mbox->u.mqe.un.unreg_fcfi, fcfi);
}
/**
* lpfc_resume_rpi - Initialize the RESUME_RPI mailbox command
* @mbox: pointer to lpfc mbox command to initialize.
* @ndlp: The nodelist structure that describes the RPI to resume.
*
* The RESUME_RPI mailbox command is used to restart I/O to an RPI after a
* link event.
**/
void
lpfc_resume_rpi(struct lpfcMboxq *mbox, struct lpfc_nodelist *ndlp)
{
struct lpfc_hba *phba = ndlp->phba;
struct lpfc_mbx_resume_rpi *resume_rpi;
memset(mbox, 0, sizeof(*mbox));
resume_rpi = &mbox->u.mqe.un.resume_rpi;
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_RESUME_RPI);
bf_set(lpfc_resume_rpi_index, resume_rpi,
phba->sli4_hba.rpi_ids[ndlp->nlp_rpi]);
bf_set(lpfc_resume_rpi_ii, resume_rpi, RESUME_INDEX_RPI);
resume_rpi->event_tag = ndlp->phba->fc_eventTag;
}
/**
* lpfc_supported_pages - Initialize the PORT_CAPABILITIES supported pages
* mailbox command.
* @mbox: pointer to lpfc mbox command to initialize.
*
* The PORT_CAPABILITIES supported pages mailbox command is issued to
* retrieve the particular feature pages supported by the port.
**/
void
lpfc_supported_pages(struct lpfcMboxq *mbox)
{
struct lpfc_mbx_supp_pages *supp_pages;
memset(mbox, 0, sizeof(*mbox));
supp_pages = &mbox->u.mqe.un.supp_pages;
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_PORT_CAPABILITIES);
bf_set(cpn, supp_pages, LPFC_SUPP_PAGES);
}
/**
* lpfc_pc_sli4_params - Initialize the PORT_CAPABILITIES SLI4 Params mbox cmd.
* @mbox: pointer to lpfc mbox command to initialize.
*
* The PORT_CAPABILITIES SLI4 parameters mailbox command is issued to
* retrieve the particular SLI4 features supported by the port.
**/
void
lpfc_pc_sli4_params(struct lpfcMboxq *mbox)
{
struct lpfc_mbx_pc_sli4_params *sli4_params;
memset(mbox, 0, sizeof(*mbox));
sli4_params = &mbox->u.mqe.un.sli4_params;
bf_set(lpfc_mqe_command, &mbox->u.mqe, MBX_PORT_CAPABILITIES);
bf_set(cpn, sli4_params, LPFC_SLI4_PARAMETERS);
}