linux_dsm_epyc7002/drivers/scsi/lpfc/lpfc_els.c

6362 lines
203 KiB
C
Raw Normal View History

/*******************************************************************
* This file is part of the Emulex Linux Device Driver for *
* Fibre Channel Host Bus Adapters. *
* Copyright (C) 2004-2008 Emulex. All rights reserved. *
* EMULEX and SLI are trademarks of Emulex. *
* www.emulex.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. *
*******************************************************************/
[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 13:52:54 +07:00
/* See Fibre Channel protocol T11 FC-LS for details */
#include <linux/blkdev.h>
#include <linux/pci.h>
#include <linux/interrupt.h>
#include <scsi/scsi.h>
#include <scsi/scsi_device.h>
#include <scsi/scsi_host.h>
#include <scsi/scsi_transport_fc.h>
#include "lpfc_hw.h"
#include "lpfc_sli.h"
#include "lpfc_disc.h"
#include "lpfc_scsi.h"
#include "lpfc.h"
#include "lpfc_logmsg.h"
#include "lpfc_crtn.h"
#include "lpfc_vport.h"
#include "lpfc_debugfs.h"
static int lpfc_els_retry(struct lpfc_hba *, struct lpfc_iocbq *,
struct lpfc_iocbq *);
static void lpfc_cmpl_fabric_iocb(struct lpfc_hba *, struct lpfc_iocbq *,
struct lpfc_iocbq *);
static void lpfc_fabric_abort_vport(struct lpfc_vport *vport);
static int lpfc_issue_els_fdisc(struct lpfc_vport *vport,
struct lpfc_nodelist *ndlp, uint8_t retry);
static int lpfc_issue_fabric_iocb(struct lpfc_hba *phba,
struct lpfc_iocbq *iocb);
static void lpfc_register_new_vport(struct lpfc_hba *phba,
struct lpfc_vport *vport,
struct lpfc_nodelist *ndlp);
static int lpfc_max_els_tries = 3;
/**
* lpfc_els_chk_latt: Check host link attention event for a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine checks whether there is an outstanding host link
* attention event during the discovery process with the @vport. It is done
* by reading the HBA's Host Attention (HA) register. If there is any host
* link attention events during this @vport's discovery process, the @vport
* shall be marked as FC_ABORT_DISCOVERY, a host link attention clear shall
* be issued if the link state is not already in host link cleared state,
* and a return code shall indicate whether the host link attention event
* had happened.
*
* Note that, if either the host link is in state LPFC_LINK_DOWN or @vport
* state in LPFC_VPORT_READY, the request for checking host link attention
* event will be ignored and a return code shall indicate no host link
* attention event had happened.
*
* Return codes
* 0 - no host link attention event happened
* 1 - host link attention event happened
**/
int
lpfc_els_chk_latt(struct lpfc_vport *vport)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
uint32_t ha_copy;
if (vport->port_state >= LPFC_VPORT_READY ||
phba->link_state == LPFC_LINK_DOWN)
return 0;
/* Read the HBA Host Attention Register */
ha_copy = readl(phba->HAregaddr);
if (!(ha_copy & HA_LATT))
return 0;
/* Pending Link Event during Discovery */
lpfc_printf_vlog(vport, KERN_ERR, LOG_DISCOVERY,
"0237 Pending Link Event during "
"Discovery: State x%x\n",
phba->pport->port_state);
/* CLEAR_LA should re-enable link attention events and
* we should then imediately take a LATT event. The
* LATT processing should call lpfc_linkdown() which
* will cleanup any left over in-progress discovery
* events.
*/
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_ABORT_DISCOVERY;
spin_unlock_irq(shost->host_lock);
if (phba->link_state != LPFC_CLEAR_LA)
lpfc_issue_clear_la(phba, vport);
return 1;
}
/**
* lpfc_prep_els_iocb: Allocate and prepare a lpfc iocb data structure.
* @vport: pointer to a host virtual N_Port data structure.
* @expectRsp: flag indicating whether response is expected.
* @cmdSize: size of the ELS command.
* @retry: number of retries to the command IOCB when it fails.
* @ndlp: pointer to a node-list data structure.
* @did: destination identifier.
* @elscmd: the ELS command code.
*
* This routine is used for allocating a lpfc-IOCB data structure from
* the driver lpfc-IOCB free-list and prepare the IOCB with the parameters
* passed into the routine for discovery state machine to issue an Extended
* Link Service (ELS) commands. It is a generic lpfc-IOCB allocation
* and preparation routine that is used by all the discovery state machine
* routines and the ELS command-specific fields will be later set up by
* the individual discovery machine routines after calling this routine
* allocating and preparing a generic IOCB data structure. It fills in the
* Buffer Descriptor Entries (BDEs), allocates buffers for both command
* payload and response payload (if expected). The reference count on the
* ndlp is incremented by 1 and the reference to the ndlp is put into
* context1 of the IOCB data structure for this IOCB to hold the ndlp
* reference for the command's callback function to access later.
*
* Return code
* Pointer to the newly allocated/prepared els iocb data structure
* NULL - when els iocb data structure allocation/preparation failed
**/
static struct lpfc_iocbq *
lpfc_prep_els_iocb(struct lpfc_vport *vport, uint8_t expectRsp,
uint16_t cmdSize, uint8_t retry,
struct lpfc_nodelist *ndlp, uint32_t did,
uint32_t elscmd)
{
struct lpfc_hba *phba = vport->phba;
struct lpfc_iocbq *elsiocb;
struct lpfc_dmabuf *pcmd, *prsp, *pbuflist;
struct ulp_bde64 *bpl;
IOCB_t *icmd;
if (!lpfc_is_link_up(phba))
return NULL;
/* Allocate buffer for command iocb */
elsiocb = lpfc_sli_get_iocbq(phba);
if (elsiocb == NULL)
return NULL;
icmd = &elsiocb->iocb;
/* fill in BDEs for command */
/* Allocate buffer for command payload */
pcmd = kmalloc(sizeof(struct lpfc_dmabuf), GFP_KERNEL);
if (pcmd)
pcmd->virt = lpfc_mbuf_alloc(phba, MEM_PRI, &pcmd->phys);
if (!pcmd || !pcmd->virt)
goto els_iocb_free_pcmb_exit;
INIT_LIST_HEAD(&pcmd->list);
/* Allocate buffer for response payload */
if (expectRsp) {
prsp = kmalloc(sizeof(struct lpfc_dmabuf), GFP_KERNEL);
if (prsp)
prsp->virt = lpfc_mbuf_alloc(phba, MEM_PRI,
&prsp->phys);
if (!prsp || !prsp->virt)
goto els_iocb_free_prsp_exit;
INIT_LIST_HEAD(&prsp->list);
} else
prsp = NULL;
/* Allocate buffer for Buffer ptr list */
pbuflist = kmalloc(sizeof(struct lpfc_dmabuf), GFP_KERNEL);
if (pbuflist)
pbuflist->virt = lpfc_mbuf_alloc(phba, MEM_PRI,
&pbuflist->phys);
if (!pbuflist || !pbuflist->virt)
goto els_iocb_free_pbuf_exit;
INIT_LIST_HEAD(&pbuflist->list);
icmd->un.elsreq64.bdl.addrHigh = putPaddrHigh(pbuflist->phys);
icmd->un.elsreq64.bdl.addrLow = putPaddrLow(pbuflist->phys);
icmd->un.elsreq64.bdl.bdeFlags = BUFF_TYPE_BDL;
icmd->un.elsreq64.remoteID = did; /* DID */
if (expectRsp) {
icmd->un.elsreq64.bdl.bdeSize = (2 * sizeof(struct ulp_bde64));
icmd->ulpCommand = CMD_ELS_REQUEST64_CR;
icmd->ulpTimeout = phba->fc_ratov * 2;
} else {
icmd->un.elsreq64.bdl.bdeSize = sizeof(struct ulp_bde64);
icmd->ulpCommand = CMD_XMIT_ELS_RSP64_CX;
}
icmd->ulpBdeCount = 1;
icmd->ulpLe = 1;
icmd->ulpClass = CLASS3;
if (phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) {
icmd->un.elsreq64.myID = vport->fc_myDID;
/* For ELS_REQUEST64_CR, use the VPI by default */
icmd->ulpContext = vport->vpi;
icmd->ulpCt_h = 0;
icmd->ulpCt_l = 1;
}
bpl = (struct ulp_bde64 *) pbuflist->virt;
bpl->addrLow = le32_to_cpu(putPaddrLow(pcmd->phys));
bpl->addrHigh = le32_to_cpu(putPaddrHigh(pcmd->phys));
bpl->tus.f.bdeSize = cmdSize;
bpl->tus.f.bdeFlags = 0;
bpl->tus.w = le32_to_cpu(bpl->tus.w);
if (expectRsp) {
bpl++;
bpl->addrLow = le32_to_cpu(putPaddrLow(prsp->phys));
bpl->addrHigh = le32_to_cpu(putPaddrHigh(prsp->phys));
bpl->tus.f.bdeSize = FCELSSIZE;
bpl->tus.f.bdeFlags = BUFF_USE_RCV;
bpl->tus.w = le32_to_cpu(bpl->tus.w);
}
/* prevent preparing iocb with NULL ndlp reference */
elsiocb->context1 = lpfc_nlp_get(ndlp);
if (!elsiocb->context1)
goto els_iocb_free_pbuf_exit;
elsiocb->context2 = pcmd;
elsiocb->context3 = pbuflist;
elsiocb->retry = retry;
elsiocb->vport = vport;
elsiocb->drvrTimeout = (phba->fc_ratov << 1) + LPFC_DRVR_TIMEOUT;
if (prsp) {
list_add(&prsp->list, &pcmd->list);
}
if (expectRsp) {
/* Xmit ELS command <elsCmd> to remote NPORT <did> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0116 Xmit ELS command x%x to remote "
"NPORT x%x I/O tag: x%x, port state: x%x\n",
elscmd, did, elsiocb->iotag,
vport->port_state);
} else {
/* Xmit ELS response <elsCmd> to remote NPORT <did> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0117 Xmit ELS response x%x to remote "
"NPORT x%x I/O tag: x%x, size: x%x\n",
elscmd, ndlp->nlp_DID, elsiocb->iotag,
cmdSize);
}
return elsiocb;
els_iocb_free_pbuf_exit:
lpfc_mbuf_free(phba, prsp->virt, prsp->phys);
kfree(pbuflist);
els_iocb_free_prsp_exit:
lpfc_mbuf_free(phba, pcmd->virt, pcmd->phys);
kfree(prsp);
els_iocb_free_pcmb_exit:
kfree(pcmd);
lpfc_sli_release_iocbq(phba, elsiocb);
return NULL;
}
/**
* lpfc_issue_fabric_reglogin: Issue fabric registration login for a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine issues a fabric registration login for a @vport. An
* active ndlp node with Fabric_DID must already exist for this @vport.
* The routine invokes two mailbox commands to carry out fabric registration
* login through the HBA firmware: the first mailbox command requests the
* HBA to perform link configuration for the @vport; and the second mailbox
* command requests the HBA to perform the actual fabric registration login
* with the @vport.
*
* Return code
* 0 - successfully issued fabric registration login for @vport
* -ENXIO -- failed to issue fabric registration login for @vport
**/
static int
lpfc_issue_fabric_reglogin(struct lpfc_vport *vport)
{
struct lpfc_hba *phba = vport->phba;
LPFC_MBOXQ_t *mbox;
struct lpfc_dmabuf *mp;
struct lpfc_nodelist *ndlp;
struct serv_parm *sp;
int rc;
int err = 0;
sp = &phba->fc_fabparam;
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp || !NLP_CHK_NODE_ACT(ndlp)) {
err = 1;
goto fail;
}
mbox = mempool_alloc(phba->mbox_mem_pool, GFP_KERNEL);
if (!mbox) {
err = 2;
goto fail;
}
vport->port_state = LPFC_FABRIC_CFG_LINK;
lpfc_config_link(phba, mbox);
mbox->mbox_cmpl = lpfc_sli_def_mbox_cmpl;
mbox->vport = vport;
rc = lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT);
if (rc == MBX_NOT_FINISHED) {
err = 3;
goto fail_free_mbox;
}
mbox = mempool_alloc(phba->mbox_mem_pool, GFP_KERNEL);
if (!mbox) {
err = 4;
goto fail;
}
rc = lpfc_reg_login(phba, vport->vpi, Fabric_DID, (uint8_t *)sp, mbox,
0);
if (rc) {
err = 5;
goto fail_free_mbox;
}
mbox->mbox_cmpl = lpfc_mbx_cmpl_fabric_reg_login;
mbox->vport = vport;
/* increment the reference count on ndlp to hold reference
* for the callback routine.
*/
mbox->context2 = lpfc_nlp_get(ndlp);
rc = lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT);
if (rc == MBX_NOT_FINISHED) {
err = 6;
goto fail_issue_reg_login;
}
return 0;
fail_issue_reg_login:
/* decrement the reference count on ndlp just incremented
* for the failed mbox command.
*/
lpfc_nlp_put(ndlp);
mp = (struct lpfc_dmabuf *) mbox->context1;
lpfc_mbuf_free(phba, mp->virt, mp->phys);
kfree(mp);
fail_free_mbox:
mempool_free(mbox, phba->mbox_mem_pool);
fail:
lpfc_vport_set_state(vport, FC_VPORT_FAILED);
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0249 Cannot issue Register Fabric login: Err %d\n", err);
return -ENXIO;
}
/**
* lpfc_cmpl_els_flogi_fabric: Completion function for flogi to a fabric port.
* @vport: pointer to a host virtual N_Port data structure.
* @ndlp: pointer to a node-list data structure.
* @sp: pointer to service parameter data structure.
* @irsp: pointer to the IOCB within the lpfc response IOCB.
*
* This routine is invoked by the lpfc_cmpl_els_flogi() completion callback
* function to handle the completion of a Fabric Login (FLOGI) into a fabric
* port in a fabric topology. It properly sets up the parameters to the @ndlp
* from the IOCB response. It also check the newly assigned N_Port ID to the
* @vport against the previously assigned N_Port ID. If it is different from
* the previously assigned Destination ID (DID), the lpfc_unreg_rpi() routine
* is invoked on all the remaining nodes with the @vport to unregister the
* Remote Port Indicators (RPIs). Finally, the lpfc_issue_fabric_reglogin()
* is invoked to register login to the fabric.
*
* Return code
* 0 - Success (currently, always return 0)
**/
static int
lpfc_cmpl_els_flogi_fabric(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
struct serv_parm *sp, IOCB_t *irsp)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
struct lpfc_nodelist *np;
struct lpfc_nodelist *next_np;
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_FABRIC;
spin_unlock_irq(shost->host_lock);
phba->fc_edtov = be32_to_cpu(sp->cmn.e_d_tov);
if (sp->cmn.edtovResolution) /* E_D_TOV ticks are in nanoseconds */
phba->fc_edtov = (phba->fc_edtov + 999999) / 1000000;
phba->fc_ratov = (be32_to_cpu(sp->cmn.w2.r_a_tov) + 999) / 1000;
if (phba->fc_topology == TOPOLOGY_LOOP) {
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_PUBLIC_LOOP;
spin_unlock_irq(shost->host_lock);
} else {
/*
* If we are a N-port connected to a Fabric, fixup sparam's so
* logins to devices on remote loops work.
*/
vport->fc_sparam.cmn.altBbCredit = 1;
}
vport->fc_myDID = irsp->un.ulpWord[4] & Mask_DID;
memcpy(&ndlp->nlp_portname, &sp->portName, sizeof(struct lpfc_name));
memcpy(&ndlp->nlp_nodename, &sp->nodeName, sizeof(struct lpfc_name));
ndlp->nlp_class_sup = 0;
if (sp->cls1.classValid)
ndlp->nlp_class_sup |= FC_COS_CLASS1;
if (sp->cls2.classValid)
ndlp->nlp_class_sup |= FC_COS_CLASS2;
if (sp->cls3.classValid)
ndlp->nlp_class_sup |= FC_COS_CLASS3;
if (sp->cls4.classValid)
ndlp->nlp_class_sup |= FC_COS_CLASS4;
ndlp->nlp_maxframe = ((sp->cmn.bbRcvSizeMsb & 0x0F) << 8) |
sp->cmn.bbRcvSizeLsb;
memcpy(&phba->fc_fabparam, sp, sizeof(struct serv_parm));
if (phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) {
if (sp->cmn.response_multiple_NPort) {
lpfc_printf_vlog(vport, KERN_WARNING,
LOG_ELS | LOG_VPORT,
"1816 FLOGI NPIV supported, "
"response data 0x%x\n",
sp->cmn.response_multiple_NPort);
phba->link_flag |= LS_NPIV_FAB_SUPPORTED;
} else {
/* Because we asked f/w for NPIV it still expects us
to call reg_vnpid atleast for the physcial host */
lpfc_printf_vlog(vport, KERN_WARNING,
LOG_ELS | LOG_VPORT,
"1817 Fabric does not support NPIV "
"- configuring single port mode.\n");
phba->link_flag &= ~LS_NPIV_FAB_SUPPORTED;
}
}
if ((vport->fc_prevDID != vport->fc_myDID) &&
!(vport->fc_flag & FC_VPORT_NEEDS_REG_VPI)) {
/* If our NportID changed, we need to ensure all
* remaining NPORTs get unreg_login'ed.
*/
list_for_each_entry_safe(np, next_np,
&vport->fc_nodes, nlp_listp) {
if (!NLP_CHK_NODE_ACT(ndlp))
continue;
if ((np->nlp_state != NLP_STE_NPR_NODE) ||
!(np->nlp_flag & NLP_NPR_ADISC))
continue;
spin_lock_irq(shost->host_lock);
np->nlp_flag &= ~NLP_NPR_ADISC;
spin_unlock_irq(shost->host_lock);
lpfc_unreg_rpi(vport, np);
}
if (phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) {
lpfc_mbx_unreg_vpi(vport);
[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 13:52:54 +07:00
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_VPORT_NEEDS_REG_VPI;
[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 13:52:54 +07:00
spin_unlock_irq(shost->host_lock);
}
}
lpfc_nlp_set_state(vport, ndlp, NLP_STE_REG_LOGIN_ISSUE);
if (phba->sli3_options & LPFC_SLI3_NPIV_ENABLED &&
vport->fc_flag & FC_VPORT_NEEDS_REG_VPI) {
lpfc_register_new_vport(phba, vport, ndlp);
return 0;
}
lpfc_issue_fabric_reglogin(vport);
return 0;
}
/**
* lpfc_cmpl_els_flogi_nport: Completion function for flogi to an N_Port.
* @vport: pointer to a host virtual N_Port data structure.
* @ndlp: pointer to a node-list data structure.
* @sp: pointer to service parameter data structure.
*
* This routine is invoked by the lpfc_cmpl_els_flogi() completion callback
* function to handle the completion of a Fabric Login (FLOGI) into an N_Port
* in a point-to-point topology. First, the @vport's N_Port Name is compared
* with the received N_Port Name: if the @vport's N_Port Name is greater than
* the received N_Port Name lexicographically, this node shall assign local
* N_Port ID (PT2PT_LocalID: 1) and remote N_Port ID (PT2PT_RemoteID: 2) and
* will send out Port Login (PLOGI) with the N_Port IDs assigned. Otherwise,
* this node shall just wait for the remote node to issue PLOGI and assign
* N_Port IDs.
*
* Return code
* 0 - Success
* -ENXIO - Fail
**/
static int
lpfc_cmpl_els_flogi_nport(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
struct serv_parm *sp)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
LPFC_MBOXQ_t *mbox;
int rc;
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~(FC_FABRIC | FC_PUBLIC_LOOP);
spin_unlock_irq(shost->host_lock);
phba->fc_edtov = FF_DEF_EDTOV;
phba->fc_ratov = FF_DEF_RATOV;
rc = memcmp(&vport->fc_portname, &sp->portName,
sizeof(vport->fc_portname));
if (rc >= 0) {
/* This side will initiate the PLOGI */
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_PT2PT_PLOGI;
spin_unlock_irq(shost->host_lock);
/*
* N_Port ID cannot be 0, set our to LocalID the other
* side will be RemoteID.
*/
/* not equal */
if (rc)
vport->fc_myDID = PT2PT_LocalID;
mbox = mempool_alloc(phba->mbox_mem_pool, GFP_KERNEL);
if (!mbox)
goto fail;
lpfc_config_link(phba, mbox);
mbox->mbox_cmpl = lpfc_sli_def_mbox_cmpl;
mbox->vport = vport;
rc = lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT);
if (rc == MBX_NOT_FINISHED) {
mempool_free(mbox, phba->mbox_mem_pool);
goto fail;
}
/* Decrement ndlp reference count indicating that ndlp can be
* safely released when other references to it are done.
*/
lpfc_nlp_put(ndlp);
ndlp = lpfc_findnode_did(vport, PT2PT_RemoteID);
if (!ndlp) {
/*
* Cannot find existing Fabric ndlp, so allocate a
* new one
*/
ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
if (!ndlp)
goto fail;
lpfc_nlp_init(vport, ndlp, PT2PT_RemoteID);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp,
NLP_STE_UNUSED_NODE);
if(!ndlp)
goto fail;
}
memcpy(&ndlp->nlp_portname, &sp->portName,
sizeof(struct lpfc_name));
memcpy(&ndlp->nlp_nodename, &sp->nodeName,
sizeof(struct lpfc_name));
/* Set state will put ndlp onto node list if not already done */
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
} else
/* This side will wait for the PLOGI, decrement ndlp reference
* count indicating that ndlp can be released when other
* references to it are done.
*/
lpfc_nlp_put(ndlp);
[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 13:52:54 +07:00
/* If we are pt2pt with another NPort, force NPIV off! */
phba->sli3_options &= ~LPFC_SLI3_NPIV_ENABLED;
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_PT2PT;
spin_unlock_irq(shost->host_lock);
/* Start discovery - this should just do CLEAR_LA */
lpfc_disc_start(vport);
return 0;
fail:
return -ENXIO;
}
/**
* lpfc_cmpl_els_flogi: Completion callback function for flogi.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is the top-level completion callback function for issuing
* a Fabric Login (FLOGI) command. If the response IOCB reported error,
* the lpfc_els_retry() routine shall be invoked to retry the FLOGI. If
* retry has been made (either immediately or delayed with lpfc_els_retry()
* returning 1), the command IOCB will be released and function returned.
* If the retry attempt has been given up (possibly reach the maximum
* number of retries), one additional decrement of ndlp reference shall be
* invoked before going out after releasing the command IOCB. This will
* actually release the remote node (Note, lpfc_els_free_iocb() will also
* invoke one decrement of ndlp reference count). If no error reported in
* the IOCB status, the command Port ID field is used to determine whether
* this is a point-to-point topology or a fabric topology: if the Port ID
* field is assigned, it is a fabric topology; otherwise, it is a
* point-to-point topology. The routine lpfc_cmpl_els_flogi_fabric() or
* lpfc_cmpl_els_flogi_nport() shall be invoked accordingly to handle the
* specific topology completion conditions.
**/
static void
lpfc_cmpl_els_flogi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_vport *vport = cmdiocb->vport;
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
IOCB_t *irsp = &rspiocb->iocb;
struct lpfc_nodelist *ndlp = cmdiocb->context1;
struct lpfc_dmabuf *pcmd = cmdiocb->context2, *prsp;
struct serv_parm *sp;
int rc;
/* Check to see if link went down during discovery */
if (lpfc_els_chk_latt(vport)) {
/* One additional decrement on node reference count to
* trigger the release of the node
*/
lpfc_nlp_put(ndlp);
goto out;
}
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"FLOGI cmpl: status:x%x/x%x state:x%x",
irsp->ulpStatus, irsp->un.ulpWord[4],
vport->port_state);
if (irsp->ulpStatus) {
/* Check for retry */
if (lpfc_els_retry(phba, cmdiocb, rspiocb))
goto out;
/* FLOGI failed, so there is no fabric */
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~(FC_FABRIC | FC_PUBLIC_LOOP);
spin_unlock_irq(shost->host_lock);
/* If private loop, then allow max outstanding els to be
* LPFC_MAX_DISC_THREADS (32). Scanning in the case of no
* alpa map would take too long otherwise.
*/
if (phba->alpa_map[0] == 0) {
vport->cfg_discovery_threads = LPFC_MAX_DISC_THREADS;
}
/* FLOGI failure */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0100 FLOGI failure Data: x%x x%x "
"x%x\n",
irsp->ulpStatus, irsp->un.ulpWord[4],
irsp->ulpTimeout);
goto flogifail;
}
/*
* The FLogI succeeded. Sync the data for the CPU before
* accessing it.
*/
prsp = list_get_first(&pcmd->list, struct lpfc_dmabuf, list);
sp = prsp->virt + sizeof(uint32_t);
/* FLOGI completes successfully */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0101 FLOGI completes sucessfully "
"Data: x%x x%x x%x x%x\n",
irsp->un.ulpWord[4], sp->cmn.e_d_tov,
sp->cmn.w2.r_a_tov, sp->cmn.edtovResolution);
if (vport->port_state == LPFC_FLOGI) {
/*
* If Common Service Parameters indicate Nport
* we are point to point, if Fport we are Fabric.
*/
if (sp->cmn.fPort)
rc = lpfc_cmpl_els_flogi_fabric(vport, ndlp, sp, irsp);
else
rc = lpfc_cmpl_els_flogi_nport(vport, ndlp, sp);
if (!rc)
goto out;
}
flogifail:
lpfc_nlp_put(ndlp);
if (!lpfc_error_lost_link(irsp)) {
/* FLOGI failed, so just use loop map to make discovery list */
lpfc_disc_list_loopmap(vport);
/* Start discovery */
lpfc_disc_start(vport);
} else if (((irsp->ulpStatus != IOSTAT_LOCAL_REJECT) ||
((irsp->un.ulpWord[4] != IOERR_SLI_ABORTED) &&
(irsp->un.ulpWord[4] != IOERR_SLI_DOWN))) &&
(phba->link_state != LPFC_CLEAR_LA)) {
/* If FLOGI failed enable link interrupt. */
lpfc_issue_clear_la(phba, vport);
}
out:
lpfc_els_free_iocb(phba, cmdiocb);
}
/**
* lpfc_issue_els_flogi: Issue an flogi iocb command for a vport.
* @vport: pointer to a host virtual N_Port data structure.
* @ndlp: pointer to a node-list data structure.
* @retry: number of retries to the command IOCB.
*
* This routine issues a Fabric Login (FLOGI) Request ELS command
* for a @vport. The initiator service parameters are put into the payload
* of the FLOGI Request IOCB and the top-level callback function pointer
* to lpfc_cmpl_els_flogi() routine is put to the IOCB completion callback
* function field. The lpfc_issue_fabric_iocb routine is invoked to send
* out FLOGI ELS command with one outstanding fabric IOCB at a time.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the FLOGI ELS command.
*
* Return code
* 0 - successfully issued flogi iocb for @vport
* 1 - failed to issue flogi iocb for @vport
**/
static int
lpfc_issue_els_flogi(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
uint8_t retry)
{
struct lpfc_hba *phba = vport->phba;
struct serv_parm *sp;
IOCB_t *icmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli_ring *pring;
uint8_t *pcmd;
uint16_t cmdsize;
uint32_t tmo;
int rc;
pring = &phba->sli.ring[LPFC_ELS_RING];
cmdsize = (sizeof(uint32_t) + sizeof(struct serv_parm));
elsiocb = lpfc_prep_els_iocb(vport, 1, cmdsize, retry, ndlp,
ndlp->nlp_DID, ELS_CMD_FLOGI);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
/* For FLOGI request, remainder of payload is service parameters */
*((uint32_t *) (pcmd)) = ELS_CMD_FLOGI;
pcmd += sizeof(uint32_t);
memcpy(pcmd, &vport->fc_sparam, sizeof(struct serv_parm));
sp = (struct serv_parm *) pcmd;
/* Setup CSPs accordingly for Fabric */
sp->cmn.e_d_tov = 0;
sp->cmn.w2.r_a_tov = 0;
sp->cls1.classValid = 0;
sp->cls2.seqDelivery = 1;
sp->cls3.seqDelivery = 1;
if (sp->cmn.fcphLow < FC_PH3)
sp->cmn.fcphLow = FC_PH3;
if (sp->cmn.fcphHigh < FC_PH3)
sp->cmn.fcphHigh = FC_PH3;
if (phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) {
sp->cmn.request_multiple_Nport = 1;
/* For FLOGI, Let FLOGI rsp set the NPortID for VPI 0 */
icmd->ulpCt_h = 1;
icmd->ulpCt_l = 0;
}
if (phba->fc_topology != TOPOLOGY_LOOP) {
icmd->un.elsreq64.myID = 0;
icmd->un.elsreq64.fl = 1;
}
tmo = phba->fc_ratov;
phba->fc_ratov = LPFC_DISC_FLOGI_TMO;
lpfc_set_disctmo(vport);
phba->fc_ratov = tmo;
phba->fc_stat.elsXmitFLOGI++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_flogi;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"Issue FLOGI: opt:x%x",
phba->sli3_options, 0, 0);
rc = lpfc_issue_fabric_iocb(phba, elsiocb);
if (rc == IOCB_ERROR) {
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_els_abort_flogi: Abort all outstanding flogi iocbs.
* @phba: pointer to lpfc hba data structure.
*
* This routine aborts all the outstanding Fabric Login (FLOGI) IOCBs
* with a @phba. This routine walks all the outstanding IOCBs on the txcmplq
* list and issues an abort IOCB commond on each outstanding IOCB that
* contains a active Fabric_DID ndlp. Note that this function is to issue
* the abort IOCB command on all the outstanding IOCBs, thus when this
* function returns, it does not guarantee all the IOCBs are actually aborted.
*
* Return code
* 0 - Sucessfully issued abort iocb on all outstanding flogis (Always 0)
**/
int
lpfc_els_abort_flogi(struct lpfc_hba *phba)
{
struct lpfc_sli_ring *pring;
struct lpfc_iocbq *iocb, *next_iocb;
struct lpfc_nodelist *ndlp;
IOCB_t *icmd;
/* Abort outstanding I/O on NPort <nlp_DID> */
lpfc_printf_log(phba, KERN_INFO, LOG_DISCOVERY,
"0201 Abort outstanding I/O on NPort x%x\n",
Fabric_DID);
pring = &phba->sli.ring[LPFC_ELS_RING];
/*
* Check the txcmplq for an iocb that matches the nport the driver is
* searching for.
*/
spin_lock_irq(&phba->hbalock);
list_for_each_entry_safe(iocb, next_iocb, &pring->txcmplq, list) {
icmd = &iocb->iocb;
if (icmd->ulpCommand == CMD_ELS_REQUEST64_CR &&
icmd->un.elsreq64.bdl.ulpIoTag32) {
ndlp = (struct lpfc_nodelist *)(iocb->context1);
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
if (ndlp && NLP_CHK_NODE_ACT(ndlp) &&
(ndlp->nlp_DID == Fabric_DID))
lpfc_sli_issue_abort_iotag(phba, pring, iocb);
}
}
spin_unlock_irq(&phba->hbalock);
return 0;
}
/**
* lpfc_initial_flogi: Issue an initial fabric login for a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine issues an initial Fabric Login (FLOGI) for the @vport
* specified. It first searches the ndlp with the Fabric_DID (0xfffffe) from
* the @vport's ndlp list. If no such ndlp found, it will create an ndlp and
* put it into the @vport's ndlp list. If an inactive ndlp found on the list,
* it will just be enabled and made active. The lpfc_issue_els_flogi() routine
* is then invoked with the @vport and the ndlp to perform the FLOGI for the
* @vport.
*
* Return code
* 0 - failed to issue initial flogi for @vport
* 1 - successfully issued initial flogi for @vport
**/
int
lpfc_initial_flogi(struct lpfc_vport *vport)
{
struct lpfc_hba *phba = vport->phba;
struct lpfc_nodelist *ndlp;
vport->port_state = LPFC_FLOGI;
lpfc_set_disctmo(vport);
/* First look for the Fabric ndlp */
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
if (!ndlp)
return 0;
lpfc_nlp_init(vport, ndlp, Fabric_DID);
/* Put ndlp onto node list */
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
/* re-setup ndlp without removing from node list */
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
if (!ndlp)
return 0;
}
if (lpfc_issue_els_flogi(vport, ndlp, 0))
/* This decrement of reference count to node shall kick off
* the release of the node.
*/
lpfc_nlp_put(ndlp);
return 1;
}
/**
* lpfc_initial_fdisc: Issue an initial fabric discovery for a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine issues an initial Fabric Discover (FDISC) for the @vport
* specified. It first searches the ndlp with the Fabric_DID (0xfffffe) from
* the @vport's ndlp list. If no such ndlp found, it will create an ndlp and
* put it into the @vport's ndlp list. If an inactive ndlp found on the list,
* it will just be enabled and made active. The lpfc_issue_els_fdisc() routine
* is then invoked with the @vport and the ndlp to perform the FDISC for the
* @vport.
*
* Return code
* 0 - failed to issue initial fdisc for @vport
* 1 - successfully issued initial fdisc for @vport
**/
int
lpfc_initial_fdisc(struct lpfc_vport *vport)
{
struct lpfc_hba *phba = vport->phba;
struct lpfc_nodelist *ndlp;
/* First look for the Fabric ndlp */
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
if (!ndlp)
return 0;
lpfc_nlp_init(vport, ndlp, Fabric_DID);
/* Put ndlp onto node list */
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
/* re-setup ndlp without removing from node list */
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
if (!ndlp)
return 0;
}
if (lpfc_issue_els_fdisc(vport, ndlp, 0)) {
/* decrement node reference count to trigger the release of
* the node.
*/
lpfc_nlp_put(ndlp);
return 0;
}
return 1;
}
/**
* lpfc_more_plogi: Check and issue remaining plogis for a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine checks whether there are more remaining Port Logins
* (PLOGI) to be issued for the @vport. If so, it will invoke the routine
* lpfc_els_disc_plogi() to go through the Node Port Recovery (NPR) nodes
* to issue ELS PLOGIs up to the configured discover threads with the
* @vport (@vport->cfg_discovery_threads). The function also decrement
* the @vport's num_disc_node by 1 if it is not already 0.
**/
void
lpfc_more_plogi(struct lpfc_vport *vport)
{
int sentplogi;
if (vport->num_disc_nodes)
vport->num_disc_nodes--;
/* Continue discovery with <num_disc_nodes> PLOGIs to go */
lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
"0232 Continue discovery with %d PLOGIs to go "
"Data: x%x x%x x%x\n",
vport->num_disc_nodes, vport->fc_plogi_cnt,
vport->fc_flag, vport->port_state);
/* Check to see if there are more PLOGIs to be sent */
if (vport->fc_flag & FC_NLP_MORE)
/* go thru NPR nodes and issue any remaining ELS PLOGIs */
sentplogi = lpfc_els_disc_plogi(vport);
return;
}
/**
* lpfc_plogi_confirm_nport: Confirm pologi wwpn matches stored ndlp.
* @phba: pointer to lpfc hba data structure.
* @prsp: pointer to response IOCB payload.
* @ndlp: pointer to a node-list data structure.
*
* This routine checks and indicates whether the WWPN of an N_Port, retrieved
* from a PLOGI, matches the WWPN that is stored in the @ndlp for that N_POrt.
* The following cases are considered N_Port confirmed:
* 1) The N_Port is a Fabric ndlp; 2) The @ndlp is on vport list and matches
* the WWPN of the N_Port logged into; 3) The @ndlp is not on vport list but
* it does not have WWPN assigned either. If the WWPN is confirmed, the
* pointer to the @ndlp will be returned. If the WWPN is not confirmed:
* 1) if there is a node on vport list other than the @ndlp with the same
* WWPN of the N_Port PLOGI logged into, the lpfc_unreg_rpi() will be invoked
* on that node to release the RPI associated with the node; 2) if there is
* no node found on vport list with the same WWPN of the N_Port PLOGI logged
* into, a new node shall be allocated (or activated). In either case, the
* parameters of the @ndlp shall be copied to the new_ndlp, the @ndlp shall
* be released and the new_ndlp shall be put on to the vport node list and
* its pointer returned as the confirmed node.
*
* Note that before the @ndlp got "released", the keepDID from not-matching
* or inactive "new_ndlp" on the vport node list is assigned to the nlp_DID
* of the @ndlp. This is because the release of @ndlp is actually to put it
* into an inactive state on the vport node list and the vport node list
* management algorithm does not allow two node with a same DID.
*
* Return code
* pointer to the PLOGI N_Port @ndlp
**/
static struct lpfc_nodelist *
lpfc_plogi_confirm_nport(struct lpfc_hba *phba, uint32_t *prsp,
struct lpfc_nodelist *ndlp)
{
struct lpfc_vport *vport = ndlp->vport;
struct lpfc_nodelist *new_ndlp;
struct lpfc_rport_data *rdata;
struct fc_rport *rport;
struct serv_parm *sp;
uint8_t name[sizeof(struct lpfc_name)];
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
uint32_t rc, keepDID = 0;
/* Fabric nodes can have the same WWPN so we don't bother searching
* by WWPN. Just return the ndlp that was given to us.
*/
if (ndlp->nlp_type & NLP_FABRIC)
return ndlp;
sp = (struct serv_parm *) ((uint8_t *) prsp + sizeof(uint32_t));
memset(name, 0, sizeof(struct lpfc_name));
/* Now we find out if the NPort we are logging into, matches the WWPN
* we have for that ndlp. If not, we have some work to do.
*/
new_ndlp = lpfc_findnode_wwpn(vport, &sp->portName);
if (new_ndlp == ndlp && NLP_CHK_NODE_ACT(new_ndlp))
return ndlp;
if (!new_ndlp) {
rc = memcmp(&ndlp->nlp_portname, name,
sizeof(struct lpfc_name));
if (!rc)
return ndlp;
new_ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_ATOMIC);
if (!new_ndlp)
return ndlp;
lpfc_nlp_init(vport, new_ndlp, ndlp->nlp_DID);
} else if (!NLP_CHK_NODE_ACT(new_ndlp)) {
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
rc = memcmp(&ndlp->nlp_portname, name,
sizeof(struct lpfc_name));
if (!rc)
return ndlp;
new_ndlp = lpfc_enable_node(vport, new_ndlp,
NLP_STE_UNUSED_NODE);
if (!new_ndlp)
return ndlp;
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
keepDID = new_ndlp->nlp_DID;
} else
keepDID = new_ndlp->nlp_DID;
lpfc_unreg_rpi(vport, new_ndlp);
new_ndlp->nlp_DID = ndlp->nlp_DID;
new_ndlp->nlp_prev_state = ndlp->nlp_prev_state;
if (ndlp->nlp_flag & NLP_NPR_2B_DISC)
new_ndlp->nlp_flag |= NLP_NPR_2B_DISC;
ndlp->nlp_flag &= ~NLP_NPR_2B_DISC;
/* Set state will put new_ndlp on to node list if not already done */
lpfc_nlp_set_state(vport, new_ndlp, ndlp->nlp_state);
/* Move this back to NPR state */
if (memcmp(&ndlp->nlp_portname, name, sizeof(struct lpfc_name)) == 0) {
/* The new_ndlp is replacing ndlp totally, so we need
* to put ndlp on UNUSED list and try to free it.
*/
/* Fix up the rport accordingly */
rport = ndlp->rport;
if (rport) {
rdata = rport->dd_data;
if (rdata->pnode == ndlp) {
lpfc_nlp_put(ndlp);
ndlp->rport = NULL;
rdata->pnode = lpfc_nlp_get(new_ndlp);
new_ndlp->rport = rport;
}
new_ndlp->nlp_type = ndlp->nlp_type;
}
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
/* We shall actually free the ndlp with both nlp_DID and
* nlp_portname fields equals 0 to avoid any ndlp on the
* nodelist never to be used.
*/
if (ndlp->nlp_DID == 0) {
spin_lock_irq(&phba->ndlp_lock);
NLP_SET_FREE_REQ(ndlp);
spin_unlock_irq(&phba->ndlp_lock);
}
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
/* Two ndlps cannot have the same did on the nodelist */
ndlp->nlp_DID = keepDID;
lpfc_drop_node(vport, ndlp);
}
else {
lpfc_unreg_rpi(vport, ndlp);
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
/* Two ndlps cannot have the same did */
ndlp->nlp_DID = keepDID;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
}
return new_ndlp;
}
/**
* lpfc_end_rscn: Check and handle more rscn for a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine checks whether more Registration State Change
* Notifications (RSCNs) came in while the discovery state machine was in
* the FC_RSCN_MODE. If so, the lpfc_els_handle_rscn() routine will be
* invoked to handle the additional RSCNs for the @vport. Otherwise, the
* FC_RSCN_MODE bit will be cleared with the @vport to mark as the end of
* handling the RSCNs.
**/
void
lpfc_end_rscn(struct lpfc_vport *vport)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
if (vport->fc_flag & FC_RSCN_MODE) {
/*
* Check to see if more RSCNs came in while we were
* processing this one.
*/
if (vport->fc_rscn_id_cnt ||
(vport->fc_flag & FC_RSCN_DISCOVERY) != 0)
lpfc_els_handle_rscn(vport);
else {
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~FC_RSCN_MODE;
spin_unlock_irq(shost->host_lock);
}
}
}
/**
* lpfc_cmpl_els_plogi: Completion callback function for plogi.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is the completion callback function for issuing the Port
* Login (PLOGI) command. For PLOGI completion, there must be an active
* ndlp on the vport node list that matches the remote node ID from the
* PLOGI reponse IOCB. If such ndlp does not exist, the PLOGI is simply
* ignored and command IOCB released. The PLOGI response IOCB status is
* checked for error conditons. If there is error status reported, PLOGI
* retry shall be attempted by invoking the lpfc_els_retry() routine.
* Otherwise, the lpfc_plogi_confirm_nport() routine shall be invoked on
* the ndlp and the NLP_EVT_CMPL_PLOGI state to the Discover State Machine
* (DSM) is set for this PLOGI completion. Finally, it checks whether
* there are additional N_Port nodes with the vport that need to perform
* PLOGI. If so, the lpfc_more_plogi() routine is invoked to issue addition
* PLOGIs.
**/
static void
lpfc_cmpl_els_plogi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_vport *vport = cmdiocb->vport;
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
IOCB_t *irsp;
struct lpfc_nodelist *ndlp;
struct lpfc_dmabuf *prsp;
int disc, rc, did, type;
/* we pass cmdiocb to state machine which needs rspiocb as well */
cmdiocb->context_un.rsp_iocb = rspiocb;
irsp = &rspiocb->iocb;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"PLOGI cmpl: status:x%x/x%x did:x%x",
irsp->ulpStatus, irsp->un.ulpWord[4],
irsp->un.elsreq64.remoteID);
ndlp = lpfc_findnode_did(vport, irsp->un.elsreq64.remoteID);
if (!ndlp || !NLP_CHK_NODE_ACT(ndlp)) {
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0136 PLOGI completes to NPort x%x "
"with no ndlp. Data: x%x x%x x%x\n",
irsp->un.elsreq64.remoteID,
irsp->ulpStatus, irsp->un.ulpWord[4],
irsp->ulpIoTag);
goto out;
}
/* Since ndlp can be freed in the disc state machine, note if this node
* is being used during discovery.
*/
spin_lock_irq(shost->host_lock);
disc = (ndlp->nlp_flag & NLP_NPR_2B_DISC);
ndlp->nlp_flag &= ~NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
rc = 0;
/* PLOGI completes to NPort <nlp_DID> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0102 PLOGI completes to NPort x%x "
"Data: x%x x%x x%x x%x x%x\n",
ndlp->nlp_DID, irsp->ulpStatus, irsp->un.ulpWord[4],
irsp->ulpTimeout, disc, vport->num_disc_nodes);
/* Check to see if link went down during discovery */
if (lpfc_els_chk_latt(vport)) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
goto out;
}
/* ndlp could be freed in DSM, save these values now */
type = ndlp->nlp_type;
did = ndlp->nlp_DID;
if (irsp->ulpStatus) {
/* Check for retry */
if (lpfc_els_retry(phba, cmdiocb, rspiocb)) {
/* ELS command is being retried */
if (disc) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
}
goto out;
}
/* PLOGI failed */
/* Do not call DSM for lpfc_els_abort'ed ELS cmds */
if (lpfc_error_lost_link(irsp))
rc = NLP_STE_FREED_NODE;
else
rc = lpfc_disc_state_machine(vport, ndlp, cmdiocb,
NLP_EVT_CMPL_PLOGI);
} else {
/* Good status, call state machine */
prsp = list_entry(((struct lpfc_dmabuf *)
cmdiocb->context2)->list.next,
struct lpfc_dmabuf, list);
ndlp = lpfc_plogi_confirm_nport(phba, prsp->virt, ndlp);
rc = lpfc_disc_state_machine(vport, ndlp, cmdiocb,
NLP_EVT_CMPL_PLOGI);
}
if (disc && vport->num_disc_nodes) {
/* Check to see if there are more PLOGIs to be sent */
lpfc_more_plogi(vport);
if (vport->num_disc_nodes == 0) {
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~FC_NDISC_ACTIVE;
spin_unlock_irq(shost->host_lock);
lpfc_can_disctmo(vport);
lpfc_end_rscn(vport);
}
}
out:
lpfc_els_free_iocb(phba, cmdiocb);
return;
}
/**
* lpfc_issue_els_plogi: Issue an plogi iocb command for a vport.
* @vport: pointer to a host virtual N_Port data structure.
* @did: destination port identifier.
* @retry: number of retries to the command IOCB.
*
* This routine issues a Port Login (PLOGI) command to a remote N_Port
* (with the @did) for a @vport. Before issuing a PLOGI to a remote N_Port,
* the ndlp with the remote N_Port DID must exist on the @vport's ndlp list.
* This routine constructs the proper feilds of the PLOGI IOCB and invokes
* the lpfc_sli_issue_iocb() routine to send out PLOGI ELS command.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the PLOGI ELS command.
*
* Return code
* 0 - Successfully issued a plogi for @vport
* 1 - failed to issue a plogi for @vport
**/
int
lpfc_issue_els_plogi(struct lpfc_vport *vport, uint32_t did, uint8_t retry)
{
struct lpfc_hba *phba = vport->phba;
struct serv_parm *sp;
IOCB_t *icmd;
struct lpfc_nodelist *ndlp;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli_ring *pring;
struct lpfc_sli *psli;
uint8_t *pcmd;
uint16_t cmdsize;
int ret;
psli = &phba->sli;
pring = &psli->ring[LPFC_ELS_RING]; /* ELS ring */
ndlp = lpfc_findnode_did(vport, did);
if (ndlp && !NLP_CHK_NODE_ACT(ndlp))
ndlp = NULL;
/* If ndlp is not NULL, we will bump the reference count on it */
cmdsize = (sizeof(uint32_t) + sizeof(struct serv_parm));
elsiocb = lpfc_prep_els_iocb(vport, 1, cmdsize, retry, ndlp, did,
ELS_CMD_PLOGI);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
/* For PLOGI request, remainder of payload is service parameters */
*((uint32_t *) (pcmd)) = ELS_CMD_PLOGI;
pcmd += sizeof(uint32_t);
memcpy(pcmd, &vport->fc_sparam, sizeof(struct serv_parm));
sp = (struct serv_parm *) pcmd;
if (sp->cmn.fcphLow < FC_PH_4_3)
sp->cmn.fcphLow = FC_PH_4_3;
if (sp->cmn.fcphHigh < FC_PH3)
sp->cmn.fcphHigh = FC_PH3;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"Issue PLOGI: did:x%x",
did, 0, 0);
phba->fc_stat.elsXmitPLOGI++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_plogi;
ret = lpfc_sli_issue_iocb(phba, pring, elsiocb, 0);
if (ret == IOCB_ERROR) {
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_cmpl_els_prli: Completion callback function for prli.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is the completion callback function for a Process Login
* (PRLI) ELS command. The PRLI response IOCB status is checked for error
* status. If there is error status reported, PRLI retry shall be attempted
* by invoking the lpfc_els_retry() routine. Otherwise, the state
* NLP_EVT_CMPL_PRLI is sent to the Discover State Machine (DSM) for this
* ndlp to mark the PRLI completion.
**/
static void
lpfc_cmpl_els_prli(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_vport *vport = cmdiocb->vport;
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
IOCB_t *irsp;
struct lpfc_sli *psli;
struct lpfc_nodelist *ndlp;
psli = &phba->sli;
/* we pass cmdiocb to state machine which needs rspiocb as well */
cmdiocb->context_un.rsp_iocb = rspiocb;
irsp = &(rspiocb->iocb);
ndlp = (struct lpfc_nodelist *) cmdiocb->context1;
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~NLP_PRLI_SND;
spin_unlock_irq(shost->host_lock);
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"PRLI cmpl: status:x%x/x%x did:x%x",
irsp->ulpStatus, irsp->un.ulpWord[4],
ndlp->nlp_DID);
/* PRLI completes to NPort <nlp_DID> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0103 PRLI completes to NPort x%x "
"Data: x%x x%x x%x x%x\n",
ndlp->nlp_DID, irsp->ulpStatus, irsp->un.ulpWord[4],
irsp->ulpTimeout, vport->num_disc_nodes);
vport->fc_prli_sent--;
/* Check to see if link went down during discovery */
if (lpfc_els_chk_latt(vport))
goto out;
if (irsp->ulpStatus) {
/* Check for retry */
if (lpfc_els_retry(phba, cmdiocb, rspiocb)) {
/* ELS command is being retried */
goto out;
}
/* PRLI failed */
/* Do not call DSM for lpfc_els_abort'ed ELS cmds */
if (lpfc_error_lost_link(irsp))
goto out;
else
lpfc_disc_state_machine(vport, ndlp, cmdiocb,
NLP_EVT_CMPL_PRLI);
} else
/* Good status, call state machine */
lpfc_disc_state_machine(vport, ndlp, cmdiocb,
NLP_EVT_CMPL_PRLI);
out:
lpfc_els_free_iocb(phba, cmdiocb);
return;
}
/**
* lpfc_issue_els_prli: Issue a prli iocb command for a vport.
* @vport: pointer to a host virtual N_Port data structure.
* @ndlp: pointer to a node-list data structure.
* @retry: number of retries to the command IOCB.
*
* This routine issues a Process Login (PRLI) ELS command for the
* @vport. The PRLI service parameters are set up in the payload of the
* PRLI Request command and the pointer to lpfc_cmpl_els_prli() routine
* is put to the IOCB completion callback func field before invoking the
* routine lpfc_sli_issue_iocb() to send out PRLI command.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the PRLI ELS command.
*
* Return code
* 0 - successfully issued prli iocb command for @vport
* 1 - failed to issue prli iocb command for @vport
**/
int
lpfc_issue_els_prli(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
uint8_t retry)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
PRLI *npr;
IOCB_t *icmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli_ring *pring;
struct lpfc_sli *psli;
uint8_t *pcmd;
uint16_t cmdsize;
psli = &phba->sli;
pring = &psli->ring[LPFC_ELS_RING]; /* ELS ring */
cmdsize = (sizeof(uint32_t) + sizeof(PRLI));
elsiocb = lpfc_prep_els_iocb(vport, 1, cmdsize, retry, ndlp,
ndlp->nlp_DID, ELS_CMD_PRLI);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
/* For PRLI request, remainder of payload is service parameters */
memset(pcmd, 0, (sizeof(PRLI) + sizeof(uint32_t)));
*((uint32_t *) (pcmd)) = ELS_CMD_PRLI;
pcmd += sizeof(uint32_t);
/* For PRLI, remainder of payload is PRLI parameter page */
npr = (PRLI *) pcmd;
/*
* If our firmware version is 3.20 or later,
* set the following bits for FC-TAPE support.
*/
if (phba->vpd.rev.feaLevelHigh >= 0x02) {
npr->ConfmComplAllowed = 1;
npr->Retry = 1;
npr->TaskRetryIdReq = 1;
}
npr->estabImagePair = 1;
npr->readXferRdyDis = 1;
/* For FCP support */
npr->prliType = PRLI_FCP_TYPE;
npr->initiatorFunc = 1;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"Issue PRLI: did:x%x",
ndlp->nlp_DID, 0, 0);
phba->fc_stat.elsXmitPRLI++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_prli;
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_PRLI_SND;
spin_unlock_irq(shost->host_lock);
if (lpfc_sli_issue_iocb(phba, pring, elsiocb, 0) == IOCB_ERROR) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~NLP_PRLI_SND;
spin_unlock_irq(shost->host_lock);
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
vport->fc_prli_sent++;
return 0;
}
/**
* lpfc_rscn_disc: Perform rscn discovery for a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine performs Registration State Change Notification (RSCN)
* discovery for a @vport. If the @vport's node port recovery count is not
* zero, it will invoke the lpfc_els_disc_plogi() to perform PLOGI for all
* the nodes that need recovery. If none of the PLOGI were needed through
* the lpfc_els_disc_plogi() routine, the lpfc_end_rscn() routine shall be
* invoked to check and handle possible more RSCN came in during the period
* of processing the current ones.
**/
static void
lpfc_rscn_disc(struct lpfc_vport *vport)
{
lpfc_can_disctmo(vport);
/* RSCN discovery */
/* go thru NPR nodes and issue ELS PLOGIs */
if (vport->fc_npr_cnt)
if (lpfc_els_disc_plogi(vport))
return;
lpfc_end_rscn(vport);
}
/**
* lpfc_adisc_done: Complete the adisc phase of discovery.
* @vport: pointer to lpfc_vport hba data structure that finished all ADISCs.
*
* This function is called when the final ADISC is completed during discovery.
* This function handles clearing link attention or issuing reg_vpi depending
* on whether npiv is enabled. This function also kicks off the PLOGI phase of
* discovery.
* This function is called with no locks held.
**/
static void
lpfc_adisc_done(struct lpfc_vport *vport)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
/*
* For NPIV, cmpl_reg_vpi will set port_state to READY,
* and continue discovery.
*/
if ((phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) &&
!(vport->fc_flag & FC_RSCN_MODE)) {
lpfc_issue_reg_vpi(phba, vport);
return;
}
/*
* For SLI2, we need to set port_state to READY
* and continue discovery.
*/
if (vport->port_state < LPFC_VPORT_READY) {
/* If we get here, there is nothing to ADISC */
if (vport->port_type == LPFC_PHYSICAL_PORT)
lpfc_issue_clear_la(phba, vport);
if (!(vport->fc_flag & FC_ABORT_DISCOVERY)) {
vport->num_disc_nodes = 0;
/* go thru NPR list, issue ELS PLOGIs */
if (vport->fc_npr_cnt)
lpfc_els_disc_plogi(vport);
if (!vport->num_disc_nodes) {
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~FC_NDISC_ACTIVE;
spin_unlock_irq(shost->host_lock);
lpfc_can_disctmo(vport);
lpfc_end_rscn(vport);
}
}
vport->port_state = LPFC_VPORT_READY;
} else
lpfc_rscn_disc(vport);
}
/**
* lpfc_more_adisc: Issue more adisc as needed.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine determines whether there are more ndlps on a @vport
* node list need to have Address Discover (ADISC) issued. If so, it will
* invoke the lpfc_els_disc_adisc() routine to issue ADISC on the @vport's
* remaining nodes which need to have ADISC sent.
**/
void
lpfc_more_adisc(struct lpfc_vport *vport)
{
int sentadisc;
if (vport->num_disc_nodes)
vport->num_disc_nodes--;
/* Continue discovery with <num_disc_nodes> ADISCs to go */
lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
"0210 Continue discovery with %d ADISCs to go "
"Data: x%x x%x x%x\n",
vport->num_disc_nodes, vport->fc_adisc_cnt,
vport->fc_flag, vport->port_state);
/* Check to see if there are more ADISCs to be sent */
if (vport->fc_flag & FC_NLP_MORE) {
lpfc_set_disctmo(vport);
/* go thru NPR nodes and issue any remaining ELS ADISCs */
sentadisc = lpfc_els_disc_adisc(vport);
}
if (!vport->num_disc_nodes)
lpfc_adisc_done(vport);
return;
}
/**
* lpfc_cmpl_els_adisc: Completion callback function for adisc.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is the completion function for issuing the Address Discover
* (ADISC) command. It first checks to see whether link went down during
* the discovery process. If so, the node will be marked as node port
* recovery for issuing discover IOCB by the link attention handler and
* exit. Otherwise, the response status is checked. If error was reported
* in the response status, the ADISC command shall be retried by invoking
* the lpfc_els_retry() routine. Otherwise, if no error was reported in
* the response status, the state machine is invoked to set transition
* with respect to NLP_EVT_CMPL_ADISC event.
**/
static void
lpfc_cmpl_els_adisc(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_vport *vport = cmdiocb->vport;
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
IOCB_t *irsp;
struct lpfc_nodelist *ndlp;
int disc;
/* we pass cmdiocb to state machine which needs rspiocb as well */
cmdiocb->context_un.rsp_iocb = rspiocb;
irsp = &(rspiocb->iocb);
ndlp = (struct lpfc_nodelist *) cmdiocb->context1;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"ADISC cmpl: status:x%x/x%x did:x%x",
irsp->ulpStatus, irsp->un.ulpWord[4],
ndlp->nlp_DID);
/* Since ndlp can be freed in the disc state machine, note if this node
* is being used during discovery.
*/
spin_lock_irq(shost->host_lock);
disc = (ndlp->nlp_flag & NLP_NPR_2B_DISC);
ndlp->nlp_flag &= ~(NLP_ADISC_SND | NLP_NPR_2B_DISC);
spin_unlock_irq(shost->host_lock);
/* ADISC completes to NPort <nlp_DID> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0104 ADISC completes to NPort x%x "
"Data: x%x x%x x%x x%x x%x\n",
ndlp->nlp_DID, irsp->ulpStatus, irsp->un.ulpWord[4],
irsp->ulpTimeout, disc, vport->num_disc_nodes);
/* Check to see if link went down during discovery */
if (lpfc_els_chk_latt(vport)) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
goto out;
}
if (irsp->ulpStatus) {
/* Check for retry */
if (lpfc_els_retry(phba, cmdiocb, rspiocb)) {
/* ELS command is being retried */
if (disc) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
lpfc_set_disctmo(vport);
}
goto out;
}
/* ADISC failed */
/* Do not call DSM for lpfc_els_abort'ed ELS cmds */
if (!lpfc_error_lost_link(irsp))
lpfc_disc_state_machine(vport, ndlp, cmdiocb,
NLP_EVT_CMPL_ADISC);
} else
/* Good status, call state machine */
lpfc_disc_state_machine(vport, ndlp, cmdiocb,
NLP_EVT_CMPL_ADISC);
/* Check to see if there are more ADISCs to be sent */
if (disc && vport->num_disc_nodes)
lpfc_more_adisc(vport);
out:
lpfc_els_free_iocb(phba, cmdiocb);
return;
}
/**
* lpfc_issue_els_adisc: Issue an address discover iocb to an node on a vport.
* @vport: pointer to a virtual N_Port data structure.
* @ndlp: pointer to a node-list data structure.
* @retry: number of retries to the command IOCB.
*
* This routine issues an Address Discover (ADISC) for an @ndlp on a
* @vport. It prepares the payload of the ADISC ELS command, updates the
* and states of the ndlp, and invokes the lpfc_sli_issue_iocb() routine
* to issue the ADISC ELS command.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the ADISC ELS command.
*
* Return code
* 0 - successfully issued adisc
* 1 - failed to issue adisc
**/
int
lpfc_issue_els_adisc(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
uint8_t retry)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
ADISC *ap;
IOCB_t *icmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli *psli = &phba->sli;
struct lpfc_sli_ring *pring = &psli->ring[LPFC_ELS_RING];
uint8_t *pcmd;
uint16_t cmdsize;
cmdsize = (sizeof(uint32_t) + sizeof(ADISC));
elsiocb = lpfc_prep_els_iocb(vport, 1, cmdsize, retry, ndlp,
ndlp->nlp_DID, ELS_CMD_ADISC);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
/* For ADISC request, remainder of payload is service parameters */
*((uint32_t *) (pcmd)) = ELS_CMD_ADISC;
pcmd += sizeof(uint32_t);
/* Fill in ADISC payload */
ap = (ADISC *) pcmd;
ap->hardAL_PA = phba->fc_pref_ALPA;
memcpy(&ap->portName, &vport->fc_portname, sizeof(struct lpfc_name));
memcpy(&ap->nodeName, &vport->fc_nodename, sizeof(struct lpfc_name));
ap->DID = be32_to_cpu(vport->fc_myDID);
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"Issue ADISC: did:x%x",
ndlp->nlp_DID, 0, 0);
phba->fc_stat.elsXmitADISC++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_adisc;
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_ADISC_SND;
spin_unlock_irq(shost->host_lock);
if (lpfc_sli_issue_iocb(phba, pring, elsiocb, 0) == IOCB_ERROR) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~NLP_ADISC_SND;
spin_unlock_irq(shost->host_lock);
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_cmpl_els_logo: Completion callback function for logo.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is the completion function for issuing the ELS Logout (LOGO)
* command. If no error status was reported from the LOGO response, the
* state machine of the associated ndlp shall be invoked for transition with
* respect to NLP_EVT_CMPL_LOGO event. Otherwise, if error status was reported,
* the lpfc_els_retry() routine will be invoked to retry the LOGO command.
**/
static void
lpfc_cmpl_els_logo(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_nodelist *ndlp = (struct lpfc_nodelist *) cmdiocb->context1;
struct lpfc_vport *vport = ndlp->vport;
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
IOCB_t *irsp;
struct lpfc_sli *psli;
psli = &phba->sli;
/* we pass cmdiocb to state machine which needs rspiocb as well */
cmdiocb->context_un.rsp_iocb = rspiocb;
irsp = &(rspiocb->iocb);
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~NLP_LOGO_SND;
spin_unlock_irq(shost->host_lock);
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"LOGO cmpl: status:x%x/x%x did:x%x",
irsp->ulpStatus, irsp->un.ulpWord[4],
ndlp->nlp_DID);
/* LOGO completes to NPort <nlp_DID> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0105 LOGO completes to NPort x%x "
"Data: x%x x%x x%x x%x\n",
ndlp->nlp_DID, irsp->ulpStatus, irsp->un.ulpWord[4],
irsp->ulpTimeout, vport->num_disc_nodes);
/* Check to see if link went down during discovery */
if (lpfc_els_chk_latt(vport))
goto out;
if (ndlp->nlp_flag & NLP_TARGET_REMOVE) {
/* NLP_EVT_DEVICE_RM should unregister the RPI
* which should abort all outstanding IOs.
*/
lpfc_disc_state_machine(vport, ndlp, cmdiocb,
NLP_EVT_DEVICE_RM);
goto out;
}
if (irsp->ulpStatus) {
/* Check for retry */
if (lpfc_els_retry(phba, cmdiocb, rspiocb))
/* ELS command is being retried */
goto out;
/* LOGO failed */
/* Do not call DSM for lpfc_els_abort'ed ELS cmds */
if (lpfc_error_lost_link(irsp))
goto out;
else
lpfc_disc_state_machine(vport, ndlp, cmdiocb,
NLP_EVT_CMPL_LOGO);
} else
/* Good status, call state machine.
* This will unregister the rpi if needed.
*/
lpfc_disc_state_machine(vport, ndlp, cmdiocb,
NLP_EVT_CMPL_LOGO);
out:
lpfc_els_free_iocb(phba, cmdiocb);
return;
}
/**
* lpfc_issue_els_logo: Issue a logo to an node on a vport.
* @vport: pointer to a virtual N_Port data structure.
* @ndlp: pointer to a node-list data structure.
* @retry: number of retries to the command IOCB.
*
* This routine constructs and issues an ELS Logout (LOGO) iocb command
* to a remote node, referred by an @ndlp on a @vport. It constructs the
* payload of the IOCB, properly sets up the @ndlp state, and invokes the
* lpfc_sli_issue_iocb() routine to send out the LOGO ELS command.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the LOGO ELS command.
*
* Return code
* 0 - successfully issued logo
* 1 - failed to issue logo
**/
int
lpfc_issue_els_logo(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
uint8_t retry)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
IOCB_t *icmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli_ring *pring;
struct lpfc_sli *psli;
uint8_t *pcmd;
uint16_t cmdsize;
int rc;
psli = &phba->sli;
pring = &psli->ring[LPFC_ELS_RING];
spin_lock_irq(shost->host_lock);
if (ndlp->nlp_flag & NLP_LOGO_SND) {
spin_unlock_irq(shost->host_lock);
return 0;
}
spin_unlock_irq(shost->host_lock);
cmdsize = (2 * sizeof(uint32_t)) + sizeof(struct lpfc_name);
elsiocb = lpfc_prep_els_iocb(vport, 1, cmdsize, retry, ndlp,
ndlp->nlp_DID, ELS_CMD_LOGO);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_LOGO;
pcmd += sizeof(uint32_t);
/* Fill in LOGO payload */
*((uint32_t *) (pcmd)) = be32_to_cpu(vport->fc_myDID);
pcmd += sizeof(uint32_t);
memcpy(pcmd, &vport->fc_portname, sizeof(struct lpfc_name));
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"Issue LOGO: did:x%x",
ndlp->nlp_DID, 0, 0);
phba->fc_stat.elsXmitLOGO++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_logo;
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_LOGO_SND;
spin_unlock_irq(shost->host_lock);
rc = lpfc_sli_issue_iocb(phba, pring, elsiocb, 0);
if (rc == IOCB_ERROR) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~NLP_LOGO_SND;
spin_unlock_irq(shost->host_lock);
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_cmpl_els_cmd: Completion callback function for generic els command.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is a generic completion callback function for ELS commands.
* Specifically, it is the callback function which does not need to perform
* any command specific operations. It is currently used by the ELS command
* issuing routines for the ELS State Change Request (SCR),
* lpfc_issue_els_scr(), and the ELS Fibre Channel Address Resolution
* Protocol Response (FARPR) routine, lpfc_issue_els_farpr(). Other than
* certain debug loggings, this callback function simply invokes the
* lpfc_els_chk_latt() routine to check whether link went down during the
* discovery process.
**/
static void
lpfc_cmpl_els_cmd(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_vport *vport = cmdiocb->vport;
IOCB_t *irsp;
irsp = &rspiocb->iocb;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"ELS cmd cmpl: status:x%x/x%x did:x%x",
irsp->ulpStatus, irsp->un.ulpWord[4],
irsp->un.elsreq64.remoteID);
/* ELS cmd tag <ulpIoTag> completes */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0106 ELS cmd tag x%x completes Data: x%x x%x x%x\n",
irsp->ulpIoTag, irsp->ulpStatus,
irsp->un.ulpWord[4], irsp->ulpTimeout);
/* Check to see if link went down during discovery */
lpfc_els_chk_latt(vport);
lpfc_els_free_iocb(phba, cmdiocb);
return;
}
/**
* lpfc_issue_els_scr: Issue a scr to an node on a vport.
* @vport: pointer to a host virtual N_Port data structure.
* @nportid: N_Port identifier to the remote node.
* @retry: number of retries to the command IOCB.
*
* This routine issues a State Change Request (SCR) to a fabric node
* on a @vport. The remote node @nportid is passed into the function. It
* first search the @vport node list to find the matching ndlp. If no such
* ndlp is found, a new ndlp shall be created for this (SCR) purpose. An
* IOCB is allocated, payload prepared, and the lpfc_sli_issue_iocb()
* routine is invoked to send the SCR IOCB.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the SCR ELS command.
*
* Return code
* 0 - Successfully issued scr command
* 1 - Failed to issue scr command
**/
int
lpfc_issue_els_scr(struct lpfc_vport *vport, uint32_t nportid, uint8_t retry)
{
struct lpfc_hba *phba = vport->phba;
IOCB_t *icmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli_ring *pring;
struct lpfc_sli *psli;
uint8_t *pcmd;
uint16_t cmdsize;
struct lpfc_nodelist *ndlp;
psli = &phba->sli;
pring = &psli->ring[LPFC_ELS_RING]; /* ELS ring */
cmdsize = (sizeof(uint32_t) + sizeof(SCR));
ndlp = lpfc_findnode_did(vport, nportid);
if (!ndlp) {
ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
if (!ndlp)
return 1;
lpfc_nlp_init(vport, ndlp, nportid);
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
if (!ndlp)
return 1;
}
elsiocb = lpfc_prep_els_iocb(vport, 1, cmdsize, retry, ndlp,
ndlp->nlp_DID, ELS_CMD_SCR);
if (!elsiocb) {
/* This will trigger the release of the node just
* allocated
*/
lpfc_nlp_put(ndlp);
return 1;
}
icmd = &elsiocb->iocb;
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_SCR;
pcmd += sizeof(uint32_t);
/* For SCR, remainder of payload is SCR parameter page */
memset(pcmd, 0, sizeof(SCR));
((SCR *) pcmd)->Function = SCR_FUNC_FULL;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"Issue SCR: did:x%x",
ndlp->nlp_DID, 0, 0);
phba->fc_stat.elsXmitSCR++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_cmd;
if (lpfc_sli_issue_iocb(phba, pring, elsiocb, 0) == IOCB_ERROR) {
/* The additional lpfc_nlp_put will cause the following
* lpfc_els_free_iocb routine to trigger the rlease of
* the node.
*/
lpfc_nlp_put(ndlp);
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
/* This will cause the callback-function lpfc_cmpl_els_cmd to
* trigger the release of node.
*/
lpfc_nlp_put(ndlp);
return 0;
}
/**
* lpfc_issue_els_farpr: Issue a farp to an node on a vport.
* @vport: pointer to a host virtual N_Port data structure.
* @nportid: N_Port identifier to the remote node.
* @retry: number of retries to the command IOCB.
*
* This routine issues a Fibre Channel Address Resolution Response
* (FARPR) to a node on a vport. The remote node N_Port identifier (@nportid)
* is passed into the function. It first search the @vport node list to find
* the matching ndlp. If no such ndlp is found, a new ndlp shall be created
* for this (FARPR) purpose. An IOCB is allocated, payload prepared, and the
* lpfc_sli_issue_iocb() routine is invoked to send the FARPR ELS command.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the PARPR ELS command.
*
* Return code
* 0 - Successfully issued farpr command
* 1 - Failed to issue farpr command
**/
static int
lpfc_issue_els_farpr(struct lpfc_vport *vport, uint32_t nportid, uint8_t retry)
{
struct lpfc_hba *phba = vport->phba;
IOCB_t *icmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli_ring *pring;
struct lpfc_sli *psli;
FARP *fp;
uint8_t *pcmd;
uint32_t *lp;
uint16_t cmdsize;
struct lpfc_nodelist *ondlp;
struct lpfc_nodelist *ndlp;
psli = &phba->sli;
pring = &psli->ring[LPFC_ELS_RING]; /* ELS ring */
cmdsize = (sizeof(uint32_t) + sizeof(FARP));
ndlp = lpfc_findnode_did(vport, nportid);
if (!ndlp) {
ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
if (!ndlp)
return 1;
lpfc_nlp_init(vport, ndlp, nportid);
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
if (!ndlp)
return 1;
}
elsiocb = lpfc_prep_els_iocb(vport, 1, cmdsize, retry, ndlp,
ndlp->nlp_DID, ELS_CMD_RNID);
if (!elsiocb) {
/* This will trigger the release of the node just
* allocated
*/
lpfc_nlp_put(ndlp);
return 1;
}
icmd = &elsiocb->iocb;
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_FARPR;
pcmd += sizeof(uint32_t);
/* Fill in FARPR payload */
fp = (FARP *) (pcmd);
memset(fp, 0, sizeof(FARP));
lp = (uint32_t *) pcmd;
*lp++ = be32_to_cpu(nportid);
*lp++ = be32_to_cpu(vport->fc_myDID);
fp->Rflags = 0;
fp->Mflags = (FARP_MATCH_PORT | FARP_MATCH_NODE);
memcpy(&fp->RportName, &vport->fc_portname, sizeof(struct lpfc_name));
memcpy(&fp->RnodeName, &vport->fc_nodename, sizeof(struct lpfc_name));
ondlp = lpfc_findnode_did(vport, nportid);
if (ondlp && NLP_CHK_NODE_ACT(ondlp)) {
memcpy(&fp->OportName, &ondlp->nlp_portname,
sizeof(struct lpfc_name));
memcpy(&fp->OnodeName, &ondlp->nlp_nodename,
sizeof(struct lpfc_name));
}
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"Issue FARPR: did:x%x",
ndlp->nlp_DID, 0, 0);
phba->fc_stat.elsXmitFARPR++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_cmd;
if (lpfc_sli_issue_iocb(phba, pring, elsiocb, 0) == IOCB_ERROR) {
/* The additional lpfc_nlp_put will cause the following
* lpfc_els_free_iocb routine to trigger the release of
* the node.
*/
lpfc_nlp_put(ndlp);
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
/* This will cause the callback-function lpfc_cmpl_els_cmd to
* trigger the release of the node.
*/
lpfc_nlp_put(ndlp);
return 0;
}
/**
* lpfc_cancel_retry_delay_tmo: Cancel the timer with delayed iocb-cmd retry.
* @vport: pointer to a host virtual N_Port data structure.
* @nlp: pointer to a node-list data structure.
*
* This routine cancels the timer with a delayed IOCB-command retry for
* a @vport's @ndlp. It stops the timer for the delayed function retrial and
* removes the ELS retry event if it presents. In addition, if the
* NLP_NPR_2B_DISC bit is set in the @nlp's nlp_flag bitmap, ADISC IOCB
* commands are sent for the @vport's nodes that require issuing discovery
* ADISC.
**/
void
lpfc_cancel_retry_delay_tmo(struct lpfc_vport *vport, struct lpfc_nodelist *nlp)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_work_evt *evtp;
if (!(nlp->nlp_flag & NLP_DELAY_TMO))
return;
spin_lock_irq(shost->host_lock);
nlp->nlp_flag &= ~NLP_DELAY_TMO;
spin_unlock_irq(shost->host_lock);
del_timer_sync(&nlp->nlp_delayfunc);
nlp->nlp_last_elscmd = 0;
if (!list_empty(&nlp->els_retry_evt.evt_listp)) {
list_del_init(&nlp->els_retry_evt.evt_listp);
/* Decrement nlp reference count held for the delayed retry */
evtp = &nlp->els_retry_evt;
lpfc_nlp_put((struct lpfc_nodelist *)evtp->evt_arg1);
}
if (nlp->nlp_flag & NLP_NPR_2B_DISC) {
spin_lock_irq(shost->host_lock);
nlp->nlp_flag &= ~NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
if (vport->num_disc_nodes) {
if (vport->port_state < LPFC_VPORT_READY) {
/* Check if there are more ADISCs to be sent */
lpfc_more_adisc(vport);
} else {
/* Check if there are more PLOGIs to be sent */
lpfc_more_plogi(vport);
if (vport->num_disc_nodes == 0) {
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~FC_NDISC_ACTIVE;
spin_unlock_irq(shost->host_lock);
lpfc_can_disctmo(vport);
lpfc_end_rscn(vport);
}
}
}
}
return;
}
/**
* lpfc_els_retry_delay: Timer function with a ndlp delayed function timer.
* @ptr: holder for the pointer to the timer function associated data (ndlp).
*
* This routine is invoked by the ndlp delayed-function timer to check
* whether there is any pending ELS retry event(s) with the node. If not, it
* simply returns. Otherwise, if there is at least one ELS delayed event, it
* adds the delayed events to the HBA work list and invokes the
* lpfc_worker_wake_up() routine to wake up worker thread to process the
* event. Note that lpfc_nlp_get() is called before posting the event to
* the work list to hold reference count of ndlp so that it guarantees the
* reference to ndlp will still be available when the worker thread gets
* to the event associated with the ndlp.
**/
void
lpfc_els_retry_delay(unsigned long ptr)
{
struct lpfc_nodelist *ndlp = (struct lpfc_nodelist *) ptr;
struct lpfc_vport *vport = ndlp->vport;
struct lpfc_hba *phba = vport->phba;
unsigned long flags;
struct lpfc_work_evt *evtp = &ndlp->els_retry_evt;
spin_lock_irqsave(&phba->hbalock, flags);
if (!list_empty(&evtp->evt_listp)) {
spin_unlock_irqrestore(&phba->hbalock, flags);
return;
}
/* We need to hold the node by incrementing the reference
* count until the queued work is done
*/
evtp->evt_arg1 = lpfc_nlp_get(ndlp);
if (evtp->evt_arg1) {
evtp->evt = LPFC_EVT_ELS_RETRY;
list_add_tail(&evtp->evt_listp, &phba->work_list);
lpfc_worker_wake_up(phba);
}
spin_unlock_irqrestore(&phba->hbalock, flags);
return;
}
/**
* lpfc_els_retry_delay_handler: Work thread handler for ndlp delayed function.
* @ndlp: pointer to a node-list data structure.
*
* This routine is the worker-thread handler for processing the @ndlp delayed
* event(s), posted by the lpfc_els_retry_delay() routine. It simply retrieves
* the last ELS command from the associated ndlp and invokes the proper ELS
* function according to the delayed ELS command to retry the command.
**/
void
lpfc_els_retry_delay_handler(struct lpfc_nodelist *ndlp)
{
struct lpfc_vport *vport = ndlp->vport;
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
uint32_t cmd, did, retry;
spin_lock_irq(shost->host_lock);
did = ndlp->nlp_DID;
cmd = ndlp->nlp_last_elscmd;
ndlp->nlp_last_elscmd = 0;
if (!(ndlp->nlp_flag & NLP_DELAY_TMO)) {
spin_unlock_irq(shost->host_lock);
return;
}
ndlp->nlp_flag &= ~NLP_DELAY_TMO;
spin_unlock_irq(shost->host_lock);
/*
* If a discovery event readded nlp_delayfunc after timer
* firing and before processing the timer, cancel the
* nlp_delayfunc.
*/
del_timer_sync(&ndlp->nlp_delayfunc);
retry = ndlp->nlp_retry;
switch (cmd) {
case ELS_CMD_FLOGI:
lpfc_issue_els_flogi(vport, ndlp, retry);
break;
case ELS_CMD_PLOGI:
if (!lpfc_issue_els_plogi(vport, ndlp->nlp_DID, retry)) {
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_PLOGI_ISSUE);
}
break;
case ELS_CMD_ADISC:
if (!lpfc_issue_els_adisc(vport, ndlp, retry)) {
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_ADISC_ISSUE);
}
break;
case ELS_CMD_PRLI:
if (!lpfc_issue_els_prli(vport, ndlp, retry)) {
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_PRLI_ISSUE);
}
break;
case ELS_CMD_LOGO:
if (!lpfc_issue_els_logo(vport, ndlp, retry)) {
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
}
break;
case ELS_CMD_FDISC:
lpfc_issue_els_fdisc(vport, ndlp, retry);
break;
}
return;
}
/**
* lpfc_els_retry: Make retry decision on an els command iocb.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine makes a retry decision on an ELS command IOCB, which has
* failed. The following ELS IOCBs use this function for retrying the command
* when previously issued command responsed with error status: FLOGI, PLOGI,
* PRLI, ADISC, LOGO, and FDISC. Based on the ELS command type and the
* returned error status, it makes the decision whether a retry shall be
* issued for the command, and whether a retry shall be made immediately or
* delayed. In the former case, the corresponding ELS command issuing-function
* is called to retry the command. In the later case, the ELS command shall
* be posted to the ndlp delayed event and delayed function timer set to the
* ndlp for the delayed command issusing.
*
* Return code
* 0 - No retry of els command is made
* 1 - Immediate or delayed retry of els command is made
**/
static int
lpfc_els_retry(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_vport *vport = cmdiocb->vport;
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
IOCB_t *irsp = &rspiocb->iocb;
struct lpfc_nodelist *ndlp = (struct lpfc_nodelist *) cmdiocb->context1;
struct lpfc_dmabuf *pcmd = (struct lpfc_dmabuf *) cmdiocb->context2;
uint32_t *elscmd;
struct ls_rjt stat;
int retry = 0, maxretry = lpfc_max_els_tries, delay = 0;
int logerr = 0;
uint32_t cmd = 0;
uint32_t did;
/* Note: context2 may be 0 for internal driver abort
* of delays ELS command.
*/
if (pcmd && pcmd->virt) {
elscmd = (uint32_t *) (pcmd->virt);
cmd = *elscmd++;
}
if (ndlp && NLP_CHK_NODE_ACT(ndlp))
did = ndlp->nlp_DID;
else {
/* We should only hit this case for retrying PLOGI */
did = irsp->un.elsreq64.remoteID;
ndlp = lpfc_findnode_did(vport, did);
if ((!ndlp || !NLP_CHK_NODE_ACT(ndlp))
&& (cmd != ELS_CMD_PLOGI))
return 1;
}
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"Retry ELS: wd7:x%x wd4:x%x did:x%x",
*(((uint32_t *) irsp) + 7), irsp->un.ulpWord[4], ndlp->nlp_DID);
switch (irsp->ulpStatus) {
case IOSTAT_FCP_RSP_ERROR:
case IOSTAT_REMOTE_STOP:
break;
case IOSTAT_LOCAL_REJECT:
switch ((irsp->un.ulpWord[4] & 0xff)) {
case IOERR_LOOP_OPEN_FAILURE:
if (cmd == ELS_CMD_PLOGI && cmdiocb->retry == 0)
delay = 1000;
retry = 1;
break;
case IOERR_ILLEGAL_COMMAND:
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0124 Retry illegal cmd x%x "
"retry:x%x delay:x%x\n",
cmd, cmdiocb->retry, delay);
retry = 1;
/* All command's retry policy */
maxretry = 8;
if (cmdiocb->retry > 2)
delay = 1000;
break;
case IOERR_NO_RESOURCES:
logerr = 1; /* HBA out of resources */
retry = 1;
if (cmdiocb->retry > 100)
delay = 100;
maxretry = 250;
break;
case IOERR_ILLEGAL_FRAME:
delay = 100;
retry = 1;
break;
case IOERR_SEQUENCE_TIMEOUT:
case IOERR_INVALID_RPI:
retry = 1;
break;
}
break;
case IOSTAT_NPORT_RJT:
case IOSTAT_FABRIC_RJT:
if (irsp->un.ulpWord[4] & RJT_UNAVAIL_TEMP) {
retry = 1;
break;
}
break;
case IOSTAT_NPORT_BSY:
case IOSTAT_FABRIC_BSY:
logerr = 1; /* Fabric / Remote NPort out of resources */
retry = 1;
break;
case IOSTAT_LS_RJT:
stat.un.lsRjtError = be32_to_cpu(irsp->un.ulpWord[4]);
/* Added for Vendor specifc support
* Just keep retrying for these Rsn / Exp codes
*/
switch (stat.un.b.lsRjtRsnCode) {
case LSRJT_UNABLE_TPC:
if (stat.un.b.lsRjtRsnCodeExp ==
LSEXP_CMD_IN_PROGRESS) {
if (cmd == ELS_CMD_PLOGI) {
delay = 1000;
maxretry = 48;
}
retry = 1;
break;
}
if (cmd == ELS_CMD_PLOGI) {
delay = 1000;
maxretry = lpfc_max_els_tries + 1;
retry = 1;
break;
}
if ((phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) &&
(cmd == ELS_CMD_FDISC) &&
(stat.un.b.lsRjtRsnCodeExp == LSEXP_OUT_OF_RESOURCE)){
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0125 FDISC Failed (x%x). "
"Fabric out of resources\n",
stat.un.lsRjtError);
lpfc_vport_set_state(vport,
FC_VPORT_NO_FABRIC_RSCS);
}
break;
case LSRJT_LOGICAL_BSY:
if ((cmd == ELS_CMD_PLOGI) ||
(cmd == ELS_CMD_PRLI)) {
delay = 1000;
maxretry = 48;
} else if (cmd == ELS_CMD_FDISC) {
/* FDISC retry policy */
maxretry = 48;
if (cmdiocb->retry >= 32)
delay = 1000;
}
retry = 1;
break;
case LSRJT_LOGICAL_ERR:
/* There are some cases where switches return this
* error when they are not ready and should be returning
* Logical Busy. We should delay every time.
*/
if (cmd == ELS_CMD_FDISC &&
stat.un.b.lsRjtRsnCodeExp == LSEXP_PORT_LOGIN_REQ) {
maxretry = 3;
delay = 1000;
retry = 1;
break;
}
case LSRJT_PROTOCOL_ERR:
if ((phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) &&
(cmd == ELS_CMD_FDISC) &&
((stat.un.b.lsRjtRsnCodeExp == LSEXP_INVALID_PNAME) ||
(stat.un.b.lsRjtRsnCodeExp == LSEXP_INVALID_NPORT_ID))
) {
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0123 FDISC Failed (x%x). "
"Fabric Detected Bad WWN\n",
stat.un.lsRjtError);
lpfc_vport_set_state(vport,
FC_VPORT_FABRIC_REJ_WWN);
}
break;
}
break;
case IOSTAT_INTERMED_RSP:
case IOSTAT_BA_RJT:
break;
default:
break;
}
if (did == FDMI_DID)
retry = 1;
if ((cmd == ELS_CMD_FLOGI) &&
(phba->fc_topology != TOPOLOGY_LOOP) &&
!lpfc_error_lost_link(irsp)) {
/* FLOGI retry policy */
retry = 1;
maxretry = 48;
if (cmdiocb->retry >= 32)
delay = 1000;
}
if ((++cmdiocb->retry) >= maxretry) {
phba->fc_stat.elsRetryExceeded++;
retry = 0;
}
if ((vport->load_flag & FC_UNLOADING) != 0)
retry = 0;
if (retry) {
/* Retry ELS command <elsCmd> to remote NPORT <did> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0107 Retry ELS command x%x to remote "
"NPORT x%x Data: x%x x%x\n",
cmd, did, cmdiocb->retry, delay);
if (((cmd == ELS_CMD_PLOGI) || (cmd == ELS_CMD_ADISC)) &&
((irsp->ulpStatus != IOSTAT_LOCAL_REJECT) ||
((irsp->un.ulpWord[4] & 0xff) != IOERR_NO_RESOURCES))) {
/* Don't reset timer for no resources */
/* If discovery / RSCN timer is running, reset it */
if (timer_pending(&vport->fc_disctmo) ||
(vport->fc_flag & FC_RSCN_MODE))
lpfc_set_disctmo(vport);
}
phba->fc_stat.elsXmitRetry++;
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
if (ndlp && NLP_CHK_NODE_ACT(ndlp) && delay) {
phba->fc_stat.elsDelayRetry++;
ndlp->nlp_retry = cmdiocb->retry;
/* delay is specified in milliseconds */
mod_timer(&ndlp->nlp_delayfunc,
jiffies + msecs_to_jiffies(delay));
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_DELAY_TMO;
spin_unlock_irq(shost->host_lock);
ndlp->nlp_prev_state = ndlp->nlp_state;
if (cmd == ELS_CMD_PRLI)
lpfc_nlp_set_state(vport, ndlp,
NLP_STE_REG_LOGIN_ISSUE);
else
lpfc_nlp_set_state(vport, ndlp,
NLP_STE_NPR_NODE);
ndlp->nlp_last_elscmd = cmd;
return 1;
}
switch (cmd) {
case ELS_CMD_FLOGI:
lpfc_issue_els_flogi(vport, ndlp, cmdiocb->retry);
return 1;
case ELS_CMD_FDISC:
lpfc_issue_els_fdisc(vport, ndlp, cmdiocb->retry);
return 1;
case ELS_CMD_PLOGI:
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
if (ndlp && NLP_CHK_NODE_ACT(ndlp)) {
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp,
NLP_STE_PLOGI_ISSUE);
}
lpfc_issue_els_plogi(vport, did, cmdiocb->retry);
return 1;
case ELS_CMD_ADISC:
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_ADISC_ISSUE);
lpfc_issue_els_adisc(vport, ndlp, cmdiocb->retry);
return 1;
case ELS_CMD_PRLI:
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_PRLI_ISSUE);
lpfc_issue_els_prli(vport, ndlp, cmdiocb->retry);
return 1;
case ELS_CMD_LOGO:
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
lpfc_issue_els_logo(vport, ndlp, cmdiocb->retry);
return 1;
}
}
/* No retry ELS command <elsCmd> to remote NPORT <did> */
if (logerr) {
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0137 No retry ELS command x%x to remote "
"NPORT x%x: Out of Resources: Error:x%x/%x\n",
cmd, did, irsp->ulpStatus,
irsp->un.ulpWord[4]);
}
else {
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0108 No retry ELS command x%x to remote "
"NPORT x%x Retried:%d Error:x%x/%x\n",
cmd, did, cmdiocb->retry, irsp->ulpStatus,
irsp->un.ulpWord[4]);
}
return 0;
}
/**
* lpfc_els_free_data: Free lpfc dma buffer and data structure with an iocb.
* @phba: pointer to lpfc hba data structure.
* @buf_ptr1: pointer to the lpfc DMA buffer data structure.
*
* This routine releases the lpfc DMA (Direct Memory Access) buffer(s)
* associated with a command IOCB back to the lpfc DMA buffer pool. It first
* checks to see whether there is a lpfc DMA buffer associated with the
* response of the command IOCB. If so, it will be released before releasing
* the lpfc DMA buffer associated with the IOCB itself.
*
* Return code
* 0 - Successfully released lpfc DMA buffer (currently, always return 0)
**/
[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 13:52:54 +07:00
static int
lpfc_els_free_data(struct lpfc_hba *phba, struct lpfc_dmabuf *buf_ptr1)
{
struct lpfc_dmabuf *buf_ptr;
/* Free the response before processing the command. */
if (!list_empty(&buf_ptr1->list)) {
list_remove_head(&buf_ptr1->list, buf_ptr,
struct lpfc_dmabuf,
list);
lpfc_mbuf_free(phba, buf_ptr->virt, buf_ptr->phys);
kfree(buf_ptr);
}
lpfc_mbuf_free(phba, buf_ptr1->virt, buf_ptr1->phys);
kfree(buf_ptr1);
return 0;
}
/**
* lpfc_els_free_bpl: Free lpfc dma buffer and data structure with bpl.
* @phba: pointer to lpfc hba data structure.
* @buf_ptr: pointer to the lpfc dma buffer data structure.
*
* This routine releases the lpfc Direct Memory Access (DMA) buffer
* associated with a Buffer Pointer List (BPL) back to the lpfc DMA buffer
* pool.
*
* Return code
* 0 - Successfully released lpfc DMA buffer (currently, always return 0)
**/
[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 13:52:54 +07:00
static int
lpfc_els_free_bpl(struct lpfc_hba *phba, struct lpfc_dmabuf *buf_ptr)
{
lpfc_mbuf_free(phba, buf_ptr->virt, buf_ptr->phys);
kfree(buf_ptr);
return 0;
}
/**
* lpfc_els_free_iocb: Free a command iocb and its associated resources.
* @phba: pointer to lpfc hba data structure.
* @elsiocb: pointer to lpfc els command iocb data structure.
*
* This routine frees a command IOCB and its associated resources. The
* command IOCB data structure contains the reference to various associated
* resources, these fields must be set to NULL if the associated reference
* not present:
* context1 - reference to ndlp
* context2 - reference to cmd
* context2->next - reference to rsp
* context3 - reference to bpl
*
* It first properly decrements the reference count held on ndlp for the
* IOCB completion callback function. If LPFC_DELAY_MEM_FREE flag is not
* set, it invokes the lpfc_els_free_data() routine to release the Direct
* Memory Access (DMA) buffers associated with the IOCB. Otherwise, it
* adds the DMA buffer the @phba data structure for the delayed release.
* If reference to the Buffer Pointer List (BPL) is present, the
* lpfc_els_free_bpl() routine is invoked to release the DMA memory
* associated with BPL. Finally, the lpfc_sli_release_iocbq() routine is
* invoked to release the IOCB data structure back to @phba IOCBQ list.
*
* Return code
* 0 - Success (currently, always return 0)
**/
int
lpfc_els_free_iocb(struct lpfc_hba *phba, struct lpfc_iocbq *elsiocb)
{
struct lpfc_dmabuf *buf_ptr, *buf_ptr1;
struct lpfc_nodelist *ndlp;
ndlp = (struct lpfc_nodelist *)elsiocb->context1;
if (ndlp) {
if (ndlp->nlp_flag & NLP_DEFER_RM) {
lpfc_nlp_put(ndlp);
/* If the ndlp is not being used by another discovery
* thread, free it.
*/
if (!lpfc_nlp_not_used(ndlp)) {
/* If ndlp is being used by another discovery
* thread, just clear NLP_DEFER_RM
*/
ndlp->nlp_flag &= ~NLP_DEFER_RM;
}
}
else
lpfc_nlp_put(ndlp);
elsiocb->context1 = NULL;
}
/* context2 = cmd, context2->next = rsp, context3 = bpl */
if (elsiocb->context2) {
if (elsiocb->iocb_flag & LPFC_DELAY_MEM_FREE) {
/* Firmware could still be in progress of DMAing
* payload, so don't free data buffer till after
* a hbeat.
*/
elsiocb->iocb_flag &= ~LPFC_DELAY_MEM_FREE;
buf_ptr = elsiocb->context2;
elsiocb->context2 = NULL;
if (buf_ptr) {
buf_ptr1 = NULL;
spin_lock_irq(&phba->hbalock);
if (!list_empty(&buf_ptr->list)) {
list_remove_head(&buf_ptr->list,
buf_ptr1, struct lpfc_dmabuf,
list);
INIT_LIST_HEAD(&buf_ptr1->list);
list_add_tail(&buf_ptr1->list,
&phba->elsbuf);
phba->elsbuf_cnt++;
}
INIT_LIST_HEAD(&buf_ptr->list);
list_add_tail(&buf_ptr->list, &phba->elsbuf);
phba->elsbuf_cnt++;
spin_unlock_irq(&phba->hbalock);
}
} else {
buf_ptr1 = (struct lpfc_dmabuf *) elsiocb->context2;
lpfc_els_free_data(phba, buf_ptr1);
}
}
if (elsiocb->context3) {
buf_ptr = (struct lpfc_dmabuf *) elsiocb->context3;
lpfc_els_free_bpl(phba, buf_ptr);
}
lpfc_sli_release_iocbq(phba, elsiocb);
return 0;
}
/**
* lpfc_cmpl_els_logo_acc: Completion callback function to logo acc response.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is the completion callback function to the Logout (LOGO)
* Accept (ACC) Response ELS command. This routine is invoked to indicate
* the completion of the LOGO process. It invokes the lpfc_nlp_not_used() to
* release the ndlp if it has the last reference remaining (reference count
* is 1). If succeeded (meaning ndlp released), it sets the IOCB context1
* field to NULL to inform the following lpfc_els_free_iocb() routine no
* ndlp reference count needs to be decremented. Otherwise, the ndlp
* reference use-count shall be decremented by the lpfc_els_free_iocb()
* routine. Finally, the lpfc_els_free_iocb() is invoked to release the
* IOCB data structure.
**/
static void
lpfc_cmpl_els_logo_acc(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_nodelist *ndlp = (struct lpfc_nodelist *) cmdiocb->context1;
struct lpfc_vport *vport = cmdiocb->vport;
IOCB_t *irsp;
irsp = &rspiocb->iocb;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_RSP,
"ACC LOGO cmpl: status:x%x/x%x did:x%x",
irsp->ulpStatus, irsp->un.ulpWord[4], ndlp->nlp_DID);
/* ACC to LOGO completes to NPort <nlp_DID> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0109 ACC to LOGO completes to NPort x%x "
"Data: x%x x%x x%x\n",
ndlp->nlp_DID, ndlp->nlp_flag, ndlp->nlp_state,
ndlp->nlp_rpi);
if (ndlp->nlp_state == NLP_STE_NPR_NODE) {
/* NPort Recovery mode or node is just allocated */
if (!lpfc_nlp_not_used(ndlp)) {
/* If the ndlp is being used by another discovery
* thread, just unregister the RPI.
*/
lpfc_unreg_rpi(vport, ndlp);
} else {
/* Indicate the node has already released, should
* not reference to it from within lpfc_els_free_iocb.
*/
cmdiocb->context1 = NULL;
}
}
lpfc_els_free_iocb(phba, cmdiocb);
return;
}
/**
* lpfc_mbx_cmpl_dflt_rpi: Completion callbk func for unreg dflt rpi mbox cmd.
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* This routine is the completion callback function for unregister default
* RPI (Remote Port Index) mailbox command to the @phba. It simply releases
* the associated lpfc Direct Memory Access (DMA) buffer back to the pool and
* decrements the ndlp reference count held for this completion callback
* function. After that, it invokes the lpfc_nlp_not_used() to check
* whether there is only one reference left on the ndlp. If so, it will
* perform one more decrement and trigger the release of the ndlp.
**/
void
lpfc_mbx_cmpl_dflt_rpi(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
{
struct lpfc_dmabuf *mp = (struct lpfc_dmabuf *) (pmb->context1);
struct lpfc_nodelist *ndlp = (struct lpfc_nodelist *) pmb->context2;
pmb->context1 = NULL;
lpfc_mbuf_free(phba, mp->virt, mp->phys);
kfree(mp);
mempool_free(pmb, phba->mbox_mem_pool);
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
if (ndlp && NLP_CHK_NODE_ACT(ndlp)) {
lpfc_nlp_put(ndlp);
/* This is the end of the default RPI cleanup logic for this
* ndlp. If no other discovery threads are using this ndlp.
* we should free all resources associated with it.
*/
lpfc_nlp_not_used(ndlp);
}
return;
}
/**
* lpfc_cmpl_els_rsp: Completion callback function for els response iocb cmd.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is the completion callback function for ELS Response IOCB
* command. In normal case, this callback function just properly sets the
* nlp_flag bitmap in the ndlp data structure, if the mbox command reference
* field in the command IOCB is not NULL, the referred mailbox command will
* be send out, and then invokes the lpfc_els_free_iocb() routine to release
* the IOCB. Under error conditions, such as when a LS_RJT is returned or a
* link down event occurred during the discovery, the lpfc_nlp_not_used()
* routine shall be invoked trying to release the ndlp if no other threads
* are currently referring it.
**/
static void
lpfc_cmpl_els_rsp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_nodelist *ndlp = (struct lpfc_nodelist *) cmdiocb->context1;
struct lpfc_vport *vport = ndlp ? ndlp->vport : NULL;
struct Scsi_Host *shost = vport ? lpfc_shost_from_vport(vport) : NULL;
IOCB_t *irsp;
uint8_t *pcmd;
LPFC_MBOXQ_t *mbox = NULL;
struct lpfc_dmabuf *mp = NULL;
uint32_t ls_rjt = 0;
irsp = &rspiocb->iocb;
if (cmdiocb->context_un.mbox)
mbox = cmdiocb->context_un.mbox;
/* First determine if this is a LS_RJT cmpl. Note, this callback
* function can have cmdiocb->contest1 (ndlp) field set to NULL.
*/
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) cmdiocb->context2)->virt);
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
if (ndlp && NLP_CHK_NODE_ACT(ndlp) &&
(*((uint32_t *) (pcmd)) == ELS_CMD_LS_RJT)) {
/* A LS_RJT associated with Default RPI cleanup has its own
* seperate code path.
*/
if (!(ndlp->nlp_flag & NLP_RM_DFLT_RPI))
ls_rjt = 1;
}
/* Check to see if link went down during discovery */
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
if (!ndlp || !NLP_CHK_NODE_ACT(ndlp) || lpfc_els_chk_latt(vport)) {
if (mbox) {
mp = (struct lpfc_dmabuf *) mbox->context1;
if (mp) {
lpfc_mbuf_free(phba, mp->virt, mp->phys);
kfree(mp);
}
mempool_free(mbox, phba->mbox_mem_pool);
}
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
if (ndlp && NLP_CHK_NODE_ACT(ndlp) &&
(ndlp->nlp_flag & NLP_RM_DFLT_RPI))
if (lpfc_nlp_not_used(ndlp)) {
ndlp = NULL;
/* Indicate the node has already released,
* should not reference to it from within
* the routine lpfc_els_free_iocb.
*/
cmdiocb->context1 = NULL;
}
goto out;
}
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_RSP,
"ELS rsp cmpl: status:x%x/x%x did:x%x",
irsp->ulpStatus, irsp->un.ulpWord[4],
cmdiocb->iocb.un.elsreq64.remoteID);
/* ELS response tag <ulpIoTag> completes */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0110 ELS response tag x%x completes "
"Data: x%x x%x x%x x%x x%x x%x x%x\n",
cmdiocb->iocb.ulpIoTag, rspiocb->iocb.ulpStatus,
rspiocb->iocb.un.ulpWord[4], rspiocb->iocb.ulpTimeout,
ndlp->nlp_DID, ndlp->nlp_flag, ndlp->nlp_state,
ndlp->nlp_rpi);
if (mbox) {
if ((rspiocb->iocb.ulpStatus == 0)
&& (ndlp->nlp_flag & NLP_ACC_REGLOGIN)) {
lpfc_unreg_rpi(vport, ndlp);
/* Increment reference count to ndlp to hold the
* reference to ndlp for the callback function.
*/
mbox->context2 = lpfc_nlp_get(ndlp);
mbox->vport = vport;
if (ndlp->nlp_flag & NLP_RM_DFLT_RPI) {
mbox->mbox_flag |= LPFC_MBX_IMED_UNREG;
mbox->mbox_cmpl = lpfc_mbx_cmpl_dflt_rpi;
}
else {
mbox->mbox_cmpl = lpfc_mbx_cmpl_reg_login;
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp,
NLP_STE_REG_LOGIN_ISSUE);
}
if (lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT)
!= MBX_NOT_FINISHED)
goto out;
else
/* Decrement the ndlp reference count we
* set for this failed mailbox command.
*/
lpfc_nlp_put(ndlp);
/* ELS rsp: Cannot issue reg_login for <NPortid> */
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0138 ELS rsp: Cannot issue reg_login for x%x "
"Data: x%x x%x x%x\n",
ndlp->nlp_DID, ndlp->nlp_flag, ndlp->nlp_state,
ndlp->nlp_rpi);
if (lpfc_nlp_not_used(ndlp)) {
ndlp = NULL;
/* Indicate node has already been released,
* should not reference to it from within
* the routine lpfc_els_free_iocb.
*/
cmdiocb->context1 = NULL;
}
} else {
/* Do not drop node for lpfc_els_abort'ed ELS cmds */
if (!lpfc_error_lost_link(irsp) &&
ndlp->nlp_flag & NLP_ACC_REGLOGIN) {
if (lpfc_nlp_not_used(ndlp)) {
ndlp = NULL;
/* Indicate node has already been
* released, should not reference
* to it from within the routine
* lpfc_els_free_iocb.
*/
cmdiocb->context1 = NULL;
}
}
}
mp = (struct lpfc_dmabuf *) mbox->context1;
if (mp) {
lpfc_mbuf_free(phba, mp->virt, mp->phys);
kfree(mp);
}
mempool_free(mbox, phba->mbox_mem_pool);
}
out:
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
if (ndlp && NLP_CHK_NODE_ACT(ndlp)) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~(NLP_ACC_REGLOGIN | NLP_RM_DFLT_RPI);
spin_unlock_irq(shost->host_lock);
/* If the node is not being used by another discovery thread,
* and we are sending a reject, we are done with it.
* Release driver reference count here and free associated
* resources.
*/
if (ls_rjt)
if (lpfc_nlp_not_used(ndlp))
/* Indicate node has already been released,
* should not reference to it from within
* the routine lpfc_els_free_iocb.
*/
cmdiocb->context1 = NULL;
}
lpfc_els_free_iocb(phba, cmdiocb);
return;
}
/**
* lpfc_els_rsp_acc: Prepare and issue an acc response iocb command.
* @vport: pointer to a host virtual N_Port data structure.
* @flag: the els command code to be accepted.
* @oldiocb: pointer to the original lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
* @mbox: pointer to the driver internal queue element for mailbox command.
*
* This routine prepares and issues an Accept (ACC) response IOCB
* command. It uses the @flag to properly set up the IOCB field for the
* specific ACC response command to be issued and invokes the
* lpfc_sli_issue_iocb() routine to send out ACC response IOCB. If a
* @mbox pointer is passed in, it will be put into the context_un.mbox
* field of the IOCB for the completion callback function to issue the
* mailbox command to the HBA later when callback is invoked.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the corresponding response ELS IOCB command.
*
* Return code
* 0 - Successfully issued acc response
* 1 - Failed to issue acc response
**/
int
lpfc_els_rsp_acc(struct lpfc_vport *vport, uint32_t flag,
struct lpfc_iocbq *oldiocb, struct lpfc_nodelist *ndlp,
LPFC_MBOXQ_t *mbox)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
IOCB_t *icmd;
IOCB_t *oldcmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli_ring *pring;
struct lpfc_sli *psli;
uint8_t *pcmd;
uint16_t cmdsize;
int rc;
ELS_PKT *els_pkt_ptr;
psli = &phba->sli;
pring = &psli->ring[LPFC_ELS_RING]; /* ELS ring */
oldcmd = &oldiocb->iocb;
switch (flag) {
case ELS_CMD_ACC:
cmdsize = sizeof(uint32_t);
elsiocb = lpfc_prep_els_iocb(vport, 0, cmdsize, oldiocb->retry,
ndlp, ndlp->nlp_DID, ELS_CMD_ACC);
if (!elsiocb) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~NLP_LOGO_ACC;
spin_unlock_irq(shost->host_lock);
return 1;
}
icmd = &elsiocb->iocb;
icmd->ulpContext = oldcmd->ulpContext; /* Xri */
pcmd = (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_ACC;
pcmd += sizeof(uint32_t);
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_RSP,
"Issue ACC: did:x%x flg:x%x",
ndlp->nlp_DID, ndlp->nlp_flag, 0);
break;
case ELS_CMD_PLOGI:
cmdsize = (sizeof(struct serv_parm) + sizeof(uint32_t));
elsiocb = lpfc_prep_els_iocb(vport, 0, cmdsize, oldiocb->retry,
ndlp, ndlp->nlp_DID, ELS_CMD_ACC);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
icmd->ulpContext = oldcmd->ulpContext; /* Xri */
pcmd = (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
if (mbox)
elsiocb->context_un.mbox = mbox;
*((uint32_t *) (pcmd)) = ELS_CMD_ACC;
pcmd += sizeof(uint32_t);
memcpy(pcmd, &vport->fc_sparam, sizeof(struct serv_parm));
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_RSP,
"Issue ACC PLOGI: did:x%x flg:x%x",
ndlp->nlp_DID, ndlp->nlp_flag, 0);
break;
case ELS_CMD_PRLO:
cmdsize = sizeof(uint32_t) + sizeof(PRLO);
elsiocb = lpfc_prep_els_iocb(vport, 0, cmdsize, oldiocb->retry,
ndlp, ndlp->nlp_DID, ELS_CMD_PRLO);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
icmd->ulpContext = oldcmd->ulpContext; /* Xri */
pcmd = (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
memcpy(pcmd, ((struct lpfc_dmabuf *) oldiocb->context2)->virt,
sizeof(uint32_t) + sizeof(PRLO));
*((uint32_t *) (pcmd)) = ELS_CMD_PRLO_ACC;
els_pkt_ptr = (ELS_PKT *) pcmd;
els_pkt_ptr->un.prlo.acceptRspCode = PRLO_REQ_EXECUTED;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_RSP,
"Issue ACC PRLO: did:x%x flg:x%x",
ndlp->nlp_DID, ndlp->nlp_flag, 0);
break;
default:
return 1;
}
/* Xmit ELS ACC response tag <ulpIoTag> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0128 Xmit ELS ACC response tag x%x, XRI: x%x, "
"DID: x%x, nlp_flag: x%x nlp_state: x%x RPI: x%x\n",
elsiocb->iotag, elsiocb->iocb.ulpContext,
ndlp->nlp_DID, ndlp->nlp_flag, ndlp->nlp_state,
ndlp->nlp_rpi);
if (ndlp->nlp_flag & NLP_LOGO_ACC) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~NLP_LOGO_ACC;
spin_unlock_irq(shost->host_lock);
elsiocb->iocb_cmpl = lpfc_cmpl_els_logo_acc;
} else {
elsiocb->iocb_cmpl = lpfc_cmpl_els_rsp;
}
phba->fc_stat.elsXmitACC++;
rc = lpfc_sli_issue_iocb(phba, pring, elsiocb, 0);
if (rc == IOCB_ERROR) {
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_els_rsp_reject: Propare and issue a rjt response iocb command.
* @vport: pointer to a virtual N_Port data structure.
* @rejectError:
* @oldiocb: pointer to the original lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
* @mbox: pointer to the driver internal queue element for mailbox command.
*
* This routine prepares and issue an Reject (RJT) response IOCB
* command. If a @mbox pointer is passed in, it will be put into the
* context_un.mbox field of the IOCB for the completion callback function
* to issue to the HBA later.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the reject response ELS IOCB command.
*
* Return code
* 0 - Successfully issued reject response
* 1 - Failed to issue reject response
**/
int
lpfc_els_rsp_reject(struct lpfc_vport *vport, uint32_t rejectError,
struct lpfc_iocbq *oldiocb, struct lpfc_nodelist *ndlp,
LPFC_MBOXQ_t *mbox)
{
struct lpfc_hba *phba = vport->phba;
IOCB_t *icmd;
IOCB_t *oldcmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli_ring *pring;
struct lpfc_sli *psli;
uint8_t *pcmd;
uint16_t cmdsize;
int rc;
psli = &phba->sli;
pring = &psli->ring[LPFC_ELS_RING]; /* ELS ring */
cmdsize = 2 * sizeof(uint32_t);
elsiocb = lpfc_prep_els_iocb(vport, 0, cmdsize, oldiocb->retry, ndlp,
ndlp->nlp_DID, ELS_CMD_LS_RJT);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
oldcmd = &oldiocb->iocb;
icmd->ulpContext = oldcmd->ulpContext; /* Xri */
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_LS_RJT;
pcmd += sizeof(uint32_t);
*((uint32_t *) (pcmd)) = rejectError;
if (mbox)
elsiocb->context_un.mbox = mbox;
/* Xmit ELS RJT <err> response tag <ulpIoTag> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0129 Xmit ELS RJT x%x response tag x%x "
"xri x%x, did x%x, nlp_flag x%x, nlp_state x%x, "
"rpi x%x\n",
rejectError, elsiocb->iotag,
elsiocb->iocb.ulpContext, ndlp->nlp_DID,
ndlp->nlp_flag, ndlp->nlp_state, ndlp->nlp_rpi);
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_RSP,
"Issue LS_RJT: did:x%x flg:x%x err:x%x",
ndlp->nlp_DID, ndlp->nlp_flag, rejectError);
phba->fc_stat.elsXmitLSRJT++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_rsp;
rc = lpfc_sli_issue_iocb(phba, pring, elsiocb, 0);
if (rc == IOCB_ERROR) {
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_els_rsp_adisc_acc: Prepare and issue acc response to adisc iocb cmd.
* @vport: pointer to a virtual N_Port data structure.
* @oldiocb: pointer to the original lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine prepares and issues an Accept (ACC) response to Address
* Discover (ADISC) ELS command. It simply prepares the payload of the IOCB
* and invokes the lpfc_sli_issue_iocb() routine to send out the command.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the ADISC Accept response ELS IOCB command.
*
* Return code
* 0 - Successfully issued acc adisc response
* 1 - Failed to issue adisc acc response
**/
int
lpfc_els_rsp_adisc_acc(struct lpfc_vport *vport, struct lpfc_iocbq *oldiocb,
struct lpfc_nodelist *ndlp)
{
struct lpfc_hba *phba = vport->phba;
struct lpfc_sli *psli = &phba->sli;
struct lpfc_sli_ring *pring = &psli->ring[LPFC_ELS_RING];
ADISC *ap;
IOCB_t *icmd, *oldcmd;
struct lpfc_iocbq *elsiocb;
uint8_t *pcmd;
uint16_t cmdsize;
int rc;
cmdsize = sizeof(uint32_t) + sizeof(ADISC);
elsiocb = lpfc_prep_els_iocb(vport, 0, cmdsize, oldiocb->retry, ndlp,
ndlp->nlp_DID, ELS_CMD_ACC);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
oldcmd = &oldiocb->iocb;
icmd->ulpContext = oldcmd->ulpContext; /* Xri */
/* Xmit ADISC ACC response tag <ulpIoTag> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0130 Xmit ADISC ACC response iotag x%x xri: "
"x%x, did x%x, nlp_flag x%x, nlp_state x%x rpi x%x\n",
elsiocb->iotag, elsiocb->iocb.ulpContext,
ndlp->nlp_DID, ndlp->nlp_flag, ndlp->nlp_state,
ndlp->nlp_rpi);
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_ACC;
pcmd += sizeof(uint32_t);
ap = (ADISC *) (pcmd);
ap->hardAL_PA = phba->fc_pref_ALPA;
memcpy(&ap->portName, &vport->fc_portname, sizeof(struct lpfc_name));
memcpy(&ap->nodeName, &vport->fc_nodename, sizeof(struct lpfc_name));
ap->DID = be32_to_cpu(vport->fc_myDID);
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_RSP,
"Issue ACC ADISC: did:x%x flg:x%x",
ndlp->nlp_DID, ndlp->nlp_flag, 0);
phba->fc_stat.elsXmitACC++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_rsp;
rc = lpfc_sli_issue_iocb(phba, pring, elsiocb, 0);
if (rc == IOCB_ERROR) {
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_els_rsp_prli_acc: Prepare and issue acc response to prli iocb cmd.
* @vport: pointer to a virtual N_Port data structure.
* @oldiocb: pointer to the original lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine prepares and issues an Accept (ACC) response to Process
* Login (PRLI) ELS command. It simply prepares the payload of the IOCB
* and invokes the lpfc_sli_issue_iocb() routine to send out the command.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the PRLI Accept response ELS IOCB command.
*
* Return code
* 0 - Successfully issued acc prli response
* 1 - Failed to issue acc prli response
**/
int
lpfc_els_rsp_prli_acc(struct lpfc_vport *vport, struct lpfc_iocbq *oldiocb,
struct lpfc_nodelist *ndlp)
{
struct lpfc_hba *phba = vport->phba;
PRLI *npr;
lpfc_vpd_t *vpd;
IOCB_t *icmd;
IOCB_t *oldcmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli_ring *pring;
struct lpfc_sli *psli;
uint8_t *pcmd;
uint16_t cmdsize;
int rc;
psli = &phba->sli;
pring = &psli->ring[LPFC_ELS_RING]; /* ELS ring */
cmdsize = sizeof(uint32_t) + sizeof(PRLI);
elsiocb = lpfc_prep_els_iocb(vport, 0, cmdsize, oldiocb->retry, ndlp,
ndlp->nlp_DID, (ELS_CMD_ACC | (ELS_CMD_PRLI & ~ELS_RSP_MASK)));
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
oldcmd = &oldiocb->iocb;
icmd->ulpContext = oldcmd->ulpContext; /* Xri */
/* Xmit PRLI ACC response tag <ulpIoTag> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0131 Xmit PRLI ACC response tag x%x xri x%x, "
"did x%x, nlp_flag x%x, nlp_state x%x, rpi x%x\n",
elsiocb->iotag, elsiocb->iocb.ulpContext,
ndlp->nlp_DID, ndlp->nlp_flag, ndlp->nlp_state,
ndlp->nlp_rpi);
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = (ELS_CMD_ACC | (ELS_CMD_PRLI & ~ELS_RSP_MASK));
pcmd += sizeof(uint32_t);
/* For PRLI, remainder of payload is PRLI parameter page */
memset(pcmd, 0, sizeof(PRLI));
npr = (PRLI *) pcmd;
vpd = &phba->vpd;
/*
* If the remote port is a target and our firmware version is 3.20 or
* later, set the following bits for FC-TAPE support.
*/
if ((ndlp->nlp_type & NLP_FCP_TARGET) &&
(vpd->rev.feaLevelHigh >= 0x02)) {
npr->ConfmComplAllowed = 1;
npr->Retry = 1;
npr->TaskRetryIdReq = 1;
}
npr->acceptRspCode = PRLI_REQ_EXECUTED;
npr->estabImagePair = 1;
npr->readXferRdyDis = 1;
npr->ConfmComplAllowed = 1;
npr->prliType = PRLI_FCP_TYPE;
npr->initiatorFunc = 1;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_RSP,
"Issue ACC PRLI: did:x%x flg:x%x",
ndlp->nlp_DID, ndlp->nlp_flag, 0);
phba->fc_stat.elsXmitACC++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_rsp;
rc = lpfc_sli_issue_iocb(phba, pring, elsiocb, 0);
if (rc == IOCB_ERROR) {
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_els_rsp_rnid_acc: Issue rnid acc response iocb command.
* @vport: pointer to a virtual N_Port data structure.
* @format: rnid command format.
* @oldiocb: pointer to the original lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine issues a Request Node Identification Data (RNID) Accept
* (ACC) response. It constructs the RNID ACC response command according to
* the proper @format and then calls the lpfc_sli_issue_iocb() routine to
* issue the response. Note that this command does not need to hold the ndlp
* reference count for the callback. So, the ndlp reference count taken by
* the lpfc_prep_els_iocb() routine is put back and the context1 field of
* IOCB is set to NULL to indicate to the lpfc_els_free_iocb() routine that
* there is no ndlp reference available.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function. However, for the RNID Accept Response ELS command,
* this is undone later by this routine after the IOCB is allocated.
*
* Return code
* 0 - Successfully issued acc rnid response
* 1 - Failed to issue acc rnid response
**/
static int
lpfc_els_rsp_rnid_acc(struct lpfc_vport *vport, uint8_t format,
struct lpfc_iocbq *oldiocb, struct lpfc_nodelist *ndlp)
{
struct lpfc_hba *phba = vport->phba;
RNID *rn;
IOCB_t *icmd, *oldcmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli_ring *pring;
struct lpfc_sli *psli;
uint8_t *pcmd;
uint16_t cmdsize;
int rc;
psli = &phba->sli;
pring = &psli->ring[LPFC_ELS_RING];
cmdsize = sizeof(uint32_t) + sizeof(uint32_t)
+ (2 * sizeof(struct lpfc_name));
if (format)
cmdsize += sizeof(RNID_TOP_DISC);
elsiocb = lpfc_prep_els_iocb(vport, 0, cmdsize, oldiocb->retry, ndlp,
ndlp->nlp_DID, ELS_CMD_ACC);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
oldcmd = &oldiocb->iocb;
icmd->ulpContext = oldcmd->ulpContext; /* Xri */
/* Xmit RNID ACC response tag <ulpIoTag> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0132 Xmit RNID ACC response tag x%x xri x%x\n",
elsiocb->iotag, elsiocb->iocb.ulpContext);
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_ACC;
pcmd += sizeof(uint32_t);
memset(pcmd, 0, sizeof(RNID));
rn = (RNID *) (pcmd);
rn->Format = format;
rn->CommonLen = (2 * sizeof(struct lpfc_name));
memcpy(&rn->portName, &vport->fc_portname, sizeof(struct lpfc_name));
memcpy(&rn->nodeName, &vport->fc_nodename, sizeof(struct lpfc_name));
switch (format) {
case 0:
rn->SpecificLen = 0;
break;
case RNID_TOPOLOGY_DISC:
rn->SpecificLen = sizeof(RNID_TOP_DISC);
memcpy(&rn->un.topologyDisc.portName,
&vport->fc_portname, sizeof(struct lpfc_name));
rn->un.topologyDisc.unitType = RNID_HBA;
rn->un.topologyDisc.physPort = 0;
rn->un.topologyDisc.attachedNodes = 0;
break;
default:
rn->CommonLen = 0;
rn->SpecificLen = 0;
break;
}
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_RSP,
"Issue ACC RNID: did:x%x flg:x%x",
ndlp->nlp_DID, ndlp->nlp_flag, 0);
phba->fc_stat.elsXmitACC++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_rsp;
lpfc_nlp_put(ndlp);
elsiocb->context1 = NULL; /* Don't need ndlp for cmpl,
* it could be freed */
rc = lpfc_sli_issue_iocb(phba, pring, elsiocb, 0);
if (rc == IOCB_ERROR) {
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_els_disc_adisc: Issue remaining adisc iocbs to npr nodes of a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine issues Address Discover (ADISC) ELS commands to those
* N_Ports which are in node port recovery state and ADISC has not been issued
* for the @vport. Each time an ELS ADISC IOCB is issued by invoking the
* lpfc_issue_els_adisc() routine, the per @vport number of discover count
* (num_disc_nodes) shall be incremented. If the num_disc_nodes reaches a
* pre-configured threshold (cfg_discovery_threads), the @vport fc_flag will
* be marked with FC_NLP_MORE bit and the process of issuing remaining ADISC
* IOCBs quit for later pick up. On the other hand, after walking through
* all the ndlps with the @vport and there is none ADISC IOCB issued, the
* @vport fc_flag shall be cleared with FC_NLP_MORE bit indicating there is
* no more ADISC need to be sent.
*
* Return code
* The number of N_Ports with adisc issued.
**/
int
lpfc_els_disc_adisc(struct lpfc_vport *vport)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_nodelist *ndlp, *next_ndlp;
int sentadisc = 0;
/* go thru NPR nodes and issue any remaining ELS ADISCs */
list_for_each_entry_safe(ndlp, next_ndlp, &vport->fc_nodes, nlp_listp) {
if (!NLP_CHK_NODE_ACT(ndlp))
continue;
if (ndlp->nlp_state == NLP_STE_NPR_NODE &&
(ndlp->nlp_flag & NLP_NPR_2B_DISC) != 0 &&
(ndlp->nlp_flag & NLP_NPR_ADISC) != 0) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~NLP_NPR_ADISC;
spin_unlock_irq(shost->host_lock);
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_ADISC_ISSUE);
lpfc_issue_els_adisc(vport, ndlp, 0);
sentadisc++;
vport->num_disc_nodes++;
if (vport->num_disc_nodes >=
vport->cfg_discovery_threads) {
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_NLP_MORE;
spin_unlock_irq(shost->host_lock);
break;
}
}
}
if (sentadisc == 0) {
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~FC_NLP_MORE;
spin_unlock_irq(shost->host_lock);
}
return sentadisc;
}
/**
* lpfc_els_disc_plogi: Issue plogi for all npr nodes of a vport before adisc.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine issues Port Login (PLOGI) ELS commands to all the N_Ports
* which are in node port recovery state, with a @vport. Each time an ELS
* ADISC PLOGI IOCB is issued by invoking the lpfc_issue_els_plogi() routine,
* the per @vport number of discover count (num_disc_nodes) shall be
* incremented. If the num_disc_nodes reaches a pre-configured threshold
* (cfg_discovery_threads), the @vport fc_flag will be marked with FC_NLP_MORE
* bit set and quit the process of issuing remaining ADISC PLOGIN IOCBs for
* later pick up. On the other hand, after walking through all the ndlps with
* the @vport and there is none ADISC PLOGI IOCB issued, the @vport fc_flag
* shall be cleared with the FC_NLP_MORE bit indicating there is no more ADISC
* PLOGI need to be sent.
*
* Return code
* The number of N_Ports with plogi issued.
**/
int
lpfc_els_disc_plogi(struct lpfc_vport *vport)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_nodelist *ndlp, *next_ndlp;
int sentplogi = 0;
/* go thru NPR nodes and issue any remaining ELS PLOGIs */
list_for_each_entry_safe(ndlp, next_ndlp, &vport->fc_nodes, nlp_listp) {
if (!NLP_CHK_NODE_ACT(ndlp))
continue;
if (ndlp->nlp_state == NLP_STE_NPR_NODE &&
(ndlp->nlp_flag & NLP_NPR_2B_DISC) != 0 &&
(ndlp->nlp_flag & NLP_DELAY_TMO) == 0 &&
(ndlp->nlp_flag & NLP_NPR_ADISC) == 0) {
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_PLOGI_ISSUE);
lpfc_issue_els_plogi(vport, ndlp->nlp_DID, 0);
sentplogi++;
vport->num_disc_nodes++;
if (vport->num_disc_nodes >=
vport->cfg_discovery_threads) {
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_NLP_MORE;
spin_unlock_irq(shost->host_lock);
break;
}
}
}
if (sentplogi) {
lpfc_set_disctmo(vport);
}
else {
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~FC_NLP_MORE;
spin_unlock_irq(shost->host_lock);
}
return sentplogi;
}
/**
* lpfc_els_flush_rscn: Clean up any rscn activities with a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine cleans up any Registration State Change Notification
* (RSCN) activity with a @vport. Note that the fc_rscn_flush flag of the
* @vport together with the host_lock is used to prevent multiple thread
* trying to access the RSCN array on a same @vport at the same time.
**/
void
lpfc_els_flush_rscn(struct lpfc_vport *vport)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
int i;
spin_lock_irq(shost->host_lock);
if (vport->fc_rscn_flush) {
/* Another thread is walking fc_rscn_id_list on this vport */
spin_unlock_irq(shost->host_lock);
return;
}
/* Indicate we are walking lpfc_els_flush_rscn on this vport */
vport->fc_rscn_flush = 1;
spin_unlock_irq(shost->host_lock);
for (i = 0; i < vport->fc_rscn_id_cnt; i++) {
lpfc_in_buf_free(phba, vport->fc_rscn_id_list[i]);
vport->fc_rscn_id_list[i] = NULL;
}
spin_lock_irq(shost->host_lock);
vport->fc_rscn_id_cnt = 0;
vport->fc_flag &= ~(FC_RSCN_MODE | FC_RSCN_DISCOVERY);
spin_unlock_irq(shost->host_lock);
lpfc_can_disctmo(vport);
/* Indicate we are done walking this fc_rscn_id_list */
vport->fc_rscn_flush = 0;
}
/**
* lpfc_rscn_payload_check: Check whether there is a pending rscn to a did.
* @vport: pointer to a host virtual N_Port data structure.
* @did: remote destination port identifier.
*
* This routine checks whether there is any pending Registration State
* Configuration Notification (RSCN) to a @did on @vport.
*
* Return code
* None zero - The @did matched with a pending rscn
* 0 - not able to match @did with a pending rscn
**/
int
lpfc_rscn_payload_check(struct lpfc_vport *vport, uint32_t did)
{
D_ID ns_did;
D_ID rscn_did;
uint32_t *lp;
uint32_t payload_len, i;
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
ns_did.un.word = did;
/* Never match fabric nodes for RSCNs */
if ((did & Fabric_DID_MASK) == Fabric_DID_MASK)
return 0;
/* If we are doing a FULL RSCN rediscovery, match everything */
if (vport->fc_flag & FC_RSCN_DISCOVERY)
return did;
spin_lock_irq(shost->host_lock);
if (vport->fc_rscn_flush) {
/* Another thread is walking fc_rscn_id_list on this vport */
spin_unlock_irq(shost->host_lock);
return 0;
}
/* Indicate we are walking fc_rscn_id_list on this vport */
vport->fc_rscn_flush = 1;
spin_unlock_irq(shost->host_lock);
for (i = 0; i < vport->fc_rscn_id_cnt; i++) {
lp = vport->fc_rscn_id_list[i]->virt;
payload_len = be32_to_cpu(*lp++ & ~ELS_CMD_MASK);
payload_len -= sizeof(uint32_t); /* take off word 0 */
while (payload_len) {
rscn_did.un.word = be32_to_cpu(*lp++);
payload_len -= sizeof(uint32_t);
switch (rscn_did.un.b.resv) {
case 0: /* Single N_Port ID effected */
if (ns_did.un.word == rscn_did.un.word)
goto return_did_out;
break;
case 1: /* Whole N_Port Area effected */
if ((ns_did.un.b.domain == rscn_did.un.b.domain)
&& (ns_did.un.b.area == rscn_did.un.b.area))
goto return_did_out;
break;
case 2: /* Whole N_Port Domain effected */
if (ns_did.un.b.domain == rscn_did.un.b.domain)
goto return_did_out;
break;
default:
/* Unknown Identifier in RSCN node */
lpfc_printf_vlog(vport, KERN_ERR, LOG_DISCOVERY,
"0217 Unknown Identifier in "
"RSCN payload Data: x%x\n",
rscn_did.un.word);
case 3: /* Whole Fabric effected */
goto return_did_out;
}
}
}
/* Indicate we are done with walking fc_rscn_id_list on this vport */
vport->fc_rscn_flush = 0;
return 0;
return_did_out:
/* Indicate we are done with walking fc_rscn_id_list on this vport */
vport->fc_rscn_flush = 0;
return did;
}
/**
* lpfc_rscn_recovery_check: Send recovery event to vport nodes matching rscn
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine sends recovery (NLP_EVT_DEVICE_RECOVERY) event to the
* state machine for a @vport's nodes that are with pending RSCN (Registration
* State Change Notification).
*
* Return code
* 0 - Successful (currently alway return 0)
**/
static int
lpfc_rscn_recovery_check(struct lpfc_vport *vport)
{
struct lpfc_nodelist *ndlp = NULL;
/* Move all affected nodes by pending RSCNs to NPR state. */
list_for_each_entry(ndlp, &vport->fc_nodes, nlp_listp) {
if (!NLP_CHK_NODE_ACT(ndlp) ||
(ndlp->nlp_state == NLP_STE_UNUSED_NODE) ||
!lpfc_rscn_payload_check(vport, ndlp->nlp_DID))
continue;
lpfc_disc_state_machine(vport, ndlp, NULL,
NLP_EVT_DEVICE_RECOVERY);
lpfc_cancel_retry_delay_tmo(vport, ndlp);
}
return 0;
}
/**
* lpfc_els_rcv_rscn: Process an unsolicited rscn iocb.
* @vport: pointer to a host virtual N_Port data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine processes an unsolicited RSCN (Registration State Change
* Notification) IOCB. First, the payload of the unsolicited RSCN is walked
* to invoke fc_host_post_event() routine to the FC transport layer. If the
* discover state machine is about to begin discovery, it just accepts the
* RSCN and the discovery process will satisfy the RSCN. If this RSCN only
* contains N_Port IDs for other vports on this HBA, it just accepts the
* RSCN and ignore processing it. If the state machine is in the recovery
* state, the fc_rscn_id_list of this @vport is walked and the
* lpfc_rscn_recovery_check() routine is invoked to send recovery event for
* all nodes that match RSCN payload. Otherwise, the lpfc_els_handle_rscn()
* routine is invoked to handle the RSCN event.
*
* Return code
* 0 - Just sent the acc response
* 1 - Sent the acc response and waited for name server completion
**/
static int
lpfc_els_rcv_rscn(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
struct lpfc_nodelist *ndlp)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
struct lpfc_dmabuf *pcmd;
uint32_t *lp, *datap;
IOCB_t *icmd;
uint32_t payload_len, length, nportid, *cmd;
int rscn_cnt;
int rscn_id = 0, hba_id = 0;
int i;
icmd = &cmdiocb->iocb;
pcmd = (struct lpfc_dmabuf *) cmdiocb->context2;
lp = (uint32_t *) pcmd->virt;
payload_len = be32_to_cpu(*lp++ & ~ELS_CMD_MASK);
payload_len -= sizeof(uint32_t); /* take off word 0 */
/* RSCN received */
lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
"0214 RSCN received Data: x%x x%x x%x x%x\n",
vport->fc_flag, payload_len, *lp,
vport->fc_rscn_id_cnt);
for (i = 0; i < payload_len/sizeof(uint32_t); i++)
fc_host_post_event(shost, fc_get_event_number(),
FCH_EVT_RSCN, lp[i]);
/* If we are about to begin discovery, just ACC the RSCN.
* Discovery processing will satisfy it.
*/
if (vport->port_state <= LPFC_NS_QRY) {
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV RSCN ignore: did:x%x/ste:x%x flg:x%x",
ndlp->nlp_DID, vport->port_state, ndlp->nlp_flag);
lpfc_els_rsp_acc(vport, ELS_CMD_ACC, cmdiocb, ndlp, NULL);
return 0;
}
/* If this RSCN just contains NPortIDs for other vports on this HBA,
* just ACC and ignore it.
*/
if ((phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) &&
!(vport->cfg_peer_port_login)) {
i = payload_len;
datap = lp;
while (i > 0) {
nportid = *datap++;
nportid = ((be32_to_cpu(nportid)) & Mask_DID);
i -= sizeof(uint32_t);
rscn_id++;
if (lpfc_find_vport_by_did(phba, nportid))
hba_id++;
}
if (rscn_id == hba_id) {
/* ALL NPortIDs in RSCN are on HBA */
lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
"0214 Ignore RSCN "
"Data: x%x x%x x%x x%x\n",
vport->fc_flag, payload_len,
*lp, vport->fc_rscn_id_cnt);
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV RSCN vport: did:x%x/ste:x%x flg:x%x",
ndlp->nlp_DID, vport->port_state,
ndlp->nlp_flag);
lpfc_els_rsp_acc(vport, ELS_CMD_ACC, cmdiocb,
ndlp, NULL);
return 0;
}
}
spin_lock_irq(shost->host_lock);
if (vport->fc_rscn_flush) {
/* Another thread is walking fc_rscn_id_list on this vport */
spin_unlock_irq(shost->host_lock);
vport->fc_flag |= FC_RSCN_DISCOVERY;
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
/* Send back ACC */
lpfc_els_rsp_acc(vport, ELS_CMD_ACC, cmdiocb, ndlp, NULL);
return 0;
}
/* Indicate we are walking fc_rscn_id_list on this vport */
vport->fc_rscn_flush = 1;
spin_unlock_irq(shost->host_lock);
/* Get the array count after sucessfully have the token */
rscn_cnt = vport->fc_rscn_id_cnt;
/* If we are already processing an RSCN, save the received
* RSCN payload buffer, cmdiocb->context2 to process later.
*/
if (vport->fc_flag & (FC_RSCN_MODE | FC_NDISC_ACTIVE)) {
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV RSCN defer: did:x%x/ste:x%x flg:x%x",
ndlp->nlp_DID, vport->port_state, ndlp->nlp_flag);
[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 13:52:54 +07:00
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_RSCN_DEFERRED;
if ((rscn_cnt < FC_MAX_HOLD_RSCN) &&
!(vport->fc_flag & FC_RSCN_DISCOVERY)) {
vport->fc_flag |= FC_RSCN_MODE;
spin_unlock_irq(shost->host_lock);
if (rscn_cnt) {
cmd = vport->fc_rscn_id_list[rscn_cnt-1]->virt;
length = be32_to_cpu(*cmd & ~ELS_CMD_MASK);
}
if ((rscn_cnt) &&
(payload_len + length <= LPFC_BPL_SIZE)) {
*cmd &= ELS_CMD_MASK;
*cmd |= cpu_to_be32(payload_len + length);
memcpy(((uint8_t *)cmd) + length, lp,
payload_len);
} else {
vport->fc_rscn_id_list[rscn_cnt] = pcmd;
vport->fc_rscn_id_cnt++;
/* If we zero, cmdiocb->context2, the calling
* routine will not try to free it.
*/
cmdiocb->context2 = NULL;
}
/* Deferred RSCN */
lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
"0235 Deferred RSCN "
"Data: x%x x%x x%x\n",
vport->fc_rscn_id_cnt, vport->fc_flag,
vport->port_state);
} else {
vport->fc_flag |= FC_RSCN_DISCOVERY;
spin_unlock_irq(shost->host_lock);
/* ReDiscovery RSCN */
lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
"0234 ReDiscovery RSCN "
"Data: x%x x%x x%x\n",
vport->fc_rscn_id_cnt, vport->fc_flag,
vport->port_state);
}
/* Indicate we are done walking fc_rscn_id_list on this vport */
vport->fc_rscn_flush = 0;
/* Send back ACC */
lpfc_els_rsp_acc(vport, ELS_CMD_ACC, cmdiocb, ndlp, NULL);
/* send RECOVERY event for ALL nodes that match RSCN payload */
lpfc_rscn_recovery_check(vport);
[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 13:52:54 +07:00
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~FC_RSCN_DEFERRED;
[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 13:52:54 +07:00
spin_unlock_irq(shost->host_lock);
return 0;
}
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV RSCN: did:x%x/ste:x%x flg:x%x",
ndlp->nlp_DID, vport->port_state, ndlp->nlp_flag);
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_RSCN_MODE;
spin_unlock_irq(shost->host_lock);
vport->fc_rscn_id_list[vport->fc_rscn_id_cnt++] = pcmd;
/* Indicate we are done walking fc_rscn_id_list on this vport */
vport->fc_rscn_flush = 0;
/*
* If we zero, cmdiocb->context2, the calling routine will
* not try to free it.
*/
cmdiocb->context2 = NULL;
lpfc_set_disctmo(vport);
/* Send back ACC */
lpfc_els_rsp_acc(vport, ELS_CMD_ACC, cmdiocb, ndlp, NULL);
/* send RECOVERY event for ALL nodes that match RSCN payload */
lpfc_rscn_recovery_check(vport);
return lpfc_els_handle_rscn(vport);
}
/**
* lpfc_els_handle_rscn: Handle rscn for a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine handles the Registration State Configuration Notification
* (RSCN) for a @vport. If login to NameServer does not exist, a new ndlp shall
* be created and a Port Login (PLOGI) to the NameServer is issued. Otherwise,
* if the ndlp to NameServer exists, a Common Transport (CT) command to the
* NameServer shall be issued. If CT command to the NameServer fails to be
* issued, the lpfc_els_flush_rscn() routine shall be invoked to clean up any
* RSCN activities with the @vport.
*
* Return code
* 0 - Cleaned up rscn on the @vport
* 1 - Wait for plogi to name server before proceed
**/
int
lpfc_els_handle_rscn(struct lpfc_vport *vport)
{
struct lpfc_nodelist *ndlp;
struct lpfc_hba *phba = vport->phba;
/* Ignore RSCN if the port is being torn down. */
if (vport->load_flag & FC_UNLOADING) {
lpfc_els_flush_rscn(vport);
return 0;
}
/* Start timer for RSCN processing */
lpfc_set_disctmo(vport);
/* RSCN processed */
lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
"0215 RSCN processed Data: x%x x%x x%x x%x\n",
vport->fc_flag, 0, vport->fc_rscn_id_cnt,
vport->port_state);
/* To process RSCN, first compare RSCN data with NameServer */
vport->fc_ns_retry = 0;
vport->num_disc_nodes = 0;
ndlp = lpfc_findnode_did(vport, NameServer_DID);
if (ndlp && NLP_CHK_NODE_ACT(ndlp)
&& ndlp->nlp_state == NLP_STE_UNMAPPED_NODE) {
/* Good ndlp, issue CT Request to NameServer */
if (lpfc_ns_cmd(vport, SLI_CTNS_GID_FT, 0, 0) == 0)
/* Wait for NameServer query cmpl before we can
continue */
return 1;
} else {
/* If login to NameServer does not exist, issue one */
/* Good status, issue PLOGI to NameServer */
ndlp = lpfc_findnode_did(vport, NameServer_DID);
if (ndlp && NLP_CHK_NODE_ACT(ndlp))
/* Wait for NameServer login cmpl before we can
continue */
return 1;
if (ndlp) {
ndlp = lpfc_enable_node(vport, ndlp,
NLP_STE_PLOGI_ISSUE);
if (!ndlp) {
lpfc_els_flush_rscn(vport);
return 0;
}
ndlp->nlp_prev_state = NLP_STE_UNUSED_NODE;
} else {
ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
if (!ndlp) {
lpfc_els_flush_rscn(vport);
return 0;
}
lpfc_nlp_init(vport, ndlp, NameServer_DID);
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_PLOGI_ISSUE);
}
ndlp->nlp_type |= NLP_FABRIC;
lpfc_issue_els_plogi(vport, NameServer_DID, 0);
/* Wait for NameServer login cmpl before we can
* continue
*/
return 1;
}
lpfc_els_flush_rscn(vport);
return 0;
}
/**
* lpfc_els_rcv_flogi: Process an unsolicited flogi iocb.
* @vport: pointer to a host virtual N_Port data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine processes Fabric Login (FLOGI) IOCB received as an ELS
* unsolicited event. An unsolicited FLOGI can be received in a point-to-
* point topology. As an unsolicited FLOGI should not be received in a loop
* mode, any unsolicited FLOGI received in loop mode shall be ignored. The
* lpfc_check_sparm() routine is invoked to check the parameters in the
* unsolicited FLOGI. If parameters validation failed, the routine
* lpfc_els_rsp_reject() shall be called with reject reason code set to
* LSEXP_SPARM_OPTIONS to reject the FLOGI. Otherwise, the Port WWN in the
* FLOGI shall be compared with the Port WWN of the @vport to determine who
* will initiate PLOGI. The higher lexicographical value party shall has
* higher priority (as the winning port) and will initiate PLOGI and
* communicate Port_IDs (Addresses) for both nodes in PLOGI. The result
* of this will be marked in the @vport fc_flag field with FC_PT2PT_PLOGI
* and then the lpfc_els_rsp_acc() routine is invoked to accept the FLOGI.
*
* Return code
* 0 - Successfully processed the unsolicited flogi
* 1 - Failed to process the unsolicited flogi
**/
static int
lpfc_els_rcv_flogi(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
struct lpfc_nodelist *ndlp)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
struct lpfc_dmabuf *pcmd = (struct lpfc_dmabuf *) cmdiocb->context2;
uint32_t *lp = (uint32_t *) pcmd->virt;
IOCB_t *icmd = &cmdiocb->iocb;
struct serv_parm *sp;
LPFC_MBOXQ_t *mbox;
struct ls_rjt stat;
uint32_t cmd, did;
int rc;
cmd = *lp++;
sp = (struct serv_parm *) lp;
/* FLOGI received */
lpfc_set_disctmo(vport);
if (phba->fc_topology == TOPOLOGY_LOOP) {
/* We should never receive a FLOGI in loop mode, ignore it */
did = icmd->un.elsreq64.remoteID;
/* An FLOGI ELS command <elsCmd> was received from DID <did> in
Loop Mode */
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0113 An FLOGI ELS command x%x was "
"received from DID x%x in Loop Mode\n",
cmd, did);
return 1;
}
did = Fabric_DID;
if ((lpfc_check_sparm(vport, ndlp, sp, CLASS3))) {
/* For a FLOGI we accept, then if our portname is greater
* then the remote portname we initiate Nport login.
*/
rc = memcmp(&vport->fc_portname, &sp->portName,
sizeof(struct lpfc_name));
if (!rc) {
mbox = mempool_alloc(phba->mbox_mem_pool, GFP_KERNEL);
if (!mbox)
return 1;
lpfc_linkdown(phba);
lpfc_init_link(phba, mbox,
phba->cfg_topology,
phba->cfg_link_speed);
mbox->mb.un.varInitLnk.lipsr_AL_PA = 0;
mbox->mbox_cmpl = lpfc_sli_def_mbox_cmpl;
mbox->vport = vport;
rc = lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT);
lpfc_set_loopback_flag(phba);
if (rc == MBX_NOT_FINISHED) {
mempool_free(mbox, phba->mbox_mem_pool);
}
return 1;
} else if (rc > 0) { /* greater than */
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_PT2PT_PLOGI;
spin_unlock_irq(shost->host_lock);
}
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_PT2PT;
vport->fc_flag &= ~(FC_FABRIC | FC_PUBLIC_LOOP);
spin_unlock_irq(shost->host_lock);
} else {
/* Reject this request because invalid parameters */
stat.un.b.lsRjtRsvd0 = 0;
stat.un.b.lsRjtRsnCode = LSRJT_UNABLE_TPC;
stat.un.b.lsRjtRsnCodeExp = LSEXP_SPARM_OPTIONS;
stat.un.b.vendorUnique = 0;
lpfc_els_rsp_reject(vport, stat.un.lsRjtError, cmdiocb, ndlp,
NULL);
return 1;
}
/* Send back ACC */
lpfc_els_rsp_acc(vport, ELS_CMD_PLOGI, cmdiocb, ndlp, NULL);
return 0;
}
/**
* lpfc_els_rcv_rnid: Process an unsolicited rnid iocb.
* @vport: pointer to a host virtual N_Port data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine processes Request Node Identification Data (RNID) IOCB
* received as an ELS unsolicited event. Only when the RNID specified format
* 0x0 or 0xDF (Topology Discovery Specific Node Identification Data)
* present, this routine will invoke the lpfc_els_rsp_rnid_acc() routine to
* Accept (ACC) the RNID ELS command. All the other RNID formats are
* rejected by invoking the lpfc_els_rsp_reject() routine.
*
* Return code
* 0 - Successfully processed rnid iocb (currently always return 0)
**/
static int
lpfc_els_rcv_rnid(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
struct lpfc_nodelist *ndlp)
{
struct lpfc_dmabuf *pcmd;
uint32_t *lp;
IOCB_t *icmd;
RNID *rn;
struct ls_rjt stat;
uint32_t cmd, did;
icmd = &cmdiocb->iocb;
did = icmd->un.elsreq64.remoteID;
pcmd = (struct lpfc_dmabuf *) cmdiocb->context2;
lp = (uint32_t *) pcmd->virt;
cmd = *lp++;
rn = (RNID *) lp;
/* RNID received */
switch (rn->Format) {
case 0:
case RNID_TOPOLOGY_DISC:
/* Send back ACC */
lpfc_els_rsp_rnid_acc(vport, rn->Format, cmdiocb, ndlp);
break;
default:
/* Reject this request because format not supported */
stat.un.b.lsRjtRsvd0 = 0;
stat.un.b.lsRjtRsnCode = LSRJT_UNABLE_TPC;
stat.un.b.lsRjtRsnCodeExp = LSEXP_CANT_GIVE_DATA;
stat.un.b.vendorUnique = 0;
lpfc_els_rsp_reject(vport, stat.un.lsRjtError, cmdiocb, ndlp,
NULL);
}
return 0;
}
/**
* lpfc_els_rcv_lirr: Process an unsolicited lirr iocb.
* @vport: pointer to a host virtual N_Port data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine processes a Link Incident Report Registration(LIRR) IOCB
* received as an ELS unsolicited event. Currently, this function just invokes
* the lpfc_els_rsp_reject() routine to reject the LIRR IOCB unconditionally.
*
* Return code
* 0 - Successfully processed lirr iocb (currently always return 0)
**/
static int
lpfc_els_rcv_lirr(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
struct lpfc_nodelist *ndlp)
{
struct ls_rjt stat;
/* For now, unconditionally reject this command */
stat.un.b.lsRjtRsvd0 = 0;
stat.un.b.lsRjtRsnCode = LSRJT_UNABLE_TPC;
stat.un.b.lsRjtRsnCodeExp = LSEXP_CANT_GIVE_DATA;
stat.un.b.vendorUnique = 0;
lpfc_els_rsp_reject(vport, stat.un.lsRjtError, cmdiocb, ndlp, NULL);
return 0;
}
/**
* lpfc_els_rsp_rps_acc: Completion callbk func for MBX_READ_LNK_STAT mbox cmd.
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* This routine is the completion callback function for the MBX_READ_LNK_STAT
* mailbox command. This callback function is to actually send the Accept
* (ACC) response to a Read Port Status (RPS) unsolicited IOCB event. It
* collects the link statistics from the completion of the MBX_READ_LNK_STAT
* mailbox command, constructs the RPS response with the link statistics
* collected, and then invokes the lpfc_sli_issue_iocb() routine to send ACC
* response to the RPS.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the RPS Accept Response ELS IOCB command.
*
**/
static void
lpfc_els_rsp_rps_acc(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
{
struct lpfc_sli *psli = &phba->sli;
struct lpfc_sli_ring *pring = &psli->ring[LPFC_ELS_RING];
MAILBOX_t *mb;
IOCB_t *icmd;
RPS_RSP *rps_rsp;
uint8_t *pcmd;
struct lpfc_iocbq *elsiocb;
struct lpfc_nodelist *ndlp;
uint16_t xri, status;
uint32_t cmdsize;
mb = &pmb->mb;
ndlp = (struct lpfc_nodelist *) pmb->context2;
xri = (uint16_t) ((unsigned long)(pmb->context1));
pmb->context1 = NULL;
pmb->context2 = NULL;
if (mb->mbxStatus) {
mempool_free(pmb, phba->mbox_mem_pool);
return;
}
cmdsize = sizeof(RPS_RSP) + sizeof(uint32_t);
mempool_free(pmb, phba->mbox_mem_pool);
elsiocb = lpfc_prep_els_iocb(phba->pport, 0, cmdsize,
lpfc_max_els_tries, ndlp,
ndlp->nlp_DID, ELS_CMD_ACC);
/* Decrement the ndlp reference count from previous mbox command */
lpfc_nlp_put(ndlp);
if (!elsiocb)
return;
icmd = &elsiocb->iocb;
icmd->ulpContext = xri;
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_ACC;
pcmd += sizeof(uint32_t); /* Skip past command */
rps_rsp = (RPS_RSP *)pcmd;
if (phba->fc_topology != TOPOLOGY_LOOP)
status = 0x10;
else
status = 0x8;
if (phba->pport->fc_flag & FC_FABRIC)
status |= 0x4;
rps_rsp->rsvd1 = 0;
[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 13:52:54 +07:00
rps_rsp->portStatus = cpu_to_be16(status);
rps_rsp->linkFailureCnt = cpu_to_be32(mb->un.varRdLnk.linkFailureCnt);
rps_rsp->lossSyncCnt = cpu_to_be32(mb->un.varRdLnk.lossSyncCnt);
rps_rsp->lossSignalCnt = cpu_to_be32(mb->un.varRdLnk.lossSignalCnt);
rps_rsp->primSeqErrCnt = cpu_to_be32(mb->un.varRdLnk.primSeqErrCnt);
rps_rsp->invalidXmitWord = cpu_to_be32(mb->un.varRdLnk.invalidXmitWord);
rps_rsp->crcCnt = cpu_to_be32(mb->un.varRdLnk.crcCnt);
/* Xmit ELS RPS ACC response tag <ulpIoTag> */
lpfc_printf_vlog(ndlp->vport, KERN_INFO, LOG_ELS,
"0118 Xmit ELS RPS ACC response tag x%x xri x%x, "
"did x%x, nlp_flag x%x, nlp_state x%x, rpi x%x\n",
elsiocb->iotag, elsiocb->iocb.ulpContext,
ndlp->nlp_DID, ndlp->nlp_flag, ndlp->nlp_state,
ndlp->nlp_rpi);
elsiocb->iocb_cmpl = lpfc_cmpl_els_rsp;
phba->fc_stat.elsXmitACC++;
if (lpfc_sli_issue_iocb(phba, pring, elsiocb, 0) == IOCB_ERROR)
lpfc_els_free_iocb(phba, elsiocb);
return;
}
/**
* lpfc_els_rcv_rps: Process an unsolicited rps iocb.
* @vport: pointer to a host virtual N_Port data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine processes Read Port Status (RPS) IOCB received as an
* ELS unsolicited event. It first checks the remote port state. If the
* remote port is not in NLP_STE_UNMAPPED_NODE state or NLP_STE_MAPPED_NODE
* state, it invokes the lpfc_els_rsp_reject() routine to send the reject
* response. Otherwise, it issue the MBX_READ_LNK_STAT mailbox command
* for reading the HBA link statistics. It is for the callback function,
* lpfc_els_rsp_rps_acc(), set to the MBX_READ_LNK_STAT mailbox command
* to actually sending out RPS Accept (ACC) response.
*
* Return codes
* 0 - Successfully processed rps iocb (currently always return 0)
**/
static int
lpfc_els_rcv_rps(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
struct lpfc_nodelist *ndlp)
{
struct lpfc_hba *phba = vport->phba;
uint32_t *lp;
uint8_t flag;
LPFC_MBOXQ_t *mbox;
struct lpfc_dmabuf *pcmd;
RPS *rps;
struct ls_rjt stat;
if ((ndlp->nlp_state != NLP_STE_UNMAPPED_NODE) &&
(ndlp->nlp_state != NLP_STE_MAPPED_NODE))
/* reject the unsolicited RPS request and done with it */
goto reject_out;
pcmd = (struct lpfc_dmabuf *) cmdiocb->context2;
lp = (uint32_t *) pcmd->virt;
flag = (be32_to_cpu(*lp++) & 0xf);
rps = (RPS *) lp;
if ((flag == 0) ||
((flag == 1) && (be32_to_cpu(rps->un.portNum) == 0)) ||
((flag == 2) && (memcmp(&rps->un.portName, &vport->fc_portname,
sizeof(struct lpfc_name)) == 0))) {
printk("Fix me....\n");
dump_stack();
mbox = mempool_alloc(phba->mbox_mem_pool, GFP_ATOMIC);
if (mbox) {
lpfc_read_lnk_stat(phba, mbox);
mbox->context1 =
(void *)((unsigned long) cmdiocb->iocb.ulpContext);
mbox->context2 = lpfc_nlp_get(ndlp);
mbox->vport = vport;
mbox->mbox_cmpl = lpfc_els_rsp_rps_acc;
if (lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT)
!= MBX_NOT_FINISHED)
/* Mbox completion will send ELS Response */
return 0;
/* Decrement reference count used for the failed mbox
* command.
*/
lpfc_nlp_put(ndlp);
mempool_free(mbox, phba->mbox_mem_pool);
}
}
reject_out:
/* issue rejection response */
stat.un.b.lsRjtRsvd0 = 0;
stat.un.b.lsRjtRsnCode = LSRJT_UNABLE_TPC;
stat.un.b.lsRjtRsnCodeExp = LSEXP_CANT_GIVE_DATA;
stat.un.b.vendorUnique = 0;
lpfc_els_rsp_reject(vport, stat.un.lsRjtError, cmdiocb, ndlp, NULL);
return 0;
}
/**
* lpfc_els_rsp_rpl_acc: Issue an accept rpl els command.
* @vport: pointer to a host virtual N_Port data structure.
* @cmdsize: size of the ELS command.
* @oldiocb: pointer to the original lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine issuees an Accept (ACC) Read Port List (RPL) ELS command.
* It is to be called by the lpfc_els_rcv_rpl() routine to accept the RPL.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the RPL Accept Response ELS command.
*
* Return code
* 0 - Successfully issued ACC RPL ELS command
* 1 - Failed to issue ACC RPL ELS command
**/
static int
lpfc_els_rsp_rpl_acc(struct lpfc_vport *vport, uint16_t cmdsize,
struct lpfc_iocbq *oldiocb, struct lpfc_nodelist *ndlp)
{
struct lpfc_hba *phba = vport->phba;
IOCB_t *icmd, *oldcmd;
RPL_RSP rpl_rsp;
struct lpfc_iocbq *elsiocb;
struct lpfc_sli *psli = &phba->sli;
struct lpfc_sli_ring *pring = &psli->ring[LPFC_ELS_RING];
uint8_t *pcmd;
elsiocb = lpfc_prep_els_iocb(vport, 0, cmdsize, oldiocb->retry, ndlp,
ndlp->nlp_DID, ELS_CMD_ACC);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
oldcmd = &oldiocb->iocb;
icmd->ulpContext = oldcmd->ulpContext; /* Xri */
pcmd = (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_ACC;
pcmd += sizeof(uint16_t);
*((uint16_t *)(pcmd)) = be16_to_cpu(cmdsize);
pcmd += sizeof(uint16_t);
/* Setup the RPL ACC payload */
rpl_rsp.listLen = be32_to_cpu(1);
rpl_rsp.index = 0;
rpl_rsp.port_num_blk.portNum = 0;
rpl_rsp.port_num_blk.portID = be32_to_cpu(vport->fc_myDID);
memcpy(&rpl_rsp.port_num_blk.portName, &vport->fc_portname,
sizeof(struct lpfc_name));
memcpy(pcmd, &rpl_rsp, cmdsize - sizeof(uint32_t));
/* Xmit ELS RPL ACC response tag <ulpIoTag> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0120 Xmit ELS RPL ACC response tag x%x "
"xri x%x, did x%x, nlp_flag x%x, nlp_state x%x, "
"rpi x%x\n",
elsiocb->iotag, elsiocb->iocb.ulpContext,
ndlp->nlp_DID, ndlp->nlp_flag, ndlp->nlp_state,
ndlp->nlp_rpi);
elsiocb->iocb_cmpl = lpfc_cmpl_els_rsp;
phba->fc_stat.elsXmitACC++;
if (lpfc_sli_issue_iocb(phba, pring, elsiocb, 0) == IOCB_ERROR) {
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_els_rcv_rpl: Process an unsolicited rpl iocb.
* @vport: pointer to a host virtual N_Port data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine processes Read Port List (RPL) IOCB received as an ELS
* unsolicited event. It first checks the remote port state. If the remote
* port is not in NLP_STE_UNMAPPED_NODE and NLP_STE_MAPPED_NODE states, it
* invokes the lpfc_els_rsp_reject() routine to send reject response.
* Otherwise, this routine then invokes the lpfc_els_rsp_rpl_acc() routine
* to accept the RPL.
*
* Return code
* 0 - Successfully processed rpl iocb (currently always return 0)
**/
static int
lpfc_els_rcv_rpl(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
struct lpfc_nodelist *ndlp)
{
struct lpfc_dmabuf *pcmd;
uint32_t *lp;
uint32_t maxsize;
uint16_t cmdsize;
RPL *rpl;
struct ls_rjt stat;
if ((ndlp->nlp_state != NLP_STE_UNMAPPED_NODE) &&
(ndlp->nlp_state != NLP_STE_MAPPED_NODE)) {
/* issue rejection response */
stat.un.b.lsRjtRsvd0 = 0;
stat.un.b.lsRjtRsnCode = LSRJT_UNABLE_TPC;
stat.un.b.lsRjtRsnCodeExp = LSEXP_CANT_GIVE_DATA;
stat.un.b.vendorUnique = 0;
lpfc_els_rsp_reject(vport, stat.un.lsRjtError, cmdiocb, ndlp,
NULL);
/* rejected the unsolicited RPL request and done with it */
return 0;
}
pcmd = (struct lpfc_dmabuf *) cmdiocb->context2;
lp = (uint32_t *) pcmd->virt;
rpl = (RPL *) (lp + 1);
maxsize = be32_to_cpu(rpl->maxsize);
/* We support only one port */
if ((rpl->index == 0) &&
((maxsize == 0) ||
((maxsize * sizeof(uint32_t)) >= sizeof(RPL_RSP)))) {
cmdsize = sizeof(uint32_t) + sizeof(RPL_RSP);
} else {
cmdsize = sizeof(uint32_t) + maxsize * sizeof(uint32_t);
}
lpfc_els_rsp_rpl_acc(vport, cmdsize, cmdiocb, ndlp);
return 0;
}
/**
* lpfc_els_rcv_farp: Process an unsolicited farp request els command.
* @vport: pointer to a virtual N_Port data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine processes Fibre Channel Address Resolution Protocol
* (FARP) Request IOCB received as an ELS unsolicited event. Currently,
* the lpfc driver only supports matching on WWPN or WWNN for FARP. As such,
* FARP_MATCH_PORT flag and FARP_MATCH_NODE flag are checked against the
* Match Flag in the FARP request IOCB: if FARP_MATCH_PORT flag is set, the
* remote PortName is compared against the FC PortName stored in the @vport
* data structure; if FARP_MATCH_NODE flag is set, the remote NodeName is
* compared against the FC NodeName stored in the @vport data structure.
* If any of these matches and the FARP_REQUEST_FARPR flag is set in the
* FARP request IOCB Response Flag, the lpfc_issue_els_farpr() routine is
* invoked to send out FARP Response to the remote node. Before sending the
* FARP Response, however, the FARP_REQUEST_PLOGI flag is check in the FARP
* request IOCB Response Flag and, if it is set, the lpfc_issue_els_plogi()
* routine is invoked to log into the remote port first.
*
* Return code
* 0 - Either the FARP Match Mode not supported or successfully processed
**/
static int
lpfc_els_rcv_farp(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
struct lpfc_nodelist *ndlp)
{
struct lpfc_dmabuf *pcmd;
uint32_t *lp;
IOCB_t *icmd;
FARP *fp;
uint32_t cmd, cnt, did;
icmd = &cmdiocb->iocb;
did = icmd->un.elsreq64.remoteID;
pcmd = (struct lpfc_dmabuf *) cmdiocb->context2;
lp = (uint32_t *) pcmd->virt;
cmd = *lp++;
fp = (FARP *) lp;
/* FARP-REQ received from DID <did> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0601 FARP-REQ received from DID x%x\n", did);
/* We will only support match on WWPN or WWNN */
if (fp->Mflags & ~(FARP_MATCH_NODE | FARP_MATCH_PORT)) {
return 0;
}
cnt = 0;
/* If this FARP command is searching for my portname */
if (fp->Mflags & FARP_MATCH_PORT) {
if (memcmp(&fp->RportName, &vport->fc_portname,
sizeof(struct lpfc_name)) == 0)
cnt = 1;
}
/* If this FARP command is searching for my nodename */
if (fp->Mflags & FARP_MATCH_NODE) {
if (memcmp(&fp->RnodeName, &vport->fc_nodename,
sizeof(struct lpfc_name)) == 0)
cnt = 1;
}
if (cnt) {
if ((ndlp->nlp_state == NLP_STE_UNMAPPED_NODE) ||
(ndlp->nlp_state == NLP_STE_MAPPED_NODE)) {
/* Log back into the node before sending the FARP. */
if (fp->Rflags & FARP_REQUEST_PLOGI) {
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp,
NLP_STE_PLOGI_ISSUE);
lpfc_issue_els_plogi(vport, ndlp->nlp_DID, 0);
}
/* Send a FARP response to that node */
if (fp->Rflags & FARP_REQUEST_FARPR)
lpfc_issue_els_farpr(vport, did, 0);
}
}
return 0;
}
/**
* lpfc_els_rcv_farpr: Process an unsolicited farp response iocb.
* @vport: pointer to a host virtual N_Port data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine processes Fibre Channel Address Resolution Protocol
* Response (FARPR) IOCB received as an ELS unsolicited event. It simply
* invokes the lpfc_els_rsp_acc() routine to the remote node to accept
* the FARP response request.
*
* Return code
* 0 - Successfully processed FARPR IOCB (currently always return 0)
**/
static int
lpfc_els_rcv_farpr(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
struct lpfc_nodelist *ndlp)
{
struct lpfc_dmabuf *pcmd;
uint32_t *lp;
IOCB_t *icmd;
uint32_t cmd, did;
icmd = &cmdiocb->iocb;
did = icmd->un.elsreq64.remoteID;
pcmd = (struct lpfc_dmabuf *) cmdiocb->context2;
lp = (uint32_t *) pcmd->virt;
cmd = *lp++;
/* FARP-RSP received from DID <did> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0600 FARP-RSP received from DID x%x\n", did);
/* ACCEPT the Farp resp request */
lpfc_els_rsp_acc(vport, ELS_CMD_ACC, cmdiocb, ndlp, NULL);
return 0;
}
/**
* lpfc_els_rcv_fan: Process an unsolicited fan iocb command.
* @vport: pointer to a host virtual N_Port data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @fan_ndlp: pointer to a node-list data structure.
*
* This routine processes a Fabric Address Notification (FAN) IOCB
* command received as an ELS unsolicited event. The FAN ELS command will
* only be processed on a physical port (i.e., the @vport represents the
* physical port). The fabric NodeName and PortName from the FAN IOCB are
* compared against those in the phba data structure. If any of those is
* different, the lpfc_initial_flogi() routine is invoked to initialize
* Fabric Login (FLOGI) to the fabric to start the discover over. Otherwise,
* if both of those are identical, the lpfc_issue_fabric_reglogin() routine
* is invoked to register login to the fabric.
*
* Return code
* 0 - Successfully processed fan iocb (currently always return 0).
**/
static int
lpfc_els_rcv_fan(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
struct lpfc_nodelist *fan_ndlp)
{
struct lpfc_hba *phba = vport->phba;
uint32_t *lp;
FAN *fp;
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS, "0265 FAN received\n");
lp = (uint32_t *)((struct lpfc_dmabuf *)cmdiocb->context2)->virt;
fp = (FAN *) ++lp;
/* FAN received; Fan does not have a reply sequence */
if ((vport == phba->pport) &&
(vport->port_state == LPFC_LOCAL_CFG_LINK)) {
if ((memcmp(&phba->fc_fabparam.nodeName, &fp->FnodeName,
sizeof(struct lpfc_name))) ||
(memcmp(&phba->fc_fabparam.portName, &fp->FportName,
sizeof(struct lpfc_name)))) {
/* This port has switched fabrics. FLOGI is required */
lpfc_initial_flogi(vport);
} else {
/* FAN verified - skip FLOGI */
vport->fc_myDID = vport->fc_prevDID;
lpfc_issue_fabric_reglogin(vport);
}
}
return 0;
}
/**
* lpfc_els_timeout: Handler funciton to the els timer.
* @ptr: holder for the timer function associated data.
*
* This routine is invoked by the ELS timer after timeout. It posts the ELS
* timer timeout event by setting the WORKER_ELS_TMO bit to the work port
* event bitmap and then invokes the lpfc_worker_wake_up() routine to wake
* up the worker thread. It is for the worker thread to invoke the routine
* lpfc_els_timeout_handler() to work on the posted event WORKER_ELS_TMO.
**/
void
lpfc_els_timeout(unsigned long ptr)
{
struct lpfc_vport *vport = (struct lpfc_vport *) ptr;
struct lpfc_hba *phba = vport->phba;
uint32_t tmo_posted;
unsigned long iflag;
spin_lock_irqsave(&vport->work_port_lock, iflag);
tmo_posted = vport->work_port_events & WORKER_ELS_TMO;
if (!tmo_posted)
vport->work_port_events |= WORKER_ELS_TMO;
spin_unlock_irqrestore(&vport->work_port_lock, iflag);
if (!tmo_posted)
lpfc_worker_wake_up(phba);
return;
}
/**
* lpfc_els_timeout_handler: Process an els timeout event.
* @vport: pointer to a virtual N_Port data structure.
*
* This routine is the actual handler function that processes an ELS timeout
* event. It walks the ELS ring to get and abort all the IOCBs (except the
* ABORT/CLOSE/FARP/FARPR/FDISC), which are associated with the @vport by
* invoking the lpfc_sli_issue_abort_iotag() routine.
**/
void
lpfc_els_timeout_handler(struct lpfc_vport *vport)
{
struct lpfc_hba *phba = vport->phba;
struct lpfc_sli_ring *pring;
struct lpfc_iocbq *tmp_iocb, *piocb;
IOCB_t *cmd = NULL;
struct lpfc_dmabuf *pcmd;
uint32_t els_command = 0;
uint32_t timeout;
uint32_t remote_ID = 0xffffffff;
/* If the timer is already canceled do nothing */
if ((vport->work_port_events & WORKER_ELS_TMO) == 0) {
return;
}
spin_lock_irq(&phba->hbalock);
timeout = (uint32_t)(phba->fc_ratov << 1);
pring = &phba->sli.ring[LPFC_ELS_RING];
list_for_each_entry_safe(piocb, tmp_iocb, &pring->txcmplq, list) {
cmd = &piocb->iocb;
if ((piocb->iocb_flag & LPFC_IO_LIBDFC) != 0 ||
piocb->iocb.ulpCommand == CMD_ABORT_XRI_CN ||
piocb->iocb.ulpCommand == CMD_CLOSE_XRI_CN)
continue;
if (piocb->vport != vport)
continue;
pcmd = (struct lpfc_dmabuf *) piocb->context2;
if (pcmd)
els_command = *(uint32_t *) (pcmd->virt);
if (els_command == ELS_CMD_FARP ||
els_command == ELS_CMD_FARPR ||
els_command == ELS_CMD_FDISC)
continue;
if (piocb->drvrTimeout > 0) {
if (piocb->drvrTimeout >= timeout)
piocb->drvrTimeout -= timeout;
else
piocb->drvrTimeout = 0;
continue;
}
remote_ID = 0xffffffff;
if (cmd->ulpCommand != CMD_GEN_REQUEST64_CR)
remote_ID = cmd->un.elsreq64.remoteID;
else {
struct lpfc_nodelist *ndlp;
ndlp = __lpfc_findnode_rpi(vport, cmd->ulpContext);
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
if (ndlp && NLP_CHK_NODE_ACT(ndlp))
remote_ID = ndlp->nlp_DID;
}
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0127 ELS timeout Data: x%x x%x x%x "
"x%x\n", els_command,
remote_ID, cmd->ulpCommand, cmd->ulpIoTag);
lpfc_sli_issue_abort_iotag(phba, pring, piocb);
}
spin_unlock_irq(&phba->hbalock);
if (phba->sli.ring[LPFC_ELS_RING].txcmplq_cnt)
mod_timer(&vport->els_tmofunc, jiffies + HZ * timeout);
}
/**
* lpfc_els_flush_cmd: Clean up the outstanding els commands to a vport.
* @vport: pointer to a host virtual N_Port data structure.
*
* This routine is used to clean up all the outstanding ELS commands on a
* @vport. It first aborts the @vport by invoking lpfc_fabric_abort_vport()
* routine. After that, it walks the ELS transmit queue to remove all the
* IOCBs with the @vport other than the QUE_RING and ABORT/CLOSE IOCBs. For
* the IOCBs with a non-NULL completion callback function, the callback
* function will be invoked with the status set to IOSTAT_LOCAL_REJECT and
* un.ulpWord[4] set to IOERR_SLI_ABORTED. For IOCBs with a NULL completion
* callback function, the IOCB will simply be released. Finally, it walks
* the ELS transmit completion queue to issue an abort IOCB to any transmit
* completion queue IOCB that is associated with the @vport and is not
* an IOCB from libdfc (i.e., the management plane IOCBs that are not
* part of the discovery state machine) out to HBA by invoking the
* lpfc_sli_issue_abort_iotag() routine. Note that this function issues the
* abort IOCB to any transmit completion queueed IOCB, it does not guarantee
* the IOCBs are aborted when this function returns.
**/
void
lpfc_els_flush_cmd(struct lpfc_vport *vport)
{
LIST_HEAD(completions);
struct lpfc_hba *phba = vport->phba;
struct lpfc_sli_ring *pring = &phba->sli.ring[LPFC_ELS_RING];
struct lpfc_iocbq *tmp_iocb, *piocb;
IOCB_t *cmd = NULL;
lpfc_fabric_abort_vport(vport);
spin_lock_irq(&phba->hbalock);
list_for_each_entry_safe(piocb, tmp_iocb, &pring->txq, list) {
cmd = &piocb->iocb;
if (piocb->iocb_flag & LPFC_IO_LIBDFC) {
continue;
}
/* Do not flush out the QUE_RING and ABORT/CLOSE iocbs */
if (cmd->ulpCommand == CMD_QUE_RING_BUF_CN ||
cmd->ulpCommand == CMD_QUE_RING_BUF64_CN ||
cmd->ulpCommand == CMD_CLOSE_XRI_CN ||
cmd->ulpCommand == CMD_ABORT_XRI_CN)
continue;
if (piocb->vport != vport)
continue;
list_move_tail(&piocb->list, &completions);
pring->txq_cnt--;
}
list_for_each_entry_safe(piocb, tmp_iocb, &pring->txcmplq, list) {
if (piocb->iocb_flag & LPFC_IO_LIBDFC) {
continue;
}
if (piocb->vport != vport)
continue;
lpfc_sli_issue_abort_iotag(phba, pring, piocb);
}
spin_unlock_irq(&phba->hbalock);
while (!list_empty(&completions)) {
piocb = list_get_first(&completions, struct lpfc_iocbq, list);
cmd = &piocb->iocb;
list_del_init(&piocb->list);
if (!piocb->iocb_cmpl)
lpfc_sli_release_iocbq(phba, piocb);
else {
cmd->ulpStatus = IOSTAT_LOCAL_REJECT;
cmd->un.ulpWord[4] = IOERR_SLI_ABORTED;
(piocb->iocb_cmpl) (phba, piocb, piocb);
}
}
return;
}
/**
* lpfc_els_flush_all_cmd: Clean up all the outstanding els commands to a HBA.
* @phba: pointer to lpfc hba data structure.
*
* This routine is used to clean up all the outstanding ELS commands on a
* @phba. It first aborts the @phba by invoking the lpfc_fabric_abort_hba()
* routine. After that, it walks the ELS transmit queue to remove all the
* IOCBs to the @phba other than the QUE_RING and ABORT/CLOSE IOCBs. For
* the IOCBs with the completion callback function associated, the callback
* function will be invoked with the status set to IOSTAT_LOCAL_REJECT and
* un.ulpWord[4] set to IOERR_SLI_ABORTED. For IOCBs without the completion
* callback function associated, the IOCB will simply be released. Finally,
* it walks the ELS transmit completion queue to issue an abort IOCB to any
* transmit completion queue IOCB that is not an IOCB from libdfc (i.e., the
* management plane IOCBs that are not part of the discovery state machine)
* out to HBA by invoking the lpfc_sli_issue_abort_iotag() routine.
**/
void
lpfc_els_flush_all_cmd(struct lpfc_hba *phba)
{
LIST_HEAD(completions);
struct lpfc_sli_ring *pring = &phba->sli.ring[LPFC_ELS_RING];
struct lpfc_iocbq *tmp_iocb, *piocb;
IOCB_t *cmd = NULL;
lpfc_fabric_abort_hba(phba);
spin_lock_irq(&phba->hbalock);
list_for_each_entry_safe(piocb, tmp_iocb, &pring->txq, list) {
cmd = &piocb->iocb;
if (piocb->iocb_flag & LPFC_IO_LIBDFC)
continue;
/* Do not flush out the QUE_RING and ABORT/CLOSE iocbs */
if (cmd->ulpCommand == CMD_QUE_RING_BUF_CN ||
cmd->ulpCommand == CMD_QUE_RING_BUF64_CN ||
cmd->ulpCommand == CMD_CLOSE_XRI_CN ||
cmd->ulpCommand == CMD_ABORT_XRI_CN)
continue;
list_move_tail(&piocb->list, &completions);
pring->txq_cnt--;
}
list_for_each_entry_safe(piocb, tmp_iocb, &pring->txcmplq, list) {
if (piocb->iocb_flag & LPFC_IO_LIBDFC)
continue;
lpfc_sli_issue_abort_iotag(phba, pring, piocb);
}
spin_unlock_irq(&phba->hbalock);
while (!list_empty(&completions)) {
piocb = list_get_first(&completions, struct lpfc_iocbq, list);
cmd = &piocb->iocb;
list_del_init(&piocb->list);
if (!piocb->iocb_cmpl)
lpfc_sli_release_iocbq(phba, piocb);
else {
cmd->ulpStatus = IOSTAT_LOCAL_REJECT;
cmd->un.ulpWord[4] = IOERR_SLI_ABORTED;
(piocb->iocb_cmpl) (phba, piocb, piocb);
}
}
return;
}
/**
* lpfc_els_unsol_buffer: Process an unsolicited event data buffer.
* @phba: pointer to lpfc hba data structure.
* @pring: pointer to a SLI ring.
* @vport: pointer to a host virtual N_Port data structure.
* @elsiocb: pointer to lpfc els command iocb data structure.
*
* This routine is used for processing the IOCB associated with a unsolicited
* event. It first determines whether there is an existing ndlp that matches
* the DID from the unsolicited IOCB. If not, it will create a new one with
* the DID from the unsolicited IOCB. The ELS command from the unsolicited
* IOCB is then used to invoke the proper routine and to set up proper state
* of the discovery state machine.
**/
static void
lpfc_els_unsol_buffer(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
struct lpfc_vport *vport, struct lpfc_iocbq *elsiocb)
{
struct Scsi_Host *shost;
struct lpfc_nodelist *ndlp;
struct ls_rjt stat;
uint32_t *payload;
uint32_t cmd, did, newnode, rjt_err = 0;
IOCB_t *icmd = &elsiocb->iocb;
if (!vport || !(elsiocb->context2))
goto dropit;
newnode = 0;
payload = ((struct lpfc_dmabuf *)elsiocb->context2)->virt;
cmd = *payload;
if ((phba->sli3_options & LPFC_SLI3_HBQ_ENABLED) == 0)
lpfc_post_buffer(phba, pring, 1);
did = icmd->un.rcvels.remoteID;
if (icmd->ulpStatus) {
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV Unsol ELS: status:x%x/x%x did:x%x",
icmd->ulpStatus, icmd->un.ulpWord[4], did);
goto dropit;
}
/* Check to see if link went down during discovery */
if (lpfc_els_chk_latt(vport))
goto dropit;
/* Ignore traffic recevied during vport shutdown. */
if (vport->load_flag & FC_UNLOADING)
goto dropit;
ndlp = lpfc_findnode_did(vport, did);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
if (!ndlp)
goto dropit;
lpfc_nlp_init(vport, ndlp, did);
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
newnode = 1;
if ((did & Fabric_DID_MASK) == Fabric_DID_MASK)
ndlp->nlp_type |= NLP_FABRIC;
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp,
NLP_STE_UNUSED_NODE);
if (!ndlp)
goto dropit;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
newnode = 1;
if ((did & Fabric_DID_MASK) == Fabric_DID_MASK)
ndlp->nlp_type |= NLP_FABRIC;
} else if (ndlp->nlp_state == NLP_STE_UNUSED_NODE) {
/* This is similar to the new node path */
ndlp = lpfc_nlp_get(ndlp);
if (!ndlp)
goto dropit;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
newnode = 1;
}
phba->fc_stat.elsRcvFrame++;
if (elsiocb->context1)
lpfc_nlp_put(elsiocb->context1);
elsiocb->context1 = lpfc_nlp_get(ndlp);
elsiocb->vport = vport;
if ((cmd & ELS_CMD_MASK) == ELS_CMD_RSCN) {
cmd &= ELS_CMD_MASK;
}
/* ELS command <elsCmd> received from NPORT <did> */
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0112 ELS command x%x received from NPORT x%x "
"Data: x%x\n", cmd, did, vport->port_state);
switch (cmd) {
case ELS_CMD_PLOGI:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV PLOGI: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvPLOGI++;
ndlp = lpfc_plogi_confirm_nport(phba, payload, ndlp);
if (vport->port_state < LPFC_DISC_AUTH) {
if (!(phba->pport->fc_flag & FC_PT2PT) ||
(phba->pport->fc_flag & FC_PT2PT_PLOGI)) {
rjt_err = LSRJT_UNABLE_TPC;
break;
}
/* We get here, and drop thru, if we are PT2PT with
* another NPort and the other side has initiated
* the PLOGI before responding to our FLOGI.
*/
}
shost = lpfc_shost_from_vport(vport);
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~NLP_TARGET_REMOVE;
spin_unlock_irq(shost->host_lock);
lpfc_disc_state_machine(vport, ndlp, elsiocb,
NLP_EVT_RCV_PLOGI);
break;
case ELS_CMD_FLOGI:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV FLOGI: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvFLOGI++;
lpfc_els_rcv_flogi(vport, elsiocb, ndlp);
if (newnode)
lpfc_nlp_put(ndlp);
break;
case ELS_CMD_LOGO:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV LOGO: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvLOGO++;
if (vport->port_state < LPFC_DISC_AUTH) {
rjt_err = LSRJT_UNABLE_TPC;
break;
}
lpfc_disc_state_machine(vport, ndlp, elsiocb, NLP_EVT_RCV_LOGO);
break;
case ELS_CMD_PRLO:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV PRLO: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvPRLO++;
if (vport->port_state < LPFC_DISC_AUTH) {
rjt_err = LSRJT_UNABLE_TPC;
break;
}
lpfc_disc_state_machine(vport, ndlp, elsiocb, NLP_EVT_RCV_PRLO);
break;
case ELS_CMD_RSCN:
phba->fc_stat.elsRcvRSCN++;
lpfc_els_rcv_rscn(vport, elsiocb, ndlp);
if (newnode)
lpfc_nlp_put(ndlp);
break;
case ELS_CMD_ADISC:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV ADISC: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvADISC++;
if (vport->port_state < LPFC_DISC_AUTH) {
rjt_err = LSRJT_UNABLE_TPC;
break;
}
lpfc_disc_state_machine(vport, ndlp, elsiocb,
NLP_EVT_RCV_ADISC);
break;
case ELS_CMD_PDISC:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV PDISC: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvPDISC++;
if (vport->port_state < LPFC_DISC_AUTH) {
rjt_err = LSRJT_UNABLE_TPC;
break;
}
lpfc_disc_state_machine(vport, ndlp, elsiocb,
NLP_EVT_RCV_PDISC);
break;
case ELS_CMD_FARPR:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV FARPR: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvFARPR++;
lpfc_els_rcv_farpr(vport, elsiocb, ndlp);
break;
case ELS_CMD_FARP:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV FARP: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvFARP++;
lpfc_els_rcv_farp(vport, elsiocb, ndlp);
break;
case ELS_CMD_FAN:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV FAN: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvFAN++;
lpfc_els_rcv_fan(vport, elsiocb, ndlp);
break;
case ELS_CMD_PRLI:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV PRLI: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvPRLI++;
if (vport->port_state < LPFC_DISC_AUTH) {
rjt_err = LSRJT_UNABLE_TPC;
break;
}
lpfc_disc_state_machine(vport, ndlp, elsiocb, NLP_EVT_RCV_PRLI);
break;
case ELS_CMD_LIRR:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV LIRR: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvLIRR++;
lpfc_els_rcv_lirr(vport, elsiocb, ndlp);
if (newnode)
lpfc_nlp_put(ndlp);
break;
case ELS_CMD_RPS:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV RPS: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvRPS++;
lpfc_els_rcv_rps(vport, elsiocb, ndlp);
if (newnode)
lpfc_nlp_put(ndlp);
break;
case ELS_CMD_RPL:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV RPL: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvRPL++;
lpfc_els_rcv_rpl(vport, elsiocb, ndlp);
if (newnode)
lpfc_nlp_put(ndlp);
break;
case ELS_CMD_RNID:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV RNID: did:x%x/ste:x%x flg:x%x",
did, vport->port_state, ndlp->nlp_flag);
phba->fc_stat.elsRcvRNID++;
lpfc_els_rcv_rnid(vport, elsiocb, ndlp);
if (newnode)
lpfc_nlp_put(ndlp);
break;
default:
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
"RCV ELS cmd: cmd:x%x did:x%x/ste:x%x",
cmd, did, vport->port_state);
/* Unsupported ELS command, reject */
rjt_err = LSRJT_INVALID_CMD;
/* Unknown ELS command <elsCmd> received from NPORT <did> */
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0115 Unknown ELS command x%x "
"received from NPORT x%x\n", cmd, did);
if (newnode)
lpfc_nlp_put(ndlp);
break;
}
/* check if need to LS_RJT received ELS cmd */
if (rjt_err) {
memset(&stat, 0, sizeof(stat));
stat.un.b.lsRjtRsnCode = rjt_err;
stat.un.b.lsRjtRsnCodeExp = LSEXP_NOTHING_MORE;
lpfc_els_rsp_reject(vport, stat.un.lsRjtError, elsiocb, ndlp,
NULL);
}
return;
dropit:
if (vport && !(vport->load_flag & FC_UNLOADING))
lpfc_printf_log(phba, KERN_ERR, LOG_ELS,
"(%d):0111 Dropping received ELS cmd "
"Data: x%x x%x x%x\n",
vport->vpi, icmd->ulpStatus,
icmd->un.ulpWord[4], icmd->ulpTimeout);
phba->fc_stat.elsRcvDrop++;
}
/**
* lpfc_find_vport_by_vpid: Find a vport on a HBA through vport identifier.
* @phba: pointer to lpfc hba data structure.
* @vpi: host virtual N_Port identifier.
*
* This routine finds a vport on a HBA (referred by @phba) through a
* @vpi. The function walks the HBA's vport list and returns the address
* of the vport with the matching @vpi.
*
* Return code
* NULL - No vport with the matching @vpi found
* Otherwise - Address to the vport with the matching @vpi.
**/
static struct lpfc_vport *
lpfc_find_vport_by_vpid(struct lpfc_hba *phba, uint16_t vpi)
{
struct lpfc_vport *vport;
unsigned long flags;
spin_lock_irqsave(&phba->hbalock, flags);
list_for_each_entry(vport, &phba->port_list, listentry) {
if (vport->vpi == vpi) {
spin_unlock_irqrestore(&phba->hbalock, flags);
return vport;
}
}
spin_unlock_irqrestore(&phba->hbalock, flags);
return NULL;
}
/**
* lpfc_els_unsol_event: Process an unsolicited event from an els sli ring.
* @phba: pointer to lpfc hba data structure.
* @pring: pointer to a SLI ring.
* @elsiocb: pointer to lpfc els iocb data structure.
*
* This routine is used to process an unsolicited event received from a SLI
* (Service Level Interface) ring. The actual processing of the data buffer
* associated with the unsolicited event is done by invoking the routine
* lpfc_els_unsol_buffer() after properly set up the iocb buffer from the
* SLI ring on which the unsolicited event was received.
**/
void
lpfc_els_unsol_event(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
struct lpfc_iocbq *elsiocb)
{
struct lpfc_vport *vport = phba->pport;
IOCB_t *icmd = &elsiocb->iocb;
dma_addr_t paddr;
struct lpfc_dmabuf *bdeBuf1 = elsiocb->context2;
struct lpfc_dmabuf *bdeBuf2 = elsiocb->context3;
elsiocb->context2 = NULL;
elsiocb->context3 = NULL;
if (icmd->ulpStatus == IOSTAT_NEED_BUFFER) {
lpfc_sli_hbqbuf_add_hbqs(phba, LPFC_ELS_HBQ);
} else if (icmd->ulpStatus == IOSTAT_LOCAL_REJECT &&
(icmd->un.ulpWord[4] & 0xff) == IOERR_RCV_BUFFER_WAITING) {
phba->fc_stat.NoRcvBuf++;
/* Not enough posted buffers; Try posting more buffers */
if (!(phba->sli3_options & LPFC_SLI3_HBQ_ENABLED))
lpfc_post_buffer(phba, pring, 0);
return;
}
if ((phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) &&
(icmd->ulpCommand == CMD_IOCB_RCV_ELS64_CX ||
icmd->ulpCommand == CMD_IOCB_RCV_SEQ64_CX)) {
if (icmd->unsli3.rcvsli3.vpi == 0xffff)
vport = phba->pport;
else {
uint16_t vpi = icmd->unsli3.rcvsli3.vpi;
vport = lpfc_find_vport_by_vpid(phba, vpi);
}
}
/* If there are no BDEs associated
* with this IOCB, there is nothing to do.
*/
if (icmd->ulpBdeCount == 0)
return;
/* type of ELS cmd is first 32bit word
* in packet
*/
if (phba->sli3_options & LPFC_SLI3_HBQ_ENABLED) {
elsiocb->context2 = bdeBuf1;
} else {
paddr = getPaddr(icmd->un.cont64[0].addrHigh,
icmd->un.cont64[0].addrLow);
elsiocb->context2 = lpfc_sli_ringpostbuf_get(phba, pring,
paddr);
}
lpfc_els_unsol_buffer(phba, pring, vport, elsiocb);
/*
* The different unsolicited event handlers would tell us
* if they are done with "mp" by setting context2 to NULL.
*/
lpfc_nlp_put(elsiocb->context1);
elsiocb->context1 = NULL;
if (elsiocb->context2) {
lpfc_in_buf_free(phba, (struct lpfc_dmabuf *)elsiocb->context2);
elsiocb->context2 = NULL;
}
/* RCV_ELS64_CX provide for 2 BDEs - process 2nd if included */
if ((phba->sli3_options & LPFC_SLI3_HBQ_ENABLED) &&
icmd->ulpBdeCount == 2) {
elsiocb->context2 = bdeBuf2;
lpfc_els_unsol_buffer(phba, pring, vport, elsiocb);
/* free mp if we are done with it */
if (elsiocb->context2) {
lpfc_in_buf_free(phba, elsiocb->context2);
elsiocb->context2 = NULL;
}
}
}
/**
* lpfc_do_scr_ns_plogi: Issue a plogi to the name server for scr.
* @phba: pointer to lpfc hba data structure.
* @vport: pointer to a virtual N_Port data structure.
*
* This routine issues a Port Login (PLOGI) to the Name Server with
* State Change Request (SCR) for a @vport. This routine will create an
* ndlp for the Name Server associated to the @vport if such node does
* not already exist. The PLOGI to Name Server is issued by invoking the
* lpfc_issue_els_plogi() routine. If Fabric-Device Management Interface
* (FDMI) is configured to the @vport, a FDMI node will be created and
* the PLOGI to FDMI is issued by invoking lpfc_issue_els_plogi() routine.
**/
void
lpfc_do_scr_ns_plogi(struct lpfc_hba *phba, struct lpfc_vport *vport)
{
struct lpfc_nodelist *ndlp, *ndlp_fdmi;
ndlp = lpfc_findnode_did(vport, NameServer_DID);
if (!ndlp) {
ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
if (!ndlp) {
if (phba->fc_topology == TOPOLOGY_LOOP) {
lpfc_disc_start(vport);
return;
}
lpfc_vport_set_state(vport, FC_VPORT_FAILED);
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0251 NameServer login: no memory\n");
return;
}
lpfc_nlp_init(vport, ndlp, NameServer_DID);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
if (!ndlp) {
if (phba->fc_topology == TOPOLOGY_LOOP) {
lpfc_disc_start(vport);
return;
}
lpfc_vport_set_state(vport, FC_VPORT_FAILED);
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0348 NameServer login: node freed\n");
return;
}
}
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
ndlp->nlp_type |= NLP_FABRIC;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_PLOGI_ISSUE);
if (lpfc_issue_els_plogi(vport, ndlp->nlp_DID, 0)) {
lpfc_vport_set_state(vport, FC_VPORT_FAILED);
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0252 Cannot issue NameServer login\n");
return;
}
if (vport->cfg_fdmi_on) {
ndlp_fdmi = mempool_alloc(phba->nlp_mem_pool,
GFP_KERNEL);
if (ndlp_fdmi) {
lpfc_nlp_init(vport, ndlp_fdmi, FDMI_DID);
ndlp_fdmi->nlp_type |= NLP_FABRIC;
[SCSI] lpfc 8.2.6 : Multiple discovery fixes Multiple Discovery Fixes: - Fix race on discovery due to link events coinciding with vport_delete. - Use NLP_FABRIC state to filter out switch-based pseudo initiators that reuse the same WWNs. - Correct erroneous setting of DID=0 in lpfc_matchdid() - Correct extra reference count that was in the lookup path for the remoteid from an unsolicited ELS. - Correct double-free bug in els abort path. - Correct FDMI server discovery logic for switch that return a WWN of 0. - Fix bugs in ndlp mgmt when a node changes address - Correct bug that did not delete RSCNs for vports upon link transitions - Fix "0216 Link event during NS query" error which pops up when vports are swapped to different switch ports. - Add sanity checks on ndlp structures - Fix devloss log message to dump WWN correctly - Hold off mgmt commands that were interferring with discovery mailbox cmds - Remove unnecessary FC_ESTABLISH_LINK logic. - Correct some race conditions in the worker thread, resulting in devloss: - Clear the work_port_events field before handling the work port events - Clear the deferred ring event before handling a deferred ring event - Hold the hba lock when waking up the work thread - Send an acc for the rscn even when we aren't going to handle it - Fix locking behavior that was not properly protecting the ACTIVE flag, thus allowing mailbox command order to shift. Signed-off-by: James Smart <james.smart@emulex.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-04-07 21:15:56 +07:00
lpfc_nlp_set_state(vport, ndlp_fdmi,
NLP_STE_PLOGI_ISSUE);
lpfc_issue_els_plogi(vport, ndlp_fdmi->nlp_DID,
0);
}
}
return;
}
/**
* lpfc_cmpl_reg_new_vport: Completion callback function to register new vport.
* @phba: pointer to lpfc hba data structure.
* @pmb: pointer to the driver internal queue element for mailbox command.
*
* This routine is the completion callback function to register new vport
* mailbox command. If the new vport mailbox command completes successfully,
* the fabric registration login shall be performed on physical port (the
* new vport created is actually a physical port, with VPI 0) or the port
* login to Name Server for State Change Request (SCR) will be performed
* on virtual port (real virtual port, with VPI greater than 0).
**/
static void
lpfc_cmpl_reg_new_vport(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
{
struct lpfc_vport *vport = pmb->vport;
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_nodelist *ndlp = (struct lpfc_nodelist *) pmb->context2;
MAILBOX_t *mb = &pmb->mb;
[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 13:52:54 +07:00
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~FC_VPORT_NEEDS_REG_VPI;
[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 13:52:54 +07:00
spin_unlock_irq(shost->host_lock);
if (mb->mbxStatus) {
lpfc_printf_vlog(vport, KERN_ERR, LOG_MBOX,
"0915 Register VPI failed: 0x%x\n",
mb->mbxStatus);
switch (mb->mbxStatus) {
case 0x11: /* unsupported feature */
case 0x9603: /* max_vpi exceeded */
case 0x9602: /* Link event since CLEAR_LA */
/* giving up on vport registration */
lpfc_vport_set_state(vport, FC_VPORT_FAILED);
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~(FC_FABRIC | FC_PUBLIC_LOOP);
spin_unlock_irq(shost->host_lock);
lpfc_can_disctmo(vport);
break;
default:
/* Try to recover from this error */
lpfc_mbx_unreg_vpi(vport);
[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 13:52:54 +07:00
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_VPORT_NEEDS_REG_VPI;
[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 13:52:54 +07:00
spin_unlock_irq(shost->host_lock);
if (vport->port_type == LPFC_PHYSICAL_PORT)
lpfc_initial_flogi(vport);
else
lpfc_initial_fdisc(vport);
break;
}
} else {
if (vport == phba->pport)
lpfc_issue_fabric_reglogin(vport);
else
lpfc_do_scr_ns_plogi(phba, vport);
}
/* Now, we decrement the ndlp reference count held for this
* callback function
*/
lpfc_nlp_put(ndlp);
mempool_free(pmb, phba->mbox_mem_pool);
return;
}
/**
* lpfc_register_new_vport: Register a new vport with a HBA.
* @phba: pointer to lpfc hba data structure.
* @vport: pointer to a host virtual N_Port data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine registers the @vport as a new virtual port with a HBA.
* It is done through a registering vpi mailbox command.
**/
static void
lpfc_register_new_vport(struct lpfc_hba *phba, struct lpfc_vport *vport,
struct lpfc_nodelist *ndlp)
{
[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 13:52:54 +07:00
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
LPFC_MBOXQ_t *mbox;
mbox = mempool_alloc(phba->mbox_mem_pool, GFP_KERNEL);
if (mbox) {
lpfc_reg_vpi(phba, vport->vpi, vport->fc_myDID, mbox);
mbox->vport = vport;
mbox->context2 = lpfc_nlp_get(ndlp);
mbox->mbox_cmpl = lpfc_cmpl_reg_new_vport;
if (lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT)
== MBX_NOT_FINISHED) {
/* mailbox command not success, decrement ndlp
* reference count for this command
*/
lpfc_nlp_put(ndlp);
mempool_free(mbox, phba->mbox_mem_pool);
lpfc_printf_vlog(vport, KERN_ERR, LOG_MBOX,
"0253 Register VPI: Can't send mbox\n");
goto mbox_err_exit;
}
} else {
lpfc_printf_vlog(vport, KERN_ERR, LOG_MBOX,
"0254 Register VPI: no memory\n");
goto mbox_err_exit;
}
return;
mbox_err_exit:
lpfc_vport_set_state(vport, FC_VPORT_FAILED);
spin_lock_irq(shost->host_lock);
vport->fc_flag &= ~FC_VPORT_NEEDS_REG_VPI;
spin_unlock_irq(shost->host_lock);
return;
}
/**
* lpfc_cmpl_els_fdisc: Completion function for fdisc iocb command.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is the completion callback function to a Fabric Discover
* (FDISC) ELS command. Since all the FDISC ELS commands are issued
* single threaded, each FDISC completion callback function will reset
* the discovery timer for all vports such that the timers will not get
* unnecessary timeout. The function checks the FDISC IOCB status. If error
* detected, the vport will be set to FC_VPORT_FAILED state. Otherwise,the
* vport will set to FC_VPORT_ACTIVE state. It then checks whether the DID
* assigned to the vport has been changed with the completion of the FDISC
* command. If so, both RPI (Remote Port Index) and VPI (Virtual Port Index)
* are unregistered from the HBA, and then the lpfc_register_new_vport()
* routine is invoked to register new vport with the HBA. Otherwise, the
* lpfc_do_scr_ns_plogi() routine is invoked to issue a PLOGI to the Name
* Server for State Change Request (SCR).
**/
static void
lpfc_cmpl_els_fdisc(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_vport *vport = cmdiocb->vport;
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_nodelist *ndlp = (struct lpfc_nodelist *) cmdiocb->context1;
struct lpfc_nodelist *np;
struct lpfc_nodelist *next_np;
IOCB_t *irsp = &rspiocb->iocb;
struct lpfc_iocbq *piocb;
lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
"0123 FDISC completes. x%x/x%x prevDID: x%x\n",
irsp->ulpStatus, irsp->un.ulpWord[4],
vport->fc_prevDID);
/* Since all FDISCs are being single threaded, we
* must reset the discovery timer for ALL vports
* waiting to send FDISC when one completes.
*/
list_for_each_entry(piocb, &phba->fabric_iocb_list, list) {
lpfc_set_disctmo(piocb->vport);
}
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"FDISC cmpl: status:x%x/x%x prevdid:x%x",
irsp->ulpStatus, irsp->un.ulpWord[4], vport->fc_prevDID);
if (irsp->ulpStatus) {
/* Check for retry */
if (lpfc_els_retry(phba, cmdiocb, rspiocb))
goto out;
/* FDISC failed */
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0124 FDISC failed. (%d/%d)\n",
irsp->ulpStatus, irsp->un.ulpWord[4]);
if (vport->fc_vport->vport_state == FC_VPORT_INITIALIZING)
lpfc_vport_set_state(vport, FC_VPORT_FAILED);
lpfc_nlp_put(ndlp);
/* giving up on FDISC. Cancel discovery timer */
lpfc_can_disctmo(vport);
} else {
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_FABRIC;
if (vport->phba->fc_topology == TOPOLOGY_LOOP)
vport->fc_flag |= FC_PUBLIC_LOOP;
spin_unlock_irq(shost->host_lock);
vport->fc_myDID = irsp->un.ulpWord[4] & Mask_DID;
lpfc_vport_set_state(vport, FC_VPORT_ACTIVE);
if ((vport->fc_prevDID != vport->fc_myDID) &&
!(vport->fc_flag & FC_VPORT_NEEDS_REG_VPI)) {
/* If our NportID changed, we need to ensure all
* remaining NPORTs get unreg_login'ed so we can
* issue unreg_vpi.
*/
list_for_each_entry_safe(np, next_np,
&vport->fc_nodes, nlp_listp) {
if (!NLP_CHK_NODE_ACT(ndlp) ||
(np->nlp_state != NLP_STE_NPR_NODE) ||
!(np->nlp_flag & NLP_NPR_ADISC))
continue;
spin_lock_irq(shost->host_lock);
np->nlp_flag &= ~NLP_NPR_ADISC;
spin_unlock_irq(shost->host_lock);
lpfc_unreg_rpi(vport, np);
}
lpfc_mbx_unreg_vpi(vport);
[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 13:52:54 +07:00
spin_lock_irq(shost->host_lock);
vport->fc_flag |= FC_VPORT_NEEDS_REG_VPI;
[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 13:52:54 +07:00
spin_unlock_irq(shost->host_lock);
}
if (vport->fc_flag & FC_VPORT_NEEDS_REG_VPI)
lpfc_register_new_vport(phba, vport, ndlp);
else
lpfc_do_scr_ns_plogi(phba, vport);
/* Unconditionaly kick off releasing fabric node for vports */
lpfc_nlp_put(ndlp);
}
out:
lpfc_els_free_iocb(phba, cmdiocb);
}
/**
* lpfc_issue_els_fdisc: Issue a fdisc iocb command.
* @vport: pointer to a virtual N_Port data structure.
* @ndlp: pointer to a node-list data structure.
* @retry: number of retries to the command IOCB.
*
* This routine prepares and issues a Fabric Discover (FDISC) IOCB to
* a remote node (@ndlp) off a @vport. It uses the lpfc_issue_fabric_iocb()
* routine to issue the IOCB, which makes sure only one outstanding fabric
* IOCB will be sent off HBA at any given time.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the FDISC ELS command.
*
* Return code
* 0 - Successfully issued fdisc iocb command
* 1 - Failed to issue fdisc iocb command
**/
static int
lpfc_issue_els_fdisc(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
uint8_t retry)
{
struct lpfc_hba *phba = vport->phba;
IOCB_t *icmd;
struct lpfc_iocbq *elsiocb;
struct serv_parm *sp;
uint8_t *pcmd;
uint16_t cmdsize;
int did = ndlp->nlp_DID;
int rc;
cmdsize = (sizeof(uint32_t) + sizeof(struct serv_parm));
elsiocb = lpfc_prep_els_iocb(vport, 1, cmdsize, retry, ndlp, did,
ELS_CMD_FDISC);
if (!elsiocb) {
lpfc_vport_set_state(vport, FC_VPORT_FAILED);
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0255 Issue FDISC: no IOCB\n");
return 1;
}
icmd = &elsiocb->iocb;
icmd->un.elsreq64.myID = 0;
icmd->un.elsreq64.fl = 1;
/* For FDISC, Let FDISC rsp set the NPortID for this VPI */
icmd->ulpCt_h = 1;
icmd->ulpCt_l = 0;
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_FDISC;
pcmd += sizeof(uint32_t); /* CSP Word 1 */
memcpy(pcmd, &vport->phba->pport->fc_sparam, sizeof(struct serv_parm));
sp = (struct serv_parm *) pcmd;
/* Setup CSPs accordingly for Fabric */
sp->cmn.e_d_tov = 0;
sp->cmn.w2.r_a_tov = 0;
sp->cls1.classValid = 0;
sp->cls2.seqDelivery = 1;
sp->cls3.seqDelivery = 1;
pcmd += sizeof(uint32_t); /* CSP Word 2 */
pcmd += sizeof(uint32_t); /* CSP Word 3 */
pcmd += sizeof(uint32_t); /* CSP Word 4 */
pcmd += sizeof(uint32_t); /* Port Name */
memcpy(pcmd, &vport->fc_portname, 8);
pcmd += sizeof(uint32_t); /* Node Name */
pcmd += sizeof(uint32_t); /* Node Name */
memcpy(pcmd, &vport->fc_nodename, 8);
lpfc_set_disctmo(vport);
phba->fc_stat.elsXmitFDISC++;
elsiocb->iocb_cmpl = lpfc_cmpl_els_fdisc;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"Issue FDISC: did:x%x",
did, 0, 0);
rc = lpfc_issue_fabric_iocb(phba, elsiocb);
if (rc == IOCB_ERROR) {
lpfc_els_free_iocb(phba, elsiocb);
lpfc_vport_set_state(vport, FC_VPORT_FAILED);
lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
"0256 Issue FDISC: Cannot send IOCB\n");
return 1;
}
lpfc_vport_set_state(vport, FC_VPORT_INITIALIZING);
vport->port_state = LPFC_FDISC;
return 0;
}
/**
* lpfc_cmpl_els_npiv_logo: Completion function with vport logo.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is the completion callback function to the issuing of a LOGO
* ELS command off a vport. It frees the command IOCB and then decrement the
* reference count held on ndlp for this completion function, indicating that
* the reference to the ndlp is no long needed. Note that the
* lpfc_els_free_iocb() routine decrements the ndlp reference held for this
* callback function and an additional explicit ndlp reference decrementation
* will trigger the actual release of the ndlp.
**/
static void
lpfc_cmpl_els_npiv_logo(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct lpfc_vport *vport = cmdiocb->vport;
IOCB_t *irsp;
struct lpfc_nodelist *ndlp;
ndlp = (struct lpfc_nodelist *)cmdiocb->context1;
irsp = &rspiocb->iocb;
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"LOGO npiv cmpl: status:x%x/x%x did:x%x",
irsp->ulpStatus, irsp->un.ulpWord[4], irsp->un.rcvels.remoteID);
lpfc_els_free_iocb(phba, cmdiocb);
vport->unreg_vpi_cmpl = VPORT_ERROR;
/* Trigger the release of the ndlp after logo */
lpfc_nlp_put(ndlp);
}
/**
* lpfc_issue_els_npiv_logo: Issue a logo off a vport.
* @vport: pointer to a virtual N_Port data structure.
* @ndlp: pointer to a node-list data structure.
*
* This routine issues a LOGO ELS command to an @ndlp off a @vport.
*
* Note that, in lpfc_prep_els_iocb() routine, the reference count of ndlp
* will be incremented by 1 for holding the ndlp and the reference to ndlp
* will be stored into the context1 field of the IOCB for the completion
* callback function to the LOGO ELS command.
*
* Return codes
* 0 - Successfully issued logo off the @vport
* 1 - Failed to issue logo off the @vport
**/
int
lpfc_issue_els_npiv_logo(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct lpfc_hba *phba = vport->phba;
struct lpfc_sli_ring *pring = &phba->sli.ring[LPFC_ELS_RING];
IOCB_t *icmd;
struct lpfc_iocbq *elsiocb;
uint8_t *pcmd;
uint16_t cmdsize;
cmdsize = 2 * sizeof(uint32_t) + sizeof(struct lpfc_name);
elsiocb = lpfc_prep_els_iocb(vport, 1, cmdsize, 0, ndlp, ndlp->nlp_DID,
ELS_CMD_LOGO);
if (!elsiocb)
return 1;
icmd = &elsiocb->iocb;
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
*((uint32_t *) (pcmd)) = ELS_CMD_LOGO;
pcmd += sizeof(uint32_t);
/* Fill in LOGO payload */
*((uint32_t *) (pcmd)) = be32_to_cpu(vport->fc_myDID);
pcmd += sizeof(uint32_t);
memcpy(pcmd, &vport->fc_portname, sizeof(struct lpfc_name));
lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_CMD,
"Issue LOGO npiv did:x%x flg:x%x",
ndlp->nlp_DID, ndlp->nlp_flag, 0);
elsiocb->iocb_cmpl = lpfc_cmpl_els_npiv_logo;
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_LOGO_SND;
spin_unlock_irq(shost->host_lock);
if (lpfc_sli_issue_iocb(phba, pring, elsiocb, 0) == IOCB_ERROR) {
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag &= ~NLP_LOGO_SND;
spin_unlock_irq(shost->host_lock);
lpfc_els_free_iocb(phba, elsiocb);
return 1;
}
return 0;
}
/**
* lpfc_fabric_block_timeout: Handler function to the fabric block timer.
* @ptr: holder for the timer function associated data.
*
* This routine is invoked by the fabric iocb block timer after
* timeout. It posts the fabric iocb block timeout event by setting the
* WORKER_FABRIC_BLOCK_TMO bit to work port event bitmap and then invokes
* lpfc_worker_wake_up() routine to wake up the worker thread. It is for
* the worker thread to invoke the lpfc_unblock_fabric_iocbs() on the
* posted event WORKER_FABRIC_BLOCK_TMO.
**/
void
lpfc_fabric_block_timeout(unsigned long ptr)
{
struct lpfc_hba *phba = (struct lpfc_hba *) ptr;
unsigned long iflags;
uint32_t tmo_posted;
spin_lock_irqsave(&phba->pport->work_port_lock, iflags);
tmo_posted = phba->pport->work_port_events & WORKER_FABRIC_BLOCK_TMO;
if (!tmo_posted)
phba->pport->work_port_events |= WORKER_FABRIC_BLOCK_TMO;
spin_unlock_irqrestore(&phba->pport->work_port_lock, iflags);
if (!tmo_posted)
lpfc_worker_wake_up(phba);
return;
}
/**
* lpfc_resume_fabric_iocbs: Issue a fabric iocb from driver internal list.
* @phba: pointer to lpfc hba data structure.
*
* This routine issues one fabric iocb from the driver internal list to
* the HBA. It first checks whether it's ready to issue one fabric iocb to
* the HBA (whether there is no outstanding fabric iocb). If so, it shall
* remove one pending fabric iocb from the driver internal list and invokes
* lpfc_sli_issue_iocb() routine to send the fabric iocb to the HBA.
**/
static void
lpfc_resume_fabric_iocbs(struct lpfc_hba *phba)
{
struct lpfc_iocbq *iocb;
unsigned long iflags;
int ret;
struct lpfc_sli_ring *pring = &phba->sli.ring[LPFC_ELS_RING];
IOCB_t *cmd;
repeat:
iocb = NULL;
spin_lock_irqsave(&phba->hbalock, iflags);
/* Post any pending iocb to the SLI layer */
if (atomic_read(&phba->fabric_iocb_count) == 0) {
list_remove_head(&phba->fabric_iocb_list, iocb, typeof(*iocb),
list);
if (iocb)
/* Increment fabric iocb count to hold the position */
atomic_inc(&phba->fabric_iocb_count);
}
spin_unlock_irqrestore(&phba->hbalock, iflags);
if (iocb) {
iocb->fabric_iocb_cmpl = iocb->iocb_cmpl;
iocb->iocb_cmpl = lpfc_cmpl_fabric_iocb;
iocb->iocb_flag |= LPFC_IO_FABRIC;
lpfc_debugfs_disc_trc(iocb->vport, LPFC_DISC_TRC_ELS_CMD,
"Fabric sched1: ste:x%x",
iocb->vport->port_state, 0, 0);
ret = lpfc_sli_issue_iocb(phba, pring, iocb, 0);
if (ret == IOCB_ERROR) {
iocb->iocb_cmpl = iocb->fabric_iocb_cmpl;
iocb->fabric_iocb_cmpl = NULL;
iocb->iocb_flag &= ~LPFC_IO_FABRIC;
cmd = &iocb->iocb;
cmd->ulpStatus = IOSTAT_LOCAL_REJECT;
cmd->un.ulpWord[4] = IOERR_SLI_ABORTED;
iocb->iocb_cmpl(phba, iocb, iocb);
atomic_dec(&phba->fabric_iocb_count);
goto repeat;
}
}
return;
}
/**
* lpfc_unblock_fabric_iocbs: Unblock issuing fabric iocb command.
* @phba: pointer to lpfc hba data structure.
*
* This routine unblocks the issuing fabric iocb command. The function
* will clear the fabric iocb block bit and then invoke the routine
* lpfc_resume_fabric_iocbs() to issue one of the pending fabric iocb
* from the driver internal fabric iocb list.
**/
void
lpfc_unblock_fabric_iocbs(struct lpfc_hba *phba)
{
clear_bit(FABRIC_COMANDS_BLOCKED, &phba->bit_flags);
lpfc_resume_fabric_iocbs(phba);
return;
}
/**
* lpfc_block_fabric_iocbs: Block issuing fabric iocb command.
* @phba: pointer to lpfc hba data structure.
*
* This routine blocks the issuing fabric iocb for a specified amount of
* time (currently 100 ms). This is done by set the fabric iocb block bit
* and set up a timeout timer for 100ms. When the block bit is set, no more
* fabric iocb will be issued out of the HBA.
**/
static void
lpfc_block_fabric_iocbs(struct lpfc_hba *phba)
{
int blocked;
blocked = test_and_set_bit(FABRIC_COMANDS_BLOCKED, &phba->bit_flags);
/* Start a timer to unblock fabric iocbs after 100ms */
if (!blocked)
mod_timer(&phba->fabric_block_timer, jiffies + HZ/10 );
return;
}
/**
* lpfc_cmpl_fabric_iocb: Completion callback function for fabric iocb.
* @phba: pointer to lpfc hba data structure.
* @cmdiocb: pointer to lpfc command iocb data structure.
* @rspiocb: pointer to lpfc response iocb data structure.
*
* This routine is the callback function that is put to the fabric iocb's
* callback function pointer (iocb->iocb_cmpl). The original iocb's callback
* function pointer has been stored in iocb->fabric_iocb_cmpl. This callback
* function first restores and invokes the original iocb's callback function
* and then invokes the lpfc_resume_fabric_iocbs() routine to issue the next
* fabric bound iocb from the driver internal fabric iocb list onto the wire.
**/
static void
lpfc_cmpl_fabric_iocb(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
struct lpfc_iocbq *rspiocb)
{
struct ls_rjt stat;
if ((cmdiocb->iocb_flag & LPFC_IO_FABRIC) != LPFC_IO_FABRIC)
BUG();
switch (rspiocb->iocb.ulpStatus) {
case IOSTAT_NPORT_RJT:
case IOSTAT_FABRIC_RJT:
if (rspiocb->iocb.un.ulpWord[4] & RJT_UNAVAIL_TEMP) {
lpfc_block_fabric_iocbs(phba);
}
break;
case IOSTAT_NPORT_BSY:
case IOSTAT_FABRIC_BSY:
lpfc_block_fabric_iocbs(phba);
break;
case IOSTAT_LS_RJT:
stat.un.lsRjtError =
be32_to_cpu(rspiocb->iocb.un.ulpWord[4]);
if ((stat.un.b.lsRjtRsnCode == LSRJT_UNABLE_TPC) ||
(stat.un.b.lsRjtRsnCode == LSRJT_LOGICAL_BSY))
lpfc_block_fabric_iocbs(phba);
break;
}
if (atomic_read(&phba->fabric_iocb_count) == 0)
BUG();
cmdiocb->iocb_cmpl = cmdiocb->fabric_iocb_cmpl;
cmdiocb->fabric_iocb_cmpl = NULL;
cmdiocb->iocb_flag &= ~LPFC_IO_FABRIC;
cmdiocb->iocb_cmpl(phba, cmdiocb, rspiocb);
atomic_dec(&phba->fabric_iocb_count);
if (!test_bit(FABRIC_COMANDS_BLOCKED, &phba->bit_flags)) {
/* Post any pending iocbs to HBA */
lpfc_resume_fabric_iocbs(phba);
}
}
/**
* lpfc_issue_fabric_iocb: Issue a fabric iocb command.
* @phba: pointer to lpfc hba data structure.
* @iocb: pointer to lpfc command iocb data structure.
*
* This routine is used as the top-level API for issuing a fabric iocb command
* such as FLOGI and FDISC. To accommodate certain switch fabric, this driver
* function makes sure that only one fabric bound iocb will be outstanding at
* any given time. As such, this function will first check to see whether there
* is already an outstanding fabric iocb on the wire. If so, it will put the
* newly issued iocb onto the driver internal fabric iocb list, waiting to be
* issued later. Otherwise, it will issue the iocb on the wire and update the
* fabric iocb count it indicate that there is one fabric iocb on the wire.
*
* Note, this implementation has a potential sending out fabric IOCBs out of
* order. The problem is caused by the construction of the "ready" boolen does
* not include the condition that the internal fabric IOCB list is empty. As
* such, it is possible a fabric IOCB issued by this routine might be "jump"
* ahead of the fabric IOCBs in the internal list.
*
* Return code
* IOCB_SUCCESS - either fabric iocb put on the list or issued successfully
* IOCB_ERROR - failed to issue fabric iocb
**/
static int
lpfc_issue_fabric_iocb(struct lpfc_hba *phba, struct lpfc_iocbq *iocb)
{
unsigned long iflags;
struct lpfc_sli_ring *pring = &phba->sli.ring[LPFC_ELS_RING];
int ready;
int ret;
if (atomic_read(&phba->fabric_iocb_count) > 1)
BUG();
spin_lock_irqsave(&phba->hbalock, iflags);
ready = atomic_read(&phba->fabric_iocb_count) == 0 &&
!test_bit(FABRIC_COMANDS_BLOCKED, &phba->bit_flags);
if (ready)
/* Increment fabric iocb count to hold the position */
atomic_inc(&phba->fabric_iocb_count);
spin_unlock_irqrestore(&phba->hbalock, iflags);
if (ready) {
iocb->fabric_iocb_cmpl = iocb->iocb_cmpl;
iocb->iocb_cmpl = lpfc_cmpl_fabric_iocb;
iocb->iocb_flag |= LPFC_IO_FABRIC;
lpfc_debugfs_disc_trc(iocb->vport, LPFC_DISC_TRC_ELS_CMD,
"Fabric sched2: ste:x%x",
iocb->vport->port_state, 0, 0);
ret = lpfc_sli_issue_iocb(phba, pring, iocb, 0);
if (ret == IOCB_ERROR) {
iocb->iocb_cmpl = iocb->fabric_iocb_cmpl;
iocb->fabric_iocb_cmpl = NULL;
iocb->iocb_flag &= ~LPFC_IO_FABRIC;
atomic_dec(&phba->fabric_iocb_count);
}
} else {
spin_lock_irqsave(&phba->hbalock, iflags);
list_add_tail(&iocb->list, &phba->fabric_iocb_list);
spin_unlock_irqrestore(&phba->hbalock, iflags);
ret = IOCB_SUCCESS;
}
return ret;
}
/**
* lpfc_fabric_abort_vport: Abort a vport's iocbs from driver fabric iocb list.
* @vport: pointer to a virtual N_Port data structure.
*
* This routine aborts all the IOCBs associated with a @vport from the
* driver internal fabric IOCB list. The list contains fabric IOCBs to be
* issued to the ELS IOCB ring. This abort function walks the fabric IOCB
* list, removes each IOCB associated with the @vport off the list, set the
* status feild to IOSTAT_LOCAL_REJECT, and invokes the callback function
* associated with the IOCB.
**/
static void lpfc_fabric_abort_vport(struct lpfc_vport *vport)
{
LIST_HEAD(completions);
struct lpfc_hba *phba = vport->phba;
struct lpfc_iocbq *tmp_iocb, *piocb;
IOCB_t *cmd;
spin_lock_irq(&phba->hbalock);
list_for_each_entry_safe(piocb, tmp_iocb, &phba->fabric_iocb_list,
list) {
if (piocb->vport != vport)
continue;
list_move_tail(&piocb->list, &completions);
}
spin_unlock_irq(&phba->hbalock);
while (!list_empty(&completions)) {
piocb = list_get_first(&completions, struct lpfc_iocbq, list);
list_del_init(&piocb->list);
cmd = &piocb->iocb;
cmd->ulpStatus = IOSTAT_LOCAL_REJECT;
cmd->un.ulpWord[4] = IOERR_SLI_ABORTED;
(piocb->iocb_cmpl) (phba, piocb, piocb);
}
}
/**
* lpfc_fabric_abort_nport: Abort a ndlp's iocbs from driver fabric iocb list.
* @ndlp: pointer to a node-list data structure.
*
* This routine aborts all the IOCBs associated with an @ndlp from the
* driver internal fabric IOCB list. The list contains fabric IOCBs to be
* issued to the ELS IOCB ring. This abort function walks the fabric IOCB
* list, removes each IOCB associated with the @ndlp off the list, set the
* status feild to IOSTAT_LOCAL_REJECT, and invokes the callback function
* associated with the IOCB.
**/
void lpfc_fabric_abort_nport(struct lpfc_nodelist *ndlp)
{
LIST_HEAD(completions);
struct lpfc_hba *phba = ndlp->vport->phba;
struct lpfc_iocbq *tmp_iocb, *piocb;
struct lpfc_sli_ring *pring = &phba->sli.ring[LPFC_ELS_RING];
IOCB_t *cmd;
spin_lock_irq(&phba->hbalock);
list_for_each_entry_safe(piocb, tmp_iocb, &phba->fabric_iocb_list,
list) {
if ((lpfc_check_sli_ndlp(phba, pring, piocb, ndlp))) {
list_move_tail(&piocb->list, &completions);
}
}
spin_unlock_irq(&phba->hbalock);
while (!list_empty(&completions)) {
piocb = list_get_first(&completions, struct lpfc_iocbq, list);
list_del_init(&piocb->list);
cmd = &piocb->iocb;
cmd->ulpStatus = IOSTAT_LOCAL_REJECT;
cmd->un.ulpWord[4] = IOERR_SLI_ABORTED;
(piocb->iocb_cmpl) (phba, piocb, piocb);
}
}
/**
* lpfc_fabric_abort_hba: Abort all iocbs on driver fabric iocb list.
* @phba: pointer to lpfc hba data structure.
*
* This routine aborts all the IOCBs currently on the driver internal
* fabric IOCB list. The list contains fabric IOCBs to be issued to the ELS
* IOCB ring. This function takes the entire IOCB list off the fabric IOCB
* list, removes IOCBs off the list, set the status feild to
* IOSTAT_LOCAL_REJECT, and invokes the callback function associated with
* the IOCB.
**/
void lpfc_fabric_abort_hba(struct lpfc_hba *phba)
{
LIST_HEAD(completions);
struct lpfc_iocbq *piocb;
IOCB_t *cmd;
spin_lock_irq(&phba->hbalock);
list_splice_init(&phba->fabric_iocb_list, &completions);
spin_unlock_irq(&phba->hbalock);
while (!list_empty(&completions)) {
piocb = list_get_first(&completions, struct lpfc_iocbq, list);
list_del_init(&piocb->list);
cmd = &piocb->iocb;
cmd->ulpStatus = IOSTAT_LOCAL_REJECT;
cmd->un.ulpWord[4] = IOERR_SLI_ABORTED;
(piocb->iocb_cmpl) (phba, piocb, piocb);
}
}