alistair23-linux/drivers/scsi/bfa/rport_ftrs.c
Jing Huang c507341713 [SCSI] bfa: fix rport speed setting
When a rport goes offline, its speed setting was not reset. Subsequently, if
the rport was not deleted due to it coming back online within rport del
timeout, previously discovered speed would continue to show up. The fix is to
reset the speed when processing rport offline transition.

In rport attributes, rport's with unknown speed were indicated as TRL
enforced.  The right thing do to would be to use TRL default speed to
determine if TRL is enforced, when TRL is enabled.

Signed-off-by: Jing Huang <huangj@brocade.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
2010-07-27 12:04:10 -05:00

380 lines
9.9 KiB
C

/*
* Copyright (c) 2005-2009 Brocade Communications Systems, Inc.
* All rights reserved
* www.brocade.com
*
* Linux driver for Brocade Fibre Channel Host Bus Adapter.
*
* This program is free software; you can redistribute it and/or modify it
* under the terms of the GNU General Public License (GPL) Version 2 as
* published by the Free Software Foundation
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*/
/**
* rport_ftrs.c Remote port features (RPF) implementation.
*/
#include <bfa.h>
#include <bfa_svc.h>
#include "fcbuild.h"
#include "fcs_rport.h"
#include "fcs_lport.h"
#include "fcs_trcmod.h"
#include "fcs_fcxp.h"
#include "fcs.h"
BFA_TRC_FILE(FCS, RPORT_FTRS);
#define BFA_FCS_RPF_RETRIES (3)
#define BFA_FCS_RPF_RETRY_TIMEOUT (1000) /* 1 sec (In millisecs) */
static void bfa_fcs_rpf_send_rpsc2(void *rport_cbarg,
struct bfa_fcxp_s *fcxp_alloced);
static void bfa_fcs_rpf_rpsc2_response(void *fcsarg,
struct bfa_fcxp_s *fcxp, void *cbarg,
bfa_status_t req_status, u32 rsp_len,
u32 resid_len,
struct fchs_s *rsp_fchs);
static void bfa_fcs_rpf_timeout(void *arg);
/**
* fcs_rport_ftrs_sm FCS rport state machine events
*/
enum rpf_event {
RPFSM_EVENT_RPORT_OFFLINE = 1, /* Rport offline */
RPFSM_EVENT_RPORT_ONLINE = 2, /* Rport online */
RPFSM_EVENT_FCXP_SENT = 3, /* Frame from has been sent */
RPFSM_EVENT_TIMEOUT = 4, /* Rport SM timeout event */
RPFSM_EVENT_RPSC_COMP = 5,
RPFSM_EVENT_RPSC_FAIL = 6,
RPFSM_EVENT_RPSC_ERROR = 7,
};
static void bfa_fcs_rpf_sm_uninit(struct bfa_fcs_rpf_s *rpf,
enum rpf_event event);
static void bfa_fcs_rpf_sm_rpsc_sending(struct bfa_fcs_rpf_s *rpf,
enum rpf_event event);
static void bfa_fcs_rpf_sm_rpsc(struct bfa_fcs_rpf_s *rpf,
enum rpf_event event);
static void bfa_fcs_rpf_sm_rpsc_retry(struct bfa_fcs_rpf_s *rpf,
enum rpf_event event);
static void bfa_fcs_rpf_sm_offline(struct bfa_fcs_rpf_s *rpf,
enum rpf_event event);
static void bfa_fcs_rpf_sm_online(struct bfa_fcs_rpf_s *rpf,
enum rpf_event event);
static void
bfa_fcs_rpf_sm_uninit(struct bfa_fcs_rpf_s *rpf, enum rpf_event event)
{
struct bfa_fcs_rport_s *rport = rpf->rport;
struct bfa_fcs_fabric_s *fabric = &rport->fcs->fabric;
bfa_trc(rport->fcs, rport->pwwn);
bfa_trc(rport->fcs, rport->pid);
bfa_trc(rport->fcs, event);
switch (event) {
case RPFSM_EVENT_RPORT_ONLINE:
/* Send RPSC2 to a Brocade fabric only. */
if ((!BFA_FCS_PID_IS_WKA(rport->pid)) &&
((bfa_lps_is_brcd_fabric(rport->port->fabric->lps)) ||
(bfa_fcs_fabric_get_switch_oui(fabric) ==
BFA_FCS_BRCD_SWITCH_OUI))) {
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_rpsc_sending);
rpf->rpsc_retries = 0;
bfa_fcs_rpf_send_rpsc2(rpf, NULL);
}
break;
case RPFSM_EVENT_RPORT_OFFLINE:
break;
default:
bfa_sm_fault(rport->fcs, event);
}
}
static void
bfa_fcs_rpf_sm_rpsc_sending(struct bfa_fcs_rpf_s *rpf, enum rpf_event event)
{
struct bfa_fcs_rport_s *rport = rpf->rport;
bfa_trc(rport->fcs, event);
switch (event) {
case RPFSM_EVENT_FCXP_SENT:
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_rpsc);
break;
case RPFSM_EVENT_RPORT_OFFLINE:
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_offline);
bfa_fcxp_walloc_cancel(rport->fcs->bfa, &rpf->fcxp_wqe);
rpf->rpsc_retries = 0;
break;
default:
bfa_sm_fault(rport->fcs, event);
}
}
static void
bfa_fcs_rpf_sm_rpsc(struct bfa_fcs_rpf_s *rpf, enum rpf_event event)
{
struct bfa_fcs_rport_s *rport = rpf->rport;
bfa_trc(rport->fcs, rport->pid);
bfa_trc(rport->fcs, event);
switch (event) {
case RPFSM_EVENT_RPSC_COMP:
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_online);
/* Update speed info in f/w via BFA */
if (rpf->rpsc_speed != BFA_PPORT_SPEED_UNKNOWN)
bfa_rport_speed(rport->bfa_rport, rpf->rpsc_speed);
else if (rpf->assigned_speed != BFA_PPORT_SPEED_UNKNOWN)
bfa_rport_speed(rport->bfa_rport, rpf->assigned_speed);
break;
case RPFSM_EVENT_RPSC_FAIL:
/* RPSC not supported by rport */
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_online);
break;
case RPFSM_EVENT_RPSC_ERROR:
/* need to retry...delayed a bit. */
if (rpf->rpsc_retries++ < BFA_FCS_RPF_RETRIES) {
bfa_timer_start(rport->fcs->bfa, &rpf->timer,
bfa_fcs_rpf_timeout, rpf,
BFA_FCS_RPF_RETRY_TIMEOUT);
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_rpsc_retry);
} else {
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_online);
}
break;
case RPFSM_EVENT_RPORT_OFFLINE:
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_offline);
bfa_fcxp_discard(rpf->fcxp);
rpf->rpsc_retries = 0;
break;
default:
bfa_sm_fault(rport->fcs, event);
}
}
static void
bfa_fcs_rpf_sm_rpsc_retry(struct bfa_fcs_rpf_s *rpf, enum rpf_event event)
{
struct bfa_fcs_rport_s *rport = rpf->rport;
bfa_trc(rport->fcs, rport->pid);
bfa_trc(rport->fcs, event);
switch (event) {
case RPFSM_EVENT_TIMEOUT:
/* re-send the RPSC */
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_rpsc_sending);
bfa_fcs_rpf_send_rpsc2(rpf, NULL);
break;
case RPFSM_EVENT_RPORT_OFFLINE:
bfa_timer_stop(&rpf->timer);
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_offline);
rpf->rpsc_retries = 0;
break;
default:
bfa_sm_fault(rport->fcs, event);
}
}
static void
bfa_fcs_rpf_sm_online(struct bfa_fcs_rpf_s *rpf, enum rpf_event event)
{
struct bfa_fcs_rport_s *rport = rpf->rport;
bfa_trc(rport->fcs, rport->pwwn);
bfa_trc(rport->fcs, rport->pid);
bfa_trc(rport->fcs, event);
switch (event) {
case RPFSM_EVENT_RPORT_OFFLINE:
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_offline);
rpf->rpsc_retries = 0;
break;
default:
bfa_sm_fault(rport->fcs, event);
}
}
static void
bfa_fcs_rpf_sm_offline(struct bfa_fcs_rpf_s *rpf, enum rpf_event event)
{
struct bfa_fcs_rport_s *rport = rpf->rport;
bfa_trc(rport->fcs, rport->pwwn);
bfa_trc(rport->fcs, rport->pid);
bfa_trc(rport->fcs, event);
switch (event) {
case RPFSM_EVENT_RPORT_ONLINE:
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_rpsc_sending);
bfa_fcs_rpf_send_rpsc2(rpf, NULL);
break;
case RPFSM_EVENT_RPORT_OFFLINE:
break;
default:
bfa_sm_fault(rport->fcs, event);
}
}
/**
* Called when Rport is created.
*/
void bfa_fcs_rpf_init(struct bfa_fcs_rport_s *rport)
{
struct bfa_fcs_rpf_s *rpf = &rport->rpf;
bfa_trc(rport->fcs, rport->pid);
rpf->rport = rport;
bfa_sm_set_state(rpf, bfa_fcs_rpf_sm_uninit);
}
/**
* Called when Rport becomes online
*/
void bfa_fcs_rpf_rport_online(struct bfa_fcs_rport_s *rport)
{
bfa_trc(rport->fcs, rport->pid);
if (__fcs_min_cfg(rport->port->fcs))
return;
if (bfa_fcs_fabric_is_switched(rport->port->fabric))
bfa_sm_send_event(&rport->rpf, RPFSM_EVENT_RPORT_ONLINE);
}
/**
* Called when Rport becomes offline
*/
void bfa_fcs_rpf_rport_offline(struct bfa_fcs_rport_s *rport)
{
bfa_trc(rport->fcs, rport->pid);
if (__fcs_min_cfg(rport->port->fcs))
return;
rport->rpf.rpsc_speed = 0;
bfa_sm_send_event(&rport->rpf, RPFSM_EVENT_RPORT_OFFLINE);
}
static void
bfa_fcs_rpf_timeout(void *arg)
{
struct bfa_fcs_rpf_s *rpf = (struct bfa_fcs_rpf_s *) arg;
struct bfa_fcs_rport_s *rport = rpf->rport;
bfa_trc(rport->fcs, rport->pid);
bfa_sm_send_event(rpf, RPFSM_EVENT_TIMEOUT);
}
static void
bfa_fcs_rpf_send_rpsc2(void *rpf_cbarg, struct bfa_fcxp_s *fcxp_alloced)
{
struct bfa_fcs_rpf_s *rpf = (struct bfa_fcs_rpf_s *)rpf_cbarg;
struct bfa_fcs_rport_s *rport = rpf->rport;
struct bfa_fcs_port_s *port = rport->port;
struct fchs_s fchs;
int len;
struct bfa_fcxp_s *fcxp;
bfa_trc(rport->fcs, rport->pwwn);
fcxp = fcxp_alloced ? fcxp_alloced : bfa_fcs_fcxp_alloc(port->fcs);
if (!fcxp) {
bfa_fcxp_alloc_wait(port->fcs->bfa, &rpf->fcxp_wqe,
bfa_fcs_rpf_send_rpsc2, rpf);
return;
}
rpf->fcxp = fcxp;
len = fc_rpsc2_build(&fchs, bfa_fcxp_get_reqbuf(fcxp), rport->pid,
bfa_fcs_port_get_fcid(port), &rport->pid, 1);
bfa_fcxp_send(fcxp, NULL, port->fabric->vf_id, port->lp_tag, BFA_FALSE,
FC_CLASS_3, len, &fchs, bfa_fcs_rpf_rpsc2_response,
rpf, FC_MAX_PDUSZ, FC_ELS_TOV);
rport->stats.rpsc_sent++;
bfa_sm_send_event(rpf, RPFSM_EVENT_FCXP_SENT);
}
static void
bfa_fcs_rpf_rpsc2_response(void *fcsarg, struct bfa_fcxp_s *fcxp, void *cbarg,
bfa_status_t req_status, u32 rsp_len,
u32 resid_len, struct fchs_s *rsp_fchs)
{
struct bfa_fcs_rpf_s *rpf = (struct bfa_fcs_rpf_s *) cbarg;
struct bfa_fcs_rport_s *rport = rpf->rport;
struct fc_ls_rjt_s *ls_rjt;
struct fc_rpsc2_acc_s *rpsc2_acc;
u16 num_ents;
bfa_trc(rport->fcs, req_status);
if (req_status != BFA_STATUS_OK) {
bfa_trc(rport->fcs, req_status);
if (req_status == BFA_STATUS_ETIMER)
rport->stats.rpsc_failed++;
bfa_sm_send_event(rpf, RPFSM_EVENT_RPSC_ERROR);
return;
}
rpsc2_acc = (struct fc_rpsc2_acc_s *) BFA_FCXP_RSP_PLD(fcxp);
if (rpsc2_acc->els_cmd == FC_ELS_ACC) {
rport->stats.rpsc_accs++;
num_ents = bfa_os_ntohs(rpsc2_acc->num_pids);
bfa_trc(rport->fcs, num_ents);
if (num_ents > 0) {
bfa_assert(rpsc2_acc->port_info[0].pid != rport->pid);
bfa_trc(rport->fcs,
bfa_os_ntohs(rpsc2_acc->port_info[0].pid));
bfa_trc(rport->fcs,
bfa_os_ntohs(rpsc2_acc->port_info[0].speed));
bfa_trc(rport->fcs,
bfa_os_ntohs(rpsc2_acc->port_info[0].index));
bfa_trc(rport->fcs,
rpsc2_acc->port_info[0].type);
if (rpsc2_acc->port_info[0].speed == 0) {
bfa_sm_send_event(rpf, RPFSM_EVENT_RPSC_ERROR);
return;
}
rpf->rpsc_speed = fc_rpsc_operspeed_to_bfa_speed(
bfa_os_ntohs(rpsc2_acc->port_info[0].speed));
bfa_sm_send_event(rpf, RPFSM_EVENT_RPSC_COMP);
}
} else {
ls_rjt = (struct fc_ls_rjt_s *) BFA_FCXP_RSP_PLD(fcxp);
bfa_trc(rport->fcs, ls_rjt->reason_code);
bfa_trc(rport->fcs, ls_rjt->reason_code_expl);
rport->stats.rpsc_rejects++;
if (ls_rjt->reason_code == FC_LS_RJT_RSN_CMD_NOT_SUPP)
bfa_sm_send_event(rpf, RPFSM_EVENT_RPSC_FAIL);
else
bfa_sm_send_event(rpf, RPFSM_EVENT_RPSC_ERROR);
}
}