linux_dsm_epyc7002/drivers/scsi/qla2xxx/qla_isr.c
Sawan Chandak 718abbdca7 qla2xxx: Restore physical port WWPN only, when port down detected for FA-WWPN port.
For FA-WWPN is enabled port, if NPIV created on that port and,
if port link is brought down, then WWPN was restored from flash for both
physical and NPIV port. This will result in NPIV port and physical port
sharing same WWPN. Any application refreshing ports information  will
not be able to scan NPIV port because of this behavior. So while restoring WWPN,
only restore physical port WWPN.

Signed-off-by: Sawan Chandak <sawan.chandak@qlogic.com>
Signed-off-by: Himanshu Madhani <himanshu.madhani@qlogic.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: James Bottomley <JBottomley@Odin.com>
2015-04-10 08:43:07 -07:00

3211 lines
86 KiB
C

/*
* QLogic Fibre Channel HBA Driver
* Copyright (c) 2003-2014 QLogic Corporation
*
* See LICENSE.qla2xxx for copyright and licensing details.
*/
#include "qla_def.h"
#include "qla_target.h"
#include <linux/delay.h>
#include <linux/slab.h>
#include <scsi/scsi_tcq.h>
#include <scsi/scsi_bsg_fc.h>
#include <scsi/scsi_eh.h>
static void qla2x00_mbx_completion(scsi_qla_host_t *, uint16_t);
static void qla2x00_status_entry(scsi_qla_host_t *, struct rsp_que *, void *);
static void qla2x00_status_cont_entry(struct rsp_que *, sts_cont_entry_t *);
static void qla2x00_error_entry(scsi_qla_host_t *, struct rsp_que *,
sts_entry_t *);
/**
* qla2100_intr_handler() - Process interrupts for the ISP2100 and ISP2200.
* @irq:
* @dev_id: SCSI driver HA context
*
* Called by system whenever the host adapter generates an interrupt.
*
* Returns handled flag.
*/
irqreturn_t
qla2100_intr_handler(int irq, void *dev_id)
{
scsi_qla_host_t *vha;
struct qla_hw_data *ha;
struct device_reg_2xxx __iomem *reg;
int status;
unsigned long iter;
uint16_t hccr;
uint16_t mb[4];
struct rsp_que *rsp;
unsigned long flags;
rsp = (struct rsp_que *) dev_id;
if (!rsp) {
ql_log(ql_log_info, NULL, 0x505d,
"%s: NULL response queue pointer.\n", __func__);
return (IRQ_NONE);
}
ha = rsp->hw;
reg = &ha->iobase->isp;
status = 0;
spin_lock_irqsave(&ha->hardware_lock, flags);
vha = pci_get_drvdata(ha->pdev);
for (iter = 50; iter--; ) {
hccr = RD_REG_WORD(&reg->hccr);
if (qla2x00_check_reg16_for_disconnect(vha, hccr))
break;
if (hccr & HCCR_RISC_PAUSE) {
if (pci_channel_offline(ha->pdev))
break;
/*
* Issue a "HARD" reset in order for the RISC interrupt
* bit to be cleared. Schedule a big hammer to get
* out of the RISC PAUSED state.
*/
WRT_REG_WORD(&reg->hccr, HCCR_RESET_RISC);
RD_REG_WORD(&reg->hccr);
ha->isp_ops->fw_dump(vha, 1);
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
break;
} else if ((RD_REG_WORD(&reg->istatus) & ISR_RISC_INT) == 0)
break;
if (RD_REG_WORD(&reg->semaphore) & BIT_0) {
WRT_REG_WORD(&reg->hccr, HCCR_CLR_RISC_INT);
RD_REG_WORD(&reg->hccr);
/* Get mailbox data. */
mb[0] = RD_MAILBOX_REG(ha, reg, 0);
if (mb[0] > 0x3fff && mb[0] < 0x8000) {
qla2x00_mbx_completion(vha, mb[0]);
status |= MBX_INTERRUPT;
} else if (mb[0] > 0x7fff && mb[0] < 0xc000) {
mb[1] = RD_MAILBOX_REG(ha, reg, 1);
mb[2] = RD_MAILBOX_REG(ha, reg, 2);
mb[3] = RD_MAILBOX_REG(ha, reg, 3);
qla2x00_async_event(vha, rsp, mb);
} else {
/*EMPTY*/
ql_dbg(ql_dbg_async, vha, 0x5025,
"Unrecognized interrupt type (%d).\n",
mb[0]);
}
/* Release mailbox registers. */
WRT_REG_WORD(&reg->semaphore, 0);
RD_REG_WORD(&reg->semaphore);
} else {
qla2x00_process_response_queue(rsp);
WRT_REG_WORD(&reg->hccr, HCCR_CLR_RISC_INT);
RD_REG_WORD(&reg->hccr);
}
}
qla2x00_handle_mbx_completion(ha, status);
spin_unlock_irqrestore(&ha->hardware_lock, flags);
return (IRQ_HANDLED);
}
bool
qla2x00_check_reg32_for_disconnect(scsi_qla_host_t *vha, uint32_t reg)
{
/* Check for PCI disconnection */
if (reg == 0xffffffff) {
if (!test_and_set_bit(PFLG_DISCONNECTED, &vha->pci_flags) &&
!test_bit(PFLG_DRIVER_REMOVING, &vha->pci_flags) &&
!test_bit(PFLG_DRIVER_PROBING, &vha->pci_flags)) {
/*
* Schedule this (only once) on the default system
* workqueue so that all the adapter workqueues and the
* DPC thread can be shutdown cleanly.
*/
schedule_work(&vha->hw->board_disable);
}
return true;
} else
return false;
}
bool
qla2x00_check_reg16_for_disconnect(scsi_qla_host_t *vha, uint16_t reg)
{
return qla2x00_check_reg32_for_disconnect(vha, 0xffff0000 | reg);
}
/**
* qla2300_intr_handler() - Process interrupts for the ISP23xx and ISP63xx.
* @irq:
* @dev_id: SCSI driver HA context
*
* Called by system whenever the host adapter generates an interrupt.
*
* Returns handled flag.
*/
irqreturn_t
qla2300_intr_handler(int irq, void *dev_id)
{
scsi_qla_host_t *vha;
struct device_reg_2xxx __iomem *reg;
int status;
unsigned long iter;
uint32_t stat;
uint16_t hccr;
uint16_t mb[4];
struct rsp_que *rsp;
struct qla_hw_data *ha;
unsigned long flags;
rsp = (struct rsp_que *) dev_id;
if (!rsp) {
ql_log(ql_log_info, NULL, 0x5058,
"%s: NULL response queue pointer.\n", __func__);
return (IRQ_NONE);
}
ha = rsp->hw;
reg = &ha->iobase->isp;
status = 0;
spin_lock_irqsave(&ha->hardware_lock, flags);
vha = pci_get_drvdata(ha->pdev);
for (iter = 50; iter--; ) {
stat = RD_REG_DWORD(&reg->u.isp2300.host_status);
if (qla2x00_check_reg32_for_disconnect(vha, stat))
break;
if (stat & HSR_RISC_PAUSED) {
if (unlikely(pci_channel_offline(ha->pdev)))
break;
hccr = RD_REG_WORD(&reg->hccr);
if (hccr & (BIT_15 | BIT_13 | BIT_11 | BIT_8))
ql_log(ql_log_warn, vha, 0x5026,
"Parity error -- HCCR=%x, Dumping "
"firmware.\n", hccr);
else
ql_log(ql_log_warn, vha, 0x5027,
"RISC paused -- HCCR=%x, Dumping "
"firmware.\n", hccr);
/*
* Issue a "HARD" reset in order for the RISC
* interrupt bit to be cleared. Schedule a big
* hammer to get out of the RISC PAUSED state.
*/
WRT_REG_WORD(&reg->hccr, HCCR_RESET_RISC);
RD_REG_WORD(&reg->hccr);
ha->isp_ops->fw_dump(vha, 1);
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
break;
} else if ((stat & HSR_RISC_INT) == 0)
break;
switch (stat & 0xff) {
case 0x1:
case 0x2:
case 0x10:
case 0x11:
qla2x00_mbx_completion(vha, MSW(stat));
status |= MBX_INTERRUPT;
/* Release mailbox registers. */
WRT_REG_WORD(&reg->semaphore, 0);
break;
case 0x12:
mb[0] = MSW(stat);
mb[1] = RD_MAILBOX_REG(ha, reg, 1);
mb[2] = RD_MAILBOX_REG(ha, reg, 2);
mb[3] = RD_MAILBOX_REG(ha, reg, 3);
qla2x00_async_event(vha, rsp, mb);
break;
case 0x13:
qla2x00_process_response_queue(rsp);
break;
case 0x15:
mb[0] = MBA_CMPLT_1_16BIT;
mb[1] = MSW(stat);
qla2x00_async_event(vha, rsp, mb);
break;
case 0x16:
mb[0] = MBA_SCSI_COMPLETION;
mb[1] = MSW(stat);
mb[2] = RD_MAILBOX_REG(ha, reg, 2);
qla2x00_async_event(vha, rsp, mb);
break;
default:
ql_dbg(ql_dbg_async, vha, 0x5028,
"Unrecognized interrupt type (%d).\n", stat & 0xff);
break;
}
WRT_REG_WORD(&reg->hccr, HCCR_CLR_RISC_INT);
RD_REG_WORD_RELAXED(&reg->hccr);
}
qla2x00_handle_mbx_completion(ha, status);
spin_unlock_irqrestore(&ha->hardware_lock, flags);
return (IRQ_HANDLED);
}
/**
* qla2x00_mbx_completion() - Process mailbox command completions.
* @ha: SCSI driver HA context
* @mb0: Mailbox0 register
*/
static void
qla2x00_mbx_completion(scsi_qla_host_t *vha, uint16_t mb0)
{
uint16_t cnt;
uint32_t mboxes;
uint16_t __iomem *wptr;
struct qla_hw_data *ha = vha->hw;
struct device_reg_2xxx __iomem *reg = &ha->iobase->isp;
/* Read all mbox registers? */
mboxes = (1 << ha->mbx_count) - 1;
if (!ha->mcp)
ql_dbg(ql_dbg_async, vha, 0x5001, "MBX pointer ERROR.\n");
else
mboxes = ha->mcp->in_mb;
/* Load return mailbox registers. */
ha->flags.mbox_int = 1;
ha->mailbox_out[0] = mb0;
mboxes >>= 1;
wptr = (uint16_t __iomem *)MAILBOX_REG(ha, reg, 1);
for (cnt = 1; cnt < ha->mbx_count; cnt++) {
if (IS_QLA2200(ha) && cnt == 8)
wptr = (uint16_t __iomem *)MAILBOX_REG(ha, reg, 8);
if ((cnt == 4 || cnt == 5) && (mboxes & BIT_0))
ha->mailbox_out[cnt] = qla2x00_debounce_register(wptr);
else if (mboxes & BIT_0)
ha->mailbox_out[cnt] = RD_REG_WORD(wptr);
wptr++;
mboxes >>= 1;
}
}
static void
qla81xx_idc_event(scsi_qla_host_t *vha, uint16_t aen, uint16_t descr)
{
static char *event[] =
{ "Complete", "Request Notification", "Time Extension" };
int rval;
struct device_reg_24xx __iomem *reg24 = &vha->hw->iobase->isp24;
struct device_reg_82xx __iomem *reg82 = &vha->hw->iobase->isp82;
uint16_t __iomem *wptr;
uint16_t cnt, timeout, mb[QLA_IDC_ACK_REGS];
/* Seed data -- mailbox1 -> mailbox7. */
if (IS_QLA81XX(vha->hw) || IS_QLA83XX(vha->hw))
wptr = (uint16_t __iomem *)&reg24->mailbox1;
else if (IS_QLA8044(vha->hw))
wptr = (uint16_t __iomem *)&reg82->mailbox_out[1];
else
return;
for (cnt = 0; cnt < QLA_IDC_ACK_REGS; cnt++, wptr++)
mb[cnt] = RD_REG_WORD(wptr);
ql_dbg(ql_dbg_async, vha, 0x5021,
"Inter-Driver Communication %s -- "
"%04x %04x %04x %04x %04x %04x %04x.\n",
event[aen & 0xff], mb[0], mb[1], mb[2], mb[3],
mb[4], mb[5], mb[6]);
switch (aen) {
/* Handle IDC Error completion case. */
case MBA_IDC_COMPLETE:
if (mb[1] >> 15) {
vha->hw->flags.idc_compl_status = 1;
if (vha->hw->notify_dcbx_comp && !vha->vp_idx)
complete(&vha->hw->dcbx_comp);
}
break;
case MBA_IDC_NOTIFY:
/* Acknowledgement needed? [Notify && non-zero timeout]. */
timeout = (descr >> 8) & 0xf;
ql_dbg(ql_dbg_async, vha, 0x5022,
"%lu Inter-Driver Communication %s -- ACK timeout=%d.\n",
vha->host_no, event[aen & 0xff], timeout);
if (!timeout)
return;
rval = qla2x00_post_idc_ack_work(vha, mb);
if (rval != QLA_SUCCESS)
ql_log(ql_log_warn, vha, 0x5023,
"IDC failed to post ACK.\n");
break;
case MBA_IDC_TIME_EXT:
vha->hw->idc_extend_tmo = descr;
ql_dbg(ql_dbg_async, vha, 0x5087,
"%lu Inter-Driver Communication %s -- "
"Extend timeout by=%d.\n",
vha->host_no, event[aen & 0xff], vha->hw->idc_extend_tmo);
break;
}
}
#define LS_UNKNOWN 2
const char *
qla2x00_get_link_speed_str(struct qla_hw_data *ha, uint16_t speed)
{
static const char *const link_speeds[] = {
"1", "2", "?", "4", "8", "16", "32", "10"
};
#define QLA_LAST_SPEED 7
if (IS_QLA2100(ha) || IS_QLA2200(ha))
return link_speeds[0];
else if (speed == 0x13)
return link_speeds[QLA_LAST_SPEED];
else if (speed < QLA_LAST_SPEED)
return link_speeds[speed];
else
return link_speeds[LS_UNKNOWN];
}
static void
qla83xx_handle_8200_aen(scsi_qla_host_t *vha, uint16_t *mb)
{
struct qla_hw_data *ha = vha->hw;
/*
* 8200 AEN Interpretation:
* mb[0] = AEN code
* mb[1] = AEN Reason code
* mb[2] = LSW of Peg-Halt Status-1 Register
* mb[6] = MSW of Peg-Halt Status-1 Register
* mb[3] = LSW of Peg-Halt Status-2 register
* mb[7] = MSW of Peg-Halt Status-2 register
* mb[4] = IDC Device-State Register value
* mb[5] = IDC Driver-Presence Register value
*/
ql_dbg(ql_dbg_async, vha, 0x506b, "AEN Code: mb[0] = 0x%x AEN reason: "
"mb[1] = 0x%x PH-status1: mb[2] = 0x%x PH-status1: mb[6] = 0x%x.\n",
mb[0], mb[1], mb[2], mb[6]);
ql_dbg(ql_dbg_async, vha, 0x506c, "PH-status2: mb[3] = 0x%x "
"PH-status2: mb[7] = 0x%x Device-State: mb[4] = 0x%x "
"Drv-Presence: mb[5] = 0x%x.\n", mb[3], mb[7], mb[4], mb[5]);
if (mb[1] & (IDC_PEG_HALT_STATUS_CHANGE | IDC_NIC_FW_REPORTED_FAILURE |
IDC_HEARTBEAT_FAILURE)) {
ha->flags.nic_core_hung = 1;
ql_log(ql_log_warn, vha, 0x5060,
"83XX: F/W Error Reported: Check if reset required.\n");
if (mb[1] & IDC_PEG_HALT_STATUS_CHANGE) {
uint32_t protocol_engine_id, fw_err_code, err_level;
/*
* IDC_PEG_HALT_STATUS_CHANGE interpretation:
* - PEG-Halt Status-1 Register:
* (LSW = mb[2], MSW = mb[6])
* Bits 0-7 = protocol-engine ID
* Bits 8-28 = f/w error code
* Bits 29-31 = Error-level
* Error-level 0x1 = Non-Fatal error
* Error-level 0x2 = Recoverable Fatal error
* Error-level 0x4 = UnRecoverable Fatal error
* - PEG-Halt Status-2 Register:
* (LSW = mb[3], MSW = mb[7])
*/
protocol_engine_id = (mb[2] & 0xff);
fw_err_code = (((mb[2] & 0xff00) >> 8) |
((mb[6] & 0x1fff) << 8));
err_level = ((mb[6] & 0xe000) >> 13);
ql_log(ql_log_warn, vha, 0x5061, "PegHalt Status-1 "
"Register: protocol_engine_id=0x%x "
"fw_err_code=0x%x err_level=0x%x.\n",
protocol_engine_id, fw_err_code, err_level);
ql_log(ql_log_warn, vha, 0x5062, "PegHalt Status-2 "
"Register: 0x%x%x.\n", mb[7], mb[3]);
if (err_level == ERR_LEVEL_NON_FATAL) {
ql_log(ql_log_warn, vha, 0x5063,
"Not a fatal error, f/w has recovered "
"iteself.\n");
} else if (err_level == ERR_LEVEL_RECOVERABLE_FATAL) {
ql_log(ql_log_fatal, vha, 0x5064,
"Recoverable Fatal error: Chip reset "
"required.\n");
qla83xx_schedule_work(vha,
QLA83XX_NIC_CORE_RESET);
} else if (err_level == ERR_LEVEL_UNRECOVERABLE_FATAL) {
ql_log(ql_log_fatal, vha, 0x5065,
"Unrecoverable Fatal error: Set FAILED "
"state, reboot required.\n");
qla83xx_schedule_work(vha,
QLA83XX_NIC_CORE_UNRECOVERABLE);
}
}
if (mb[1] & IDC_NIC_FW_REPORTED_FAILURE) {
uint16_t peg_fw_state, nw_interface_link_up;
uint16_t nw_interface_signal_detect, sfp_status;
uint16_t htbt_counter, htbt_monitor_enable;
uint16_t sfp_additonal_info, sfp_multirate;
uint16_t sfp_tx_fault, link_speed, dcbx_status;
/*
* IDC_NIC_FW_REPORTED_FAILURE interpretation:
* - PEG-to-FC Status Register:
* (LSW = mb[2], MSW = mb[6])
* Bits 0-7 = Peg-Firmware state
* Bit 8 = N/W Interface Link-up
* Bit 9 = N/W Interface signal detected
* Bits 10-11 = SFP Status
* SFP Status 0x0 = SFP+ transceiver not expected
* SFP Status 0x1 = SFP+ transceiver not present
* SFP Status 0x2 = SFP+ transceiver invalid
* SFP Status 0x3 = SFP+ transceiver present and
* valid
* Bits 12-14 = Heartbeat Counter
* Bit 15 = Heartbeat Monitor Enable
* Bits 16-17 = SFP Additional Info
* SFP info 0x0 = Unregocnized transceiver for
* Ethernet
* SFP info 0x1 = SFP+ brand validation failed
* SFP info 0x2 = SFP+ speed validation failed
* SFP info 0x3 = SFP+ access error
* Bit 18 = SFP Multirate
* Bit 19 = SFP Tx Fault
* Bits 20-22 = Link Speed
* Bits 23-27 = Reserved
* Bits 28-30 = DCBX Status
* DCBX Status 0x0 = DCBX Disabled
* DCBX Status 0x1 = DCBX Enabled
* DCBX Status 0x2 = DCBX Exchange error
* Bit 31 = Reserved
*/
peg_fw_state = (mb[2] & 0x00ff);
nw_interface_link_up = ((mb[2] & 0x0100) >> 8);
nw_interface_signal_detect = ((mb[2] & 0x0200) >> 9);
sfp_status = ((mb[2] & 0x0c00) >> 10);
htbt_counter = ((mb[2] & 0x7000) >> 12);
htbt_monitor_enable = ((mb[2] & 0x8000) >> 15);
sfp_additonal_info = (mb[6] & 0x0003);
sfp_multirate = ((mb[6] & 0x0004) >> 2);
sfp_tx_fault = ((mb[6] & 0x0008) >> 3);
link_speed = ((mb[6] & 0x0070) >> 4);
dcbx_status = ((mb[6] & 0x7000) >> 12);
ql_log(ql_log_warn, vha, 0x5066,
"Peg-to-Fc Status Register:\n"
"peg_fw_state=0x%x, nw_interface_link_up=0x%x, "
"nw_interface_signal_detect=0x%x"
"\nsfp_statis=0x%x.\n ", peg_fw_state,
nw_interface_link_up, nw_interface_signal_detect,
sfp_status);
ql_log(ql_log_warn, vha, 0x5067,
"htbt_counter=0x%x, htbt_monitor_enable=0x%x, "
"sfp_additonal_info=0x%x, sfp_multirate=0x%x.\n ",
htbt_counter, htbt_monitor_enable,
sfp_additonal_info, sfp_multirate);
ql_log(ql_log_warn, vha, 0x5068,
"sfp_tx_fault=0x%x, link_state=0x%x, "
"dcbx_status=0x%x.\n", sfp_tx_fault, link_speed,
dcbx_status);
qla83xx_schedule_work(vha, QLA83XX_NIC_CORE_RESET);
}
if (mb[1] & IDC_HEARTBEAT_FAILURE) {
ql_log(ql_log_warn, vha, 0x5069,
"Heartbeat Failure encountered, chip reset "
"required.\n");
qla83xx_schedule_work(vha, QLA83XX_NIC_CORE_RESET);
}
}
if (mb[1] & IDC_DEVICE_STATE_CHANGE) {
ql_log(ql_log_info, vha, 0x506a,
"IDC Device-State changed = 0x%x.\n", mb[4]);
if (ha->flags.nic_core_reset_owner)
return;
qla83xx_schedule_work(vha, MBA_IDC_AEN);
}
}
int
qla2x00_is_a_vp_did(scsi_qla_host_t *vha, uint32_t rscn_entry)
{
struct qla_hw_data *ha = vha->hw;
scsi_qla_host_t *vp;
uint32_t vp_did;
unsigned long flags;
int ret = 0;
if (!ha->num_vhosts)
return ret;
spin_lock_irqsave(&ha->vport_slock, flags);
list_for_each_entry(vp, &ha->vp_list, list) {
vp_did = vp->d_id.b24;
if (vp_did == rscn_entry) {
ret = 1;
break;
}
}
spin_unlock_irqrestore(&ha->vport_slock, flags);
return ret;
}
/**
* qla2x00_async_event() - Process aynchronous events.
* @ha: SCSI driver HA context
* @mb: Mailbox registers (0 - 3)
*/
void
qla2x00_async_event(scsi_qla_host_t *vha, struct rsp_que *rsp, uint16_t *mb)
{
uint16_t handle_cnt;
uint16_t cnt, mbx;
uint32_t handles[5];
struct qla_hw_data *ha = vha->hw;
struct device_reg_2xxx __iomem *reg = &ha->iobase->isp;
struct device_reg_24xx __iomem *reg24 = &ha->iobase->isp24;
struct device_reg_82xx __iomem *reg82 = &ha->iobase->isp82;
uint32_t rscn_entry, host_pid, tmp_pid;
unsigned long flags;
fc_port_t *fcport = NULL;
/* Setup to process RIO completion. */
handle_cnt = 0;
if (IS_CNA_CAPABLE(ha))
goto skip_rio;
switch (mb[0]) {
case MBA_SCSI_COMPLETION:
handles[0] = le32_to_cpu((uint32_t)((mb[2] << 16) | mb[1]));
handle_cnt = 1;
break;
case MBA_CMPLT_1_16BIT:
handles[0] = mb[1];
handle_cnt = 1;
mb[0] = MBA_SCSI_COMPLETION;
break;
case MBA_CMPLT_2_16BIT:
handles[0] = mb[1];
handles[1] = mb[2];
handle_cnt = 2;
mb[0] = MBA_SCSI_COMPLETION;
break;
case MBA_CMPLT_3_16BIT:
handles[0] = mb[1];
handles[1] = mb[2];
handles[2] = mb[3];
handle_cnt = 3;
mb[0] = MBA_SCSI_COMPLETION;
break;
case MBA_CMPLT_4_16BIT:
handles[0] = mb[1];
handles[1] = mb[2];
handles[2] = mb[3];
handles[3] = (uint32_t)RD_MAILBOX_REG(ha, reg, 6);
handle_cnt = 4;
mb[0] = MBA_SCSI_COMPLETION;
break;
case MBA_CMPLT_5_16BIT:
handles[0] = mb[1];
handles[1] = mb[2];
handles[2] = mb[3];
handles[3] = (uint32_t)RD_MAILBOX_REG(ha, reg, 6);
handles[4] = (uint32_t)RD_MAILBOX_REG(ha, reg, 7);
handle_cnt = 5;
mb[0] = MBA_SCSI_COMPLETION;
break;
case MBA_CMPLT_2_32BIT:
handles[0] = le32_to_cpu((uint32_t)((mb[2] << 16) | mb[1]));
handles[1] = le32_to_cpu(
((uint32_t)(RD_MAILBOX_REG(ha, reg, 7) << 16)) |
RD_MAILBOX_REG(ha, reg, 6));
handle_cnt = 2;
mb[0] = MBA_SCSI_COMPLETION;
break;
default:
break;
}
skip_rio:
switch (mb[0]) {
case MBA_SCSI_COMPLETION: /* Fast Post */
if (!vha->flags.online)
break;
for (cnt = 0; cnt < handle_cnt; cnt++)
qla2x00_process_completed_request(vha, rsp->req,
handles[cnt]);
break;
case MBA_RESET: /* Reset */
ql_dbg(ql_dbg_async, vha, 0x5002,
"Asynchronous RESET.\n");
set_bit(RESET_MARKER_NEEDED, &vha->dpc_flags);
break;
case MBA_SYSTEM_ERR: /* System Error */
mbx = (IS_QLA81XX(ha) || IS_QLA83XX(ha) || IS_QLA27XX(ha)) ?
RD_REG_WORD(&reg24->mailbox7) : 0;
ql_log(ql_log_warn, vha, 0x5003,
"ISP System Error - mbx1=%xh mbx2=%xh mbx3=%xh "
"mbx7=%xh.\n", mb[1], mb[2], mb[3], mbx);
ha->isp_ops->fw_dump(vha, 1);
if (IS_FWI2_CAPABLE(ha)) {
if (mb[1] == 0 && mb[2] == 0) {
ql_log(ql_log_fatal, vha, 0x5004,
"Unrecoverable Hardware Error: adapter "
"marked OFFLINE!\n");
vha->flags.online = 0;
vha->device_flags |= DFLG_DEV_FAILED;
} else {
/* Check to see if MPI timeout occurred */
if ((mbx & MBX_3) && (ha->port_no == 0))
set_bit(MPI_RESET_NEEDED,
&vha->dpc_flags);
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
}
} else if (mb[1] == 0) {
ql_log(ql_log_fatal, vha, 0x5005,
"Unrecoverable Hardware Error: adapter marked "
"OFFLINE!\n");
vha->flags.online = 0;
vha->device_flags |= DFLG_DEV_FAILED;
} else
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
break;
case MBA_REQ_TRANSFER_ERR: /* Request Transfer Error */
ql_log(ql_log_warn, vha, 0x5006,
"ISP Request Transfer Error (%x).\n", mb[1]);
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
break;
case MBA_RSP_TRANSFER_ERR: /* Response Transfer Error */
ql_log(ql_log_warn, vha, 0x5007,
"ISP Response Transfer Error.\n");
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
break;
case MBA_WAKEUP_THRES: /* Request Queue Wake-up */
ql_dbg(ql_dbg_async, vha, 0x5008,
"Asynchronous WAKEUP_THRES.\n");
break;
case MBA_LIP_OCCURRED: /* Loop Initialization Procedure */
ql_dbg(ql_dbg_async, vha, 0x5009,
"LIP occurred (%x).\n", mb[1]);
if (atomic_read(&vha->loop_state) != LOOP_DOWN) {
atomic_set(&vha->loop_state, LOOP_DOWN);
atomic_set(&vha->loop_down_timer, LOOP_DOWN_TIME);
qla2x00_mark_all_devices_lost(vha, 1);
}
if (vha->vp_idx) {
atomic_set(&vha->vp_state, VP_FAILED);
fc_vport_set_state(vha->fc_vport, FC_VPORT_FAILED);
}
set_bit(REGISTER_FC4_NEEDED, &vha->dpc_flags);
set_bit(REGISTER_FDMI_NEEDED, &vha->dpc_flags);
vha->flags.management_server_logged_in = 0;
qla2x00_post_aen_work(vha, FCH_EVT_LIP, mb[1]);
break;
case MBA_LOOP_UP: /* Loop Up Event */
if (IS_QLA2100(ha) || IS_QLA2200(ha))
ha->link_data_rate = PORT_SPEED_1GB;
else
ha->link_data_rate = mb[1];
ql_log(ql_log_info, vha, 0x500a,
"LOOP UP detected (%s Gbps).\n",
qla2x00_get_link_speed_str(ha, ha->link_data_rate));
vha->flags.management_server_logged_in = 0;
qla2x00_post_aen_work(vha, FCH_EVT_LINKUP, ha->link_data_rate);
break;
case MBA_LOOP_DOWN: /* Loop Down Event */
mbx = (IS_QLA81XX(ha) || IS_QLA8031(ha))
? RD_REG_WORD(&reg24->mailbox4) : 0;
mbx = (IS_P3P_TYPE(ha)) ? RD_REG_WORD(&reg82->mailbox_out[4])
: mbx;
ql_log(ql_log_info, vha, 0x500b,
"LOOP DOWN detected (%x %x %x %x).\n",
mb[1], mb[2], mb[3], mbx);
if (atomic_read(&vha->loop_state) != LOOP_DOWN) {
atomic_set(&vha->loop_state, LOOP_DOWN);
atomic_set(&vha->loop_down_timer, LOOP_DOWN_TIME);
/*
* In case of loop down, restore WWPN from
* NVRAM in case of FA-WWPN capable ISP
* Restore for Physical Port only
*/
if (!vha->vp_idx) {
if (ha->flags.fawwpn_enabled) {
void *wwpn = ha->init_cb->port_name;
memcpy(vha->port_name, wwpn, WWN_SIZE);
fc_host_port_name(vha->host) =
wwn_to_u64(vha->port_name);
ql_dbg(ql_dbg_init + ql_dbg_verbose,
vha, 0x0144, "LOOP DOWN detected,"
"restore WWPN %016llx\n",
wwn_to_u64(vha->port_name));
}
clear_bit(VP_CONFIG_OK, &vha->vp_flags);
}
vha->device_flags |= DFLG_NO_CABLE;
qla2x00_mark_all_devices_lost(vha, 1);
}
if (vha->vp_idx) {
atomic_set(&vha->vp_state, VP_FAILED);
fc_vport_set_state(vha->fc_vport, FC_VPORT_FAILED);
}
vha->flags.management_server_logged_in = 0;
ha->link_data_rate = PORT_SPEED_UNKNOWN;
qla2x00_post_aen_work(vha, FCH_EVT_LINKDOWN, 0);
break;
case MBA_LIP_RESET: /* LIP reset occurred */
ql_dbg(ql_dbg_async, vha, 0x500c,
"LIP reset occurred (%x).\n", mb[1]);
if (atomic_read(&vha->loop_state) != LOOP_DOWN) {
atomic_set(&vha->loop_state, LOOP_DOWN);
atomic_set(&vha->loop_down_timer, LOOP_DOWN_TIME);
qla2x00_mark_all_devices_lost(vha, 1);
}
if (vha->vp_idx) {
atomic_set(&vha->vp_state, VP_FAILED);
fc_vport_set_state(vha->fc_vport, FC_VPORT_FAILED);
}
set_bit(RESET_MARKER_NEEDED, &vha->dpc_flags);
ha->operating_mode = LOOP;
vha->flags.management_server_logged_in = 0;
qla2x00_post_aen_work(vha, FCH_EVT_LIPRESET, mb[1]);
break;
/* case MBA_DCBX_COMPLETE: */
case MBA_POINT_TO_POINT: /* Point-to-Point */
if (IS_QLA2100(ha))
break;
if (IS_CNA_CAPABLE(ha)) {
ql_dbg(ql_dbg_async, vha, 0x500d,
"DCBX Completed -- %04x %04x %04x.\n",
mb[1], mb[2], mb[3]);
if (ha->notify_dcbx_comp && !vha->vp_idx)
complete(&ha->dcbx_comp);
} else
ql_dbg(ql_dbg_async, vha, 0x500e,
"Asynchronous P2P MODE received.\n");
/*
* Until there's a transition from loop down to loop up, treat
* this as loop down only.
*/
if (atomic_read(&vha->loop_state) != LOOP_DOWN) {
atomic_set(&vha->loop_state, LOOP_DOWN);
if (!atomic_read(&vha->loop_down_timer))
atomic_set(&vha->loop_down_timer,
LOOP_DOWN_TIME);
qla2x00_mark_all_devices_lost(vha, 1);
}
if (vha->vp_idx) {
atomic_set(&vha->vp_state, VP_FAILED);
fc_vport_set_state(vha->fc_vport, FC_VPORT_FAILED);
}
if (!(test_bit(ABORT_ISP_ACTIVE, &vha->dpc_flags)))
set_bit(RESET_MARKER_NEEDED, &vha->dpc_flags);
set_bit(REGISTER_FC4_NEEDED, &vha->dpc_flags);
set_bit(REGISTER_FDMI_NEEDED, &vha->dpc_flags);
ha->flags.gpsc_supported = 1;
vha->flags.management_server_logged_in = 0;
break;
case MBA_CHG_IN_CONNECTION: /* Change in connection mode */
if (IS_QLA2100(ha))
break;
ql_dbg(ql_dbg_async, vha, 0x500f,
"Configuration change detected: value=%x.\n", mb[1]);
if (atomic_read(&vha->loop_state) != LOOP_DOWN) {
atomic_set(&vha->loop_state, LOOP_DOWN);
if (!atomic_read(&vha->loop_down_timer))
atomic_set(&vha->loop_down_timer,
LOOP_DOWN_TIME);
qla2x00_mark_all_devices_lost(vha, 1);
}
if (vha->vp_idx) {
atomic_set(&vha->vp_state, VP_FAILED);
fc_vport_set_state(vha->fc_vport, FC_VPORT_FAILED);
}
set_bit(LOOP_RESYNC_NEEDED, &vha->dpc_flags);
set_bit(LOCAL_LOOP_UPDATE, &vha->dpc_flags);
break;
case MBA_PORT_UPDATE: /* Port database update */
/*
* Handle only global and vn-port update events
*
* Relevant inputs:
* mb[1] = N_Port handle of changed port
* OR 0xffff for global event
* mb[2] = New login state
* 7 = Port logged out
* mb[3] = LSB is vp_idx, 0xff = all vps
*
* Skip processing if:
* Event is global, vp_idx is NOT all vps,
* vp_idx does not match
* Event is not global, vp_idx does not match
*/
if (IS_QLA2XXX_MIDTYPE(ha) &&
((mb[1] == 0xffff && (mb[3] & 0xff) != 0xff) ||
(mb[1] != 0xffff)) && vha->vp_idx != (mb[3] & 0xff))
break;
/* Global event -- port logout or port unavailable. */
if (mb[1] == 0xffff && mb[2] == 0x7) {
ql_dbg(ql_dbg_async, vha, 0x5010,
"Port unavailable %04x %04x %04x.\n",
mb[1], mb[2], mb[3]);
ql_log(ql_log_warn, vha, 0x505e,
"Link is offline.\n");
if (atomic_read(&vha->loop_state) != LOOP_DOWN) {
atomic_set(&vha->loop_state, LOOP_DOWN);
atomic_set(&vha->loop_down_timer,
LOOP_DOWN_TIME);
vha->device_flags |= DFLG_NO_CABLE;
qla2x00_mark_all_devices_lost(vha, 1);
}
if (vha->vp_idx) {
atomic_set(&vha->vp_state, VP_FAILED);
fc_vport_set_state(vha->fc_vport,
FC_VPORT_FAILED);
qla2x00_mark_all_devices_lost(vha, 1);
}
vha->flags.management_server_logged_in = 0;
ha->link_data_rate = PORT_SPEED_UNKNOWN;
break;
}
/*
* If PORT UPDATE is global (received LIP_OCCURRED/LIP_RESET
* event etc. earlier indicating loop is down) then process
* it. Otherwise ignore it and Wait for RSCN to come in.
*/
atomic_set(&vha->loop_down_timer, 0);
if (atomic_read(&vha->loop_state) != LOOP_DOWN &&
atomic_read(&vha->loop_state) != LOOP_DEAD) {
ql_dbg(ql_dbg_async, vha, 0x5011,
"Asynchronous PORT UPDATE ignored %04x/%04x/%04x.\n",
mb[1], mb[2], mb[3]);
qlt_async_event(mb[0], vha, mb);
break;
}
ql_dbg(ql_dbg_async, vha, 0x5012,
"Port database changed %04x %04x %04x.\n",
mb[1], mb[2], mb[3]);
/*
* Mark all devices as missing so we will login again.
*/
atomic_set(&vha->loop_state, LOOP_UP);
qla2x00_mark_all_devices_lost(vha, 1);
if (vha->vp_idx == 0 && !qla_ini_mode_enabled(vha))
set_bit(SCR_PENDING, &vha->dpc_flags);
set_bit(LOOP_RESYNC_NEEDED, &vha->dpc_flags);
set_bit(LOCAL_LOOP_UPDATE, &vha->dpc_flags);
set_bit(VP_CONFIG_OK, &vha->vp_flags);
qlt_async_event(mb[0], vha, mb);
break;
case MBA_RSCN_UPDATE: /* State Change Registration */
/* Check if the Vport has issued a SCR */
if (vha->vp_idx && test_bit(VP_SCR_NEEDED, &vha->vp_flags))
break;
/* Only handle SCNs for our Vport index. */
if (ha->flags.npiv_supported && vha->vp_idx != (mb[3] & 0xff))
break;
ql_dbg(ql_dbg_async, vha, 0x5013,
"RSCN database changed -- %04x %04x %04x.\n",
mb[1], mb[2], mb[3]);
rscn_entry = ((mb[1] & 0xff) << 16) | mb[2];
host_pid = (vha->d_id.b.domain << 16) | (vha->d_id.b.area << 8)
| vha->d_id.b.al_pa;
if (rscn_entry == host_pid) {
ql_dbg(ql_dbg_async, vha, 0x5014,
"Ignoring RSCN update to local host "
"port ID (%06x).\n", host_pid);
break;
}
/* Ignore reserved bits from RSCN-payload. */
rscn_entry = ((mb[1] & 0x3ff) << 16) | mb[2];
/* Skip RSCNs for virtual ports on the same physical port */
if (qla2x00_is_a_vp_did(vha, rscn_entry))
break;
/*
* Search for the rport related to this RSCN entry and mark it
* as lost.
*/
list_for_each_entry(fcport, &vha->vp_fcports, list) {
if (atomic_read(&fcport->state) != FCS_ONLINE)
continue;
tmp_pid = fcport->d_id.b24;
if (fcport->d_id.b24 == rscn_entry) {
qla2x00_mark_device_lost(vha, fcport, 0, 0);
break;
}
}
atomic_set(&vha->loop_down_timer, 0);
vha->flags.management_server_logged_in = 0;
set_bit(LOOP_RESYNC_NEEDED, &vha->dpc_flags);
set_bit(RSCN_UPDATE, &vha->dpc_flags);
qla2x00_post_aen_work(vha, FCH_EVT_RSCN, rscn_entry);
break;
/* case MBA_RIO_RESPONSE: */
case MBA_ZIO_RESPONSE:
ql_dbg(ql_dbg_async, vha, 0x5015,
"[R|Z]IO update completion.\n");
if (IS_FWI2_CAPABLE(ha))
qla24xx_process_response_queue(vha, rsp);
else
qla2x00_process_response_queue(rsp);
break;
case MBA_DISCARD_RND_FRAME:
ql_dbg(ql_dbg_async, vha, 0x5016,
"Discard RND Frame -- %04x %04x %04x.\n",
mb[1], mb[2], mb[3]);
break;
case MBA_TRACE_NOTIFICATION:
ql_dbg(ql_dbg_async, vha, 0x5017,
"Trace Notification -- %04x %04x.\n", mb[1], mb[2]);
break;
case MBA_ISP84XX_ALERT:
ql_dbg(ql_dbg_async, vha, 0x5018,
"ISP84XX Alert Notification -- %04x %04x %04x.\n",
mb[1], mb[2], mb[3]);
spin_lock_irqsave(&ha->cs84xx->access_lock, flags);
switch (mb[1]) {
case A84_PANIC_RECOVERY:
ql_log(ql_log_info, vha, 0x5019,
"Alert 84XX: panic recovery %04x %04x.\n",
mb[2], mb[3]);
break;
case A84_OP_LOGIN_COMPLETE:
ha->cs84xx->op_fw_version = mb[3] << 16 | mb[2];
ql_log(ql_log_info, vha, 0x501a,
"Alert 84XX: firmware version %x.\n",
ha->cs84xx->op_fw_version);
break;
case A84_DIAG_LOGIN_COMPLETE:
ha->cs84xx->diag_fw_version = mb[3] << 16 | mb[2];
ql_log(ql_log_info, vha, 0x501b,
"Alert 84XX: diagnostic firmware version %x.\n",
ha->cs84xx->diag_fw_version);
break;
case A84_GOLD_LOGIN_COMPLETE:
ha->cs84xx->diag_fw_version = mb[3] << 16 | mb[2];
ha->cs84xx->fw_update = 1;
ql_log(ql_log_info, vha, 0x501c,
"Alert 84XX: gold firmware version %x.\n",
ha->cs84xx->gold_fw_version);
break;
default:
ql_log(ql_log_warn, vha, 0x501d,
"Alert 84xx: Invalid Alert %04x %04x %04x.\n",
mb[1], mb[2], mb[3]);
}
spin_unlock_irqrestore(&ha->cs84xx->access_lock, flags);
break;
case MBA_DCBX_START:
ql_dbg(ql_dbg_async, vha, 0x501e,
"DCBX Started -- %04x %04x %04x.\n",
mb[1], mb[2], mb[3]);
break;
case MBA_DCBX_PARAM_UPDATE:
ql_dbg(ql_dbg_async, vha, 0x501f,
"DCBX Parameters Updated -- %04x %04x %04x.\n",
mb[1], mb[2], mb[3]);
break;
case MBA_FCF_CONF_ERR:
ql_dbg(ql_dbg_async, vha, 0x5020,
"FCF Configuration Error -- %04x %04x %04x.\n",
mb[1], mb[2], mb[3]);
break;
case MBA_IDC_NOTIFY:
if (IS_QLA8031(vha->hw) || IS_QLA8044(ha)) {
mb[4] = RD_REG_WORD(&reg24->mailbox4);
if (((mb[2] & 0x7fff) == MBC_PORT_RESET ||
(mb[2] & 0x7fff) == MBC_SET_PORT_CONFIG) &&
(mb[4] & INTERNAL_LOOPBACK_MASK) != 0) {
set_bit(ISP_QUIESCE_NEEDED, &vha->dpc_flags);
/*
* Extend loop down timer since port is active.
*/
if (atomic_read(&vha->loop_state) == LOOP_DOWN)
atomic_set(&vha->loop_down_timer,
LOOP_DOWN_TIME);
qla2xxx_wake_dpc(vha);
}
}
case MBA_IDC_COMPLETE:
if (ha->notify_lb_portup_comp && !vha->vp_idx)
complete(&ha->lb_portup_comp);
/* Fallthru */
case MBA_IDC_TIME_EXT:
if (IS_QLA81XX(vha->hw) || IS_QLA8031(vha->hw) ||
IS_QLA8044(ha))
qla81xx_idc_event(vha, mb[0], mb[1]);
break;
case MBA_IDC_AEN:
mb[4] = RD_REG_WORD(&reg24->mailbox4);
mb[5] = RD_REG_WORD(&reg24->mailbox5);
mb[6] = RD_REG_WORD(&reg24->mailbox6);
mb[7] = RD_REG_WORD(&reg24->mailbox7);
qla83xx_handle_8200_aen(vha, mb);
break;
case MBA_DPORT_DIAGNOSTICS:
ql_dbg(ql_dbg_async, vha, 0x5052,
"D-Port Diagnostics: %04x %04x=%s\n", mb[0], mb[1],
mb[1] == 0 ? "start" :
mb[1] == 1 ? "done (ok)" :
mb[1] == 2 ? "done (error)" : "other");
break;
default:
ql_dbg(ql_dbg_async, vha, 0x5057,
"Unknown AEN:%04x %04x %04x %04x\n",
mb[0], mb[1], mb[2], mb[3]);
}
qlt_async_event(mb[0], vha, mb);
if (!vha->vp_idx && ha->num_vhosts)
qla2x00_alert_all_vps(rsp, mb);
}
/**
* qla2x00_process_completed_request() - Process a Fast Post response.
* @ha: SCSI driver HA context
* @index: SRB index
*/
void
qla2x00_process_completed_request(struct scsi_qla_host *vha,
struct req_que *req, uint32_t index)
{
srb_t *sp;
struct qla_hw_data *ha = vha->hw;
/* Validate handle. */
if (index >= req->num_outstanding_cmds) {
ql_log(ql_log_warn, vha, 0x3014,
"Invalid SCSI command index (%x).\n", index);
if (IS_P3P_TYPE(ha))
set_bit(FCOE_CTX_RESET_NEEDED, &vha->dpc_flags);
else
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
return;
}
sp = req->outstanding_cmds[index];
if (sp) {
/* Free outstanding command slot. */
req->outstanding_cmds[index] = NULL;
/* Save ISP completion status */
sp->done(ha, sp, DID_OK << 16);
} else {
ql_log(ql_log_warn, vha, 0x3016, "Invalid SCSI SRB.\n");
if (IS_P3P_TYPE(ha))
set_bit(FCOE_CTX_RESET_NEEDED, &vha->dpc_flags);
else
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
}
}
srb_t *
qla2x00_get_sp_from_handle(scsi_qla_host_t *vha, const char *func,
struct req_que *req, void *iocb)
{
struct qla_hw_data *ha = vha->hw;
sts_entry_t *pkt = iocb;
srb_t *sp = NULL;
uint16_t index;
index = LSW(pkt->handle);
if (index >= req->num_outstanding_cmds) {
ql_log(ql_log_warn, vha, 0x5031,
"Invalid command index (%x).\n", index);
if (IS_P3P_TYPE(ha))
set_bit(FCOE_CTX_RESET_NEEDED, &vha->dpc_flags);
else
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
goto done;
}
sp = req->outstanding_cmds[index];
if (!sp) {
ql_log(ql_log_warn, vha, 0x5032,
"Invalid completion handle (%x) -- timed-out.\n", index);
return sp;
}
if (sp->handle != index) {
ql_log(ql_log_warn, vha, 0x5033,
"SRB handle (%x) mismatch %x.\n", sp->handle, index);
return NULL;
}
req->outstanding_cmds[index] = NULL;
done:
return sp;
}
static void
qla2x00_mbx_iocb_entry(scsi_qla_host_t *vha, struct req_que *req,
struct mbx_entry *mbx)
{
const char func[] = "MBX-IOCB";
const char *type;
fc_port_t *fcport;
srb_t *sp;
struct srb_iocb *lio;
uint16_t *data;
uint16_t status;
sp = qla2x00_get_sp_from_handle(vha, func, req, mbx);
if (!sp)
return;
lio = &sp->u.iocb_cmd;
type = sp->name;
fcport = sp->fcport;
data = lio->u.logio.data;
data[0] = MBS_COMMAND_ERROR;
data[1] = lio->u.logio.flags & SRB_LOGIN_RETRIED ?
QLA_LOGIO_LOGIN_RETRIED : 0;
if (mbx->entry_status) {
ql_dbg(ql_dbg_async, vha, 0x5043,
"Async-%s error entry - hdl=%x portid=%02x%02x%02x "
"entry-status=%x status=%x state-flag=%x "
"status-flags=%x.\n", type, sp->handle,
fcport->d_id.b.domain, fcport->d_id.b.area,
fcport->d_id.b.al_pa, mbx->entry_status,
le16_to_cpu(mbx->status), le16_to_cpu(mbx->state_flags),
le16_to_cpu(mbx->status_flags));
ql_dump_buffer(ql_dbg_async + ql_dbg_buffer, vha, 0x5029,
(uint8_t *)mbx, sizeof(*mbx));
goto logio_done;
}
status = le16_to_cpu(mbx->status);
if (status == 0x30 && sp->type == SRB_LOGIN_CMD &&
le16_to_cpu(mbx->mb0) == MBS_COMMAND_COMPLETE)
status = 0;
if (!status && le16_to_cpu(mbx->mb0) == MBS_COMMAND_COMPLETE) {
ql_dbg(ql_dbg_async, vha, 0x5045,
"Async-%s complete - hdl=%x portid=%02x%02x%02x mbx1=%x.\n",
type, sp->handle, fcport->d_id.b.domain,
fcport->d_id.b.area, fcport->d_id.b.al_pa,
le16_to_cpu(mbx->mb1));
data[0] = MBS_COMMAND_COMPLETE;
if (sp->type == SRB_LOGIN_CMD) {
fcport->port_type = FCT_TARGET;
if (le16_to_cpu(mbx->mb1) & BIT_0)
fcport->port_type = FCT_INITIATOR;
else if (le16_to_cpu(mbx->mb1) & BIT_1)
fcport->flags |= FCF_FCP2_DEVICE;
}
goto logio_done;
}
data[0] = le16_to_cpu(mbx->mb0);
switch (data[0]) {
case MBS_PORT_ID_USED:
data[1] = le16_to_cpu(mbx->mb1);
break;
case MBS_LOOP_ID_USED:
break;
default:
data[0] = MBS_COMMAND_ERROR;
break;
}
ql_log(ql_log_warn, vha, 0x5046,
"Async-%s failed - hdl=%x portid=%02x%02x%02x status=%x "
"mb0=%x mb1=%x mb2=%x mb6=%x mb7=%x.\n", type, sp->handle,
fcport->d_id.b.domain, fcport->d_id.b.area, fcport->d_id.b.al_pa,
status, le16_to_cpu(mbx->mb0), le16_to_cpu(mbx->mb1),
le16_to_cpu(mbx->mb2), le16_to_cpu(mbx->mb6),
le16_to_cpu(mbx->mb7));
logio_done:
sp->done(vha, sp, 0);
}
static void
qla2x00_ct_entry(scsi_qla_host_t *vha, struct req_que *req,
sts_entry_t *pkt, int iocb_type)
{
const char func[] = "CT_IOCB";
const char *type;
srb_t *sp;
struct fc_bsg_job *bsg_job;
uint16_t comp_status;
int res;
sp = qla2x00_get_sp_from_handle(vha, func, req, pkt);
if (!sp)
return;
bsg_job = sp->u.bsg_job;
type = "ct pass-through";
comp_status = le16_to_cpu(pkt->comp_status);
/* return FC_CTELS_STATUS_OK and leave the decoding of the ELS/CT
* fc payload to the caller
*/
bsg_job->reply->reply_data.ctels_reply.status = FC_CTELS_STATUS_OK;
bsg_job->reply_len = sizeof(struct fc_bsg_reply);
if (comp_status != CS_COMPLETE) {
if (comp_status == CS_DATA_UNDERRUN) {
res = DID_OK << 16;
bsg_job->reply->reply_payload_rcv_len =
le16_to_cpu(((sts_entry_t *)pkt)->rsp_info_len);
ql_log(ql_log_warn, vha, 0x5048,
"CT pass-through-%s error "
"comp_status-status=0x%x total_byte = 0x%x.\n",
type, comp_status,
bsg_job->reply->reply_payload_rcv_len);
} else {
ql_log(ql_log_warn, vha, 0x5049,
"CT pass-through-%s error "
"comp_status-status=0x%x.\n", type, comp_status);
res = DID_ERROR << 16;
bsg_job->reply->reply_payload_rcv_len = 0;
}
ql_dump_buffer(ql_dbg_async + ql_dbg_buffer, vha, 0x5035,
(uint8_t *)pkt, sizeof(*pkt));
} else {
res = DID_OK << 16;
bsg_job->reply->reply_payload_rcv_len =
bsg_job->reply_payload.payload_len;
bsg_job->reply_len = 0;
}
sp->done(vha, sp, res);
}
static void
qla24xx_els_ct_entry(scsi_qla_host_t *vha, struct req_que *req,
struct sts_entry_24xx *pkt, int iocb_type)
{
const char func[] = "ELS_CT_IOCB";
const char *type;
srb_t *sp;
struct fc_bsg_job *bsg_job;
uint16_t comp_status;
uint32_t fw_status[3];
uint8_t* fw_sts_ptr;
int res;
sp = qla2x00_get_sp_from_handle(vha, func, req, pkt);
if (!sp)
return;
bsg_job = sp->u.bsg_job;
type = NULL;
switch (sp->type) {
case SRB_ELS_CMD_RPT:
case SRB_ELS_CMD_HST:
type = "els";
break;
case SRB_CT_CMD:
type = "ct pass-through";
break;
default:
ql_dbg(ql_dbg_user, vha, 0x503e,
"Unrecognized SRB: (%p) type=%d.\n", sp, sp->type);
return;
}
comp_status = fw_status[0] = le16_to_cpu(pkt->comp_status);
fw_status[1] = le16_to_cpu(((struct els_sts_entry_24xx*)pkt)->error_subcode_1);
fw_status[2] = le16_to_cpu(((struct els_sts_entry_24xx*)pkt)->error_subcode_2);
/* return FC_CTELS_STATUS_OK and leave the decoding of the ELS/CT
* fc payload to the caller
*/
bsg_job->reply->reply_data.ctels_reply.status = FC_CTELS_STATUS_OK;
bsg_job->reply_len = sizeof(struct fc_bsg_reply) + sizeof(fw_status);
if (comp_status != CS_COMPLETE) {
if (comp_status == CS_DATA_UNDERRUN) {
res = DID_OK << 16;
bsg_job->reply->reply_payload_rcv_len =
le16_to_cpu(((struct els_sts_entry_24xx *)pkt)->total_byte_count);
ql_dbg(ql_dbg_user, vha, 0x503f,
"ELS-CT pass-through-%s error hdl=%x comp_status-status=0x%x "
"error subcode 1=0x%x error subcode 2=0x%x total_byte = 0x%x.\n",
type, sp->handle, comp_status, fw_status[1], fw_status[2],
le16_to_cpu(((struct els_sts_entry_24xx *)
pkt)->total_byte_count));
fw_sts_ptr = ((uint8_t*)bsg_job->req->sense) + sizeof(struct fc_bsg_reply);
memcpy( fw_sts_ptr, fw_status, sizeof(fw_status));
}
else {
ql_dbg(ql_dbg_user, vha, 0x5040,
"ELS-CT pass-through-%s error hdl=%x comp_status-status=0x%x "
"error subcode 1=0x%x error subcode 2=0x%x.\n",
type, sp->handle, comp_status,
le16_to_cpu(((struct els_sts_entry_24xx *)
pkt)->error_subcode_1),
le16_to_cpu(((struct els_sts_entry_24xx *)
pkt)->error_subcode_2));
res = DID_ERROR << 16;
bsg_job->reply->reply_payload_rcv_len = 0;
fw_sts_ptr = ((uint8_t*)bsg_job->req->sense) + sizeof(struct fc_bsg_reply);
memcpy( fw_sts_ptr, fw_status, sizeof(fw_status));
}
ql_dump_buffer(ql_dbg_user + ql_dbg_buffer, vha, 0x5056,
(uint8_t *)pkt, sizeof(*pkt));
}
else {
res = DID_OK << 16;
bsg_job->reply->reply_payload_rcv_len = bsg_job->reply_payload.payload_len;
bsg_job->reply_len = 0;
}
sp->done(vha, sp, res);
}
static void
qla24xx_logio_entry(scsi_qla_host_t *vha, struct req_que *req,
struct logio_entry_24xx *logio)
{
const char func[] = "LOGIO-IOCB";
const char *type;
fc_port_t *fcport;
srb_t *sp;
struct srb_iocb *lio;
uint16_t *data;
uint32_t iop[2];
sp = qla2x00_get_sp_from_handle(vha, func, req, logio);
if (!sp)
return;
lio = &sp->u.iocb_cmd;
type = sp->name;
fcport = sp->fcport;
data = lio->u.logio.data;
data[0] = MBS_COMMAND_ERROR;
data[1] = lio->u.logio.flags & SRB_LOGIN_RETRIED ?
QLA_LOGIO_LOGIN_RETRIED : 0;
if (logio->entry_status) {
ql_log(ql_log_warn, fcport->vha, 0x5034,
"Async-%s error entry - hdl=%x"
"portid=%02x%02x%02x entry-status=%x.\n",
type, sp->handle, fcport->d_id.b.domain,
fcport->d_id.b.area, fcport->d_id.b.al_pa,
logio->entry_status);
ql_dump_buffer(ql_dbg_async + ql_dbg_buffer, vha, 0x504d,
(uint8_t *)logio, sizeof(*logio));
goto logio_done;
}
if (le16_to_cpu(logio->comp_status) == CS_COMPLETE) {
ql_dbg(ql_dbg_async, fcport->vha, 0x5036,
"Async-%s complete - hdl=%x portid=%02x%02x%02x "
"iop0=%x.\n", type, sp->handle, fcport->d_id.b.domain,
fcport->d_id.b.area, fcport->d_id.b.al_pa,
le32_to_cpu(logio->io_parameter[0]));
data[0] = MBS_COMMAND_COMPLETE;
if (sp->type != SRB_LOGIN_CMD)
goto logio_done;
iop[0] = le32_to_cpu(logio->io_parameter[0]);
if (iop[0] & BIT_4) {
fcport->port_type = FCT_TARGET;
if (iop[0] & BIT_8)
fcport->flags |= FCF_FCP2_DEVICE;
} else if (iop[0] & BIT_5)
fcport->port_type = FCT_INITIATOR;
if (iop[0] & BIT_7)
fcport->flags |= FCF_CONF_COMP_SUPPORTED;
if (logio->io_parameter[7] || logio->io_parameter[8])
fcport->supported_classes |= FC_COS_CLASS2;
if (logio->io_parameter[9] || logio->io_parameter[10])
fcport->supported_classes |= FC_COS_CLASS3;
goto logio_done;
}
iop[0] = le32_to_cpu(logio->io_parameter[0]);
iop[1] = le32_to_cpu(logio->io_parameter[1]);
switch (iop[0]) {
case LSC_SCODE_PORTID_USED:
data[0] = MBS_PORT_ID_USED;
data[1] = LSW(iop[1]);
break;
case LSC_SCODE_NPORT_USED:
data[0] = MBS_LOOP_ID_USED;
break;
default:
data[0] = MBS_COMMAND_ERROR;
break;
}
ql_dbg(ql_dbg_async, fcport->vha, 0x5037,
"Async-%s failed - hdl=%x portid=%02x%02x%02x comp=%x "
"iop0=%x iop1=%x.\n", type, sp->handle, fcport->d_id.b.domain,
fcport->d_id.b.area, fcport->d_id.b.al_pa,
le16_to_cpu(logio->comp_status),
le32_to_cpu(logio->io_parameter[0]),
le32_to_cpu(logio->io_parameter[1]));
logio_done:
sp->done(vha, sp, 0);
}
static void
qla24xx_tm_iocb_entry(scsi_qla_host_t *vha, struct req_que *req, void *tsk)
{
const char func[] = "TMF-IOCB";
const char *type;
fc_port_t *fcport;
srb_t *sp;
struct srb_iocb *iocb;
struct sts_entry_24xx *sts = (struct sts_entry_24xx *)tsk;
sp = qla2x00_get_sp_from_handle(vha, func, req, tsk);
if (!sp)
return;
iocb = &sp->u.iocb_cmd;
type = sp->name;
fcport = sp->fcport;
iocb->u.tmf.data = QLA_SUCCESS;
if (sts->entry_status) {
ql_log(ql_log_warn, fcport->vha, 0x5038,
"Async-%s error - hdl=%x entry-status(%x).\n",
type, sp->handle, sts->entry_status);
iocb->u.tmf.data = QLA_FUNCTION_FAILED;
} else if (sts->comp_status != __constant_cpu_to_le16(CS_COMPLETE)) {
ql_log(ql_log_warn, fcport->vha, 0x5039,
"Async-%s error - hdl=%x completion status(%x).\n",
type, sp->handle, sts->comp_status);
iocb->u.tmf.data = QLA_FUNCTION_FAILED;
} else if ((le16_to_cpu(sts->scsi_status) &
SS_RESPONSE_INFO_LEN_VALID)) {
if (le32_to_cpu(sts->rsp_data_len) < 4) {
ql_log(ql_log_warn, fcport->vha, 0x503b,
"Async-%s error - hdl=%x not enough response(%d).\n",
type, sp->handle, sts->rsp_data_len);
} else if (sts->data[3]) {
ql_log(ql_log_warn, fcport->vha, 0x503c,
"Async-%s error - hdl=%x response(%x).\n",
type, sp->handle, sts->data[3]);
iocb->u.tmf.data = QLA_FUNCTION_FAILED;
}
}
if (iocb->u.tmf.data != QLA_SUCCESS)
ql_dump_buffer(ql_dbg_async + ql_dbg_buffer, vha, 0x5055,
(uint8_t *)sts, sizeof(*sts));
sp->done(vha, sp, 0);
}
/**
* qla2x00_process_response_queue() - Process response queue entries.
* @ha: SCSI driver HA context
*/
void
qla2x00_process_response_queue(struct rsp_que *rsp)
{
struct scsi_qla_host *vha;
struct qla_hw_data *ha = rsp->hw;
struct device_reg_2xxx __iomem *reg = &ha->iobase->isp;
sts_entry_t *pkt;
uint16_t handle_cnt;
uint16_t cnt;
vha = pci_get_drvdata(ha->pdev);
if (!vha->flags.online)
return;
while (rsp->ring_ptr->signature != RESPONSE_PROCESSED) {
pkt = (sts_entry_t *)rsp->ring_ptr;
rsp->ring_index++;
if (rsp->ring_index == rsp->length) {
rsp->ring_index = 0;
rsp->ring_ptr = rsp->ring;
} else {
rsp->ring_ptr++;
}
if (pkt->entry_status != 0) {
qla2x00_error_entry(vha, rsp, pkt);
((response_t *)pkt)->signature = RESPONSE_PROCESSED;
wmb();
continue;
}
switch (pkt->entry_type) {
case STATUS_TYPE:
qla2x00_status_entry(vha, rsp, pkt);
break;
case STATUS_TYPE_21:
handle_cnt = ((sts21_entry_t *)pkt)->handle_count;
for (cnt = 0; cnt < handle_cnt; cnt++) {
qla2x00_process_completed_request(vha, rsp->req,
((sts21_entry_t *)pkt)->handle[cnt]);
}
break;
case STATUS_TYPE_22:
handle_cnt = ((sts22_entry_t *)pkt)->handle_count;
for (cnt = 0; cnt < handle_cnt; cnt++) {
qla2x00_process_completed_request(vha, rsp->req,
((sts22_entry_t *)pkt)->handle[cnt]);
}
break;
case STATUS_CONT_TYPE:
qla2x00_status_cont_entry(rsp, (sts_cont_entry_t *)pkt);
break;
case MBX_IOCB_TYPE:
qla2x00_mbx_iocb_entry(vha, rsp->req,
(struct mbx_entry *)pkt);
break;
case CT_IOCB_TYPE:
qla2x00_ct_entry(vha, rsp->req, pkt, CT_IOCB_TYPE);
break;
default:
/* Type Not Supported. */
ql_log(ql_log_warn, vha, 0x504a,
"Received unknown response pkt type %x "
"entry status=%x.\n",
pkt->entry_type, pkt->entry_status);
break;
}
((response_t *)pkt)->signature = RESPONSE_PROCESSED;
wmb();
}
/* Adjust ring index */
WRT_REG_WORD(ISP_RSP_Q_OUT(ha, reg), rsp->ring_index);
}
static inline void
qla2x00_handle_sense(srb_t *sp, uint8_t *sense_data, uint32_t par_sense_len,
uint32_t sense_len, struct rsp_que *rsp, int res)
{
struct scsi_qla_host *vha = sp->fcport->vha;
struct scsi_cmnd *cp = GET_CMD_SP(sp);
uint32_t track_sense_len;
if (sense_len >= SCSI_SENSE_BUFFERSIZE)
sense_len = SCSI_SENSE_BUFFERSIZE;
SET_CMD_SENSE_LEN(sp, sense_len);
SET_CMD_SENSE_PTR(sp, cp->sense_buffer);
track_sense_len = sense_len;
if (sense_len > par_sense_len)
sense_len = par_sense_len;
memcpy(cp->sense_buffer, sense_data, sense_len);
SET_CMD_SENSE_PTR(sp, cp->sense_buffer + sense_len);
track_sense_len -= sense_len;
SET_CMD_SENSE_LEN(sp, track_sense_len);
if (track_sense_len != 0) {
rsp->status_srb = sp;
cp->result = res;
}
if (sense_len) {
ql_dbg(ql_dbg_io + ql_dbg_buffer, vha, 0x301c,
"Check condition Sense data, nexus%ld:%d:%llu cmd=%p.\n",
sp->fcport->vha->host_no, cp->device->id, cp->device->lun,
cp);
ql_dump_buffer(ql_dbg_io + ql_dbg_buffer, vha, 0x302b,
cp->sense_buffer, sense_len);
}
}
struct scsi_dif_tuple {
__be16 guard; /* Checksum */
__be16 app_tag; /* APPL identifier */
__be32 ref_tag; /* Target LBA or indirect LBA */
};
/*
* Checks the guard or meta-data for the type of error
* detected by the HBA. In case of errors, we set the
* ASC/ASCQ fields in the sense buffer with ILLEGAL_REQUEST
* to indicate to the kernel that the HBA detected error.
*/
static inline int
qla2x00_handle_dif_error(srb_t *sp, struct sts_entry_24xx *sts24)
{
struct scsi_qla_host *vha = sp->fcport->vha;
struct scsi_cmnd *cmd = GET_CMD_SP(sp);
uint8_t *ap = &sts24->data[12];
uint8_t *ep = &sts24->data[20];
uint32_t e_ref_tag, a_ref_tag;
uint16_t e_app_tag, a_app_tag;
uint16_t e_guard, a_guard;
/*
* swab32 of the "data" field in the beginning of qla2x00_status_entry()
* would make guard field appear at offset 2
*/
a_guard = le16_to_cpu(*(uint16_t *)(ap + 2));
a_app_tag = le16_to_cpu(*(uint16_t *)(ap + 0));
a_ref_tag = le32_to_cpu(*(uint32_t *)(ap + 4));
e_guard = le16_to_cpu(*(uint16_t *)(ep + 2));
e_app_tag = le16_to_cpu(*(uint16_t *)(ep + 0));
e_ref_tag = le32_to_cpu(*(uint32_t *)(ep + 4));
ql_dbg(ql_dbg_io, vha, 0x3023,
"iocb(s) %p Returned STATUS.\n", sts24);
ql_dbg(ql_dbg_io, vha, 0x3024,
"DIF ERROR in cmd 0x%x lba 0x%llx act ref"
" tag=0x%x, exp ref_tag=0x%x, act app tag=0x%x, exp app"
" tag=0x%x, act guard=0x%x, exp guard=0x%x.\n",
cmd->cmnd[0], (u64)scsi_get_lba(cmd), a_ref_tag, e_ref_tag,
a_app_tag, e_app_tag, a_guard, e_guard);
/*
* Ignore sector if:
* For type 3: ref & app tag is all 'f's
* For type 0,1,2: app tag is all 'f's
*/
if ((a_app_tag == 0xffff) &&
((scsi_get_prot_type(cmd) != SCSI_PROT_DIF_TYPE3) ||
(a_ref_tag == 0xffffffff))) {
uint32_t blocks_done, resid;
sector_t lba_s = scsi_get_lba(cmd);
/* 2TB boundary case covered automatically with this */
blocks_done = e_ref_tag - (uint32_t)lba_s + 1;
resid = scsi_bufflen(cmd) - (blocks_done *
cmd->device->sector_size);
scsi_set_resid(cmd, resid);
cmd->result = DID_OK << 16;
/* Update protection tag */
if (scsi_prot_sg_count(cmd)) {
uint32_t i, j = 0, k = 0, num_ent;
struct scatterlist *sg;
struct sd_dif_tuple *spt;
/* Patch the corresponding protection tags */
scsi_for_each_prot_sg(cmd, sg,
scsi_prot_sg_count(cmd), i) {
num_ent = sg_dma_len(sg) / 8;
if (k + num_ent < blocks_done) {
k += num_ent;
continue;
}
j = blocks_done - k - 1;
k = blocks_done;
break;
}
if (k != blocks_done) {
ql_log(ql_log_warn, vha, 0x302f,
"unexpected tag values tag:lba=%x:%llx)\n",
e_ref_tag, (unsigned long long)lba_s);
return 1;
}
spt = page_address(sg_page(sg)) + sg->offset;
spt += j;
spt->app_tag = 0xffff;
if (scsi_get_prot_type(cmd) == SCSI_PROT_DIF_TYPE3)
spt->ref_tag = 0xffffffff;
}
return 0;
}
/* check guard */
if (e_guard != a_guard) {
scsi_build_sense_buffer(1, cmd->sense_buffer, ILLEGAL_REQUEST,
0x10, 0x1);
set_driver_byte(cmd, DRIVER_SENSE);
set_host_byte(cmd, DID_ABORT);
cmd->result |= SAM_STAT_CHECK_CONDITION << 1;
return 1;
}
/* check ref tag */
if (e_ref_tag != a_ref_tag) {
scsi_build_sense_buffer(1, cmd->sense_buffer, ILLEGAL_REQUEST,
0x10, 0x3);
set_driver_byte(cmd, DRIVER_SENSE);
set_host_byte(cmd, DID_ABORT);
cmd->result |= SAM_STAT_CHECK_CONDITION << 1;
return 1;
}
/* check appl tag */
if (e_app_tag != a_app_tag) {
scsi_build_sense_buffer(1, cmd->sense_buffer, ILLEGAL_REQUEST,
0x10, 0x2);
set_driver_byte(cmd, DRIVER_SENSE);
set_host_byte(cmd, DID_ABORT);
cmd->result |= SAM_STAT_CHECK_CONDITION << 1;
return 1;
}
return 1;
}
static void
qla25xx_process_bidir_status_iocb(scsi_qla_host_t *vha, void *pkt,
struct req_que *req, uint32_t index)
{
struct qla_hw_data *ha = vha->hw;
srb_t *sp;
uint16_t comp_status;
uint16_t scsi_status;
uint16_t thread_id;
uint32_t rval = EXT_STATUS_OK;
struct fc_bsg_job *bsg_job = NULL;
sts_entry_t *sts;
struct sts_entry_24xx *sts24;
sts = (sts_entry_t *) pkt;
sts24 = (struct sts_entry_24xx *) pkt;
/* Validate handle. */
if (index >= req->num_outstanding_cmds) {
ql_log(ql_log_warn, vha, 0x70af,
"Invalid SCSI completion handle 0x%x.\n", index);
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
return;
}
sp = req->outstanding_cmds[index];
if (sp) {
/* Free outstanding command slot. */
req->outstanding_cmds[index] = NULL;
bsg_job = sp->u.bsg_job;
} else {
ql_log(ql_log_warn, vha, 0x70b0,
"Req:%d: Invalid ISP SCSI completion handle(0x%x)\n",
req->id, index);
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
return;
}
if (IS_FWI2_CAPABLE(ha)) {
comp_status = le16_to_cpu(sts24->comp_status);
scsi_status = le16_to_cpu(sts24->scsi_status) & SS_MASK;
} else {
comp_status = le16_to_cpu(sts->comp_status);
scsi_status = le16_to_cpu(sts->scsi_status) & SS_MASK;
}
thread_id = bsg_job->request->rqst_data.h_vendor.vendor_cmd[1];
switch (comp_status) {
case CS_COMPLETE:
if (scsi_status == 0) {
bsg_job->reply->reply_payload_rcv_len =
bsg_job->reply_payload.payload_len;
vha->qla_stats.input_bytes +=
bsg_job->reply->reply_payload_rcv_len;
vha->qla_stats.input_requests++;
rval = EXT_STATUS_OK;
}
goto done;
case CS_DATA_OVERRUN:
ql_dbg(ql_dbg_user, vha, 0x70b1,
"Command completed with date overrun thread_id=%d\n",
thread_id);
rval = EXT_STATUS_DATA_OVERRUN;
break;
case CS_DATA_UNDERRUN:
ql_dbg(ql_dbg_user, vha, 0x70b2,
"Command completed with date underrun thread_id=%d\n",
thread_id);
rval = EXT_STATUS_DATA_UNDERRUN;
break;
case CS_BIDIR_RD_OVERRUN:
ql_dbg(ql_dbg_user, vha, 0x70b3,
"Command completed with read data overrun thread_id=%d\n",
thread_id);
rval = EXT_STATUS_DATA_OVERRUN;
break;
case CS_BIDIR_RD_WR_OVERRUN:
ql_dbg(ql_dbg_user, vha, 0x70b4,
"Command completed with read and write data overrun "
"thread_id=%d\n", thread_id);
rval = EXT_STATUS_DATA_OVERRUN;
break;
case CS_BIDIR_RD_OVERRUN_WR_UNDERRUN:
ql_dbg(ql_dbg_user, vha, 0x70b5,
"Command completed with read data over and write data "
"underrun thread_id=%d\n", thread_id);
rval = EXT_STATUS_DATA_OVERRUN;
break;
case CS_BIDIR_RD_UNDERRUN:
ql_dbg(ql_dbg_user, vha, 0x70b6,
"Command completed with read data data underrun "
"thread_id=%d\n", thread_id);
rval = EXT_STATUS_DATA_UNDERRUN;
break;
case CS_BIDIR_RD_UNDERRUN_WR_OVERRUN:
ql_dbg(ql_dbg_user, vha, 0x70b7,
"Command completed with read data under and write data "
"overrun thread_id=%d\n", thread_id);
rval = EXT_STATUS_DATA_UNDERRUN;
break;
case CS_BIDIR_RD_WR_UNDERRUN:
ql_dbg(ql_dbg_user, vha, 0x70b8,
"Command completed with read and write data underrun "
"thread_id=%d\n", thread_id);
rval = EXT_STATUS_DATA_UNDERRUN;
break;
case CS_BIDIR_DMA:
ql_dbg(ql_dbg_user, vha, 0x70b9,
"Command completed with data DMA error thread_id=%d\n",
thread_id);
rval = EXT_STATUS_DMA_ERR;
break;
case CS_TIMEOUT:
ql_dbg(ql_dbg_user, vha, 0x70ba,
"Command completed with timeout thread_id=%d\n",
thread_id);
rval = EXT_STATUS_TIMEOUT;
break;
default:
ql_dbg(ql_dbg_user, vha, 0x70bb,
"Command completed with completion status=0x%x "
"thread_id=%d\n", comp_status, thread_id);
rval = EXT_STATUS_ERR;
break;
}
bsg_job->reply->reply_payload_rcv_len = 0;
done:
/* Return the vendor specific reply to API */
bsg_job->reply->reply_data.vendor_reply.vendor_rsp[0] = rval;
bsg_job->reply_len = sizeof(struct fc_bsg_reply);
/* Always return DID_OK, bsg will send the vendor specific response
* in this case only */
sp->done(vha, sp, (DID_OK << 6));
}
/**
* qla2x00_status_entry() - Process a Status IOCB entry.
* @ha: SCSI driver HA context
* @pkt: Entry pointer
*/
static void
qla2x00_status_entry(scsi_qla_host_t *vha, struct rsp_que *rsp, void *pkt)
{
srb_t *sp;
fc_port_t *fcport;
struct scsi_cmnd *cp;
sts_entry_t *sts;
struct sts_entry_24xx *sts24;
uint16_t comp_status;
uint16_t scsi_status;
uint16_t ox_id;
uint8_t lscsi_status;
int32_t resid;
uint32_t sense_len, par_sense_len, rsp_info_len, resid_len,
fw_resid_len;
uint8_t *rsp_info, *sense_data;
struct qla_hw_data *ha = vha->hw;
uint32_t handle;
uint16_t que;
struct req_que *req;
int logit = 1;
int res = 0;
uint16_t state_flags = 0;
uint16_t retry_delay = 0;
sts = (sts_entry_t *) pkt;
sts24 = (struct sts_entry_24xx *) pkt;
if (IS_FWI2_CAPABLE(ha)) {
comp_status = le16_to_cpu(sts24->comp_status);
scsi_status = le16_to_cpu(sts24->scsi_status) & SS_MASK;
state_flags = le16_to_cpu(sts24->state_flags);
} else {
comp_status = le16_to_cpu(sts->comp_status);
scsi_status = le16_to_cpu(sts->scsi_status) & SS_MASK;
}
handle = (uint32_t) LSW(sts->handle);
que = MSW(sts->handle);
req = ha->req_q_map[que];
/* Check for invalid queue pointer */
if (req == NULL ||
que >= find_first_zero_bit(ha->req_qid_map, ha->max_req_queues)) {
ql_dbg(ql_dbg_io, vha, 0x3059,
"Invalid status handle (0x%x): Bad req pointer. req=%p, "
"que=%u.\n", sts->handle, req, que);
return;
}
/* Validate handle. */
if (handle < req->num_outstanding_cmds)
sp = req->outstanding_cmds[handle];
else
sp = NULL;
if (sp == NULL) {
ql_dbg(ql_dbg_io, vha, 0x3017,
"Invalid status handle (0x%x).\n", sts->handle);
if (!test_bit(ABORT_ISP_ACTIVE, &vha->dpc_flags)) {
if (IS_P3P_TYPE(ha))
set_bit(FCOE_CTX_RESET_NEEDED, &vha->dpc_flags);
else
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
qla2xxx_wake_dpc(vha);
}
return;
}
if (unlikely((state_flags & BIT_1) && (sp->type == SRB_BIDI_CMD))) {
qla25xx_process_bidir_status_iocb(vha, pkt, req, handle);
return;
}
/* Task Management completion. */
if (sp->type == SRB_TM_CMD) {
qla24xx_tm_iocb_entry(vha, req, pkt);
return;
}
/* Fast path completion. */
if (comp_status == CS_COMPLETE && scsi_status == 0) {
qla2x00_process_completed_request(vha, req, handle);
return;
}
req->outstanding_cmds[handle] = NULL;
cp = GET_CMD_SP(sp);
if (cp == NULL) {
ql_dbg(ql_dbg_io, vha, 0x3018,
"Command already returned (0x%x/%p).\n",
sts->handle, sp);
return;
}
lscsi_status = scsi_status & STATUS_MASK;
fcport = sp->fcport;
ox_id = 0;
sense_len = par_sense_len = rsp_info_len = resid_len =
fw_resid_len = 0;
if (IS_FWI2_CAPABLE(ha)) {
if (scsi_status & SS_SENSE_LEN_VALID)
sense_len = le32_to_cpu(sts24->sense_len);
if (scsi_status & SS_RESPONSE_INFO_LEN_VALID)
rsp_info_len = le32_to_cpu(sts24->rsp_data_len);
if (scsi_status & (SS_RESIDUAL_UNDER | SS_RESIDUAL_OVER))
resid_len = le32_to_cpu(sts24->rsp_residual_count);
if (comp_status == CS_DATA_UNDERRUN)
fw_resid_len = le32_to_cpu(sts24->residual_len);
rsp_info = sts24->data;
sense_data = sts24->data;
host_to_fcp_swap(sts24->data, sizeof(sts24->data));
ox_id = le16_to_cpu(sts24->ox_id);
par_sense_len = sizeof(sts24->data);
/* Valid values of the retry delay timer are 0x1-0xffef */
if (sts24->retry_delay > 0 && sts24->retry_delay < 0xfff1)
retry_delay = sts24->retry_delay;
} else {
if (scsi_status & SS_SENSE_LEN_VALID)
sense_len = le16_to_cpu(sts->req_sense_length);
if (scsi_status & SS_RESPONSE_INFO_LEN_VALID)
rsp_info_len = le16_to_cpu(sts->rsp_info_len);
resid_len = le32_to_cpu(sts->residual_length);
rsp_info = sts->rsp_info;
sense_data = sts->req_sense_data;
par_sense_len = sizeof(sts->req_sense_data);
}
/* Check for any FCP transport errors. */
if (scsi_status & SS_RESPONSE_INFO_LEN_VALID) {
/* Sense data lies beyond any FCP RESPONSE data. */
if (IS_FWI2_CAPABLE(ha)) {
sense_data += rsp_info_len;
par_sense_len -= rsp_info_len;
}
if (rsp_info_len > 3 && rsp_info[3]) {
ql_dbg(ql_dbg_io, fcport->vha, 0x3019,
"FCP I/O protocol failure (0x%x/0x%x).\n",
rsp_info_len, rsp_info[3]);
res = DID_BUS_BUSY << 16;
goto out;
}
}
/* Check for overrun. */
if (IS_FWI2_CAPABLE(ha) && comp_status == CS_COMPLETE &&
scsi_status & SS_RESIDUAL_OVER)
comp_status = CS_DATA_OVERRUN;
/*
* Check retry_delay_timer value if we receive a busy or
* queue full.
*/
if (lscsi_status == SAM_STAT_TASK_SET_FULL ||
lscsi_status == SAM_STAT_BUSY)
qla2x00_set_retry_delay_timestamp(fcport, retry_delay);
/*
* Based on Host and scsi status generate status code for Linux
*/
switch (comp_status) {
case CS_COMPLETE:
case CS_QUEUE_FULL:
if (scsi_status == 0) {
res = DID_OK << 16;
break;
}
if (scsi_status & (SS_RESIDUAL_UNDER | SS_RESIDUAL_OVER)) {
resid = resid_len;
scsi_set_resid(cp, resid);
if (!lscsi_status &&
((unsigned)(scsi_bufflen(cp) - resid) <
cp->underflow)) {
ql_dbg(ql_dbg_io, fcport->vha, 0x301a,
"Mid-layer underflow "
"detected (0x%x of 0x%x bytes).\n",
resid, scsi_bufflen(cp));
res = DID_ERROR << 16;
break;
}
}
res = DID_OK << 16 | lscsi_status;
if (lscsi_status == SAM_STAT_TASK_SET_FULL) {
ql_dbg(ql_dbg_io, fcport->vha, 0x301b,
"QUEUE FULL detected.\n");
break;
}
logit = 0;
if (lscsi_status != SS_CHECK_CONDITION)
break;
memset(cp->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE);
if (!(scsi_status & SS_SENSE_LEN_VALID))
break;
qla2x00_handle_sense(sp, sense_data, par_sense_len, sense_len,
rsp, res);
break;
case CS_DATA_UNDERRUN:
/* Use F/W calculated residual length. */
resid = IS_FWI2_CAPABLE(ha) ? fw_resid_len : resid_len;
scsi_set_resid(cp, resid);
if (scsi_status & SS_RESIDUAL_UNDER) {
if (IS_FWI2_CAPABLE(ha) && fw_resid_len != resid_len) {
ql_dbg(ql_dbg_io, fcport->vha, 0x301d,
"Dropped frame(s) detected "
"(0x%x of 0x%x bytes).\n",
resid, scsi_bufflen(cp));
res = DID_ERROR << 16 | lscsi_status;
goto check_scsi_status;
}
if (!lscsi_status &&
((unsigned)(scsi_bufflen(cp) - resid) <
cp->underflow)) {
ql_dbg(ql_dbg_io, fcport->vha, 0x301e,
"Mid-layer underflow "
"detected (0x%x of 0x%x bytes).\n",
resid, scsi_bufflen(cp));
res = DID_ERROR << 16;
break;
}
} else if (lscsi_status != SAM_STAT_TASK_SET_FULL &&
lscsi_status != SAM_STAT_BUSY) {
/*
* scsi status of task set and busy are considered to be
* task not completed.
*/
ql_dbg(ql_dbg_io, fcport->vha, 0x301f,
"Dropped frame(s) detected (0x%x "
"of 0x%x bytes).\n", resid,
scsi_bufflen(cp));
res = DID_ERROR << 16 | lscsi_status;
goto check_scsi_status;
} else {
ql_dbg(ql_dbg_io, fcport->vha, 0x3030,
"scsi_status: 0x%x, lscsi_status: 0x%x\n",
scsi_status, lscsi_status);
}
res = DID_OK << 16 | lscsi_status;
logit = 0;
check_scsi_status:
/*
* Check to see if SCSI Status is non zero. If so report SCSI
* Status.
*/
if (lscsi_status != 0) {
if (lscsi_status == SAM_STAT_TASK_SET_FULL) {
ql_dbg(ql_dbg_io, fcport->vha, 0x3020,
"QUEUE FULL detected.\n");
logit = 1;
break;
}
if (lscsi_status != SS_CHECK_CONDITION)
break;
memset(cp->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE);
if (!(scsi_status & SS_SENSE_LEN_VALID))
break;
qla2x00_handle_sense(sp, sense_data, par_sense_len,
sense_len, rsp, res);
}
break;
case CS_PORT_LOGGED_OUT:
case CS_PORT_CONFIG_CHG:
case CS_PORT_BUSY:
case CS_INCOMPLETE:
case CS_PORT_UNAVAILABLE:
case CS_TIMEOUT:
case CS_RESET:
/*
* We are going to have the fc class block the rport
* while we try to recover so instruct the mid layer
* to requeue until the class decides how to handle this.
*/
res = DID_TRANSPORT_DISRUPTED << 16;
if (comp_status == CS_TIMEOUT) {
if (IS_FWI2_CAPABLE(ha))
break;
else if ((le16_to_cpu(sts->status_flags) &
SF_LOGOUT_SENT) == 0)
break;
}
ql_dbg(ql_dbg_io, fcport->vha, 0x3021,
"Port to be marked lost on fcport=%02x%02x%02x, current "
"port state= %s.\n", fcport->d_id.b.domain,
fcport->d_id.b.area, fcport->d_id.b.al_pa,
port_state_str[atomic_read(&fcport->state)]);
if (atomic_read(&fcport->state) == FCS_ONLINE)
qla2x00_mark_device_lost(fcport->vha, fcport, 1, 1);
break;
case CS_ABORTED:
res = DID_RESET << 16;
break;
case CS_DIF_ERROR:
logit = qla2x00_handle_dif_error(sp, sts24);
res = cp->result;
break;
case CS_TRANSPORT:
res = DID_ERROR << 16;
if (!IS_PI_SPLIT_DET_CAPABLE(ha))
break;
if (state_flags & BIT_4)
scmd_printk(KERN_WARNING, cp,
"Unsupported device '%s' found.\n",
cp->device->vendor);
break;
default:
res = DID_ERROR << 16;
break;
}
out:
if (logit)
ql_dbg(ql_dbg_io, fcport->vha, 0x3022,
"FCP command status: 0x%x-0x%x (0x%x) nexus=%ld:%d:%llu "
"portid=%02x%02x%02x oxid=0x%x cdb=%10phN len=0x%x "
"rsp_info=0x%x resid=0x%x fw_resid=0x%x.\n",
comp_status, scsi_status, res, vha->host_no,
cp->device->id, cp->device->lun, fcport->d_id.b.domain,
fcport->d_id.b.area, fcport->d_id.b.al_pa, ox_id,
cp->cmnd, scsi_bufflen(cp), rsp_info_len,
resid_len, fw_resid_len);
if (rsp->status_srb == NULL)
sp->done(ha, sp, res);
}
/**
* qla2x00_status_cont_entry() - Process a Status Continuations entry.
* @ha: SCSI driver HA context
* @pkt: Entry pointer
*
* Extended sense data.
*/
static void
qla2x00_status_cont_entry(struct rsp_que *rsp, sts_cont_entry_t *pkt)
{
uint8_t sense_sz = 0;
struct qla_hw_data *ha = rsp->hw;
struct scsi_qla_host *vha = pci_get_drvdata(ha->pdev);
srb_t *sp = rsp->status_srb;
struct scsi_cmnd *cp;
uint32_t sense_len;
uint8_t *sense_ptr;
if (!sp || !GET_CMD_SENSE_LEN(sp))
return;
sense_len = GET_CMD_SENSE_LEN(sp);
sense_ptr = GET_CMD_SENSE_PTR(sp);
cp = GET_CMD_SP(sp);
if (cp == NULL) {
ql_log(ql_log_warn, vha, 0x3025,
"cmd is NULL: already returned to OS (sp=%p).\n", sp);
rsp->status_srb = NULL;
return;
}
if (sense_len > sizeof(pkt->data))
sense_sz = sizeof(pkt->data);
else
sense_sz = sense_len;
/* Move sense data. */
if (IS_FWI2_CAPABLE(ha))
host_to_fcp_swap(pkt->data, sizeof(pkt->data));
memcpy(sense_ptr, pkt->data, sense_sz);
ql_dump_buffer(ql_dbg_io + ql_dbg_buffer, vha, 0x302c,
sense_ptr, sense_sz);
sense_len -= sense_sz;
sense_ptr += sense_sz;
SET_CMD_SENSE_PTR(sp, sense_ptr);
SET_CMD_SENSE_LEN(sp, sense_len);
/* Place command on done queue. */
if (sense_len == 0) {
rsp->status_srb = NULL;
sp->done(ha, sp, cp->result);
}
}
/**
* qla2x00_error_entry() - Process an error entry.
* @ha: SCSI driver HA context
* @pkt: Entry pointer
*/
static void
qla2x00_error_entry(scsi_qla_host_t *vha, struct rsp_que *rsp, sts_entry_t *pkt)
{
srb_t *sp;
struct qla_hw_data *ha = vha->hw;
const char func[] = "ERROR-IOCB";
uint16_t que = MSW(pkt->handle);
struct req_que *req = NULL;
int res = DID_ERROR << 16;
ql_dbg(ql_dbg_async, vha, 0x502a,
"type of error status in response: 0x%x\n", pkt->entry_status);
if (que >= ha->max_req_queues || !ha->req_q_map[que])
goto fatal;
req = ha->req_q_map[que];
if (pkt->entry_status & RF_BUSY)
res = DID_BUS_BUSY << 16;
sp = qla2x00_get_sp_from_handle(vha, func, req, pkt);
if (sp) {
sp->done(ha, sp, res);
return;
}
fatal:
ql_log(ql_log_warn, vha, 0x5030,
"Error entry - invalid handle/queue.\n");
if (IS_P3P_TYPE(ha))
set_bit(FCOE_CTX_RESET_NEEDED, &vha->dpc_flags);
else
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
qla2xxx_wake_dpc(vha);
}
/**
* qla24xx_mbx_completion() - Process mailbox command completions.
* @ha: SCSI driver HA context
* @mb0: Mailbox0 register
*/
static void
qla24xx_mbx_completion(scsi_qla_host_t *vha, uint16_t mb0)
{
uint16_t cnt;
uint32_t mboxes;
uint16_t __iomem *wptr;
struct qla_hw_data *ha = vha->hw;
struct device_reg_24xx __iomem *reg = &ha->iobase->isp24;
/* Read all mbox registers? */
mboxes = (1 << ha->mbx_count) - 1;
if (!ha->mcp)
ql_dbg(ql_dbg_async, vha, 0x504e, "MBX pointer ERROR.\n");
else
mboxes = ha->mcp->in_mb;
/* Load return mailbox registers. */
ha->flags.mbox_int = 1;
ha->mailbox_out[0] = mb0;
mboxes >>= 1;
wptr = (uint16_t __iomem *)&reg->mailbox1;
for (cnt = 1; cnt < ha->mbx_count; cnt++) {
if (mboxes & BIT_0)
ha->mailbox_out[cnt] = RD_REG_WORD(wptr);
mboxes >>= 1;
wptr++;
}
}
static void
qla24xx_abort_iocb_entry(scsi_qla_host_t *vha, struct req_que *req,
struct abort_entry_24xx *pkt)
{
const char func[] = "ABT_IOCB";
srb_t *sp;
struct srb_iocb *abt;
sp = qla2x00_get_sp_from_handle(vha, func, req, pkt);
if (!sp)
return;
abt = &sp->u.iocb_cmd;
abt->u.abt.comp_status = le32_to_cpu(pkt->nport_handle);
sp->done(vha, sp, 0);
}
/**
* qla24xx_process_response_queue() - Process response queue entries.
* @ha: SCSI driver HA context
*/
void qla24xx_process_response_queue(struct scsi_qla_host *vha,
struct rsp_que *rsp)
{
struct sts_entry_24xx *pkt;
struct qla_hw_data *ha = vha->hw;
if (!vha->flags.online)
return;
while (rsp->ring_ptr->signature != RESPONSE_PROCESSED) {
pkt = (struct sts_entry_24xx *)rsp->ring_ptr;
rsp->ring_index++;
if (rsp->ring_index == rsp->length) {
rsp->ring_index = 0;
rsp->ring_ptr = rsp->ring;
} else {
rsp->ring_ptr++;
}
if (pkt->entry_status != 0) {
qla2x00_error_entry(vha, rsp, (sts_entry_t *) pkt);
if (qlt_24xx_process_response_error(vha, pkt))
goto process_err;
((response_t *)pkt)->signature = RESPONSE_PROCESSED;
wmb();
continue;
}
process_err:
switch (pkt->entry_type) {
case STATUS_TYPE:
qla2x00_status_entry(vha, rsp, pkt);
break;
case STATUS_CONT_TYPE:
qla2x00_status_cont_entry(rsp, (sts_cont_entry_t *)pkt);
break;
case VP_RPT_ID_IOCB_TYPE:
qla24xx_report_id_acquisition(vha,
(struct vp_rpt_id_entry_24xx *)pkt);
break;
case LOGINOUT_PORT_IOCB_TYPE:
qla24xx_logio_entry(vha, rsp->req,
(struct logio_entry_24xx *)pkt);
break;
case CT_IOCB_TYPE:
qla24xx_els_ct_entry(vha, rsp->req, pkt, CT_IOCB_TYPE);
break;
case ELS_IOCB_TYPE:
qla24xx_els_ct_entry(vha, rsp->req, pkt, ELS_IOCB_TYPE);
break;
case ABTS_RECV_24XX:
/* ensure that the ATIO queue is empty */
qlt_24xx_process_atio_queue(vha);
case ABTS_RESP_24XX:
case CTIO_TYPE7:
case NOTIFY_ACK_TYPE:
case CTIO_CRC2:
qlt_response_pkt_all_vps(vha, (response_t *)pkt);
break;
case MARKER_TYPE:
/* Do nothing in this case, this check is to prevent it
* from falling into default case
*/
break;
case ABORT_IOCB_TYPE:
qla24xx_abort_iocb_entry(vha, rsp->req,
(struct abort_entry_24xx *)pkt);
break;
default:
/* Type Not Supported. */
ql_dbg(ql_dbg_async, vha, 0x5042,
"Received unknown response pkt type %x "
"entry status=%x.\n",
pkt->entry_type, pkt->entry_status);
break;
}
((response_t *)pkt)->signature = RESPONSE_PROCESSED;
wmb();
}
/* Adjust ring index */
if (IS_P3P_TYPE(ha)) {
struct device_reg_82xx __iomem *reg = &ha->iobase->isp82;
WRT_REG_DWORD(&reg->rsp_q_out[0], rsp->ring_index);
} else
WRT_REG_DWORD(rsp->rsp_q_out, rsp->ring_index);
}
static void
qla2xxx_check_risc_status(scsi_qla_host_t *vha)
{
int rval;
uint32_t cnt;
struct qla_hw_data *ha = vha->hw;
struct device_reg_24xx __iomem *reg = &ha->iobase->isp24;
if (!IS_QLA25XX(ha) && !IS_QLA81XX(ha) && !IS_QLA83XX(ha) &&
!IS_QLA27XX(ha))
return;
rval = QLA_SUCCESS;
WRT_REG_DWORD(&reg->iobase_addr, 0x7C00);
RD_REG_DWORD(&reg->iobase_addr);
WRT_REG_DWORD(&reg->iobase_window, 0x0001);
for (cnt = 10000; (RD_REG_DWORD(&reg->iobase_window) & BIT_0) == 0 &&
rval == QLA_SUCCESS; cnt--) {
if (cnt) {
WRT_REG_DWORD(&reg->iobase_window, 0x0001);
udelay(10);
} else
rval = QLA_FUNCTION_TIMEOUT;
}
if (rval == QLA_SUCCESS)
goto next_test;
rval = QLA_SUCCESS;
WRT_REG_DWORD(&reg->iobase_window, 0x0003);
for (cnt = 100; (RD_REG_DWORD(&reg->iobase_window) & BIT_0) == 0 &&
rval == QLA_SUCCESS; cnt--) {
if (cnt) {
WRT_REG_DWORD(&reg->iobase_window, 0x0003);
udelay(10);
} else
rval = QLA_FUNCTION_TIMEOUT;
}
if (rval != QLA_SUCCESS)
goto done;
next_test:
if (RD_REG_DWORD(&reg->iobase_c8) & BIT_3)
ql_log(ql_log_info, vha, 0x504c,
"Additional code -- 0x55AA.\n");
done:
WRT_REG_DWORD(&reg->iobase_window, 0x0000);
RD_REG_DWORD(&reg->iobase_window);
}
/**
* qla24xx_intr_handler() - Process interrupts for the ISP23xx and ISP24xx.
* @irq:
* @dev_id: SCSI driver HA context
*
* Called by system whenever the host adapter generates an interrupt.
*
* Returns handled flag.
*/
irqreturn_t
qla24xx_intr_handler(int irq, void *dev_id)
{
scsi_qla_host_t *vha;
struct qla_hw_data *ha;
struct device_reg_24xx __iomem *reg;
int status;
unsigned long iter;
uint32_t stat;
uint32_t hccr;
uint16_t mb[8];
struct rsp_que *rsp;
unsigned long flags;
rsp = (struct rsp_que *) dev_id;
if (!rsp) {
ql_log(ql_log_info, NULL, 0x5059,
"%s: NULL response queue pointer.\n", __func__);
return IRQ_NONE;
}
ha = rsp->hw;
reg = &ha->iobase->isp24;
status = 0;
if (unlikely(pci_channel_offline(ha->pdev)))
return IRQ_HANDLED;
spin_lock_irqsave(&ha->hardware_lock, flags);
vha = pci_get_drvdata(ha->pdev);
for (iter = 50; iter--; ) {
stat = RD_REG_DWORD(&reg->host_status);
if (qla2x00_check_reg32_for_disconnect(vha, stat))
break;
if (stat & HSRX_RISC_PAUSED) {
if (unlikely(pci_channel_offline(ha->pdev)))
break;
hccr = RD_REG_DWORD(&reg->hccr);
ql_log(ql_log_warn, vha, 0x504b,
"RISC paused -- HCCR=%x, Dumping firmware.\n",
hccr);
qla2xxx_check_risc_status(vha);
ha->isp_ops->fw_dump(vha, 1);
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
break;
} else if ((stat & HSRX_RISC_INT) == 0)
break;
switch (stat & 0xff) {
case INTR_ROM_MB_SUCCESS:
case INTR_ROM_MB_FAILED:
case INTR_MB_SUCCESS:
case INTR_MB_FAILED:
qla24xx_mbx_completion(vha, MSW(stat));
status |= MBX_INTERRUPT;
break;
case INTR_ASYNC_EVENT:
mb[0] = MSW(stat);
mb[1] = RD_REG_WORD(&reg->mailbox1);
mb[2] = RD_REG_WORD(&reg->mailbox2);
mb[3] = RD_REG_WORD(&reg->mailbox3);
qla2x00_async_event(vha, rsp, mb);
break;
case INTR_RSP_QUE_UPDATE:
case INTR_RSP_QUE_UPDATE_83XX:
qla24xx_process_response_queue(vha, rsp);
break;
case INTR_ATIO_QUE_UPDATE:
qlt_24xx_process_atio_queue(vha);
break;
case INTR_ATIO_RSP_QUE_UPDATE:
qlt_24xx_process_atio_queue(vha);
qla24xx_process_response_queue(vha, rsp);
break;
default:
ql_dbg(ql_dbg_async, vha, 0x504f,
"Unrecognized interrupt type (%d).\n", stat * 0xff);
break;
}
WRT_REG_DWORD(&reg->hccr, HCCRX_CLR_RISC_INT);
RD_REG_DWORD_RELAXED(&reg->hccr);
if (unlikely(IS_QLA83XX(ha) && (ha->pdev->revision == 1)))
ndelay(3500);
}
qla2x00_handle_mbx_completion(ha, status);
spin_unlock_irqrestore(&ha->hardware_lock, flags);
return IRQ_HANDLED;
}
static irqreturn_t
qla24xx_msix_rsp_q(int irq, void *dev_id)
{
struct qla_hw_data *ha;
struct rsp_que *rsp;
struct device_reg_24xx __iomem *reg;
struct scsi_qla_host *vha;
unsigned long flags;
uint32_t stat = 0;
rsp = (struct rsp_que *) dev_id;
if (!rsp) {
ql_log(ql_log_info, NULL, 0x505a,
"%s: NULL response queue pointer.\n", __func__);
return IRQ_NONE;
}
ha = rsp->hw;
reg = &ha->iobase->isp24;
spin_lock_irqsave(&ha->hardware_lock, flags);
vha = pci_get_drvdata(ha->pdev);
/*
* Use host_status register to check to PCI disconnection before we
* we process the response queue.
*/
stat = RD_REG_DWORD(&reg->host_status);
if (qla2x00_check_reg32_for_disconnect(vha, stat))
goto out;
qla24xx_process_response_queue(vha, rsp);
if (!ha->flags.disable_msix_handshake) {
WRT_REG_DWORD(&reg->hccr, HCCRX_CLR_RISC_INT);
RD_REG_DWORD_RELAXED(&reg->hccr);
}
out:
spin_unlock_irqrestore(&ha->hardware_lock, flags);
return IRQ_HANDLED;
}
static irqreturn_t
qla25xx_msix_rsp_q(int irq, void *dev_id)
{
struct qla_hw_data *ha;
scsi_qla_host_t *vha;
struct rsp_que *rsp;
struct device_reg_24xx __iomem *reg;
unsigned long flags;
uint32_t hccr = 0;
rsp = (struct rsp_que *) dev_id;
if (!rsp) {
ql_log(ql_log_info, NULL, 0x505b,
"%s: NULL response queue pointer.\n", __func__);
return IRQ_NONE;
}
ha = rsp->hw;
vha = pci_get_drvdata(ha->pdev);
/* Clear the interrupt, if enabled, for this response queue */
if (!ha->flags.disable_msix_handshake) {
reg = &ha->iobase->isp24;
spin_lock_irqsave(&ha->hardware_lock, flags);
WRT_REG_DWORD(&reg->hccr, HCCRX_CLR_RISC_INT);
hccr = RD_REG_DWORD_RELAXED(&reg->hccr);
spin_unlock_irqrestore(&ha->hardware_lock, flags);
}
if (qla2x00_check_reg32_for_disconnect(vha, hccr))
goto out;
queue_work_on((int) (rsp->id - 1), ha->wq, &rsp->q_work);
out:
return IRQ_HANDLED;
}
static irqreturn_t
qla24xx_msix_default(int irq, void *dev_id)
{
scsi_qla_host_t *vha;
struct qla_hw_data *ha;
struct rsp_que *rsp;
struct device_reg_24xx __iomem *reg;
int status;
uint32_t stat;
uint32_t hccr;
uint16_t mb[8];
unsigned long flags;
rsp = (struct rsp_que *) dev_id;
if (!rsp) {
ql_log(ql_log_info, NULL, 0x505c,
"%s: NULL response queue pointer.\n", __func__);
return IRQ_NONE;
}
ha = rsp->hw;
reg = &ha->iobase->isp24;
status = 0;
spin_lock_irqsave(&ha->hardware_lock, flags);
vha = pci_get_drvdata(ha->pdev);
do {
stat = RD_REG_DWORD(&reg->host_status);
if (qla2x00_check_reg32_for_disconnect(vha, stat))
break;
if (stat & HSRX_RISC_PAUSED) {
if (unlikely(pci_channel_offline(ha->pdev)))
break;
hccr = RD_REG_DWORD(&reg->hccr);
ql_log(ql_log_info, vha, 0x5050,
"RISC paused -- HCCR=%x, Dumping firmware.\n",
hccr);
qla2xxx_check_risc_status(vha);
ha->isp_ops->fw_dump(vha, 1);
set_bit(ISP_ABORT_NEEDED, &vha->dpc_flags);
break;
} else if ((stat & HSRX_RISC_INT) == 0)
break;
switch (stat & 0xff) {
case INTR_ROM_MB_SUCCESS:
case INTR_ROM_MB_FAILED:
case INTR_MB_SUCCESS:
case INTR_MB_FAILED:
qla24xx_mbx_completion(vha, MSW(stat));
status |= MBX_INTERRUPT;
break;
case INTR_ASYNC_EVENT:
mb[0] = MSW(stat);
mb[1] = RD_REG_WORD(&reg->mailbox1);
mb[2] = RD_REG_WORD(&reg->mailbox2);
mb[3] = RD_REG_WORD(&reg->mailbox3);
qla2x00_async_event(vha, rsp, mb);
break;
case INTR_RSP_QUE_UPDATE:
case INTR_RSP_QUE_UPDATE_83XX:
qla24xx_process_response_queue(vha, rsp);
break;
case INTR_ATIO_QUE_UPDATE:
qlt_24xx_process_atio_queue(vha);
break;
case INTR_ATIO_RSP_QUE_UPDATE:
qlt_24xx_process_atio_queue(vha);
qla24xx_process_response_queue(vha, rsp);
break;
default:
ql_dbg(ql_dbg_async, vha, 0x5051,
"Unrecognized interrupt type (%d).\n", stat & 0xff);
break;
}
WRT_REG_DWORD(&reg->hccr, HCCRX_CLR_RISC_INT);
} while (0);
qla2x00_handle_mbx_completion(ha, status);
spin_unlock_irqrestore(&ha->hardware_lock, flags);
return IRQ_HANDLED;
}
/* Interrupt handling helpers. */
struct qla_init_msix_entry {
const char *name;
irq_handler_t handler;
};
static struct qla_init_msix_entry msix_entries[3] = {
{ "qla2xxx (default)", qla24xx_msix_default },
{ "qla2xxx (rsp_q)", qla24xx_msix_rsp_q },
{ "qla2xxx (multiq)", qla25xx_msix_rsp_q },
};
static struct qla_init_msix_entry qla82xx_msix_entries[2] = {
{ "qla2xxx (default)", qla82xx_msix_default },
{ "qla2xxx (rsp_q)", qla82xx_msix_rsp_q },
};
static struct qla_init_msix_entry qla83xx_msix_entries[3] = {
{ "qla2xxx (default)", qla24xx_msix_default },
{ "qla2xxx (rsp_q)", qla24xx_msix_rsp_q },
{ "qla2xxx (atio_q)", qla83xx_msix_atio_q },
};
static void
qla24xx_disable_msix(struct qla_hw_data *ha)
{
int i;
struct qla_msix_entry *qentry;
scsi_qla_host_t *vha = pci_get_drvdata(ha->pdev);
for (i = 0; i < ha->msix_count; i++) {
qentry = &ha->msix_entries[i];
if (qentry->have_irq)
free_irq(qentry->vector, qentry->rsp);
}
pci_disable_msix(ha->pdev);
kfree(ha->msix_entries);
ha->msix_entries = NULL;
ha->flags.msix_enabled = 0;
ql_dbg(ql_dbg_init, vha, 0x0042,
"Disabled the MSI.\n");
}
static int
qla24xx_enable_msix(struct qla_hw_data *ha, struct rsp_que *rsp)
{
#define MIN_MSIX_COUNT 2
#define ATIO_VECTOR 2
int i, ret;
struct msix_entry *entries;
struct qla_msix_entry *qentry;
scsi_qla_host_t *vha = pci_get_drvdata(ha->pdev);
entries = kzalloc(sizeof(struct msix_entry) * ha->msix_count,
GFP_KERNEL);
if (!entries) {
ql_log(ql_log_warn, vha, 0x00bc,
"Failed to allocate memory for msix_entry.\n");
return -ENOMEM;
}
for (i = 0; i < ha->msix_count; i++)
entries[i].entry = i;
ret = pci_enable_msix_range(ha->pdev,
entries, MIN_MSIX_COUNT, ha->msix_count);
if (ret < 0) {
ql_log(ql_log_fatal, vha, 0x00c7,
"MSI-X: Failed to enable support, "
"giving up -- %d/%d.\n",
ha->msix_count, ret);
goto msix_out;
} else if (ret < ha->msix_count) {
ql_log(ql_log_warn, vha, 0x00c6,
"MSI-X: Failed to enable support "
"-- %d/%d\n Retry with %d vectors.\n",
ha->msix_count, ret, ret);
}
ha->msix_count = ret;
ha->max_rsp_queues = ha->msix_count - 1;
ha->msix_entries = kzalloc(sizeof(struct qla_msix_entry) *
ha->msix_count, GFP_KERNEL);
if (!ha->msix_entries) {
ql_log(ql_log_fatal, vha, 0x00c8,
"Failed to allocate memory for ha->msix_entries.\n");
ret = -ENOMEM;
goto msix_out;
}
ha->flags.msix_enabled = 1;
for (i = 0; i < ha->msix_count; i++) {
qentry = &ha->msix_entries[i];
qentry->vector = entries[i].vector;
qentry->entry = entries[i].entry;
qentry->have_irq = 0;
qentry->rsp = NULL;
}
/* Enable MSI-X vectors for the base queue */
for (i = 0; i < 2; i++) {
qentry = &ha->msix_entries[i];
if (IS_P3P_TYPE(ha))
ret = request_irq(qentry->vector,
qla82xx_msix_entries[i].handler,
0, qla82xx_msix_entries[i].name, rsp);
else
ret = request_irq(qentry->vector,
msix_entries[i].handler,
0, msix_entries[i].name, rsp);
if (ret)
goto msix_register_fail;
qentry->have_irq = 1;
qentry->rsp = rsp;
rsp->msix = qentry;
}
/*
* If target mode is enable, also request the vector for the ATIO
* queue.
*/
if (QLA_TGT_MODE_ENABLED() && IS_ATIO_MSIX_CAPABLE(ha)) {
qentry = &ha->msix_entries[ATIO_VECTOR];
ret = request_irq(qentry->vector,
qla83xx_msix_entries[ATIO_VECTOR].handler,
0, qla83xx_msix_entries[ATIO_VECTOR].name, rsp);
qentry->have_irq = 1;
qentry->rsp = rsp;
rsp->msix = qentry;
}
msix_register_fail:
if (ret) {
ql_log(ql_log_fatal, vha, 0x00cb,
"MSI-X: unable to register handler -- %x/%d.\n",
qentry->vector, ret);
qla24xx_disable_msix(ha);
ha->mqenable = 0;
goto msix_out;
}
/* Enable MSI-X vector for response queue update for queue 0 */
if (IS_QLA83XX(ha) || IS_QLA27XX(ha)) {
if (ha->msixbase && ha->mqiobase &&
(ha->max_rsp_queues > 1 || ha->max_req_queues > 1))
ha->mqenable = 1;
} else
if (ha->mqiobase
&& (ha->max_rsp_queues > 1 || ha->max_req_queues > 1))
ha->mqenable = 1;
ql_dbg(ql_dbg_multiq, vha, 0xc005,
"mqiobase=%p, max_rsp_queues=%d, max_req_queues=%d.\n",
ha->mqiobase, ha->max_rsp_queues, ha->max_req_queues);
ql_dbg(ql_dbg_init, vha, 0x0055,
"mqiobase=%p, max_rsp_queues=%d, max_req_queues=%d.\n",
ha->mqiobase, ha->max_rsp_queues, ha->max_req_queues);
msix_out:
kfree(entries);
return ret;
}
int
qla2x00_request_irqs(struct qla_hw_data *ha, struct rsp_que *rsp)
{
int ret = QLA_FUNCTION_FAILED;
device_reg_t *reg = ha->iobase;
scsi_qla_host_t *vha = pci_get_drvdata(ha->pdev);
/* If possible, enable MSI-X. */
if (!IS_QLA2432(ha) && !IS_QLA2532(ha) && !IS_QLA8432(ha) &&
!IS_CNA_CAPABLE(ha) && !IS_QLA2031(ha) && !IS_QLAFX00(ha) &&
!IS_QLA27XX(ha))
goto skip_msi;
if (ha->pdev->subsystem_vendor == PCI_VENDOR_ID_HP &&
(ha->pdev->subsystem_device == 0x7040 ||
ha->pdev->subsystem_device == 0x7041 ||
ha->pdev->subsystem_device == 0x1705)) {
ql_log(ql_log_warn, vha, 0x0034,
"MSI-X: Unsupported ISP 2432 SSVID/SSDID (0x%X,0x%X).\n",
ha->pdev->subsystem_vendor,
ha->pdev->subsystem_device);
goto skip_msi;
}
if (IS_QLA2432(ha) && (ha->pdev->revision < QLA_MSIX_CHIP_REV_24XX)) {
ql_log(ql_log_warn, vha, 0x0035,
"MSI-X; Unsupported ISP2432 (0x%X, 0x%X).\n",
ha->pdev->revision, QLA_MSIX_CHIP_REV_24XX);
goto skip_msix;
}
ret = qla24xx_enable_msix(ha, rsp);
if (!ret) {
ql_dbg(ql_dbg_init, vha, 0x0036,
"MSI-X: Enabled (0x%X, 0x%X).\n",
ha->chip_revision, ha->fw_attributes);
goto clear_risc_ints;
}
skip_msix:
ql_log(ql_log_info, vha, 0x0037,
"Falling back-to MSI mode -%d.\n", ret);
if (!IS_QLA24XX(ha) && !IS_QLA2532(ha) && !IS_QLA8432(ha) &&
!IS_QLA8001(ha) && !IS_P3P_TYPE(ha) && !IS_QLAFX00(ha) &&
!IS_QLA27XX(ha))
goto skip_msi;
ret = pci_enable_msi(ha->pdev);
if (!ret) {
ql_dbg(ql_dbg_init, vha, 0x0038,
"MSI: Enabled.\n");
ha->flags.msi_enabled = 1;
} else
ql_log(ql_log_warn, vha, 0x0039,
"Falling back-to INTa mode -- %d.\n", ret);
skip_msi:
/* Skip INTx on ISP82xx. */
if (!ha->flags.msi_enabled && IS_QLA82XX(ha))
return QLA_FUNCTION_FAILED;
ret = request_irq(ha->pdev->irq, ha->isp_ops->intr_handler,
ha->flags.msi_enabled ? 0 : IRQF_SHARED,
QLA2XXX_DRIVER_NAME, rsp);
if (ret) {
ql_log(ql_log_warn, vha, 0x003a,
"Failed to reserve interrupt %d already in use.\n",
ha->pdev->irq);
goto fail;
} else if (!ha->flags.msi_enabled) {
ql_dbg(ql_dbg_init, vha, 0x0125,
"INTa mode: Enabled.\n");
ha->flags.mr_intr_valid = 1;
}
clear_risc_ints:
if (IS_FWI2_CAPABLE(ha) || IS_QLAFX00(ha))
goto fail;
spin_lock_irq(&ha->hardware_lock);
WRT_REG_WORD(&reg->isp.semaphore, 0);
spin_unlock_irq(&ha->hardware_lock);
fail:
return ret;
}
void
qla2x00_free_irqs(scsi_qla_host_t *vha)
{
struct qla_hw_data *ha = vha->hw;
struct rsp_que *rsp;
/*
* We need to check that ha->rsp_q_map is valid in case we are called
* from a probe failure context.
*/
if (!ha->rsp_q_map || !ha->rsp_q_map[0])
return;
rsp = ha->rsp_q_map[0];
if (ha->flags.msix_enabled)
qla24xx_disable_msix(ha);
else if (ha->flags.msi_enabled) {
free_irq(ha->pdev->irq, rsp);
pci_disable_msi(ha->pdev);
} else
free_irq(ha->pdev->irq, rsp);
}
int qla25xx_request_irq(struct rsp_que *rsp)
{
struct qla_hw_data *ha = rsp->hw;
struct qla_init_msix_entry *intr = &msix_entries[2];
struct qla_msix_entry *msix = rsp->msix;
scsi_qla_host_t *vha = pci_get_drvdata(ha->pdev);
int ret;
ret = request_irq(msix->vector, intr->handler, 0, intr->name, rsp);
if (ret) {
ql_log(ql_log_fatal, vha, 0x00e6,
"MSI-X: Unable to register handler -- %x/%d.\n",
msix->vector, ret);
return ret;
}
msix->have_irq = 1;
msix->rsp = rsp;
return ret;
}