asgardius/android_kernel_motorola_sm6225
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
android_kernel_motorola_sm6225/drivers/scsi/qla2xxx/qla_isr.c

1819 lines
47 KiB
C
Raw Normal View History

Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
/*
[SCSI] qla2xxx: Update license. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-10-27 20:10:08 +02:00
* QLogic Fibre Channel HBA Driver
* Copyright (c) 2003-2005 QLogic Corporation
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
*
[SCSI] qla2xxx: Update license. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-10-27 20:10:08 +02:00
* See LICENSE.qla2xxx for copyright and licensing details.
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
*/
#include "qla_def.h"
[SCSI] qla2xxx: Correct QUEUE_FULL handling. - Drop queue-depths across all luns for a given fcport during TASK_SET_FULL statuses. - Ramp-up I/Os after throttling. - Consolidate completion-status handling of CS_QUEUE_FULL with CS_COMPLETE as ISP24xx firmware no longer reports CS_QUEUE_FULL. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-10-13 18:33:39 +02:00
#include <scsi/scsi_tcq.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
static void qla2x00_mbx_completion(scsi_qla_host_t *, uint16_t);
static void qla2x00_process_completed_request(struct scsi_qla_host *, uint32_t);
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
static void qla2x00_status_entry(scsi_qla_host_t *, void *);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
static void qla2x00_status_cont_entry(scsi_qla_host_t *, sts_cont_entry_t *);
static void qla2x00_error_entry(scsi_qla_host_t *, sts_entry_t *);
static void qla2x00_ms_entry(scsi_qla_host_t *, ms_iocb_entry_t *);
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
static void qla24xx_ms_entry(scsi_qla_host_t *, struct ct_entry_24xx *);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
/**
* 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
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 15:55:46 +02:00
qla2100_intr_handler(int irq, void *dev_id)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
{
scsi_qla_host_t *ha;
[SCSI] qla2xxx: Add ISP24xx definitions. Add ISP24xx definitions. Add requisite structure definitions and #define's for ISP24xx support. Also drop volatile modifiers from device_reg_* register layouts as the members are never really accessed, only their offsets within the layout are used during reads and writes. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:30:26 +02:00
struct device_reg_2xxx __iomem *reg;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
int status;
unsigned long flags;
unsigned long iter;
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
uint16_t mb[4];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
ha = (scsi_qla_host_t *) dev_id;
if (!ha) {
printk(KERN_INFO
"%s(): NULL host pointer\n", __func__);
return (IRQ_NONE);
}
[SCSI] qla2xxx: Add ISP24xx definitions. Add ISP24xx definitions. Add requisite structure definitions and #define's for ISP24xx support. Also drop volatile modifiers from device_reg_* register layouts as the members are never really accessed, only their offsets within the layout are used during reads and writes. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:30:26 +02:00
reg = &ha->iobase->isp;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
status = 0;
spin_lock_irqsave(&ha->hardware_lock, flags);
for (iter = 50; iter--; ) {
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. */
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
mb[0] = RD_MAILBOX_REG(ha, reg, 0);
if (mb[0] > 0x3fff && mb[0] < 0x8000) {
qla2x00_mbx_completion(ha, mb[0]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
status |= MBX_INTERRUPT;
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
} 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(ha, mb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
} else {
/*EMPTY*/
DEBUG2(printk("scsi(%ld): Unrecognized "
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
"interrupt type (%d).\n",
ha->host_no, mb[0]));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
/* Release mailbox registers. */
WRT_REG_WORD(&reg->semaphore, 0);
RD_REG_WORD(&reg->semaphore);
} else {
qla2x00_process_response_queue(ha);
WRT_REG_WORD(&reg->hccr, HCCR_CLR_RISC_INT);
RD_REG_WORD(&reg->hccr);
}
}
spin_unlock_irqrestore(&ha->hardware_lock, flags);
if (test_bit(MBX_INTR_WAIT, &ha->mbx_cmd_flags) &&
(status & MBX_INTERRUPT) && ha->flags.mbox_int) {
set_bit(MBX_INTERRUPT, &ha->mbx_cmd_flags);
up(&ha->mbx_intr_sem);
}
return (IRQ_HANDLED);
}
/**
* 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
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 15:55:46 +02:00
qla2300_intr_handler(int irq, void *dev_id)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
{
scsi_qla_host_t *ha;
[SCSI] qla2xxx: Add ISP24xx definitions. Add ISP24xx definitions. Add requisite structure definitions and #define's for ISP24xx support. Also drop volatile modifiers from device_reg_* register layouts as the members are never really accessed, only their offsets within the layout are used during reads and writes. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:30:26 +02:00
struct device_reg_2xxx __iomem *reg;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
int status;
unsigned long flags;
unsigned long iter;
uint32_t stat;
uint16_t hccr;
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
uint16_t mb[4];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
ha = (scsi_qla_host_t *) dev_id;
if (!ha) {
printk(KERN_INFO
"%s(): NULL host pointer\n", __func__);
return (IRQ_NONE);
}
[SCSI] qla2xxx: Add ISP24xx definitions. Add ISP24xx definitions. Add requisite structure definitions and #define's for ISP24xx support. Also drop volatile modifiers from device_reg_* register layouts as the members are never really accessed, only their offsets within the layout are used during reads and writes. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:30:26 +02:00
reg = &ha->iobase->isp;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
status = 0;
spin_lock_irqsave(&ha->hardware_lock, flags);
for (iter = 50; iter--; ) {
stat = RD_REG_DWORD(&reg->u.isp2300.host_status);
if (stat & HSR_RISC_PAUSED) {
hccr = RD_REG_WORD(&reg->hccr);
if (hccr & (BIT_15 | BIT_13 | BIT_11 | BIT_8))
[SCSI] qla2xxx: Perform a fw-dump when an ISP23xx RISC-paused state is detected. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-12-14 04:20:31 +01:00
qla_printk(KERN_INFO, ha, "Parity error -- "
"HCCR=%x, Dumping firmware!\n", hccr);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
else
[SCSI] qla2xxx: Perform a fw-dump when an ISP23xx RISC-paused state is detected. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-12-14 04:20:31 +01:00
qla_printk(KERN_INFO, ha, "RISC paused -- "
"HCCR=%x, Dumping firmware!\n", hccr);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
/*
* Issue a "HARD" reset in order for the RISC
* interrupt bit to be cleared. Schedule a big
* hammmer to get out of the RISC PAUSED state.
*/
WRT_REG_WORD(&reg->hccr, HCCR_RESET_RISC);
RD_REG_WORD(&reg->hccr);
[SCSI] qla2xxx: Perform a fw-dump when an ISP23xx RISC-paused state is detected. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-12-14 04:20:31 +01:00
[SCSI] qla2xxx: Re-factor isp_operations to static structures. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:06:00 +02:00
ha->isp_ops->fw_dump(ha, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
break;
} else if ((stat & HSR_RISC_INT) == 0)
break;
switch (stat & 0xff) {
case 0x1:
case 0x2:
case 0x10:
case 0x11:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
qla2x00_mbx_completion(ha, MSW(stat));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
status |= MBX_INTERRUPT;
/* Release mailbox registers. */
WRT_REG_WORD(&reg->semaphore, 0);
break;
case 0x12:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
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(ha, mb);
break;
case 0x13:
qla2x00_process_response_queue(ha);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
break;
case 0x15:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
mb[0] = MBA_CMPLT_1_16BIT;
mb[1] = MSW(stat);
qla2x00_async_event(ha, mb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
break;
case 0x16:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
mb[0] = MBA_SCSI_COMPLETION;
mb[1] = MSW(stat);
mb[2] = RD_MAILBOX_REG(ha, reg, 2);
qla2x00_async_event(ha, mb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
break;
default:
DEBUG2(printk("scsi(%ld): Unrecognized interrupt type "
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
"(%d).\n",
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
ha->host_no, stat & 0xff));
break;
}
WRT_REG_WORD(&reg->hccr, HCCR_CLR_RISC_INT);
RD_REG_WORD_RELAXED(&reg->hccr);
}
spin_unlock_irqrestore(&ha->hardware_lock, flags);
if (test_bit(MBX_INTR_WAIT, &ha->mbx_cmd_flags) &&
(status & MBX_INTERRUPT) && ha->flags.mbox_int) {
set_bit(MBX_INTERRUPT, &ha->mbx_cmd_flags);
up(&ha->mbx_intr_sem);
}
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 *ha, uint16_t mb0)
{
uint16_t cnt;
uint16_t __iomem *wptr;
[SCSI] qla2xxx: Add ISP24xx definitions. Add ISP24xx definitions. Add requisite structure definitions and #define's for ISP24xx support. Also drop volatile modifiers from device_reg_* register layouts as the members are never really accessed, only their offsets within the layout are used during reads and writes. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:30:26 +02:00
struct device_reg_2xxx __iomem *reg = &ha->iobase->isp;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
/* Load return mailbox registers. */
ha->flags.mbox_int = 1;
ha->mailbox_out[0] = mb0;
wptr = (uint16_t __iomem *)MAILBOX_REG(ha, reg, 1);
for (cnt = 1; cnt < ha->mbx_count; cnt++) {
[SCSI] qla2xxx: Code scrubbing. Code scrubbing. - Remove trailing whitespace from driver files. - Remove unused #defines and inlines. - Standardize on C comments (// -> /* */) Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:32:07 +02:00
if (IS_QLA2200(ha) && cnt == 8)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
wptr = (uint16_t __iomem *)MAILBOX_REG(ha, reg, 8);
if (cnt == 4 || cnt == 5)
ha->mailbox_out[cnt] = qla2x00_debounce_register(wptr);
else
ha->mailbox_out[cnt] = RD_REG_WORD(wptr);
[SCSI] qla2xxx: Code scrubbing. Code scrubbing. - Remove trailing whitespace from driver files. - Remove unused #defines and inlines. - Standardize on C comments (// -> /* */) Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:32:07 +02:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
wptr++;
}
if (ha->mcp) {
DEBUG3(printk("%s(%ld): Got mailbox completion. cmd=%x.\n",
__func__, ha->host_no, ha->mcp->mb[0]));
} else {
DEBUG2_3(printk("%s(%ld): MBX pointer ERROR!\n",
__func__, ha->host_no));
}
}
/**
* qla2x00_async_event() - Process aynchronous events.
* @ha: SCSI driver HA context
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
* @mb: Mailbox registers (0 - 3)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
*/
[SCSI] qla2xxx: add support for NPIV Following patch adds support for NPIV (N-Port ID Virtualization) to the qla2xxx. - supported within switched-fabric topologies only. - supports up to 63 virtual ports on each physical port. Signed-off-by: Seokmann Ju <seokmann.ju@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-05 22:16:51 +02:00
void
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
qla2x00_async_event(scsi_qla_host_t *ha, uint16_t *mb)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
{
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
#define LS_UNKNOWN 2
static char *link_speeds[5] = { "1", "2", "?", "4", "10" };
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
char *link_speed;
uint16_t handle_cnt;
uint16_t cnt;
uint32_t handles[5];
[SCSI] qla2xxx: Add ISP24xx definitions. Add ISP24xx definitions. Add requisite structure definitions and #define's for ISP24xx support. Also drop volatile modifiers from device_reg_* register layouts as the members are never really accessed, only their offsets within the layout are used during reads and writes. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:30:26 +02:00
struct device_reg_2xxx __iomem *reg = &ha->iobase->isp;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
uint32_t rscn_entry, host_pid;
uint8_t rscn_queue_index;
/* Setup to process RIO completion. */
handle_cnt = 0;
switch (mb[0]) {
case MBA_SCSI_COMPLETION:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
handles[0] = le32_to_cpu((uint32_t)((mb[2] << 16) | mb[1]));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
handle_cnt = 1;
break;
case MBA_CMPLT_1_16BIT:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
handles[0] = mb[1];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
handle_cnt = 1;
mb[0] = MBA_SCSI_COMPLETION;
break;
case MBA_CMPLT_2_16BIT:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
handles[0] = mb[1];
handles[1] = mb[2];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
handle_cnt = 2;
mb[0] = MBA_SCSI_COMPLETION;
break;
case MBA_CMPLT_3_16BIT:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
handles[0] = mb[1];
handles[1] = mb[2];
handles[2] = mb[3];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
handle_cnt = 3;
mb[0] = MBA_SCSI_COMPLETION;
break;
case MBA_CMPLT_4_16BIT:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
handles[0] = mb[1];
handles[1] = mb[2];
handles[2] = mb[3];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
handles[3] = (uint32_t)RD_MAILBOX_REG(ha, reg, 6);
handle_cnt = 4;
mb[0] = MBA_SCSI_COMPLETION;
break;
case MBA_CMPLT_5_16BIT:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
handles[0] = mb[1];
handles[1] = mb[2];
handles[2] = mb[3];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
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:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
handles[0] = le32_to_cpu((uint32_t)((mb[2] << 16) | mb[1]));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
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;
}
switch (mb[0]) {
case MBA_SCSI_COMPLETION: /* Fast Post */
if (!ha->flags.online)
break;
for (cnt = 0; cnt < handle_cnt; cnt++)
qla2x00_process_completed_request(ha, handles[cnt]);
break;
case MBA_RESET: /* Reset */
DEBUG2(printk("scsi(%ld): Asynchronous RESET.\n", ha->host_no));
set_bit(RESET_MARKER_NEEDED, &ha->dpc_flags);
break;
case MBA_SYSTEM_ERR: /* System Error */
mb[1] = RD_MAILBOX_REG(ha, reg, 1);
mb[2] = RD_MAILBOX_REG(ha, reg, 2);
mb[3] = RD_MAILBOX_REG(ha, reg, 3);
qla_printk(KERN_INFO, ha,
"ISP System Error - mbx1=%xh mbx2=%xh mbx3=%xh.\n",
mb[1], mb[2], mb[3]);
[SCSI] qla2xxx: Re-factor isp_operations to static structures. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:06:00 +02:00
ha->isp_ops->fw_dump(ha, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
[SCSI] qla2xxx: Generalize FW-Interface-2 support. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:05:56 +02:00
if (IS_FWI2_CAPABLE(ha)) {
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
if (mb[1] == 0 && mb[2] == 0) {
qla_printk(KERN_ERR, ha,
"Unrecoverable Hardware Error: adapter "
"marked OFFLINE!\n");
ha->flags.online = 0;
} else
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
} else if (mb[1] == 0) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
qla_printk(KERN_INFO, ha,
"Unrecoverable Hardware Error: adapter marked "
"OFFLINE!\n");
ha->flags.online = 0;
} else
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
break;
case MBA_REQ_TRANSFER_ERR: /* Request Transfer Error */
DEBUG2(printk("scsi(%ld): ISP Request Transfer Error.\n",
ha->host_no));
qla_printk(KERN_WARNING, ha, "ISP Request Transfer Error.\n");
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
break;
case MBA_RSP_TRANSFER_ERR: /* Response Transfer Error */
DEBUG2(printk("scsi(%ld): ISP Response Transfer Error.\n",
ha->host_no));
qla_printk(KERN_WARNING, ha, "ISP Response Transfer Error.\n");
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
break;
case MBA_WAKEUP_THRES: /* Request Queue Wake-up */
DEBUG2(printk("scsi(%ld): Asynchronous WAKEUP_THRES.\n",
ha->host_no));
break;
case MBA_LIP_OCCURRED: /* Loop Initialization Procedure */
DEBUG2(printk("scsi(%ld): LIP occured (%x).\n", ha->host_no,
mb[1]));
qla_printk(KERN_INFO, ha, "LIP occured (%x).\n", mb[1]);
if (atomic_read(&ha->loop_state) != LOOP_DOWN) {
atomic_set(&ha->loop_state, LOOP_DOWN);
atomic_set(&ha->loop_down_timer, LOOP_DOWN_TIME);
[SCSI] qla2xxx: Correct synchronization issues during rport addition/deletion. The driver can typically detect port-loss during an interrupt context (i.e. via interrogation of a status IOCB's completion status [CS_PORT_LOGGED_OUT]. Due to the calling requirements of the fc_rport APIs, the driver would defer removal of the device to the default workqueue. If the work-item was preceded by an event which caused the port to obtain visibility (relogin successful, target re-logged into the topology), deferred removal could inadvertently drop the rport. The code also no longer defers removal via the default workqueue, instead opting for use of the driver's own DPC thread. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-01-20 23:53:13 +01:00
qla2x00_mark_all_devices_lost(ha, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
[SCSI] qla2xxx: add support for NPIV Following patch adds support for NPIV (N-Port ID Virtualization) to the qla2xxx. - supported within switched-fabric topologies only. - supports up to 63 virtual ports on each physical port. Signed-off-by: Seokmann Ju <seokmann.ju@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-05 22:16:51 +02:00
if (ha->parent) {
atomic_set(&ha->vp_state, VP_FAILED);
fc_vport_set_state(ha->fc_vport, FC_VPORT_FAILED);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
set_bit(REGISTER_FC4_NEEDED, &ha->dpc_flags);
ha->flags.management_server_logged_in = 0;
break;
case MBA_LOOP_UP: /* Loop Up Event */
if (IS_QLA2100(ha) || IS_QLA2200(ha)) {
link_speed = link_speeds[0];
[SCSI] qla2xxx: Add iIDMA support. iIDMA (Intelligent Interleaved Direct Memory Access) allows for the HBA hardware to send FC frames at the rate at which they can be received by a target device. By taking advantage of the higher link rate, the HBA can maximize bandwidth utilization in a heterogeneous multi-speed SAN. Within a fabric topology, port speed detection is done via a Name Server command (GFPN_ID) followed by a Fabric Management command (GPSC). In an FCAL/N2N topology, port speed is based on the HBA link-rate. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-10-02 21:00:43 +02:00
ha->link_data_rate = PORT_SPEED_1GB;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
} else {
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
link_speed = link_speeds[LS_UNKNOWN];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (mb[1] < 5)
link_speed = link_speeds[mb[1]];
ha->link_data_rate = mb[1];
}
DEBUG2(printk("scsi(%ld): Asynchronous LOOP UP (%s Gbps).\n",
ha->host_no, link_speed));
qla_printk(KERN_INFO, ha, "LOOP UP detected (%s Gbps).\n",
link_speed);
ha->flags.management_server_logged_in = 0;
break;
case MBA_LOOP_DOWN: /* Loop Down Event */
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
DEBUG2(printk("scsi(%ld): Asynchronous LOOP DOWN (%x).\n",
ha->host_no, mb[1]));
qla_printk(KERN_INFO, ha, "LOOP DOWN detected (%x).\n", mb[1]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (atomic_read(&ha->loop_state) != LOOP_DOWN) {
atomic_set(&ha->loop_state, LOOP_DOWN);
atomic_set(&ha->loop_down_timer, LOOP_DOWN_TIME);
ha->device_flags |= DFLG_NO_CABLE;
[SCSI] qla2xxx: Correct synchronization issues during rport addition/deletion. The driver can typically detect port-loss during an interrupt context (i.e. via interrogation of a status IOCB's completion status [CS_PORT_LOGGED_OUT]. Due to the calling requirements of the fc_rport APIs, the driver would defer removal of the device to the default workqueue. If the work-item was preceded by an event which caused the port to obtain visibility (relogin successful, target re-logged into the topology), deferred removal could inadvertently drop the rport. The code also no longer defers removal via the default workqueue, instead opting for use of the driver's own DPC thread. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-01-20 23:53:13 +01:00
qla2x00_mark_all_devices_lost(ha, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
[SCSI] qla2xxx: add support for NPIV Following patch adds support for NPIV (N-Port ID Virtualization) to the qla2xxx. - supported within switched-fabric topologies only. - supports up to 63 virtual ports on each physical port. Signed-off-by: Seokmann Ju <seokmann.ju@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-05 22:16:51 +02:00
if (ha->parent) {
atomic_set(&ha->vp_state, VP_FAILED);
fc_vport_set_state(ha->fc_vport, FC_VPORT_FAILED);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
ha->flags.management_server_logged_in = 0;
[SCSI] qla2xxx: Add iIDMA support. iIDMA (Intelligent Interleaved Direct Memory Access) allows for the HBA hardware to send FC frames at the rate at which they can be received by a target device. By taking advantage of the higher link rate, the HBA can maximize bandwidth utilization in a heterogeneous multi-speed SAN. Within a fabric topology, port speed detection is done via a Name Server command (GFPN_ID) followed by a Fabric Management command (GPSC). In an FCAL/N2N topology, port speed is based on the HBA link-rate. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-10-02 21:00:43 +02:00
ha->link_data_rate = PORT_SPEED_UNKNOWN;
[SCSI] qla2xxx: Add FDMI support. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-08-27 04:08:30 +02:00
if (ql2xfdmienable)
set_bit(REGISTER_FDMI_NEEDED, &ha->dpc_flags);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
break;
case MBA_LIP_RESET: /* LIP reset occurred */
DEBUG2(printk("scsi(%ld): Asynchronous LIP RESET (%x).\n",
ha->host_no, mb[1]));
qla_printk(KERN_INFO, ha,
"LIP reset occured (%x).\n", mb[1]);
if (atomic_read(&ha->loop_state) != LOOP_DOWN) {
atomic_set(&ha->loop_state, LOOP_DOWN);
atomic_set(&ha->loop_down_timer, LOOP_DOWN_TIME);
[SCSI] qla2xxx: Correct synchronization issues during rport addition/deletion. The driver can typically detect port-loss during an interrupt context (i.e. via interrogation of a status IOCB's completion status [CS_PORT_LOGGED_OUT]. Due to the calling requirements of the fc_rport APIs, the driver would defer removal of the device to the default workqueue. If the work-item was preceded by an event which caused the port to obtain visibility (relogin successful, target re-logged into the topology), deferred removal could inadvertently drop the rport. The code also no longer defers removal via the default workqueue, instead opting for use of the driver's own DPC thread. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-01-20 23:53:13 +01:00
qla2x00_mark_all_devices_lost(ha, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
[SCSI] qla2xxx: add support for NPIV Following patch adds support for NPIV (N-Port ID Virtualization) to the qla2xxx. - supported within switched-fabric topologies only. - supports up to 63 virtual ports on each physical port. Signed-off-by: Seokmann Ju <seokmann.ju@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-05 22:16:51 +02:00
if (ha->parent) {
atomic_set(&ha->vp_state, VP_FAILED);
fc_vport_set_state(ha->fc_vport, FC_VPORT_FAILED);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
set_bit(RESET_MARKER_NEEDED, &ha->dpc_flags);
ha->operating_mode = LOOP;
ha->flags.management_server_logged_in = 0;
break;
case MBA_POINT_TO_POINT: /* Point-to-Point */
if (IS_QLA2100(ha))
break;
DEBUG2(printk("scsi(%ld): Asynchronous P2P MODE received.\n",
ha->host_no));
/*
* Until there's a transition from loop down to loop up, treat
* this as loop down only.
*/
if (atomic_read(&ha->loop_state) != LOOP_DOWN) {
atomic_set(&ha->loop_state, LOOP_DOWN);
if (!atomic_read(&ha->loop_down_timer))
atomic_set(&ha->loop_down_timer,
LOOP_DOWN_TIME);
[SCSI] qla2xxx: Correct synchronization issues during rport addition/deletion. The driver can typically detect port-loss during an interrupt context (i.e. via interrogation of a status IOCB's completion status [CS_PORT_LOGGED_OUT]. Due to the calling requirements of the fc_rport APIs, the driver would defer removal of the device to the default workqueue. If the work-item was preceded by an event which caused the port to obtain visibility (relogin successful, target re-logged into the topology), deferred removal could inadvertently drop the rport. The code also no longer defers removal via the default workqueue, instead opting for use of the driver's own DPC thread. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-01-20 23:53:13 +01:00
qla2x00_mark_all_devices_lost(ha, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
[SCSI] qla2xxx: add support for NPIV Following patch adds support for NPIV (N-Port ID Virtualization) to the qla2xxx. - supported within switched-fabric topologies only. - supports up to 63 virtual ports on each physical port. Signed-off-by: Seokmann Ju <seokmann.ju@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-05 22:16:51 +02:00
if (ha->parent) {
atomic_set(&ha->vp_state, VP_FAILED);
fc_vport_set_state(ha->fc_vport, FC_VPORT_FAILED);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (!(test_bit(ABORT_ISP_ACTIVE, &ha->dpc_flags))) {
set_bit(RESET_MARKER_NEEDED, &ha->dpc_flags);
}
set_bit(REGISTER_FC4_NEEDED, &ha->dpc_flags);
[SCSI] qla2xxx: Detect GPSC capabilities within a fabric. Disable subsequent GPSC queries if Fabric Management services do not support the operation. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-12-14 04:20:28 +01:00
ha->flags.gpsc_supported = 1;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
break;
case MBA_CHG_IN_CONNECTION: /* Change in connection mode */
if (IS_QLA2100(ha))
break;
DEBUG2(printk("scsi(%ld): Asynchronous Change In Connection "
"received.\n",
ha->host_no));
qla_printk(KERN_INFO, ha,
"Configuration change detected: value=%x.\n", mb[1]);
if (atomic_read(&ha->loop_state) != LOOP_DOWN) {
[SCSI] qla2xxx: Code scrubbing. Code scrubbing. - Remove trailing whitespace from driver files. - Remove unused #defines and inlines. - Standardize on C comments (// -> /* */) Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:32:07 +02:00
atomic_set(&ha->loop_state, LOOP_DOWN);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (!atomic_read(&ha->loop_down_timer))
atomic_set(&ha->loop_down_timer,
LOOP_DOWN_TIME);
[SCSI] qla2xxx: Correct synchronization issues during rport addition/deletion. The driver can typically detect port-loss during an interrupt context (i.e. via interrogation of a status IOCB's completion status [CS_PORT_LOGGED_OUT]. Due to the calling requirements of the fc_rport APIs, the driver would defer removal of the device to the default workqueue. If the work-item was preceded by an event which caused the port to obtain visibility (relogin successful, target re-logged into the topology), deferred removal could inadvertently drop the rport. The code also no longer defers removal via the default workqueue, instead opting for use of the driver's own DPC thread. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-01-20 23:53:13 +01:00
qla2x00_mark_all_devices_lost(ha, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
[SCSI] qla2xxx: add support for NPIV Following patch adds support for NPIV (N-Port ID Virtualization) to the qla2xxx. - supported within switched-fabric topologies only. - supports up to 63 virtual ports on each physical port. Signed-off-by: Seokmann Ju <seokmann.ju@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-05 22:16:51 +02:00
if (ha->parent) {
atomic_set(&ha->vp_state, VP_FAILED);
fc_vport_set_state(ha->fc_vport, FC_VPORT_FAILED);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
set_bit(LOOP_RESYNC_NEEDED, &ha->dpc_flags);
set_bit(LOCAL_LOOP_UPDATE, &ha->dpc_flags);
break;
case MBA_PORT_UPDATE: /* Port database update */
/*
* If PORT UPDATE is global (recieved LIP_OCCURED/LIP_RESET
* event etc. earlier indicating loop is down) then process
* it. Otherwise ignore it and Wait for RSCN to come in.
*/
atomic_set(&ha->loop_down_timer, 0);
if (atomic_read(&ha->loop_state) != LOOP_DOWN &&
atomic_read(&ha->loop_state) != LOOP_DEAD) {
DEBUG2(printk("scsi(%ld): Asynchronous PORT UPDATE "
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
"ignored %04x/%04x/%04x.\n", ha->host_no, mb[1],
mb[2], mb[3]));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
break;
}
DEBUG2(printk("scsi(%ld): Asynchronous PORT UPDATE.\n",
ha->host_no));
DEBUG(printk(KERN_INFO
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
"scsi(%ld): Port database changed %04x %04x %04x.\n",
ha->host_no, mb[1], mb[2], mb[3]));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
/*
* Mark all devices as missing so we will login again.
*/
atomic_set(&ha->loop_state, LOOP_UP);
[SCSI] qla2xxx: Correct synchronization issues during rport addition/deletion. The driver can typically detect port-loss during an interrupt context (i.e. via interrogation of a status IOCB's completion status [CS_PORT_LOGGED_OUT]. Due to the calling requirements of the fc_rport APIs, the driver would defer removal of the device to the default workqueue. If the work-item was preceded by an event which caused the port to obtain visibility (relogin successful, target re-logged into the topology), deferred removal could inadvertently drop the rport. The code also no longer defers removal via the default workqueue, instead opting for use of the driver's own DPC thread. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-01-20 23:53:13 +01:00
qla2x00_mark_all_devices_lost(ha, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
ha->flags.rscn_queue_overflow = 1;
set_bit(LOOP_RESYNC_NEEDED, &ha->dpc_flags);
set_bit(LOCAL_LOOP_UPDATE, &ha->dpc_flags);
break;
case MBA_RSCN_UPDATE: /* State Change Registration */
[SCSI] qla2xxx: add support for NPIV Following patch adds support for NPIV (N-Port ID Virtualization) to the qla2xxx. - supported within switched-fabric topologies only. - supports up to 63 virtual ports on each physical port. Signed-off-by: Seokmann Ju <seokmann.ju@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-05 22:16:51 +02:00
/* Check if the Vport has issued a SCR */
if (ha->parent && test_bit(VP_SCR_NEEDED, &ha->vp_flags))
break;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
DEBUG2(printk("scsi(%ld): Asynchronous RSCR UPDATE.\n",
ha->host_no));
DEBUG(printk(KERN_INFO
"scsi(%ld): RSCN database changed -- %04x %04x.\n",
ha->host_no, mb[1], mb[2]));
rscn_entry = (mb[1] << 16) | mb[2];
host_pid = (ha->d_id.b.domain << 16) | (ha->d_id.b.area << 8) |
ha->d_id.b.al_pa;
if (rscn_entry == host_pid) {
DEBUG(printk(KERN_INFO
"scsi(%ld): Ignoring RSCN update to local host "
"port ID (%06x)\n",
ha->host_no, host_pid));
break;
}
rscn_queue_index = ha->rscn_in_ptr + 1;
if (rscn_queue_index == MAX_RSCN_COUNT)
rscn_queue_index = 0;
if (rscn_queue_index != ha->rscn_out_ptr) {
ha->rscn_queue[ha->rscn_in_ptr] = rscn_entry;
ha->rscn_in_ptr = rscn_queue_index;
} else {
ha->flags.rscn_queue_overflow = 1;
}
atomic_set(&ha->loop_state, LOOP_UPDATE);
atomic_set(&ha->loop_down_timer, 0);
ha->flags.management_server_logged_in = 0;
set_bit(LOOP_RESYNC_NEEDED, &ha->dpc_flags);
set_bit(RSCN_UPDATE, &ha->dpc_flags);
break;
/* case MBA_RIO_RESPONSE: */
case MBA_ZIO_RESPONSE:
DEBUG2(printk("scsi(%ld): [R|Z]IO update completion.\n",
ha->host_no));
DEBUG(printk(KERN_INFO
"scsi(%ld): [R|Z]IO update completion.\n",
ha->host_no));
[SCSI] qla2xxx: Generalize FW-Interface-2 support. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:05:56 +02:00
if (IS_FWI2_CAPABLE(ha))
[SCSI] qla2xxx: Add support to dynamically enable/disable ZIO. ISP23xx and ISP24xx chips have support for an adaptive method of posting SCSI command completions for multiple SCSI commands during a single system interrupt. SCSI commands are placed on the system response queue without interrupting the host until 1) a delay timer expires; or 2) a SCSI command completes with an error. As long as the host software (qla2xxx) services the response queue for completions (this polling is done during queuecommand()) within the 'delay timer' period, the firmware will not generate system interrupt. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-10-27 20:09:48 +02:00
qla24xx_process_response_queue(ha);
else
qla2x00_process_response_queue(ha);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
break;
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
case MBA_DISCARD_RND_FRAME:
DEBUG2(printk("scsi(%ld): Discard RND Frame -- %04x %04x "
"%04x.\n", ha->host_no, mb[1], mb[2], mb[3]));
break;
[SCSI] qla2xxx: Log Trace/Diagonostic asynchronous events. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-08-01 22:48:14 +02:00
case MBA_TRACE_NOTIFICATION:
DEBUG2(printk("scsi(%ld): Trace Notification -- %04x %04x.\n",
ha->host_no, mb[1], mb[2]));
break;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
[SCSI] qla2xxx: add support for NPIV Following patch adds support for NPIV (N-Port ID Virtualization) to the qla2xxx. - supported within switched-fabric topologies only. - supports up to 63 virtual ports on each physical port. Signed-off-by: Seokmann Ju <seokmann.ju@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-05 22:16:51 +02:00
if (!ha->parent && ha->num_vhosts)
qla2x00_alert_all_vps(ha, mb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
[SCSI] qla2xxx: Correct QUEUE_FULL handling. - Drop queue-depths across all luns for a given fcport during TASK_SET_FULL statuses. - Ramp-up I/Os after throttling. - Consolidate completion-status handling of CS_QUEUE_FULL with CS_COMPLETE as ISP24xx firmware no longer reports CS_QUEUE_FULL. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-10-13 18:33:39 +02:00
static void
qla2x00_adjust_sdev_qdepth_up(struct scsi_device *sdev, void *data)
{
fc_port_t *fcport = data;
if (fcport->ha->max_q_depth <= sdev->queue_depth)
return;
if (sdev->ordered_tags)
scsi_adjust_queue_depth(sdev, MSG_ORDERED_TAG,
sdev->queue_depth + 1);
else
scsi_adjust_queue_depth(sdev, MSG_SIMPLE_TAG,
sdev->queue_depth + 1);
fcport->last_ramp_up = jiffies;
DEBUG2(qla_printk(KERN_INFO, fcport->ha,
"scsi(%ld:%d:%d:%d): Queue depth adjusted-up to %d.\n",
fcport->ha->host_no, sdev->channel, sdev->id, sdev->lun,
sdev->queue_depth));
}
static void
qla2x00_adjust_sdev_qdepth_down(struct scsi_device *sdev, void *data)
{
fc_port_t *fcport = data;
if (!scsi_track_queue_full(sdev, sdev->queue_depth - 1))
return;
DEBUG2(qla_printk(KERN_INFO, fcport->ha,
"scsi(%ld:%d:%d:%d): Queue depth adjusted-down to %d.\n",
fcport->ha->host_no, sdev->channel, sdev->id, sdev->lun,
sdev->queue_depth));
}
static inline void
qla2x00_ramp_up_queue_depth(scsi_qla_host_t *ha, srb_t *sp)
{
fc_port_t *fcport;
struct scsi_device *sdev;
sdev = sp->cmd->device;
if (sdev->queue_depth >= ha->max_q_depth)
return;
fcport = sp->fcport;
if (time_before(jiffies,
fcport->last_ramp_up + ql2xqfullrampup * HZ))
return;
if (time_before(jiffies,
fcport->last_queue_full + ql2xqfullrampup * HZ))
return;
starget_for_each_device(sdev->sdev_target, fcport,
qla2x00_adjust_sdev_qdepth_up);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
/**
* qla2x00_process_completed_request() - Process a Fast Post response.
* @ha: SCSI driver HA context
* @index: SRB index
*/
static void
qla2x00_process_completed_request(struct scsi_qla_host *ha, uint32_t index)
{
srb_t *sp;
/* Validate handle. */
if (index >= MAX_OUTSTANDING_COMMANDS) {
DEBUG2(printk("scsi(%ld): Invalid SCSI completion handle %d.\n",
ha->host_no, index));
qla_printk(KERN_WARNING, ha,
"Invalid SCSI completion handle %d.\n", index);
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
return;
}
sp = ha->outstanding_cmds[index];
if (sp) {
/* Free outstanding command slot. */
ha->outstanding_cmds[index] = NULL;
CMD_COMPL_STATUS(sp->cmd) = 0L;
CMD_SCSI_STATUS(sp->cmd) = 0L;
/* Save ISP completion status */
sp->cmd->result = DID_OK << 16;
[SCSI] qla2xxx: Correct QUEUE_FULL handling. - Drop queue-depths across all luns for a given fcport during TASK_SET_FULL statuses. - Ramp-up I/Os after throttling. - Consolidate completion-status handling of CS_QUEUE_FULL with CS_COMPLETE as ISP24xx firmware no longer reports CS_QUEUE_FULL. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-10-13 18:33:39 +02:00
qla2x00_ramp_up_queue_depth(ha, sp);
[PATCH] qla2xxx: remove internal queuing... Remove internal command queuing from the driver. As is, this driver cannot tolerate cable-pulls as I/Os will begin to fail by the upper layers. o Should be used in conjuction with the 11-fc_rport_adds_2.diff patch. o Removes qla_listops.h file -- no longer needed. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-17 22:02:26 +02:00
qla2x00_sp_compl(ha, sp);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
} else {
DEBUG2(printk("scsi(%ld): Invalid ISP SCSI completion handle\n",
ha->host_no));
qla_printk(KERN_WARNING, ha,
"Invalid ISP SCSI completion handle\n");
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
}
}
/**
* qla2x00_process_response_queue() - Process response queue entries.
* @ha: SCSI driver HA context
*/
void
qla2x00_process_response_queue(struct scsi_qla_host *ha)
{
[SCSI] qla2xxx: Add ISP24xx definitions. Add ISP24xx definitions. Add requisite structure definitions and #define's for ISP24xx support. Also drop volatile modifiers from device_reg_* register layouts as the members are never really accessed, only their offsets within the layout are used during reads and writes. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:30:26 +02:00
struct device_reg_2xxx __iomem *reg = &ha->iobase->isp;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
sts_entry_t *pkt;
uint16_t handle_cnt;
uint16_t cnt;
if (!ha->flags.online)
return;
while (ha->response_ring_ptr->signature != RESPONSE_PROCESSED) {
pkt = (sts_entry_t *)ha->response_ring_ptr;
ha->rsp_ring_index++;
if (ha->rsp_ring_index == ha->response_q_length) {
ha->rsp_ring_index = 0;
ha->response_ring_ptr = ha->response_ring;
} else {
ha->response_ring_ptr++;
}
if (pkt->entry_status != 0) {
DEBUG3(printk(KERN_INFO
"scsi(%ld): Process error entry.\n", ha->host_no));
qla2x00_error_entry(ha, pkt);
((response_t *)pkt)->signature = RESPONSE_PROCESSED;
wmb();
continue;
}
switch (pkt->entry_type) {
case STATUS_TYPE:
qla2x00_status_entry(ha, 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(ha,
((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(ha,
((sts22_entry_t *)pkt)->handle[cnt]);
}
break;
case STATUS_CONT_TYPE:
qla2x00_status_cont_entry(ha, (sts_cont_entry_t *)pkt);
break;
case MS_IOCB_TYPE:
qla2x00_ms_entry(ha, (ms_iocb_entry_t *)pkt);
break;
default:
/* Type Not Supported. */
DEBUG4(printk(KERN_WARNING
"scsi(%ld): Received unknown response pkt type %x "
"entry status=%x.\n",
ha->host_no, 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), ha->rsp_ring_index);
}
/**
* qla2x00_status_entry() - Process a Status IOCB entry.
* @ha: SCSI driver HA context
* @pkt: Entry pointer
*/
static void
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
qla2x00_status_entry(scsi_qla_host_t *ha, void *pkt)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
{
srb_t *sp;
fc_port_t *fcport;
struct scsi_cmnd *cp;
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
sts_entry_t *sts;
struct sts_entry_24xx *sts24;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
uint16_t comp_status;
uint16_t scsi_status;
uint8_t lscsi_status;
int32_t resid;
[SCSI] qla2xxx: Use FW calculated residual count for underrun handling. With ISP24XX and ISP54XX parts. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-05-18 00:08:55 +02:00
uint32_t sense_len, rsp_info_len, resid_len, fw_resid_len;
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
uint8_t *rsp_info, *sense_data;
sts = (sts_entry_t *) pkt;
sts24 = (struct sts_entry_24xx *) pkt;
[SCSI] qla2xxx: Generalize FW-Interface-2 support. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:05:56 +02:00
if (IS_FWI2_CAPABLE(ha)) {
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
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;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
/* Fast path completion. */
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
if (comp_status == CS_COMPLETE && scsi_status == 0) {
qla2x00_process_completed_request(ha, sts->handle);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
return;
}
/* Validate handle. */
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
if (sts->handle < MAX_OUTSTANDING_COMMANDS) {
sp = ha->outstanding_cmds[sts->handle];
ha->outstanding_cmds[sts->handle] = NULL;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
} else
sp = NULL;
if (sp == NULL) {
DEBUG2(printk("scsi(%ld): Status Entry invalid handle.\n",
ha->host_no));
qla_printk(KERN_WARNING, ha, "Status Entry invalid handle.\n");
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
[SCSI] qla2xxx: use kthread_ API Use the kthread_ API instead of opencoding lots of hairy code for kernel thread creation and teardown. Also switch from semaphore-based thread wakeup to wake_up_process. Signed-off-by: Christoph Hellwig <hch@lst.de> Acked-By: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-02-14 18:46:22 +01:00
qla2xxx_wake_dpc(ha);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
return;
}
cp = sp->cmd;
if (cp == NULL) {
DEBUG2(printk("scsi(%ld): Command already returned back to OS "
[SCSI] qla2xxx: Drop unused driver cruft. - structure definitions. - structure members. - #define's. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-05-18 00:09:22 +02:00
"pkt->handle=%d sp=%p.\n", ha->host_no, sts->handle, sp));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
qla_printk(KERN_WARNING, ha,
"Command is NULL: already returned to OS (sp=%p)\n", sp);
return;
}
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
lscsi_status = scsi_status & STATUS_MASK;
CMD_ENTRY_STATUS(cp) = sts->entry_status;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
CMD_COMPL_STATUS(cp) = comp_status;
CMD_SCSI_STATUS(cp) = scsi_status;
[PATCH] qla2xxx: remove lun discovery codes... Remove internal lun discovery routines and support structures. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-17 22:06:53 +02:00
fcport = sp->fcport;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
[SCSI] qla2xxx: Use FW calculated residual count for underrun handling. With ISP24XX and ISP54XX parts. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-05-18 00:08:55 +02:00
sense_len = rsp_info_len = resid_len = fw_resid_len = 0;
[SCSI] qla2xxx: Generalize FW-Interface-2 support. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:05:56 +02:00
if (IS_FWI2_CAPABLE(ha)) {
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
sense_len = le32_to_cpu(sts24->sense_len);
rsp_info_len = le32_to_cpu(sts24->rsp_data_len);
resid_len = le32_to_cpu(sts24->rsp_residual_count);
[SCSI] qla2xxx: Use FW calculated residual count for underrun handling. With ISP24XX and ISP54XX parts. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-05-18 00:08:55 +02:00
fw_resid_len = le32_to_cpu(sts24->residual_len);
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
rsp_info = sts24->data;
sense_data = sts24->data;
host_to_fcp_swap(sts24->data, sizeof(sts24->data));
} else {
sense_len = le16_to_cpu(sts->req_sense_length);
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;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
/* Check for any FCP transport errors. */
if (scsi_status & SS_RESPONSE_INFO_LEN_VALID) {
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
/* Sense data lies beyond any FCP RESPONSE data. */
[SCSI] qla2xxx: Generalize FW-Interface-2 support. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:05:56 +02:00
if (IS_FWI2_CAPABLE(ha))
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
sense_data += rsp_info_len;
if (rsp_info_len > 3 && rsp_info[3]) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
DEBUG2(printk("scsi(%ld:%d:%d:%d) FCP I/O protocol "
"failure (%x/%02x%02x%02x%02x%02x%02x%02x%02x)..."
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
"retrying command\n", ha->host_no,
cp->device->channel, cp->device->id,
cp->device->lun, rsp_info_len, rsp_info[0],
rsp_info[1], rsp_info[2], rsp_info[3], rsp_info[4],
rsp_info[5], rsp_info[6], rsp_info[7]));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
cp->result = DID_BUS_BUSY << 16;
[PATCH] qla2xxx: remove internal queuing... Remove internal command queuing from the driver. As is, this driver cannot tolerate cable-pulls as I/Os will begin to fail by the upper layers. o Should be used in conjuction with the 11-fc_rport_adds_2.diff patch. o Removes qla_listops.h file -- no longer needed. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-17 22:02:26 +02:00
qla2x00_sp_compl(ha, sp);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
return;
}
}
/*
* Based on Host and scsi status generate status code for Linux
*/
switch (comp_status) {
case CS_COMPLETE:
[SCSI] qla2xxx: Correct QUEUE_FULL handling. - Drop queue-depths across all luns for a given fcport during TASK_SET_FULL statuses. - Ramp-up I/Os after throttling. - Consolidate completion-status handling of CS_QUEUE_FULL with CS_COMPLETE as ISP24xx firmware no longer reports CS_QUEUE_FULL. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-10-13 18:33:39 +02:00
case CS_QUEUE_FULL:
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (scsi_status == 0) {
cp->result = DID_OK << 16;
break;
}
if (scsi_status & (SS_RESIDUAL_UNDER | SS_RESIDUAL_OVER)) {
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
resid = resid_len;
[SCSI] qla2xxx: convert to use the data buffer accessors - remove the unnecessary map_single path. - convert to use the new accessors for the sg lists and the parameters. Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Acked-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-25 18:55:38 +02:00
scsi_set_resid(cp, resid);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
CMD_RESID_LEN(cp) = resid;
[SCSI] qla2xxx: Correct short-WRITE status handling. Properly check FC_RESID for any non-transfered bytes regardless of firmware completion status. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-12-06 19:58:06 +01:00
if (!lscsi_status &&
[SCSI] qla2xxx: convert to use the data buffer accessors - remove the unnecessary map_single path. - convert to use the new accessors for the sg lists and the parameters. Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Acked-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-25 18:55:38 +02:00
((unsigned)(scsi_bufflen(cp) - resid) <
[SCSI] qla2xxx: Correct short-WRITE status handling. Properly check FC_RESID for any non-transfered bytes regardless of firmware completion status. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-12-06 19:58:06 +01:00
cp->underflow)) {
qla_printk(KERN_INFO, ha,
[SCSI] qla2xxx: convert to use the data buffer accessors - remove the unnecessary map_single path. - convert to use the new accessors for the sg lists and the parameters. Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Acked-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-25 18:55:38 +02:00
"scsi(%ld:%d:%d:%d): Mid-layer underflow "
"detected (%x of %x bytes)...returning "
"error status.\n", ha->host_no,
cp->device->channel, cp->device->id,
cp->device->lun, resid,
scsi_bufflen(cp));
[SCSI] qla2xxx: Correct short-WRITE status handling. Properly check FC_RESID for any non-transfered bytes regardless of firmware completion status. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-12-06 19:58:06 +01:00
cp->result = DID_ERROR << 16;
break;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
cp->result = DID_OK << 16 | lscsi_status;
[SCSI] qla2xxx: Correct QUEUE_FULL handling. - Drop queue-depths across all luns for a given fcport during TASK_SET_FULL statuses. - Ramp-up I/Os after throttling. - Consolidate completion-status handling of CS_QUEUE_FULL with CS_COMPLETE as ISP24xx firmware no longer reports CS_QUEUE_FULL. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-10-13 18:33:39 +02:00
if (lscsi_status == SAM_STAT_TASK_SET_FULL) {
DEBUG2(printk(KERN_INFO
"scsi(%ld): QUEUE FULL status detected "
"0x%x-0x%x.\n", ha->host_no, comp_status,
scsi_status));
/* Adjust queue depth for all luns on the port. */
fcport->last_queue_full = jiffies;
starget_for_each_device(cp->device->sdev_target,
fcport, qla2x00_adjust_sdev_qdepth_down);
break;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (lscsi_status != SS_CHECK_CONDITION)
break;
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
/* Copy Sense Data into sense buffer. */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
memset(cp->sense_buffer, 0, sizeof(cp->sense_buffer));
if (!(scsi_status & SS_SENSE_LEN_VALID))
break;
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
if (sense_len >= sizeof(cp->sense_buffer))
sense_len = sizeof(cp->sense_buffer);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
CMD_ACTUAL_SNSLEN(cp) = sense_len;
sp->request_sense_length = sense_len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
sp->request_sense_ptr = cp->sense_buffer;
if (sp->request_sense_length > 32)
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
sense_len = 32;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
memcpy(cp->sense_buffer, sense_data, sense_len);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
sp->request_sense_ptr += sense_len;
sp->request_sense_length -= sense_len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (sp->request_sense_length != 0)
ha->status_srb = sp;
DEBUG5(printk("%s(): Check condition Sense data, "
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
"scsi(%ld:%d:%d:%d) cmd=%p pid=%ld\n", __func__,
ha->host_no, cp->device->channel, cp->device->id,
cp->device->lun, cp, cp->serial_number));
if (sense_len)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
DEBUG5(qla2x00_dump_buffer(cp->sense_buffer,
CMD_ACTUAL_SNSLEN(cp)));
break;
case CS_DATA_UNDERRUN:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
resid = resid_len;
[SCSI] qla2xxx: Use FW calculated residual count for underrun handling. With ISP24XX and ISP54XX parts. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-05-18 00:08:55 +02:00
/* Use F/W calculated residual length. */
[SCSI] qla2xxx: Generalize FW-Interface-2 support. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:05:56 +02:00
if (IS_FWI2_CAPABLE(ha))
[SCSI] qla2xxx: Use FW calculated residual count for underrun handling. With ISP24XX and ISP54XX parts. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-05-18 00:08:55 +02:00
resid = fw_resid_len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (scsi_status & SS_RESIDUAL_UNDER) {
[SCSI] qla2xxx: convert to use the data buffer accessors - remove the unnecessary map_single path. - convert to use the new accessors for the sg lists and the parameters. Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Acked-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-25 18:55:38 +02:00
scsi_set_resid(cp, resid);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
CMD_RESID_LEN(cp) = resid;
[SCSI] qla2xxx: Drop noisy 'UNDERRUN' status message. There's no point in displaying the message during a valid underrun case. Limit the message to potentially problematic cases. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-01-14 02:04:59 +01:00
} else {
DEBUG2(printk(KERN_INFO
"scsi(%ld:%d:%d) UNDERRUN status detected "
[SCSI] qla2xxx: Use FW calculated residual count for underrun handling. With ISP24XX and ISP54XX parts. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-05-18 00:08:55 +02:00
"0x%x-0x%x. resid=0x%x fw_resid=0x%x cdb=0x%x "
"os_underflow=0x%x\n", ha->host_no,
cp->device->id, cp->device->lun, comp_status,
scsi_status, resid_len, resid, cp->cmnd[0],
cp->underflow));
[SCSI] qla2xxx: Drop noisy 'UNDERRUN' status message. There's no point in displaying the message during a valid underrun case. Limit the message to potentially problematic cases. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-01-14 02:04:59 +01:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
/*
[SCSI] qla2xxx: Code scrubbing. Code scrubbing. - Remove trailing whitespace from driver files. - Remove unused #defines and inlines. - Standardize on C comments (// -> /* */) Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:32:07 +02:00
* Check to see if SCSI Status is non zero. If so report SCSI
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
* Status.
*/
if (lscsi_status != 0) {
cp->result = DID_OK << 16 | lscsi_status;
[SCSI] qla2xxx: Enable queue-full throttling when UNDERRUN detected. As ISP24xx firmware can return a CS_DATA_UNDERRUN completion status when the storage has returned a SAM_STAT_TASK_SET_FULL scsi-status. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-01-29 19:22:27 +01:00
if (lscsi_status == SAM_STAT_TASK_SET_FULL) {
DEBUG2(printk(KERN_INFO
"scsi(%ld): QUEUE FULL status detected "
"0x%x-0x%x.\n", ha->host_no, comp_status,
scsi_status));
/*
* Adjust queue depth for all luns on the
* port.
*/
fcport->last_queue_full = jiffies;
starget_for_each_device(
cp->device->sdev_target, fcport,
qla2x00_adjust_sdev_qdepth_down);
break;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (lscsi_status != SS_CHECK_CONDITION)
break;
/* Copy Sense Data into sense buffer */
memset(cp->sense_buffer, 0, sizeof(cp->sense_buffer));
if (!(scsi_status & SS_SENSE_LEN_VALID))
break;
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
if (sense_len >= sizeof(cp->sense_buffer))
sense_len = sizeof(cp->sense_buffer);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
CMD_ACTUAL_SNSLEN(cp) = sense_len;
sp->request_sense_length = sense_len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
sp->request_sense_ptr = cp->sense_buffer;
[SCSI] qla2xxx: Code scrubbing. Code scrubbing. - Remove trailing whitespace from driver files. - Remove unused #defines and inlines. - Standardize on C comments (// -> /* */) Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:32:07 +02:00
if (sp->request_sense_length > 32)
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
sense_len = 32;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
memcpy(cp->sense_buffer, sense_data, sense_len);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
sp->request_sense_ptr += sense_len;
sp->request_sense_length -= sense_len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (sp->request_sense_length != 0)
ha->status_srb = sp;
DEBUG5(printk("%s(): Check condition Sense data, "
"scsi(%ld:%d:%d:%d) cmd=%p pid=%ld\n",
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
__func__, ha->host_no, cp->device->channel,
cp->device->id, cp->device->lun, cp,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
cp->serial_number));
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
[SCSI] qla2xxx: Validate mid-layer 'underflow' during check-condition handling. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:05:59 +02:00
/*
* In case of a Underrun condition, set both the lscsi
* status and the completion status to appropriate
* values.
*/
if (resid &&
((unsigned)(cp->request_bufflen - resid) <
cp->underflow)) {
DEBUG2(qla_printk(KERN_INFO, ha,
"scsi(%ld:%d:%d:%d): Mid-layer underflow "
"detected (%x of %x bytes)...returning "
"error status.\n", ha->host_no,
cp->device->channel, cp->device->id,
cp->device->lun, resid,
cp->request_bufflen));
cp->result = DID_ERROR << 16 | lscsi_status;
}
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
if (sense_len)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
DEBUG5(qla2x00_dump_buffer(cp->sense_buffer,
CMD_ACTUAL_SNSLEN(cp)));
} else {
/*
* If RISC reports underrun and target does not report
* it then we must have a lost frame, so tell upper
* layer to retry it by reporting a bus busy.
*/
if (!(scsi_status & SS_RESIDUAL_UNDER)) {
DEBUG2(printk("scsi(%ld:%d:%d:%d) Dropped "
[SCSI] qla2xxx: convert to use the data buffer accessors - remove the unnecessary map_single path. - convert to use the new accessors for the sg lists and the parameters. Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Acked-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-25 18:55:38 +02:00
"frame(s) detected (%x of %x bytes)..."
"retrying command.\n", ha->host_no,
cp->device->channel, cp->device->id,
cp->device->lun, resid,
scsi_bufflen(cp)));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
cp->result = DID_BUS_BUSY << 16;
break;
}
/* Handle mid-layer underflow */
[SCSI] qla2xxx: convert to use the data buffer accessors - remove the unnecessary map_single path. - convert to use the new accessors for the sg lists and the parameters. Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Acked-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-25 18:55:38 +02:00
if ((unsigned)(scsi_bufflen(cp) - resid) <
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
cp->underflow) {
qla_printk(KERN_INFO, ha,
[SCSI] qla2xxx: convert to use the data buffer accessors - remove the unnecessary map_single path. - convert to use the new accessors for the sg lists and the parameters. Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Acked-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-25 18:55:38 +02:00
"scsi(%ld:%d:%d:%d): Mid-layer underflow "
"detected (%x of %x bytes)...returning "
"error status.\n", ha->host_no,
cp->device->channel, cp->device->id,
cp->device->lun, resid,
scsi_bufflen(cp));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
cp->result = DID_ERROR << 16;
break;
}
/* Everybody online, looking good... */
cp->result = DID_OK << 16;
}
break;
case CS_DATA_OVERRUN:
DEBUG2(printk(KERN_INFO
"scsi(%ld:%d:%d): OVERRUN status detected 0x%x-0x%x\n",
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
ha->host_no, cp->device->id, cp->device->lun, comp_status,
scsi_status));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
DEBUG2(printk(KERN_INFO
"CDB: 0x%x 0x%x 0x%x 0x%x 0x%x 0x%x\n",
cp->cmnd[0], cp->cmnd[1], cp->cmnd[2], cp->cmnd[3],
cp->cmnd[4], cp->cmnd[5]));
DEBUG2(printk(KERN_INFO
"PID=0x%lx req=0x%x xtra=0x%x -- returning DID_ERROR "
"status!\n",
[SCSI] qla2xxx: convert to use the data buffer accessors - remove the unnecessary map_single path. - convert to use the new accessors for the sg lists and the parameters. Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Acked-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-25 18:55:38 +02:00
cp->serial_number, scsi_bufflen(cp), resid_len));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
cp->result = DID_ERROR << 16;
break;
case CS_PORT_LOGGED_OUT:
case CS_PORT_CONFIG_CHG:
case CS_PORT_BUSY:
case CS_INCOMPLETE:
case CS_PORT_UNAVAILABLE:
/*
* If the port is in Target Down state, return all IOs for this
* Target with DID_NO_CONNECT ELSE Queue the IOs in the
* retry_queue.
*/
DEBUG2(printk("scsi(%ld:%d:%d): status_entry: Port Down "
"pid=%ld, compl status=0x%x, port state=0x%x\n",
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
ha->host_no, cp->device->id, cp->device->lun,
cp->serial_number, comp_status,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
atomic_read(&fcport->state)));
[PATCH] qla2xxx: remove internal queuing... Remove internal command queuing from the driver. As is, this driver cannot tolerate cable-pulls as I/Os will begin to fail by the upper layers. o Should be used in conjuction with the 11-fc_rport_adds_2.diff patch. o Removes qla_listops.h file -- no longer needed. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-17 22:02:26 +02:00
cp->result = DID_BUS_BUSY << 16;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
if (atomic_read(&fcport->state) == FCS_ONLINE) {
[SCSI] qla2xxx: Correct synchronization issues during rport addition/deletion. The driver can typically detect port-loss during an interrupt context (i.e. via interrogation of a status IOCB's completion status [CS_PORT_LOGGED_OUT]. Due to the calling requirements of the fc_rport APIs, the driver would defer removal of the device to the default workqueue. If the work-item was preceded by an event which caused the port to obtain visibility (relogin successful, target re-logged into the topology), deferred removal could inadvertently drop the rport. The code also no longer defers removal via the default workqueue, instead opting for use of the driver's own DPC thread. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-01-20 23:53:13 +01:00
qla2x00_mark_device_lost(ha, fcport, 1, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
break;
case CS_RESET:
DEBUG2(printk(KERN_INFO
"scsi(%ld): RESET status detected 0x%x-0x%x.\n",
ha->host_no, comp_status, scsi_status));
[PATCH] qla2xxx: remove internal queuing... Remove internal command queuing from the driver. As is, this driver cannot tolerate cable-pulls as I/Os will begin to fail by the upper layers. o Should be used in conjuction with the 11-fc_rport_adds_2.diff patch. o Removes qla_listops.h file -- no longer needed. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-17 22:02:26 +02:00
cp->result = DID_RESET << 16;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
break;
case CS_ABORTED:
[SCSI] qla2xxx: Code scrubbing. Code scrubbing. - Remove trailing whitespace from driver files. - Remove unused #defines and inlines. - Standardize on C comments (// -> /* */) Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:32:07 +02:00
/*
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
* hv2.19.12 - DID_ABORT does not retry the request if we
* aborted this request then abort otherwise it must be a
* reset.
*/
DEBUG2(printk(KERN_INFO
"scsi(%ld): ABORT status detected 0x%x-0x%x.\n",
ha->host_no, comp_status, scsi_status));
cp->result = DID_RESET << 16;
break;
case CS_TIMEOUT:
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
cp->result = DID_BUS_BUSY << 16;
[SCSI] qla2xxx: Generalize FW-Interface-2 support. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:05:56 +02:00
if (IS_FWI2_CAPABLE(ha)) {
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
DEBUG2(printk(KERN_INFO
"scsi(%ld:%d:%d:%d): TIMEOUT status detected "
"0x%x-0x%x\n", ha->host_no, cp->device->channel,
cp->device->id, cp->device->lun, comp_status,
scsi_status));
break;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
DEBUG2(printk(KERN_INFO
"scsi(%ld:%d:%d:%d): TIMEOUT status detected 0x%x-0x%x "
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
"sflags=%x.\n", ha->host_no, cp->device->channel,
cp->device->id, cp->device->lun, comp_status, scsi_status,
le16_to_cpu(sts->status_flags)));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
/* Check to see if logout occurred. */
if ((le16_to_cpu(sts->status_flags) & SF_LOGOUT_SENT))
[SCSI] qla2xxx: Correct synchronization issues during rport addition/deletion. The driver can typically detect port-loss during an interrupt context (i.e. via interrogation of a status IOCB's completion status [CS_PORT_LOGGED_OUT]. Due to the calling requirements of the fc_rport APIs, the driver would defer removal of the device to the default workqueue. If the work-item was preceded by an event which caused the port to obtain visibility (relogin successful, target re-logged into the topology), deferred removal could inadvertently drop the rport. The code also no longer defers removal via the default workqueue, instead opting for use of the driver's own DPC thread. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-01-20 23:53:13 +01:00
qla2x00_mark_device_lost(ha, fcport, 1, 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
break;
default:
DEBUG3(printk("scsi(%ld): Error detected (unknown status) "
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
"0x%x-0x%x.\n", ha->host_no, comp_status, scsi_status));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
qla_printk(KERN_INFO, ha,
"Unknown status detected 0x%x-0x%x.\n",
comp_status, scsi_status);
cp->result = DID_ERROR << 16;
break;
}
/* Place command on done queue. */
if (ha->status_srb == NULL)
[PATCH] qla2xxx: remove internal queuing... Remove internal command queuing from the driver. As is, this driver cannot tolerate cable-pulls as I/Os will begin to fail by the upper layers. o Should be used in conjuction with the 11-fc_rport_adds_2.diff patch. o Removes qla_listops.h file -- no longer needed. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-17 22:02:26 +02:00
qla2x00_sp_compl(ha, sp);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
/**
* 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(scsi_qla_host_t *ha, sts_cont_entry_t *pkt)
{
uint8_t sense_sz = 0;
srb_t *sp = ha->status_srb;
struct scsi_cmnd *cp;
if (sp != NULL && sp->request_sense_length != 0) {
cp = sp->cmd;
if (cp == NULL) {
DEBUG2(printk("%s(): Cmd already returned back to OS "
[SCSI] qla2xxx: Drop unused driver cruft. - structure definitions. - structure members. - #define's. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-05-18 00:09:22 +02:00
"sp=%p.\n", __func__, sp));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
qla_printk(KERN_INFO, ha,
"cmd is NULL: already returned to OS (sp=%p)\n",
[SCSI] qla2xxx: Code scrubbing. Code scrubbing. - Remove trailing whitespace from driver files. - Remove unused #defines and inlines. - Standardize on C comments (// -> /* */) Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:32:07 +02:00
sp);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
ha->status_srb = NULL;
return;
}
if (sp->request_sense_length > sizeof(pkt->data)) {
sense_sz = sizeof(pkt->data);
} else {
sense_sz = sp->request_sense_length;
}
/* Move sense data. */
[SCSI] qla2xxx: Generalize FW-Interface-2 support. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:05:56 +02:00
if (IS_FWI2_CAPABLE(ha))
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
host_to_fcp_swap(pkt->data, sizeof(pkt->data));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
memcpy(sp->request_sense_ptr, pkt->data, sense_sz);
DEBUG5(qla2x00_dump_buffer(sp->request_sense_ptr, sense_sz));
sp->request_sense_ptr += sense_sz;
sp->request_sense_length -= sense_sz;
/* Place command on done queue. */
if (sp->request_sense_length == 0) {
ha->status_srb = NULL;
[PATCH] qla2xxx: remove internal queuing... Remove internal command queuing from the driver. As is, this driver cannot tolerate cable-pulls as I/Os will begin to fail by the upper layers. o Should be used in conjuction with the 11-fc_rport_adds_2.diff patch. o Removes qla_listops.h file -- no longer needed. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-17 22:02:26 +02:00
qla2x00_sp_compl(ha, sp);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
}
}
/**
* qla2x00_error_entry() - Process an error entry.
* @ha: SCSI driver HA context
* @pkt: Entry pointer
*/
static void
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
qla2x00_error_entry(scsi_qla_host_t *ha, sts_entry_t *pkt)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
{
srb_t *sp;
#if defined(QL_DEBUG_LEVEL_2)
if (pkt->entry_status & RF_INV_E_ORDER)
qla_printk(KERN_ERR, ha, "%s: Invalid Entry Order\n", __func__);
else if (pkt->entry_status & RF_INV_E_COUNT)
qla_printk(KERN_ERR, ha, "%s: Invalid Entry Count\n", __func__);
else if (pkt->entry_status & RF_INV_E_PARAM)
[SCSI] qla2xxx: Code scrubbing. Code scrubbing. - Remove trailing whitespace from driver files. - Remove unused #defines and inlines. - Standardize on C comments (// -> /* */) Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:32:07 +02:00
qla_printk(KERN_ERR, ha,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
"%s: Invalid Entry Parameter\n", __func__);
else if (pkt->entry_status & RF_INV_E_TYPE)
qla_printk(KERN_ERR, ha, "%s: Invalid Entry Type\n", __func__);
else if (pkt->entry_status & RF_BUSY)
qla_printk(KERN_ERR, ha, "%s: Busy\n", __func__);
else
qla_printk(KERN_ERR, ha, "%s: UNKNOWN flag error\n", __func__);
#endif
/* Validate handle. */
if (pkt->handle < MAX_OUTSTANDING_COMMANDS)
sp = ha->outstanding_cmds[pkt->handle];
else
sp = NULL;
if (sp) {
/* Free outstanding command slot. */
ha->outstanding_cmds[pkt->handle] = NULL;
[SCSI] remove some dead code in qla2xxx Original from: Christoph Hellwig <hch@lst.de> Modified and Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-23 08:47:27 +02:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
/* Bad payload or header */
if (pkt->entry_status &
(RF_INV_E_ORDER | RF_INV_E_COUNT |
RF_INV_E_PARAM | RF_INV_E_TYPE)) {
sp->cmd->result = DID_ERROR << 16;
} else if (pkt->entry_status & RF_BUSY) {
sp->cmd->result = DID_BUS_BUSY << 16;
} else {
sp->cmd->result = DID_ERROR << 16;
}
[PATCH] qla2xxx: remove internal queuing... Remove internal command queuing from the driver. As is, this driver cannot tolerate cable-pulls as I/Os will begin to fail by the upper layers. o Should be used in conjuction with the 11-fc_rport_adds_2.diff patch. o Removes qla_listops.h file -- no longer needed. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-17 22:02:26 +02:00
qla2x00_sp_compl(ha, sp);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
} else if (pkt->entry_type == COMMAND_A64_TYPE || pkt->entry_type ==
COMMAND_TYPE || pkt->entry_type == COMMAND_TYPE_7) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
DEBUG2(printk("scsi(%ld): Error entry - invalid handle\n",
ha->host_no));
qla_printk(KERN_WARNING, ha,
"Error entry - invalid handle\n");
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
[SCSI] qla2xxx: use kthread_ API Use the kthread_ API instead of opencoding lots of hairy code for kernel thread creation and teardown. Also switch from semaphore-based thread wakeup to wake_up_process. Signed-off-by: Christoph Hellwig <hch@lst.de> Acked-By: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-02-14 18:46:22 +01:00
qla2xxx_wake_dpc(ha);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
}
/**
* qla2x00_ms_entry() - Process a Management Server entry.
* @ha: SCSI driver HA context
* @index: Response queue out pointer
*/
static void
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
qla2x00_ms_entry(scsi_qla_host_t *ha, ms_iocb_entry_t *pkt)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
{
srb_t *sp;
DEBUG3(printk("%s(%ld): pkt=%p pkthandle=%d.\n",
__func__, ha->host_no, pkt, pkt->handle1));
/* Validate handle. */
if (pkt->handle1 < MAX_OUTSTANDING_COMMANDS)
sp = ha->outstanding_cmds[pkt->handle1];
else
sp = NULL;
if (sp == NULL) {
DEBUG2(printk("scsi(%ld): MS entry - invalid handle\n",
ha->host_no));
qla_printk(KERN_WARNING, ha, "MS entry - invalid handle\n");
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
return;
}
CMD_COMPL_STATUS(sp->cmd) = le16_to_cpu(pkt->status);
CMD_ENTRY_STATUS(sp->cmd) = pkt->entry_status;
/* Free outstanding command slot. */
ha->outstanding_cmds[pkt->handle1] = NULL;
[PATCH] qla2xxx: remove internal queuing... Remove internal command queuing from the driver. As is, this driver cannot tolerate cable-pulls as I/Os will begin to fail by the upper layers. o Should be used in conjuction with the 11-fc_rport_adds_2.diff patch. o Removes qla_listops.h file -- no longer needed. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-17 22:02:26 +02:00
qla2x00_sp_compl(ha, sp);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 00:20:36 +02:00
}
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
/**
* qla24xx_mbx_completion() - Process mailbox command completions.
* @ha: SCSI driver HA context
* @mb0: Mailbox0 register
*/
static void
qla24xx_mbx_completion(scsi_qla_host_t *ha, uint16_t mb0)
{
uint16_t cnt;
uint16_t __iomem *wptr;
struct device_reg_24xx __iomem *reg = &ha->iobase->isp24;
/* Load return mailbox registers. */
ha->flags.mbox_int = 1;
ha->mailbox_out[0] = mb0;
wptr = (uint16_t __iomem *)&reg->mailbox1;
for (cnt = 1; cnt < ha->mbx_count; cnt++) {
ha->mailbox_out[cnt] = RD_REG_WORD(wptr);
wptr++;
}
if (ha->mcp) {
DEBUG3(printk("%s(%ld): Got mailbox completion. cmd=%x.\n",
__func__, ha->host_no, ha->mcp->mb[0]));
} else {
DEBUG2_3(printk("%s(%ld): MBX pointer ERROR!\n",
__func__, ha->host_no));
}
}
/**
* qla24xx_process_response_queue() - Process response queue entries.
* @ha: SCSI driver HA context
*/
void
qla24xx_process_response_queue(struct scsi_qla_host *ha)
{
struct device_reg_24xx __iomem *reg = &ha->iobase->isp24;
struct sts_entry_24xx *pkt;
if (!ha->flags.online)
return;
while (ha->response_ring_ptr->signature != RESPONSE_PROCESSED) {
pkt = (struct sts_entry_24xx *)ha->response_ring_ptr;
ha->rsp_ring_index++;
if (ha->rsp_ring_index == ha->response_q_length) {
ha->rsp_ring_index = 0;
ha->response_ring_ptr = ha->response_ring;
} else {
ha->response_ring_ptr++;
}
if (pkt->entry_status != 0) {
DEBUG3(printk(KERN_INFO
"scsi(%ld): Process error entry.\n", ha->host_no));
qla2x00_error_entry(ha, (sts_entry_t *) pkt);
((response_t *)pkt)->signature = RESPONSE_PROCESSED;
wmb();
continue;
}
switch (pkt->entry_type) {
case STATUS_TYPE:
qla2x00_status_entry(ha, pkt);
break;
case STATUS_CONT_TYPE:
qla2x00_status_cont_entry(ha, (sts_cont_entry_t *)pkt);
break;
case MS_IOCB_TYPE:
qla24xx_ms_entry(ha, (struct ct_entry_24xx *)pkt);
break;
[SCSI] qla2xxx: add support for NPIV Following patch adds support for NPIV (N-Port ID Virtualization) to the qla2xxx. - supported within switched-fabric topologies only. - supports up to 63 virtual ports on each physical port. Signed-off-by: Seokmann Ju <seokmann.ju@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-05 22:16:51 +02:00
case VP_RPT_ID_IOCB_TYPE:
qla24xx_report_id_acquisition(ha,
(struct vp_rpt_id_entry_24xx *)pkt);
break;
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
default:
/* Type Not Supported. */
DEBUG4(printk(KERN_WARNING
"scsi(%ld): Received unknown response pkt type %x "
"entry status=%x.\n",
ha->host_no, pkt->entry_type, pkt->entry_status));
break;
}
((response_t *)pkt)->signature = RESPONSE_PROCESSED;
wmb();
}
/* Adjust ring index */
WRT_REG_DWORD(&reg->rsp_q_out, ha->rsp_ring_index);
}
/**
* qla24xx_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
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 15:55:46 +02:00
qla24xx_intr_handler(int irq, void *dev_id)
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
{
scsi_qla_host_t *ha;
struct device_reg_24xx __iomem *reg;
int status;
unsigned long flags;
unsigned long iter;
uint32_t stat;
uint32_t hccr;
uint16_t mb[4];
ha = (scsi_qla_host_t *) dev_id;
if (!ha) {
printk(KERN_INFO
"%s(): NULL host pointer\n", __func__);
return IRQ_NONE;
}
reg = &ha->iobase->isp24;
status = 0;
spin_lock_irqsave(&ha->hardware_lock, flags);
for (iter = 50; iter--; ) {
stat = RD_REG_DWORD(&reg->host_status);
if (stat & HSRX_RISC_PAUSED) {
hccr = RD_REG_DWORD(&reg->hccr);
qla_printk(KERN_INFO, ha, "RISC paused -- HCCR=%x, "
"Dumping firmware!\n", hccr);
[SCSI] qla2xxx: Re-factor isp_operations to static structures. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:06:00 +02:00
ha->isp_ops->fw_dump(ha, 1);
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
break;
} else if ((stat & HSRX_RISC_INT) == 0)
break;
switch (stat & 0xff) {
case 0x1:
case 0x2:
case 0x10:
case 0x11:
qla24xx_mbx_completion(ha, MSW(stat));
status |= MBX_INTERRUPT;
break;
case 0x12:
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(ha, mb);
break;
case 0x13:
qla24xx_process_response_queue(ha);
break;
default:
DEBUG2(printk("scsi(%ld): Unrecognized interrupt type "
"(%d).\n",
ha->host_no, stat & 0xff));
break;
}
WRT_REG_DWORD(&reg->hccr, HCCRX_CLR_RISC_INT);
RD_REG_DWORD_RELAXED(&reg->hccr);
}
spin_unlock_irqrestore(&ha->hardware_lock, flags);
if (test_bit(MBX_INTR_WAIT, &ha->mbx_cmd_flags) &&
(status & MBX_INTERRUPT) && ha->flags.mbox_int) {
set_bit(MBX_INTERRUPT, &ha->mbx_cmd_flags);
up(&ha->mbx_intr_sem);
}
return IRQ_HANDLED;
}
/**
* qla24xx_ms_entry() - Process a Management Server entry.
* @ha: SCSI driver HA context
* @index: Response queue out pointer
*/
static void
qla24xx_ms_entry(scsi_qla_host_t *ha, struct ct_entry_24xx *pkt)
{
srb_t *sp;
DEBUG3(printk("%s(%ld): pkt=%p pkthandle=%d.\n",
__func__, ha->host_no, pkt, pkt->handle));
[SCSI] qla2xxx: Cleanup DEBUG macro usage. - macro usage statements should terminate with a ';' - remove unused macros. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-06-24 01:11:05 +02:00
DEBUG9(printk("%s: ct pkt dump:\n", __func__));
DEBUG9(qla2x00_dump_buffer((void *)pkt, sizeof(struct ct_entry_24xx)));
[SCSI] qla2xxx: Add ISP24xx ISR routines. Add ISP24xx ISR routines. Add appropriate glue-code for ISP24xx support -- this included generalizing some of the core handling routines (qla2x00_async_event() [pull-up retrieval of mailbox values] and qla2x00_status_entry()]. Fixup 2100/2300 ISRs to handle the new conventions. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-07-06 19:31:27 +02:00
/* Validate handle. */
if (pkt->handle < MAX_OUTSTANDING_COMMANDS)
sp = ha->outstanding_cmds[pkt->handle];
else
sp = NULL;
if (sp == NULL) {
DEBUG2(printk("scsi(%ld): MS entry - invalid handle\n",
ha->host_no));
DEBUG10(printk("scsi(%ld): MS entry - invalid handle\n",
ha->host_no));
qla_printk(KERN_WARNING, ha, "MS entry - invalid handle %d\n",
pkt->handle);
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
return;
}
CMD_COMPL_STATUS(sp->cmd) = le16_to_cpu(pkt->comp_status);
CMD_ENTRY_STATUS(sp->cmd) = pkt->entry_status;
/* Free outstanding command slot. */
ha->outstanding_cmds[pkt->handle] = NULL;
qla2x00_sp_compl(ha, sp);
}
[SCSI] qla2xxx: Add MSI-X support. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-01-29 19:22:19 +01:00
static irqreturn_t
qla24xx_msix_rsp_q(int irq, void *dev_id)
{
scsi_qla_host_t *ha;
struct device_reg_24xx __iomem *reg;
unsigned long flags;
ha = dev_id;
reg = &ha->iobase->isp24;
spin_lock_irqsave(&ha->hardware_lock, flags);
qla24xx_process_response_queue(ha);
WRT_REG_DWORD(&reg->hccr, HCCRX_CLR_RISC_INT);
RD_REG_DWORD_RELAXED(&reg->hccr);
spin_unlock_irqrestore(&ha->hardware_lock, flags);
return IRQ_HANDLED;
}
static irqreturn_t
qla24xx_msix_default(int irq, void *dev_id)
{
scsi_qla_host_t *ha;
struct device_reg_24xx __iomem *reg;
int status;
unsigned long flags;
unsigned long iter;
uint32_t stat;
uint32_t hccr;
uint16_t mb[4];
ha = dev_id;
reg = &ha->iobase->isp24;
status = 0;
spin_lock_irqsave(&ha->hardware_lock, flags);
for (iter = 50; iter--; ) {
stat = RD_REG_DWORD(&reg->host_status);
if (stat & HSRX_RISC_PAUSED) {
hccr = RD_REG_DWORD(&reg->hccr);
qla_printk(KERN_INFO, ha, "RISC paused -- HCCR=%x, "
"Dumping firmware!\n", hccr);
[SCSI] qla2xxx: Re-factor isp_operations to static structures. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:06:00 +02:00
ha->isp_ops->fw_dump(ha, 1);
[SCSI] qla2xxx: Add MSI-X support. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-01-29 19:22:19 +01:00
set_bit(ISP_ABORT_NEEDED, &ha->dpc_flags);
break;
} else if ((stat & HSRX_RISC_INT) == 0)
break;
switch (stat & 0xff) {
case 0x1:
case 0x2:
case 0x10:
case 0x11:
qla24xx_mbx_completion(ha, MSW(stat));
status |= MBX_INTERRUPT;
break;
case 0x12:
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(ha, mb);
break;
case 0x13:
qla24xx_process_response_queue(ha);
break;
default:
DEBUG2(printk("scsi(%ld): Unrecognized interrupt type "
"(%d).\n",
ha->host_no, stat & 0xff));
break;
}
WRT_REG_DWORD(&reg->hccr, HCCRX_CLR_RISC_INT);
RD_REG_DWORD_RELAXED(&reg->hccr);
}
spin_unlock_irqrestore(&ha->hardware_lock, flags);
if (test_bit(MBX_INTR_WAIT, &ha->mbx_cmd_flags) &&
(status & MBX_INTERRUPT) && ha->flags.mbox_int) {
set_bit(MBX_INTERRUPT, &ha->mbx_cmd_flags);
up(&ha->mbx_intr_sem);
}
return IRQ_HANDLED;
}
/* Interrupt handling helpers. */
struct qla_init_msix_entry {
uint16_t entry;
uint16_t index;
const char *name;
[SCSI] aacraid,qla2xxx: use irq_handler_t where appropriate Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Acked-by: "Salyzyn, Mark" <mark_salyzyn@adaptec.com> Acked-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-23 23:41:44 +02:00
irq_handler_t handler;
[SCSI] qla2xxx: Add MSI-X support. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-01-29 19:22:19 +01:00
};
static struct qla_init_msix_entry imsix_entries[QLA_MSIX_ENTRIES] = {
{ QLA_MSIX_DEFAULT, QLA_MIDX_DEFAULT,
"qla2xxx (default)", qla24xx_msix_default },
{ QLA_MSIX_RSP_Q, QLA_MIDX_RSP_Q,
"qla2xxx (rsp_q)", qla24xx_msix_rsp_q },
};
static void
qla24xx_disable_msix(scsi_qla_host_t *ha)
{
int i;
struct qla_msix_entry *qentry;
for (i = 0; i < QLA_MSIX_ENTRIES; i++) {
qentry = &ha->msix_entries[imsix_entries[i].index];
if (qentry->have_irq)
free_irq(qentry->msix_vector, ha);
}
pci_disable_msix(ha->pdev);
}
static int
qla24xx_enable_msix(scsi_qla_host_t *ha)
{
int i, ret;
struct msix_entry entries[QLA_MSIX_ENTRIES];
struct qla_msix_entry *qentry;
for (i = 0; i < QLA_MSIX_ENTRIES; i++)
entries[i].entry = imsix_entries[i].entry;
ret = pci_enable_msix(ha->pdev, entries, ARRAY_SIZE(entries));
if (ret) {
qla_printk(KERN_WARNING, ha,
"MSI-X: Failed to enable support -- %d/%d\n",
QLA_MSIX_ENTRIES, ret);
goto msix_out;
}
ha->flags.msix_enabled = 1;
for (i = 0; i < QLA_MSIX_ENTRIES; i++) {
qentry = &ha->msix_entries[imsix_entries[i].index];
qentry->msix_vector = entries[i].vector;
qentry->msix_entry = entries[i].entry;
qentry->have_irq = 0;
ret = request_irq(qentry->msix_vector,
imsix_entries[i].handler, 0, imsix_entries[i].name, ha);
if (ret) {
qla_printk(KERN_WARNING, ha,
"MSI-X: Unable to register handler -- %x/%d.\n",
imsix_entries[i].index, ret);
qla24xx_disable_msix(ha);
goto msix_out;
}
qentry->have_irq = 1;
}
msix_out:
return ret;
}
int
qla2x00_request_irqs(scsi_qla_host_t *ha)
{
int ret;
/* If possible, enable MSI-X. */
if (!IS_QLA2432(ha))
goto skip_msix;
if (ha->chip_revision < QLA_MSIX_CHIP_REV_24XX ||
!QLA_MSIX_FW_MODE_1(ha->fw_attributes)) {
DEBUG2(qla_printk(KERN_WARNING, ha,
"MSI-X: Unsupported ISP2432 (0x%X, 0x%X).\n",
ha->chip_revision, ha->fw_attributes));
goto skip_msix;
}
ret = qla24xx_enable_msix(ha);
if (!ret) {
DEBUG2(qla_printk(KERN_INFO, ha,
"MSI-X: Enabled (0x%X, 0x%X).\n", ha->chip_revision,
ha->fw_attributes));
return ret;
}
qla_printk(KERN_WARNING, ha,
"MSI-X: Falling back-to INTa mode -- %d.\n", ret);
skip_msix:
[SCSI] qla2xxx: Add MSI support. Supported ISP types include ISP2422 and revision-2 type ISP2432 chips. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-07 16:43:02 +02:00
if (!IS_QLA24XX(ha))
goto skip_msi;
ret = pci_enable_msi(ha->pdev);
if (!ret) {
DEBUG2(qla_printk(KERN_INFO, ha, "MSI: Enabled.\n"));
ha->flags.msi_enabled = 1;
}
skip_msi:
[SCSI] qla2xxx: Re-factor isp_operations to static structures. In preparation for new ISP types. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-07-20 00:06:00 +02:00
ret = request_irq(ha->pdev->irq, ha->isp_ops->intr_handler,
[SCSI] qla2xxx: Add MSI-X support. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-01-29 19:22:19 +01:00
IRQF_DISABLED|IRQF_SHARED, QLA2XXX_DRIVER_NAME, ha);
[SCSI] qla2xxx: Handle IRQ-0 assignments by the system. No restriction should be placed on the IRQ number assigned to a given ISP. Original code incorrectly assumed a non-zero IRQ number assignment by the system. In these circumstances the proper freeing of the IRQ (via free_irq()) would not take place. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-01-29 19:22:20 +01:00
if (!ret) {
ha->flags.inta_enabled = 1;
ha->host->irq = ha->pdev->irq;
} else {
[SCSI] qla2xxx: Add MSI-X support. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-01-29 19:22:19 +01:00
qla_printk(KERN_WARNING, ha,
"Failed to reserve interrupt %d already in use.\n",
ha->pdev->irq);
}
return ret;
}
void
qla2x00_free_irqs(scsi_qla_host_t *ha)
{
if (ha->flags.msix_enabled)
qla24xx_disable_msix(ha);
[SCSI] qla2xxx: Add MSI support. Supported ISP types include ISP2422 and revision-2 type ISP2432 chips. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-07 16:43:02 +02:00
else if (ha->flags.inta_enabled) {
[SCSI] qla2xxx: Add MSI-X support. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-01-29 19:22:19 +01:00
free_irq(ha->host->irq, ha);
[SCSI] qla2xxx: Add MSI support. Supported ISP types include ISP2422 and revision-2 type ISP2432 chips. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-05-07 16:43:02 +02:00
pci_disable_msi(ha->pdev);
}
[SCSI] qla2xxx: Add MSI-X support. Signed-off-by: Andrew Vasquez <andrew.vasquez@qlogic.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-01-29 19:22:19 +01:00
}
Reference in a new issue Copy permalink