linux/drivers/scsi/bfa/bfad_im.c

1253 lines
31 KiB
C
Raw Normal View History

/*
* Copyright (c) 2005-2009 Brocade Communications Systems, Inc.
* All rights reserved
* www.brocade.com
*
* Linux driver for Brocade Fibre Channel Host Bus Adapter.
*
* This program is free software; you can redistribute it and/or modify it
* under the terms of the GNU General Public License (GPL) Version 2 as
* published by the Free Software Foundation
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*/
/**
* bfad_im.c Linux driver IM module.
*/
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/slab.h>
#include "bfad_drv.h"
#include "bfad_im.h"
#include "bfad_trcmod.h"
#include "bfa_cb_ioim_macros.h"
#include <fcb/bfa_fcb_fcpim.h>
BFA_TRC_FILE(LDRV, IM);
DEFINE_IDR(bfad_im_port_index);
struct scsi_transport_template *bfad_im_scsi_transport_template;
struct scsi_transport_template *bfad_im_scsi_vport_transport_template;
static void bfad_im_itnim_work_handler(struct work_struct *work);
static int bfad_im_queuecommand(struct scsi_cmnd *cmnd,
void (*done)(struct scsi_cmnd *));
static int bfad_im_slave_alloc(struct scsi_device *sdev);
void
bfa_cb_ioim_done(void *drv, struct bfad_ioim_s *dio,
enum bfi_ioim_status io_status, u8 scsi_status,
int sns_len, u8 *sns_info, s32 residue)
{
struct scsi_cmnd *cmnd = (struct scsi_cmnd *)dio;
struct bfad_s *bfad = drv;
struct bfad_itnim_data_s *itnim_data;
struct bfad_itnim_s *itnim;
u8 host_status = DID_OK;
switch (io_status) {
case BFI_IOIM_STS_OK:
bfa_trc(bfad, scsi_status);
scsi_set_resid(cmnd, 0);
if (sns_len > 0) {
bfa_trc(bfad, sns_len);
if (sns_len > SCSI_SENSE_BUFFERSIZE)
sns_len = SCSI_SENSE_BUFFERSIZE;
memcpy(cmnd->sense_buffer, sns_info, sns_len);
}
if (residue > 0) {
bfa_trc(bfad, residue);
scsi_set_resid(cmnd, residue);
if (!sns_len && (scsi_status == SAM_STAT_GOOD) &&
(scsi_bufflen(cmnd) - residue) <
cmnd->underflow) {
bfa_trc(bfad, 0);
host_status = DID_ERROR;
}
}
cmnd->result = ScsiResult(host_status, scsi_status);
break;
case BFI_IOIM_STS_ABORTED:
case BFI_IOIM_STS_TIMEDOUT:
case BFI_IOIM_STS_PATHTOV:
default:
cmnd->result = ScsiResult(DID_ERROR, 0);
}
/* Unmap DMA, if host is NULL, it means a scsi passthru cmd */
if (cmnd->device->host != NULL)
scsi_dma_unmap(cmnd);
cmnd->host_scribble = NULL;
bfa_trc(bfad, cmnd->result);
itnim_data = cmnd->device->hostdata;
if (itnim_data) {
itnim = itnim_data->itnim;
if (!cmnd->result && itnim &&
(bfa_lun_queue_depth > cmnd->device->queue_depth)) {
/* Queue depth adjustment for good status completion */
bfad_os_ramp_up_qdepth(itnim, cmnd->device);
} else if (cmnd->result == SAM_STAT_TASK_SET_FULL && itnim) {
/* qfull handling */
bfad_os_handle_qfull(itnim, cmnd->device);
}
}
cmnd->scsi_done(cmnd);
}
void
bfa_cb_ioim_good_comp(void *drv, struct bfad_ioim_s *dio)
{
struct scsi_cmnd *cmnd = (struct scsi_cmnd *)dio;
struct bfad_itnim_data_s *itnim_data;
struct bfad_itnim_s *itnim;
cmnd->result = ScsiResult(DID_OK, SCSI_STATUS_GOOD);
/* Unmap DMA, if host is NULL, it means a scsi passthru cmd */
if (cmnd->device->host != NULL)
scsi_dma_unmap(cmnd);
cmnd->host_scribble = NULL;
/* Queue depth adjustment */
if (bfa_lun_queue_depth > cmnd->device->queue_depth) {
itnim_data = cmnd->device->hostdata;
if (itnim_data) {
itnim = itnim_data->itnim;
if (itnim)
bfad_os_ramp_up_qdepth(itnim, cmnd->device);
}
}
cmnd->scsi_done(cmnd);
}
void
bfa_cb_ioim_abort(void *drv, struct bfad_ioim_s *dio)
{
struct scsi_cmnd *cmnd = (struct scsi_cmnd *)dio;
struct bfad_s *bfad = drv;
cmnd->result = ScsiResult(DID_ERROR, 0);
/* Unmap DMA, if host is NULL, it means a scsi passthru cmd */
if (cmnd->device->host != NULL)
scsi_dma_unmap(cmnd);
bfa_trc(bfad, cmnd->result);
cmnd->host_scribble = NULL;
}
void
bfa_cb_tskim_done(void *bfad, struct bfad_tskim_s *dtsk,
enum bfi_tskim_status tsk_status)
{
struct scsi_cmnd *cmnd = (struct scsi_cmnd *)dtsk;
wait_queue_head_t *wq;
cmnd->SCp.Status |= tsk_status << 1;
set_bit(IO_DONE_BIT, (unsigned long *)&cmnd->SCp.Status);
wq = (wait_queue_head_t *) cmnd->SCp.ptr;
cmnd->SCp.ptr = NULL;
if (wq)
wake_up(wq);
}
void
bfa_cb_ioim_resfree(void *drv)
{
}
/**
* Scsi_Host_template SCSI host template
*/
/**
* Scsi_Host template entry, returns BFAD PCI info.
*/
static const char *
bfad_im_info(struct Scsi_Host *shost)
{
static char bfa_buf[256];
struct bfad_im_port_s *im_port =
(struct bfad_im_port_s *) shost->hostdata[0];
struct bfad_s *bfad = im_port->bfad;
char model[BFA_ADAPTER_MODEL_NAME_LEN];
bfa_get_adapter_model(&bfad->bfa, model);
memset(bfa_buf, 0, sizeof(bfa_buf));
snprintf(bfa_buf, sizeof(bfa_buf),
"Brocade FC/FCOE Adapter, " "model: %s hwpath: %s driver: %s",
model, bfad->pci_name, BFAD_DRIVER_VERSION);
return bfa_buf;
}
/**
* Scsi_Host template entry, aborts the specified SCSI command.
*
* Returns: SUCCESS or FAILED.
*/
static int
bfad_im_abort_handler(struct scsi_cmnd *cmnd)
{
struct Scsi_Host *shost = cmnd->device->host;
struct bfad_im_port_s *im_port =
(struct bfad_im_port_s *) shost->hostdata[0];
struct bfad_s *bfad = im_port->bfad;
struct bfa_ioim_s *hal_io;
unsigned long flags;
u32 timeout;
int rc = FAILED;
spin_lock_irqsave(&bfad->bfad_lock, flags);
hal_io = (struct bfa_ioim_s *) cmnd->host_scribble;
if (!hal_io) {
/* IO has been completed, retrun success */
rc = SUCCESS;
goto out;
}
if (hal_io->dio != (struct bfad_ioim_s *) cmnd) {
rc = FAILED;
goto out;
}
bfa_trc(bfad, hal_io->iotag);
bfa_log(bfad->logmod, BFA_LOG_LINUX_SCSI_ABORT,
im_port->shost->host_no, cmnd, hal_io->iotag);
bfa_ioim_abort(hal_io);
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
/* Need to wait until the command get aborted */
timeout = 10;
while ((struct bfa_ioim_s *) cmnd->host_scribble == hal_io) {
set_current_state(TASK_UNINTERRUPTIBLE);
schedule_timeout(timeout);
if (timeout < 4 * HZ)
timeout *= 2;
}
cmnd->scsi_done(cmnd);
bfa_trc(bfad, hal_io->iotag);
bfa_log(bfad->logmod, BFA_LOG_LINUX_SCSI_ABORT_COMP,
im_port->shost->host_no, cmnd, hal_io->iotag);
return SUCCESS;
out:
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
return rc;
}
static bfa_status_t
bfad_im_target_reset_send(struct bfad_s *bfad, struct scsi_cmnd *cmnd,
struct bfad_itnim_s *itnim)
{
struct bfa_tskim_s *tskim;
struct bfa_itnim_s *bfa_itnim;
bfa_status_t rc = BFA_STATUS_OK;
tskim = bfa_tskim_alloc(&bfad->bfa, (struct bfad_tskim_s *) cmnd);
if (!tskim) {
BFA_DEV_PRINTF(bfad, BFA_ERR,
"target reset, fail to allocate tskim\n");
rc = BFA_STATUS_FAILED;
goto out;
}
/*
* Set host_scribble to NULL to avoid aborting a task command if
* happens.
*/
cmnd->host_scribble = NULL;
cmnd->SCp.Status = 0;
bfa_itnim = bfa_fcs_itnim_get_halitn(&itnim->fcs_itnim);
bfa_tskim_start(tskim, bfa_itnim, (lun_t)0,
FCP_TM_TARGET_RESET, BFAD_TARGET_RESET_TMO);
out:
return rc;
}
/**
* Scsi_Host template entry, resets a LUN and abort its all commands.
*
* Returns: SUCCESS or FAILED.
*
*/
static int
bfad_im_reset_lun_handler(struct scsi_cmnd *cmnd)
{
struct Scsi_Host *shost = cmnd->device->host;
struct bfad_im_port_s *im_port =
(struct bfad_im_port_s *) shost->hostdata[0];
struct bfad_itnim_data_s *itnim_data = cmnd->device->hostdata;
struct bfad_s *bfad = im_port->bfad;
struct bfa_tskim_s *tskim;
struct bfad_itnim_s *itnim;
struct bfa_itnim_s *bfa_itnim;
DECLARE_WAIT_QUEUE_HEAD(wq);
int rc = SUCCESS;
unsigned long flags;
enum bfi_tskim_status task_status;
spin_lock_irqsave(&bfad->bfad_lock, flags);
itnim = itnim_data->itnim;
if (!itnim) {
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
rc = FAILED;
goto out;
}
tskim = bfa_tskim_alloc(&bfad->bfa, (struct bfad_tskim_s *) cmnd);
if (!tskim) {
BFA_DEV_PRINTF(bfad, BFA_ERR,
"LUN reset, fail to allocate tskim");
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
rc = FAILED;
goto out;
}
/**
* Set host_scribble to NULL to avoid aborting a task command
* if happens.
*/
cmnd->host_scribble = NULL;
cmnd->SCp.ptr = (char *)&wq;
cmnd->SCp.Status = 0;
bfa_itnim = bfa_fcs_itnim_get_halitn(&itnim->fcs_itnim);
bfa_tskim_start(tskim, bfa_itnim,
bfad_int_to_lun(cmnd->device->lun),
FCP_TM_LUN_RESET, BFAD_LUN_RESET_TMO);
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
wait_event(wq, test_bit(IO_DONE_BIT,
(unsigned long *)&cmnd->SCp.Status));
task_status = cmnd->SCp.Status >> 1;
if (task_status != BFI_TSKIM_STS_OK) {
BFA_DEV_PRINTF(bfad, BFA_ERR, "LUN reset failure, status: %d\n",
task_status);
rc = FAILED;
}
out:
return rc;
}
/**
* Scsi_Host template entry, resets the bus and abort all commands.
*/
static int
bfad_im_reset_bus_handler(struct scsi_cmnd *cmnd)
{
struct Scsi_Host *shost = cmnd->device->host;
struct bfad_im_port_s *im_port =
(struct bfad_im_port_s *) shost->hostdata[0];
struct bfad_s *bfad = im_port->bfad;
struct bfad_itnim_s *itnim;
unsigned long flags;
u32 i, rc, err_cnt = 0;
DECLARE_WAIT_QUEUE_HEAD(wq);
enum bfi_tskim_status task_status;
spin_lock_irqsave(&bfad->bfad_lock, flags);
for (i = 0; i < MAX_FCP_TARGET; i++) {
itnim = bfad_os_get_itnim(im_port, i);
if (itnim) {
cmnd->SCp.ptr = (char *)&wq;
rc = bfad_im_target_reset_send(bfad, cmnd, itnim);
if (rc != BFA_STATUS_OK) {
err_cnt++;
continue;
}
/* wait target reset to complete */
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
wait_event(wq, test_bit(IO_DONE_BIT,
(unsigned long *)&cmnd->SCp.Status));
spin_lock_irqsave(&bfad->bfad_lock, flags);
task_status = cmnd->SCp.Status >> 1;
if (task_status != BFI_TSKIM_STS_OK) {
BFA_DEV_PRINTF(bfad, BFA_ERR,
"target reset failure,"
" status: %d\n", task_status);
err_cnt++;
}
}
}
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
if (err_cnt)
return FAILED;
return SUCCESS;
}
/**
* Scsi_Host template entry slave_destroy.
*/
static void
bfad_im_slave_destroy(struct scsi_device *sdev)
{
sdev->hostdata = NULL;
return;
}
/**
* BFA FCS itnim callbacks
*/
/**
* BFA FCS itnim alloc callback, after successful PRLI
* Context: Interrupt
*/
void
bfa_fcb_itnim_alloc(struct bfad_s *bfad, struct bfa_fcs_itnim_s **itnim,
struct bfad_itnim_s **itnim_drv)
{
*itnim_drv = kzalloc(sizeof(struct bfad_itnim_s), GFP_ATOMIC);
if (*itnim_drv == NULL)
return;
(*itnim_drv)->im = bfad->im;
*itnim = &(*itnim_drv)->fcs_itnim;
(*itnim_drv)->state = ITNIM_STATE_NONE;
/*
* Initiaze the itnim_work
*/
INIT_WORK(&(*itnim_drv)->itnim_work, bfad_im_itnim_work_handler);
bfad->bfad_flags |= BFAD_RPORT_ONLINE;
}
/**
* BFA FCS itnim free callback.
* Context: Interrupt. bfad_lock is held
*/
void
bfa_fcb_itnim_free(struct bfad_s *bfad, struct bfad_itnim_s *itnim_drv)
{
struct bfad_port_s *port;
wwn_t wwpn;
u32 fcid;
char wwpn_str[32], fcid_str[16];
/* online to free state transtion should not happen */
bfa_assert(itnim_drv->state != ITNIM_STATE_ONLINE);
itnim_drv->queue_work = 1;
/* offline request is not yet done, use the same request to free */
if (itnim_drv->state == ITNIM_STATE_OFFLINE_PENDING)
itnim_drv->queue_work = 0;
itnim_drv->state = ITNIM_STATE_FREE;
port = bfa_fcs_itnim_get_drvport(&itnim_drv->fcs_itnim);
itnim_drv->im_port = port->im_port;
wwpn = bfa_fcs_itnim_get_pwwn(&itnim_drv->fcs_itnim);
fcid = bfa_fcs_itnim_get_fcid(&itnim_drv->fcs_itnim);
wwn2str(wwpn_str, wwpn);
fcid2str(fcid_str, fcid);
bfa_log(bfad->logmod, BFA_LOG_LINUX_ITNIM_FREE,
port->im_port->shost->host_no,
fcid_str, wwpn_str);
bfad_os_itnim_process(itnim_drv);
}
/**
* BFA FCS itnim online callback.
* Context: Interrupt. bfad_lock is held
*/
void
bfa_fcb_itnim_online(struct bfad_itnim_s *itnim_drv)
{
struct bfad_port_s *port;
itnim_drv->bfa_itnim = bfa_fcs_itnim_get_halitn(&itnim_drv->fcs_itnim);
port = bfa_fcs_itnim_get_drvport(&itnim_drv->fcs_itnim);
itnim_drv->state = ITNIM_STATE_ONLINE;
itnim_drv->queue_work = 1;
itnim_drv->im_port = port->im_port;
bfad_os_itnim_process(itnim_drv);
}
/**
* BFA FCS itnim offline callback.
* Context: Interrupt. bfad_lock is held
*/
void
bfa_fcb_itnim_offline(struct bfad_itnim_s *itnim_drv)
{
struct bfad_port_s *port;
struct bfad_s *bfad;
port = bfa_fcs_itnim_get_drvport(&itnim_drv->fcs_itnim);
bfad = port->bfad;
if ((bfad->pport.flags & BFAD_PORT_DELETE) ||
(port->flags & BFAD_PORT_DELETE)) {
itnim_drv->state = ITNIM_STATE_OFFLINE;
return;
}
itnim_drv->im_port = port->im_port;
itnim_drv->state = ITNIM_STATE_OFFLINE_PENDING;
itnim_drv->queue_work = 1;
bfad_os_itnim_process(itnim_drv);
}
/**
* BFA FCS itnim timeout callback.
* Context: Interrupt. bfad_lock is held
*/
void bfa_fcb_itnim_tov(struct bfad_itnim_s *itnim)
{
itnim->state = ITNIM_STATE_TIMEOUT;
}
/**
* Allocate a Scsi_Host for a port.
*/
int
bfad_im_scsi_host_alloc(struct bfad_s *bfad, struct bfad_im_port_s *im_port,
struct device *dev)
{
int error = 1;
mutex_lock(&bfad_mutex);
if (!idr_pre_get(&bfad_im_port_index, GFP_KERNEL)) {
mutex_unlock(&bfad_mutex);
printk(KERN_WARNING "idr_pre_get failure\n");
goto out;
}
error = idr_get_new(&bfad_im_port_index, im_port,
&im_port->idr_id);
if (error) {
mutex_unlock(&bfad_mutex);
printk(KERN_WARNING "idr_get_new failure\n");
goto out;
}
mutex_unlock(&bfad_mutex);
im_port->shost = bfad_os_scsi_host_alloc(im_port, bfad);
if (!im_port->shost) {
error = 1;
goto out_free_idr;
}
im_port->shost->hostdata[0] = (unsigned long)im_port;
im_port->shost->unique_id = im_port->idr_id;
im_port->shost->this_id = -1;
im_port->shost->max_id = MAX_FCP_TARGET;
im_port->shost->max_lun = MAX_FCP_LUN;
im_port->shost->max_cmd_len = 16;
im_port->shost->can_queue = bfad->cfg_data.ioc_queue_depth;
if (im_port->port->pvb_type == BFAD_PORT_PHYS_BASE)
im_port->shost->transportt = bfad_im_scsi_transport_template;
else
im_port->shost->transportt =
bfad_im_scsi_vport_transport_template;
error = scsi_add_host_with_dma(im_port->shost, dev, &bfad->pcidev->dev);
if (error) {
printk(KERN_WARNING "scsi_add_host failure %d\n", error);
goto out_fc_rel;
}
/* setup host fixed attribute if the lk supports */
bfad_os_fc_host_init(im_port);
return 0;
out_fc_rel:
scsi_host_put(im_port->shost);
im_port->shost = NULL;
out_free_idr:
mutex_lock(&bfad_mutex);
idr_remove(&bfad_im_port_index, im_port->idr_id);
mutex_unlock(&bfad_mutex);
out:
return error;
}
void
bfad_im_scsi_host_free(struct bfad_s *bfad, struct bfad_im_port_s *im_port)
{
bfa_trc(bfad, bfad->inst_no);
bfa_log(bfad->logmod, BFA_LOG_LINUX_SCSI_HOST_FREE,
im_port->shost->host_no);
fc_remove_host(im_port->shost);
scsi_remove_host(im_port->shost);
scsi_host_put(im_port->shost);
mutex_lock(&bfad_mutex);
idr_remove(&bfad_im_port_index, im_port->idr_id);
mutex_unlock(&bfad_mutex);
}
static void
bfad_im_port_delete_handler(struct work_struct *work)
{
struct bfad_im_port_s *im_port =
container_of(work, struct bfad_im_port_s, port_delete_work);
struct bfad_s *bfad = im_port->bfad;
if (im_port->port->pvb_type != BFAD_PORT_PHYS_BASE) {
im_port->flags |= BFAD_PORT_DELETE;
fc_vport_terminate(im_port->fc_vport);
atomic_dec(&bfad->wq_reqcnt);
}
}
bfa_status_t
bfad_im_port_new(struct bfad_s *bfad, struct bfad_port_s *port)
{
int rc = BFA_STATUS_OK;
struct bfad_im_port_s *im_port;
im_port = kzalloc(sizeof(struct bfad_im_port_s), GFP_ATOMIC);
if (im_port == NULL) {
rc = BFA_STATUS_ENOMEM;
goto ext;
}
port->im_port = im_port;
im_port->port = port;
im_port->bfad = bfad;
INIT_WORK(&im_port->port_delete_work, bfad_im_port_delete_handler);
INIT_LIST_HEAD(&im_port->itnim_mapped_list);
INIT_LIST_HEAD(&im_port->binding_list);
ext:
return rc;
}
void
bfad_im_port_delete(struct bfad_s *bfad, struct bfad_port_s *port)
{
struct bfad_im_port_s *im_port = port->im_port;
if (im_port->port->pvb_type != BFAD_PORT_PHYS_BASE) {
atomic_inc(&bfad->wq_reqcnt);
queue_work(bfad->im->drv_workq,
&im_port->port_delete_work);
}
}
void
bfad_im_port_clean(struct bfad_im_port_s *im_port)
{
struct bfad_fcp_binding *bp, *bp_new;
unsigned long flags;
struct bfad_s *bfad = im_port->bfad;
spin_lock_irqsave(&bfad->bfad_lock, flags);
list_for_each_entry_safe(bp, bp_new, &im_port->binding_list,
list_entry) {
list_del(&bp->list_entry);
kfree(bp);
}
/* the itnim_mapped_list must be empty at this time */
bfa_assert(list_empty(&im_port->itnim_mapped_list));
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
}
void
bfad_im_port_online(struct bfad_s *bfad, struct bfad_port_s *port)
{
}
void
bfad_im_port_offline(struct bfad_s *bfad, struct bfad_port_s *port)
{
}
bfa_status_t
bfad_im_probe(struct bfad_s *bfad)
{
struct bfad_im_s *im;
bfa_status_t rc = BFA_STATUS_OK;
im = kzalloc(sizeof(struct bfad_im_s), GFP_KERNEL);
if (im == NULL) {
rc = BFA_STATUS_ENOMEM;
goto ext;
}
bfad->im = im;
im->bfad = bfad;
if (bfad_os_thread_workq(bfad) != BFA_STATUS_OK) {
kfree(im);
rc = BFA_STATUS_FAILED;
}
ext:
return rc;
}
void
bfad_im_probe_undo(struct bfad_s *bfad)
{
if (bfad->im) {
while (atomic_read(&bfad->wq_reqcnt)) {
printk(KERN_INFO "bfa %s: waiting workq processing,"
" wq_reqcnt:%x\n", bfad->pci_name,
atomic_read(&bfad->wq_reqcnt));
schedule_timeout_uninterruptible(HZ);
}
bfad_os_destroy_workq(bfad->im);
kfree(bfad->im);
bfad->im = NULL;
}
}
/**
* Call back function to handle IO redirection state change
*/
void
bfa_cb_ioredirect_state_change(void *hcb_bfad, bfa_boolean_t ioredirect)
{
/* Do nothing */
}
struct Scsi_Host *
bfad_os_scsi_host_alloc(struct bfad_im_port_s *im_port, struct bfad_s *bfad)
{
struct scsi_host_template *sht;
if (im_port->port->pvb_type == BFAD_PORT_PHYS_BASE)
sht = &bfad_im_scsi_host_template;
else
sht = &bfad_im_vport_template;
sht->sg_tablesize = bfad->cfg_data.io_max_sge;
return scsi_host_alloc(sht, sizeof(unsigned long));
}
void
bfad_os_scsi_host_free(struct bfad_s *bfad, struct bfad_im_port_s *im_port)
{
if (!(im_port->flags & BFAD_PORT_DELETE))
flush_workqueue(bfad->im->drv_workq);
bfad_im_scsi_host_free(im_port->bfad, im_port);
bfad_im_port_clean(im_port);
kfree(im_port);
}
void
bfad_os_destroy_workq(struct bfad_im_s *im)
{
if (im && im->drv_workq) {
destroy_workqueue(im->drv_workq);
im->drv_workq = NULL;
}
}
bfa_status_t
bfad_os_thread_workq(struct bfad_s *bfad)
{
struct bfad_im_s *im = bfad->im;
bfa_trc(bfad, 0);
snprintf(im->drv_workq_name, BFAD_KOBJ_NAME_LEN, "bfad_wq_%d",
bfad->inst_no);
im->drv_workq = create_singlethread_workqueue(im->drv_workq_name);
if (!im->drv_workq)
return BFA_STATUS_FAILED;
return BFA_STATUS_OK;
}
/**
* Scsi_Host template entry.
*
* Description:
* OS entry point to adjust the queue_depths on a per-device basis.
* Called once per device during the bus scan.
* Return non-zero if fails.
*/
static int
bfad_im_slave_configure(struct scsi_device *sdev)
{
if (sdev->tagged_supported)
scsi_activate_tcq(sdev, bfa_lun_queue_depth);
else
scsi_deactivate_tcq(sdev, bfa_lun_queue_depth);
return 0;
}
struct scsi_host_template bfad_im_scsi_host_template = {
.module = THIS_MODULE,
.name = BFAD_DRIVER_NAME,
.info = bfad_im_info,
.queuecommand = bfad_im_queuecommand,
.eh_abort_handler = bfad_im_abort_handler,
.eh_device_reset_handler = bfad_im_reset_lun_handler,
.eh_bus_reset_handler = bfad_im_reset_bus_handler,
.slave_alloc = bfad_im_slave_alloc,
.slave_configure = bfad_im_slave_configure,
.slave_destroy = bfad_im_slave_destroy,
.this_id = -1,
.sg_tablesize = BFAD_IO_MAX_SGE,
.cmd_per_lun = 3,
.use_clustering = ENABLE_CLUSTERING,
.shost_attrs = bfad_im_host_attrs,
.max_sectors = 0xFFFF,
};
struct scsi_host_template bfad_im_vport_template = {
.module = THIS_MODULE,
.name = BFAD_DRIVER_NAME,
.info = bfad_im_info,
.queuecommand = bfad_im_queuecommand,
.eh_abort_handler = bfad_im_abort_handler,
.eh_device_reset_handler = bfad_im_reset_lun_handler,
.eh_bus_reset_handler = bfad_im_reset_bus_handler,
.slave_alloc = bfad_im_slave_alloc,
.slave_configure = bfad_im_slave_configure,
.slave_destroy = bfad_im_slave_destroy,
.this_id = -1,
.sg_tablesize = BFAD_IO_MAX_SGE,
.cmd_per_lun = 3,
.use_clustering = ENABLE_CLUSTERING,
.shost_attrs = bfad_im_vport_attrs,
.max_sectors = 0xFFFF,
};
void
bfad_im_probe_post(struct bfad_im_s *im)
{
flush_workqueue(im->drv_workq);
}
bfa_status_t
bfad_im_module_init(void)
{
bfad_im_scsi_transport_template =
fc_attach_transport(&bfad_im_fc_function_template);
if (!bfad_im_scsi_transport_template)
return BFA_STATUS_ENOMEM;
bfad_im_scsi_vport_transport_template =
fc_attach_transport(&bfad_im_vport_fc_function_template);
if (!bfad_im_scsi_vport_transport_template) {
fc_release_transport(bfad_im_scsi_transport_template);
return BFA_STATUS_ENOMEM;
}
return BFA_STATUS_OK;
}
void
bfad_im_module_exit(void)
{
if (bfad_im_scsi_transport_template)
fc_release_transport(bfad_im_scsi_transport_template);
if (bfad_im_scsi_vport_transport_template)
fc_release_transport(bfad_im_scsi_vport_transport_template);
}
void
bfad_os_itnim_process(struct bfad_itnim_s *itnim_drv)
{
struct bfad_im_s *im = itnim_drv->im;
if (itnim_drv->queue_work)
queue_work(im->drv_workq, &itnim_drv->itnim_work);
}
void
bfad_os_ramp_up_qdepth(struct bfad_itnim_s *itnim, struct scsi_device *sdev)
{
struct scsi_device *tmp_sdev;
if (((jiffies - itnim->last_ramp_up_time) >
BFA_QUEUE_FULL_RAMP_UP_TIME * HZ) &&
((jiffies - itnim->last_queue_full_time) >
BFA_QUEUE_FULL_RAMP_UP_TIME * HZ)) {
shost_for_each_device(tmp_sdev, sdev->host) {
if (bfa_lun_queue_depth > tmp_sdev->queue_depth) {
if (tmp_sdev->id != sdev->id)
continue;
if (tmp_sdev->ordered_tags)
scsi_adjust_queue_depth(tmp_sdev,
MSG_ORDERED_TAG,
tmp_sdev->queue_depth + 1);
else
scsi_adjust_queue_depth(tmp_sdev,
MSG_SIMPLE_TAG,
tmp_sdev->queue_depth + 1);
itnim->last_ramp_up_time = jiffies;
}
}
}
}
void
bfad_os_handle_qfull(struct bfad_itnim_s *itnim, struct scsi_device *sdev)
{
struct scsi_device *tmp_sdev;
itnim->last_queue_full_time = jiffies;
shost_for_each_device(tmp_sdev, sdev->host) {
if (tmp_sdev->id != sdev->id)
continue;
scsi_track_queue_full(tmp_sdev, tmp_sdev->queue_depth - 1);
}
}
struct bfad_itnim_s *
bfad_os_get_itnim(struct bfad_im_port_s *im_port, int id)
{
struct bfad_itnim_s *itnim = NULL;
/* Search the mapped list for this target ID */
list_for_each_entry(itnim, &im_port->itnim_mapped_list, list_entry) {
if (id == itnim->scsi_tgt_id)
return itnim;
}
return NULL;
}
/**
* Scsi_Host template entry slave_alloc
*/
static int
bfad_im_slave_alloc(struct scsi_device *sdev)
{
struct fc_rport *rport = starget_to_rport(scsi_target(sdev));
if (!rport || fc_remote_port_chkready(rport))
return -ENXIO;
sdev->hostdata = rport->dd_data;
return 0;
}
void
bfad_os_fc_host_init(struct bfad_im_port_s *im_port)
{
struct Scsi_Host *host = im_port->shost;
struct bfad_s *bfad = im_port->bfad;
struct bfad_port_s *port = im_port->port;
struct bfa_pport_attr_s pattr;
char model[BFA_ADAPTER_MODEL_NAME_LEN];
char fw_ver[BFA_VERSION_LEN];
fc_host_node_name(host) =
bfa_os_htonll((bfa_fcs_port_get_nwwn(port->fcs_port)));
fc_host_port_name(host) =
bfa_os_htonll((bfa_fcs_port_get_pwwn(port->fcs_port)));
fc_host_max_npiv_vports(host) = bfa_lps_get_max_vport(&bfad->bfa);
fc_host_supported_classes(host) = FC_COS_CLASS3;
memset(fc_host_supported_fc4s(host), 0,
sizeof(fc_host_supported_fc4s(host)));
if (bfad_supported_fc4s & (BFA_PORT_ROLE_FCP_IM | BFA_PORT_ROLE_FCP_TM))
/* For FCP type 0x08 */
fc_host_supported_fc4s(host)[2] = 1;
if (bfad_supported_fc4s & BFA_PORT_ROLE_FCP_IPFC)
/* For LLC/SNAP type 0x05 */
fc_host_supported_fc4s(host)[3] = 0x20;
/* For fibre channel services type 0x20 */
fc_host_supported_fc4s(host)[7] = 1;
bfa_get_adapter_model(&bfad->bfa, model);
bfa_get_adapter_fw_ver(&bfad->bfa, fw_ver);
sprintf(fc_host_symbolic_name(host), "Brocade %s FV%s DV%s",
model, fw_ver, BFAD_DRIVER_VERSION);
fc_host_supported_speeds(host) = 0;
fc_host_supported_speeds(host) |=
FC_PORTSPEED_8GBIT | FC_PORTSPEED_4GBIT | FC_PORTSPEED_2GBIT |
FC_PORTSPEED_1GBIT;
bfa_fcport_get_attr(&bfad->bfa, &pattr);
fc_host_maxframe_size(host) = pattr.pport_cfg.maxfrsize;
}
static void
bfad_im_fc_rport_add(struct bfad_im_port_s *im_port, struct bfad_itnim_s *itnim)
{
struct fc_rport_identifiers rport_ids;
struct fc_rport *fc_rport;
struct bfad_itnim_data_s *itnim_data;
rport_ids.node_name =
bfa_os_htonll(bfa_fcs_itnim_get_nwwn(&itnim->fcs_itnim));
rport_ids.port_name =
bfa_os_htonll(bfa_fcs_itnim_get_pwwn(&itnim->fcs_itnim));
rport_ids.port_id =
bfa_os_hton3b(bfa_fcs_itnim_get_fcid(&itnim->fcs_itnim));
rport_ids.roles = FC_RPORT_ROLE_UNKNOWN;
itnim->fc_rport = fc_rport =
fc_remote_port_add(im_port->shost, 0, &rport_ids);
if (!fc_rport)
return;
fc_rport->maxframe_size =
bfa_fcs_itnim_get_maxfrsize(&itnim->fcs_itnim);
fc_rport->supported_classes = bfa_fcs_itnim_get_cos(&itnim->fcs_itnim);
itnim_data = fc_rport->dd_data;
itnim_data->itnim = itnim;
rport_ids.roles |= FC_RPORT_ROLE_FCP_TARGET;
if (rport_ids.roles != FC_RPORT_ROLE_UNKNOWN)
fc_remote_port_rolechg(fc_rport, rport_ids.roles);
if ((fc_rport->scsi_target_id != -1)
&& (fc_rport->scsi_target_id < MAX_FCP_TARGET))
itnim->scsi_tgt_id = fc_rport->scsi_target_id;
return;
}
/**
* Work queue handler using FC transport service
* Context: kernel
*/
static void
bfad_im_itnim_work_handler(struct work_struct *work)
{
struct bfad_itnim_s *itnim = container_of(work, struct bfad_itnim_s,
itnim_work);
struct bfad_im_s *im = itnim->im;
struct bfad_s *bfad = im->bfad;
struct bfad_im_port_s *im_port;
unsigned long flags;
struct fc_rport *fc_rport;
wwn_t wwpn;
u32 fcid;
char wwpn_str[32], fcid_str[16];
spin_lock_irqsave(&bfad->bfad_lock, flags);
im_port = itnim->im_port;
bfa_trc(bfad, itnim->state);
switch (itnim->state) {
case ITNIM_STATE_ONLINE:
if (!itnim->fc_rport) {
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
bfad_im_fc_rport_add(im_port, itnim);
spin_lock_irqsave(&bfad->bfad_lock, flags);
wwpn = bfa_fcs_itnim_get_pwwn(&itnim->fcs_itnim);
fcid = bfa_fcs_itnim_get_fcid(&itnim->fcs_itnim);
wwn2str(wwpn_str, wwpn);
fcid2str(fcid_str, fcid);
list_add_tail(&itnim->list_entry,
&im_port->itnim_mapped_list);
bfa_log(bfad->logmod, BFA_LOG_LINUX_ITNIM_ONLINE,
im_port->shost->host_no,
itnim->scsi_tgt_id,
fcid_str, wwpn_str);
} else {
printk(KERN_WARNING
"%s: itnim %llx is already in online state\n",
__func__,
bfa_fcs_itnim_get_pwwn(&itnim->fcs_itnim));
}
break;
case ITNIM_STATE_OFFLINE_PENDING:
itnim->state = ITNIM_STATE_OFFLINE;
if (itnim->fc_rport) {
fc_rport = itnim->fc_rport;
((struct bfad_itnim_data_s *)
fc_rport->dd_data)->itnim = NULL;
itnim->fc_rport = NULL;
if (!(im_port->port->flags & BFAD_PORT_DELETE)) {
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
fc_rport->dev_loss_tmo =
bfa_fcpim_path_tov_get(&bfad->bfa) + 1;
fc_remote_port_delete(fc_rport);
spin_lock_irqsave(&bfad->bfad_lock, flags);
}
wwpn = bfa_fcs_itnim_get_pwwn(&itnim->fcs_itnim);
fcid = bfa_fcs_itnim_get_fcid(&itnim->fcs_itnim);
wwn2str(wwpn_str, wwpn);
fcid2str(fcid_str, fcid);
list_del(&itnim->list_entry);
bfa_log(bfad->logmod, BFA_LOG_LINUX_ITNIM_OFFLINE,
im_port->shost->host_no,
itnim->scsi_tgt_id,
fcid_str, wwpn_str);
}
break;
case ITNIM_STATE_FREE:
if (itnim->fc_rport) {
fc_rport = itnim->fc_rport;
((struct bfad_itnim_data_s *)
fc_rport->dd_data)->itnim = NULL;
itnim->fc_rport = NULL;
if (!(im_port->port->flags & BFAD_PORT_DELETE)) {
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
fc_rport->dev_loss_tmo =
bfa_fcpim_path_tov_get(&bfad->bfa) + 1;
fc_remote_port_delete(fc_rport);
spin_lock_irqsave(&bfad->bfad_lock, flags);
}
list_del(&itnim->list_entry);
}
kfree(itnim);
break;
default:
bfa_assert(0);
break;
}
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
}
/**
* Scsi_Host template entry, queue a SCSI command to the BFAD.
*/
static int
bfad_im_queuecommand(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
{
struct bfad_im_port_s *im_port =
(struct bfad_im_port_s *) cmnd->device->host->hostdata[0];
struct bfad_s *bfad = im_port->bfad;
struct bfad_itnim_data_s *itnim_data = cmnd->device->hostdata;
struct bfad_itnim_s *itnim;
struct bfa_ioim_s *hal_io;
unsigned long flags;
int rc;
s16 sg_cnt = 0;
struct fc_rport *rport = starget_to_rport(scsi_target(cmnd->device));
rc = fc_remote_port_chkready(rport);
if (rc) {
cmnd->result = rc;
done(cmnd);
return 0;
}
sg_cnt = scsi_dma_map(cmnd);
if (sg_cnt < 0)
return SCSI_MLQUEUE_HOST_BUSY;
cmnd->scsi_done = done;
spin_lock_irqsave(&bfad->bfad_lock, flags);
if (!(bfad->bfad_flags & BFAD_HAL_START_DONE)) {
printk(KERN_WARNING
"bfad%d, queuecommand %p %x failed, BFA stopped\n",
bfad->inst_no, cmnd, cmnd->cmnd[0]);
cmnd->result = ScsiResult(DID_NO_CONNECT, 0);
goto out_fail_cmd;
}
itnim = itnim_data->itnim;
if (!itnim) {
cmnd->result = ScsiResult(DID_IMM_RETRY, 0);
goto out_fail_cmd;
}
hal_io = bfa_ioim_alloc(&bfad->bfa, (struct bfad_ioim_s *) cmnd,
itnim->bfa_itnim, sg_cnt);
if (!hal_io) {
printk(KERN_WARNING "hal_io failure\n");
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
scsi_dma_unmap(cmnd);
return SCSI_MLQUEUE_HOST_BUSY;
}
cmnd->host_scribble = (char *)hal_io;
bfa_trc_fp(bfad, hal_io->iotag);
bfa_ioim_start(hal_io);
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
return 0;
out_fail_cmd:
spin_unlock_irqrestore(&bfad->bfad_lock, flags);
scsi_dma_unmap(cmnd);
if (done)
done(cmnd);
return 0;
}
void
bfad_os_rport_online_wait(struct bfad_s *bfad)
{
int i;
int rport_delay = 10;
for (i = 0; !(bfad->bfad_flags & BFAD_PORT_ONLINE)
&& i < bfa_linkup_delay; i++)
schedule_timeout_uninterruptible(HZ);
if (bfad->bfad_flags & BFAD_PORT_ONLINE) {
rport_delay = rport_delay < bfa_linkup_delay ?
rport_delay : bfa_linkup_delay;
for (i = 0; !(bfad->bfad_flags & BFAD_RPORT_ONLINE)
&& i < rport_delay; i++)
schedule_timeout_uninterruptible(HZ);
if (rport_delay > 0 && (bfad->bfad_flags & BFAD_RPORT_ONLINE))
schedule_timeout_uninterruptible(rport_delay * HZ);
}
}
int
bfad_os_get_linkup_delay(struct bfad_s *bfad)
{
u8 nwwns = 0;
wwn_t wwns[BFA_PREBOOT_BOOTLUN_MAX];
int ldelay;
/*
* Querying for the boot target port wwns
* -- read from boot information in flash.
* If nwwns > 0 => boot over SAN and set bfa_linkup_delay = 30
* else => local boot machine set bfa_linkup_delay = 10
*/
bfa_iocfc_get_bootwwns(&bfad->bfa, &nwwns, wwns);
if (nwwns > 0) {
/* If boot over SAN; linkup_delay = 30sec */
ldelay = 30;
} else {
/* If local boot; linkup_delay = 10sec */
ldelay = 0;
}
return ldelay;
}