openkylin
/
linux
mirror of https://gitee.com/openkylin/linux.git
9
0
Fork 0
Code Issues Projects Releases Wiki Activity
linux/drivers/s390/scsi/zfcp_fsf.c

4813 lines
149 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 06:20:36 +08:00
/*
*
* linux/drivers/s390/scsi/zfcp_fsf.c
*
* FCP adapter driver for IBM eServer zSeries
*
* (C) Copyright IBM Corp. 2002, 2004
*
* Author(s): Martin Peschke <mpeschke@de.ibm.com>
* Raimund Schroeder <raimund.schroeder@de.ibm.com>
* Aron Zeh
* Wolfgang Taphorn
* Stefan Bader <stefan.bader@de.ibm.com>
* Heiko Carstens <heiko.carstens@de.ibm.com>
* Andreas Herrmann <aherrman@de.ibm.com>
* Volker Sameske <sameske@de.ibm.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2, or (at your option)
* any later version.
*
* 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.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
#define ZFCP_FSF_C_REVISION "$Revision: 1.92 $"
#include "zfcp_ext.h"
static int zfcp_fsf_exchange_config_data_handler(struct zfcp_fsf_req *);
static void zfcp_fsf_exchange_port_data_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_open_port_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_close_port_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_close_physical_port_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_open_unit_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_close_unit_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_send_fcp_command_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_send_fcp_command_task_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_send_fcp_command_task_management_handler(
struct zfcp_fsf_req *);
static int zfcp_fsf_abort_fcp_command_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_status_read_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_send_ct_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_send_els_handler(struct zfcp_fsf_req *);
static int zfcp_fsf_control_file_handler(struct zfcp_fsf_req *);
static inline int zfcp_fsf_req_sbal_check(
unsigned long *, struct zfcp_qdio_queue *, int);
static inline int zfcp_use_one_sbal(
struct scatterlist *, int, struct scatterlist *, int);
static struct zfcp_fsf_req *zfcp_fsf_req_alloc(mempool_t *, int);
static int zfcp_fsf_req_send(struct zfcp_fsf_req *, struct timer_list *);
static int zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *);
static int zfcp_fsf_fsfstatus_eval(struct zfcp_fsf_req *);
static int zfcp_fsf_fsfstatus_qual_eval(struct zfcp_fsf_req *);
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
static void zfcp_fsf_link_down_info_eval(struct zfcp_adapter *,
struct fsf_link_down_info *);
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 06:20:36 +08:00
static int zfcp_fsf_req_dispatch(struct zfcp_fsf_req *);
static void zfcp_fsf_req_dismiss(struct zfcp_fsf_req *);
/* association between FSF command and FSF QTCB type */
static u32 fsf_qtcb_type[] = {
[FSF_QTCB_FCP_CMND] = FSF_IO_COMMAND,
[FSF_QTCB_ABORT_FCP_CMND] = FSF_SUPPORT_COMMAND,
[FSF_QTCB_OPEN_PORT_WITH_DID] = FSF_SUPPORT_COMMAND,
[FSF_QTCB_OPEN_LUN] = FSF_SUPPORT_COMMAND,
[FSF_QTCB_CLOSE_LUN] = FSF_SUPPORT_COMMAND,
[FSF_QTCB_CLOSE_PORT] = FSF_SUPPORT_COMMAND,
[FSF_QTCB_CLOSE_PHYSICAL_PORT] = FSF_SUPPORT_COMMAND,
[FSF_QTCB_SEND_ELS] = FSF_SUPPORT_COMMAND,
[FSF_QTCB_SEND_GENERIC] = FSF_SUPPORT_COMMAND,
[FSF_QTCB_EXCHANGE_CONFIG_DATA] = FSF_CONFIG_COMMAND,
[FSF_QTCB_EXCHANGE_PORT_DATA] = FSF_PORT_COMMAND,
[FSF_QTCB_DOWNLOAD_CONTROL_FILE] = FSF_SUPPORT_COMMAND,
[FSF_QTCB_UPLOAD_CONTROL_FILE] = FSF_SUPPORT_COMMAND
};
static const char zfcp_act_subtable_type[5][8] = {
"unknown", "OS", "WWPN", "DID", "LUN"
};
/****************************************************************/
/*************** FSF related Functions *************************/
/****************************************************************/
#define ZFCP_LOG_AREA ZFCP_LOG_AREA_FSF
/*
* function: zfcp_fsf_req_alloc
*
* purpose: Obtains an fsf_req and potentially a qtcb (for all but
* unsolicited requests) via helper functions
* Does some initial fsf request set-up.
*
* returns: pointer to allocated fsf_req if successfull
* NULL otherwise
*
* locks: none
*
*/
static struct zfcp_fsf_req *
zfcp_fsf_req_alloc(mempool_t *pool, int req_flags)
{
size_t size;
void *ptr;
struct zfcp_fsf_req *fsf_req = NULL;
if (req_flags & ZFCP_REQ_NO_QTCB)
size = sizeof(struct zfcp_fsf_req);
else
size = sizeof(struct zfcp_fsf_req_pool_element);
if (likely(pool != NULL))
ptr = mempool_alloc(pool, GFP_ATOMIC);
else
ptr = kmalloc(size, GFP_ATOMIC);
if (unlikely(NULL == ptr))
goto out;
memset(ptr, 0, size);
if (req_flags & ZFCP_REQ_NO_QTCB) {
fsf_req = (struct zfcp_fsf_req *) ptr;
} else {
fsf_req = &((struct zfcp_fsf_req_pool_element *) ptr)->fsf_req;
fsf_req->qtcb =
&((struct zfcp_fsf_req_pool_element *) ptr)->qtcb;
}
fsf_req->pool = pool;
out:
return fsf_req;
}
/*
* function: zfcp_fsf_req_free
*
* purpose: Frees the memory of an fsf_req (and potentially a qtcb) or
* returns it into the pool via helper functions.
*
* returns: sod all
*
* locks: none
*/
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
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 06:20:36 +08:00
zfcp_fsf_req_free(struct zfcp_fsf_req *fsf_req)
{
if (likely(fsf_req->pool != NULL))
mempool_free(fsf_req, fsf_req->pool);
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
else
kfree(fsf_req);
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 06:20:36 +08:00
}
/*
* function:
*
* purpose:
*
* returns:
*
* note: qdio queues shall be down (no ongoing inbound processing)
*/
int
zfcp_fsf_req_dismiss_all(struct zfcp_adapter *adapter)
{
struct zfcp_fsf_req *fsf_req, *tmp;
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
unsigned long flags;
LIST_HEAD(remove_queue);
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 06:20:36 +08:00
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
spin_lock_irqsave(&adapter->fsf_req_list_lock, flags);
list_splice_init(&adapter->fsf_req_list_head, &remove_queue);
atomic_set(&adapter->fsf_reqs_active, 0);
spin_unlock_irqrestore(&adapter->fsf_req_list_lock, 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 06:20:36 +08:00
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
list_for_each_entry_safe(fsf_req, tmp, &remove_queue, list) {
list_del(&fsf_req->list);
zfcp_fsf_req_dismiss(fsf_req);
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 06:20:36 +08:00
}
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
return 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 06:20:36 +08:00
}
/*
* function:
*
* purpose:
*
* returns:
*/
static void
zfcp_fsf_req_dismiss(struct zfcp_fsf_req *fsf_req)
{
fsf_req->status |= ZFCP_STATUS_FSFREQ_DISMISSED;
zfcp_fsf_req_complete(fsf_req);
}
/*
* function: zfcp_fsf_req_complete
*
* purpose: Updates active counts and timers for openfcp-reqs
* May cleanup request after req_eval returns
*
* returns: 0 - success
* !0 - failure
*
* context:
*/
int
zfcp_fsf_req_complete(struct zfcp_fsf_req *fsf_req)
{
int retval = 0;
int cleanup;
if (unlikely(fsf_req->fsf_command == FSF_QTCB_UNSOLICITED_STATUS)) {
ZFCP_LOG_DEBUG("Status read response received\n");
/*
* Note: all cleanup handling is done in the callchain of
* the function call-chain below.
*/
zfcp_fsf_status_read_handler(fsf_req);
goto out;
} else
zfcp_fsf_protstatus_eval(fsf_req);
/*
* fsf_req may be deleted due to waking up functions, so
* cleanup is saved here and used later
*/
if (likely(fsf_req->status & ZFCP_STATUS_FSFREQ_CLEANUP))
cleanup = 1;
else
cleanup = 0;
fsf_req->status |= ZFCP_STATUS_FSFREQ_COMPLETED;
/* cleanup request if requested by initiator */
if (likely(cleanup)) {
ZFCP_LOG_TRACE("removing FSF request %p\n", fsf_req);
/*
* lock must not be held here since it will be
* grabed by the called routine, too
*/
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
zfcp_fsf_req_free(fsf_req);
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 06:20:36 +08:00
} else {
/* notify initiator waiting for the requests completion */
ZFCP_LOG_TRACE("waking initiator of FSF request %p\n",fsf_req);
/*
* FIXME: Race! We must not access fsf_req here as it might have been
* cleaned up already due to the set ZFCP_STATUS_FSFREQ_COMPLETED
* flag. It's an improbable case. But, we have the same paranoia for
* the cleanup flag already.
* Might better be handled using complete()?
* (setting the flag and doing wakeup ought to be atomic
* with regard to checking the flag as long as waitqueue is
* part of the to be released structure)
*/
wake_up(&fsf_req->completion_wq);
}
out:
return retval;
}
/*
* function: zfcp_fsf_protstatus_eval
*
* purpose: evaluates the QTCB of the finished FSF request
* and initiates appropriate actions
* (usually calling FSF command specific handlers)
*
* returns:
*
* context:
*
* locks:
*/
static int
zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req)
{
int retval = 0;
struct zfcp_adapter *adapter = fsf_req->adapter;
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
struct fsf_qtcb *qtcb = fsf_req->qtcb;
union fsf_prot_status_qual *prot_status_qual =
&qtcb->prefix.prot_status_qual;
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 06:20:36 +08:00
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_hba_dbf_event_fsf_response(fsf_req);
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 06:20:36 +08:00
if (fsf_req->status & ZFCP_STATUS_FSFREQ_DISMISSED) {
ZFCP_LOG_DEBUG("fsf_req 0x%lx has been dismissed\n",
(unsigned long) fsf_req);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR |
ZFCP_STATUS_FSFREQ_RETRY; /* only for SCSI cmnds. */
goto skip_protstatus;
}
/* log additional information provided by FSF (if any) */
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
if (unlikely(qtcb->header.log_length)) {
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 06:20:36 +08:00
/* do not trust them ;-) */
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
if (qtcb->header.log_start > sizeof(struct fsf_qtcb)) {
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 06:20:36 +08:00
ZFCP_LOG_NORMAL
("bug: ULP (FSF logging) log data starts "
"beyond end of packet header. Ignored. "
"(start=%i, size=%li)\n",
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
qtcb->header.log_start,
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 06:20:36 +08:00
sizeof(struct fsf_qtcb));
goto forget_log;
}
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
if ((size_t) (qtcb->header.log_start + qtcb->header.log_length)
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 06:20:36 +08:00
> sizeof(struct fsf_qtcb)) {
ZFCP_LOG_NORMAL("bug: ULP (FSF logging) log data ends "
"beyond end of packet header. Ignored. "
"(start=%i, length=%i, size=%li)\n",
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
qtcb->header.log_start,
qtcb->header.log_length,
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 06:20:36 +08:00
sizeof(struct fsf_qtcb));
goto forget_log;
}
ZFCP_LOG_TRACE("ULP log data: \n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_TRACE,
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
(char *) qtcb + qtcb->header.log_start,
qtcb->header.log_length);
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 06:20:36 +08:00
}
forget_log:
/* evaluate FSF Protocol Status */
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
switch (qtcb->prefix.prot_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 06:20:36 +08:00
case FSF_PROT_GOOD:
case FSF_PROT_FSF_STATUS_PRESENTED:
break;
case FSF_PROT_QTCB_VERSION_ERROR:
ZFCP_LOG_NORMAL("error: The adapter %s contains "
"microcode of version 0x%x, the device driver "
"only supports 0x%x. Aborting.\n",
zfcp_get_busid_by_adapter(adapter),
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
prot_status_qual->version_error.fsf_version,
ZFCP_QTCB_VERSION);
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 06:20:36 +08:00
zfcp_erp_adapter_shutdown(adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PROT_SEQ_NUMB_ERROR:
ZFCP_LOG_NORMAL("bug: Sequence number mismatch between "
"driver (0x%x) and adapter %s (0x%x). "
"Restarting all operations on this adapter.\n",
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
qtcb->prefix.req_seq_no,
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 06:20:36 +08:00
zfcp_get_busid_by_adapter(adapter),
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
prot_status_qual->sequence_error.exp_req_seq_no);
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 06:20:36 +08:00
zfcp_erp_adapter_reopen(adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_RETRY;
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PROT_UNSUPP_QTCB_TYPE:
ZFCP_LOG_NORMAL("error: Packet header type used by the "
"device driver is incompatible with "
"that used on adapter %s. "
"Stopping all operations on this adapter.\n",
zfcp_get_busid_by_adapter(adapter));
zfcp_erp_adapter_shutdown(adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PROT_HOST_CONNECTION_INITIALIZING:
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
atomic_set_mask(ZFCP_STATUS_ADAPTER_HOST_CON_INIT,
&(adapter->status));
break;
case FSF_PROT_DUPLICATE_REQUEST_ID:
ZFCP_LOG_NORMAL("bug: The request identifier 0x%Lx "
"to the adapter %s is ambiguous. "
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
"Stopping all operations on this adapter.\n",
*(unsigned long long*)
(&qtcb->bottom.support.req_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 06:20:36 +08:00
zfcp_get_busid_by_adapter(adapter));
zfcp_erp_adapter_shutdown(adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PROT_LINK_DOWN:
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
zfcp_fsf_link_down_info_eval(adapter,
&prot_status_qual->link_down_info);
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 06:20:36 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PROT_REEST_QUEUE:
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
ZFCP_LOG_NORMAL("The local link to adapter with "
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 06:20:36 +08:00
"%s was re-plugged. "
"Re-starting operations on this adapter.\n",
zfcp_get_busid_by_adapter(adapter));
/* All ports should be marked as ready to run again */
zfcp_erp_modify_adapter_status(adapter,
ZFCP_STATUS_COMMON_RUNNING,
ZFCP_SET);
zfcp_erp_adapter_reopen(adapter,
ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED
| ZFCP_STATUS_COMMON_ERP_FAILED);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PROT_ERROR_STATE:
ZFCP_LOG_NORMAL("error: The adapter %s "
"has entered the error state. "
"Restarting all operations on this "
"adapter.\n",
zfcp_get_busid_by_adapter(adapter));
zfcp_erp_adapter_reopen(adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_RETRY;
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_NORMAL("bug: Transfer protocol status information "
"provided by the adapter %s "
"is not compatible with the device driver. "
"Stopping all operations on this adapter. "
"(debug info 0x%x).\n",
zfcp_get_busid_by_adapter(adapter),
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
qtcb->prefix.prot_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 06:20:36 +08:00
zfcp_erp_adapter_shutdown(adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
}
skip_protstatus:
/*
* always call specific handlers to give them a chance to do
* something meaningful even in error cases
*/
zfcp_fsf_fsfstatus_eval(fsf_req);
return retval;
}
/*
* function: zfcp_fsf_fsfstatus_eval
*
* purpose: evaluates FSF status of completed FSF request
* and acts accordingly
*
* returns:
*/
static int
zfcp_fsf_fsfstatus_eval(struct zfcp_fsf_req *fsf_req)
{
int retval = 0;
if (unlikely(fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR)) {
goto skip_fsfstatus;
}
/* evaluate FSF Status */
switch (fsf_req->qtcb->header.fsf_status) {
case FSF_UNKNOWN_COMMAND:
ZFCP_LOG_NORMAL("bug: Command issued by the device driver is "
"not known by the adapter %s "
"Stopping all operations on this adapter. "
"(debug info 0x%x).\n",
zfcp_get_busid_by_adapter(fsf_req->adapter),
fsf_req->qtcb->header.fsf_command);
zfcp_erp_adapter_shutdown(fsf_req->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_FCP_RSP_AVAILABLE:
ZFCP_LOG_DEBUG("FCP Sense data will be presented to the "
"SCSI stack.\n");
break;
case FSF_ADAPTER_STATUS_AVAILABLE:
zfcp_fsf_fsfstatus_qual_eval(fsf_req);
break;
}
skip_fsfstatus:
/*
* always call specific handlers to give them a chance to do
* something meaningful even in error cases
*/
zfcp_fsf_req_dispatch(fsf_req);
return retval;
}
/*
* function: zfcp_fsf_fsfstatus_qual_eval
*
* purpose: evaluates FSF status-qualifier of completed FSF request
* and acts accordingly
*
* returns:
*/
static int
zfcp_fsf_fsfstatus_qual_eval(struct zfcp_fsf_req *fsf_req)
{
int retval = 0;
switch (fsf_req->qtcb->header.fsf_status_qual.word[0]) {
case FSF_SQ_FCP_RSP_AVAILABLE:
break;
case FSF_SQ_RETRY_IF_POSSIBLE:
/* The SCSI-stack may now issue retries or escalate */
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_COMMAND_ABORTED:
/* Carry the aborted state on to upper layer */
fsf_req->status |= ZFCP_STATUS_FSFREQ_ABORTED;
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_NO_RECOM:
ZFCP_LOG_NORMAL("bug: No recommendation could be given for a"
"problem on the adapter %s "
"Stopping all operations on this adapter. ",
zfcp_get_busid_by_adapter(fsf_req->adapter));
zfcp_erp_adapter_shutdown(fsf_req->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_ULP_PROGRAMMING_ERROR:
ZFCP_LOG_NORMAL("error: not enough SBALs for data transfer "
"(adapter %s)\n",
zfcp_get_busid_by_adapter(fsf_req->adapter));
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_INVOKE_LINK_TEST_PROCEDURE:
case FSF_SQ_NO_RETRY_POSSIBLE:
case FSF_SQ_ULP_DEPENDENT_ERP_REQUIRED:
/* dealt with in the respective functions */
break;
default:
ZFCP_LOG_NORMAL("bug: Additional status info could "
"not be interpreted properly.\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_NORMAL,
(char *) &fsf_req->qtcb->header.fsf_status_qual,
sizeof (union fsf_status_qual));
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
}
return retval;
}
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
/**
* zfcp_fsf_link_down_info_eval - evaluate link down information block
*/
static void
zfcp_fsf_link_down_info_eval(struct zfcp_adapter *adapter,
struct fsf_link_down_info *link_down)
{
switch (link_down->error_code) {
case FSF_PSQ_LINK_NO_LIGHT:
ZFCP_LOG_NORMAL("The local link to adapter %s is down "
"(no light detected)\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_WRAP_PLUG:
ZFCP_LOG_NORMAL("The local link to adapter %s is down "
"(wrap plug detected)\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_NO_FCP:
ZFCP_LOG_NORMAL("The local link to adapter %s is down "
"(adjacent node on link does not support FCP)\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_FIRMWARE_UPDATE:
ZFCP_LOG_NORMAL("The local link to adapter %s is down "
"(firmware update in progress)\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_INVALID_WWPN:
ZFCP_LOG_NORMAL("The local link to adapter %s is down "
"(duplicate or invalid WWPN detected)\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_NO_NPIV_SUPPORT:
ZFCP_LOG_NORMAL("The local link to adapter %s is down "
"(no support for NPIV by Fabric)\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_NO_FCP_RESOURCES:
ZFCP_LOG_NORMAL("The local link to adapter %s is down "
"(out of resource in FCP daughtercard)\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_NO_FABRIC_RESOURCES:
ZFCP_LOG_NORMAL("The local link to adapter %s is down "
"(out of resource in Fabric)\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_FABRIC_LOGIN_UNABLE:
ZFCP_LOG_NORMAL("The local link to adapter %s is down "
"(unable to Fabric login)\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_WWPN_ASSIGNMENT_CORRUPTED:
ZFCP_LOG_NORMAL("WWPN assignment file corrupted on adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_MODE_TABLE_CURRUPTED:
ZFCP_LOG_NORMAL("Mode table corrupted on adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_PSQ_LINK_NO_WWPN_ASSIGNMENT:
ZFCP_LOG_NORMAL("No WWPN for assignment table on adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
break;
default:
ZFCP_LOG_NORMAL("The local link to adapter %s is down "
"(warning: unknown reason code %d)\n",
zfcp_get_busid_by_adapter(adapter),
link_down->error_code);
}
if (adapter->connection_features & FSF_FEATURE_NPIV_MODE)
ZFCP_LOG_DEBUG("Debug information to link down: "
"primary_status=0x%02x "
"ioerr_code=0x%02x "
"action_code=0x%02x "
"reason_code=0x%02x "
"explanation_code=0x%02x "
"vendor_specific_code=0x%02x\n",
link_down->primary_status,
link_down->ioerr_code,
link_down->action_code,
link_down->reason_code,
link_down->explanation_code,
link_down->vendor_specific_code);
if (!atomic_test_mask(ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED,
&adapter->status)) {
atomic_set_mask(ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED,
&adapter->status);
switch (link_down->error_code) {
case FSF_PSQ_LINK_NO_LIGHT:
case FSF_PSQ_LINK_WRAP_PLUG:
case FSF_PSQ_LINK_NO_FCP:
case FSF_PSQ_LINK_FIRMWARE_UPDATE:
zfcp_erp_adapter_reopen(adapter, 0);
break;
default:
zfcp_erp_adapter_failed(adapter);
}
}
}
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 06:20:36 +08:00
/*
* function: zfcp_fsf_req_dispatch
*
* purpose: calls the appropriate command specific handler
*
* returns:
*/
static int
zfcp_fsf_req_dispatch(struct zfcp_fsf_req *fsf_req)
{
struct zfcp_erp_action *erp_action = fsf_req->erp_action;
struct zfcp_adapter *adapter = fsf_req->adapter;
int retval = 0;
switch (fsf_req->fsf_command) {
case FSF_QTCB_FCP_CMND:
zfcp_fsf_send_fcp_command_handler(fsf_req);
break;
case FSF_QTCB_ABORT_FCP_CMND:
zfcp_fsf_abort_fcp_command_handler(fsf_req);
break;
case FSF_QTCB_SEND_GENERIC:
zfcp_fsf_send_ct_handler(fsf_req);
break;
case FSF_QTCB_OPEN_PORT_WITH_DID:
zfcp_fsf_open_port_handler(fsf_req);
break;
case FSF_QTCB_OPEN_LUN:
zfcp_fsf_open_unit_handler(fsf_req);
break;
case FSF_QTCB_CLOSE_LUN:
zfcp_fsf_close_unit_handler(fsf_req);
break;
case FSF_QTCB_CLOSE_PORT:
zfcp_fsf_close_port_handler(fsf_req);
break;
case FSF_QTCB_CLOSE_PHYSICAL_PORT:
zfcp_fsf_close_physical_port_handler(fsf_req);
break;
case FSF_QTCB_EXCHANGE_CONFIG_DATA:
zfcp_fsf_exchange_config_data_handler(fsf_req);
break;
case FSF_QTCB_EXCHANGE_PORT_DATA:
zfcp_fsf_exchange_port_data_handler(fsf_req);
break;
case FSF_QTCB_SEND_ELS:
zfcp_fsf_send_els_handler(fsf_req);
break;
case FSF_QTCB_DOWNLOAD_CONTROL_FILE:
zfcp_fsf_control_file_handler(fsf_req);
break;
case FSF_QTCB_UPLOAD_CONTROL_FILE:
zfcp_fsf_control_file_handler(fsf_req);
break;
default:
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
ZFCP_LOG_NORMAL("bug: Command issued by the device driver is "
"not supported by the adapter %s\n",
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_get_busid_by_adapter(adapter));
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 06:20:36 +08:00
if (fsf_req->fsf_command != fsf_req->qtcb->header.fsf_command)
ZFCP_LOG_NORMAL
("bug: Command issued by the device driver differs "
"from the command returned by the adapter %s "
"(debug info 0x%x, 0x%x).\n",
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_get_busid_by_adapter(adapter),
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 06:20:36 +08:00
fsf_req->fsf_command,
fsf_req->qtcb->header.fsf_command);
}
if (!erp_action)
return retval;
zfcp_erp_async_handler(erp_action, 0);
return retval;
}
/*
* function: zfcp_fsf_status_read
*
* purpose: initiates a Status Read command at the specified adapter
*
* returns:
*/
int
zfcp_fsf_status_read(struct zfcp_adapter *adapter, int req_flags)
{
struct zfcp_fsf_req *fsf_req;
struct fsf_status_read_buffer *status_buffer;
unsigned long lock_flags;
volatile struct qdio_buffer_element *sbale;
int retval = 0;
/* setup new FSF request */
retval = zfcp_fsf_req_create(adapter, FSF_QTCB_UNSOLICITED_STATUS,
req_flags | ZFCP_REQ_NO_QTCB,
adapter->pool.fsf_req_status_read,
&lock_flags, &fsf_req);
if (retval < 0) {
ZFCP_LOG_INFO("error: Could not create unsolicited status "
"buffer for adapter %s.\n",
zfcp_get_busid_by_adapter(adapter));
goto failed_req_create;
}
sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0);
sbale[0].flags |= SBAL_FLAGS0_TYPE_STATUS;
sbale[2].flags |= SBAL_FLAGS_LAST_ENTRY;
fsf_req->sbale_curr = 2;
status_buffer =
mempool_alloc(adapter->pool.data_status_read, GFP_ATOMIC);
if (!status_buffer) {
ZFCP_LOG_NORMAL("bug: could not get some buffer\n");
goto failed_buf;
}
memset(status_buffer, 0, sizeof (struct fsf_status_read_buffer));
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
fsf_req->data = (unsigned long) status_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 06:20:36 +08:00
/* insert pointer to respective buffer */
sbale = zfcp_qdio_sbale_curr(fsf_req);
sbale->addr = (void *) status_buffer;
sbale->length = sizeof(struct fsf_status_read_buffer);
/* start QDIO request for this FSF request */
retval = zfcp_fsf_req_send(fsf_req, NULL);
if (retval) {
ZFCP_LOG_DEBUG("error: Could not set-up unsolicited status "
"environment.\n");
goto failed_req_send;
}
ZFCP_LOG_TRACE("Status Read request initiated (adapter%s)\n",
zfcp_get_busid_by_adapter(adapter));
goto out;
failed_req_send:
mempool_free(status_buffer, adapter->pool.data_status_read);
failed_buf:
zfcp_fsf_req_free(fsf_req);
failed_req_create:
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_hba_dbf_event_fsf_unsol("fail", adapter, 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 06:20:36 +08:00
out:
write_unlock_irqrestore(&adapter->request_queue.queue_lock, lock_flags);
return retval;
}
static int
zfcp_fsf_status_read_port_closed(struct zfcp_fsf_req *fsf_req)
{
struct fsf_status_read_buffer *status_buffer;
struct zfcp_adapter *adapter;
struct zfcp_port *port;
unsigned long flags;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
status_buffer = (struct fsf_status_read_buffer *) fsf_req->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 06:20:36 +08:00
adapter = fsf_req->adapter;
read_lock_irqsave(&zfcp_data.config_lock, flags);
list_for_each_entry(port, &adapter->port_list_head, list)
if (port->d_id == (status_buffer->d_id & ZFCP_DID_MASK))
break;
read_unlock_irqrestore(&zfcp_data.config_lock, flags);
if (!port || (port->d_id != (status_buffer->d_id & ZFCP_DID_MASK))) {
ZFCP_LOG_NORMAL("bug: Reopen port indication received for"
"nonexisting port with d_id 0x%08x on "
"adapter %s. Ignored.\n",
status_buffer->d_id & ZFCP_DID_MASK,
zfcp_get_busid_by_adapter(adapter));
goto out;
}
switch (status_buffer->status_subtype) {
case FSF_STATUS_READ_SUB_CLOSE_PHYS_PORT:
debug_text_event(adapter->erp_dbf, 3, "unsol_pc_phys:");
zfcp_erp_port_reopen(port, 0);
break;
case FSF_STATUS_READ_SUB_ERROR_PORT:
debug_text_event(adapter->erp_dbf, 1, "unsol_pc_err:");
zfcp_erp_port_shutdown(port, 0);
break;
default:
debug_text_event(adapter->erp_dbf, 0, "unsol_unk_sub:");
debug_exception(adapter->erp_dbf, 0,
&status_buffer->status_subtype, sizeof (u32));
ZFCP_LOG_NORMAL("bug: Undefined status subtype received "
"for a reopen indication on port with "
"d_id 0x%08x on the adapter %s. "
"Ignored. (debug info 0x%x)\n",
status_buffer->d_id,
zfcp_get_busid_by_adapter(adapter),
status_buffer->status_subtype);
}
out:
return 0;
}
/*
* function: zfcp_fsf_status_read_handler
*
* purpose: is called for finished Open Port command
*
* returns:
*/
static int
zfcp_fsf_status_read_handler(struct zfcp_fsf_req *fsf_req)
{
int retval = 0;
struct zfcp_adapter *adapter = fsf_req->adapter;
struct fsf_status_read_buffer *status_buffer =
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
(struct fsf_status_read_buffer *) fsf_req->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 06:20:36 +08:00
if (fsf_req->status & ZFCP_STATUS_FSFREQ_DISMISSED) {
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_hba_dbf_event_fsf_unsol("dism", adapter, status_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 06:20:36 +08:00
mempool_free(status_buffer, adapter->pool.data_status_read);
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
zfcp_fsf_req_free(fsf_req);
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 06:20:36 +08:00
goto out;
}
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_hba_dbf_event_fsf_unsol("read", adapter, status_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 06:20:36 +08:00
switch (status_buffer->status_type) {
case FSF_STATUS_READ_PORT_CLOSED:
zfcp_fsf_status_read_port_closed(fsf_req);
break;
case FSF_STATUS_READ_INCOMING_ELS:
zfcp_fsf_incoming_els(fsf_req);
break;
case FSF_STATUS_READ_SENSE_DATA_AVAIL:
ZFCP_LOG_INFO("unsolicited sense data received (adapter %s)\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_STATUS_READ_BIT_ERROR_THRESHOLD:
ZFCP_LOG_NORMAL("Bit error threshold data received:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_NORMAL,
(char *) status_buffer,
sizeof (struct fsf_status_read_buffer));
break;
case FSF_STATUS_READ_LINK_DOWN:
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
switch (status_buffer->status_subtype) {
case FSF_STATUS_READ_SUB_NO_PHYSICAL_LINK:
ZFCP_LOG_INFO("Physical link to adapter %s is down\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_STATUS_READ_SUB_FDISC_FAILED:
ZFCP_LOG_INFO("Local link to adapter %s is down "
"due to failed FDISC login\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 06:20:36 +08:00
zfcp_get_busid_by_adapter(adapter));
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
break;
case FSF_STATUS_READ_SUB_FIRMWARE_UPDATE:
ZFCP_LOG_INFO("Local link to adapter %s is down "
"due to firmware update on adapter\n",
zfcp_get_busid_by_adapter(adapter));
break;
default:
ZFCP_LOG_INFO("Local link to adapter %s is down "
"due to unknown reason\n",
zfcp_get_busid_by_adapter(adapter));
};
zfcp_fsf_link_down_info_eval(adapter,
(struct fsf_link_down_info *) &status_buffer->payload);
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 06:20:36 +08:00
break;
case FSF_STATUS_READ_LINK_UP:
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
ZFCP_LOG_NORMAL("Local link to adapter %s was replugged. "
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 06:20:36 +08:00
"Restarting operations on this adapter\n",
zfcp_get_busid_by_adapter(adapter));
/* All ports should be marked as ready to run again */
zfcp_erp_modify_adapter_status(adapter,
ZFCP_STATUS_COMMON_RUNNING,
ZFCP_SET);
zfcp_erp_adapter_reopen(adapter,
ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED
| ZFCP_STATUS_COMMON_ERP_FAILED);
break;
case FSF_STATUS_READ_CFDC_UPDATED:
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
ZFCP_LOG_NORMAL("CFDC has been updated on the adapter %s\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 06:20:36 +08:00
zfcp_get_busid_by_adapter(adapter));
zfcp_erp_adapter_access_changed(adapter);
break;
case FSF_STATUS_READ_CFDC_HARDENED:
switch (status_buffer->status_subtype) {
case FSF_STATUS_READ_SUB_CFDC_HARDENED_ON_SE:
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
ZFCP_LOG_NORMAL("CFDC of adapter %s saved on SE\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 06:20:36 +08:00
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_STATUS_READ_SUB_CFDC_HARDENED_ON_SE2:
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
ZFCP_LOG_NORMAL("CFDC of adapter %s has been copied "
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 06:20:36 +08:00
"to the secondary SE\n",
zfcp_get_busid_by_adapter(adapter));
break;
default:
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
ZFCP_LOG_NORMAL("CFDC of adapter %s has been hardened\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 06:20:36 +08:00
zfcp_get_busid_by_adapter(adapter));
}
break;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
case FSF_STATUS_READ_FEATURE_UPDATE_ALERT:
debug_text_event(adapter->erp_dbf, 2, "unsol_features:");
ZFCP_LOG_INFO("List of supported features on adapter %s has "
"been changed from 0x%08X to 0x%08X\n",
zfcp_get_busid_by_adapter(adapter),
*(u32*) (status_buffer->payload + 4),
*(u32*) (status_buffer->payload));
adapter->adapter_features = *(u32*) status_buffer->payload;
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 06:20:36 +08:00
default:
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
ZFCP_LOG_NORMAL("warning: An unsolicited status packet of 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 06:20:36 +08:00
"type was received (debug info 0x%x)\n",
status_buffer->status_type);
ZFCP_LOG_DEBUG("Dump of status_read_buffer %p:\n",
status_buffer);
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) status_buffer,
sizeof (struct fsf_status_read_buffer));
break;
}
mempool_free(status_buffer, adapter->pool.data_status_read);
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
zfcp_fsf_req_free(fsf_req);
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 06:20:36 +08:00
/*
* recycle buffer and start new request repeat until outbound
* queue is empty or adapter shutdown is requested
*/
/*
* FIXME(qdio):
* we may wait in the req_create for 5s during shutdown, so
* qdio_cleanup will have to wait at least that long before returning
* with failure to allow us a proper cleanup under all circumstances
*/
/*
* FIXME:
* allocation failure possible? (Is this code needed?)
*/
retval = zfcp_fsf_status_read(adapter, 0);
if (retval < 0) {
ZFCP_LOG_INFO("Failed to create unsolicited status read "
"request for the adapter %s.\n",
zfcp_get_busid_by_adapter(adapter));
/* temporary fix to avoid status read buffer shortage */
adapter->status_read_failed++;
if ((ZFCP_STATUS_READS_RECOM - adapter->status_read_failed)
< ZFCP_STATUS_READ_FAILED_THRESHOLD) {
ZFCP_LOG_INFO("restart adapter %s due to status read "
"buffer shortage\n",
zfcp_get_busid_by_adapter(adapter));
zfcp_erp_adapter_reopen(adapter, 0);
}
}
out:
return retval;
}
/*
* function: zfcp_fsf_abort_fcp_command
*
* purpose: tells FSF to abort a running SCSI command
*
* returns: address of initiated FSF request
* NULL - request could not be initiated
*
* FIXME(design): should be watched by a timeout !!!
* FIXME(design) shouldn't this be modified to return an int
* also...don't know how though
*/
struct zfcp_fsf_req *
zfcp_fsf_abort_fcp_command(unsigned long old_req_id,
struct zfcp_adapter *adapter,
struct zfcp_unit *unit, int req_flags)
{
volatile struct qdio_buffer_element *sbale;
unsigned long lock_flags;
struct zfcp_fsf_req *fsf_req = NULL;
int retval = 0;
/* setup new FSF request */
retval = zfcp_fsf_req_create(adapter, FSF_QTCB_ABORT_FCP_CMND,
req_flags, adapter->pool.fsf_req_abort,
&lock_flags, &fsf_req);
if (retval < 0) {
ZFCP_LOG_INFO("error: Failed to create an abort command "
"request for lun 0x%016Lx on port 0x%016Lx "
"on adapter %s.\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_adapter(adapter));
goto out;
}
sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0);
sbale[0].flags |= SBAL_FLAGS0_TYPE_READ;
sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
fsf_req->data = (unsigned long) unit;
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 06:20:36 +08:00
/* set handles of unit and its parent port in QTCB */
fsf_req->qtcb->header.lun_handle = unit->handle;
fsf_req->qtcb->header.port_handle = unit->port->handle;
/* set handle of request which should be aborted */
fsf_req->qtcb->bottom.support.req_handle = (u64) old_req_id;
/* start QDIO request for this FSF request */
zfcp_fsf_start_scsi_er_timer(adapter);
retval = zfcp_fsf_req_send(fsf_req, NULL);
if (retval) {
del_timer(&adapter->scsi_er_timer);
ZFCP_LOG_INFO("error: Failed to send abort command request "
"on adapter %s, port 0x%016Lx, unit 0x%016Lx\n",
zfcp_get_busid_by_adapter(adapter),
unit->port->wwpn, unit->fcp_lun);
zfcp_fsf_req_free(fsf_req);
fsf_req = NULL;
goto out;
}
ZFCP_LOG_DEBUG("Abort FCP Command request initiated "
"(adapter%s, port d_id=0x%08x, "
"unit x%016Lx, old_req_id=0x%lx)\n",
zfcp_get_busid_by_adapter(adapter),
unit->port->d_id,
unit->fcp_lun, old_req_id);
out:
write_unlock_irqrestore(&adapter->request_queue.queue_lock, lock_flags);
return fsf_req;
}
/*
* function: zfcp_fsf_abort_fcp_command_handler
*
* purpose: is called for finished Abort FCP Command request
*
* returns:
*/
static int
zfcp_fsf_abort_fcp_command_handler(struct zfcp_fsf_req *new_fsf_req)
{
int retval = -EINVAL;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
struct zfcp_unit *unit;
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 06:20:36 +08:00
unsigned char status_qual =
new_fsf_req->qtcb->header.fsf_status_qual.word[0];
del_timer(&new_fsf_req->adapter->scsi_er_timer);
if (new_fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) {
/* do not set ZFCP_STATUS_FSFREQ_ABORTSUCCEEDED */
goto skip_fsfstatus;
}
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
unit = (struct zfcp_unit *) new_fsf_req->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 06:20:36 +08:00
/* evaluate FSF status in QTCB */
switch (new_fsf_req->qtcb->header.fsf_status) {
case FSF_PORT_HANDLE_NOT_VALID:
if (status_qual >> 4 != status_qual % 0xf) {
debug_text_event(new_fsf_req->adapter->erp_dbf, 3,
"fsf_s_phand_nv0");
/*
* In this case a command that was sent prior to a port
* reopen was aborted (handles are different). This is
* fine.
*/
} else {
ZFCP_LOG_INFO("Temporary port identifier 0x%x for "
"port 0x%016Lx on adapter %s invalid. "
"This may happen occasionally.\n",
unit->port->handle,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
ZFCP_LOG_INFO("status qualifier:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_INFO,
(char *) &new_fsf_req->qtcb->header.
fsf_status_qual,
sizeof (union fsf_status_qual));
/* Let's hope this sorts out the mess */
debug_text_event(new_fsf_req->adapter->erp_dbf, 1,
"fsf_s_phand_nv1");
zfcp_erp_adapter_reopen(unit->port->adapter, 0);
new_fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
}
break;
case FSF_LUN_HANDLE_NOT_VALID:
if (status_qual >> 4 != status_qual % 0xf) {
/* 2 */
debug_text_event(new_fsf_req->adapter->erp_dbf, 3,
"fsf_s_lhand_nv0");
/*
* In this case a command that was sent prior to a unit
* reopen was aborted (handles are different).
* This is fine.
*/
} else {
ZFCP_LOG_INFO
("Warning: Temporary LUN identifier 0x%x of LUN "
"0x%016Lx on port 0x%016Lx on adapter %s is "
"invalid. This may happen in rare cases. "
"Trying to re-establish link.\n",
unit->handle,
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
ZFCP_LOG_DEBUG("Status qualifier data:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &new_fsf_req->qtcb->header.
fsf_status_qual,
sizeof (union fsf_status_qual));
/* Let's hope this sorts out the mess */
debug_text_event(new_fsf_req->adapter->erp_dbf, 1,
"fsf_s_lhand_nv1");
zfcp_erp_port_reopen(unit->port, 0);
new_fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
}
break;
case FSF_FCP_COMMAND_DOES_NOT_EXIST:
retval = 0;
debug_text_event(new_fsf_req->adapter->erp_dbf, 3,
"fsf_s_no_exist");
new_fsf_req->status |= ZFCP_STATUS_FSFREQ_ABORTNOTNEEDED;
break;
case FSF_PORT_BOXED:
ZFCP_LOG_INFO("Remote port 0x%016Lx on adapter %s needs to "
"be reopened\n", unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
debug_text_event(new_fsf_req->adapter->erp_dbf, 2,
"fsf_s_pboxed");
[SCSI] zfcp: fix handling of port boxed and lun boxed fsf states From: Maxim Shchetynin <maxim@de.ibm.com> Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:23:57 +08:00
zfcp_erp_port_boxed(unit->port);
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 06:20:36 +08:00
new_fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR
| ZFCP_STATUS_FSFREQ_RETRY;
break;
case FSF_LUN_BOXED:
ZFCP_LOG_INFO(
"unit 0x%016Lx on port 0x%016Lx on adapter %s needs "
"to be reopened\n",
unit->fcp_lun, unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
debug_text_event(new_fsf_req->adapter->erp_dbf, 1, "fsf_s_lboxed");
[SCSI] zfcp: fix handling of port boxed and lun boxed fsf states From: Maxim Shchetynin <maxim@de.ibm.com> Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:23:57 +08:00
zfcp_erp_unit_boxed(unit);
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 06:20:36 +08:00
new_fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR
| ZFCP_STATUS_FSFREQ_RETRY;
break;
case FSF_ADAPTER_STATUS_AVAILABLE:
switch (new_fsf_req->qtcb->header.fsf_status_qual.word[0]) {
case FSF_SQ_INVOKE_LINK_TEST_PROCEDURE:
debug_text_event(new_fsf_req->adapter->erp_dbf, 1,
"fsf_sq_ltest");
[SCSI] zfcp: fix: reopen port only if link-test fails From: Maxim Shchetynin <maxim@de.ibm.com> Reopen a remote port only if the link-test fails. This avoids that a port is unnecessarily reopened. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:16:27 +08:00
zfcp_test_link(unit->port);
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 06:20:36 +08:00
new_fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_ULP_DEPENDENT_ERP_REQUIRED:
/* SCSI stack will escalate */
debug_text_event(new_fsf_req->adapter->erp_dbf, 1,
"fsf_sq_ulp");
new_fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_NORMAL
("bug: Wrong status qualifier 0x%x arrived.\n",
new_fsf_req->qtcb->header.fsf_status_qual.word[0]);
debug_text_event(new_fsf_req->adapter->erp_dbf, 0,
"fsf_sq_inval:");
debug_exception(new_fsf_req->adapter->erp_dbf, 0,
&new_fsf_req->qtcb->header.
fsf_status_qual.word[0], sizeof (u32));
break;
}
break;
case FSF_GOOD:
retval = 0;
new_fsf_req->status |= ZFCP_STATUS_FSFREQ_ABORTSUCCEEDED;
break;
default:
ZFCP_LOG_NORMAL("bug: An unknown FSF Status was presented "
"(debug info 0x%x)\n",
new_fsf_req->qtcb->header.fsf_status);
debug_text_event(new_fsf_req->adapter->erp_dbf, 0,
"fsf_s_inval:");
debug_exception(new_fsf_req->adapter->erp_dbf, 0,
&new_fsf_req->qtcb->header.fsf_status,
sizeof (u32));
break;
}
skip_fsfstatus:
return retval;
}
/**
* zfcp_use_one_sbal - checks whether req buffer and resp bother each fit into
* one SBALE
* Two scatter-gather lists are passed, one for the reqeust and one for the
* response.
*/
static inline int
zfcp_use_one_sbal(struct scatterlist *req, int req_count,
struct scatterlist *resp, int resp_count)
{
return ((req_count == 1) &&
(resp_count == 1) &&
(((unsigned long) zfcp_sg_to_address(&req[0]) &
PAGE_MASK) ==
((unsigned long) (zfcp_sg_to_address(&req[0]) +
req[0].length - 1) & PAGE_MASK)) &&
(((unsigned long) zfcp_sg_to_address(&resp[0]) &
PAGE_MASK) ==
((unsigned long) (zfcp_sg_to_address(&resp[0]) +
resp[0].length - 1) & PAGE_MASK)));
}
/**
* zfcp_fsf_send_ct - initiate a Generic Service request (FC-GS)
* @ct: pointer to struct zfcp_send_ct which conatins all needed data for
* the request
* @pool: pointer to memory pool, if non-null this pool is used to allocate
* a struct zfcp_fsf_req
* @erp_action: pointer to erp_action, if non-null the Generic Service request
* is sent within error recovery
*/
int
zfcp_fsf_send_ct(struct zfcp_send_ct *ct, mempool_t *pool,
struct zfcp_erp_action *erp_action)
{
volatile struct qdio_buffer_element *sbale;
struct zfcp_port *port;
struct zfcp_adapter *adapter;
struct zfcp_fsf_req *fsf_req;
unsigned long lock_flags;
int bytes;
int ret = 0;
port = ct->port;
adapter = port->adapter;
ret = zfcp_fsf_req_create(adapter, FSF_QTCB_SEND_GENERIC,
ZFCP_WAIT_FOR_SBAL | ZFCP_REQ_AUTO_CLEANUP,
pool, &lock_flags, &fsf_req);
if (ret < 0) {
ZFCP_LOG_INFO("error: Could not create CT request (FC-GS) for "
"adapter: %s\n",
zfcp_get_busid_by_adapter(adapter));
goto failed_req;
}
if (erp_action != NULL) {
erp_action->fsf_req = fsf_req;
fsf_req->erp_action = erp_action;
}
sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0);
if (zfcp_use_one_sbal(ct->req, ct->req_count,
ct->resp, ct->resp_count)){
/* both request buffer and response buffer
fit into one sbale each */
sbale[0].flags |= SBAL_FLAGS0_TYPE_WRITE_READ;
sbale[2].addr = zfcp_sg_to_address(&ct->req[0]);
sbale[2].length = ct->req[0].length;
sbale[3].addr = zfcp_sg_to_address(&ct->resp[0]);
sbale[3].length = ct->resp[0].length;
sbale[3].flags |= SBAL_FLAGS_LAST_ENTRY;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
} else if (adapter->adapter_features &
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 06:20:36 +08:00
FSF_FEATURE_ELS_CT_CHAINED_SBALS) {
/* try to use chained SBALs */
bytes = zfcp_qdio_sbals_from_sg(fsf_req,
SBAL_FLAGS0_TYPE_WRITE_READ,
ct->req, ct->req_count,
ZFCP_MAX_SBALS_PER_CT_REQ);
if (bytes <= 0) {
ZFCP_LOG_INFO("error: creation of CT request failed "
"on adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
if (bytes == 0)
ret = -ENOMEM;
else
ret = bytes;
goto failed_send;
}
fsf_req->qtcb->bottom.support.req_buf_length = bytes;
fsf_req->sbale_curr = ZFCP_LAST_SBALE_PER_SBAL;
bytes = zfcp_qdio_sbals_from_sg(fsf_req,
SBAL_FLAGS0_TYPE_WRITE_READ,
ct->resp, ct->resp_count,
ZFCP_MAX_SBALS_PER_CT_REQ);
if (bytes <= 0) {
ZFCP_LOG_INFO("error: creation of CT request failed "
"on adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
if (bytes == 0)
ret = -ENOMEM;
else
ret = bytes;
goto failed_send;
}
fsf_req->qtcb->bottom.support.resp_buf_length = bytes;
} else {
/* reject send generic request */
ZFCP_LOG_INFO(
"error: microcode does not support chained SBALs,"
"CT request too big (adapter %s)\n",
zfcp_get_busid_by_adapter(adapter));
ret = -EOPNOTSUPP;
goto failed_send;
}
/* settings in QTCB */
fsf_req->qtcb->header.port_handle = port->handle;
fsf_req->qtcb->bottom.support.service_class = adapter->fc_service_class;
fsf_req->qtcb->bottom.support.timeout = ct->timeout;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
fsf_req->data = (unsigned long) ct;
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 06:20:36 +08:00
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_san_dbf_event_ct_request(fsf_req);
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 06:20:36 +08:00
/* start QDIO request for this FSF request */
ret = zfcp_fsf_req_send(fsf_req, ct->timer);
if (ret) {
ZFCP_LOG_DEBUG("error: initiation of CT request failed "
"(adapter %s, port 0x%016Lx)\n",
zfcp_get_busid_by_adapter(adapter), port->wwpn);
goto failed_send;
}
ZFCP_LOG_DEBUG("CT request initiated (adapter %s, port 0x%016Lx)\n",
zfcp_get_busid_by_adapter(adapter), port->wwpn);
goto out;
failed_send:
zfcp_fsf_req_free(fsf_req);
if (erp_action != NULL) {
erp_action->fsf_req = NULL;
}
failed_req:
out:
write_unlock_irqrestore(&adapter->request_queue.queue_lock,
lock_flags);
return ret;
}
/**
* zfcp_fsf_send_ct_handler - handler for Generic Service requests
* @fsf_req: pointer to struct zfcp_fsf_req
*
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
* Data specific for the Generic Service request is passed using
* fsf_req->data. There we find the pointer to struct zfcp_send_ct.
* Usually a specific handler for the CT request is called which is
* found in this structure.
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 06:20:36 +08:00
*/
static int
zfcp_fsf_send_ct_handler(struct zfcp_fsf_req *fsf_req)
{
struct zfcp_port *port;
struct zfcp_adapter *adapter;
struct zfcp_send_ct *send_ct;
struct fsf_qtcb_header *header;
struct fsf_qtcb_bottom_support *bottom;
int retval = -EINVAL;
u16 subtable, rule, counter;
adapter = fsf_req->adapter;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
send_ct = (struct zfcp_send_ct *) fsf_req->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 06:20:36 +08:00
port = send_ct->port;
header = &fsf_req->qtcb->header;
bottom = &fsf_req->qtcb->bottom.support;
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR)
goto skip_fsfstatus;
/* evaluate FSF status in QTCB */
switch (header->fsf_status) {
case FSF_GOOD:
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_san_dbf_event_ct_response(fsf_req);
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 06:20:36 +08:00
retval = 0;
break;
case FSF_SERVICE_CLASS_NOT_SUPPORTED:
if (adapter->fc_service_class <= 3) {
ZFCP_LOG_INFO("error: adapter %s does not support fc "
"class %d.\n",
zfcp_get_busid_by_port(port),
adapter->fc_service_class);
} else {
ZFCP_LOG_INFO("bug: The fibre channel class at the "
"adapter %s is invalid. "
"(debug info %d)\n",
zfcp_get_busid_by_port(port),
adapter->fc_service_class);
}
/* stop operation for this adapter */
debug_text_exception(adapter->erp_dbf, 0, "fsf_s_class_nsup");
zfcp_erp_adapter_shutdown(adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_ADAPTER_STATUS_AVAILABLE:
switch (header->fsf_status_qual.word[0]){
case FSF_SQ_INVOKE_LINK_TEST_PROCEDURE:
/* reopening link to port */
debug_text_event(adapter->erp_dbf, 1, "fsf_sq_ltest");
zfcp_test_link(port);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_ULP_DEPENDENT_ERP_REQUIRED:
/* ERP strategy will escalate */
debug_text_event(adapter->erp_dbf, 1, "fsf_sq_ulp");
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_INFO("bug: Wrong status qualifier 0x%x "
"arrived.\n",
header->fsf_status_qual.word[0]);
break;
}
break;
case FSF_ACCESS_DENIED:
ZFCP_LOG_NORMAL("access denied, cannot send generic service "
"command (adapter %s, port d_id=0x%08x)\n",
zfcp_get_busid_by_port(port), port->d_id);
for (counter = 0; counter < 2; counter++) {
subtable = header->fsf_status_qual.halfword[counter * 2];
rule = header->fsf_status_qual.halfword[counter * 2 + 1];
switch (subtable) {
case FSF_SQ_CFDC_SUBTABLE_OS:
case FSF_SQ_CFDC_SUBTABLE_PORT_WWPN:
case FSF_SQ_CFDC_SUBTABLE_PORT_DID:
case FSF_SQ_CFDC_SUBTABLE_LUN:
ZFCP_LOG_INFO("Access denied (%s rule %d)\n",
zfcp_act_subtable_type[subtable], rule);
break;
}
}
debug_text_event(adapter->erp_dbf, 1, "fsf_s_access");
zfcp_erp_port_access_denied(port);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_GENERIC_COMMAND_REJECTED:
ZFCP_LOG_INFO("generic service command rejected "
"(adapter %s, port d_id=0x%08x)\n",
zfcp_get_busid_by_port(port), port->d_id);
ZFCP_LOG_INFO("status qualifier:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_INFO,
(char *) &header->fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(adapter->erp_dbf, 1, "fsf_s_gcom_rej");
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PORT_HANDLE_NOT_VALID:
ZFCP_LOG_DEBUG("Temporary port identifier 0x%x for port "
"0x%016Lx on adapter %s invalid. This may "
"happen occasionally.\n", port->handle,
port->wwpn, zfcp_get_busid_by_port(port));
ZFCP_LOG_INFO("status qualifier:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_INFO,
(char *) &header->fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(adapter->erp_dbf, 1, "fsf_s_phandle_nv");
zfcp_erp_adapter_reopen(adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PORT_BOXED:
ZFCP_LOG_INFO("port needs to be reopened "
"(adapter %s, port d_id=0x%08x)\n",
zfcp_get_busid_by_port(port), port->d_id);
debug_text_event(adapter->erp_dbf, 2, "fsf_s_pboxed");
[SCSI] zfcp: fix handling of port boxed and lun boxed fsf states From: Maxim Shchetynin <maxim@de.ibm.com> Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:23:57 +08:00
zfcp_erp_port_boxed(port);
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 06:20:36 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR
| ZFCP_STATUS_FSFREQ_RETRY;
break;
/* following states should never occure, all cases avoided
in zfcp_fsf_send_ct - but who knows ... */
case FSF_PAYLOAD_SIZE_MISMATCH:
ZFCP_LOG_INFO("payload size mismatch (adapter: %s, "
"req_buf_length=%d, resp_buf_length=%d)\n",
zfcp_get_busid_by_adapter(adapter),
bottom->req_buf_length, bottom->resp_buf_length);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_REQUEST_SIZE_TOO_LARGE:
ZFCP_LOG_INFO("request size too large (adapter: %s, "
"req_buf_length=%d)\n",
zfcp_get_busid_by_adapter(adapter),
bottom->req_buf_length);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_RESPONSE_SIZE_TOO_LARGE:
ZFCP_LOG_INFO("response size too large (adapter: %s, "
"resp_buf_length=%d)\n",
zfcp_get_busid_by_adapter(adapter),
bottom->resp_buf_length);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SBAL_MISMATCH:
ZFCP_LOG_INFO("SBAL mismatch (adapter: %s, req_buf_length=%d, "
"resp_buf_length=%d)\n",
zfcp_get_busid_by_adapter(adapter),
bottom->req_buf_length, bottom->resp_buf_length);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_NORMAL("bug: An unknown FSF Status was presented "
"(debug info 0x%x)\n", header->fsf_status);
debug_text_event(adapter->erp_dbf, 0, "fsf_sq_inval:");
debug_exception(adapter->erp_dbf, 0,
&header->fsf_status_qual.word[0], sizeof (u32));
break;
}
skip_fsfstatus:
send_ct->status = retval;
if (send_ct->handler != NULL)
send_ct->handler(send_ct->handler_data);
return retval;
}
/**
* zfcp_fsf_send_els - initiate an ELS command (FC-FS)
* @els: pointer to struct zfcp_send_els which contains all needed data for
* the command.
*/
int
zfcp_fsf_send_els(struct zfcp_send_els *els)
{
volatile struct qdio_buffer_element *sbale;
struct zfcp_fsf_req *fsf_req;
fc_id_t d_id;
struct zfcp_adapter *adapter;
unsigned long lock_flags;
int bytes;
int ret = 0;
d_id = els->d_id;
adapter = els->adapter;
ret = zfcp_fsf_req_create(adapter, FSF_QTCB_SEND_ELS,
ZFCP_REQ_AUTO_CLEANUP,
NULL, &lock_flags, &fsf_req);
if (ret < 0) {
ZFCP_LOG_INFO("error: creation of ELS request failed "
"(adapter %s, port d_id: 0x%08x)\n",
zfcp_get_busid_by_adapter(adapter), d_id);
goto failed_req;
}
sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0);
if (zfcp_use_one_sbal(els->req, els->req_count,
els->resp, els->resp_count)){
/* both request buffer and response buffer
fit into one sbale each */
sbale[0].flags |= SBAL_FLAGS0_TYPE_WRITE_READ;
sbale[2].addr = zfcp_sg_to_address(&els->req[0]);
sbale[2].length = els->req[0].length;
sbale[3].addr = zfcp_sg_to_address(&els->resp[0]);
sbale[3].length = els->resp[0].length;
sbale[3].flags |= SBAL_FLAGS_LAST_ENTRY;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
} else if (adapter->adapter_features &
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 06:20:36 +08:00
FSF_FEATURE_ELS_CT_CHAINED_SBALS) {
/* try to use chained SBALs */
bytes = zfcp_qdio_sbals_from_sg(fsf_req,
SBAL_FLAGS0_TYPE_WRITE_READ,
els->req, els->req_count,
ZFCP_MAX_SBALS_PER_ELS_REQ);
if (bytes <= 0) {
ZFCP_LOG_INFO("error: creation of ELS request failed "
"(adapter %s, port d_id: 0x%08x)\n",
zfcp_get_busid_by_adapter(adapter), d_id);
if (bytes == 0) {
ret = -ENOMEM;
} else {
ret = bytes;
}
goto failed_send;
}
fsf_req->qtcb->bottom.support.req_buf_length = bytes;
fsf_req->sbale_curr = ZFCP_LAST_SBALE_PER_SBAL;
bytes = zfcp_qdio_sbals_from_sg(fsf_req,
SBAL_FLAGS0_TYPE_WRITE_READ,
els->resp, els->resp_count,
ZFCP_MAX_SBALS_PER_ELS_REQ);
if (bytes <= 0) {
ZFCP_LOG_INFO("error: creation of ELS request failed "
"(adapter %s, port d_id: 0x%08x)\n",
zfcp_get_busid_by_adapter(adapter), d_id);
if (bytes == 0) {
ret = -ENOMEM;
} else {
ret = bytes;
}
goto failed_send;
}
fsf_req->qtcb->bottom.support.resp_buf_length = bytes;
} else {
/* reject request */
ZFCP_LOG_INFO("error: microcode does not support chained SBALs"
", ELS request too big (adapter %s, "
"port d_id: 0x%08x)\n",
zfcp_get_busid_by_adapter(adapter), d_id);
ret = -EOPNOTSUPP;
goto failed_send;
}
/* settings in QTCB */
fsf_req->qtcb->bottom.support.d_id = d_id;
fsf_req->qtcb->bottom.support.service_class = adapter->fc_service_class;
fsf_req->qtcb->bottom.support.timeout = ZFCP_ELS_TIMEOUT;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
fsf_req->data = (unsigned long) els;
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 06:20:36 +08:00
sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0);
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_san_dbf_event_els_request(fsf_req);
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 06:20:36 +08:00
/* start QDIO request for this FSF request */
ret = zfcp_fsf_req_send(fsf_req, els->timer);
if (ret) {
ZFCP_LOG_DEBUG("error: initiation of ELS request failed "
"(adapter %s, port d_id: 0x%08x)\n",
zfcp_get_busid_by_adapter(adapter), d_id);
goto failed_send;
}
ZFCP_LOG_DEBUG("ELS request initiated (adapter %s, port d_id: "
"0x%08x)\n", zfcp_get_busid_by_adapter(adapter), d_id);
goto out;
failed_send:
zfcp_fsf_req_free(fsf_req);
failed_req:
out:
write_unlock_irqrestore(&adapter->request_queue.queue_lock,
lock_flags);
return ret;
}
/**
* zfcp_fsf_send_els_handler - handler for ELS commands
* @fsf_req: pointer to struct zfcp_fsf_req
*
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
* Data specific for the ELS command is passed using
* fsf_req->data. There we find the pointer to struct zfcp_send_els.
* Usually a specific handler for the ELS command is called which is
* found in this structure.
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 06:20:36 +08:00
*/
static int zfcp_fsf_send_els_handler(struct zfcp_fsf_req *fsf_req)
{
struct zfcp_adapter *adapter;
struct zfcp_port *port;
[SCSI] zfcp: fix: problem in send_els_handler when D_ID assignment changes From: Maxim Shchetynin <maxim@de.ibm.com> Fixes a bug in zfcp_send_els_handler. If D_ID assignments for ports are changing between initiation of one ELS request and its completion the wrong port might be accessed in the completion for that ELS request. Thus a pointer to the port has to be passed for ELS requests to identify the port structure if required. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:18:56 +08:00
fc_id_t d_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 06:20:36 +08:00
struct fsf_qtcb_header *header;
struct fsf_qtcb_bottom_support *bottom;
struct zfcp_send_els *send_els;
int retval = -EINVAL;
u16 subtable, rule, counter;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
send_els = (struct zfcp_send_els *) fsf_req->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 06:20:36 +08:00
adapter = send_els->adapter;
[SCSI] zfcp: fix: problem in send_els_handler when D_ID assignment changes From: Maxim Shchetynin <maxim@de.ibm.com> Fixes a bug in zfcp_send_els_handler. If D_ID assignments for ports are changing between initiation of one ELS request and its completion the wrong port might be accessed in the completion for that ELS request. Thus a pointer to the port has to be passed for ELS requests to identify the port structure if required. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:18:56 +08:00
port = send_els->port;
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 06:20:36 +08:00
d_id = send_els->d_id;
header = &fsf_req->qtcb->header;
bottom = &fsf_req->qtcb->bottom.support;
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR)
goto skip_fsfstatus;
switch (header->fsf_status) {
case FSF_GOOD:
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_san_dbf_event_els_response(fsf_req);
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 06:20:36 +08:00
retval = 0;
break;
case FSF_SERVICE_CLASS_NOT_SUPPORTED:
if (adapter->fc_service_class <= 3) {
ZFCP_LOG_INFO("error: adapter %s does "
"not support fibrechannel class %d.\n",
zfcp_get_busid_by_adapter(adapter),
adapter->fc_service_class);
} else {
ZFCP_LOG_INFO("bug: The fibrechannel class at "
"adapter %s is invalid. "
"(debug info %d)\n",
zfcp_get_busid_by_adapter(adapter),
adapter->fc_service_class);
}
/* stop operation for this adapter */
debug_text_exception(adapter->erp_dbf, 0, "fsf_s_class_nsup");
zfcp_erp_adapter_shutdown(adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_ADAPTER_STATUS_AVAILABLE:
switch (header->fsf_status_qual.word[0]){
case FSF_SQ_INVOKE_LINK_TEST_PROCEDURE:
debug_text_event(adapter->erp_dbf, 1, "fsf_sq_ltest");
[SCSI] zfcp: fix: problem in send_els_handler when D_ID assignment changes From: Maxim Shchetynin <maxim@de.ibm.com> Fixes a bug in zfcp_send_els_handler. If D_ID assignments for ports are changing between initiation of one ELS request and its completion the wrong port might be accessed in the completion for that ELS request. Thus a pointer to the port has to be passed for ELS requests to identify the port structure if required. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:18:56 +08:00
if (port && (send_els->ls_code != ZFCP_LS_ADISC))
zfcp_test_link(port);
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 06:20:36 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_ULP_DEPENDENT_ERP_REQUIRED:
debug_text_event(adapter->erp_dbf, 1, "fsf_sq_ulp");
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval =
zfcp_handle_els_rjt(header->fsf_status_qual.word[1],
(struct zfcp_ls_rjt_par *)
&header->fsf_status_qual.word[2]);
break;
case FSF_SQ_RETRY_IF_POSSIBLE:
debug_text_event(adapter->erp_dbf, 1, "fsf_sq_retry");
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_INFO("bug: Wrong status qualifier 0x%x\n",
header->fsf_status_qual.word[0]);
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_INFO,
(char*)header->fsf_status_qual.word, 16);
}
break;
case FSF_ELS_COMMAND_REJECTED:
ZFCP_LOG_INFO("ELS has been rejected because command filter "
"prohibited sending "
"(adapter: %s, port d_id: 0x%08x)\n",
zfcp_get_busid_by_adapter(adapter), d_id);
break;
case FSF_PAYLOAD_SIZE_MISMATCH:
ZFCP_LOG_INFO(
"ELS request size and ELS response size must be either "
"both 0, or both greater than 0 "
"(adapter: %s, req_buf_length=%d resp_buf_length=%d)\n",
zfcp_get_busid_by_adapter(adapter),
bottom->req_buf_length,
bottom->resp_buf_length);
break;
case FSF_REQUEST_SIZE_TOO_LARGE:
ZFCP_LOG_INFO(
"Length of the ELS request buffer, "
"specified in QTCB bottom, "
"exceeds the size of the buffers "
"that have been allocated for ELS request data "
"(adapter: %s, req_buf_length=%d)\n",
zfcp_get_busid_by_adapter(adapter),
bottom->req_buf_length);
break;
case FSF_RESPONSE_SIZE_TOO_LARGE:
ZFCP_LOG_INFO(
"Length of the ELS response buffer, "
"specified in QTCB bottom, "
"exceeds the size of the buffers "
"that have been allocated for ELS response data "
"(adapter: %s, resp_buf_length=%d)\n",
zfcp_get_busid_by_adapter(adapter),
bottom->resp_buf_length);
break;
case FSF_SBAL_MISMATCH:
/* should never occure, avoided in zfcp_fsf_send_els */
ZFCP_LOG_INFO("SBAL mismatch (adapter: %s, req_buf_length=%d, "
"resp_buf_length=%d)\n",
zfcp_get_busid_by_adapter(adapter),
bottom->req_buf_length, bottom->resp_buf_length);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_ACCESS_DENIED:
ZFCP_LOG_NORMAL("access denied, cannot send ELS command "
"(adapter %s, port d_id=0x%08x)\n",
zfcp_get_busid_by_adapter(adapter), d_id);
for (counter = 0; counter < 2; counter++) {
subtable = header->fsf_status_qual.halfword[counter * 2];
rule = header->fsf_status_qual.halfword[counter * 2 + 1];
switch (subtable) {
case FSF_SQ_CFDC_SUBTABLE_OS:
case FSF_SQ_CFDC_SUBTABLE_PORT_WWPN:
case FSF_SQ_CFDC_SUBTABLE_PORT_DID:
case FSF_SQ_CFDC_SUBTABLE_LUN:
ZFCP_LOG_INFO("Access denied (%s rule %d)\n",
zfcp_act_subtable_type[subtable], rule);
break;
}
}
debug_text_event(adapter->erp_dbf, 1, "fsf_s_access");
if (port != NULL)
zfcp_erp_port_access_denied(port);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_NORMAL(
"bug: An unknown FSF Status was presented "
"(adapter: %s, fsf_status=0x%08x)\n",
zfcp_get_busid_by_adapter(adapter),
header->fsf_status);
debug_text_event(adapter->erp_dbf, 0, "fsf_sq_inval");
debug_exception(adapter->erp_dbf, 0,
&header->fsf_status_qual.word[0], sizeof(u32));
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
}
skip_fsfstatus:
send_els->status = retval;
if (send_els->handler != 0)
send_els->handler(send_els->handler_data);
return retval;
}
/*
* function:
*
* purpose:
*
* returns: address of initiated FSF request
* NULL - request could not be initiated
*/
int
zfcp_fsf_exchange_config_data(struct zfcp_erp_action *erp_action)
{
volatile struct qdio_buffer_element *sbale;
unsigned long lock_flags;
int retval = 0;
/* setup new FSF request */
retval = zfcp_fsf_req_create(erp_action->adapter,
FSF_QTCB_EXCHANGE_CONFIG_DATA,
ZFCP_REQ_AUTO_CLEANUP,
erp_action->adapter->pool.fsf_req_erp,
&lock_flags, &(erp_action->fsf_req));
if (retval < 0) {
ZFCP_LOG_INFO("error: Could not create exchange configuration "
"data request for adapter %s.\n",
zfcp_get_busid_by_adapter(erp_action->adapter));
goto out;
}
sbale = zfcp_qdio_sbale_req(erp_action->fsf_req,
erp_action->fsf_req->sbal_curr, 0);
sbale[0].flags |= SBAL_FLAGS0_TYPE_READ;
sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY;
erp_action->fsf_req->erp_action = erp_action;
erp_action->fsf_req->qtcb->bottom.config.feature_selection =
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
FSF_FEATURE_CFDC |
FSF_FEATURE_LUN_SHARING |
FSF_FEATURE_UPDATE_ALERT;
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 06:20:36 +08:00
/* start QDIO request for this FSF request */
retval = zfcp_fsf_req_send(erp_action->fsf_req, &erp_action->timer);
if (retval) {
ZFCP_LOG_INFO
("error: Could not send exchange configuration data "
"command on the adapter %s\n",
zfcp_get_busid_by_adapter(erp_action->adapter));
zfcp_fsf_req_free(erp_action->fsf_req);
erp_action->fsf_req = NULL;
goto out;
}
ZFCP_LOG_DEBUG("exchange configuration data request initiated "
"(adapter %s)\n",
zfcp_get_busid_by_adapter(erp_action->adapter));
out:
write_unlock_irqrestore(&erp_action->adapter->request_queue.queue_lock,
lock_flags);
return retval;
}
/**
* zfcp_fsf_exchange_config_evaluate
* @fsf_req: fsf_req which belongs to xchg config data request
* @xchg_ok: specifies if xchg config data was incomplete or complete (0/1)
*
* returns: -EIO on error, 0 otherwise
*/
static int
zfcp_fsf_exchange_config_evaluate(struct zfcp_fsf_req *fsf_req, int xchg_ok)
{
struct fsf_qtcb_bottom_config *bottom;
struct zfcp_adapter *adapter = fsf_req->adapter;
bottom = &fsf_req->qtcb->bottom.config;
ZFCP_LOG_DEBUG("low/high QTCB version 0x%x/0x%x of FSF\n",
bottom->low_qtcb_version, bottom->high_qtcb_version);
adapter->fsf_lic_version = bottom->lic_version;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
adapter->adapter_features = bottom->adapter_features;
adapter->connection_features = bottom->connection_features;
zfcp: add point-2-point support From: Andreas Herrmann <aherrman@de.ibm.com> This patch mainly introduces support for point-2-point topology. From: Heiko Carstens <heiko.carstens@de.ibm.com> From: Maxim Shchetynin <maxim@de.ibm.com> From: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-11 12:04:28 +08:00
adapter->peer_wwpn = 0;
adapter->peer_wwnn = 0;
adapter->peer_d_id = 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 06:20:36 +08:00
if (xchg_ok) {
adapter->wwnn = bottom->nport_serv_param.wwnn;
adapter->wwpn = bottom->nport_serv_param.wwpn;
adapter->s_id = bottom->s_id & ZFCP_DID_MASK;
adapter->fc_topology = bottom->fc_topology;
adapter->fc_link_speed = bottom->fc_link_speed;
adapter->hydra_version = bottom->adapter_type;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
if (adapter->physical_wwpn == 0)
adapter->physical_wwpn = adapter->wwpn;
if (adapter->physical_s_id == 0)
adapter->physical_s_id = adapter->s_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 06:20:36 +08:00
} else {
adapter->wwnn = 0;
adapter->wwpn = 0;
adapter->s_id = 0;
adapter->fc_topology = 0;
adapter->fc_link_speed = 0;
adapter->hydra_version = 0;
}
zfcp: add point-2-point support From: Andreas Herrmann <aherrman@de.ibm.com> This patch mainly introduces support for point-2-point topology. From: Heiko Carstens <heiko.carstens@de.ibm.com> From: Maxim Shchetynin <maxim@de.ibm.com> From: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-11 12:04:28 +08:00
if (adapter->fc_topology == FSF_TOPO_P2P) {
adapter->peer_d_id = bottom->peer_d_id & ZFCP_DID_MASK;
adapter->peer_wwpn = bottom->plogi_payload.wwpn;
adapter->peer_wwnn = bottom->plogi_payload.wwnn;
}
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
if (adapter->adapter_features & FSF_FEATURE_HBAAPI_MANAGEMENT) {
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 06:20:36 +08:00
adapter->hardware_version = bottom->hardware_version;
memcpy(adapter->serial_number, bottom->serial_number, 17);
EBCASC(adapter->serial_number, sizeof(adapter->serial_number));
}
zfcp: add point-2-point support From: Andreas Herrmann <aherrman@de.ibm.com> This patch mainly introduces support for point-2-point topology. From: Heiko Carstens <heiko.carstens@de.ibm.com> From: Maxim Shchetynin <maxim@de.ibm.com> From: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-11 12:04:28 +08:00
ZFCP_LOG_NORMAL("The adapter %s reported the following characteristics:\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 06:20:36 +08:00
"WWNN 0x%016Lx, "
"WWPN 0x%016Lx, "
"S_ID 0x%08x,\n"
"adapter version 0x%x, "
"LIC version 0x%x, "
"FC link speed %d Gb/s\n",
zfcp_get_busid_by_adapter(adapter),
adapter->wwnn,
adapter->wwpn,
(unsigned int) adapter->s_id,
adapter->hydra_version,
adapter->fsf_lic_version,
adapter->fc_link_speed);
if (ZFCP_QTCB_VERSION < bottom->low_qtcb_version) {
ZFCP_LOG_NORMAL("error: the adapter %s "
"only supports newer control block "
"versions in comparison to this device "
"driver (try updated device driver)\n",
zfcp_get_busid_by_adapter(adapter));
debug_text_event(adapter->erp_dbf, 0, "low_qtcb_ver");
zfcp_erp_adapter_shutdown(adapter, 0);
return -EIO;
}
if (ZFCP_QTCB_VERSION > bottom->high_qtcb_version) {
ZFCP_LOG_NORMAL("error: the adapter %s "
"only supports older control block "
"versions than this device driver uses"
"(consider a microcode upgrade)\n",
zfcp_get_busid_by_adapter(adapter));
debug_text_event(adapter->erp_dbf, 0, "high_qtcb_ver");
zfcp_erp_adapter_shutdown(adapter, 0);
return -EIO;
}
[PATCH] zfcp: add rports to enable scsi_add_device to work again This patch fixes a severe problem with 2.6.13-rc7. Due to recent SCSI changes it is not possible to add any LUNs to the zfcp device driver anymore. With registration of remote ports this is fixed. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Acked-by: James Bottomley <jejb@steeleye.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-08-28 02:07:54 +08:00
zfcp_set_fc_host_attrs(adapter);
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 06:20:36 +08:00
return 0;
}
/*
* function: zfcp_fsf_exchange_config_data_handler
*
* purpose: is called for finished Exchange Configuration Data command
*
* returns:
*/
static int
zfcp_fsf_exchange_config_data_handler(struct zfcp_fsf_req *fsf_req)
{
struct fsf_qtcb_bottom_config *bottom;
struct zfcp_adapter *adapter = fsf_req->adapter;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
struct fsf_qtcb *qtcb = fsf_req->qtcb;
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 06:20:36 +08:00
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR)
return -EIO;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
switch (qtcb->header.fsf_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 06:20:36 +08:00
case FSF_GOOD:
if (zfcp_fsf_exchange_config_evaluate(fsf_req, 1))
return -EIO;
switch (adapter->fc_topology) {
case FSF_TOPO_P2P:
zfcp: add point-2-point support From: Andreas Herrmann <aherrman@de.ibm.com> This patch mainly introduces support for point-2-point topology. From: Heiko Carstens <heiko.carstens@de.ibm.com> From: Maxim Shchetynin <maxim@de.ibm.com> From: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-11 12:04:28 +08:00
ZFCP_LOG_NORMAL("Point-to-Point fibrechannel "
"configuration detected at adapter %s\n"
"Peer WWNN 0x%016llx, "
"peer WWPN 0x%016llx, "
"peer d_id 0x%06x\n",
zfcp_get_busid_by_adapter(adapter),
adapter->peer_wwnn,
adapter->peer_wwpn,
adapter->peer_d_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 06:20:36 +08:00
debug_text_event(fsf_req->adapter->erp_dbf, 0,
"top-p-to-p");
zfcp: add point-2-point support From: Andreas Herrmann <aherrman@de.ibm.com> This patch mainly introduces support for point-2-point topology. From: Heiko Carstens <heiko.carstens@de.ibm.com> From: Maxim Shchetynin <maxim@de.ibm.com> From: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-11 12:04:28 +08:00
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 06:20:36 +08:00
case FSF_TOPO_AL:
ZFCP_LOG_NORMAL("error: Arbitrated loop fibrechannel "
"topology detected at adapter %s "
"unsupported, shutting down adapter\n",
zfcp_get_busid_by_adapter(adapter));
debug_text_event(fsf_req->adapter->erp_dbf, 0,
"top-al");
zfcp_erp_adapter_shutdown(adapter, 0);
return -EIO;
case FSF_TOPO_FABRIC:
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
ZFCP_LOG_NORMAL("Switched fabric fibrechannel "
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 06:20:36 +08:00
"network detected at adapter %s.\n",
zfcp_get_busid_by_adapter(adapter));
break;
default:
ZFCP_LOG_NORMAL("bug: The fibrechannel topology "
"reported by the exchange "
"configuration command for "
"the adapter %s is not "
"of a type known to the zfcp "
"driver, shutting down adapter\n",
zfcp_get_busid_by_adapter(adapter));
zfcp: add point-2-point support From: Andreas Herrmann <aherrman@de.ibm.com> This patch mainly introduces support for point-2-point topology. From: Heiko Carstens <heiko.carstens@de.ibm.com> From: Maxim Shchetynin <maxim@de.ibm.com> From: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-11 12:04:28 +08:00
adapter->fc_topology = FSF_TOPO_ERROR;
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 06:20:36 +08:00
debug_text_exception(fsf_req->adapter->erp_dbf, 0,
"unknown-topo");
zfcp_erp_adapter_shutdown(adapter, 0);
return -EIO;
}
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
bottom = &qtcb->bottom.config;
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 06:20:36 +08:00
if (bottom->max_qtcb_size < sizeof(struct fsf_qtcb)) {
ZFCP_LOG_NORMAL("bug: Maximum QTCB size (%d bytes) "
"allowed by the adapter %s "
"is lower than the minimum "
"required by the driver (%ld bytes).\n",
bottom->max_qtcb_size,
zfcp_get_busid_by_adapter(adapter),
sizeof(struct fsf_qtcb));
debug_text_event(fsf_req->adapter->erp_dbf, 0,
"qtcb-size");
debug_event(fsf_req->adapter->erp_dbf, 0,
&bottom->max_qtcb_size, sizeof (u32));
zfcp_erp_adapter_shutdown(adapter, 0);
return -EIO;
}
atomic_set_mask(ZFCP_STATUS_ADAPTER_XCONFIG_OK,
&adapter->status);
break;
case FSF_EXCHANGE_CONFIG_DATA_INCOMPLETE:
debug_text_event(adapter->erp_dbf, 0, "xchg-inco");
if (zfcp_fsf_exchange_config_evaluate(fsf_req, 0))
return -EIO;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
atomic_set_mask(ZFCP_STATUS_ADAPTER_XCONFIG_OK, &adapter->status);
zfcp_fsf_link_down_info_eval(adapter,
&qtcb->header.fsf_status_qual.link_down_info);
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 06:20:36 +08:00
break;
default:
debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf-stat-ng");
debug_event(fsf_req->adapter->erp_dbf, 0,
&fsf_req->qtcb->header.fsf_status, sizeof (u32));
zfcp_erp_adapter_shutdown(adapter, 0);
return -EIO;
}
return 0;
}
/**
* zfcp_fsf_exchange_port_data - request information about local port
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
* @erp_action: ERP action for the adapter for which port data is requested
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 06:20:36 +08:00
* @adapter: for which port data is requested
* @data: response to exchange port data request
*/
int
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
zfcp_fsf_exchange_port_data(struct zfcp_erp_action *erp_action,
struct zfcp_adapter *adapter,
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 06:20:36 +08:00
struct fsf_qtcb_bottom_port *data)
{
volatile struct qdio_buffer_element *sbale;
int retval = 0;
unsigned long lock_flags;
struct zfcp_fsf_req *fsf_req;
struct timer_list *timer;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
if (!(adapter->adapter_features & FSF_FEATURE_HBAAPI_MANAGEMENT)) {
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 06:20:36 +08:00
ZFCP_LOG_INFO("error: exchange port data "
"command not supported by adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
return -EOPNOTSUPP;
}
timer = kmalloc(sizeof(struct timer_list), GFP_KERNEL);
if (!timer)
return -ENOMEM;
/* setup new FSF request */
retval = zfcp_fsf_req_create(adapter, FSF_QTCB_EXCHANGE_PORT_DATA,
0, 0, &lock_flags, &fsf_req);
if (retval < 0) {
ZFCP_LOG_INFO("error: Out of resources. Could not create an "
"exchange port data request for"
"the adapter %s.\n",
zfcp_get_busid_by_adapter(adapter));
write_unlock_irqrestore(&adapter->request_queue.queue_lock,
lock_flags);
goto out;
}
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
if (erp_action) {
erp_action->fsf_req = fsf_req;
fsf_req->erp_action = erp_action;
}
if (data)
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
fsf_req->data = (unsigned long) 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 06:20:36 +08:00
sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0);
sbale[0].flags |= SBAL_FLAGS0_TYPE_READ;
sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY;
init_timer(timer);
timer->function = zfcp_fsf_request_timeout_handler;
timer->data = (unsigned long) adapter;
timer->expires = ZFCP_FSF_REQUEST_TIMEOUT;
retval = zfcp_fsf_req_send(fsf_req, timer);
if (retval) {
ZFCP_LOG_INFO("error: Could not send an exchange port data "
"command on the adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
zfcp_fsf_req_free(fsf_req);
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
if (erp_action)
erp_action->fsf_req = 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 06:20:36 +08:00
write_unlock_irqrestore(&adapter->request_queue.queue_lock,
lock_flags);
goto out;
}
ZFCP_LOG_DEBUG("Exchange Port Data request initiated (adapter %s)\n",
zfcp_get_busid_by_adapter(adapter));
write_unlock_irqrestore(&adapter->request_queue.queue_lock,
lock_flags);
wait_event(fsf_req->completion_wq,
fsf_req->status & ZFCP_STATUS_FSFREQ_COMPLETED);
del_timer_sync(timer);
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
zfcp_fsf_req_free(fsf_req);
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 06:20:36 +08:00
out:
kfree(timer);
return retval;
}
/**
* zfcp_fsf_exchange_port_data_handler - handler for exchange_port_data request
* @fsf_req: pointer to struct zfcp_fsf_req
*/
static void
zfcp_fsf_exchange_port_data_handler(struct zfcp_fsf_req *fsf_req)
{
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
struct zfcp_adapter *adapter = fsf_req->adapter;
struct fsf_qtcb *qtcb = fsf_req->qtcb;
struct fsf_qtcb_bottom_port *bottom, *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 06:20:36 +08:00
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR)
return;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
switch (qtcb->header.fsf_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 06:20:36 +08:00
case FSF_GOOD:
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
atomic_set_mask(ZFCP_STATUS_ADAPTER_XPORT_OK, &adapter->status);
bottom = &qtcb->bottom.port;
data = (struct fsf_qtcb_bottom_port*) fsf_req->data;
if (data)
memcpy(data, bottom, sizeof(struct fsf_qtcb_bottom_port));
if (adapter->connection_features & FSF_FEATURE_NPIV_MODE) {
adapter->physical_wwpn = bottom->wwpn;
adapter->physical_s_id = bottom->fc_port_id;
} else {
adapter->physical_wwpn = adapter->wwpn;
adapter->physical_s_id = adapter->s_id;
}
break;
case FSF_EXCHANGE_CONFIG_DATA_INCOMPLETE:
atomic_set_mask(ZFCP_STATUS_ADAPTER_XPORT_OK, &adapter->status);
zfcp_fsf_link_down_info_eval(adapter,
&qtcb->header.fsf_status_qual.link_down_info);
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 06:20:36 +08:00
break;
default:
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
debug_text_event(adapter->erp_dbf, 0, "xchg-port-ng");
debug_event(adapter->erp_dbf, 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 06:20:36 +08:00
&fsf_req->qtcb->header.fsf_status, sizeof(u32));
}
}
/*
* function: zfcp_fsf_open_port
*
* purpose:
*
* returns: address of initiated FSF request
* NULL - request could not be initiated
*/
int
zfcp_fsf_open_port(struct zfcp_erp_action *erp_action)
{
volatile struct qdio_buffer_element *sbale;
unsigned long lock_flags;
int retval = 0;
/* setup new FSF request */
retval = zfcp_fsf_req_create(erp_action->adapter,
FSF_QTCB_OPEN_PORT_WITH_DID,
ZFCP_WAIT_FOR_SBAL | ZFCP_REQ_AUTO_CLEANUP,
erp_action->adapter->pool.fsf_req_erp,
&lock_flags, &(erp_action->fsf_req));
if (retval < 0) {
ZFCP_LOG_INFO("error: Could not create open port request "
"for port 0x%016Lx on adapter %s.\n",
erp_action->port->wwpn,
zfcp_get_busid_by_adapter(erp_action->adapter));
goto out;
}
sbale = zfcp_qdio_sbale_req(erp_action->fsf_req,
erp_action->fsf_req->sbal_curr, 0);
sbale[0].flags |= SBAL_FLAGS0_TYPE_READ;
sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY;
erp_action->fsf_req->qtcb->bottom.support.d_id = erp_action->port->d_id;
atomic_set_mask(ZFCP_STATUS_COMMON_OPENING, &erp_action->port->status);
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
erp_action->fsf_req->data = (unsigned long) erp_action->port;
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 06:20:36 +08:00
erp_action->fsf_req->erp_action = erp_action;
/* start QDIO request for this FSF request */
retval = zfcp_fsf_req_send(erp_action->fsf_req, &erp_action->timer);
if (retval) {
ZFCP_LOG_INFO("error: Could not send open port request for "
"port 0x%016Lx on adapter %s.\n",
erp_action->port->wwpn,
zfcp_get_busid_by_adapter(erp_action->adapter));
zfcp_fsf_req_free(erp_action->fsf_req);
erp_action->fsf_req = NULL;
goto out;
}
ZFCP_LOG_DEBUG("open port request initiated "
"(adapter %s, port 0x%016Lx)\n",
zfcp_get_busid_by_adapter(erp_action->adapter),
erp_action->port->wwpn);
out:
write_unlock_irqrestore(&erp_action->adapter->request_queue.queue_lock,
lock_flags);
return retval;
}
/*
* function: zfcp_fsf_open_port_handler
*
* purpose: is called for finished Open Port command
*
* returns:
*/
static int
zfcp_fsf_open_port_handler(struct zfcp_fsf_req *fsf_req)
{
int retval = -EINVAL;
struct zfcp_port *port;
struct fsf_plogi *plogi;
struct fsf_qtcb_header *header;
u16 subtable, rule, counter;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
port = (struct zfcp_port *) fsf_req->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 06:20:36 +08:00
header = &fsf_req->qtcb->header;
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) {
/* don't change port status in our bookkeeping */
goto skip_fsfstatus;
}
/* evaluate FSF status in QTCB */
switch (header->fsf_status) {
case FSF_PORT_ALREADY_OPEN:
ZFCP_LOG_NORMAL("bug: remote port 0x%016Lx on adapter %s "
"is already open.\n",
port->wwpn, zfcp_get_busid_by_port(port));
debug_text_exception(fsf_req->adapter->erp_dbf, 0,
"fsf_s_popen");
/*
* This is a bug, however operation should continue normally
* if it is simply ignored
*/
break;
case FSF_ACCESS_DENIED:
ZFCP_LOG_NORMAL("Access denied, cannot open port 0x%016Lx "
"on adapter %s\n",
port->wwpn, zfcp_get_busid_by_port(port));
for (counter = 0; counter < 2; counter++) {
subtable = header->fsf_status_qual.halfword[counter * 2];
rule = header->fsf_status_qual.halfword[counter * 2 + 1];
switch (subtable) {
case FSF_SQ_CFDC_SUBTABLE_OS:
case FSF_SQ_CFDC_SUBTABLE_PORT_WWPN:
case FSF_SQ_CFDC_SUBTABLE_PORT_DID:
case FSF_SQ_CFDC_SUBTABLE_LUN:
ZFCP_LOG_INFO("Access denied (%s rule %d)\n",
zfcp_act_subtable_type[subtable], rule);
break;
}
}
debug_text_event(fsf_req->adapter->erp_dbf, 1, "fsf_s_access");
zfcp_erp_port_access_denied(port);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_MAXIMUM_NUMBER_OF_PORTS_EXCEEDED:
ZFCP_LOG_INFO("error: The FSF adapter is out of resources. "
"The remote port 0x%016Lx on adapter %s "
"could not be opened. Disabling it.\n",
port->wwpn, zfcp_get_busid_by_port(port));
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_s_max_ports");
zfcp_erp_port_failed(port);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_ADAPTER_STATUS_AVAILABLE:
switch (header->fsf_status_qual.word[0]) {
case FSF_SQ_INVOKE_LINK_TEST_PROCEDURE:
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_sq_ltest");
/* ERP strategy will escalate */
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_ULP_DEPENDENT_ERP_REQUIRED:
/* ERP strategy will escalate */
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_sq_ulp");
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_NO_RETRY_POSSIBLE:
ZFCP_LOG_NORMAL("The remote port 0x%016Lx on "
"adapter %s could not be opened. "
"Disabling it.\n",
port->wwpn,
zfcp_get_busid_by_port(port));
debug_text_exception(fsf_req->adapter->erp_dbf, 0,
"fsf_sq_no_retry");
zfcp_erp_port_failed(port);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_NORMAL
("bug: Wrong status qualifier 0x%x arrived.\n",
header->fsf_status_qual.word[0]);
debug_text_event(fsf_req->adapter->erp_dbf, 0,
"fsf_sq_inval:");
debug_exception(
fsf_req->adapter->erp_dbf, 0,
&header->fsf_status_qual.word[0],
sizeof (u32));
break;
}
break;
case FSF_GOOD:
/* save port handle assigned by FSF */
port->handle = header->port_handle;
ZFCP_LOG_INFO("The remote port 0x%016Lx via adapter %s "
"was opened, it's port handle is 0x%x\n",
port->wwpn, zfcp_get_busid_by_port(port),
port->handle);
/* mark port as open */
atomic_set_mask(ZFCP_STATUS_COMMON_OPEN |
ZFCP_STATUS_PORT_PHYS_OPEN, &port->status);
[SCSI] zfcp: fix handling of port boxed and lun boxed fsf states From: Maxim Shchetynin <maxim@de.ibm.com> Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:23:57 +08:00
atomic_clear_mask(ZFCP_STATUS_COMMON_ACCESS_DENIED |
ZFCP_STATUS_COMMON_ACCESS_BOXED,
&port->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 06:20:36 +08:00
retval = 0;
/* check whether D_ID has changed during open */
/*
* FIXME: This check is not airtight, as the FCP channel does
* not monitor closures of target port connections caused on
* the remote side. Thus, they might miss out on invalidating
* locally cached WWPNs (and other N_Port parameters) of gone
* target ports. So, our heroic attempt to make things safe
* could be undermined by 'open port' response data tagged with
* obsolete WWPNs. Another reason to monitor potential
* connection closures ourself at least (by interpreting
* incoming ELS' and unsolicited status). It just crosses my
* mind that one should be able to cross-check by means of
* another GID_PN straight after a port has been opened.
* Alternately, an ADISC/PDISC ELS should suffice, as well.
*/
plogi = (struct fsf_plogi *) fsf_req->qtcb->bottom.support.els;
if (!atomic_test_mask(ZFCP_STATUS_PORT_NO_WWPN, &port->status))
{
if (fsf_req->qtcb->bottom.support.els1_length <
((((unsigned long) &plogi->serv_param.wwpn) -
((unsigned long) plogi)) + sizeof (u64))) {
ZFCP_LOG_INFO(
"warning: insufficient length of "
"PLOGI payload (%i)\n",
fsf_req->qtcb->bottom.support.els1_length);
debug_text_event(fsf_req->adapter->erp_dbf, 0,
"fsf_s_short_plogi:");
/* skip sanity check and assume wwpn is ok */
} else {
if (plogi->serv_param.wwpn != port->wwpn) {
ZFCP_LOG_INFO("warning: d_id of port "
"0x%016Lx changed during "
"open\n", port->wwpn);
debug_text_event(
fsf_req->adapter->erp_dbf, 0,
"fsf_s_did_change:");
atomic_clear_mask(
ZFCP_STATUS_PORT_DID_DID,
&port->status);
} else
port->wwnn = plogi->serv_param.wwnn;
}
}
break;
case FSF_UNKNOWN_OP_SUBTYPE:
/* should never occure, subtype not set in zfcp_fsf_open_port */
ZFCP_LOG_INFO("unknown operation subtype (adapter: %s, "
"op_subtype=0x%x)\n",
zfcp_get_busid_by_port(port),
fsf_req->qtcb->bottom.support.operation_subtype);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_NORMAL("bug: An unknown FSF Status was presented "
"(debug info 0x%x)\n",
header->fsf_status);
debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf_s_inval:");
debug_exception(fsf_req->adapter->erp_dbf, 0,
&header->fsf_status, sizeof (u32));
break;
}
skip_fsfstatus:
atomic_clear_mask(ZFCP_STATUS_COMMON_OPENING, &port->status);
return retval;
}
/*
* function: zfcp_fsf_close_port
*
* purpose: submit FSF command "close port"
*
* returns: address of initiated FSF request
* NULL - request could not be initiated
*/
int
zfcp_fsf_close_port(struct zfcp_erp_action *erp_action)
{
volatile struct qdio_buffer_element *sbale;
unsigned long lock_flags;
int retval = 0;
/* setup new FSF request */
retval = zfcp_fsf_req_create(erp_action->adapter,
FSF_QTCB_CLOSE_PORT,
ZFCP_WAIT_FOR_SBAL | ZFCP_REQ_AUTO_CLEANUP,
erp_action->adapter->pool.fsf_req_erp,
&lock_flags, &(erp_action->fsf_req));
if (retval < 0) {
ZFCP_LOG_INFO("error: Could not create a close port request "
"for port 0x%016Lx on adapter %s.\n",
erp_action->port->wwpn,
zfcp_get_busid_by_adapter(erp_action->adapter));
goto out;
}
sbale = zfcp_qdio_sbale_req(erp_action->fsf_req,
erp_action->fsf_req->sbal_curr, 0);
sbale[0].flags |= SBAL_FLAGS0_TYPE_READ;
sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY;
atomic_set_mask(ZFCP_STATUS_COMMON_CLOSING, &erp_action->port->status);
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
erp_action->fsf_req->data = (unsigned long) erp_action->port;
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 06:20:36 +08:00
erp_action->fsf_req->erp_action = erp_action;
erp_action->fsf_req->qtcb->header.port_handle =
erp_action->port->handle;
/* start QDIO request for this FSF request */
retval = zfcp_fsf_req_send(erp_action->fsf_req, &erp_action->timer);
if (retval) {
ZFCP_LOG_INFO("error: Could not send a close port request for "
"port 0x%016Lx on adapter %s.\n",
erp_action->port->wwpn,
zfcp_get_busid_by_adapter(erp_action->adapter));
zfcp_fsf_req_free(erp_action->fsf_req);
erp_action->fsf_req = NULL;
goto out;
}
ZFCP_LOG_TRACE("close port request initiated "
"(adapter %s, port 0x%016Lx)\n",
zfcp_get_busid_by_adapter(erp_action->adapter),
erp_action->port->wwpn);
out:
write_unlock_irqrestore(&erp_action->adapter->request_queue.queue_lock,
lock_flags);
return retval;
}
/*
* function: zfcp_fsf_close_port_handler
*
* purpose: is called for finished Close Port FSF command
*
* returns:
*/
static int
zfcp_fsf_close_port_handler(struct zfcp_fsf_req *fsf_req)
{
int retval = -EINVAL;
struct zfcp_port *port;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
port = (struct zfcp_port *) fsf_req->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 06:20:36 +08:00
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) {
/* don't change port status in our bookkeeping */
goto skip_fsfstatus;
}
/* evaluate FSF status in QTCB */
switch (fsf_req->qtcb->header.fsf_status) {
case FSF_PORT_HANDLE_NOT_VALID:
ZFCP_LOG_INFO("Temporary port identifier 0x%x for port "
"0x%016Lx on adapter %s invalid. This may happen "
"occasionally.\n", port->handle,
port->wwpn, zfcp_get_busid_by_port(port));
ZFCP_LOG_DEBUG("status qualifier:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &fsf_req->qtcb->header.fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_s_phand_nv");
zfcp_erp_adapter_reopen(port->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_ADAPTER_STATUS_AVAILABLE:
/* Note: FSF has actually closed the port in this case.
* The status code is just daft. Fingers crossed for a change
*/
retval = 0;
break;
case FSF_GOOD:
ZFCP_LOG_TRACE("remote port 0x016%Lx on adapter %s closed, "
"port handle 0x%x\n", port->wwpn,
zfcp_get_busid_by_port(port), port->handle);
zfcp_erp_modify_port_status(port,
ZFCP_STATUS_COMMON_OPEN,
ZFCP_CLEAR);
retval = 0;
break;
default:
ZFCP_LOG_NORMAL("bug: An unknown FSF Status was presented "
"(debug info 0x%x)\n",
fsf_req->qtcb->header.fsf_status);
debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf_s_inval:");
debug_exception(fsf_req->adapter->erp_dbf, 0,
&fsf_req->qtcb->header.fsf_status,
sizeof (u32));
break;
}
skip_fsfstatus:
atomic_clear_mask(ZFCP_STATUS_COMMON_CLOSING, &port->status);
return retval;
}
/*
* function: zfcp_fsf_close_physical_port
*
* purpose: submit FSF command "close physical port"
*
* returns: address of initiated FSF request
* NULL - request could not be initiated
*/
int
zfcp_fsf_close_physical_port(struct zfcp_erp_action *erp_action)
{
int retval = 0;
unsigned long lock_flags;
volatile struct qdio_buffer_element *sbale;
/* setup new FSF request */
retval = zfcp_fsf_req_create(erp_action->adapter,
FSF_QTCB_CLOSE_PHYSICAL_PORT,
ZFCP_WAIT_FOR_SBAL | ZFCP_REQ_AUTO_CLEANUP,
erp_action->adapter->pool.fsf_req_erp,
&lock_flags, &erp_action->fsf_req);
if (retval < 0) {
ZFCP_LOG_INFO("error: Could not create close physical port "
"request (adapter %s, port 0x%016Lx)\n",
zfcp_get_busid_by_adapter(erp_action->adapter),
erp_action->port->wwpn);
goto out;
}
sbale = zfcp_qdio_sbale_req(erp_action->fsf_req,
erp_action->fsf_req->sbal_curr, 0);
sbale[0].flags |= SBAL_FLAGS0_TYPE_READ;
sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY;
/* mark port as being closed */
atomic_set_mask(ZFCP_STATUS_PORT_PHYS_CLOSING,
&erp_action->port->status);
/* save a pointer to this port */
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
erp_action->fsf_req->data = (unsigned long) erp_action->port;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
/* port to be closed */
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 06:20:36 +08:00
erp_action->fsf_req->qtcb->header.port_handle =
erp_action->port->handle;
erp_action->fsf_req->erp_action = erp_action;
/* start QDIO request for this FSF request */
retval = zfcp_fsf_req_send(erp_action->fsf_req, &erp_action->timer);
if (retval) {
ZFCP_LOG_INFO("error: Could not send close physical port "
"request (adapter %s, port 0x%016Lx)\n",
zfcp_get_busid_by_adapter(erp_action->adapter),
erp_action->port->wwpn);
zfcp_fsf_req_free(erp_action->fsf_req);
erp_action->fsf_req = NULL;
goto out;
}
ZFCP_LOG_TRACE("close physical port request initiated "
"(adapter %s, port 0x%016Lx)\n",
zfcp_get_busid_by_adapter(erp_action->adapter),
erp_action->port->wwpn);
out:
write_unlock_irqrestore(&erp_action->adapter->request_queue.queue_lock,
lock_flags);
return retval;
}
/*
* function: zfcp_fsf_close_physical_port_handler
*
* purpose: is called for finished Close Physical Port FSF command
*
* returns:
*/
static int
zfcp_fsf_close_physical_port_handler(struct zfcp_fsf_req *fsf_req)
{
int retval = -EINVAL;
struct zfcp_port *port;
struct zfcp_unit *unit;
struct fsf_qtcb_header *header;
u16 subtable, rule, counter;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
port = (struct zfcp_port *) fsf_req->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 06:20:36 +08:00
header = &fsf_req->qtcb->header;
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) {
/* don't change port status in our bookkeeping */
goto skip_fsfstatus;
}
/* evaluate FSF status in QTCB */
switch (header->fsf_status) {
case FSF_PORT_HANDLE_NOT_VALID:
ZFCP_LOG_INFO("Temporary port identifier 0x%x invalid"
"(adapter %s, port 0x%016Lx). "
"This may happen occasionally.\n",
port->handle,
zfcp_get_busid_by_port(port),
port->wwpn);
ZFCP_LOG_DEBUG("status qualifier:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &header->fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_s_phand_nv");
zfcp_erp_adapter_reopen(port->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_ACCESS_DENIED:
ZFCP_LOG_NORMAL("Access denied, cannot close "
"physical port 0x%016Lx on adapter %s\n",
port->wwpn, zfcp_get_busid_by_port(port));
for (counter = 0; counter < 2; counter++) {
subtable = header->fsf_status_qual.halfword[counter * 2];
rule = header->fsf_status_qual.halfword[counter * 2 + 1];
switch (subtable) {
case FSF_SQ_CFDC_SUBTABLE_OS:
case FSF_SQ_CFDC_SUBTABLE_PORT_WWPN:
case FSF_SQ_CFDC_SUBTABLE_PORT_DID:
case FSF_SQ_CFDC_SUBTABLE_LUN:
ZFCP_LOG_INFO("Access denied (%s rule %d)\n",
zfcp_act_subtable_type[subtable], rule);
break;
}
}
debug_text_event(fsf_req->adapter->erp_dbf, 1, "fsf_s_access");
zfcp_erp_port_access_denied(port);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PORT_BOXED:
ZFCP_LOG_DEBUG("The remote port 0x%016Lx on adapter "
"%s needs to be reopened but it was attempted "
"to close it physically.\n",
port->wwpn,
zfcp_get_busid_by_port(port));
debug_text_event(fsf_req->adapter->erp_dbf, 1, "fsf_s_pboxed");
[SCSI] zfcp: fix handling of port boxed and lun boxed fsf states From: Maxim Shchetynin <maxim@de.ibm.com> Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:23:57 +08:00
zfcp_erp_port_boxed(port);
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 06:20:36 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR |
ZFCP_STATUS_FSFREQ_RETRY;
break;
case FSF_ADAPTER_STATUS_AVAILABLE:
switch (header->fsf_status_qual.word[0]) {
case FSF_SQ_INVOKE_LINK_TEST_PROCEDURE:
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_sq_ltest");
/* This will now be escalated by ERP */
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_ULP_DEPENDENT_ERP_REQUIRED:
/* ERP strategy will escalate */
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_sq_ulp");
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_NORMAL
("bug: Wrong status qualifier 0x%x arrived.\n",
header->fsf_status_qual.word[0]);
debug_text_event(fsf_req->adapter->erp_dbf, 0,
"fsf_sq_inval:");
debug_exception(
fsf_req->adapter->erp_dbf, 0,
&header->fsf_status_qual.word[0], sizeof (u32));
break;
}
break;
case FSF_GOOD:
ZFCP_LOG_DEBUG("Remote port 0x%016Lx via adapter %s "
"physically closed, port handle 0x%x\n",
port->wwpn,
zfcp_get_busid_by_port(port), port->handle);
/* can't use generic zfcp_erp_modify_port_status because
* ZFCP_STATUS_COMMON_OPEN must not be reset for the port
*/
atomic_clear_mask(ZFCP_STATUS_PORT_PHYS_OPEN, &port->status);
list_for_each_entry(unit, &port->unit_list_head, list)
atomic_clear_mask(ZFCP_STATUS_COMMON_OPEN, &unit->status);
retval = 0;
break;
default:
ZFCP_LOG_NORMAL("bug: An unknown FSF Status was presented "
"(debug info 0x%x)\n",
header->fsf_status);
debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf_s_inval:");
debug_exception(fsf_req->adapter->erp_dbf, 0,
&header->fsf_status, sizeof (u32));
break;
}
skip_fsfstatus:
atomic_clear_mask(ZFCP_STATUS_PORT_PHYS_CLOSING, &port->status);
return retval;
}
/*
* function: zfcp_fsf_open_unit
*
* purpose:
*
* returns:
*
* assumptions: This routine does not check whether the associated
* remote port has already been opened. This should be
* done by calling routines. Otherwise some status
* may be presented by FSF
*/
int
zfcp_fsf_open_unit(struct zfcp_erp_action *erp_action)
{
volatile struct qdio_buffer_element *sbale;
unsigned long lock_flags;
int retval = 0;
/* setup new FSF request */
retval = zfcp_fsf_req_create(erp_action->adapter,
FSF_QTCB_OPEN_LUN,
ZFCP_WAIT_FOR_SBAL | ZFCP_REQ_AUTO_CLEANUP,
erp_action->adapter->pool.fsf_req_erp,
&lock_flags, &(erp_action->fsf_req));
if (retval < 0) {
ZFCP_LOG_INFO("error: Could not create open unit request for "
"unit 0x%016Lx on port 0x%016Lx on adapter %s.\n",
erp_action->unit->fcp_lun,
erp_action->unit->port->wwpn,
zfcp_get_busid_by_adapter(erp_action->adapter));
goto out;
}
sbale = zfcp_qdio_sbale_req(erp_action->fsf_req,
erp_action->fsf_req->sbal_curr, 0);
sbale[0].flags |= SBAL_FLAGS0_TYPE_READ;
sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY;
erp_action->fsf_req->qtcb->header.port_handle =
erp_action->port->handle;
erp_action->fsf_req->qtcb->bottom.support.fcp_lun =
erp_action->unit->fcp_lun;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
if (!(erp_action->adapter->connection_features & FSF_FEATURE_NPIV_MODE))
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 06:20:36 +08:00
erp_action->fsf_req->qtcb->bottom.support.option =
FSF_OPEN_LUN_SUPPRESS_BOXING;
atomic_set_mask(ZFCP_STATUS_COMMON_OPENING, &erp_action->unit->status);
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
erp_action->fsf_req->data = (unsigned long) erp_action->unit;
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 06:20:36 +08:00
erp_action->fsf_req->erp_action = erp_action;
/* start QDIO request for this FSF request */
retval = zfcp_fsf_req_send(erp_action->fsf_req, &erp_action->timer);
if (retval) {
ZFCP_LOG_INFO("error: Could not send an open unit request "
"on the adapter %s, port 0x%016Lx for "
"unit 0x%016Lx\n",
zfcp_get_busid_by_adapter(erp_action->adapter),
erp_action->port->wwpn,
erp_action->unit->fcp_lun);
zfcp_fsf_req_free(erp_action->fsf_req);
erp_action->fsf_req = NULL;
goto out;
}
ZFCP_LOG_TRACE("Open LUN request initiated (adapter %s, "
"port 0x%016Lx, unit 0x%016Lx)\n",
zfcp_get_busid_by_adapter(erp_action->adapter),
erp_action->port->wwpn, erp_action->unit->fcp_lun);
out:
write_unlock_irqrestore(&erp_action->adapter->request_queue.queue_lock,
lock_flags);
return retval;
}
/*
* function: zfcp_fsf_open_unit_handler
*
* purpose: is called for finished Open LUN command
*
* returns:
*/
static int
zfcp_fsf_open_unit_handler(struct zfcp_fsf_req *fsf_req)
{
int retval = -EINVAL;
struct zfcp_adapter *adapter;
struct zfcp_unit *unit;
struct fsf_qtcb_header *header;
struct fsf_qtcb_bottom_support *bottom;
struct fsf_queue_designator *queue_designator;
u16 subtable, rule, counter;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
int exclusive, readwrite;
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 06:20:36 +08:00
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
unit = (struct zfcp_unit *) fsf_req->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 06:20:36 +08:00
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) {
/* don't change unit status in our bookkeeping */
goto skip_fsfstatus;
}
adapter = fsf_req->adapter;
header = &fsf_req->qtcb->header;
bottom = &fsf_req->qtcb->bottom.support;
queue_designator = &header->fsf_status_qual.fsf_queue_designator;
atomic_clear_mask(ZFCP_STATUS_COMMON_ACCESS_DENIED |
ZFCP_STATUS_UNIT_SHARED |
ZFCP_STATUS_UNIT_READONLY,
&unit->status);
/* evaluate FSF status in QTCB */
switch (header->fsf_status) {
case FSF_PORT_HANDLE_NOT_VALID:
ZFCP_LOG_INFO("Temporary port identifier 0x%x "
"for port 0x%016Lx on adapter %s invalid "
"This may happen occasionally\n",
unit->port->handle,
unit->port->wwpn, zfcp_get_busid_by_unit(unit));
ZFCP_LOG_DEBUG("status qualifier:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &header->fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(adapter->erp_dbf, 1, "fsf_s_ph_nv");
zfcp_erp_adapter_reopen(unit->port->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_LUN_ALREADY_OPEN:
ZFCP_LOG_NORMAL("bug: Attempted to open unit 0x%016Lx on "
"remote port 0x%016Lx on adapter %s twice.\n",
unit->fcp_lun,
unit->port->wwpn, zfcp_get_busid_by_unit(unit));
debug_text_exception(adapter->erp_dbf, 0,
"fsf_s_uopen");
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_ACCESS_DENIED:
ZFCP_LOG_NORMAL("Access denied, cannot open unit 0x%016Lx on "
"remote port 0x%016Lx on adapter %s\n",
unit->fcp_lun, unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
for (counter = 0; counter < 2; counter++) {
subtable = header->fsf_status_qual.halfword[counter * 2];
rule = header->fsf_status_qual.halfword[counter * 2 + 1];
switch (subtable) {
case FSF_SQ_CFDC_SUBTABLE_OS:
case FSF_SQ_CFDC_SUBTABLE_PORT_WWPN:
case FSF_SQ_CFDC_SUBTABLE_PORT_DID:
case FSF_SQ_CFDC_SUBTABLE_LUN:
ZFCP_LOG_INFO("Access denied (%s rule %d)\n",
zfcp_act_subtable_type[subtable], rule);
break;
}
}
debug_text_event(adapter->erp_dbf, 1, "fsf_s_access");
zfcp_erp_unit_access_denied(unit);
atomic_clear_mask(ZFCP_STATUS_UNIT_SHARED, &unit->status);
atomic_clear_mask(ZFCP_STATUS_UNIT_READONLY, &unit->status);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PORT_BOXED:
ZFCP_LOG_DEBUG("The remote port 0x%016Lx on adapter %s "
"needs to be reopened\n",
unit->port->wwpn, zfcp_get_busid_by_unit(unit));
debug_text_event(adapter->erp_dbf, 2, "fsf_s_pboxed");
[SCSI] zfcp: fix handling of port boxed and lun boxed fsf states From: Maxim Shchetynin <maxim@de.ibm.com> Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:23:57 +08:00
zfcp_erp_port_boxed(unit->port);
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 06:20:36 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR |
ZFCP_STATUS_FSFREQ_RETRY;
break;
case FSF_LUN_SHARING_VIOLATION:
if (header->fsf_status_qual.word[0] != 0) {
ZFCP_LOG_NORMAL("FCP-LUN 0x%Lx at the remote port "
"with WWPN 0x%Lx "
"connected to the adapter %s "
"is already in use in LPAR%d, CSS%d\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
queue_designator->hla,
queue_designator->cssid);
} else {
subtable = header->fsf_status_qual.halfword[4];
rule = header->fsf_status_qual.halfword[5];
switch (subtable) {
case FSF_SQ_CFDC_SUBTABLE_OS:
case FSF_SQ_CFDC_SUBTABLE_PORT_WWPN:
case FSF_SQ_CFDC_SUBTABLE_PORT_DID:
case FSF_SQ_CFDC_SUBTABLE_LUN:
ZFCP_LOG_NORMAL("Access to FCP-LUN 0x%Lx at the "
"remote port with WWPN 0x%Lx "
"connected to the adapter %s "
"is denied (%s rule %d)\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
zfcp_act_subtable_type[subtable],
rule);
break;
}
}
ZFCP_LOG_DEBUG("status qualifier:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &header->fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(adapter->erp_dbf, 2,
"fsf_s_l_sh_vio");
zfcp_erp_unit_access_denied(unit);
atomic_clear_mask(ZFCP_STATUS_UNIT_SHARED, &unit->status);
atomic_clear_mask(ZFCP_STATUS_UNIT_READONLY, &unit->status);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_MAXIMUM_NUMBER_OF_LUNS_EXCEEDED:
ZFCP_LOG_INFO("error: The adapter ran out of resources. "
"There is no handle (temporary port identifier) "
"available for unit 0x%016Lx on port 0x%016Lx "
"on adapter %s\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
debug_text_event(adapter->erp_dbf, 1,
"fsf_s_max_units");
zfcp_erp_unit_failed(unit);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_ADAPTER_STATUS_AVAILABLE:
switch (header->fsf_status_qual.word[0]) {
case FSF_SQ_INVOKE_LINK_TEST_PROCEDURE:
/* Re-establish link to port */
debug_text_event(adapter->erp_dbf, 1,
"fsf_sq_ltest");
[SCSI] zfcp: fix: reopen port only if link-test fails From: Maxim Shchetynin <maxim@de.ibm.com> Reopen a remote port only if the link-test fails. This avoids that a port is unnecessarily reopened. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:16:27 +08:00
zfcp_test_link(unit->port);
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 06:20:36 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_ULP_DEPENDENT_ERP_REQUIRED:
/* ERP strategy will escalate */
debug_text_event(adapter->erp_dbf, 1,
"fsf_sq_ulp");
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_NORMAL
("bug: Wrong status qualifier 0x%x arrived.\n",
header->fsf_status_qual.word[0]);
debug_text_event(adapter->erp_dbf, 0,
"fsf_sq_inval:");
debug_exception(adapter->erp_dbf, 0,
&header->fsf_status_qual.word[0],
sizeof (u32));
}
break;
case FSF_INVALID_COMMAND_OPTION:
ZFCP_LOG_NORMAL(
"Invalid option 0x%x has been specified "
"in QTCB bottom sent to the adapter %s\n",
bottom->option,
zfcp_get_busid_by_adapter(adapter));
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EINVAL;
break;
case FSF_GOOD:
/* save LUN handle assigned by FSF */
unit->handle = header->lun_handle;
ZFCP_LOG_TRACE("unit 0x%016Lx on remote port 0x%016Lx on "
"adapter %s opened, port handle 0x%x\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
unit->handle);
/* mark unit as open */
atomic_set_mask(ZFCP_STATUS_COMMON_OPEN, &unit->status);
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
if (!(adapter->connection_features & FSF_FEATURE_NPIV_MODE) &&
(adapter->adapter_features & FSF_FEATURE_LUN_SHARING) &&
(adapter->ccw_device->id.dev_model != ZFCP_DEVICE_MODEL_PRIV)) {
exclusive = (bottom->lun_access_info &
FSF_UNIT_ACCESS_EXCLUSIVE);
readwrite = (bottom->lun_access_info &
FSF_UNIT_ACCESS_OUTBOUND_TRANSFER);
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 06:20:36 +08:00
if (!exclusive)
atomic_set_mask(ZFCP_STATUS_UNIT_SHARED,
&unit->status);
if (!readwrite) {
atomic_set_mask(ZFCP_STATUS_UNIT_READONLY,
&unit->status);
ZFCP_LOG_NORMAL("read-only access for unit "
"(adapter %s, wwpn=0x%016Lx, "
"fcp_lun=0x%016Lx)\n",
zfcp_get_busid_by_unit(unit),
unit->port->wwpn,
unit->fcp_lun);
}
if (exclusive && !readwrite) {
ZFCP_LOG_NORMAL("exclusive access of read-only "
"unit not supported\n");
zfcp_erp_unit_failed(unit);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
zfcp_erp_unit_shutdown(unit, 0);
} else if (!exclusive && readwrite) {
ZFCP_LOG_NORMAL("shared access of read-write "
"unit not supported\n");
zfcp_erp_unit_failed(unit);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
zfcp_erp_unit_shutdown(unit, 0);
}
}
retval = 0;
break;
default:
ZFCP_LOG_NORMAL("bug: An unknown FSF Status was presented "
"(debug info 0x%x)\n",
header->fsf_status);
debug_text_event(adapter->erp_dbf, 0, "fsf_s_inval:");
debug_exception(adapter->erp_dbf, 0,
&header->fsf_status, sizeof (u32));
break;
}
skip_fsfstatus:
atomic_clear_mask(ZFCP_STATUS_COMMON_OPENING, &unit->status);
return retval;
}
/*
* function: zfcp_fsf_close_unit
*
* purpose:
*
* returns: address of fsf_req - request successfully initiated
* NULL -
*
* assumptions: This routine does not check whether the associated
* remote port/lun has already been opened. This should be
* done by calling routines. Otherwise some status
* may be presented by FSF
*/
int
zfcp_fsf_close_unit(struct zfcp_erp_action *erp_action)
{
volatile struct qdio_buffer_element *sbale;
unsigned long lock_flags;
int retval = 0;
/* setup new FSF request */
retval = zfcp_fsf_req_create(erp_action->adapter,
FSF_QTCB_CLOSE_LUN,
ZFCP_WAIT_FOR_SBAL | ZFCP_REQ_AUTO_CLEANUP,
erp_action->adapter->pool.fsf_req_erp,
&lock_flags, &(erp_action->fsf_req));
if (retval < 0) {
ZFCP_LOG_INFO("error: Could not create close unit request for "
"unit 0x%016Lx on port 0x%016Lx on adapter %s.\n",
erp_action->unit->fcp_lun,
erp_action->port->wwpn,
zfcp_get_busid_by_adapter(erp_action->adapter));
goto out;
}
sbale = zfcp_qdio_sbale_req(erp_action->fsf_req,
erp_action->fsf_req->sbal_curr, 0);
sbale[0].flags |= SBAL_FLAGS0_TYPE_READ;
sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY;
erp_action->fsf_req->qtcb->header.port_handle =
erp_action->port->handle;
erp_action->fsf_req->qtcb->header.lun_handle = erp_action->unit->handle;
atomic_set_mask(ZFCP_STATUS_COMMON_CLOSING, &erp_action->unit->status);
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
erp_action->fsf_req->data = (unsigned long) erp_action->unit;
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 06:20:36 +08:00
erp_action->fsf_req->erp_action = erp_action;
/* start QDIO request for this FSF request */
retval = zfcp_fsf_req_send(erp_action->fsf_req, &erp_action->timer);
if (retval) {
ZFCP_LOG_INFO("error: Could not send a close unit request for "
"unit 0x%016Lx on port 0x%016Lx onadapter %s.\n",
erp_action->unit->fcp_lun,
erp_action->port->wwpn,
zfcp_get_busid_by_adapter(erp_action->adapter));
zfcp_fsf_req_free(erp_action->fsf_req);
erp_action->fsf_req = NULL;
goto out;
}
ZFCP_LOG_TRACE("Close LUN request initiated (adapter %s, "
"port 0x%016Lx, unit 0x%016Lx)\n",
zfcp_get_busid_by_adapter(erp_action->adapter),
erp_action->port->wwpn, erp_action->unit->fcp_lun);
out:
write_unlock_irqrestore(&erp_action->adapter->request_queue.queue_lock,
lock_flags);
return retval;
}
/*
* function: zfcp_fsf_close_unit_handler
*
* purpose: is called for finished Close LUN FSF command
*
* returns:
*/
static int
zfcp_fsf_close_unit_handler(struct zfcp_fsf_req *fsf_req)
{
int retval = -EINVAL;
struct zfcp_unit *unit;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
unit = (struct zfcp_unit *) fsf_req->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 06:20:36 +08:00
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) {
/* don't change unit status in our bookkeeping */
goto skip_fsfstatus;
}
/* evaluate FSF status in QTCB */
switch (fsf_req->qtcb->header.fsf_status) {
case FSF_PORT_HANDLE_NOT_VALID:
ZFCP_LOG_INFO("Temporary port identifier 0x%x for port "
"0x%016Lx on adapter %s invalid. This may "
"happen in rare circumstances\n",
unit->port->handle,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
ZFCP_LOG_DEBUG("status qualifier:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &fsf_req->qtcb->header.fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_s_phand_nv");
zfcp_erp_adapter_reopen(unit->port->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_LUN_HANDLE_NOT_VALID:
ZFCP_LOG_INFO("Temporary LUN identifier 0x%x of unit "
"0x%016Lx on port 0x%016Lx on adapter %s is "
"invalid. This may happen occasionally.\n",
unit->handle,
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
ZFCP_LOG_DEBUG("Status qualifier data:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &fsf_req->qtcb->header.fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_s_lhand_nv");
zfcp_erp_port_reopen(unit->port, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PORT_BOXED:
ZFCP_LOG_DEBUG("The remote port 0x%016Lx on adapter %s "
"needs to be reopened\n",
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
debug_text_event(fsf_req->adapter->erp_dbf, 2, "fsf_s_pboxed");
[SCSI] zfcp: fix handling of port boxed and lun boxed fsf states From: Maxim Shchetynin <maxim@de.ibm.com> Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:23:57 +08:00
zfcp_erp_port_boxed(unit->port);
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 06:20:36 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR |
ZFCP_STATUS_FSFREQ_RETRY;
break;
case FSF_ADAPTER_STATUS_AVAILABLE:
switch (fsf_req->qtcb->header.fsf_status_qual.word[0]) {
case FSF_SQ_INVOKE_LINK_TEST_PROCEDURE:
/* re-establish link to port */
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_sq_ltest");
[SCSI] zfcp: fix: reopen port only if link-test fails From: Maxim Shchetynin <maxim@de.ibm.com> Reopen a remote port only if the link-test fails. This avoids that a port is unnecessarily reopened. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:16:27 +08:00
zfcp_test_link(unit->port);
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 06:20:36 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SQ_ULP_DEPENDENT_ERP_REQUIRED:
/* ERP strategy will escalate */
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_sq_ulp");
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
default:
ZFCP_LOG_NORMAL
("bug: Wrong status qualifier 0x%x arrived.\n",
fsf_req->qtcb->header.fsf_status_qual.word[0]);
debug_text_event(fsf_req->adapter->erp_dbf, 0,
"fsf_sq_inval:");
debug_exception(
fsf_req->adapter->erp_dbf, 0,
&fsf_req->qtcb->header.fsf_status_qual.word[0],
sizeof (u32));
break;
}
break;
case FSF_GOOD:
ZFCP_LOG_TRACE("unit 0x%016Lx on port 0x%016Lx on adapter %s "
"closed, port handle 0x%x\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
unit->handle);
/* mark unit as closed */
atomic_clear_mask(ZFCP_STATUS_COMMON_OPEN, &unit->status);
retval = 0;
break;
default:
ZFCP_LOG_NORMAL("bug: An unknown FSF Status was presented "
"(debug info 0x%x)\n",
fsf_req->qtcb->header.fsf_status);
debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf_s_inval:");
debug_exception(fsf_req->adapter->erp_dbf, 0,
&fsf_req->qtcb->header.fsf_status,
sizeof (u32));
break;
}
skip_fsfstatus:
atomic_clear_mask(ZFCP_STATUS_COMMON_CLOSING, &unit->status);
return retval;
}
/**
* zfcp_fsf_send_fcp_command_task - initiate an FCP command (for a SCSI command)
* @adapter: adapter where scsi command is issued
* @unit: unit where command is sent to
* @scsi_cmnd: scsi command to be sent
* @timer: timer to be started when request is initiated
* @req_flags: flags for fsf_request
*/
int
zfcp_fsf_send_fcp_command_task(struct zfcp_adapter *adapter,
struct zfcp_unit *unit,
struct scsi_cmnd * scsi_cmnd,
struct timer_list *timer, int req_flags)
{
struct zfcp_fsf_req *fsf_req = NULL;
struct fcp_cmnd_iu *fcp_cmnd_iu;
unsigned int sbtype;
unsigned long lock_flags;
int real_bytes = 0;
int retval = 0;
int mask;
/* setup new FSF request */
retval = zfcp_fsf_req_create(adapter, FSF_QTCB_FCP_CMND, req_flags,
adapter->pool.fsf_req_scsi,
&lock_flags, &fsf_req);
if (unlikely(retval < 0)) {
ZFCP_LOG_DEBUG("error: Could not create FCP command request "
"for unit 0x%016Lx on port 0x%016Lx on "
"adapter %s\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_adapter(adapter));
goto failed_req_create;
}
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
zfcp_unit_get(unit);
fsf_req->unit = unit;
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 06:20:36 +08:00
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
/* associate FSF request with SCSI request (for look up on abort) */
scsi_cmnd->host_scribble = (char *) fsf_req;
/* associate SCSI command with FSF request */
fsf_req->data = (unsigned long) scsi_cmnd;
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 06:20:36 +08:00
/* set handles of unit and its parent port in QTCB */
fsf_req->qtcb->header.lun_handle = unit->handle;
fsf_req->qtcb->header.port_handle = unit->port->handle;
/* FSF does not define the structure of the FCP_CMND IU */
fcp_cmnd_iu = (struct fcp_cmnd_iu *)
&(fsf_req->qtcb->bottom.io.fcp_cmnd);
/*
* set depending on data direction:
* data direction bits in SBALE (SB Type)
* data direction bits in QTCB
* data direction bits in FCP_CMND IU
*/
switch (scsi_cmnd->sc_data_direction) {
case DMA_NONE:
fsf_req->qtcb->bottom.io.data_direction = FSF_DATADIR_CMND;
/*
* FIXME(qdio):
* what is the correct type for commands
* without 'real' data buffers?
*/
sbtype = SBAL_FLAGS0_TYPE_READ;
break;
case DMA_FROM_DEVICE:
fsf_req->qtcb->bottom.io.data_direction = FSF_DATADIR_READ;
sbtype = SBAL_FLAGS0_TYPE_READ;
fcp_cmnd_iu->rddata = 1;
break;
case DMA_TO_DEVICE:
fsf_req->qtcb->bottom.io.data_direction = FSF_DATADIR_WRITE;
sbtype = SBAL_FLAGS0_TYPE_WRITE;
fcp_cmnd_iu->wddata = 1;
break;
case DMA_BIDIRECTIONAL:
default:
/*
* dummy, catch this condition earlier
* in zfcp_scsi_queuecommand
*/
goto failed_scsi_cmnd;
}
/* set FC service class in QTCB (3 per default) */
fsf_req->qtcb->bottom.io.service_class = adapter->fc_service_class;
/* set FCP_LUN in FCP_CMND IU in QTCB */
fcp_cmnd_iu->fcp_lun = unit->fcp_lun;
mask = ZFCP_STATUS_UNIT_READONLY | ZFCP_STATUS_UNIT_SHARED;
/* set task attributes in FCP_CMND IU in QTCB */
if (likely((scsi_cmnd->device->simple_tags) ||
(atomic_test_mask(mask, &unit->status))))
fcp_cmnd_iu->task_attribute = SIMPLE_Q;
else
fcp_cmnd_iu->task_attribute = UNTAGGED;
/* set additional length of FCP_CDB in FCP_CMND IU in QTCB, if needed */
if (unlikely(scsi_cmnd->cmd_len > FCP_CDB_LENGTH)) {
fcp_cmnd_iu->add_fcp_cdb_length
= (scsi_cmnd->cmd_len - FCP_CDB_LENGTH) >> 2;
ZFCP_LOG_TRACE("SCSI CDB length is 0x%x, "
"additional FCP_CDB length is 0x%x "
"(shifted right 2 bits)\n",
scsi_cmnd->cmd_len,
fcp_cmnd_iu->add_fcp_cdb_length);
}
/*
* copy SCSI CDB (including additional length, if any) to
* FCP_CDB in FCP_CMND IU in QTCB
*/
memcpy(fcp_cmnd_iu->fcp_cdb, scsi_cmnd->cmnd, scsi_cmnd->cmd_len);
/* FCP CMND IU length in QTCB */
fsf_req->qtcb->bottom.io.fcp_cmnd_length =
sizeof (struct fcp_cmnd_iu) +
fcp_cmnd_iu->add_fcp_cdb_length + sizeof (fcp_dl_t);
/* generate SBALEs from data buffer */
real_bytes = zfcp_qdio_sbals_from_scsicmnd(fsf_req, sbtype, scsi_cmnd);
if (unlikely(real_bytes < 0)) {
if (fsf_req->sbal_number < ZFCP_MAX_SBALS_PER_REQ) {
ZFCP_LOG_DEBUG(
"Data did not fit into available buffer(s), "
"waiting for more...\n");
retval = -EIO;
} else {
ZFCP_LOG_NORMAL("error: No truncation implemented but "
"required. Shutting down unit "
"(adapter %s, port 0x%016Lx, "
"unit 0x%016Lx)\n",
zfcp_get_busid_by_unit(unit),
unit->port->wwpn,
unit->fcp_lun);
zfcp_erp_unit_shutdown(unit, 0);
retval = -EINVAL;
}
goto no_fit;
}
/* set length of FCP data length in FCP_CMND IU in QTCB */
zfcp_set_fcp_dl(fcp_cmnd_iu, real_bytes);
ZFCP_LOG_DEBUG("Sending SCSI command:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) scsi_cmnd->cmnd, scsi_cmnd->cmd_len);
/*
* start QDIO request for this FSF request
* covered by an SBALE)
*/
retval = zfcp_fsf_req_send(fsf_req, timer);
if (unlikely(retval < 0)) {
ZFCP_LOG_INFO("error: Could not send FCP command request "
"on adapter %s, port 0x%016Lx, unit 0x%016Lx\n",
zfcp_get_busid_by_adapter(adapter),
unit->port->wwpn,
unit->fcp_lun);
goto send_failed;
}
ZFCP_LOG_TRACE("Send FCP Command initiated (adapter %s, "
"port 0x%016Lx, unit 0x%016Lx)\n",
zfcp_get_busid_by_adapter(adapter),
unit->port->wwpn,
unit->fcp_lun);
goto success;
send_failed:
no_fit:
failed_scsi_cmnd:
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
zfcp_unit_put(unit);
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 06:20:36 +08:00
zfcp_fsf_req_free(fsf_req);
fsf_req = NULL;
scsi_cmnd->host_scribble = NULL;
success:
failed_req_create:
write_unlock_irqrestore(&adapter->request_queue.queue_lock, lock_flags);
return retval;
}
/*
* function: zfcp_fsf_send_fcp_command_task_management
*
* purpose:
*
* returns:
*
* FIXME(design): should be watched by a timeout!!!
* FIXME(design) shouldn't this be modified to return an int
* also...don't know how though
*
*/
struct zfcp_fsf_req *
zfcp_fsf_send_fcp_command_task_management(struct zfcp_adapter *adapter,
struct zfcp_unit *unit,
u8 tm_flags, int req_flags)
{
struct zfcp_fsf_req *fsf_req = NULL;
int retval = 0;
struct fcp_cmnd_iu *fcp_cmnd_iu;
unsigned long lock_flags;
volatile struct qdio_buffer_element *sbale;
/* setup new FSF request */
retval = zfcp_fsf_req_create(adapter, FSF_QTCB_FCP_CMND, req_flags,
adapter->pool.fsf_req_scsi,
&lock_flags, &fsf_req);
if (retval < 0) {
ZFCP_LOG_INFO("error: Could not create FCP command (task "
"management) request for adapter %s, port "
" 0x%016Lx, unit 0x%016Lx.\n",
zfcp_get_busid_by_adapter(adapter),
unit->port->wwpn, unit->fcp_lun);
goto out;
}
/*
* Used to decide on proper handler in the return path,
* could be either zfcp_fsf_send_fcp_command_task_handler or
* zfcp_fsf_send_fcp_command_task_management_handler */
fsf_req->status |= ZFCP_STATUS_FSFREQ_TASK_MANAGEMENT;
/*
* hold a pointer to the unit being target of this
* task management request
*/
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
fsf_req->data = (unsigned long) unit;
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 06:20:36 +08:00
/* set FSF related fields in QTCB */
fsf_req->qtcb->header.lun_handle = unit->handle;
fsf_req->qtcb->header.port_handle = unit->port->handle;
fsf_req->qtcb->bottom.io.data_direction = FSF_DATADIR_CMND;
fsf_req->qtcb->bottom.io.service_class = adapter->fc_service_class;
fsf_req->qtcb->bottom.io.fcp_cmnd_length =
sizeof (struct fcp_cmnd_iu) + sizeof (fcp_dl_t);
sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0);
sbale[0].flags |= SBAL_FLAGS0_TYPE_WRITE;
sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY;
/* set FCP related fields in FCP_CMND IU in QTCB */
fcp_cmnd_iu = (struct fcp_cmnd_iu *)
&(fsf_req->qtcb->bottom.io.fcp_cmnd);
fcp_cmnd_iu->fcp_lun = unit->fcp_lun;
fcp_cmnd_iu->task_management_flags = tm_flags;
/* start QDIO request for this FSF request */
zfcp_fsf_start_scsi_er_timer(adapter);
retval = zfcp_fsf_req_send(fsf_req, NULL);
if (retval) {
del_timer(&adapter->scsi_er_timer);
ZFCP_LOG_INFO("error: Could not send an FCP-command (task "
"management) on adapter %s, port 0x%016Lx for "
"unit LUN 0x%016Lx\n",
zfcp_get_busid_by_adapter(adapter),
unit->port->wwpn,
unit->fcp_lun);
zfcp_fsf_req_free(fsf_req);
fsf_req = NULL;
goto out;
}
ZFCP_LOG_TRACE("Send FCP Command (task management function) initiated "
"(adapter %s, port 0x%016Lx, unit 0x%016Lx, "
"tm_flags=0x%x)\n",
zfcp_get_busid_by_adapter(adapter),
unit->port->wwpn,
unit->fcp_lun,
tm_flags);
out:
write_unlock_irqrestore(&adapter->request_queue.queue_lock, lock_flags);
return fsf_req;
}
/*
* function: zfcp_fsf_send_fcp_command_handler
*
* purpose: is called for finished Send FCP Command
*
* returns:
*/
static int
zfcp_fsf_send_fcp_command_handler(struct zfcp_fsf_req *fsf_req)
{
int retval = -EINVAL;
struct zfcp_unit *unit;
struct fsf_qtcb_header *header;
u16 subtable, rule, counter;
header = &fsf_req->qtcb->header;
if (unlikely(fsf_req->status & ZFCP_STATUS_FSFREQ_TASK_MANAGEMENT))
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
unit = (struct zfcp_unit *) fsf_req->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 06:20:36 +08:00
else
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
unit = fsf_req->unit;
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 06:20:36 +08:00
if (unlikely(fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR)) {
/* go directly to calls of special handlers */
goto skip_fsfstatus;
}
/* evaluate FSF status in QTCB */
switch (header->fsf_status) {
case FSF_PORT_HANDLE_NOT_VALID:
ZFCP_LOG_INFO("Temporary port identifier 0x%x for port "
"0x%016Lx on adapter %s invalid\n",
unit->port->handle,
unit->port->wwpn, zfcp_get_busid_by_unit(unit));
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &header->fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_s_phand_nv");
zfcp_erp_adapter_reopen(unit->port->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_LUN_HANDLE_NOT_VALID:
ZFCP_LOG_INFO("Temporary LUN identifier 0x%x for unit "
"0x%016Lx on port 0x%016Lx on adapter %s is "
"invalid. This may happen occasionally.\n",
unit->handle,
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
ZFCP_LOG_NORMAL("Status qualifier data:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_NORMAL,
(char *) &header->fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_s_uhand_nv");
zfcp_erp_port_reopen(unit->port, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_HANDLE_MISMATCH:
ZFCP_LOG_NORMAL("bug: The port handle 0x%x has changed "
"unexpectedly. (adapter %s, port 0x%016Lx, "
"unit 0x%016Lx)\n",
unit->port->handle,
zfcp_get_busid_by_unit(unit),
unit->port->wwpn,
unit->fcp_lun);
ZFCP_LOG_NORMAL("status qualifier:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_NORMAL,
(char *) &header->fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_s_hand_mis");
zfcp_erp_adapter_reopen(unit->port->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_SERVICE_CLASS_NOT_SUPPORTED:
if (fsf_req->adapter->fc_service_class <= 3) {
ZFCP_LOG_NORMAL("error: The adapter %s does "
"not support fibrechannel class %d.\n",
zfcp_get_busid_by_unit(unit),
fsf_req->adapter->fc_service_class);
} else {
ZFCP_LOG_NORMAL("bug: The fibrechannel class at "
"adapter %s is invalid. "
"(debug info %d)\n",
zfcp_get_busid_by_unit(unit),
fsf_req->adapter->fc_service_class);
}
/* stop operation for this adapter */
debug_text_exception(fsf_req->adapter->erp_dbf, 0,
"fsf_s_class_nsup");
zfcp_erp_adapter_shutdown(unit->port->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_FCPLUN_NOT_VALID:
ZFCP_LOG_NORMAL("bug: unit 0x%016Lx on port 0x%016Lx on "
"adapter %s does not have correct unit "
"handle 0x%x\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
unit->handle);
ZFCP_LOG_DEBUG("status qualifier:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &header->fsf_status_qual,
sizeof (union fsf_status_qual));
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_s_fcp_lun_nv");
zfcp_erp_port_reopen(unit->port, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_ACCESS_DENIED:
ZFCP_LOG_NORMAL("Access denied, cannot send FCP command to "
"unit 0x%016Lx on port 0x%016Lx on "
"adapter %s\n", unit->fcp_lun, unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
for (counter = 0; counter < 2; counter++) {
subtable = header->fsf_status_qual.halfword[counter * 2];
rule = header->fsf_status_qual.halfword[counter * 2 + 1];
switch (subtable) {
case FSF_SQ_CFDC_SUBTABLE_OS:
case FSF_SQ_CFDC_SUBTABLE_PORT_WWPN:
case FSF_SQ_CFDC_SUBTABLE_PORT_DID:
case FSF_SQ_CFDC_SUBTABLE_LUN:
ZFCP_LOG_INFO("Access denied (%s rule %d)\n",
zfcp_act_subtable_type[subtable], rule);
break;
}
}
debug_text_event(fsf_req->adapter->erp_dbf, 1, "fsf_s_access");
zfcp_erp_unit_access_denied(unit);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_DIRECTION_INDICATOR_NOT_VALID:
ZFCP_LOG_INFO("bug: Invalid data direction given for unit "
"0x%016Lx on port 0x%016Lx on adapter %s "
"(debug info %d)\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
fsf_req->qtcb->bottom.io.data_direction);
/* stop operation for this adapter */
debug_text_event(fsf_req->adapter->erp_dbf, 0,
"fsf_s_dir_ind_nv");
zfcp_erp_adapter_shutdown(unit->port->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_CMND_LENGTH_NOT_VALID:
ZFCP_LOG_NORMAL
("bug: An invalid control-data-block length field "
"was found in a command for unit 0x%016Lx on port "
"0x%016Lx on adapter %s " "(debug info %d)\n",
unit->fcp_lun, unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
fsf_req->qtcb->bottom.io.fcp_cmnd_length);
/* stop operation for this adapter */
debug_text_event(fsf_req->adapter->erp_dbf, 0,
"fsf_s_cmd_len_nv");
zfcp_erp_adapter_shutdown(unit->port->adapter, 0);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
break;
case FSF_PORT_BOXED:
ZFCP_LOG_DEBUG("The remote port 0x%016Lx on adapter %s "
"needs to be reopened\n",
unit->port->wwpn, zfcp_get_busid_by_unit(unit));
debug_text_event(fsf_req->adapter->erp_dbf, 2, "fsf_s_pboxed");
[SCSI] zfcp: fix handling of port boxed and lun boxed fsf states From: Maxim Shchetynin <maxim@de.ibm.com> Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:23:57 +08:00
zfcp_erp_port_boxed(unit->port);
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 06:20:36 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR |
ZFCP_STATUS_FSFREQ_RETRY;
break;
case FSF_LUN_BOXED:
ZFCP_LOG_NORMAL("unit needs to be reopened (adapter %s, "
"wwpn=0x%016Lx, fcp_lun=0x%016Lx)\n",
zfcp_get_busid_by_unit(unit),
unit->port->wwpn, unit->fcp_lun);
debug_text_event(fsf_req->adapter->erp_dbf, 1, "fsf_s_lboxed");
[SCSI] zfcp: fix handling of port boxed and lun boxed fsf states From: Maxim Shchetynin <maxim@de.ibm.com> Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:23:57 +08:00
zfcp_erp_unit_boxed(unit);
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 06:20:36 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR
| ZFCP_STATUS_FSFREQ_RETRY;
break;
case FSF_ADAPTER_STATUS_AVAILABLE:
switch (header->fsf_status_qual.word[0]) {
case FSF_SQ_INVOKE_LINK_TEST_PROCEDURE:
/* re-establish link to port */
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_sq_ltest");
[SCSI] zfcp: fix: reopen port only if link-test fails From: Maxim Shchetynin <maxim@de.ibm.com> Reopen a remote port only if the link-test fails. This avoids that a port is unnecessarily reopened. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:16:27 +08:00
zfcp_test_link(unit->port);
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 06:20:36 +08:00
break;
case FSF_SQ_ULP_DEPENDENT_ERP_REQUIRED:
/* FIXME(hw) need proper specs for proper action */
/* let scsi stack deal with retries and escalation */
debug_text_event(fsf_req->adapter->erp_dbf, 1,
"fsf_sq_ulp");
break;
default:
ZFCP_LOG_NORMAL
[SCSI] zfcp: fix: mark fsf request failed when receiving unknown status qualifier From: Maxim Shchetynin <maxim@de.ibm.com> Correct a bug in zfcp_fsf_send_fcp_command_handler. An fsf request was not marked as failed if an unknown status qualifier was returned. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:17:44 +08:00
("Unknown status qualifier 0x%x arrived.\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 06:20:36 +08:00
header->fsf_status_qual.word[0]);
debug_text_event(fsf_req->adapter->erp_dbf, 0,
"fsf_sq_inval:");
debug_exception(fsf_req->adapter->erp_dbf, 0,
&header->fsf_status_qual.word[0],
sizeof(u32));
break;
}
[SCSI] zfcp: fix: mark fsf request failed when receiving unknown status qualifier From: Maxim Shchetynin <maxim@de.ibm.com> Correct a bug in zfcp_fsf_send_fcp_command_handler. An fsf request was not marked as failed if an unknown status qualifier was returned. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:17:44 +08:00
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
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 06:20:36 +08:00
break;
case FSF_GOOD:
break;
case FSF_FCP_RSP_AVAILABLE:
break;
default:
debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf_s_inval:");
debug_exception(fsf_req->adapter->erp_dbf, 0,
&header->fsf_status, sizeof(u32));
break;
}
skip_fsfstatus:
if (fsf_req->status & ZFCP_STATUS_FSFREQ_TASK_MANAGEMENT) {
retval =
zfcp_fsf_send_fcp_command_task_management_handler(fsf_req);
} else {
retval = zfcp_fsf_send_fcp_command_task_handler(fsf_req);
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
fsf_req->unit = NULL;
zfcp_unit_put(unit);
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 06:20:36 +08:00
}
return retval;
}
/*
* function: zfcp_fsf_send_fcp_command_task_handler
*
* purpose: evaluates FCP_RSP IU
*
* returns:
*/
static int
zfcp_fsf_send_fcp_command_task_handler(struct zfcp_fsf_req *fsf_req)
{
int retval = 0;
struct scsi_cmnd *scpnt;
struct fcp_rsp_iu *fcp_rsp_iu = (struct fcp_rsp_iu *)
&(fsf_req->qtcb->bottom.io.fcp_rsp);
struct fcp_cmnd_iu *fcp_cmnd_iu = (struct fcp_cmnd_iu *)
&(fsf_req->qtcb->bottom.io.fcp_cmnd);
u32 sns_len;
char *fcp_rsp_info = zfcp_get_fcp_rsp_info_ptr(fcp_rsp_iu);
unsigned long flags;
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
struct zfcp_unit *unit = fsf_req->unit;
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 06:20:36 +08:00
read_lock_irqsave(&fsf_req->adapter->abort_lock, flags);
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
scpnt = (struct scsi_cmnd *) fsf_req->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 06:20:36 +08:00
if (unlikely(!scpnt)) {
ZFCP_LOG_DEBUG
("Command with fsf_req %p is not associated to "
"a scsi command anymore. Aborted?\n", fsf_req);
goto out;
}
if (unlikely(fsf_req->status & ZFCP_STATUS_FSFREQ_ABORTED)) {
/* FIXME: (design) mid-layer should handle DID_ABORT like
* DID_SOFT_ERROR by retrying the request for devices
* that allow retries.
*/
ZFCP_LOG_DEBUG("Setting DID_SOFT_ERROR and SUGGEST_RETRY\n");
set_host_byte(&scpnt->result, DID_SOFT_ERROR);
set_driver_byte(&scpnt->result, SUGGEST_RETRY);
goto skip_fsfstatus;
}
if (unlikely(fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR)) {
ZFCP_LOG_DEBUG("Setting DID_ERROR\n");
set_host_byte(&scpnt->result, DID_ERROR);
goto skip_fsfstatus;
}
/* set message byte of result in SCSI command */
scpnt->result |= COMMAND_COMPLETE << 8;
/*
* copy SCSI status code of FCP_STATUS of FCP_RSP IU to status byte
* of result in SCSI command
*/
scpnt->result |= fcp_rsp_iu->scsi_status;
if (unlikely(fcp_rsp_iu->scsi_status)) {
/* DEBUG */
ZFCP_LOG_DEBUG("status for SCSI Command:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
scpnt->cmnd, scpnt->cmd_len);
ZFCP_LOG_DEBUG("SCSI status code 0x%x\n",
fcp_rsp_iu->scsi_status);
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(void *) fcp_rsp_iu, sizeof (struct fcp_rsp_iu));
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
zfcp_get_fcp_sns_info_ptr(fcp_rsp_iu),
fcp_rsp_iu->fcp_sns_len);
}
/* check FCP_RSP_INFO */
if (unlikely(fcp_rsp_iu->validity.bits.fcp_rsp_len_valid)) {
ZFCP_LOG_DEBUG("rsp_len is valid\n");
switch (fcp_rsp_info[3]) {
case RSP_CODE_GOOD:
/* ok, continue */
ZFCP_LOG_TRACE("no failure or Task Management "
"Function complete\n");
set_host_byte(&scpnt->result, DID_OK);
break;
case RSP_CODE_LENGTH_MISMATCH:
/* hardware bug */
ZFCP_LOG_NORMAL("bug: FCP response code indictates "
"that the fibrechannel protocol data "
"length differs from the burst length. "
"The problem occured on unit 0x%016Lx "
"on port 0x%016Lx on adapter %s",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
/* dump SCSI CDB as prepared by zfcp */
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &fsf_req->qtcb->
bottom.io.fcp_cmnd, FSF_FCP_CMND_SIZE);
set_host_byte(&scpnt->result, DID_ERROR);
goto skip_fsfstatus;
case RSP_CODE_FIELD_INVALID:
/* driver or hardware bug */
ZFCP_LOG_NORMAL("bug: FCP response code indictates "
"that the fibrechannel protocol data "
"fields were incorrectly set up. "
"The problem occured on the unit "
"0x%016Lx on port 0x%016Lx on "
"adapter %s",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
/* dump SCSI CDB as prepared by zfcp */
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &fsf_req->qtcb->
bottom.io.fcp_cmnd, FSF_FCP_CMND_SIZE);
set_host_byte(&scpnt->result, DID_ERROR);
goto skip_fsfstatus;
case RSP_CODE_RO_MISMATCH:
/* hardware bug */
ZFCP_LOG_NORMAL("bug: The FCP response code indicates "
"that conflicting values for the "
"fibrechannel payload offset from the "
"header were found. "
"The problem occured on unit 0x%016Lx "
"on port 0x%016Lx on adapter %s.\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
/* dump SCSI CDB as prepared by zfcp */
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &fsf_req->qtcb->
bottom.io.fcp_cmnd, FSF_FCP_CMND_SIZE);
set_host_byte(&scpnt->result, DID_ERROR);
goto skip_fsfstatus;
default:
ZFCP_LOG_NORMAL("bug: An invalid FCP response "
"code was detected for a command. "
"The problem occured on the unit "
"0x%016Lx on port 0x%016Lx on "
"adapter %s (debug info 0x%x)\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
fcp_rsp_info[3]);
/* dump SCSI CDB as prepared by zfcp */
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG,
(char *) &fsf_req->qtcb->
bottom.io.fcp_cmnd, FSF_FCP_CMND_SIZE);
set_host_byte(&scpnt->result, DID_ERROR);
zfcp: add point-2-point support From: Andreas Herrmann <aherrman@de.ibm.com> This patch mainly introduces support for point-2-point topology. From: Heiko Carstens <heiko.carstens@de.ibm.com> From: Maxim Shchetynin <maxim@de.ibm.com> From: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-11 12:04:28 +08:00
goto skip_fsfstatus;
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 06:20:36 +08:00
}
}
/* check for sense data */
if (unlikely(fcp_rsp_iu->validity.bits.fcp_sns_len_valid)) {
sns_len = FSF_FCP_RSP_SIZE -
sizeof (struct fcp_rsp_iu) + fcp_rsp_iu->fcp_rsp_len;
ZFCP_LOG_TRACE("room for %i bytes sense data in QTCB\n",
sns_len);
sns_len = min(sns_len, (u32) SCSI_SENSE_BUFFERSIZE);
ZFCP_LOG_TRACE("room for %i bytes sense data in SCSI command\n",
SCSI_SENSE_BUFFERSIZE);
sns_len = min(sns_len, fcp_rsp_iu->fcp_sns_len);
ZFCP_LOG_TRACE("scpnt->result =0x%x, command was:\n",
scpnt->result);
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_TRACE,
(void *) &scpnt->cmnd, scpnt->cmd_len);
ZFCP_LOG_TRACE("%i bytes sense data provided by FCP\n",
fcp_rsp_iu->fcp_sns_len);
memcpy(&scpnt->sense_buffer,
zfcp_get_fcp_sns_info_ptr(fcp_rsp_iu), sns_len);
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_TRACE,
(void *) &scpnt->sense_buffer, sns_len);
}
/* check for overrun */
if (unlikely(fcp_rsp_iu->validity.bits.fcp_resid_over)) {
ZFCP_LOG_INFO("A data overrun was detected for a command. "
"unit 0x%016Lx, port 0x%016Lx, adapter %s. "
"The response data length is "
"%d, the original length was %d.\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
fcp_rsp_iu->fcp_resid,
(int) zfcp_get_fcp_dl(fcp_cmnd_iu));
}
/* check for underrun */
if (unlikely(fcp_rsp_iu->validity.bits.fcp_resid_under)) {
ZFCP_LOG_INFO("A data underrun was detected for a command. "
"unit 0x%016Lx, port 0x%016Lx, adapter %s. "
"The response data length is "
"%d, the original length was %d.\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
fcp_rsp_iu->fcp_resid,
(int) zfcp_get_fcp_dl(fcp_cmnd_iu));
scpnt->resid = fcp_rsp_iu->fcp_resid;
if (scpnt->request_bufflen - scpnt->resid < scpnt->underflow)
zfcp: add point-2-point support From: Andreas Herrmann <aherrman@de.ibm.com> This patch mainly introduces support for point-2-point topology. From: Heiko Carstens <heiko.carstens@de.ibm.com> From: Maxim Shchetynin <maxim@de.ibm.com> From: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-11 12:04:28 +08:00
set_host_byte(&scpnt->result, DID_ERROR);
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 06:20:36 +08:00
}
skip_fsfstatus:
ZFCP_LOG_DEBUG("scpnt->result =0x%x\n", scpnt->result);
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
if (scpnt->result != 0)
zfcp_scsi_dbf_event_result("erro", 3, fsf_req->adapter, scpnt);
else if (scpnt->retries > 0)
zfcp_scsi_dbf_event_result("retr", 4, fsf_req->adapter, scpnt);
else
zfcp_scsi_dbf_event_result("norm", 6, fsf_req->adapter, scpnt);
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 06:20:36 +08:00
/* cleanup pointer (need this especially for abort) */
scpnt->host_scribble = NULL;
/* always call back */
(scpnt->scsi_done) (scpnt);
/*
* We must hold this lock until scsi_done has been called.
* Otherwise we may call scsi_done after abort regarding this
* command has completed.
* Note: scsi_done must not block!
*/
out:
read_unlock_irqrestore(&fsf_req->adapter->abort_lock, flags);
return retval;
}
/*
* function: zfcp_fsf_send_fcp_command_task_management_handler
*
* purpose: evaluates FCP_RSP IU
*
* returns:
*/
static int
zfcp_fsf_send_fcp_command_task_management_handler(struct zfcp_fsf_req *fsf_req)
{
int retval = 0;
struct fcp_rsp_iu *fcp_rsp_iu = (struct fcp_rsp_iu *)
&(fsf_req->qtcb->bottom.io.fcp_rsp);
char *fcp_rsp_info = zfcp_get_fcp_rsp_info_ptr(fcp_rsp_iu);
[SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:47:52 +08:00
struct zfcp_unit *unit = (struct zfcp_unit *) fsf_req->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 06:20:36 +08:00
del_timer(&fsf_req->adapter->scsi_er_timer);
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) {
fsf_req->status |= ZFCP_STATUS_FSFREQ_TMFUNCFAILED;
goto skip_fsfstatus;
}
/* check FCP_RSP_INFO */
switch (fcp_rsp_info[3]) {
case RSP_CODE_GOOD:
/* ok, continue */
ZFCP_LOG_DEBUG("no failure or Task Management "
"Function complete\n");
break;
case RSP_CODE_TASKMAN_UNSUPP:
ZFCP_LOG_NORMAL("bug: A reuested task management function "
"is not supported on the target device "
"unit 0x%016Lx, port 0x%016Lx, adapter %s\n ",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
fsf_req->status |= ZFCP_STATUS_FSFREQ_TMFUNCNOTSUPP;
break;
case RSP_CODE_TASKMAN_FAILED:
ZFCP_LOG_NORMAL("bug: A reuested task management function "
"failed to complete successfully. "
"unit 0x%016Lx, port 0x%016Lx, adapter %s.\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit));
fsf_req->status |= ZFCP_STATUS_FSFREQ_TMFUNCFAILED;
break;
default:
ZFCP_LOG_NORMAL("bug: An invalid FCP response "
"code was detected for a command. "
"unit 0x%016Lx, port 0x%016Lx, adapter %s "
"(debug info 0x%x)\n",
unit->fcp_lun,
unit->port->wwpn,
zfcp_get_busid_by_unit(unit),
fcp_rsp_info[3]);
fsf_req->status |= ZFCP_STATUS_FSFREQ_TMFUNCFAILED;
}
skip_fsfstatus:
return retval;
}
/*
* function: zfcp_fsf_control_file
*
* purpose: Initiator of the control file upload/download FSF requests
*
* returns: 0 - FSF request is successfuly created and queued
* -EOPNOTSUPP - The FCP adapter does not have Control File support
* -EINVAL - Invalid direction specified
* -ENOMEM - Insufficient memory
* -EPERM - Cannot create FSF request or place it in QDIO queue
*/
int
zfcp_fsf_control_file(struct zfcp_adapter *adapter,
struct zfcp_fsf_req **fsf_req_ptr,
u32 fsf_command,
u32 option,
struct zfcp_sg_list *sg_list)
{
struct zfcp_fsf_req *fsf_req;
struct fsf_qtcb_bottom_support *bottom;
volatile struct qdio_buffer_element *sbale;
struct timer_list *timer;
unsigned long lock_flags;
int req_flags = 0;
int direction;
int retval = 0;
[SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:51:16 +08:00
if (!(adapter->adapter_features & FSF_FEATURE_CFDC)) {
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 06:20:36 +08:00
ZFCP_LOG_INFO("cfdc not supported (adapter %s)\n",
zfcp_get_busid_by_adapter(adapter));
retval = -EOPNOTSUPP;
goto out;
}
switch (fsf_command) {
case FSF_QTCB_DOWNLOAD_CONTROL_FILE:
direction = SBAL_FLAGS0_TYPE_WRITE;
if ((option != FSF_CFDC_OPTION_FULL_ACCESS) &&
(option != FSF_CFDC_OPTION_RESTRICTED_ACCESS))
req_flags = ZFCP_WAIT_FOR_SBAL;
break;
case FSF_QTCB_UPLOAD_CONTROL_FILE:
direction = SBAL_FLAGS0_TYPE_READ;
break;
default:
ZFCP_LOG_INFO("Invalid FSF command code 0x%08x\n", fsf_command);
retval = -EINVAL;
goto out;
}
timer = kmalloc(sizeof(struct timer_list), GFP_KERNEL);
if (!timer) {
retval = -ENOMEM;
goto out;
}
retval = zfcp_fsf_req_create(adapter, fsf_command, req_flags,
NULL, &lock_flags, &fsf_req);
if (retval < 0) {
ZFCP_LOG_INFO("error: Could not create FSF request for the "
"adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
retval = -EPERM;
goto unlock_queue_lock;
}
sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0);
sbale[0].flags |= direction;
bottom = &fsf_req->qtcb->bottom.support;
bottom->operation_subtype = FSF_CFDC_OPERATION_SUBTYPE;
bottom->option = option;
if (sg_list->count > 0) {
int bytes;
bytes = zfcp_qdio_sbals_from_sg(fsf_req, direction,
sg_list->sg, sg_list->count,
ZFCP_MAX_SBALS_PER_REQ);
if (bytes != ZFCP_CFDC_MAX_CONTROL_FILE_SIZE) {
ZFCP_LOG_INFO(
"error: Could not create sufficient number of "
"SBALS for an FSF request to the adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
retval = -ENOMEM;
goto free_fsf_req;
}
} else
sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY;
init_timer(timer);
timer->function = zfcp_fsf_request_timeout_handler;
timer->data = (unsigned long) adapter;
timer->expires = ZFCP_FSF_REQUEST_TIMEOUT;
retval = zfcp_fsf_req_send(fsf_req, timer);
if (retval < 0) {
ZFCP_LOG_INFO("initiation of cfdc up/download failed"
"(adapter %s)\n",
zfcp_get_busid_by_adapter(adapter));
retval = -EPERM;
goto free_fsf_req;
}
write_unlock_irqrestore(&adapter->request_queue.queue_lock, lock_flags);
ZFCP_LOG_NORMAL("Control file %s FSF request has been sent to the "
"adapter %s\n",
fsf_command == FSF_QTCB_DOWNLOAD_CONTROL_FILE ?
"download" : "upload",
zfcp_get_busid_by_adapter(adapter));
wait_event(fsf_req->completion_wq,
fsf_req->status & ZFCP_STATUS_FSFREQ_COMPLETED);
*fsf_req_ptr = fsf_req;
del_timer_sync(timer);
goto free_timer;
free_fsf_req:
zfcp_fsf_req_free(fsf_req);
unlock_queue_lock:
write_unlock_irqrestore(&adapter->request_queue.queue_lock, lock_flags);
free_timer:
kfree(timer);
out:
return retval;
}
/*
* function: zfcp_fsf_control_file_handler
*
* purpose: Handler of the control file upload/download FSF requests
*
* returns: 0 - FSF request successfuly processed
* -EAGAIN - Operation has to be repeated because of a temporary problem
* -EACCES - There is no permission to execute an operation
* -EPERM - The control file is not in a right format
* -EIO - There is a problem with the FCP adapter
* -EINVAL - Invalid operation
* -EFAULT - User space memory I/O operation fault
*/
static int
zfcp_fsf_control_file_handler(struct zfcp_fsf_req *fsf_req)
{
struct zfcp_adapter *adapter = fsf_req->adapter;
struct fsf_qtcb_header *header = &fsf_req->qtcb->header;
struct fsf_qtcb_bottom_support *bottom = &fsf_req->qtcb->bottom.support;
int retval = 0;
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) {
retval = -EINVAL;
goto skip_fsfstatus;
}
switch (header->fsf_status) {
case FSF_GOOD:
ZFCP_LOG_NORMAL(
"The FSF request has been successfully completed "
"on the adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_OPERATION_PARTIALLY_SUCCESSFUL:
if (bottom->operation_subtype == FSF_CFDC_OPERATION_SUBTYPE) {
switch (header->fsf_status_qual.word[0]) {
zfcp: add point-2-point support From: Andreas Herrmann <aherrman@de.ibm.com> This patch mainly introduces support for point-2-point topology. From: Heiko Carstens <heiko.carstens@de.ibm.com> From: Maxim Shchetynin <maxim@de.ibm.com> From: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-04-11 12:04:28 +08:00
case FSF_SQ_CFDC_HARDENED_ON_SE:
ZFCP_LOG_NORMAL(
"CFDC on the adapter %s has being "
"hardened on primary and secondary SE\n",
zfcp_get_busid_by_adapter(adapter));
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 06:20:36 +08:00
case FSF_SQ_CFDC_COULD_NOT_HARDEN_ON_SE:
ZFCP_LOG_NORMAL(
"CFDC of the adapter %s could not "
"be saved on the SE\n",
zfcp_get_busid_by_adapter(adapter));
break;
case FSF_SQ_CFDC_COULD_NOT_HARDEN_ON_SE2:
ZFCP_LOG_NORMAL(
"CFDC of the adapter %s could not "
"be copied to the secondary SE\n",
zfcp_get_busid_by_adapter(adapter));
break;
default:
ZFCP_LOG_NORMAL(
"CFDC could not be hardened "
"on the adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
}
}
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EAGAIN;
break;
case FSF_AUTHORIZATION_FAILURE:
ZFCP_LOG_NORMAL(
"Adapter %s does not accept privileged commands\n",
zfcp_get_busid_by_adapter(adapter));
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EACCES;
break;
case FSF_CFDC_ERROR_DETECTED:
ZFCP_LOG_NORMAL(
"Error at position %d in the CFDC, "
"CFDC is discarded by the adapter %s\n",
header->fsf_status_qual.word[0],
zfcp_get_busid_by_adapter(adapter));
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EPERM;
break;
case FSF_CONTROL_FILE_UPDATE_ERROR:
ZFCP_LOG_NORMAL(
"Adapter %s cannot harden the control file, "
"file is discarded\n",
zfcp_get_busid_by_adapter(adapter));
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EIO;
break;
case FSF_CONTROL_FILE_TOO_LARGE:
ZFCP_LOG_NORMAL(
"Control file is too large, file is discarded "
"by the adapter %s\n",
zfcp_get_busid_by_adapter(adapter));
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EIO;
break;
case FSF_ACCESS_CONFLICT_DETECTED:
if (bottom->operation_subtype == FSF_CFDC_OPERATION_SUBTYPE)
ZFCP_LOG_NORMAL(
"CFDC has been discarded by the adapter %s, "
"because activation would impact "
"%d active connection(s)\n",
zfcp_get_busid_by_adapter(adapter),
header->fsf_status_qual.word[0]);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EIO;
break;
case FSF_CONFLICTS_OVERRULED:
if (bottom->operation_subtype == FSF_CFDC_OPERATION_SUBTYPE)
ZFCP_LOG_NORMAL(
"CFDC has been activated on the adapter %s, "
"but activation has impacted "
"%d active connection(s)\n",
zfcp_get_busid_by_adapter(adapter),
header->fsf_status_qual.word[0]);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EIO;
break;
case FSF_UNKNOWN_OP_SUBTYPE:
ZFCP_LOG_NORMAL("unknown operation subtype (adapter: %s, "
"op_subtype=0x%x)\n",
zfcp_get_busid_by_adapter(adapter),
bottom->operation_subtype);
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EINVAL;
break;
case FSF_INVALID_COMMAND_OPTION:
ZFCP_LOG_NORMAL(
"Invalid option 0x%x has been specified "
"in QTCB bottom sent to the adapter %s\n",
bottom->option,
zfcp_get_busid_by_adapter(adapter));
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EINVAL;
break;
default:
ZFCP_LOG_NORMAL(
"bug: An unknown/unexpected FSF status 0x%08x "
"was presented on the adapter %s\n",
header->fsf_status,
zfcp_get_busid_by_adapter(adapter));
debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf_sq_inval");
debug_exception(fsf_req->adapter->erp_dbf, 0,
&header->fsf_status_qual.word[0], sizeof(u32));
fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR;
retval = -EINVAL;
break;
}
skip_fsfstatus:
return retval;
}
static inline int
zfcp_fsf_req_sbal_check(unsigned long *flags,
struct zfcp_qdio_queue *queue, int needed)
{
write_lock_irqsave(&queue->queue_lock, *flags);
if (likely(atomic_read(&queue->free_count) >= needed))
return 1;
write_unlock_irqrestore(&queue->queue_lock, *flags);
return 0;
}
/*
* set qtcb pointer in fsf_req and initialize QTCB
*/
static inline void
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
zfcp_fsf_req_qtcb_init(struct zfcp_fsf_req *fsf_req)
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 06:20:36 +08:00
{
if (likely(fsf_req->qtcb != NULL)) {
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
fsf_req->qtcb->prefix.req_seq_no = fsf_req->adapter->fsf_req_seq_no;
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 06:20:36 +08:00
fsf_req->qtcb->prefix.req_id = (unsigned long)fsf_req;
fsf_req->qtcb->prefix.ulp_info = ZFCP_ULP_INFO_VERSION;
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
fsf_req->qtcb->prefix.qtcb_type = fsf_qtcb_type[fsf_req->fsf_command];
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 06:20:36 +08:00
fsf_req->qtcb->prefix.qtcb_version = ZFCP_QTCB_VERSION;
fsf_req->qtcb->header.req_handle = (unsigned long)fsf_req;
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
fsf_req->qtcb->header.fsf_command = fsf_req->fsf_command;
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 06:20:36 +08:00
}
}
/**
* zfcp_fsf_req_sbal_get - try to get one SBAL in the request queue
* @adapter: adapter for which request queue is examined
* @req_flags: flags indicating whether to wait for needed SBAL or not
* @lock_flags: lock_flags if queue_lock is taken
* Return: 0 on success, otherwise -EIO, or -ERESTARTSYS
* Locks: lock adapter->request_queue->queue_lock on success
*/
static int
zfcp_fsf_req_sbal_get(struct zfcp_adapter *adapter, int req_flags,
unsigned long *lock_flags)
{
long ret;
struct zfcp_qdio_queue *req_queue = &adapter->request_queue;
if (unlikely(req_flags & ZFCP_WAIT_FOR_SBAL)) {
ret = wait_event_interruptible_timeout(adapter->request_wq,
zfcp_fsf_req_sbal_check(lock_flags, req_queue, 1),
ZFCP_SBAL_TIMEOUT);
if (ret < 0)
return ret;
if (!ret)
return -EIO;
} else if (!zfcp_fsf_req_sbal_check(lock_flags, req_queue, 1))
return -EIO;
return 0;
}
/*
* function: zfcp_fsf_req_create
*
* purpose: create an FSF request at the specified adapter and
* setup common fields
*
* returns: -ENOMEM if there was insufficient memory for a request
* -EIO if no qdio buffers could be allocate to the request
* -EINVAL/-EPERM on bug conditions in req_dequeue
* 0 in success
*
* note: The created request is returned by reference.
*
* locks: lock of concerned request queue must not be held,
* but is held on completion (write, irqsave)
*/
int
zfcp_fsf_req_create(struct zfcp_adapter *adapter, u32 fsf_cmd, int req_flags,
mempool_t *pool, unsigned long *lock_flags,
struct zfcp_fsf_req **fsf_req_p)
{
volatile struct qdio_buffer_element *sbale;
struct zfcp_fsf_req *fsf_req = NULL;
int ret = 0;
struct zfcp_qdio_queue *req_queue = &adapter->request_queue;
/* allocate new FSF request */
fsf_req = zfcp_fsf_req_alloc(pool, req_flags);
if (unlikely(NULL == fsf_req)) {
ZFCP_LOG_DEBUG("error: Could not put an FSF request into"
"the outbound (send) queue.\n");
ret = -ENOMEM;
goto failed_fsf_req;
}
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
fsf_req->adapter = adapter;
fsf_req->fsf_command = fsf_cmd;
zfcp_fsf_req_qtcb_init(fsf_req);
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 06:20:36 +08:00
/* initialize waitqueue which may be used to wait on
this request completion */
init_waitqueue_head(&fsf_req->completion_wq);
ret = zfcp_fsf_req_sbal_get(adapter, req_flags, lock_flags);
if(ret < 0) {
goto failed_sbals;
}
/*
* We hold queue_lock here. Check if QDIOUP is set and let request fail
* if it is not set (see also *_open_qdio and *_close_qdio).
*/
if (!atomic_test_mask(ZFCP_STATUS_ADAPTER_QDIOUP, &adapter->status)) {
write_unlock_irqrestore(&req_queue->queue_lock, *lock_flags);
ret = -EIO;
goto failed_sbals;
}
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
if (fsf_req->qtcb) {
fsf_req->seq_no = adapter->fsf_req_seq_no;
fsf_req->qtcb->prefix.req_seq_no = adapter->fsf_req_seq_no;
}
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 06:20:36 +08:00
fsf_req->sbal_number = 1;
fsf_req->sbal_first = req_queue->free_index;
fsf_req->sbal_curr = req_queue->free_index;
fsf_req->sbale_curr = 1;
if (likely(req_flags & ZFCP_REQ_AUTO_CLEANUP)) {
fsf_req->status |= ZFCP_STATUS_FSFREQ_CLEANUP;
}
sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0);
/* setup common SBALE fields */
sbale[0].addr = fsf_req;
sbale[0].flags |= SBAL_FLAGS0_COMMAND;
if (likely(fsf_req->qtcb != NULL)) {
sbale[1].addr = (void *) fsf_req->qtcb;
sbale[1].length = sizeof(struct fsf_qtcb);
}
ZFCP_LOG_TRACE("got %i free BUFFERs starting at index %i\n",
fsf_req->sbal_number, fsf_req->sbal_first);
goto success;
failed_sbals:
/* dequeue new FSF request previously enqueued */
zfcp_fsf_req_free(fsf_req);
fsf_req = NULL;
failed_fsf_req:
write_lock_irqsave(&req_queue->queue_lock, *lock_flags);
success:
*fsf_req_p = fsf_req;
return ret;
}
/*
* function: zfcp_fsf_req_send
*
* purpose: start transfer of FSF request via QDIO
*
* returns: 0 - request transfer succesfully started
* !0 - start of request transfer failed
*/
static int
zfcp_fsf_req_send(struct zfcp_fsf_req *fsf_req, struct timer_list *timer)
{
struct zfcp_adapter *adapter;
struct zfcp_qdio_queue *req_queue;
volatile struct qdio_buffer_element *sbale;
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
int inc_seq_no;
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 06:20:36 +08:00
int new_distance_from_int;
unsigned long flags;
int retval = 0;
adapter = fsf_req->adapter;
req_queue = &adapter->request_queue,
/* FIXME(debug): remove it later */
sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_first, 0);
ZFCP_LOG_DEBUG("SBALE0 flags=0x%x\n", sbale[0].flags);
ZFCP_LOG_TRACE("HEX DUMP OF SBALE1 PAYLOAD:\n");
ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_TRACE, (char *) sbale[1].addr,
sbale[1].length);
/* put allocated FSF request at list tail */
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
spin_lock_irqsave(&adapter->fsf_req_list_lock, 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 06:20:36 +08:00
list_add_tail(&fsf_req->list, &adapter->fsf_req_list_head);
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
spin_unlock_irqrestore(&adapter->fsf_req_list_lock, 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 06:20:36 +08:00
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
inc_seq_no = (fsf_req->qtcb != 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 06:20:36 +08:00
/* figure out expiration time of timeout and start timeout */
if (unlikely(timer)) {
timer->expires += jiffies;
add_timer(timer);
}
ZFCP_LOG_TRACE("request queue of adapter %s: "
"next free SBAL is %i, %i free SBALs\n",
zfcp_get_busid_by_adapter(adapter),
req_queue->free_index,
atomic_read(&req_queue->free_count));
ZFCP_LOG_DEBUG("calling do_QDIO adapter %s, flags=0x%x, queue_no=%i, "
"index_in_queue=%i, count=%i, buffers=%p\n",
zfcp_get_busid_by_adapter(adapter),
QDIO_FLAG_SYNC_OUTPUT,
0, fsf_req->sbal_first, fsf_req->sbal_number,
&req_queue->buffer[fsf_req->sbal_first]);
/*
* adjust the number of free SBALs in request queue as well as
* position of first one
*/
atomic_sub(fsf_req->sbal_number, &req_queue->free_count);
ZFCP_LOG_TRACE("free_count=%d\n", atomic_read(&req_queue->free_count));
req_queue->free_index += fsf_req->sbal_number; /* increase */
req_queue->free_index %= QDIO_MAX_BUFFERS_PER_Q; /* wrap if needed */
new_distance_from_int = zfcp_qdio_determine_pci(req_queue, fsf_req);
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
fsf_req->issued = get_clock();
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 06:20:36 +08:00
retval = do_QDIO(adapter->ccw_device,
QDIO_FLAG_SYNC_OUTPUT,
0, fsf_req->sbal_first, fsf_req->sbal_number, NULL);
if (unlikely(retval)) {
/* Queues are down..... */
retval = -EIO;
/*
* FIXME(potential race):
* timer might be expired (absolutely unlikely)
*/
if (timer)
del_timer(timer);
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
spin_lock_irqsave(&adapter->fsf_req_list_lock, 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 06:20:36 +08:00
list_del(&fsf_req->list);
[SCSI] zfcp: fix bug during adapter shutdown Fixes a race between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check. During adapter shutdown it occurred that a request was cleaned up twice. First during its normal completion. Second when dismiss_all was called. The fix is to serialize access to fsf request list between zfcp_fsf_req_dismiss_all and zfcp_qdio_reqid_check and delete a fsf request from the list if its completion is triggered. (Additionally a rwlock was replaced by a spinlock and fsf_req_cleanup was eliminated.) Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-06-13 19:20:35 +08:00
spin_unlock_irqrestore(&adapter->fsf_req_list_lock, 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 06:20:36 +08:00
/*
* adjust the number of free SBALs in request queue as well as
* position of first one
*/
zfcp_qdio_zero_sbals(req_queue->buffer,
fsf_req->sbal_first, fsf_req->sbal_number);
atomic_add(fsf_req->sbal_number, &req_queue->free_count);
req_queue->free_index -= fsf_req->sbal_number; /* increase */
req_queue->free_index += QDIO_MAX_BUFFERS_PER_Q;
req_queue->free_index %= QDIO_MAX_BUFFERS_PER_Q; /* wrap */
ZFCP_LOG_DEBUG
("error: do_QDIO failed. Buffers could not be enqueued "
"to request queue.\n");
} else {
req_queue->distance_from_int = new_distance_from_int;
/*
* increase FSF sequence counter -
* this must only be done for request successfully enqueued to
* QDIO this rejected requests may be cleaned up by calling
* routines resulting in missing sequence counter values
* otherwise,
*/
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
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 06:20:36 +08:00
/* Don't increase for unsolicited status */
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
if (inc_seq_no)
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 06:20:36 +08:00
adapter->fsf_req_seq_no++;
[SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann <aherrman@de.ibm.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2005-09-14 03:50:38 +08:00
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 06:20:36 +08:00
/* count FSF requests pending */
atomic_inc(&adapter->fsf_reqs_active);
}
return retval;
}
#undef ZFCP_LOG_AREA