Commit 4eff4a36 authored by Andreas Herrmann's avatar Andreas Herrmann Committed by James Bottomley

[SCSI] zfcp: fix: use correct req_id in eh_abort_handler

zfcp's eh_abort_handler used the wrong request ID to
identify the request to be aborted. The bug was introduced
with commit fea9d6c7
for improved management of request IDs. The bug is
fixed with this patch.
Signed-off-by: default avatarAndreas Herrmann <aherrman@de.ibm.com>
Signed-off-by: default avatarJames Bottomley <James.Bottomley@SteelEye.com>
parent dd52e0ea
......@@ -189,6 +189,10 @@ struct zfcp_fsf_req *zfcp_reqlist_ismember(struct zfcp_adapter *adapter,
struct zfcp_fsf_req *request, *tmp;
unsigned int i;
/* 0 is reserved as an invalid req_id */
if (req_id == 0)
return NULL;
i = req_id % REQUEST_LIST_SIZE;
list_for_each_entry_safe(request, tmp, &adapter->req_list[i], list)
......
......@@ -707,7 +707,7 @@ _zfcp_scsi_dbf_event_common(const char *tag, const char *tag2, int level,
struct zfcp_adapter *adapter,
struct scsi_cmnd *scsi_cmnd,
struct zfcp_fsf_req *fsf_req,
struct zfcp_fsf_req *old_fsf_req)
unsigned long old_req_id)
{
struct zfcp_scsi_dbf_record *rec = &adapter->scsi_dbf_buf;
struct zfcp_dbf_dump *dump = (struct zfcp_dbf_dump *)rec;
......@@ -768,8 +768,7 @@ _zfcp_scsi_dbf_event_common(const char *tag, const char *tag2, int level,
rec->fsf_seqno = fsf_req->seq_no;
rec->fsf_issued = fsf_req->issued;
}
rec->type.old_fsf_reqid =
(unsigned long) old_fsf_req;
rec->type.old_fsf_reqid = old_req_id;
} else {
strncpy(dump->tag, "dump", ZFCP_DBF_TAG_SIZE);
dump->total_size = buflen;
......@@ -794,17 +793,17 @@ zfcp_scsi_dbf_event_result(const char *tag, int level,
struct zfcp_fsf_req *fsf_req)
{
_zfcp_scsi_dbf_event_common("rslt", tag, level,
adapter, scsi_cmnd, fsf_req, NULL);
adapter, scsi_cmnd, fsf_req, 0);
}
inline void
zfcp_scsi_dbf_event_abort(const char *tag, struct zfcp_adapter *adapter,
struct scsi_cmnd *scsi_cmnd,
struct zfcp_fsf_req *new_fsf_req,
struct zfcp_fsf_req *old_fsf_req)
unsigned long old_req_id)
{
_zfcp_scsi_dbf_event_common("abrt", tag, 1,
adapter, scsi_cmnd, new_fsf_req, old_fsf_req);
adapter, scsi_cmnd, new_fsf_req, old_req_id);
}
inline void
......@@ -814,7 +813,7 @@ zfcp_scsi_dbf_event_devreset(const char *tag, u8 flag, struct zfcp_unit *unit,
struct zfcp_adapter *adapter = unit->port->adapter;
_zfcp_scsi_dbf_event_common(flag == FCP_TARGET_RESET ? "trst" : "lrst",
tag, 1, adapter, scsi_cmnd, NULL, NULL);
tag, 1, adapter, scsi_cmnd, NULL, 0);
}
static int
......
......@@ -185,7 +185,7 @@ extern void zfcp_scsi_dbf_event_result(const char *, int, struct zfcp_adapter *,
struct zfcp_fsf_req *);
extern void zfcp_scsi_dbf_event_abort(const char *, struct zfcp_adapter *,
struct scsi_cmnd *, struct zfcp_fsf_req *,
struct zfcp_fsf_req *);
unsigned long);
extern void zfcp_scsi_dbf_event_devreset(const char *, u8, struct zfcp_unit *,
struct scsi_cmnd *);
extern void zfcp_reqlist_add(struct zfcp_adapter *, struct zfcp_fsf_req *);
......
......@@ -3527,7 +3527,7 @@ zfcp_fsf_send_fcp_command_task(struct zfcp_adapter *adapter,
fsf_req->unit = unit;
/* associate FSF request with SCSI request (for look up on abort) */
scsi_cmnd->host_scribble = (char *) fsf_req;
scsi_cmnd->host_scribble = (unsigned char *) fsf_req->req_id;
/* associate SCSI command with FSF request */
fsf_req->data = (unsigned long) scsi_cmnd;
......@@ -4667,7 +4667,6 @@ zfcp_fsf_req_create(struct zfcp_adapter *adapter, u32 fsf_cmd, int req_flags,
{
volatile struct qdio_buffer_element *sbale;
struct zfcp_fsf_req *fsf_req = NULL;
unsigned long flags;
int ret = 0;
struct zfcp_qdio_queue *req_queue = &adapter->request_queue;
......@@ -4684,10 +4683,10 @@ zfcp_fsf_req_create(struct zfcp_adapter *adapter, u32 fsf_cmd, int req_flags,
fsf_req->fsf_command = fsf_cmd;
INIT_LIST_HEAD(&fsf_req->list);
/* unique request id */
spin_lock_irqsave(&adapter->req_list_lock, flags);
/* this is serialized (we are holding req_queue-lock of adapter */
if (adapter->req_no == 0)
adapter->req_no++;
fsf_req->req_id = adapter->req_no++;
spin_unlock_irqrestore(&adapter->req_list_lock, flags);
zfcp_fsf_req_qtcb_init(fsf_req);
......
......@@ -378,16 +378,15 @@ zfcp_unit_lookup(struct zfcp_adapter *adapter, int channel, unsigned int id,
* will handle late commands. (Usually, the normal completion of late
* commands is ignored with respect to the running abort operation.)
*/
int
zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt)
int zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt)
{
struct Scsi_Host *scsi_host;
struct zfcp_adapter *adapter;
struct zfcp_unit *unit;
int retval = SUCCESS;
struct zfcp_fsf_req *new_fsf_req = NULL;
struct zfcp_fsf_req *old_fsf_req;
struct zfcp_fsf_req *fsf_req;
unsigned long flags;
unsigned long old_req_id;
int retval = SUCCESS;
scsi_host = scpnt->device->host;
adapter = (struct zfcp_adapter *) scsi_host->hostdata[0];
......@@ -399,55 +398,47 @@ zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt)
/* avoid race condition between late normal completion and abort */
write_lock_irqsave(&adapter->abort_lock, flags);
/*
* Check whether command has just completed and can not be aborted.
* Even if the command has just been completed late, we can access
* scpnt since the SCSI stack does not release it at least until
* this routine returns. (scpnt is parameter passed to this routine
* and must not disappear during abort even on late completion.)
*/
old_fsf_req = (struct zfcp_fsf_req *) scpnt->host_scribble;
if (!old_fsf_req) {
/* Check whether corresponding fsf_req is still pending */
spin_lock(&adapter->req_list_lock);
fsf_req = zfcp_reqlist_ismember(adapter, (unsigned long)
scpnt->host_scribble);
spin_unlock(&adapter->req_list_lock);
if (!fsf_req) {
write_unlock_irqrestore(&adapter->abort_lock, flags);
zfcp_scsi_dbf_event_abort("lte1", adapter, scpnt, NULL, NULL);
zfcp_scsi_dbf_event_abort("lte1", adapter, scpnt, NULL, 0);
retval = SUCCESS;
goto out;
}
old_fsf_req->data = 0;
old_fsf_req->status |= ZFCP_STATUS_FSFREQ_ABORTING;
fsf_req->data = 0;
fsf_req->status |= ZFCP_STATUS_FSFREQ_ABORTING;
old_req_id = fsf_req->req_id;
/* don't access old_fsf_req after releasing the abort_lock */
/* don't access old fsf_req after releasing the abort_lock */
write_unlock_irqrestore(&adapter->abort_lock, flags);
/* call FSF routine which does the abort */
new_fsf_req = zfcp_fsf_abort_fcp_command((unsigned long) old_fsf_req,
adapter, unit, 0);
if (!new_fsf_req) {
fsf_req = zfcp_fsf_abort_fcp_command(old_req_id, adapter, unit, 0);
if (!fsf_req) {
ZFCP_LOG_INFO("error: initiation of Abort FCP Cmnd failed\n");
zfcp_scsi_dbf_event_abort("nres", adapter, scpnt, NULL,
old_fsf_req);
old_req_id);
retval = FAILED;
goto out;
}
/* wait for completion of abort */
__wait_event(new_fsf_req->completion_wq,
new_fsf_req->status & ZFCP_STATUS_FSFREQ_COMPLETED);
__wait_event(fsf_req->completion_wq,
fsf_req->status & ZFCP_STATUS_FSFREQ_COMPLETED);
/* status should be valid since signals were not permitted */
if (new_fsf_req->status & ZFCP_STATUS_FSFREQ_ABORTSUCCEEDED) {
zfcp_scsi_dbf_event_abort("okay", adapter, scpnt, new_fsf_req,
NULL);
if (fsf_req->status & ZFCP_STATUS_FSFREQ_ABORTSUCCEEDED) {
zfcp_scsi_dbf_event_abort("okay", adapter, scpnt, fsf_req, 0);
retval = SUCCESS;
} else if (new_fsf_req->status & ZFCP_STATUS_FSFREQ_ABORTNOTNEEDED) {
zfcp_scsi_dbf_event_abort("lte2", adapter, scpnt, new_fsf_req,
NULL);
} else if (fsf_req->status & ZFCP_STATUS_FSFREQ_ABORTNOTNEEDED) {
zfcp_scsi_dbf_event_abort("lte2", adapter, scpnt, fsf_req, 0);
retval = SUCCESS;
} else {
zfcp_scsi_dbf_event_abort("fail", adapter, scpnt, new_fsf_req,
NULL);
zfcp_scsi_dbf_event_abort("fail", adapter, scpnt, fsf_req, 0);
retval = FAILED;
}
zfcp_fsf_req_free(new_fsf_req);
zfcp_fsf_req_free(fsf_req);
out:
return retval;
}
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment