* [PATCH-v1 00/22] lpfc updates for 11.2.0.12
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
This patch set provides a number of bug fixes, code cleanups, and error
handling, mostly in the nvme area of lpfc.
There is one new feature in the series to add Fabric Assigned WWN support.
The patches were cut against the linux-block tree, for-4.12/block branch,
and are intended to be pulled in through that tree rather than the scsi trees.
The patches are dependent on the FC nvme/nvmet patches from the following 2
series:
http://lists.infradead.org/pipermail/linux-nvme/2017-April/009250.html
http://lists.infradead.org/pipermail/linux-nvme/2017-April/009256.html
James Smart (22):
Standardize nvme SGL segment count
Fix nvme unregister port timeout.
Fix rejected nvme LS Req.
Fix log message in completion path.
Add debug messages for nvme/fcp resource allocation.
Fix spelling in comments.
Remove unused defines for NVME PostBuf.
Remove NULL ptr check before kfree.
Fix extra line print in rqpair debug print.
Fix PRLI ACC rsp for NVME
Fix driver unload/reload operation.
Fix driver usage of 128B WQEs when WQ_CREATE is V1.
Fix nvme initiator handling when not enabled.
Remove hba lock from NVMET issue WQE.
Fix driver load issues when MRQ=8
Fix crash after issuing lip reset
Fix max_sgl_segments settings for NVME / NVMET
Add Fabric assigned WWN support.
Fix implicit logo and RSCN handling for NVMET
Update ABORT processing for NVMET.
Fix Express lane queue creation.
lpfc revison 11.2.0.12
drivers/scsi/lpfc/lpfc.h | 5 +-
drivers/scsi/lpfc/lpfc_attr.c | 10 +-
drivers/scsi/lpfc/lpfc_bsg.c | 4 +
drivers/scsi/lpfc/lpfc_crtn.h | 9 +-
drivers/scsi/lpfc/lpfc_ct.c | 68 +++++--
drivers/scsi/lpfc/lpfc_debugfs.c | 69 ++++---
drivers/scsi/lpfc/lpfc_disc.h | 1 +
drivers/scsi/lpfc/lpfc_els.c | 68 +++----
drivers/scsi/lpfc/lpfc_hbadisc.c | 131 +++++++++----
drivers/scsi/lpfc/lpfc_hw.h | 3 +
drivers/scsi/lpfc/lpfc_hw4.h | 4 +
drivers/scsi/lpfc/lpfc_init.c | 202 +++++++++++++-------
drivers/scsi/lpfc/lpfc_mbox.c | 7 +-
drivers/scsi/lpfc/lpfc_nportdisc.c | 8 +-
drivers/scsi/lpfc/lpfc_nvme.c | 139 ++++++++++----
drivers/scsi/lpfc/lpfc_nvme.h | 7 +-
drivers/scsi/lpfc/lpfc_nvmet.c | 378 ++++++++++++++++++++++++++++---------
drivers/scsi/lpfc/lpfc_nvmet.h | 14 +-
drivers/scsi/lpfc/lpfc_sli.c | 22 ++-
drivers/scsi/lpfc/lpfc_sli4.h | 2 +-
drivers/scsi/lpfc/lpfc_version.h | 2 +-
drivers/scsi/lpfc/lpfc_vport.c | 3 +-
22 files changed, 814 insertions(+), 342 deletions(-)
--
2.1.0
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 01/22] Standardize nvme SGL segment count
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
Standardize default SGL segment count for nvme target and initiator
The driver needs to make them the same for clarity.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvme.h | 4 +---
drivers/scsi/lpfc/lpfc_nvmet.c | 2 +-
drivers/scsi/lpfc/lpfc_nvmet.h | 4 +---
3 files changed, 3 insertions(+), 7 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvme.h b/drivers/scsi/lpfc/lpfc_nvme.h
index 1347deb..6277796 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.h
+++ b/drivers/scsi/lpfc/lpfc_nvme.h
@@ -21,9 +21,7 @@
* included with this package. *
********************************************************************/
-#define LPFC_NVME_MIN_SEGS 16
-#define LPFC_NVME_DEFAULT_SEGS 66 /* 256K IOs - 64 + 2 */
-#define LPFC_NVME_MAX_SEGS 510
+#define LPFC_NVME_DEFAULT_SEGS (64 + 1) /* 256K IOs */
#define LPFC_NVMET_MIN_POSTBUF 16
#define LPFC_NVMET_DEFAULT_POSTBUF 1024
#define LPFC_NVMET_MAX_POSTBUF 4096
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index d488c33..c8a4941 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -709,7 +709,7 @@ lpfc_nvmet_create_targetport(struct lpfc_hba *phba)
pinfo.port_id = vport->fc_myDID;
lpfc_tgttemplate.max_hw_queues = phba->cfg_nvme_io_channel;
- lpfc_tgttemplate.max_sgl_segments = phba->cfg_sg_seg_cnt;
+ lpfc_tgttemplate.max_sgl_segments = phba->cfg_sg_seg_cnt + 1;
lpfc_tgttemplate.target_features = NVMET_FCTGTFEAT_READDATA_RSP |
NVMET_FCTGTFEAT_NEEDS_CMD_CPUSCHED |
NVMET_FCTGTFEAT_CMD_IN_ISR |
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.h b/drivers/scsi/lpfc/lpfc_nvmet.h
index 02735fc..d8bac4c 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.h
+++ b/drivers/scsi/lpfc/lpfc_nvmet.h
@@ -21,9 +21,7 @@
* included with this package. *
********************************************************************/
-#define LPFC_NVMET_MIN_SEGS 16
-#define LPFC_NVMET_DEFAULT_SEGS 64 /* 256K IOs */
-#define LPFC_NVMET_MAX_SEGS 510
+#define LPFC_NVMET_DEFAULT_SEGS (64 + 1) /* 256K IOs */
#define LPFC_NVMET_SUCCESS_LEN 12
/* Used for NVME Target */
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH-v1 01/22] Standardize nvme SGL segment count
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
Standardize default SGL segment count for nvme target and initiator
The driver needs to make them the same for clarity.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvme.h | 4 +---
drivers/scsi/lpfc/lpfc_nvmet.c | 2 +-
drivers/scsi/lpfc/lpfc_nvmet.h | 4 +---
3 files changed, 3 insertions(+), 7 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvme.h b/drivers/scsi/lpfc/lpfc_nvme.h
index 1347deb..6277796 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.h
+++ b/drivers/scsi/lpfc/lpfc_nvme.h
@@ -21,9 +21,7 @@
* included with this package. *
********************************************************************/
-#define LPFC_NVME_MIN_SEGS 16
-#define LPFC_NVME_DEFAULT_SEGS 66 /* 256K IOs - 64 + 2 */
-#define LPFC_NVME_MAX_SEGS 510
+#define LPFC_NVME_DEFAULT_SEGS (64 + 1) /* 256K IOs */
#define LPFC_NVMET_MIN_POSTBUF 16
#define LPFC_NVMET_DEFAULT_POSTBUF 1024
#define LPFC_NVMET_MAX_POSTBUF 4096
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index d488c33..c8a4941 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -709,7 +709,7 @@ lpfc_nvmet_create_targetport(struct lpfc_hba *phba)
pinfo.port_id = vport->fc_myDID;
lpfc_tgttemplate.max_hw_queues = phba->cfg_nvme_io_channel;
- lpfc_tgttemplate.max_sgl_segments = phba->cfg_sg_seg_cnt;
+ lpfc_tgttemplate.max_sgl_segments = phba->cfg_sg_seg_cnt + 1;
lpfc_tgttemplate.target_features = NVMET_FCTGTFEAT_READDATA_RSP |
NVMET_FCTGTFEAT_NEEDS_CMD_CPUSCHED |
NVMET_FCTGTFEAT_CMD_IN_ISR |
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.h b/drivers/scsi/lpfc/lpfc_nvmet.h
index 02735fc..d8bac4c 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.h
+++ b/drivers/scsi/lpfc/lpfc_nvmet.h
@@ -21,9 +21,7 @@
* included with this package. *
********************************************************************/
-#define LPFC_NVMET_MIN_SEGS 16
-#define LPFC_NVMET_DEFAULT_SEGS 64 /* 256K IOs */
-#define LPFC_NVMET_MAX_SEGS 510
+#define LPFC_NVMET_DEFAULT_SEGS (64 + 1) /* 256K IOs */
#define LPFC_NVMET_SUCCESS_LEN 12
/* Used for NVME Target */
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 01/22] Standardize nvme SGL segment count
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:15 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:15 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:20PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> Standardize default SGL segment count for nvme target and initiator
>
> The driver needs to make them the same for clarity.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 01/22] Standardize nvme SGL segment count
@ 2017-04-20 7:15 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:15 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:20PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> Standardize default SGL segment count for nvme target and initiator
>
> The driver needs to make them the same for clarity.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 02/22] Fix nvme unregister port timeout.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
During some link event testing it was observed that the
wait_for_completion_timeout in the lpfc_nvme_unregister_port
was timing out all the time.
The initiator is claiming the nvme_fc_unregister_remoteport upcall is
not completing the unregister in the time allotted.
[ 2186.151317] lpfc 0000:07:00.0: 0:(0):6169 Unreg nvme wait failed 0
The wait_for_completion_timeout returns 0 when the wait has
been outstanding for the jiffies passed by the caller. In this error
message, the nvme initiator passed value 5 - meaning 5 jiffies -
and this is just wrong.
Calculate 5 seconds in Jiffies and pass that value
from the current jiffies.
Also the log message for the unregister timeout was reduced
because timeout failure is the same as timeout.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvme.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index 0024de1..a39d72c 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -2409,6 +2409,7 @@ lpfc_nvme_unregister_port(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp)
struct lpfc_nvme_lport *lport;
struct lpfc_nvme_rport *rport;
struct nvme_fc_remote_port *remoteport;
+ unsigned long wait_tmo;
localport = vport->localport;
@@ -2451,11 +2452,12 @@ lpfc_nvme_unregister_port(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp)
* before proceeding. This guarantees the transport and driver
* have completed the unreg process.
*/
- ret = wait_for_completion_timeout(&rport->rport_unreg_done, 5);
+ wait_tmo = msecs_to_jiffies(5000);
+ ret = wait_for_completion_timeout(&rport->rport_unreg_done,
+ wait_tmo);
if (ret == 0) {
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_DISC,
- "6169 Unreg nvme wait failed %d\n",
- ret);
+ "6169 Unreg nvme wait timeout\n");
}
}
return;
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 02/22] Fix nvme unregister port timeout.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
During some link event testing it was observed that the
wait_for_completion_timeout in the lpfc_nvme_unregister_port
was timing out all the time.
The initiator is claiming the nvme_fc_unregister_remoteport upcall is
not completing the unregister in the time allotted.
[ 2186.151317] lpfc 0000:07:00.0: 0:(0):6169 Unreg nvme wait failed 0
The wait_for_completion_timeout returns 0 when the wait has
been outstanding for the jiffies passed by the caller. In this error
message, the nvme initiator passed value 5 - meaning 5 jiffies -
and this is just wrong.
Calculate 5 seconds in Jiffies and pass that value
from the current jiffies.
Also the log message for the unregister timeout was reduced
because timeout failure is the same as timeout.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvme.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index 0024de1..a39d72c 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -2409,6 +2409,7 @@ lpfc_nvme_unregister_port(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp)
struct lpfc_nvme_lport *lport;
struct lpfc_nvme_rport *rport;
struct nvme_fc_remote_port *remoteport;
+ unsigned long wait_tmo;
localport = vport->localport;
@@ -2451,11 +2452,12 @@ lpfc_nvme_unregister_port(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp)
* before proceeding. This guarantees the transport and driver
* have completed the unreg process.
*/
- ret = wait_for_completion_timeout(&rport->rport_unreg_done, 5);
+ wait_tmo = msecs_to_jiffies(5000);
+ ret = wait_for_completion_timeout(&rport->rport_unreg_done,
+ wait_tmo);
if (ret == 0) {
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_DISC,
- "6169 Unreg nvme wait failed %d\n",
- ret);
+ "6169 Unreg nvme wait timeout\n");
}
}
return;
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 02/22] Fix nvme unregister port timeout.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:15 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:15 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:21PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> During some link event testing it was observed that the
> wait_for_completion_timeout in the lpfc_nvme_unregister_port
> was timing out all the time.
>
> The initiator is claiming the nvme_fc_unregister_remoteport upcall is
> not completing the unregister in the time allotted.
> [ 2186.151317] lpfc 0000:07:00.0: 0:(0):6169 Unreg nvme wait failed 0
>
> The wait_for_completion_timeout returns 0 when the wait has
> been outstanding for the jiffies passed by the caller. In this error
> message, the nvme initiator passed value 5 - meaning 5 jiffies -
> and this is just wrong.
>
> Calculate 5 seconds in Jiffies and pass that value
> from the current jiffies.
>
> Also the log message for the unregister timeout was reduced
> because timeout failure is the same as timeout.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 02/22] Fix nvme unregister port timeout.
@ 2017-04-20 7:15 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:15 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:21PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> During some link event testing it was observed that the
> wait_for_completion_timeout in the lpfc_nvme_unregister_port
> was timing out all the time.
>
> The initiator is claiming the nvme_fc_unregister_remoteport upcall is
> not completing the unregister in the time allotted.
> [ 2186.151317] lpfc 0000:07:00.0: 0:(0):6169 Unreg nvme wait failed 0
>
> The wait_for_completion_timeout returns 0 when the wait has
> been outstanding for the jiffies passed by the caller. In this error
> message, the nvme initiator passed value 5 - meaning 5 jiffies -
> and this is just wrong.
>
> Calculate 5 seconds in Jiffies and pass that value
> from the current jiffies.
>
> Also the log message for the unregister timeout was reduced
> because timeout failure is the same as timeout.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 03/22] Fix rejected nvme LS Req.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
In this case, the NVME initiator is sending an LS REQ command on an NDLP
that is not MAPPED. The FW rejects it.
The lpfc_nvme_ls_req routine checks for a NULL ndlp pointer
but does not check the NDLP state. This allows the routine
to send an LS IO when the ndlp is disconnected.
Check the ndlp for NULL, actual node, Target and MAPPED
or Initiator and UNMAPPED. This avoids Fabric nodes getting
the Create Association or Create Connection commands. Initiators
are free to Reject either Create.
Also some of the messages numbers in lpfc_nvme_ls_req were changed because
they were already used in other log messages.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvme.c | 25 ++++++++++++++++++++-----
1 file changed, 20 insertions(+), 5 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index a39d72c..2a87428 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -417,11 +417,26 @@ lpfc_nvme_ls_req(struct nvme_fc_local_port *pnvme_lport,
vport = lport->vport;
ndlp = lpfc_findnode_did(vport, pnvme_rport->port_id);
- if (!ndlp) {
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_DISC,
- "6043 Could not find node for DID %x\n",
+ if (!ndlp || !NLP_CHK_NODE_ACT(ndlp)) {
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_IOERR,
+ "6051 DID x%06x not an active rport.\n",
pnvme_rport->port_id);
- return 1;
+ return -ENODEV;
+ }
+
+ /* The remote node has to be a mapped nvme target or an
+ * unmapped nvme initiator or it's an error.
+ */
+ if (((ndlp->nlp_type & NLP_NVME_TARGET) &&
+ (ndlp->nlp_state != NLP_STE_MAPPED_NODE)) ||
+ ((ndlp->nlp_type & NLP_NVME_INITIATOR) &&
+ (ndlp->nlp_state != NLP_STE_UNMAPPED_NODE))) {
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_IOERR,
+ "6088 DID x%06x not ready for "
+ "IO. State x%x, Type x%x\n",
+ pnvme_rport->port_id,
+ ndlp->nlp_state, ndlp->nlp_type);
+ return -ENODEV;
}
bmp = kmalloc(sizeof(struct lpfc_dmabuf), GFP_KERNEL);
if (!bmp) {
@@ -456,7 +471,7 @@ lpfc_nvme_ls_req(struct nvme_fc_local_port *pnvme_lport,
/* Expand print to include key fields. */
lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME_DISC,
- "6051 ENTER. lport %p, rport %p lsreq%p rqstlen:%d "
+ "6149 ENTER. lport %p, rport %p lsreq%p rqstlen:%d "
"rsplen:%d %pad %pad\n",
pnvme_lport, pnvme_rport,
pnvme_lsreq, pnvme_lsreq->rqstlen,
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 03/22] Fix rejected nvme LS Req.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
In this case, the NVME initiator is sending an LS REQ command on an NDLP
that is not MAPPED. The FW rejects it.
The lpfc_nvme_ls_req routine checks for a NULL ndlp pointer
but does not check the NDLP state. This allows the routine
to send an LS IO when the ndlp is disconnected.
Check the ndlp for NULL, actual node, Target and MAPPED
or Initiator and UNMAPPED. This avoids Fabric nodes getting
the Create Association or Create Connection commands. Initiators
are free to Reject either Create.
Also some of the messages numbers in lpfc_nvme_ls_req were changed because
they were already used in other log messages.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvme.c | 25 ++++++++++++++++++++-----
1 file changed, 20 insertions(+), 5 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index a39d72c..2a87428 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -417,11 +417,26 @@ lpfc_nvme_ls_req(struct nvme_fc_local_port *pnvme_lport,
vport = lport->vport;
ndlp = lpfc_findnode_did(vport, pnvme_rport->port_id);
- if (!ndlp) {
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_DISC,
- "6043 Could not find node for DID %x\n",
+ if (!ndlp || !NLP_CHK_NODE_ACT(ndlp)) {
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_IOERR,
+ "6051 DID x%06x not an active rport.\n",
pnvme_rport->port_id);
- return 1;
+ return -ENODEV;
+ }
+
+ /* The remote node has to be a mapped nvme target or an
+ * unmapped nvme initiator or it's an error.
+ */
+ if (((ndlp->nlp_type & NLP_NVME_TARGET) &&
+ (ndlp->nlp_state != NLP_STE_MAPPED_NODE)) ||
+ ((ndlp->nlp_type & NLP_NVME_INITIATOR) &&
+ (ndlp->nlp_state != NLP_STE_UNMAPPED_NODE))) {
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_IOERR,
+ "6088 DID x%06x not ready for "
+ "IO. State x%x, Type x%x\n",
+ pnvme_rport->port_id,
+ ndlp->nlp_state, ndlp->nlp_type);
+ return -ENODEV;
}
bmp = kmalloc(sizeof(struct lpfc_dmabuf), GFP_KERNEL);
if (!bmp) {
@@ -456,7 +471,7 @@ lpfc_nvme_ls_req(struct nvme_fc_local_port *pnvme_lport,
/* Expand print to include key fields. */
lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME_DISC,
- "6051 ENTER. lport %p, rport %p lsreq%p rqstlen:%d "
+ "6149 ENTER. lport %p, rport %p lsreq%p rqstlen:%d "
"rsplen:%d %pad %pad\n",
pnvme_lport, pnvme_rport,
pnvme_lsreq, pnvme_lsreq->rqstlen,
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 03/22] Fix rejected nvme LS Req.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:29 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:29 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:22PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> In this case, the NVME initiator is sending an LS REQ command on an NDLP
> that is not MAPPED. The FW rejects it.
>
> The lpfc_nvme_ls_req routine checks for a NULL ndlp pointer
> but does not check the NDLP state. This allows the routine
> to send an LS IO when the ndlp is disconnected.
>
> Check the ndlp for NULL, actual node, Target and MAPPED
> or Initiator and UNMAPPED. This avoids Fabric nodes getting
> the Create Association or Create Connection commands. Initiators
> are free to Reject either Create.
> Also some of the messages numbers in lpfc_nvme_ls_req were changed because
> they were already used in other log messages.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
> drivers/scsi/lpfc/lpfc_nvme.c | 25 ++++++++++++++++++++-----
> 1 file changed, 20 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
> index a39d72c..2a87428 100644
> --- a/drivers/scsi/lpfc/lpfc_nvme.c
> +++ b/drivers/scsi/lpfc/lpfc_nvme.c
> @@ -417,11 +417,26 @@ lpfc_nvme_ls_req(struct nvme_fc_local_port *pnvme_lport,
> vport = lport->vport;
>
> ndlp = lpfc_findnode_did(vport, pnvme_rport->port_id);
> - if (!ndlp) {
> - lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_DISC,
> - "6043 Could not find node for DID %x\n",
> + if (!ndlp || !NLP_CHK_NODE_ACT(ndlp)) {
> + lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_IOERR,
> + "6051 DID x%06x not an active rport.\n",
> pnvme_rport->port_id);
> - return 1;
> + return -ENODEV;
> + }
> +
> + /* The remote node has to be a mapped nvme target or an
> + * unmapped nvme initiator or it's an error.
> + */
> + if (((ndlp->nlp_type & NLP_NVME_TARGET) &&
> + (ndlp->nlp_state != NLP_STE_MAPPED_NODE)) ||
> + ((ndlp->nlp_type & NLP_NVME_INITIATOR) &&
> + (ndlp->nlp_state != NLP_STE_UNMAPPED_NODE))) {
> + lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_IOERR,
> + "6088 DID x%06x not ready for "
> + "IO. State x%x, Type x%x\n",
> + pnvme_rport->port_id,
> + ndlp->nlp_state, ndlp->nlp_type);
> + return -ENODEV;
This if statement is horrible to read with all those parenthesis. It really
looks more like LISP than C. IIRC I already told you this in my last review.
With this fixed up,
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread* [PATCH-v1 03/22] Fix rejected nvme LS Req.
@ 2017-04-20 7:29 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:29 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:22PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> In this case, the NVME initiator is sending an LS REQ command on an NDLP
> that is not MAPPED. The FW rejects it.
>
> The lpfc_nvme_ls_req routine checks for a NULL ndlp pointer
> but does not check the NDLP state. This allows the routine
> to send an LS IO when the ndlp is disconnected.
>
> Check the ndlp for NULL, actual node, Target and MAPPED
> or Initiator and UNMAPPED. This avoids Fabric nodes getting
> the Create Association or Create Connection commands. Initiators
> are free to Reject either Create.
> Also some of the messages numbers in lpfc_nvme_ls_req were changed because
> they were already used in other log messages.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
> drivers/scsi/lpfc/lpfc_nvme.c | 25 ++++++++++++++++++++-----
> 1 file changed, 20 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
> index a39d72c..2a87428 100644
> --- a/drivers/scsi/lpfc/lpfc_nvme.c
> +++ b/drivers/scsi/lpfc/lpfc_nvme.c
> @@ -417,11 +417,26 @@ lpfc_nvme_ls_req(struct nvme_fc_local_port *pnvme_lport,
> vport = lport->vport;
>
> ndlp = lpfc_findnode_did(vport, pnvme_rport->port_id);
> - if (!ndlp) {
> - lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_DISC,
> - "6043 Could not find node for DID %x\n",
> + if (!ndlp || !NLP_CHK_NODE_ACT(ndlp)) {
> + lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_IOERR,
> + "6051 DID x%06x not an active rport.\n",
> pnvme_rport->port_id);
> - return 1;
> + return -ENODEV;
> + }
> +
> + /* The remote node has to be a mapped nvme target or an
> + * unmapped nvme initiator or it's an error.
> + */
> + if (((ndlp->nlp_type & NLP_NVME_TARGET) &&
> + (ndlp->nlp_state != NLP_STE_MAPPED_NODE)) ||
> + ((ndlp->nlp_type & NLP_NVME_INITIATOR) &&
> + (ndlp->nlp_state != NLP_STE_UNMAPPED_NODE))) {
> + lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_IOERR,
> + "6088 DID x%06x not ready for "
> + "IO. State x%x, Type x%x\n",
> + pnvme_rport->port_id,
> + ndlp->nlp_state, ndlp->nlp_type);
> + return -ENODEV;
This if statement is horrible to read with all those parenthesis. It really
looks more like LISP than C. IIRC I already told you this in my last review.
With this fixed up,
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 04/22] Fix log message in completion path.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
In the lpfc_nvme_io_cmd_wqe_cmpl routine the driver was printing two
pointers and the DID for the rport whenever an IO completed on a now
that had transitioned to a non active state.
There is no need to print the node pointer address for a node that
is not active the DID should be enough to debug.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvme.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index 2a87428..bc6946f 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -787,9 +787,8 @@ lpfc_nvme_io_cmd_wqe_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pwqeIn,
ndlp = rport->ndlp;
if (!ndlp || !NLP_CHK_NODE_ACT(ndlp)) {
lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_IOERR,
- "6061 rport %p, ndlp %p, DID x%06x ndlp "
- "not ready.\n",
- rport, ndlp, rport->remoteport->port_id);
+ "6061 rport %p, DID x%06x node not ready.\n",
+ rport, rport->remoteport->port_id);
ndlp = lpfc_findnode_did(vport, rport->remoteport->port_id);
if (!ndlp) {
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 04/22] Fix log message in completion path.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
In the lpfc_nvme_io_cmd_wqe_cmpl routine the driver was printing two
pointers and the DID for the rport whenever an IO completed on a now
that had transitioned to a non active state.
There is no need to print the node pointer address for a node that
is not active the DID should be enough to debug.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvme.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index 2a87428..bc6946f 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -787,9 +787,8 @@ lpfc_nvme_io_cmd_wqe_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pwqeIn,
ndlp = rport->ndlp;
if (!ndlp || !NLP_CHK_NODE_ACT(ndlp)) {
lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_IOERR,
- "6061 rport %p, ndlp %p, DID x%06x ndlp "
- "not ready.\n",
- rport, ndlp, rport->remoteport->port_id);
+ "6061 rport %p, DID x%06x node not ready.\n",
+ rport, rport->remoteport->port_id);
ndlp = lpfc_findnode_did(vport, rport->remoteport->port_id);
if (!ndlp) {
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 04/22] Fix log message in completion path.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:32 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:32 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:23PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> In the lpfc_nvme_io_cmd_wqe_cmpl routine the driver was printing two
> pointers and the DID for the rport whenever an IO completed on a now
> that had transitioned to a non active state.
>
> There is no need to print the node pointer address for a node that
> is not active the DID should be enough to debug.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 04/22] Fix log message in completion path.
@ 2017-04-20 7:32 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:32 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:23PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> In the lpfc_nvme_io_cmd_wqe_cmpl routine the driver was printing two
> pointers and the DID for the rport whenever an IO completed on a now
> that had transitioned to a non active state.
>
> There is no need to print the node pointer address for a node that
> is not active the DID should be enough to debug.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 05/22] Add debug messages for nvme/fcp resource allocation.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
The xri resources are split into pools for NVME and FCP IO when NVME is
enabled. There was not message in the log that identified this allocation.
Added debug message to log XRI split.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_init.c | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 6cc561b..b56da01 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -3508,6 +3508,12 @@ lpfc_sli4_scsi_sgl_update(struct lpfc_hba *phba)
spin_unlock(&phba->scsi_buf_list_put_lock);
spin_unlock_irq(&phba->scsi_buf_list_get_lock);
+ lpfc_printf_log(phba, KERN_INFO, LOG_SLI,
+ "6060 Current allocated SCSI xri-sgl count:%d, "
+ "maximum SCSI xri count:%d (split:%d)\n",
+ phba->sli4_hba.scsi_xri_cnt,
+ phba->sli4_hba.scsi_xri_max, phba->cfg_xri_split);
+
if (phba->sli4_hba.scsi_xri_cnt > phba->sli4_hba.scsi_xri_max) {
/* max scsi xri shrinked below the allocated scsi buffers */
scsi_xri_cnt = phba->sli4_hba.scsi_xri_cnt -
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 05/22] Add debug messages for nvme/fcp resource allocation.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
The xri resources are split into pools for NVME and FCP IO when NVME is
enabled. There was not message in the log that identified this allocation.
Added debug message to log XRI split.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_init.c | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 6cc561b..b56da01 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -3508,6 +3508,12 @@ lpfc_sli4_scsi_sgl_update(struct lpfc_hba *phba)
spin_unlock(&phba->scsi_buf_list_put_lock);
spin_unlock_irq(&phba->scsi_buf_list_get_lock);
+ lpfc_printf_log(phba, KERN_INFO, LOG_SLI,
+ "6060 Current allocated SCSI xri-sgl count:%d, "
+ "maximum SCSI xri count:%d (split:%d)\n",
+ phba->sli4_hba.scsi_xri_cnt,
+ phba->sli4_hba.scsi_xri_max, phba->cfg_xri_split);
+
if (phba->sli4_hba.scsi_xri_cnt > phba->sli4_hba.scsi_xri_max) {
/* max scsi xri shrinked below the allocated scsi buffers */
scsi_xri_cnt = phba->sli4_hba.scsi_xri_cnt -
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 05/22] Add debug messages for nvme/fcp resource allocation.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:33 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:33 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:24PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> The xri resources are split into pools for NVME and FCP IO when NVME is
> enabled. There was not message in the log that identified this allocation.
>
> Added debug message to log XRI split.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 05/22] Add debug messages for nvme/fcp resource allocation.
@ 2017-04-20 7:33 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:33 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:24PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> The xri resources are split into pools for NVME and FCP IO when NVME is
> enabled. There was not message in the log that identified this allocation.
>
> Added debug message to log XRI split.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 06/22] Fix spelling in comments.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
Comment should have said Repost.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_sli.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index 1c9fa45..7087c55 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -6338,7 +6338,7 @@ lpfc_sli4_get_allocated_extnts(struct lpfc_hba *phba, uint16_t type,
}
/**
- * lpfc_sli4_repost_sgl_list - Repsot the buffers sgl pages as block
+ * lpfc_sli4_repost_sgl_list - Repost the buffers sgl pages as block
* @phba: pointer to lpfc hba data structure.
* @pring: Pointer to driver SLI ring object.
* @sgl_list: linked link of sgl buffers to post
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH-v1 06/22] Fix spelling in comments.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
Comment should have said Repost.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_sli.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index 1c9fa45..7087c55 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -6338,7 +6338,7 @@ lpfc_sli4_get_allocated_extnts(struct lpfc_hba *phba, uint16_t type,
}
/**
- * lpfc_sli4_repost_sgl_list - Repsot the buffers sgl pages as block
+ * lpfc_sli4_repost_sgl_list - Repost the buffers sgl pages as block
* @phba: pointer to lpfc hba data structure.
* @pring: Pointer to driver SLI ring object.
* @sgl_list: linked link of sgl buffers to post
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 06/22] Fix spelling in comments.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:34 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:34 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:25PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> Comment should have said Repost.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 06/22] Fix spelling in comments.
@ 2017-04-20 7:34 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:34 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:25PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> Comment should have said Repost.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 07/22] Remove unused defines for NVME PostBuf.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
These defines for the posting of buffers for nvmet target were not used.
Removing the unused defines.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvme.h | 3 ---
1 file changed, 3 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvme.h b/drivers/scsi/lpfc/lpfc_nvme.h
index 6277796..2582f46 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.h
+++ b/drivers/scsi/lpfc/lpfc_nvme.h
@@ -22,9 +22,6 @@
********************************************************************/
#define LPFC_NVME_DEFAULT_SEGS (64 + 1) /* 256K IOs */
-#define LPFC_NVMET_MIN_POSTBUF 16
-#define LPFC_NVMET_DEFAULT_POSTBUF 1024
-#define LPFC_NVMET_MAX_POSTBUF 4096
#define LPFC_NVME_WQSIZE 256
#define LPFC_NVME_ERSP_LEN 0x20
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH-v1 07/22] Remove unused defines for NVME PostBuf.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
These defines for the posting of buffers for nvmet target were not used.
Removing the unused defines.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvme.h | 3 ---
1 file changed, 3 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvme.h b/drivers/scsi/lpfc/lpfc_nvme.h
index 6277796..2582f46 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.h
+++ b/drivers/scsi/lpfc/lpfc_nvme.h
@@ -22,9 +22,6 @@
********************************************************************/
#define LPFC_NVME_DEFAULT_SEGS (64 + 1) /* 256K IOs */
-#define LPFC_NVMET_MIN_POSTBUF 16
-#define LPFC_NVMET_DEFAULT_POSTBUF 1024
-#define LPFC_NVMET_MAX_POSTBUF 4096
#define LPFC_NVME_WQSIZE 256
#define LPFC_NVME_ERSP_LEN 0x20
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 07/22] Remove unused defines for NVME PostBuf.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:34 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:34 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:26PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> These defines for the posting of buffers for nvmet target were not used.
>
> Removing the unused defines.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 07/22] Remove unused defines for NVME PostBuf.
@ 2017-04-20 7:34 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:34 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:26PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> These defines for the posting of buffers for nvmet target were not used.
>
> Removing the unused defines.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 08/22] Remove NULL ptr check before kfree.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
The check for NULL ptr is not necessary, kfree will check it.
Removing NULL ptr check.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_debugfs.c | 12 ++++--------
1 file changed, 4 insertions(+), 8 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_debugfs.c b/drivers/scsi/lpfc/lpfc_debugfs.c
index 913eed8..76857e2 100644
--- a/drivers/scsi/lpfc/lpfc_debugfs.c
+++ b/drivers/scsi/lpfc/lpfc_debugfs.c
@@ -5700,10 +5700,8 @@ lpfc_debugfs_terminate(struct lpfc_vport *vport)
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
struct lpfc_hba *phba = vport->phba;
- if (vport->disc_trc) {
- kfree(vport->disc_trc);
- vport->disc_trc = NULL;
- }
+ kfree(vport->disc_trc);
+ vport->disc_trc = NULL;
debugfs_remove(vport->debug_disc_trc); /* discovery_trace */
vport->debug_disc_trc = NULL;
@@ -5770,10 +5768,8 @@ lpfc_debugfs_terminate(struct lpfc_vport *vport)
debugfs_remove(phba->debug_readRef); /* readRef */
phba->debug_readRef = NULL;
- if (phba->slow_ring_trc) {
- kfree(phba->slow_ring_trc);
- phba->slow_ring_trc = NULL;
- }
+ kfree(phba->slow_ring_trc);
+ phba->slow_ring_trc = NULL;
/* slow_ring_trace */
debugfs_remove(phba->debug_slow_ring_trc);
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 08/22] Remove NULL ptr check before kfree.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
The check for NULL ptr is not necessary, kfree will check it.
Removing NULL ptr check.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_debugfs.c | 12 ++++--------
1 file changed, 4 insertions(+), 8 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_debugfs.c b/drivers/scsi/lpfc/lpfc_debugfs.c
index 913eed8..76857e2 100644
--- a/drivers/scsi/lpfc/lpfc_debugfs.c
+++ b/drivers/scsi/lpfc/lpfc_debugfs.c
@@ -5700,10 +5700,8 @@ lpfc_debugfs_terminate(struct lpfc_vport *vport)
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
struct lpfc_hba *phba = vport->phba;
- if (vport->disc_trc) {
- kfree(vport->disc_trc);
- vport->disc_trc = NULL;
- }
+ kfree(vport->disc_trc);
+ vport->disc_trc = NULL;
debugfs_remove(vport->debug_disc_trc); /* discovery_trace */
vport->debug_disc_trc = NULL;
@@ -5770,10 +5768,8 @@ lpfc_debugfs_terminate(struct lpfc_vport *vport)
debugfs_remove(phba->debug_readRef); /* readRef */
phba->debug_readRef = NULL;
- if (phba->slow_ring_trc) {
- kfree(phba->slow_ring_trc);
- phba->slow_ring_trc = NULL;
- }
+ kfree(phba->slow_ring_trc);
+ phba->slow_ring_trc = NULL;
/* slow_ring_trace */
debugfs_remove(phba->debug_slow_ring_trc);
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 08/22] Remove NULL ptr check before kfree.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:34 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:34 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:27PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> The check for NULL ptr is not necessary, kfree will check it.
>
> Removing NULL ptr check.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 08/22] Remove NULL ptr check before kfree.
@ 2017-04-20 7:34 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:34 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:27PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> The check for NULL ptr is not necessary, kfree will check it.
>
> Removing NULL ptr check.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 09/22] Fix extra line print in rqpair debug print.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
An extra blank line was being added the the rqpair printing.
Remove the extra line feed.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_debugfs.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_debugfs.c b/drivers/scsi/lpfc/lpfc_debugfs.c
index 76857e2..55a8d8f 100644
--- a/drivers/scsi/lpfc/lpfc_debugfs.c
+++ b/drivers/scsi/lpfc/lpfc_debugfs.c
@@ -3128,8 +3128,6 @@ __lpfc_idiag_print_rqpair(struct lpfc_queue *qp, struct lpfc_queue *datqp,
datqp->queue_id, datqp->entry_count,
datqp->entry_size, datqp->host_index,
datqp->hba_index);
- len += snprintf(pbuffer + len, LPFC_QUE_INFO_GET_BUF_SIZE - len, "\n");
-
return len;
}
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH-v1 09/22] Fix extra line print in rqpair debug print.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
An extra blank line was being added the the rqpair printing.
Remove the extra line feed.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_debugfs.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_debugfs.c b/drivers/scsi/lpfc/lpfc_debugfs.c
index 76857e2..55a8d8f 100644
--- a/drivers/scsi/lpfc/lpfc_debugfs.c
+++ b/drivers/scsi/lpfc/lpfc_debugfs.c
@@ -3128,8 +3128,6 @@ __lpfc_idiag_print_rqpair(struct lpfc_queue *qp, struct lpfc_queue *datqp,
datqp->queue_id, datqp->entry_count,
datqp->entry_size, datqp->host_index,
datqp->hba_index);
- len += snprintf(pbuffer + len, LPFC_QUE_INFO_GET_BUF_SIZE - len, "\n");
-
return len;
}
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 09/22] Fix extra line print in rqpair debug print.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:35 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:35 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:28PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> An extra blank line was being added the the rqpair printing.
>
> Remove the extra line feed.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 09/22] Fix extra line print in rqpair debug print.
@ 2017-04-20 7:35 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:35 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:28PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> An extra blank line was being added the the rqpair printing.
>
> Remove the extra line feed.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 10/22] Fix PRLI ACC rsp for NVME
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
PRLI ACC from target is FCP oriented.
Word 0 was wrong. This was noticed by another nvmet-fc vendor that
was testing the lpfc nvme-fc initiator with their target.
Verified results with analyzer.
PRLI
BC B5 56 56 22 61 04 00 00 61 00 00 01 29 00 00 20 00 00 00
00 10 FF FF 00 00 00 00 20 14 00 18 28 00 00 00 00 00 00 00
00 00 00 00 00 00 00 20 00 00 00 00 9C D8 DA C9 BC 95 75 75
ACC
BC B5 56 56 23 61 00 00 00 61 04 00 01 98 00 00 30 00 00 00
00 10 00 18 00 00 00 00 02 14 00 18 28 00 01 00 00 00 00 00
00 00 00 00 00 00 00 18 00 00 00 00 B0 6B 07 57 BC B5 75 75
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_els.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index d9c61d0..d6e0f58 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -4403,7 +4403,7 @@ lpfc_els_rsp_prli_acc(struct lpfc_vport *vport, struct lpfc_iocbq *oldiocb,
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
memset(pcmd, 0, cmdsize);
- *((uint32_t *) (pcmd)) = (ELS_CMD_ACC | (ELS_CMD_PRLI & ~ELS_RSP_MASK));
+ *((uint32_t *)(pcmd)) = elsrspcmd;
pcmd += sizeof(uint32_t);
/* For PRLI, remainder of payload is PRLI parameter page */
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH-v1 10/22] Fix PRLI ACC rsp for NVME
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
PRLI ACC from target is FCP oriented.
Word 0 was wrong. This was noticed by another nvmet-fc vendor that
was testing the lpfc nvme-fc initiator with their target.
Verified results with analyzer.
PRLI
BC B5 56 56 22 61 04 00 00 61 00 00 01 29 00 00 20 00 00 00
00 10 FF FF 00 00 00 00 20 14 00 18 28 00 00 00 00 00 00 00
00 00 00 00 00 00 00 20 00 00 00 00 9C D8 DA C9 BC 95 75 75
ACC
BC B5 56 56 23 61 00 00 00 61 04 00 01 98 00 00 30 00 00 00
00 10 00 18 00 00 00 00 02 14 00 18 28 00 01 00 00 00 00 00
00 00 00 00 00 00 00 18 00 00 00 00 B0 6B 07 57 BC B5 75 75
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_els.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index d9c61d0..d6e0f58 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -4403,7 +4403,7 @@ lpfc_els_rsp_prli_acc(struct lpfc_vport *vport, struct lpfc_iocbq *oldiocb,
pcmd = (uint8_t *) (((struct lpfc_dmabuf *) elsiocb->context2)->virt);
memset(pcmd, 0, cmdsize);
- *((uint32_t *) (pcmd)) = (ELS_CMD_ACC | (ELS_CMD_PRLI & ~ELS_RSP_MASK));
+ *((uint32_t *)(pcmd)) = elsrspcmd;
pcmd += sizeof(uint32_t);
/* For PRLI, remainder of payload is PRLI parameter page */
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 10/22] Fix PRLI ACC rsp for NVME
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:35 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:35 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:29PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> PRLI ACC from target is FCP oriented.
>
> Word 0 was wrong. This was noticed by another nvmet-fc vendor that
> was testing the lpfc nvme-fc initiator with their target.
>
> Verified results with analyzer.
>
> PRLI
> BC B5 56 56 22 61 04 00 00 61 00 00 01 29 00 00 20 00 00 00
> 00 10 FF FF 00 00 00 00 20 14 00 18 28 00 00 00 00 00 00 00
> 00 00 00 00 00 00 00 20 00 00 00 00 9C D8 DA C9 BC 95 75 75
>
> ACC
> BC B5 56 56 23 61 00 00 00 61 04 00 01 98 00 00 30 00 00 00
> 00 10 00 18 00 00 00 00 02 14 00 18 28 00 01 00 00 00 00 00
> 00 00 00 00 00 00 00 18 00 00 00 00 B0 6B 07 57 BC B5 75 75
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 10/22] Fix PRLI ACC rsp for NVME
@ 2017-04-20 7:35 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:35 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:29PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> PRLI ACC from target is FCP oriented.
>
> Word 0 was wrong. This was noticed by another nvmet-fc vendor that
> was testing the lpfc nvme-fc initiator with their target.
>
> Verified results with analyzer.
>
> PRLI
> BC B5 56 56 22 61 04 00 00 61 00 00 01 29 00 00 20 00 00 00
> 00 10 FF FF 00 00 00 00 20 14 00 18 28 00 00 00 00 00 00 00
> 00 00 00 00 00 00 00 20 00 00 00 00 9C D8 DA C9 BC 95 75 75
>
> ACC
> BC B5 56 56 23 61 00 00 00 61 04 00 01 98 00 00 30 00 00 00
> 00 10 00 18 00 00 00 00 02 14 00 18 28 00 01 00 00 00 00 00
> 00 00 00 00 00 00 00 18 00 00 00 00 B0 6B 07 57 BC B5 75 75
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 11/22] Fix driver unload/reload operation.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
There are couple of different load/unload issues fixed with this patch.
One of the issues was reported by Junichi Nomura, a patch was submitted
by Johannes Thumsrhirn which did fix one of the problems but the fix in
this patch separates the pring free from the queue free and does not set
the parameter passed in to NULL.
issues:
(1) driver could not be unloaded and reloaded without some Oops or
Panic occurring.
(2) The driver was panicking because of a corruption in the Memory
Manager when the iocb list was getting allocated.
Root cause for the memory corruption was a double free of the Work Queue
ring pointer memory - Freed once in the lpfc_sli4_queue_free when the CQ
was destroyed and again in lpfc_sli4_queue_free when the WQ was destroyed.
The pring free and the queue free were separated, the pring free was moved
to the wq destroy routine because it a better fit logically to delete the
ring with the wq.
The checkpatch flagged several alignmenet issues that were also corrected
with this patch.
The mboxq was never initialed correctly before it was used by the driver
this patch corrects that issue.
Reported-by: Junichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_init.c | 46 +++++++++++++++++++++++--------------------
drivers/scsi/lpfc/lpfc_sli.c | 7 ++++++-
2 files changed, 31 insertions(+), 22 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index b56da01..cca7f81 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -5815,6 +5815,12 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
INIT_LIST_HEAD(&phba->sli4_hba.lpfc_vfi_blk_list);
INIT_LIST_HEAD(&phba->lpfc_vpi_blk_list);
+ /* Initialize mboxq lists. If the early init routines fail
+ * these lists need to be correctly initialized.
+ */
+ INIT_LIST_HEAD(&phba->sli.mboxq);
+ INIT_LIST_HEAD(&phba->sli.mboxq_cmpl);
+
/* initialize optic_state to 0xFF */
phba->sli4_hba.lnk_info.optic_state = 0xff;
@@ -5880,6 +5886,7 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
"READ_NV, mbxStatus x%x\n",
bf_get(lpfc_mqe_command, &mboxq->u.mqe),
bf_get(lpfc_mqe_status, &mboxq->u.mqe));
+ mempool_free(mboxq, phba->mbox_mem_pool);
rc = -EIO;
goto out_free_bsmbx;
}
@@ -7805,7 +7812,7 @@ lpfc_alloc_fcp_wq_cq(struct lpfc_hba *phba, int wqidx)
/* Create Fast Path FCP WQs */
wqesize = (phba->fcp_embed_io) ?
- LPFC_WQE128_SIZE : phba->sli4_hba.wq_esize;
+ LPFC_WQE128_SIZE : phba->sli4_hba.wq_esize;
qdesc = lpfc_sli4_queue_alloc(phba, wqesize, phba->sli4_hba.wq_ecount);
if (!qdesc) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
@@ -7836,7 +7843,7 @@ int
lpfc_sli4_queue_create(struct lpfc_hba *phba)
{
struct lpfc_queue *qdesc;
- int idx, io_channel, max;
+ int idx, io_channel;
/*
* Create HBA Record arrays.
@@ -7997,15 +8004,6 @@ lpfc_sli4_queue_create(struct lpfc_hba *phba)
if (lpfc_alloc_nvme_wq_cq(phba, idx))
goto out_error;
- /* allocate MRQ CQs */
- max = phba->cfg_nvme_io_channel;
- if (max < phba->cfg_nvmet_mrq)
- max = phba->cfg_nvmet_mrq;
-
- for (idx = 0; idx < max; idx++)
- if (lpfc_alloc_nvme_wq_cq(phba, idx))
- goto out_error;
-
if (phba->nvmet_support) {
for (idx = 0; idx < phba->cfg_nvmet_mrq; idx++) {
qdesc = lpfc_sli4_queue_alloc(phba,
@@ -8227,11 +8225,11 @@ lpfc_sli4_queue_destroy(struct lpfc_hba *phba)
/* Release FCP cqs */
lpfc_sli4_release_queues(&phba->sli4_hba.fcp_cq,
- phba->cfg_fcp_io_channel);
+ phba->cfg_fcp_io_channel);
/* Release FCP wqs */
lpfc_sli4_release_queues(&phba->sli4_hba.fcp_wq,
- phba->cfg_fcp_io_channel);
+ phba->cfg_fcp_io_channel);
/* Release FCP CQ mapping array */
lpfc_sli4_release_queue_map(&phba->sli4_hba.fcp_cq_map);
@@ -8577,15 +8575,15 @@ lpfc_sli4_queue_setup(struct lpfc_hba *phba)
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
"0528 %s not allocated\n",
phba->sli4_hba.mbx_cq ?
- "Mailbox WQ" : "Mailbox CQ");
+ "Mailbox WQ" : "Mailbox CQ");
rc = -ENOMEM;
goto out_destroy;
}
rc = lpfc_create_wq_cq(phba, phba->sli4_hba.hba_eq[0],
- phba->sli4_hba.mbx_cq,
- phba->sli4_hba.mbx_wq,
- NULL, 0, LPFC_MBOX);
+ phba->sli4_hba.mbx_cq,
+ phba->sli4_hba.mbx_wq,
+ NULL, 0, LPFC_MBOX);
if (rc) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
"0529 Failed setup of mailbox WQ/CQ: rc = 0x%x\n",
@@ -10054,9 +10052,14 @@ lpfc_sli4_hba_unset(struct lpfc_hba *phba)
/* Stop kthread signal shall trigger work_done one more time */
kthread_stop(phba->worker_thread);
+ /* Unset the queues shared with the hardware then release all
+ * allocated resources.
+ */
+ lpfc_sli4_queue_unset(phba);
+ lpfc_sli4_queue_destroy(phba);
+
/* Reset SLI4 HBA FCoE function */
lpfc_pci_function_reset(phba);
- lpfc_sli4_queue_destroy(phba);
/* Stop the SLI4 device port */
phba->pport->work_port_events = 0;
@@ -10312,6 +10315,7 @@ lpfc_pci_probe_one_s3(struct pci_dev *pdev, const struct pci_device_id *pid)
}
/* Initialize and populate the iocb list per host */
+
error = lpfc_init_iocb_list(phba, LPFC_IOCB_LIST_CNT);
if (error) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
@@ -11092,7 +11096,6 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
}
/* Initialize and populate the iocb list per host */
-
lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
"2821 initialize iocb list %d.\n",
phba->cfg_iocb_cnt*1024);
@@ -11183,7 +11186,9 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
if ((phba->nvmet_support == 0) &&
(phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME)) {
/* Create NVME binding with nvme_fc_transport. This
- * ensures the vport is initialized.
+ * ensures the vport is initialized. If the localport
+ * create fails, it should not unload the driver to
+ * support field issues.
*/
error = lpfc_nvme_create_localport(vport);
if (error) {
@@ -11191,7 +11196,6 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
"6004 NVME registration failed, "
"error x%x\n",
error);
- goto out_disable_intr;
}
}
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index 7087c55..d606944 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -13758,7 +13758,10 @@ lpfc_sli4_queue_free(struct lpfc_queue *queue)
lpfc_free_rq_buffer(queue->phba, queue);
kfree(queue->rqbp);
}
- kfree(queue->pring);
+
+ if (!list_empty(&queue->wq_list))
+ list_del(&queue->wq_list);
+
kfree(queue);
return;
}
@@ -15561,6 +15564,8 @@ lpfc_wq_destroy(struct lpfc_hba *phba, struct lpfc_queue *wq)
}
/* Remove wq from any list */
list_del_init(&wq->list);
+ kfree(wq->pring);
+ wq->pring = NULL;
mempool_free(mbox, wq->phba->mbox_mem_pool);
return status;
}
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 11/22] Fix driver unload/reload operation.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
There are couple of different load/unload issues fixed with this patch.
One of the issues was reported by Junichi Nomura, a patch was submitted
by Johannes Thumsrhirn which did fix one of the problems but the fix in
this patch separates the pring free from the queue free and does not set
the parameter passed in to NULL.
issues:
(1) driver could not be unloaded and reloaded without some Oops or
Panic occurring.
(2) The driver was panicking because of a corruption in the Memory
Manager when the iocb list was getting allocated.
Root cause for the memory corruption was a double free of the Work Queue
ring pointer memory - Freed once in the lpfc_sli4_queue_free when the CQ
was destroyed and again in lpfc_sli4_queue_free when the WQ was destroyed.
The pring free and the queue free were separated, the pring free was moved
to the wq destroy routine because it a better fit logically to delete the
ring with the wq.
The checkpatch flagged several alignmenet issues that were also corrected
with this patch.
The mboxq was never initialed correctly before it was used by the driver
this patch corrects that issue.
Reported-by: Junichi Nomura <j-nomura at ce.jp.nec.com>
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_init.c | 46 +++++++++++++++++++++++--------------------
drivers/scsi/lpfc/lpfc_sli.c | 7 ++++++-
2 files changed, 31 insertions(+), 22 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index b56da01..cca7f81 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -5815,6 +5815,12 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
INIT_LIST_HEAD(&phba->sli4_hba.lpfc_vfi_blk_list);
INIT_LIST_HEAD(&phba->lpfc_vpi_blk_list);
+ /* Initialize mboxq lists. If the early init routines fail
+ * these lists need to be correctly initialized.
+ */
+ INIT_LIST_HEAD(&phba->sli.mboxq);
+ INIT_LIST_HEAD(&phba->sli.mboxq_cmpl);
+
/* initialize optic_state to 0xFF */
phba->sli4_hba.lnk_info.optic_state = 0xff;
@@ -5880,6 +5886,7 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
"READ_NV, mbxStatus x%x\n",
bf_get(lpfc_mqe_command, &mboxq->u.mqe),
bf_get(lpfc_mqe_status, &mboxq->u.mqe));
+ mempool_free(mboxq, phba->mbox_mem_pool);
rc = -EIO;
goto out_free_bsmbx;
}
@@ -7805,7 +7812,7 @@ lpfc_alloc_fcp_wq_cq(struct lpfc_hba *phba, int wqidx)
/* Create Fast Path FCP WQs */
wqesize = (phba->fcp_embed_io) ?
- LPFC_WQE128_SIZE : phba->sli4_hba.wq_esize;
+ LPFC_WQE128_SIZE : phba->sli4_hba.wq_esize;
qdesc = lpfc_sli4_queue_alloc(phba, wqesize, phba->sli4_hba.wq_ecount);
if (!qdesc) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
@@ -7836,7 +7843,7 @@ int
lpfc_sli4_queue_create(struct lpfc_hba *phba)
{
struct lpfc_queue *qdesc;
- int idx, io_channel, max;
+ int idx, io_channel;
/*
* Create HBA Record arrays.
@@ -7997,15 +8004,6 @@ lpfc_sli4_queue_create(struct lpfc_hba *phba)
if (lpfc_alloc_nvme_wq_cq(phba, idx))
goto out_error;
- /* allocate MRQ CQs */
- max = phba->cfg_nvme_io_channel;
- if (max < phba->cfg_nvmet_mrq)
- max = phba->cfg_nvmet_mrq;
-
- for (idx = 0; idx < max; idx++)
- if (lpfc_alloc_nvme_wq_cq(phba, idx))
- goto out_error;
-
if (phba->nvmet_support) {
for (idx = 0; idx < phba->cfg_nvmet_mrq; idx++) {
qdesc = lpfc_sli4_queue_alloc(phba,
@@ -8227,11 +8225,11 @@ lpfc_sli4_queue_destroy(struct lpfc_hba *phba)
/* Release FCP cqs */
lpfc_sli4_release_queues(&phba->sli4_hba.fcp_cq,
- phba->cfg_fcp_io_channel);
+ phba->cfg_fcp_io_channel);
/* Release FCP wqs */
lpfc_sli4_release_queues(&phba->sli4_hba.fcp_wq,
- phba->cfg_fcp_io_channel);
+ phba->cfg_fcp_io_channel);
/* Release FCP CQ mapping array */
lpfc_sli4_release_queue_map(&phba->sli4_hba.fcp_cq_map);
@@ -8577,15 +8575,15 @@ lpfc_sli4_queue_setup(struct lpfc_hba *phba)
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
"0528 %s not allocated\n",
phba->sli4_hba.mbx_cq ?
- "Mailbox WQ" : "Mailbox CQ");
+ "Mailbox WQ" : "Mailbox CQ");
rc = -ENOMEM;
goto out_destroy;
}
rc = lpfc_create_wq_cq(phba, phba->sli4_hba.hba_eq[0],
- phba->sli4_hba.mbx_cq,
- phba->sli4_hba.mbx_wq,
- NULL, 0, LPFC_MBOX);
+ phba->sli4_hba.mbx_cq,
+ phba->sli4_hba.mbx_wq,
+ NULL, 0, LPFC_MBOX);
if (rc) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
"0529 Failed setup of mailbox WQ/CQ: rc = 0x%x\n",
@@ -10054,9 +10052,14 @@ lpfc_sli4_hba_unset(struct lpfc_hba *phba)
/* Stop kthread signal shall trigger work_done one more time */
kthread_stop(phba->worker_thread);
+ /* Unset the queues shared with the hardware then release all
+ * allocated resources.
+ */
+ lpfc_sli4_queue_unset(phba);
+ lpfc_sli4_queue_destroy(phba);
+
/* Reset SLI4 HBA FCoE function */
lpfc_pci_function_reset(phba);
- lpfc_sli4_queue_destroy(phba);
/* Stop the SLI4 device port */
phba->pport->work_port_events = 0;
@@ -10312,6 +10315,7 @@ lpfc_pci_probe_one_s3(struct pci_dev *pdev, const struct pci_device_id *pid)
}
/* Initialize and populate the iocb list per host */
+
error = lpfc_init_iocb_list(phba, LPFC_IOCB_LIST_CNT);
if (error) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
@@ -11092,7 +11096,6 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
}
/* Initialize and populate the iocb list per host */
-
lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
"2821 initialize iocb list %d.\n",
phba->cfg_iocb_cnt*1024);
@@ -11183,7 +11186,9 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
if ((phba->nvmet_support == 0) &&
(phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME)) {
/* Create NVME binding with nvme_fc_transport. This
- * ensures the vport is initialized.
+ * ensures the vport is initialized. If the localport
+ * create fails, it should not unload the driver to
+ * support field issues.
*/
error = lpfc_nvme_create_localport(vport);
if (error) {
@@ -11191,7 +11196,6 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
"6004 NVME registration failed, "
"error x%x\n",
error);
- goto out_disable_intr;
}
}
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index 7087c55..d606944 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -13758,7 +13758,10 @@ lpfc_sli4_queue_free(struct lpfc_queue *queue)
lpfc_free_rq_buffer(queue->phba, queue);
kfree(queue->rqbp);
}
- kfree(queue->pring);
+
+ if (!list_empty(&queue->wq_list))
+ list_del(&queue->wq_list);
+
kfree(queue);
return;
}
@@ -15561,6 +15564,8 @@ lpfc_wq_destroy(struct lpfc_hba *phba, struct lpfc_queue *wq)
}
/* Remove wq from any list */
list_del_init(&wq->list);
+ kfree(wq->pring);
+ wq->pring = NULL;
mempool_free(mbox, wq->phba->mbox_mem_pool);
return status;
}
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 11/22] Fix driver unload/reload operation.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:37 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:37 UTC (permalink / raw)
To: jsmart2021, Junichi Nomura
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:30PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> There are couple of different load/unload issues fixed with this patch.
> One of the issues was reported by Junichi Nomura, a patch was submitted
> by Johannes Thumsrhirn which did fix one of the problems but the fix in
> this patch separates the pring free from the queue free and does not set
> the parameter passed in to NULL.
>
> issues:
> (1) driver could not be unloaded and reloaded without some Oops or
> Panic occurring.
> (2) The driver was panicking because of a corruption in the Memory
> Manager when the iocb list was getting allocated.
>
> Root cause for the memory corruption was a double free of the Work Queue
> ring pointer memory - Freed once in the lpfc_sli4_queue_free when the CQ
> was destroyed and again in lpfc_sli4_queue_free when the WQ was destroyed.
>
> The pring free and the queue free were separated, the pring free was moved
> to the wq destroy routine because it a better fit logically to delete the
> ring with the wq.
>
> The checkpatch flagged several alignmenet issues that were also corrected
> with this patch.
>
> The mboxq was never initialed correctly before it was used by the driver
> this patch corrects that issue.
>
> Reported-by: Junichi Nomura <j-nomura@ce.jp.nec.com>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Looks good,
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
But I'd really like to see a Tested-by from Junichi here.
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 11/22] Fix driver unload/reload operation.
@ 2017-04-20 7:37 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:37 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:30PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> There are couple of different load/unload issues fixed with this patch.
> One of the issues was reported by Junichi Nomura, a patch was submitted
> by Johannes Thumsrhirn which did fix one of the problems but the fix in
> this patch separates the pring free from the queue free and does not set
> the parameter passed in to NULL.
>
> issues:
> (1) driver could not be unloaded and reloaded without some Oops or
> Panic occurring.
> (2) The driver was panicking because of a corruption in the Memory
> Manager when the iocb list was getting allocated.
>
> Root cause for the memory corruption was a double free of the Work Queue
> ring pointer memory - Freed once in the lpfc_sli4_queue_free when the CQ
> was destroyed and again in lpfc_sli4_queue_free when the WQ was destroyed.
>
> The pring free and the queue free were separated, the pring free was moved
> to the wq destroy routine because it a better fit logically to delete the
> ring with the wq.
>
> The checkpatch flagged several alignmenet issues that were also corrected
> with this patch.
>
> The mboxq was never initialed correctly before it was used by the driver
> this patch corrects that issue.
>
> Reported-by: Junichi Nomura <j-nomura at ce.jp.nec.com>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Looks good,
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
But I'd really like to see a Tested-by from Junichi here.
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 11/22] Fix driver unload/reload operation.
2017-04-20 7:37 ` Johannes Thumshirn
@ 2017-04-20 9:08 ` Junichi Nomura
-1 siblings, 0 replies; 100+ messages in thread
From: Junichi Nomura @ 2017-04-20 9:08 UTC (permalink / raw)
To: Johannes Thumshirn, jsmart2021, linux-scsi
Cc: linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On 04/20/17 16:37, Johannes Thumshirn wrote:
> On Wed, Apr 19, 2017 at 09:46:30PM -0700, jsmart2021@gmail.com wrote:
>> From: James Smart <jsmart2021@gmail.com>
>>
>> There are couple of different load/unload issues fixed with this patch.
>> One of the issues was reported by Junichi Nomura, a patch was submitted
>> by Johannes Thumsrhirn which did fix one of the problems but the fix in
>> this patch separates the pring free from the queue free and does not set
>> the parameter passed in to NULL.
>>
>> issues:
>> (1) driver could not be unloaded and reloaded without some Oops or
>> Panic occurring.
>> (2) The driver was panicking because of a corruption in the Memory
>> Manager when the iocb list was getting allocated.
>>
>> Root cause for the memory corruption was a double free of the Work Queue
>> ring pointer memory - Freed once in the lpfc_sli4_queue_free when the CQ
>> was destroyed and again in lpfc_sli4_queue_free when the WQ was destroyed.
>>
>> The pring free and the queue free were separated, the pring free was moved
>> to the wq destroy routine because it a better fit logically to delete the
>> ring with the wq.
>>
>> The checkpatch flagged several alignmenet issues that were also corrected
>> with this patch.
>>
>> The mboxq was never initialed correctly before it was used by the driver
>> this patch corrects that issue.
>>
>> Reported-by: Junichi Nomura <j-nomura@ce.jp.nec.com>
>> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
>> Signed-off-by: James Smart <james.smart@broadcom.com>
>> ---
>
> Looks good,
> Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
>
> But I'd really like to see a Tested-by from Junichi here.
Thank you for Cc-ing to me.
I tested the patch on top of v4.11-rc7 and confirmed the reported
problem doesn't occur. So,
Tested-by: Junichi Nomura <j-nomura@ce.jp.nec.com>
--
Jun'ichi Nomura, NEC Corporation / NEC Solution Innovators, Ltd.
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 11/22] Fix driver unload/reload operation.
@ 2017-04-20 9:08 ` Junichi Nomura
0 siblings, 0 replies; 100+ messages in thread
From: Junichi Nomura @ 2017-04-20 9:08 UTC (permalink / raw)
On 04/20/17 16:37, Johannes Thumshirn wrote:
> On Wed, Apr 19, 2017@09:46:30PM -0700, jsmart2021@gmail.com wrote:
>> From: James Smart <jsmart2021 at gmail.com>
>>
>> There are couple of different load/unload issues fixed with this patch.
>> One of the issues was reported by Junichi Nomura, a patch was submitted
>> by Johannes Thumsrhirn which did fix one of the problems but the fix in
>> this patch separates the pring free from the queue free and does not set
>> the parameter passed in to NULL.
>>
>> issues:
>> (1) driver could not be unloaded and reloaded without some Oops or
>> Panic occurring.
>> (2) The driver was panicking because of a corruption in the Memory
>> Manager when the iocb list was getting allocated.
>>
>> Root cause for the memory corruption was a double free of the Work Queue
>> ring pointer memory - Freed once in the lpfc_sli4_queue_free when the CQ
>> was destroyed and again in lpfc_sli4_queue_free when the WQ was destroyed.
>>
>> The pring free and the queue free were separated, the pring free was moved
>> to the wq destroy routine because it a better fit logically to delete the
>> ring with the wq.
>>
>> The checkpatch flagged several alignmenet issues that were also corrected
>> with this patch.
>>
>> The mboxq was never initialed correctly before it was used by the driver
>> this patch corrects that issue.
>>
>> Reported-by: Junichi Nomura <j-nomura at ce.jp.nec.com>
>> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
>> Signed-off-by: James Smart <james.smart at broadcom.com>
>> ---
>
> Looks good,
> Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
>
> But I'd really like to see a Tested-by from Junichi here.
Thank you for Cc-ing to me.
I tested the patch on top of v4.11-rc7 and confirmed the reported
problem doesn't occur. So,
Tested-by: Junichi Nomura <j-nomura at ce.jp.nec.com>
--
Jun'ichi Nomura, NEC Corporation / NEC Solution Innovators, Ltd.
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 12/22] Fix driver usage of 128B WQEs when WQ_CREATE is V1.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
There are two versions of a structure for queue creation and setup that the
driver shares with FW. The driver was only treating as version 0.
Verify WQ_CREATE with 128B WQEs in V0 and V1.
Code review of another bug showed the driver passing
128B WQEs and 8 pages in WQ CREATE and V0.
Code inspection/instrumentation showed that the driver
uses V0 in WQ_CREATE and if the caller passes queue->entry_size
128B, the driver sets the hdr_version to V1 so all is good.
When I tested the V1 WQ_CREATE, the mailbox failed causing
the driver to unload.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_sli.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index d606944..1519fdf 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -14741,6 +14741,9 @@ lpfc_wq_create(struct lpfc_hba *phba, struct lpfc_queue *wq,
case LPFC_Q_CREATE_VERSION_1:
bf_set(lpfc_mbx_wq_create_wqe_count, &wq_create->u.request_1,
wq->entry_count);
+ bf_set(lpfc_mbox_hdr_version, &shdr->request,
+ LPFC_Q_CREATE_VERSION_1);
+
switch (wq->entry_size) {
default:
case 64:
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 12/22] Fix driver usage of 128B WQEs when WQ_CREATE is V1.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
There are two versions of a structure for queue creation and setup that the
driver shares with FW. The driver was only treating as version 0.
Verify WQ_CREATE with 128B WQEs in V0 and V1.
Code review of another bug showed the driver passing
128B WQEs and 8 pages in WQ CREATE and V0.
Code inspection/instrumentation showed that the driver
uses V0 in WQ_CREATE and if the caller passes queue->entry_size
128B, the driver sets the hdr_version to V1 so all is good.
When I tested the V1 WQ_CREATE, the mailbox failed causing
the driver to unload.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_sli.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index d606944..1519fdf 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -14741,6 +14741,9 @@ lpfc_wq_create(struct lpfc_hba *phba, struct lpfc_queue *wq,
case LPFC_Q_CREATE_VERSION_1:
bf_set(lpfc_mbx_wq_create_wqe_count, &wq_create->u.request_1,
wq->entry_count);
+ bf_set(lpfc_mbox_hdr_version, &shdr->request,
+ LPFC_Q_CREATE_VERSION_1);
+
switch (wq->entry_size) {
default:
case 64:
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 12/22] Fix driver usage of 128B WQEs when WQ_CREATE is V1.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:37 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:37 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:31PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> There are two versions of a structure for queue creation and setup that the
> driver shares with FW. The driver was only treating as version 0.
>
> Verify WQ_CREATE with 128B WQEs in V0 and V1.
>
> Code review of another bug showed the driver passing
> 128B WQEs and 8 pages in WQ CREATE and V0.
> Code inspection/instrumentation showed that the driver
> uses V0 in WQ_CREATE and if the caller passes queue->entry_size
> 128B, the driver sets the hdr_version to V1 so all is good.
> When I tested the V1 WQ_CREATE, the mailbox failed causing
> the driver to unload.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 12/22] Fix driver usage of 128B WQEs when WQ_CREATE is V1.
@ 2017-04-20 7:37 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:37 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:31PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> There are two versions of a structure for queue creation and setup that the
> driver shares with FW. The driver was only treating as version 0.
>
> Verify WQ_CREATE with 128B WQEs in V0 and V1.
>
> Code review of another bug showed the driver passing
> 128B WQEs and 8 pages in WQ CREATE and V0.
> Code inspection/instrumentation showed that the driver
> uses V0 in WQ_CREATE and if the caller passes queue->entry_size
> 128B, the driver sets the hdr_version to V1 so all is good.
> When I tested the V1 WQ_CREATE, the mailbox failed causing
> the driver to unload.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 13/22] Fix nvme initiator handling when not enabled.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
Fix nvme initiator handline when CONFIG_LPFC_NVME_INITIATOR is not enabled.
With update nvme upstream driver sources, loading
the driver with nvme enabled resulting in this Oops.
BUG: unable to handle kernel NULL pointer dereference at 0000000000000018
IP: lpfc_nvme_update_localport+0x23/0xd0 [lpfc]
PGD 0
Oops: 0000 [#1] SMP
CPU: 0 PID: 10256 Comm: lpfc_worker_0 Tainted
Hardware name: ...
task: ffff881028191c40 task.stack: ffff880ffdf00000
RIP: 0010:lpfc_nvme_update_localport+0x23/0xd0 [lpfc]
RSP: 0018:ffff880ffdf03c20 EFLAGS: 00010202
Cause: As the initiator driver completes discovery at different stages,
it call lpfc_nvme_update_localport to hint that the DID and role may have
changed. In the implementation of lpfc_nvme_update_localport, the driver
was not validating the localport or the lport during the execution
of the update_localport routine. With the recent upstream additions to
the driver, the create_localport routine didn't run and so the localport
was NULL causing the page-fault Oops.
Fix: Add the CONFIG_LPFC_NVME_INITIATOR preprocessor inclusions to
lpfc_nvme_update_localport to turn off all routine processing when
the running kernel does not have NVME configured. Add NULL pointer
checks on the localport and lport in lpfc_nvme_update_localport and
dump messages if they are NULL and just exit.
Also one alingment issue fixed.
Repalces the ifdef with the IS_ENABLED macro.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_attr.c | 2 +-
drivers/scsi/lpfc/lpfc_nvme.c | 15 +++++++++++++--
2 files changed, 14 insertions(+), 3 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 22819af..4b32d02 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -3335,7 +3335,7 @@ LPFC_ATTR_R(enable_fc4_type, LPFC_ENABLE_FCP,
* percentage will go to NVME.
*/
LPFC_ATTR_R(xri_split, 50, 10, 90,
- "Division of XRI resources between SCSI and NVME");
+ "Division of XRI resources between SCSI and NVME");
/*
# lpfc_log_verbose: Only turn this flag on if you are willing to risk being
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index bc6946f..f23496b 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -2263,12 +2263,23 @@ lpfc_nvme_destroy_localport(struct lpfc_vport *vport)
void
lpfc_nvme_update_localport(struct lpfc_vport *vport)
{
+#if (IS_ENABLED(CONFIG_NVME_FC))
struct nvme_fc_local_port *localport;
struct lpfc_nvme_lport *lport;
localport = vport->localport;
+ if (!localport) {
+ lpfc_printf_vlog(vport, KERN_WARNING, LOG_NVME,
+ "6710 Update NVME fail. No localport\n");
+ return;
+ }
lport = (struct lpfc_nvme_lport *)localport->private;
-
+ if (!lport) {
+ lpfc_printf_vlog(vport, KERN_WARNING, LOG_NVME,
+ "6171 Update NVME fail. localP %p, No lport\n",
+ localport);
+ return;
+ }
lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME,
"6012 Update NVME lport %p did x%x\n",
localport, vport->fc_myDID);
@@ -2282,7 +2293,7 @@ lpfc_nvme_update_localport(struct lpfc_vport *vport)
lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME_DISC,
"6030 bound lport %p to DID x%06x\n",
lport, localport->port_id);
-
+#endif
}
int
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 13/22] Fix nvme initiator handling when not enabled.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
Fix nvme initiator handline when CONFIG_LPFC_NVME_INITIATOR is not enabled.
With update nvme upstream driver sources, loading
the driver with nvme enabled resulting in this Oops.
BUG: unable to handle kernel NULL pointer dereference at 0000000000000018
IP: lpfc_nvme_update_localport+0x23/0xd0 [lpfc]
PGD 0
Oops: 0000 [#1] SMP
CPU: 0 PID: 10256 Comm: lpfc_worker_0 Tainted
Hardware name: ...
task: ffff881028191c40 task.stack: ffff880ffdf00000
RIP: 0010:lpfc_nvme_update_localport+0x23/0xd0 [lpfc]
RSP: 0018:ffff880ffdf03c20 EFLAGS: 00010202
Cause: As the initiator driver completes discovery at different stages,
it call lpfc_nvme_update_localport to hint that the DID and role may have
changed. In the implementation of lpfc_nvme_update_localport, the driver
was not validating the localport or the lport during the execution
of the update_localport routine. With the recent upstream additions to
the driver, the create_localport routine didn't run and so the localport
was NULL causing the page-fault Oops.
Fix: Add the CONFIG_LPFC_NVME_INITIATOR preprocessor inclusions to
lpfc_nvme_update_localport to turn off all routine processing when
the running kernel does not have NVME configured. Add NULL pointer
checks on the localport and lport in lpfc_nvme_update_localport and
dump messages if they are NULL and just exit.
Also one alingment issue fixed.
Repalces the ifdef with the IS_ENABLED macro.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_attr.c | 2 +-
drivers/scsi/lpfc/lpfc_nvme.c | 15 +++++++++++++--
2 files changed, 14 insertions(+), 3 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 22819af..4b32d02 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -3335,7 +3335,7 @@ LPFC_ATTR_R(enable_fc4_type, LPFC_ENABLE_FCP,
* percentage will go to NVME.
*/
LPFC_ATTR_R(xri_split, 50, 10, 90,
- "Division of XRI resources between SCSI and NVME");
+ "Division of XRI resources between SCSI and NVME");
/*
# lpfc_log_verbose: Only turn this flag on if you are willing to risk being
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index bc6946f..f23496b 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -2263,12 +2263,23 @@ lpfc_nvme_destroy_localport(struct lpfc_vport *vport)
void
lpfc_nvme_update_localport(struct lpfc_vport *vport)
{
+#if (IS_ENABLED(CONFIG_NVME_FC))
struct nvme_fc_local_port *localport;
struct lpfc_nvme_lport *lport;
localport = vport->localport;
+ if (!localport) {
+ lpfc_printf_vlog(vport, KERN_WARNING, LOG_NVME,
+ "6710 Update NVME fail. No localport\n");
+ return;
+ }
lport = (struct lpfc_nvme_lport *)localport->private;
-
+ if (!lport) {
+ lpfc_printf_vlog(vport, KERN_WARNING, LOG_NVME,
+ "6171 Update NVME fail. localP %p, No lport\n",
+ localport);
+ return;
+ }
lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME,
"6012 Update NVME lport %p did x%x\n",
localport, vport->fc_myDID);
@@ -2282,7 +2293,7 @@ lpfc_nvme_update_localport(struct lpfc_vport *vport)
lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME_DISC,
"6030 bound lport %p to DID x%06x\n",
lport, localport->port_id);
-
+#endif
}
int
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 13/22] Fix nvme initiator handling when not enabled.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:38 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:38 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:32PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> Fix nvme initiator handline when CONFIG_LPFC_NVME_INITIATOR is not enabled.
>
> With update nvme upstream driver sources, loading
> the driver with nvme enabled resulting in this Oops.
>
> BUG: unable to handle kernel NULL pointer dereference at 0000000000000018
> IP: lpfc_nvme_update_localport+0x23/0xd0 [lpfc]
> PGD 0
> Oops: 0000 [#1] SMP
> CPU: 0 PID: 10256 Comm: lpfc_worker_0 Tainted
> Hardware name: ...
> task: ffff881028191c40 task.stack: ffff880ffdf00000
> RIP: 0010:lpfc_nvme_update_localport+0x23/0xd0 [lpfc]
> RSP: 0018:ffff880ffdf03c20 EFLAGS: 00010202
>
> Cause: As the initiator driver completes discovery at different stages,
> it call lpfc_nvme_update_localport to hint that the DID and role may have
> changed. In the implementation of lpfc_nvme_update_localport, the driver
> was not validating the localport or the lport during the execution
> of the update_localport routine. With the recent upstream additions to
> the driver, the create_localport routine didn't run and so the localport
> was NULL causing the page-fault Oops.
>
> Fix: Add the CONFIG_LPFC_NVME_INITIATOR preprocessor inclusions to
> lpfc_nvme_update_localport to turn off all routine processing when
> the running kernel does not have NVME configured. Add NULL pointer
> checks on the localport and lport in lpfc_nvme_update_localport and
> dump messages if they are NULL and just exit.
> Also one alingment issue fixed.
> Repalces the ifdef with the IS_ENABLED macro.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 13/22] Fix nvme initiator handling when not enabled.
@ 2017-04-20 7:38 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:38 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:32PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> Fix nvme initiator handline when CONFIG_LPFC_NVME_INITIATOR is not enabled.
>
> With update nvme upstream driver sources, loading
> the driver with nvme enabled resulting in this Oops.
>
> BUG: unable to handle kernel NULL pointer dereference at 0000000000000018
> IP: lpfc_nvme_update_localport+0x23/0xd0 [lpfc]
> PGD 0
> Oops: 0000 [#1] SMP
> CPU: 0 PID: 10256 Comm: lpfc_worker_0 Tainted
> Hardware name: ...
> task: ffff881028191c40 task.stack: ffff880ffdf00000
> RIP: 0010:lpfc_nvme_update_localport+0x23/0xd0 [lpfc]
> RSP: 0018:ffff880ffdf03c20 EFLAGS: 00010202
>
> Cause: As the initiator driver completes discovery at different stages,
> it call lpfc_nvme_update_localport to hint that the DID and role may have
> changed. In the implementation of lpfc_nvme_update_localport, the driver
> was not validating the localport or the lport during the execution
> of the update_localport routine. With the recent upstream additions to
> the driver, the create_localport routine didn't run and so the localport
> was NULL causing the page-fault Oops.
>
> Fix: Add the CONFIG_LPFC_NVME_INITIATOR preprocessor inclusions to
> lpfc_nvme_update_localport to turn off all routine processing when
> the running kernel does not have NVME configured. Add NULL pointer
> checks on the localport and lport in lpfc_nvme_update_localport and
> dump messages if they are NULL and just exit.
> Also one alingment issue fixed.
> Repalces the ifdef with the IS_ENABLED macro.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 14/22] Remove hba lock from NVMET issue WQE.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
Unnecessary lock is taken. ring lock should be sufficient to protect the
work queue submission.
This was noticed when doing performance testing. The hbalock is not
needed to issue io to the nvme work queue.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvmet.c | 4 ----
1 file changed, 4 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index c8a4941..5a994a8 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -517,7 +517,6 @@ lpfc_nvmet_xmt_fcp_op(struct nvmet_fc_target_port *tgtport,
container_of(rsp, struct lpfc_nvmet_rcv_ctx, ctx.fcp_req);
struct lpfc_hba *phba = ctxp->phba;
struct lpfc_iocbq *nvmewqeq;
- unsigned long iflags;
int rc, id;
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
@@ -571,10 +570,7 @@ lpfc_nvmet_xmt_fcp_op(struct nvmet_fc_target_port *tgtport,
lpfc_nvmeio_data(phba, "NVMET FCP CMND: xri x%x op x%x len x%x\n",
ctxp->oxid, rsp->op, rsp->rsplen);
- /* For now we take hbalock */
- spin_lock_irqsave(&phba->hbalock, iflags);
rc = lpfc_sli4_issue_wqe(phba, LPFC_FCP_RING, nvmewqeq);
- spin_unlock_irqrestore(&phba->hbalock, iflags);
if (rc == WQE_SUCCESS) {
ctxp->flag |= LPFC_NVMET_IO_INP;
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 14/22] Remove hba lock from NVMET issue WQE.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
Unnecessary lock is taken. ring lock should be sufficient to protect the
work queue submission.
This was noticed when doing performance testing. The hbalock is not
needed to issue io to the nvme work queue.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_nvmet.c | 4 ----
1 file changed, 4 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index c8a4941..5a994a8 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -517,7 +517,6 @@ lpfc_nvmet_xmt_fcp_op(struct nvmet_fc_target_port *tgtport,
container_of(rsp, struct lpfc_nvmet_rcv_ctx, ctx.fcp_req);
struct lpfc_hba *phba = ctxp->phba;
struct lpfc_iocbq *nvmewqeq;
- unsigned long iflags;
int rc, id;
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
@@ -571,10 +570,7 @@ lpfc_nvmet_xmt_fcp_op(struct nvmet_fc_target_port *tgtport,
lpfc_nvmeio_data(phba, "NVMET FCP CMND: xri x%x op x%x len x%x\n",
ctxp->oxid, rsp->op, rsp->rsplen);
- /* For now we take hbalock */
- spin_lock_irqsave(&phba->hbalock, iflags);
rc = lpfc_sli4_issue_wqe(phba, LPFC_FCP_RING, nvmewqeq);
- spin_unlock_irqrestore(&phba->hbalock, iflags);
if (rc == WQE_SUCCESS) {
ctxp->flag |= LPFC_NVMET_IO_INP;
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 14/22] Remove hba lock from NVMET issue WQE.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:40 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:40 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:33PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> Unnecessary lock is taken. ring lock should be sufficient to protect the
> work queue submission.
>
> This was noticed when doing performance testing. The hbalock is not
> needed to issue io to the nvme work queue.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 14/22] Remove hba lock from NVMET issue WQE.
@ 2017-04-20 7:40 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:40 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:33PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> Unnecessary lock is taken. ring lock should be sufficient to protect the
> work queue submission.
>
> This was noticed when doing performance testing. The hbalock is not
> needed to issue io to the nvme work queue.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 15/22] Fix driver load issues when MRQ=8
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
The symptom is that the driver will fail to login to the fabric.
The reason is because it is out of iocb resources.
There is a one to one relationship between MRQs
(receive buffers for NVMET-FC) and iocbs and the default number of
IOCBs was not accounting for the number of MRQs that were being created.
This fix aligns the number of MRQ resources with the total resources so
that it can handle fabric events when needed.
Also the initialization of ctxlock to be on FCP commands, NOT LS commands.
And modified log messages so that the log output can be correlated with
the analyzer trace.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_init.c | 12 ++++++++----
drivers/scsi/lpfc/lpfc_nvme.c | 23 ++++++++++++-----------
drivers/scsi/lpfc/lpfc_nvmet.c | 6 +++---
3 files changed, 23 insertions(+), 18 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index cca7f81..88977b8 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -11061,7 +11061,7 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
struct lpfc_hba *phba;
struct lpfc_vport *vport = NULL;
struct Scsi_Host *shost = NULL;
- int error;
+ int error, cnt;
uint32_t cfg_mode, intr_mode;
/* Allocate memory for HBA structure */
@@ -11095,11 +11095,15 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
goto out_unset_pci_mem_s4;
}
+ cnt = phba->cfg_iocb_cnt * 1024;
+ if (phba->nvmet_support)
+ cnt += (phba->cfg_nvmet_mrq_post * phba->cfg_nvmet_mrq);
+
/* Initialize and populate the iocb list per host */
lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
- "2821 initialize iocb list %d.\n",
- phba->cfg_iocb_cnt*1024);
- error = lpfc_init_iocb_list(phba, phba->cfg_iocb_cnt*1024);
+ "2821 initialize iocb list %d total %d\n",
+ phba->cfg_iocb_cnt, cnt);
+ error = lpfc_init_iocb_list(phba, cnt);
if (error) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index f23496b..efe1242 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -1494,6 +1494,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
"io buffer. Skipping abort req.\n");
return;
}
+ nvmereq_wqe = &lpfc_nbuf->cur_iocbq;
/*
* The lpfc_nbuf and the mapped nvme_fcreq in the driver's
@@ -1507,20 +1508,19 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
"6143 NVME req mismatch: "
"lpfc_nbuf %p nvmeCmd %p, "
- "pnvme_fcreq %p. Skipping Abort\n",
+ "pnvme_fcreq %p. Skipping Abort xri x%x\n",
lpfc_nbuf, lpfc_nbuf->nvmeCmd,
- pnvme_fcreq);
+ pnvme_fcreq, nvmereq_wqe->sli4_xritag);
return;
}
/* Don't abort IOs no longer on the pending queue. */
- nvmereq_wqe = &lpfc_nbuf->cur_iocbq;
if (!(nvmereq_wqe->iocb_flag & LPFC_IO_ON_TXCMPLQ)) {
spin_unlock_irqrestore(&phba->hbalock, flags);
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
"6142 NVME IO req %p not queued - skipping "
- "abort req\n",
- pnvme_fcreq);
+ "abort req xri x%x\n",
+ pnvme_fcreq, nvmereq_wqe->sli4_xritag);
return;
}
@@ -1534,8 +1534,9 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
"6144 Outstanding NVME I/O Abort Request "
"still pending on nvme_fcreq %p, "
- "lpfc_ncmd %p\n",
- pnvme_fcreq, lpfc_nbuf);
+ "lpfc_ncmd %p xri x%x\n",
+ pnvme_fcreq, lpfc_nbuf,
+ nvmereq_wqe->sli4_xritag);
return;
}
@@ -1544,8 +1545,8 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
spin_unlock_irqrestore(&phba->hbalock, flags);
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
"6136 No available abort wqes. Skipping "
- "Abts req for nvme_fcreq %p.\n",
- pnvme_fcreq);
+ "Abts req for nvme_fcreq %p xri x%x\n",
+ pnvme_fcreq, nvmereq_wqe->sli4_xritag);
return;
}
@@ -1603,7 +1604,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
}
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
- "6138 Transport Abort NVME Request Issued for\n"
+ "6138 Transport Abort NVME Request Issued for "
"ox_id x%x on reqtag x%x\n",
nvmereq_wqe->sli4_xritag,
abts_buf->iotag);
@@ -2490,7 +2491,7 @@ lpfc_nvme_unregister_port(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp)
input_err:
#endif
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_DISC,
- "6168: State error: lport %p, rport%p FCID x%06x\n",
+ "6168 State error: lport %p, rport%p FCID x%06x\n",
vport->localport, ndlp->rport, ndlp->nlp_DID);
}
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index 5a994a8..8348bb5 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -872,7 +872,6 @@ dropit:
ctxp->wqeq = NULL;
ctxp->state = LPFC_NVMET_STE_RCV;
ctxp->rqb_buffer = (void *)nvmebuf;
- spin_lock_init(&ctxp->ctxlock);
lpfc_nvmeio_data(phba, "NVMET LS RCV: xri x%x sz %d from %06x\n",
oxid, size, sid);
@@ -981,6 +980,7 @@ lpfc_nvmet_unsol_fcp_buffer(struct lpfc_hba *phba,
ctxp->rqb_buffer = nvmebuf;
ctxp->entry_cnt = 1;
ctxp->flag = 0;
+ spin_lock_init(&ctxp->ctxlock);
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
if (phba->ktime_on) {
@@ -1003,8 +1003,8 @@ lpfc_nvmet_unsol_fcp_buffer(struct lpfc_hba *phba,
}
#endif
- lpfc_nvmeio_data(phba, "NVMET FCP RCV: xri x%x sz %d from %06x\n",
- oxid, size, sid);
+ lpfc_nvmeio_data(phba, "NVMET FCP RCV: xri x%x sz %d CPU %02x\n",
+ oxid, size, smp_processor_id());
atomic_inc(&tgtp->rcv_fcp_cmd_in);
/*
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 15/22] Fix driver load issues when MRQ=8
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
The symptom is that the driver will fail to login to the fabric.
The reason is because it is out of iocb resources.
There is a one to one relationship between MRQs
(receive buffers for NVMET-FC) and iocbs and the default number of
IOCBs was not accounting for the number of MRQs that were being created.
This fix aligns the number of MRQ resources with the total resources so
that it can handle fabric events when needed.
Also the initialization of ctxlock to be on FCP commands, NOT LS commands.
And modified log messages so that the log output can be correlated with
the analyzer trace.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_init.c | 12 ++++++++----
drivers/scsi/lpfc/lpfc_nvme.c | 23 ++++++++++++-----------
drivers/scsi/lpfc/lpfc_nvmet.c | 6 +++---
3 files changed, 23 insertions(+), 18 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index cca7f81..88977b8 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -11061,7 +11061,7 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
struct lpfc_hba *phba;
struct lpfc_vport *vport = NULL;
struct Scsi_Host *shost = NULL;
- int error;
+ int error, cnt;
uint32_t cfg_mode, intr_mode;
/* Allocate memory for HBA structure */
@@ -11095,11 +11095,15 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
goto out_unset_pci_mem_s4;
}
+ cnt = phba->cfg_iocb_cnt * 1024;
+ if (phba->nvmet_support)
+ cnt += (phba->cfg_nvmet_mrq_post * phba->cfg_nvmet_mrq);
+
/* Initialize and populate the iocb list per host */
lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
- "2821 initialize iocb list %d.\n",
- phba->cfg_iocb_cnt*1024);
- error = lpfc_init_iocb_list(phba, phba->cfg_iocb_cnt*1024);
+ "2821 initialize iocb list %d total %d\n",
+ phba->cfg_iocb_cnt, cnt);
+ error = lpfc_init_iocb_list(phba, cnt);
if (error) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index f23496b..efe1242 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -1494,6 +1494,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
"io buffer. Skipping abort req.\n");
return;
}
+ nvmereq_wqe = &lpfc_nbuf->cur_iocbq;
/*
* The lpfc_nbuf and the mapped nvme_fcreq in the driver's
@@ -1507,20 +1508,19 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
"6143 NVME req mismatch: "
"lpfc_nbuf %p nvmeCmd %p, "
- "pnvme_fcreq %p. Skipping Abort\n",
+ "pnvme_fcreq %p. Skipping Abort xri x%x\n",
lpfc_nbuf, lpfc_nbuf->nvmeCmd,
- pnvme_fcreq);
+ pnvme_fcreq, nvmereq_wqe->sli4_xritag);
return;
}
/* Don't abort IOs no longer on the pending queue. */
- nvmereq_wqe = &lpfc_nbuf->cur_iocbq;
if (!(nvmereq_wqe->iocb_flag & LPFC_IO_ON_TXCMPLQ)) {
spin_unlock_irqrestore(&phba->hbalock, flags);
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
"6142 NVME IO req %p not queued - skipping "
- "abort req\n",
- pnvme_fcreq);
+ "abort req xri x%x\n",
+ pnvme_fcreq, nvmereq_wqe->sli4_xritag);
return;
}
@@ -1534,8 +1534,9 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
"6144 Outstanding NVME I/O Abort Request "
"still pending on nvme_fcreq %p, "
- "lpfc_ncmd %p\n",
- pnvme_fcreq, lpfc_nbuf);
+ "lpfc_ncmd %p xri x%x\n",
+ pnvme_fcreq, lpfc_nbuf,
+ nvmereq_wqe->sli4_xritag);
return;
}
@@ -1544,8 +1545,8 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
spin_unlock_irqrestore(&phba->hbalock, flags);
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
"6136 No available abort wqes. Skipping "
- "Abts req for nvme_fcreq %p.\n",
- pnvme_fcreq);
+ "Abts req for nvme_fcreq %p xri x%x\n",
+ pnvme_fcreq, nvmereq_wqe->sli4_xritag);
return;
}
@@ -1603,7 +1604,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
}
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
- "6138 Transport Abort NVME Request Issued for\n"
+ "6138 Transport Abort NVME Request Issued for "
"ox_id x%x on reqtag x%x\n",
nvmereq_wqe->sli4_xritag,
abts_buf->iotag);
@@ -2490,7 +2491,7 @@ lpfc_nvme_unregister_port(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp)
input_err:
#endif
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_DISC,
- "6168: State error: lport %p, rport%p FCID x%06x\n",
+ "6168 State error: lport %p, rport%p FCID x%06x\n",
vport->localport, ndlp->rport, ndlp->nlp_DID);
}
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index 5a994a8..8348bb5 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -872,7 +872,6 @@ dropit:
ctxp->wqeq = NULL;
ctxp->state = LPFC_NVMET_STE_RCV;
ctxp->rqb_buffer = (void *)nvmebuf;
- spin_lock_init(&ctxp->ctxlock);
lpfc_nvmeio_data(phba, "NVMET LS RCV: xri x%x sz %d from %06x\n",
oxid, size, sid);
@@ -981,6 +980,7 @@ lpfc_nvmet_unsol_fcp_buffer(struct lpfc_hba *phba,
ctxp->rqb_buffer = nvmebuf;
ctxp->entry_cnt = 1;
ctxp->flag = 0;
+ spin_lock_init(&ctxp->ctxlock);
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
if (phba->ktime_on) {
@@ -1003,8 +1003,8 @@ lpfc_nvmet_unsol_fcp_buffer(struct lpfc_hba *phba,
}
#endif
- lpfc_nvmeio_data(phba, "NVMET FCP RCV: xri x%x sz %d from %06x\n",
- oxid, size, sid);
+ lpfc_nvmeio_data(phba, "NVMET FCP RCV: xri x%x sz %d CPU %02x\n",
+ oxid, size, smp_processor_id());
atomic_inc(&tgtp->rcv_fcp_cmd_in);
/*
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 15/22] Fix driver load issues when MRQ=8
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:42 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:42 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:34PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> The symptom is that the driver will fail to login to the fabric.
> The reason is because it is out of iocb resources.
>
> There is a one to one relationship between MRQs
> (receive buffers for NVMET-FC) and iocbs and the default number of
> IOCBs was not accounting for the number of MRQs that were being created.
>
> This fix aligns the number of MRQ resources with the total resources so
> that it can handle fabric events when needed.
>
> Also the initialization of ctxlock to be on FCP commands, NOT LS commands.
> And modified log messages so that the log output can be correlated with
> the analyzer trace.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
> drivers/scsi/lpfc/lpfc_init.c | 12 ++++++++----
> drivers/scsi/lpfc/lpfc_nvme.c | 23 ++++++++++++-----------
> drivers/scsi/lpfc/lpfc_nvmet.c | 6 +++---
> 3 files changed, 23 insertions(+), 18 deletions(-)
>
> diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
> index cca7f81..88977b8 100644
> --- a/drivers/scsi/lpfc/lpfc_init.c
> +++ b/drivers/scsi/lpfc/lpfc_init.c
> @@ -11061,7 +11061,7 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
> struct lpfc_hba *phba;
> struct lpfc_vport *vport = NULL;
> struct Scsi_Host *shost = NULL;
> - int error;
> + int error, cnt;
> uint32_t cfg_mode, intr_mode;
>
> /* Allocate memory for HBA structure */
> @@ -11095,11 +11095,15 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
> goto out_unset_pci_mem_s4;
> }
>
> + cnt = phba->cfg_iocb_cnt * 1024;
> + if (phba->nvmet_support)
> + cnt += (phba->cfg_nvmet_mrq_post * phba->cfg_nvmet_mrq);
Souperflous parenthesis ^
Otherwise,
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 15/22] Fix driver load issues when MRQ=8
@ 2017-04-20 7:42 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:42 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:34PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> The symptom is that the driver will fail to login to the fabric.
> The reason is because it is out of iocb resources.
>
> There is a one to one relationship between MRQs
> (receive buffers for NVMET-FC) and iocbs and the default number of
> IOCBs was not accounting for the number of MRQs that were being created.
>
> This fix aligns the number of MRQ resources with the total resources so
> that it can handle fabric events when needed.
>
> Also the initialization of ctxlock to be on FCP commands, NOT LS commands.
> And modified log messages so that the log output can be correlated with
> the analyzer trace.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
> drivers/scsi/lpfc/lpfc_init.c | 12 ++++++++----
> drivers/scsi/lpfc/lpfc_nvme.c | 23 ++++++++++++-----------
> drivers/scsi/lpfc/lpfc_nvmet.c | 6 +++---
> 3 files changed, 23 insertions(+), 18 deletions(-)
>
> diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
> index cca7f81..88977b8 100644
> --- a/drivers/scsi/lpfc/lpfc_init.c
> +++ b/drivers/scsi/lpfc/lpfc_init.c
> @@ -11061,7 +11061,7 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
> struct lpfc_hba *phba;
> struct lpfc_vport *vport = NULL;
> struct Scsi_Host *shost = NULL;
> - int error;
> + int error, cnt;
> uint32_t cfg_mode, intr_mode;
>
> /* Allocate memory for HBA structure */
> @@ -11095,11 +11095,15 @@ lpfc_pci_probe_one_s4(struct pci_dev *pdev, const struct pci_device_id *pid)
> goto out_unset_pci_mem_s4;
> }
>
> + cnt = phba->cfg_iocb_cnt * 1024;
> + if (phba->nvmet_support)
> + cnt += (phba->cfg_nvmet_mrq_post * phba->cfg_nvmet_mrq);
Souperflous parenthesis ^
Otherwise,
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 16/22] Fix crash after issuing lip reset
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
When RPI is not available, driver sends WQE with invalid RPI value and
rejected by HBA.
lpfc 0000:82:00.3: 1:3154 BLS ABORT RSP failed, data: x3/xa0320008
and
lpfc :2753 PLOGI failure DID:FFFFFA Status:x3/xa0240008
In this case, driver accesses rpi_ids array out of bounds.
Fix:
Check return value of lpfc_sli4_alloc_rpi(). Do not allocate
lpfc_nodelist entry if RPI is not available.
When RPI is not available, we will get discovery timeouts and
command drops for some of the vports as seen below.
lpfc :0273 Unexpected discovery timeout, vport State x0
lpfc :0230 Unexpected timeout, hba link state x5
lpfc :0111 Dropping received ELS cmd Data: x0 xc90c55 x0
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_bsg.c | 4 ++++
drivers/scsi/lpfc/lpfc_crtn.h | 2 +-
drivers/scsi/lpfc/lpfc_els.c | 35 ++++++++--------------------
drivers/scsi/lpfc/lpfc_hbadisc.c | 47 +++++++++++++++++++++++++++++---------
drivers/scsi/lpfc/lpfc_init.c | 49 +++++++++++++++++++++-------------------
drivers/scsi/lpfc/lpfc_sli.c | 3 +--
drivers/scsi/lpfc/lpfc_vport.c | 3 +--
7 files changed, 79 insertions(+), 64 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_bsg.c b/drivers/scsi/lpfc/lpfc_bsg.c
index 18157d2..a1686c2 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.c
+++ b/drivers/scsi/lpfc/lpfc_bsg.c
@@ -2486,6 +2486,10 @@ static int lpfcdiag_loop_self_reg(struct lpfc_hba *phba, uint16_t *rpi)
mbox, *rpi);
else {
*rpi = lpfc_sli4_alloc_rpi(phba);
+ if (*rpi == LPFC_RPI_ALLOC_ERROR) {
+ mempool_free(mbox, phba->mbox_mem_pool);
+ return -EBUSY;
+ }
status = lpfc_reg_rpi(phba, phba->pport->vpi,
phba->pport->fc_myDID,
(uint8_t *)&phba->pport->fc_sparam,
diff --git a/drivers/scsi/lpfc/lpfc_crtn.h b/drivers/scsi/lpfc/lpfc_crtn.h
index 54e6ac4..d859aff 100644
--- a/drivers/scsi/lpfc/lpfc_crtn.h
+++ b/drivers/scsi/lpfc/lpfc_crtn.h
@@ -99,7 +99,7 @@ void lpfc_issue_reg_vpi(struct lpfc_hba *, struct lpfc_vport *);
int lpfc_check_sli_ndlp(struct lpfc_hba *, struct lpfc_sli_ring *,
struct lpfc_iocbq *, struct lpfc_nodelist *);
-void lpfc_nlp_init(struct lpfc_vport *, struct lpfc_nodelist *, uint32_t);
+struct lpfc_nodelist *lpfc_nlp_init(struct lpfc_vport *vport, uint32_t did);
struct lpfc_nodelist *lpfc_nlp_get(struct lpfc_nodelist *);
int lpfc_nlp_put(struct lpfc_nodelist *);
int lpfc_nlp_not_used(struct lpfc_nodelist *ndlp);
diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index d6e0f58..ffdcac0 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -895,10 +895,9 @@ lpfc_cmpl_els_flogi_nport(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
* Cannot find existing Fabric ndlp, so allocate a
* new one
*/
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, PT2PT_RemoteID);
if (!ndlp)
goto fail;
- lpfc_nlp_init(vport, ndlp, PT2PT_RemoteID);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp,
NLP_STE_UNUSED_NODE);
@@ -1364,7 +1363,6 @@ lpfc_els_abort_flogi(struct lpfc_hba *phba)
int
lpfc_initial_flogi(struct lpfc_vport *vport)
{
- struct lpfc_hba *phba = vport->phba;
struct lpfc_nodelist *ndlp;
vport->port_state = LPFC_FLOGI;
@@ -1374,10 +1372,9 @@ lpfc_initial_flogi(struct lpfc_vport *vport)
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, Fabric_DID);
if (!ndlp)
return 0;
- lpfc_nlp_init(vport, ndlp, Fabric_DID);
/* Set the node type */
ndlp->nlp_type |= NLP_FABRIC;
/* Put ndlp onto node list */
@@ -1418,17 +1415,15 @@ lpfc_initial_flogi(struct lpfc_vport *vport)
int
lpfc_initial_fdisc(struct lpfc_vport *vport)
{
- struct lpfc_hba *phba = vport->phba;
struct lpfc_nodelist *ndlp;
/* First look for the Fabric ndlp */
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, Fabric_DID);
if (!ndlp)
return 0;
- lpfc_nlp_init(vport, ndlp, Fabric_DID);
/* Put ndlp onto node list */
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
@@ -1564,14 +1559,13 @@ lpfc_plogi_confirm_nport(struct lpfc_hba *phba, uint32_t *prsp,
phba->active_rrq_pool);
return ndlp;
}
- new_ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_ATOMIC);
+ new_ndlp = lpfc_nlp_init(vport, ndlp->nlp_DID);
if (!new_ndlp) {
if (active_rrqs_xri_bitmap)
mempool_free(active_rrqs_xri_bitmap,
phba->active_rrq_pool);
return ndlp;
}
- lpfc_nlp_init(vport, new_ndlp, ndlp->nlp_DID);
} else if (!NLP_CHK_NODE_ACT(new_ndlp)) {
rc = memcmp(&ndlp->nlp_portname, name,
sizeof(struct lpfc_name));
@@ -2845,10 +2839,9 @@ lpfc_issue_els_scr(struct lpfc_vport *vport, uint32_t nportid, uint8_t retry)
ndlp = lpfc_findnode_did(vport, nportid);
if (!ndlp) {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, nportid);
if (!ndlp)
return 1;
- lpfc_nlp_init(vport, ndlp, nportid);
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
@@ -2938,10 +2931,9 @@ lpfc_issue_els_farpr(struct lpfc_vport *vport, uint32_t nportid, uint8_t retry)
ndlp = lpfc_findnode_did(vport, nportid);
if (!ndlp) {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, nportid);
if (!ndlp)
return 1;
- lpfc_nlp_init(vport, ndlp, nportid);
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
@@ -6133,7 +6125,6 @@ int
lpfc_els_handle_rscn(struct lpfc_vport *vport)
{
struct lpfc_nodelist *ndlp;
- struct lpfc_hba *phba = vport->phba;
/* Ignore RSCN if the port is being torn down. */
if (vport->load_flag & FC_UNLOADING) {
@@ -6182,12 +6173,11 @@ lpfc_els_handle_rscn(struct lpfc_vport *vport)
}
ndlp->nlp_prev_state = NLP_STE_UNUSED_NODE;
} else {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, NameServer_DID);
if (!ndlp) {
lpfc_els_flush_rscn(vport);
return 0;
}
- lpfc_nlp_init(vport, ndlp, NameServer_DID);
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_PLOGI_ISSUE);
}
@@ -7746,11 +7736,9 @@ lpfc_els_unsol_buffer(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
ndlp = lpfc_findnode_did(vport, did);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, did);
if (!ndlp)
goto dropit;
-
- lpfc_nlp_init(vport, ndlp, did);
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
newnode = 1;
if ((did & Fabric_DID_MASK) == Fabric_DID_MASK)
@@ -8192,7 +8180,6 @@ lpfc_els_unsol_event(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
static void
lpfc_start_fdmi(struct lpfc_vport *vport)
{
- struct lpfc_hba *phba = vport->phba;
struct lpfc_nodelist *ndlp;
/* If this is the first time, allocate an ndlp and initialize
@@ -8201,9 +8188,8 @@ lpfc_start_fdmi(struct lpfc_vport *vport)
*/
ndlp = lpfc_findnode_did(vport, FDMI_DID);
if (!ndlp) {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, FDMI_DID);
if (ndlp) {
- lpfc_nlp_init(vport, ndlp, FDMI_DID);
ndlp->nlp_type |= NLP_FABRIC;
} else {
return;
@@ -8256,7 +8242,7 @@ lpfc_do_scr_ns_plogi(struct lpfc_hba *phba, struct lpfc_vport *vport)
ndlp = lpfc_findnode_did(vport, NameServer_DID);
if (!ndlp) {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, NameServer_DID);
if (!ndlp) {
if (phba->fc_topology == LPFC_TOPOLOGY_LOOP) {
lpfc_disc_start(vport);
@@ -8267,7 +8253,6 @@ lpfc_do_scr_ns_plogi(struct lpfc_hba *phba, struct lpfc_vport *vport)
"0251 NameServer login: no memory\n");
return;
}
- lpfc_nlp_init(vport, ndlp, NameServer_DID);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
if (!ndlp) {
diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c
index 180b072..d313dde 100644
--- a/drivers/scsi/lpfc/lpfc_hbadisc.c
+++ b/drivers/scsi/lpfc/lpfc_hbadisc.c
@@ -4368,10 +4368,17 @@ lpfc_enable_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
uint32_t did;
unsigned long flags;
unsigned long *active_rrqs_xri_bitmap = NULL;
+ int rpi = LPFC_RPI_ALLOC_ERROR;
if (!ndlp)
return NULL;
+ if (phba->sli_rev == LPFC_SLI_REV4) {
+ rpi = lpfc_sli4_alloc_rpi(vport->phba);
+ if (rpi == LPFC_RPI_ALLOC_ERROR)
+ return NULL;
+ }
+
spin_lock_irqsave(&phba->ndlp_lock, flags);
/* The ndlp should not be in memory free mode */
if (NLP_CHK_FREE_REQ(ndlp)) {
@@ -4381,7 +4388,7 @@ lpfc_enable_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
"usgmap:x%x refcnt:%d\n",
(void *)ndlp, ndlp->nlp_usg_map,
kref_read(&ndlp->kref));
- return NULL;
+ goto free_rpi;
}
/* The ndlp should not already be in active mode */
if (NLP_CHK_NODE_ACT(ndlp)) {
@@ -4391,7 +4398,7 @@ lpfc_enable_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
"usgmap:x%x refcnt:%d\n",
(void *)ndlp, ndlp->nlp_usg_map,
kref_read(&ndlp->kref));
- return NULL;
+ goto free_rpi;
}
/* Keep the original DID */
@@ -4409,7 +4416,7 @@ lpfc_enable_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
spin_unlock_irqrestore(&phba->ndlp_lock, flags);
if (vport->phba->sli_rev == LPFC_SLI_REV4) {
- ndlp->nlp_rpi = lpfc_sli4_alloc_rpi(vport->phba);
+ ndlp->nlp_rpi = rpi;
lpfc_printf_vlog(vport, KERN_INFO, LOG_NODE,
"0008 rpi:%x DID:%x flg:%x refcnt:%d "
"map:%x %p\n", ndlp->nlp_rpi, ndlp->nlp_DID,
@@ -4426,6 +4433,11 @@ lpfc_enable_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
"node enable: did:x%x",
ndlp->nlp_DID, 0, 0);
return ndlp;
+
+free_rpi:
+ if (phba->sli_rev == LPFC_SLI_REV4)
+ lpfc_sli4_free_rpi(vport->phba, rpi);
+ return NULL;
}
void
@@ -5107,11 +5119,9 @@ lpfc_setup_disc_node(struct lpfc_vport *vport, uint32_t did)
if ((vport->fc_flag & FC_RSCN_MODE) != 0 &&
lpfc_rscn_payload_check(vport, did) == 0)
return NULL;
- ndlp = (struct lpfc_nodelist *)
- mempool_alloc(vport->phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, did);
if (!ndlp)
return NULL;
- lpfc_nlp_init(vport, ndlp, did);
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
if (vport->phba->nvmet_support)
return ndlp;
@@ -5887,16 +5897,31 @@ lpfc_find_vport_by_vpid(struct lpfc_hba *phba, uint16_t vpi)
return NULL;
}
-void
-lpfc_nlp_init(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
- uint32_t did)
+struct lpfc_nodelist *
+lpfc_nlp_init(struct lpfc_vport *vport, uint32_t did)
{
+ struct lpfc_nodelist *ndlp;
+ int rpi = LPFC_RPI_ALLOC_ERROR;
+
+ if (vport->phba->sli_rev == LPFC_SLI_REV4) {
+ rpi = lpfc_sli4_alloc_rpi(vport->phba);
+ if (rpi == LPFC_RPI_ALLOC_ERROR)
+ return NULL;
+ }
+
+ ndlp = mempool_alloc(vport->phba->nlp_mem_pool, GFP_KERNEL);
+ if (!ndlp) {
+ if (vport->phba->sli_rev == LPFC_SLI_REV4)
+ lpfc_sli4_free_rpi(vport->phba, rpi);
+ return NULL;
+ }
+
memset(ndlp, 0, sizeof (struct lpfc_nodelist));
lpfc_initialize_node(vport, ndlp, did);
INIT_LIST_HEAD(&ndlp->nlp_listp);
if (vport->phba->sli_rev == LPFC_SLI_REV4) {
- ndlp->nlp_rpi = lpfc_sli4_alloc_rpi(vport->phba);
+ ndlp->nlp_rpi = rpi;
lpfc_printf_vlog(vport, KERN_INFO, LOG_NODE,
"0007 rpi:%x DID:%x flg:%x refcnt:%d "
"map:%x %p\n", ndlp->nlp_rpi, ndlp->nlp_DID,
@@ -5918,7 +5943,7 @@ lpfc_nlp_init(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
"node init: did:x%x",
ndlp->nlp_DID, 0, 0);
- return;
+ return ndlp;
}
/* This routine releases all resources associated with a specifc NPort's ndlp
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 88977b8..61bf14c 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -2874,34 +2874,38 @@ lpfc_sli4_node_prep(struct lpfc_hba *phba)
{
struct lpfc_nodelist *ndlp, *next_ndlp;
struct lpfc_vport **vports;
- int i;
+ int i, rpi;
+ unsigned long flags;
if (phba->sli_rev != LPFC_SLI_REV4)
return;
vports = lpfc_create_vport_work_array(phba);
- if (vports != NULL) {
- for (i = 0; i <= phba->max_vports && vports[i] != NULL; i++) {
- if (vports[i]->load_flag & FC_UNLOADING)
- continue;
+ if (vports == NULL)
+ return;
- list_for_each_entry_safe(ndlp, next_ndlp,
- &vports[i]->fc_nodes,
- nlp_listp) {
- if (NLP_CHK_NODE_ACT(ndlp)) {
- ndlp->nlp_rpi =
- lpfc_sli4_alloc_rpi(phba);
- lpfc_printf_vlog(ndlp->vport, KERN_INFO,
- LOG_NODE,
- "0009 rpi:%x DID:%x "
- "flg:%x map:%x %p\n",
- ndlp->nlp_rpi,
- ndlp->nlp_DID,
- ndlp->nlp_flag,
- ndlp->nlp_usg_map,
- ndlp);
- }
+ for (i = 0; i <= phba->max_vports && vports[i] != NULL; i++) {
+ if (vports[i]->load_flag & FC_UNLOADING)
+ continue;
+
+ list_for_each_entry_safe(ndlp, next_ndlp,
+ &vports[i]->fc_nodes,
+ nlp_listp) {
+ if (!NLP_CHK_NODE_ACT(ndlp))
+ continue;
+ rpi = lpfc_sli4_alloc_rpi(phba);
+ if (rpi == LPFC_RPI_ALLOC_ERROR) {
+ spin_lock_irqsave(&phba->ndlp_lock, flags);
+ NLP_CLR_NODE_ACT(ndlp);
+ spin_unlock_irqrestore(&phba->ndlp_lock, flags);
+ continue;
}
+ ndlp->nlp_rpi = rpi;
+ lpfc_printf_vlog(ndlp->vport, KERN_INFO, LOG_NODE,
+ "0009 rpi:%x DID:%x "
+ "flg:%x map:%x %p\n", ndlp->nlp_rpi,
+ ndlp->nlp_DID, ndlp->nlp_flag,
+ ndlp->nlp_usg_map, ndlp);
}
}
lpfc_destroy_vport_work_array(phba, vports);
@@ -4722,10 +4726,9 @@ lpfc_sli4_perform_vport_cvl(struct lpfc_vport *vport)
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, Fabric_DID);
if (!ndlp)
return 0;
- lpfc_nlp_init(vport, ndlp, Fabric_DID);
/* Set the node type */
ndlp->nlp_type |= NLP_FABRIC;
/* Put ndlp onto node list */
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index 1519fdf..a6adc2b 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -16542,14 +16542,13 @@ lpfc_sli4_seq_abort_rsp(struct lpfc_vport *vport,
ndlp = lpfc_findnode_did(vport, sid);
if (!ndlp) {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, sid);
if (!ndlp) {
lpfc_printf_vlog(vport, KERN_WARNING, LOG_ELS,
"1268 Failed to allocate ndlp for "
"oxid:x%x SID:x%x\n", oxid, sid);
return;
}
- lpfc_nlp_init(vport, ndlp, sid);
/* Put ndlp onto pport node list */
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
diff --git a/drivers/scsi/lpfc/lpfc_vport.c b/drivers/scsi/lpfc/lpfc_vport.c
index 9a0339d..c714482 100644
--- a/drivers/scsi/lpfc/lpfc_vport.c
+++ b/drivers/scsi/lpfc/lpfc_vport.c
@@ -738,10 +738,9 @@ lpfc_vport_delete(struct fc_vport *fc_vport)
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, allocate one */
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, Fabric_DID);
if (!ndlp)
goto skip_logo;
- lpfc_nlp_init(vport, ndlp, Fabric_DID);
/* Indicate free memory when release */
NLP_SET_FREE_REQ(ndlp);
} else {
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 16/22] Fix crash after issuing lip reset
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
When RPI is not available, driver sends WQE with invalid RPI value and
rejected by HBA.
lpfc 0000:82:00.3: 1:3154 BLS ABORT RSP failed, data: x3/xa0320008
and
lpfc :2753 PLOGI failure DID:FFFFFA Status:x3/xa0240008
In this case, driver accesses rpi_ids array out of bounds.
Fix:
Check return value of lpfc_sli4_alloc_rpi(). Do not allocate
lpfc_nodelist entry if RPI is not available.
When RPI is not available, we will get discovery timeouts and
command drops for some of the vports as seen below.
lpfc :0273 Unexpected discovery timeout, vport State x0
lpfc :0230 Unexpected timeout, hba link state x5
lpfc :0111 Dropping received ELS cmd Data: x0 xc90c55 x0
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_bsg.c | 4 ++++
drivers/scsi/lpfc/lpfc_crtn.h | 2 +-
drivers/scsi/lpfc/lpfc_els.c | 35 ++++++++--------------------
drivers/scsi/lpfc/lpfc_hbadisc.c | 47 +++++++++++++++++++++++++++++---------
drivers/scsi/lpfc/lpfc_init.c | 49 +++++++++++++++++++++-------------------
drivers/scsi/lpfc/lpfc_sli.c | 3 +--
drivers/scsi/lpfc/lpfc_vport.c | 3 +--
7 files changed, 79 insertions(+), 64 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_bsg.c b/drivers/scsi/lpfc/lpfc_bsg.c
index 18157d2..a1686c2 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.c
+++ b/drivers/scsi/lpfc/lpfc_bsg.c
@@ -2486,6 +2486,10 @@ static int lpfcdiag_loop_self_reg(struct lpfc_hba *phba, uint16_t *rpi)
mbox, *rpi);
else {
*rpi = lpfc_sli4_alloc_rpi(phba);
+ if (*rpi == LPFC_RPI_ALLOC_ERROR) {
+ mempool_free(mbox, phba->mbox_mem_pool);
+ return -EBUSY;
+ }
status = lpfc_reg_rpi(phba, phba->pport->vpi,
phba->pport->fc_myDID,
(uint8_t *)&phba->pport->fc_sparam,
diff --git a/drivers/scsi/lpfc/lpfc_crtn.h b/drivers/scsi/lpfc/lpfc_crtn.h
index 54e6ac4..d859aff 100644
--- a/drivers/scsi/lpfc/lpfc_crtn.h
+++ b/drivers/scsi/lpfc/lpfc_crtn.h
@@ -99,7 +99,7 @@ void lpfc_issue_reg_vpi(struct lpfc_hba *, struct lpfc_vport *);
int lpfc_check_sli_ndlp(struct lpfc_hba *, struct lpfc_sli_ring *,
struct lpfc_iocbq *, struct lpfc_nodelist *);
-void lpfc_nlp_init(struct lpfc_vport *, struct lpfc_nodelist *, uint32_t);
+struct lpfc_nodelist *lpfc_nlp_init(struct lpfc_vport *vport, uint32_t did);
struct lpfc_nodelist *lpfc_nlp_get(struct lpfc_nodelist *);
int lpfc_nlp_put(struct lpfc_nodelist *);
int lpfc_nlp_not_used(struct lpfc_nodelist *ndlp);
diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index d6e0f58..ffdcac0 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -895,10 +895,9 @@ lpfc_cmpl_els_flogi_nport(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
* Cannot find existing Fabric ndlp, so allocate a
* new one
*/
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, PT2PT_RemoteID);
if (!ndlp)
goto fail;
- lpfc_nlp_init(vport, ndlp, PT2PT_RemoteID);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp,
NLP_STE_UNUSED_NODE);
@@ -1364,7 +1363,6 @@ lpfc_els_abort_flogi(struct lpfc_hba *phba)
int
lpfc_initial_flogi(struct lpfc_vport *vport)
{
- struct lpfc_hba *phba = vport->phba;
struct lpfc_nodelist *ndlp;
vport->port_state = LPFC_FLOGI;
@@ -1374,10 +1372,9 @@ lpfc_initial_flogi(struct lpfc_vport *vport)
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, Fabric_DID);
if (!ndlp)
return 0;
- lpfc_nlp_init(vport, ndlp, Fabric_DID);
/* Set the node type */
ndlp->nlp_type |= NLP_FABRIC;
/* Put ndlp onto node list */
@@ -1418,17 +1415,15 @@ lpfc_initial_flogi(struct lpfc_vport *vport)
int
lpfc_initial_fdisc(struct lpfc_vport *vport)
{
- struct lpfc_hba *phba = vport->phba;
struct lpfc_nodelist *ndlp;
/* First look for the Fabric ndlp */
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, Fabric_DID);
if (!ndlp)
return 0;
- lpfc_nlp_init(vport, ndlp, Fabric_DID);
/* Put ndlp onto node list */
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
@@ -1564,14 +1559,13 @@ lpfc_plogi_confirm_nport(struct lpfc_hba *phba, uint32_t *prsp,
phba->active_rrq_pool);
return ndlp;
}
- new_ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_ATOMIC);
+ new_ndlp = lpfc_nlp_init(vport, ndlp->nlp_DID);
if (!new_ndlp) {
if (active_rrqs_xri_bitmap)
mempool_free(active_rrqs_xri_bitmap,
phba->active_rrq_pool);
return ndlp;
}
- lpfc_nlp_init(vport, new_ndlp, ndlp->nlp_DID);
} else if (!NLP_CHK_NODE_ACT(new_ndlp)) {
rc = memcmp(&ndlp->nlp_portname, name,
sizeof(struct lpfc_name));
@@ -2845,10 +2839,9 @@ lpfc_issue_els_scr(struct lpfc_vport *vport, uint32_t nportid, uint8_t retry)
ndlp = lpfc_findnode_did(vport, nportid);
if (!ndlp) {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, nportid);
if (!ndlp)
return 1;
- lpfc_nlp_init(vport, ndlp, nportid);
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
@@ -2938,10 +2931,9 @@ lpfc_issue_els_farpr(struct lpfc_vport *vport, uint32_t nportid, uint8_t retry)
ndlp = lpfc_findnode_did(vport, nportid);
if (!ndlp) {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, nportid);
if (!ndlp)
return 1;
- lpfc_nlp_init(vport, ndlp, nportid);
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
@@ -6133,7 +6125,6 @@ int
lpfc_els_handle_rscn(struct lpfc_vport *vport)
{
struct lpfc_nodelist *ndlp;
- struct lpfc_hba *phba = vport->phba;
/* Ignore RSCN if the port is being torn down. */
if (vport->load_flag & FC_UNLOADING) {
@@ -6182,12 +6173,11 @@ lpfc_els_handle_rscn(struct lpfc_vport *vport)
}
ndlp->nlp_prev_state = NLP_STE_UNUSED_NODE;
} else {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, NameServer_DID);
if (!ndlp) {
lpfc_els_flush_rscn(vport);
return 0;
}
- lpfc_nlp_init(vport, ndlp, NameServer_DID);
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_PLOGI_ISSUE);
}
@@ -7746,11 +7736,9 @@ lpfc_els_unsol_buffer(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
ndlp = lpfc_findnode_did(vport, did);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, did);
if (!ndlp)
goto dropit;
-
- lpfc_nlp_init(vport, ndlp, did);
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
newnode = 1;
if ((did & Fabric_DID_MASK) == Fabric_DID_MASK)
@@ -8192,7 +8180,6 @@ lpfc_els_unsol_event(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
static void
lpfc_start_fdmi(struct lpfc_vport *vport)
{
- struct lpfc_hba *phba = vport->phba;
struct lpfc_nodelist *ndlp;
/* If this is the first time, allocate an ndlp and initialize
@@ -8201,9 +8188,8 @@ lpfc_start_fdmi(struct lpfc_vport *vport)
*/
ndlp = lpfc_findnode_did(vport, FDMI_DID);
if (!ndlp) {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, FDMI_DID);
if (ndlp) {
- lpfc_nlp_init(vport, ndlp, FDMI_DID);
ndlp->nlp_type |= NLP_FABRIC;
} else {
return;
@@ -8256,7 +8242,7 @@ lpfc_do_scr_ns_plogi(struct lpfc_hba *phba, struct lpfc_vport *vport)
ndlp = lpfc_findnode_did(vport, NameServer_DID);
if (!ndlp) {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, NameServer_DID);
if (!ndlp) {
if (phba->fc_topology == LPFC_TOPOLOGY_LOOP) {
lpfc_disc_start(vport);
@@ -8267,7 +8253,6 @@ lpfc_do_scr_ns_plogi(struct lpfc_hba *phba, struct lpfc_vport *vport)
"0251 NameServer login: no memory\n");
return;
}
- lpfc_nlp_init(vport, ndlp, NameServer_DID);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_UNUSED_NODE);
if (!ndlp) {
diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c
index 180b072..d313dde 100644
--- a/drivers/scsi/lpfc/lpfc_hbadisc.c
+++ b/drivers/scsi/lpfc/lpfc_hbadisc.c
@@ -4368,10 +4368,17 @@ lpfc_enable_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
uint32_t did;
unsigned long flags;
unsigned long *active_rrqs_xri_bitmap = NULL;
+ int rpi = LPFC_RPI_ALLOC_ERROR;
if (!ndlp)
return NULL;
+ if (phba->sli_rev == LPFC_SLI_REV4) {
+ rpi = lpfc_sli4_alloc_rpi(vport->phba);
+ if (rpi == LPFC_RPI_ALLOC_ERROR)
+ return NULL;
+ }
+
spin_lock_irqsave(&phba->ndlp_lock, flags);
/* The ndlp should not be in memory free mode */
if (NLP_CHK_FREE_REQ(ndlp)) {
@@ -4381,7 +4388,7 @@ lpfc_enable_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
"usgmap:x%x refcnt:%d\n",
(void *)ndlp, ndlp->nlp_usg_map,
kref_read(&ndlp->kref));
- return NULL;
+ goto free_rpi;
}
/* The ndlp should not already be in active mode */
if (NLP_CHK_NODE_ACT(ndlp)) {
@@ -4391,7 +4398,7 @@ lpfc_enable_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
"usgmap:x%x refcnt:%d\n",
(void *)ndlp, ndlp->nlp_usg_map,
kref_read(&ndlp->kref));
- return NULL;
+ goto free_rpi;
}
/* Keep the original DID */
@@ -4409,7 +4416,7 @@ lpfc_enable_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
spin_unlock_irqrestore(&phba->ndlp_lock, flags);
if (vport->phba->sli_rev == LPFC_SLI_REV4) {
- ndlp->nlp_rpi = lpfc_sli4_alloc_rpi(vport->phba);
+ ndlp->nlp_rpi = rpi;
lpfc_printf_vlog(vport, KERN_INFO, LOG_NODE,
"0008 rpi:%x DID:%x flg:%x refcnt:%d "
"map:%x %p\n", ndlp->nlp_rpi, ndlp->nlp_DID,
@@ -4426,6 +4433,11 @@ lpfc_enable_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
"node enable: did:x%x",
ndlp->nlp_DID, 0, 0);
return ndlp;
+
+free_rpi:
+ if (phba->sli_rev == LPFC_SLI_REV4)
+ lpfc_sli4_free_rpi(vport->phba, rpi);
+ return NULL;
}
void
@@ -5107,11 +5119,9 @@ lpfc_setup_disc_node(struct lpfc_vport *vport, uint32_t did)
if ((vport->fc_flag & FC_RSCN_MODE) != 0 &&
lpfc_rscn_payload_check(vport, did) == 0)
return NULL;
- ndlp = (struct lpfc_nodelist *)
- mempool_alloc(vport->phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, did);
if (!ndlp)
return NULL;
- lpfc_nlp_init(vport, ndlp, did);
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
if (vport->phba->nvmet_support)
return ndlp;
@@ -5887,16 +5897,31 @@ lpfc_find_vport_by_vpid(struct lpfc_hba *phba, uint16_t vpi)
return NULL;
}
-void
-lpfc_nlp_init(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
- uint32_t did)
+struct lpfc_nodelist *
+lpfc_nlp_init(struct lpfc_vport *vport, uint32_t did)
{
+ struct lpfc_nodelist *ndlp;
+ int rpi = LPFC_RPI_ALLOC_ERROR;
+
+ if (vport->phba->sli_rev == LPFC_SLI_REV4) {
+ rpi = lpfc_sli4_alloc_rpi(vport->phba);
+ if (rpi == LPFC_RPI_ALLOC_ERROR)
+ return NULL;
+ }
+
+ ndlp = mempool_alloc(vport->phba->nlp_mem_pool, GFP_KERNEL);
+ if (!ndlp) {
+ if (vport->phba->sli_rev == LPFC_SLI_REV4)
+ lpfc_sli4_free_rpi(vport->phba, rpi);
+ return NULL;
+ }
+
memset(ndlp, 0, sizeof (struct lpfc_nodelist));
lpfc_initialize_node(vport, ndlp, did);
INIT_LIST_HEAD(&ndlp->nlp_listp);
if (vport->phba->sli_rev == LPFC_SLI_REV4) {
- ndlp->nlp_rpi = lpfc_sli4_alloc_rpi(vport->phba);
+ ndlp->nlp_rpi = rpi;
lpfc_printf_vlog(vport, KERN_INFO, LOG_NODE,
"0007 rpi:%x DID:%x flg:%x refcnt:%d "
"map:%x %p\n", ndlp->nlp_rpi, ndlp->nlp_DID,
@@ -5918,7 +5943,7 @@ lpfc_nlp_init(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
"node init: did:x%x",
ndlp->nlp_DID, 0, 0);
- return;
+ return ndlp;
}
/* This routine releases all resources associated with a specifc NPort's ndlp
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 88977b8..61bf14c 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -2874,34 +2874,38 @@ lpfc_sli4_node_prep(struct lpfc_hba *phba)
{
struct lpfc_nodelist *ndlp, *next_ndlp;
struct lpfc_vport **vports;
- int i;
+ int i, rpi;
+ unsigned long flags;
if (phba->sli_rev != LPFC_SLI_REV4)
return;
vports = lpfc_create_vport_work_array(phba);
- if (vports != NULL) {
- for (i = 0; i <= phba->max_vports && vports[i] != NULL; i++) {
- if (vports[i]->load_flag & FC_UNLOADING)
- continue;
+ if (vports == NULL)
+ return;
- list_for_each_entry_safe(ndlp, next_ndlp,
- &vports[i]->fc_nodes,
- nlp_listp) {
- if (NLP_CHK_NODE_ACT(ndlp)) {
- ndlp->nlp_rpi =
- lpfc_sli4_alloc_rpi(phba);
- lpfc_printf_vlog(ndlp->vport, KERN_INFO,
- LOG_NODE,
- "0009 rpi:%x DID:%x "
- "flg:%x map:%x %p\n",
- ndlp->nlp_rpi,
- ndlp->nlp_DID,
- ndlp->nlp_flag,
- ndlp->nlp_usg_map,
- ndlp);
- }
+ for (i = 0; i <= phba->max_vports && vports[i] != NULL; i++) {
+ if (vports[i]->load_flag & FC_UNLOADING)
+ continue;
+
+ list_for_each_entry_safe(ndlp, next_ndlp,
+ &vports[i]->fc_nodes,
+ nlp_listp) {
+ if (!NLP_CHK_NODE_ACT(ndlp))
+ continue;
+ rpi = lpfc_sli4_alloc_rpi(phba);
+ if (rpi == LPFC_RPI_ALLOC_ERROR) {
+ spin_lock_irqsave(&phba->ndlp_lock, flags);
+ NLP_CLR_NODE_ACT(ndlp);
+ spin_unlock_irqrestore(&phba->ndlp_lock, flags);
+ continue;
}
+ ndlp->nlp_rpi = rpi;
+ lpfc_printf_vlog(ndlp->vport, KERN_INFO, LOG_NODE,
+ "0009 rpi:%x DID:%x "
+ "flg:%x map:%x %p\n", ndlp->nlp_rpi,
+ ndlp->nlp_DID, ndlp->nlp_flag,
+ ndlp->nlp_usg_map, ndlp);
}
}
lpfc_destroy_vport_work_array(phba, vports);
@@ -4722,10 +4726,9 @@ lpfc_sli4_perform_vport_cvl(struct lpfc_vport *vport)
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, so allocate a new one */
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, Fabric_DID);
if (!ndlp)
return 0;
- lpfc_nlp_init(vport, ndlp, Fabric_DID);
/* Set the node type */
ndlp->nlp_type |= NLP_FABRIC;
/* Put ndlp onto node list */
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index 1519fdf..a6adc2b 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -16542,14 +16542,13 @@ lpfc_sli4_seq_abort_rsp(struct lpfc_vport *vport,
ndlp = lpfc_findnode_did(vport, sid);
if (!ndlp) {
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, sid);
if (!ndlp) {
lpfc_printf_vlog(vport, KERN_WARNING, LOG_ELS,
"1268 Failed to allocate ndlp for "
"oxid:x%x SID:x%x\n", oxid, sid);
return;
}
- lpfc_nlp_init(vport, ndlp, sid);
/* Put ndlp onto pport node list */
lpfc_enqueue_node(vport, ndlp);
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
diff --git a/drivers/scsi/lpfc/lpfc_vport.c b/drivers/scsi/lpfc/lpfc_vport.c
index 9a0339d..c714482 100644
--- a/drivers/scsi/lpfc/lpfc_vport.c
+++ b/drivers/scsi/lpfc/lpfc_vport.c
@@ -738,10 +738,9 @@ lpfc_vport_delete(struct fc_vport *fc_vport)
ndlp = lpfc_findnode_did(vport, Fabric_DID);
if (!ndlp) {
/* Cannot find existing Fabric ndlp, allocate one */
- ndlp = mempool_alloc(phba->nlp_mem_pool, GFP_KERNEL);
+ ndlp = lpfc_nlp_init(vport, Fabric_DID);
if (!ndlp)
goto skip_logo;
- lpfc_nlp_init(vport, ndlp, Fabric_DID);
/* Indicate free memory when release */
NLP_SET_FREE_REQ(ndlp);
} else {
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 16/22] Fix crash after issuing lip reset
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:42 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:42 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:35PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> When RPI is not available, driver sends WQE with invalid RPI value and
> rejected by HBA.
> lpfc 0000:82:00.3: 1:3154 BLS ABORT RSP failed, data: x3/xa0320008
> and
> lpfc :2753 PLOGI failure DID:FFFFFA Status:x3/xa0240008
>
> In this case, driver accesses rpi_ids array out of bounds.
>
> Fix:
> Check return value of lpfc_sli4_alloc_rpi(). Do not allocate
> lpfc_nodelist entry if RPI is not available.
>
> When RPI is not available, we will get discovery timeouts and
> command drops for some of the vports as seen below.
>
>
> lpfc :0273 Unexpected discovery timeout, vport State x0
> lpfc :0230 Unexpected timeout, hba link state x5
> lpfc :0111 Dropping received ELS cmd Data: x0 xc90c55 x0
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 16/22] Fix crash after issuing lip reset
@ 2017-04-20 7:42 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:42 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:35PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> When RPI is not available, driver sends WQE with invalid RPI value and
> rejected by HBA.
> lpfc 0000:82:00.3: 1:3154 BLS ABORT RSP failed, data: x3/xa0320008
> and
> lpfc :2753 PLOGI failure DID:FFFFFA Status:x3/xa0240008
>
> In this case, driver accesses rpi_ids array out of bounds.
>
> Fix:
> Check return value of lpfc_sli4_alloc_rpi(). Do not allocate
> lpfc_nodelist entry if RPI is not available.
>
> When RPI is not available, we will get discovery timeouts and
> command drops for some of the vports as seen below.
>
>
> lpfc :0273 Unexpected discovery timeout, vport State x0
> lpfc :0230 Unexpected timeout, hba link state x5
> lpfc :0111 Dropping received ELS cmd Data: x0 xc90c55 x0
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 17/22] Fix max_sgl_segments settings for NVME / NVMET
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
Cannot set NVME segment counts to a large number
The existing module parameter lpfc_sg_seg_cnt is used for both
SCSI and NVME.
Limit the module parameter lpfc_sg_seg_cnt to 128 with the
default being 64 for both NVME and NVMET, assuming NVME is enabled in the
driver for that port. The driver will set max_sgl_segments in the
NVME/NVMET template to lpfc_sg_seg_cnt + 1.
Looks good,
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc.h | 3 ++-
drivers/scsi/lpfc/lpfc_nvme.c | 18 ++++++++++++++----
drivers/scsi/lpfc/lpfc_nvmet.c | 19 +++++++++++++++----
3 files changed, 31 insertions(+), 9 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc.h b/drivers/scsi/lpfc/lpfc.h
index 257bbdd..0fc1450 100644
--- a/drivers/scsi/lpfc/lpfc.h
+++ b/drivers/scsi/lpfc/lpfc.h
@@ -56,7 +56,7 @@ struct lpfc_sli2_slim;
#define LPFC_MAX_SG_SEG_CNT 4096 /* sg element count per scsi cmnd */
#define LPFC_MAX_SGL_SEG_CNT 512 /* SGL element count per scsi cmnd */
#define LPFC_MAX_BPL_SEG_CNT 4096 /* BPL element count per scsi cmnd */
-#define LPFC_MIN_NVME_SEG_CNT 254
+#define LPFC_MAX_NVME_SEG_CNT 128 /* max SGL element cnt per NVME cmnd */
#define LPFC_MAX_SGE_SIZE 0x80000000 /* Maximum data allowed in a SGE */
#define LPFC_IOCB_LIST_CNT 2250 /* list of IOCBs for fast-path usage. */
@@ -781,6 +781,7 @@ struct lpfc_hba {
uint32_t cfg_nvmet_fb_size;
uint32_t cfg_total_seg_cnt;
uint32_t cfg_sg_seg_cnt;
+ uint32_t cfg_nvme_seg_cnt;
uint32_t cfg_sg_dma_buf_size;
uint64_t cfg_soft_wwnn;
uint64_t cfg_soft_wwpn;
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index efe1242..1a83ade 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -1113,12 +1113,12 @@ lpfc_nvme_prep_io_dma(struct lpfc_vport *vport,
first_data_sgl = sgl;
lpfc_ncmd->seg_cnt = nCmd->sg_cnt;
- if (lpfc_ncmd->seg_cnt > phba->cfg_sg_seg_cnt) {
+ if (lpfc_ncmd->seg_cnt > phba->cfg_nvme_seg_cnt) {
lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
"6058 Too many sg segments from "
"NVME Transport. Max %d, "
"nvmeIO sg_cnt %d\n",
- phba->cfg_sg_seg_cnt,
+ phba->cfg_nvme_seg_cnt,
lpfc_ncmd->seg_cnt);
lpfc_ncmd->seg_cnt = 0;
return 1;
@@ -2157,8 +2157,18 @@ lpfc_nvme_create_localport(struct lpfc_vport *vport)
nfcp_info.node_name = wwn_to_u64(vport->fc_nodename.u.wwn);
nfcp_info.port_name = wwn_to_u64(vport->fc_portname.u.wwn);
- /* For now need + 1 to get around NVME transport logic */
- lpfc_nvme_template.max_sgl_segments = phba->cfg_sg_seg_cnt + 1;
+ /* Limit to LPFC_MAX_NVME_SEG_CNT.
+ * For now need + 1 to get around NVME transport logic.
+ */
+ if (phba->cfg_sg_seg_cnt > LPFC_MAX_NVME_SEG_CNT) {
+ lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME | LOG_INIT,
+ "6300 Reducing sg segment cnt to %d\n",
+ LPFC_MAX_NVME_SEG_CNT);
+ phba->cfg_nvme_seg_cnt = LPFC_MAX_NVME_SEG_CNT;
+ } else {
+ phba->cfg_nvme_seg_cnt = phba->cfg_sg_seg_cnt;
+ }
+ lpfc_nvme_template.max_sgl_segments = phba->cfg_nvme_seg_cnt + 1;
lpfc_nvme_template.max_hw_queues = phba->cfg_nvme_io_channel;
/* localport is allocated from the stack, but the registration
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index 8348bb5..d0dda42 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -704,8 +704,19 @@ lpfc_nvmet_create_targetport(struct lpfc_hba *phba)
pinfo.port_name = wwn_to_u64(vport->fc_portname.u.wwn);
pinfo.port_id = vport->fc_myDID;
+ /* Limit to LPFC_MAX_NVME_SEG_CNT.
+ * For now need + 1 to get around NVME transport logic.
+ */
+ if (phba->cfg_sg_seg_cnt > LPFC_MAX_NVME_SEG_CNT) {
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME | LOG_INIT,
+ "6400 Reducing sg segment cnt to %d\n",
+ LPFC_MAX_NVME_SEG_CNT);
+ phba->cfg_nvme_seg_cnt = LPFC_MAX_NVME_SEG_CNT;
+ } else {
+ phba->cfg_nvme_seg_cnt = phba->cfg_sg_seg_cnt;
+ }
+ lpfc_tgttemplate.max_sgl_segments = phba->cfg_nvme_seg_cnt + 1;
lpfc_tgttemplate.max_hw_queues = phba->cfg_nvme_io_channel;
- lpfc_tgttemplate.max_sgl_segments = phba->cfg_sg_seg_cnt + 1;
lpfc_tgttemplate.target_features = NVMET_FCTGTFEAT_READDATA_RSP |
NVMET_FCTGTFEAT_NEEDS_CMD_CPUSCHED |
NVMET_FCTGTFEAT_CMD_IN_ISR |
@@ -1278,11 +1289,11 @@ lpfc_nvmet_prep_fcp_wqe(struct lpfc_hba *phba,
return NULL;
}
- if (rsp->sg_cnt > phba->cfg_sg_seg_cnt) {
+ if (rsp->sg_cnt > phba->cfg_nvme_seg_cnt) {
lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
"6109 lpfc_nvmet_prep_fcp_wqe: seg cnt err: "
- "NPORT x%x oxid:x%x\n",
- ctxp->sid, ctxp->oxid);
+ "NPORT x%x oxid:x%x cnt %d\n",
+ ctxp->sid, ctxp->oxid, phba->cfg_nvme_seg_cnt);
return NULL;
}
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 17/22] Fix max_sgl_segments settings for NVME / NVMET
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
Cannot set NVME segment counts to a large number
The existing module parameter lpfc_sg_seg_cnt is used for both
SCSI and NVME.
Limit the module parameter lpfc_sg_seg_cnt to 128 with the
default being 64 for both NVME and NVMET, assuming NVME is enabled in the
driver for that port. The driver will set max_sgl_segments in the
NVME/NVMET template to lpfc_sg_seg_cnt + 1.
Looks good,
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc.h | 3 ++-
drivers/scsi/lpfc/lpfc_nvme.c | 18 ++++++++++++++----
drivers/scsi/lpfc/lpfc_nvmet.c | 19 +++++++++++++++----
3 files changed, 31 insertions(+), 9 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc.h b/drivers/scsi/lpfc/lpfc.h
index 257bbdd..0fc1450 100644
--- a/drivers/scsi/lpfc/lpfc.h
+++ b/drivers/scsi/lpfc/lpfc.h
@@ -56,7 +56,7 @@ struct lpfc_sli2_slim;
#define LPFC_MAX_SG_SEG_CNT 4096 /* sg element count per scsi cmnd */
#define LPFC_MAX_SGL_SEG_CNT 512 /* SGL element count per scsi cmnd */
#define LPFC_MAX_BPL_SEG_CNT 4096 /* BPL element count per scsi cmnd */
-#define LPFC_MIN_NVME_SEG_CNT 254
+#define LPFC_MAX_NVME_SEG_CNT 128 /* max SGL element cnt per NVME cmnd */
#define LPFC_MAX_SGE_SIZE 0x80000000 /* Maximum data allowed in a SGE */
#define LPFC_IOCB_LIST_CNT 2250 /* list of IOCBs for fast-path usage. */
@@ -781,6 +781,7 @@ struct lpfc_hba {
uint32_t cfg_nvmet_fb_size;
uint32_t cfg_total_seg_cnt;
uint32_t cfg_sg_seg_cnt;
+ uint32_t cfg_nvme_seg_cnt;
uint32_t cfg_sg_dma_buf_size;
uint64_t cfg_soft_wwnn;
uint64_t cfg_soft_wwpn;
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index efe1242..1a83ade 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -1113,12 +1113,12 @@ lpfc_nvme_prep_io_dma(struct lpfc_vport *vport,
first_data_sgl = sgl;
lpfc_ncmd->seg_cnt = nCmd->sg_cnt;
- if (lpfc_ncmd->seg_cnt > phba->cfg_sg_seg_cnt) {
+ if (lpfc_ncmd->seg_cnt > phba->cfg_nvme_seg_cnt) {
lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
"6058 Too many sg segments from "
"NVME Transport. Max %d, "
"nvmeIO sg_cnt %d\n",
- phba->cfg_sg_seg_cnt,
+ phba->cfg_nvme_seg_cnt,
lpfc_ncmd->seg_cnt);
lpfc_ncmd->seg_cnt = 0;
return 1;
@@ -2157,8 +2157,18 @@ lpfc_nvme_create_localport(struct lpfc_vport *vport)
nfcp_info.node_name = wwn_to_u64(vport->fc_nodename.u.wwn);
nfcp_info.port_name = wwn_to_u64(vport->fc_portname.u.wwn);
- /* For now need + 1 to get around NVME transport logic */
- lpfc_nvme_template.max_sgl_segments = phba->cfg_sg_seg_cnt + 1;
+ /* Limit to LPFC_MAX_NVME_SEG_CNT.
+ * For now need + 1 to get around NVME transport logic.
+ */
+ if (phba->cfg_sg_seg_cnt > LPFC_MAX_NVME_SEG_CNT) {
+ lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME | LOG_INIT,
+ "6300 Reducing sg segment cnt to %d\n",
+ LPFC_MAX_NVME_SEG_CNT);
+ phba->cfg_nvme_seg_cnt = LPFC_MAX_NVME_SEG_CNT;
+ } else {
+ phba->cfg_nvme_seg_cnt = phba->cfg_sg_seg_cnt;
+ }
+ lpfc_nvme_template.max_sgl_segments = phba->cfg_nvme_seg_cnt + 1;
lpfc_nvme_template.max_hw_queues = phba->cfg_nvme_io_channel;
/* localport is allocated from the stack, but the registration
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index 8348bb5..d0dda42 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -704,8 +704,19 @@ lpfc_nvmet_create_targetport(struct lpfc_hba *phba)
pinfo.port_name = wwn_to_u64(vport->fc_portname.u.wwn);
pinfo.port_id = vport->fc_myDID;
+ /* Limit to LPFC_MAX_NVME_SEG_CNT.
+ * For now need + 1 to get around NVME transport logic.
+ */
+ if (phba->cfg_sg_seg_cnt > LPFC_MAX_NVME_SEG_CNT) {
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME | LOG_INIT,
+ "6400 Reducing sg segment cnt to %d\n",
+ LPFC_MAX_NVME_SEG_CNT);
+ phba->cfg_nvme_seg_cnt = LPFC_MAX_NVME_SEG_CNT;
+ } else {
+ phba->cfg_nvme_seg_cnt = phba->cfg_sg_seg_cnt;
+ }
+ lpfc_tgttemplate.max_sgl_segments = phba->cfg_nvme_seg_cnt + 1;
lpfc_tgttemplate.max_hw_queues = phba->cfg_nvme_io_channel;
- lpfc_tgttemplate.max_sgl_segments = phba->cfg_sg_seg_cnt + 1;
lpfc_tgttemplate.target_features = NVMET_FCTGTFEAT_READDATA_RSP |
NVMET_FCTGTFEAT_NEEDS_CMD_CPUSCHED |
NVMET_FCTGTFEAT_CMD_IN_ISR |
@@ -1278,11 +1289,11 @@ lpfc_nvmet_prep_fcp_wqe(struct lpfc_hba *phba,
return NULL;
}
- if (rsp->sg_cnt > phba->cfg_sg_seg_cnt) {
+ if (rsp->sg_cnt > phba->cfg_nvme_seg_cnt) {
lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
"6109 lpfc_nvmet_prep_fcp_wqe: seg cnt err: "
- "NPORT x%x oxid:x%x\n",
- ctxp->sid, ctxp->oxid);
+ "NPORT x%x oxid:x%x cnt %d\n",
+ ctxp->sid, ctxp->oxid, phba->cfg_nvme_seg_cnt);
return NULL;
}
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH-v1 18/22] Add Fabric assigned WWN support.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
Adding support for Fabric assigned WWPN and WWNN.
Firmware sends first FLOGI to fabric with vendor version changes.
On link up driver gets updated service parameter with FAWWN assigned port
name. Driver sends 2nd FLOGI with updated fawwpn and modifies the
vport->fc_portname in driver.
Note:
Soft wwpn will not be allowed when fawwpn is enabled.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc.h | 2 ++
drivers/scsi/lpfc/lpfc_attr.c | 8 ++++++++
drivers/scsi/lpfc/lpfc_els.c | 8 +++++---
drivers/scsi/lpfc/lpfc_hbadisc.c | 22 +++++++++++++++++-----
drivers/scsi/lpfc/lpfc_hw.h | 3 +++
drivers/scsi/lpfc/lpfc_hw4.h | 1 +
drivers/scsi/lpfc/lpfc_init.c | 31 ++++++++++++++++++++++++++++---
7 files changed, 64 insertions(+), 11 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc.h b/drivers/scsi/lpfc/lpfc.h
index 0fc1450..6d7840b 100644
--- a/drivers/scsi/lpfc/lpfc.h
+++ b/drivers/scsi/lpfc/lpfc.h
@@ -474,6 +474,8 @@ struct lpfc_vport {
unsigned long rcv_buffer_time_stamp;
uint32_t vport_flag;
#define STATIC_VPORT 1
+#define FAWWPN_SET 2
+#define FAWWPN_PARAM_CHG 4
uint16_t fdmi_num_disc;
uint32_t fdmi_hba_mask;
diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 4b32d02..e9d956d 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -2292,6 +2292,8 @@ lpfc_soft_wwn_enable_store(struct device *dev, struct device_attribute *attr,
struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
struct lpfc_hba *phba = vport->phba;
unsigned int cnt = count;
+ uint8_t vvvl = vport->fc_sparam.cmn.valid_vendor_ver_level;
+ u32 *fawwpn_key = (uint32_t *)&vport->fc_sparam.un.vendorVersion[0];
/*
* We're doing a simple sanity check for soft_wwpn setting.
@@ -2305,6 +2307,12 @@ lpfc_soft_wwn_enable_store(struct device *dev, struct device_attribute *attr,
* here. The intent is to protect against the random user or
* application that is just writing attributes.
*/
+ if (vvvl == 1 && (cpu_to_be32(*fawwpn_key)) == FAPWWN_KEY_VENDOR) {
+ lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
+ "0051 "LPFC_DRIVER_NAME" soft wwpn can not"
+ " be enabled: fawwpn is enabled\n");
+ return -EINVAL;
+ }
/* count may include a LF at end of string */
if (buf[cnt-1] == '\n')
diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index ffdcac0..35fc260 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -603,9 +603,11 @@ lpfc_check_clean_addr_bit(struct lpfc_vport *vport,
memcmp(&vport->fabric_portname, &sp->portName,
sizeof(struct lpfc_name)) ||
memcmp(&vport->fabric_nodename, &sp->nodeName,
- sizeof(struct lpfc_name)))
+ sizeof(struct lpfc_name)) ||
+ (vport->vport_flag & FAWWPN_PARAM_CHG)) {
fabric_param_changed = 1;
-
+ vport->vport_flag &= ~FAWWPN_PARAM_CHG;
+ }
/*
* Word 1 Bit 31 in common service parameter is overloaded.
* Word 1 Bit 31 in FLOGI request is multiple NPort request
@@ -8755,7 +8757,7 @@ lpfc_issue_els_fdisc(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
pcmd += sizeof(uint32_t); /* Node Name */
pcmd += sizeof(uint32_t); /* Node Name */
memcpy(pcmd, &vport->fc_nodename, 8);
-
+ memset(sp->un.vendorVersion, 0, sizeof(sp->un.vendorVersion));
lpfc_set_disctmo(vport);
phba->fc_stat.elsXmitFDISC++;
diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c
index d313dde..e824ec1 100644
--- a/drivers/scsi/lpfc/lpfc_hbadisc.c
+++ b/drivers/scsi/lpfc/lpfc_hbadisc.c
@@ -3002,6 +3002,7 @@ lpfc_mbx_cmpl_read_sparam(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
MAILBOX_t *mb = &pmb->u.mb;
struct lpfc_dmabuf *mp = (struct lpfc_dmabuf *) pmb->context1;
struct lpfc_vport *vport = pmb->vport;
+ struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct serv_parm *sp = &vport->fc_sparam;
uint32_t ed_tov;
@@ -3031,6 +3032,7 @@ lpfc_mbx_cmpl_read_sparam(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
}
lpfc_update_vport_wwn(vport);
+ fc_host_port_name(shost) = wwn_to_u64(vport->fc_portname.u.wwn);
if (vport->port_type == LPFC_PHYSICAL_PORT) {
memcpy(&phba->wwnn, &vport->fc_nodename, sizeof(phba->wwnn));
memcpy(&phba->wwpn, &vport->fc_portname, sizeof(phba->wwnn));
@@ -3309,6 +3311,7 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
struct lpfc_sli_ring *pring;
MAILBOX_t *mb = &pmb->u.mb;
struct lpfc_dmabuf *mp = (struct lpfc_dmabuf *) (pmb->context1);
+ uint8_t attn_type;
/* Unblock ELS traffic */
pring = lpfc_phba_elsring(phba);
@@ -3325,6 +3328,7 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
}
la = (struct lpfc_mbx_read_top *) &pmb->u.mb.un.varReadTop;
+ attn_type = bf_get(lpfc_mbx_read_top_att_type, la);
memcpy(&phba->alpa_map[0], mp->virt, 128);
@@ -3337,7 +3341,7 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
if (phba->fc_eventTag <= la->eventTag) {
phba->fc_stat.LinkMultiEvent++;
- if (bf_get(lpfc_mbx_read_top_att_type, la) == LPFC_ATT_LINK_UP)
+ if (attn_type == LPFC_ATT_LINK_UP)
if (phba->fc_eventTag != 0)
lpfc_linkdown(phba);
}
@@ -3353,7 +3357,7 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
}
phba->link_events++;
- if ((bf_get(lpfc_mbx_read_top_att_type, la) == LPFC_ATT_LINK_UP) &&
+ if ((attn_type == LPFC_ATT_LINK_UP) &&
!(phba->sli.sli_flag & LPFC_MENLO_MAINT)) {
phba->fc_stat.LinkUp++;
if (phba->link_flag & LS_LOOPBACK_MODE) {
@@ -3379,8 +3383,8 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
phba->wait_4_mlo_maint_flg);
}
lpfc_mbx_process_link_up(phba, la);
- } else if (bf_get(lpfc_mbx_read_top_att_type, la) ==
- LPFC_ATT_LINK_DOWN) {
+ } else if ((attn_type == LPFC_ATT_LINK_DOWN) ||
+ (attn_type == LPFC_ATT_UNEXP_WWPN)) {
phba->fc_stat.LinkDown++;
if (phba->link_flag & LS_LOOPBACK_MODE)
lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
@@ -3389,6 +3393,14 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
"Data: x%x x%x x%x\n",
la->eventTag, phba->fc_eventTag,
phba->pport->port_state, vport->fc_flag);
+ else if (attn_type == LPFC_ATT_UNEXP_WWPN)
+ lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
+ "1313 Link Down UNEXP WWPN Event x%x received "
+ "Data: x%x x%x x%x x%x x%x\n",
+ la->eventTag, phba->fc_eventTag,
+ phba->pport->port_state, vport->fc_flag,
+ bf_get(lpfc_mbx_read_top_mm, la),
+ bf_get(lpfc_mbx_read_top_fa, la));
else
lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
"1305 Link Down Event x%x received "
@@ -3400,7 +3412,7 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
lpfc_mbx_issue_link_down(phba);
}
if ((phba->sli.sli_flag & LPFC_MENLO_MAINT) &&
- ((bf_get(lpfc_mbx_read_top_att_type, la) == LPFC_ATT_LINK_UP))) {
+ (attn_type == LPFC_ATT_LINK_UP)) {
if (phba->link_state != LPFC_LINK_DOWN) {
phba->fc_stat.LinkDown++;
lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
diff --git a/drivers/scsi/lpfc/lpfc_hw.h b/drivers/scsi/lpfc/lpfc_hw.h
index 15ca214..26a5647 100644
--- a/drivers/scsi/lpfc/lpfc_hw.h
+++ b/drivers/scsi/lpfc/lpfc_hw.h
@@ -509,6 +509,8 @@ struct class_parms {
uint8_t word3Reserved2; /* Fc Word 3, bit 0: 7 */
};
+#define FAPWWN_KEY_VENDOR 0x42524344 /*valid vendor version fawwpn key*/
+
struct serv_parm { /* Structure is in Big Endian format */
struct csp cmn;
struct lpfc_name portName;
@@ -2885,6 +2887,7 @@ struct lpfc_mbx_read_top {
#define LPFC_ATT_RESERVED 0x00 /* Reserved - attType */
#define LPFC_ATT_LINK_UP 0x01 /* Link is up */
#define LPFC_ATT_LINK_DOWN 0x02 /* Link is down */
+#define LPFC_ATT_UNEXP_WWPN 0x06 /* Link is down Unexpected WWWPN */
uint32_t word3;
#define lpfc_mbx_read_top_alpa_granted_SHIFT 24
#define lpfc_mbx_read_top_alpa_granted_MASK 0x000000FF
diff --git a/drivers/scsi/lpfc/lpfc_hw4.h b/drivers/scsi/lpfc/lpfc_hw4.h
index 1527770..90499f9 100644
--- a/drivers/scsi/lpfc/lpfc_hw4.h
+++ b/drivers/scsi/lpfc/lpfc_hw4.h
@@ -3853,6 +3853,7 @@ struct lpfc_acqe_fc_la {
#define LPFC_FC_LA_TYPE_NO_HARD_ALPA 0x3
#define LPFC_FC_LA_TYPE_MDS_LINK_DOWN 0x4
#define LPFC_FC_LA_TYPE_MDS_LOOPBACK 0x5
+#define LPFC_FC_LA_TYPE_UNEXP_WWPN 0x6
#define lpfc_acqe_fc_la_port_type_SHIFT 6
#define lpfc_acqe_fc_la_port_type_MASK 0x00000003
#define lpfc_acqe_fc_la_port_type_WORD word0
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 61bf14c..8acacdf 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -335,6 +335,8 @@ lpfc_dump_wakeup_param_cmpl(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmboxq)
void
lpfc_update_vport_wwn(struct lpfc_vport *vport)
{
+ uint8_t vvvl = vport->fc_sparam.cmn.valid_vendor_ver_level;
+ u32 *fawwpn_key = (u32 *)&vport->fc_sparam.un.vendorVersion[0];
/* If the soft name exists then update it using the service params */
if (vport->phba->cfg_soft_wwnn)
u64_to_wwn(vport->phba->cfg_soft_wwnn,
@@ -354,9 +356,26 @@ lpfc_update_vport_wwn(struct lpfc_vport *vport)
memcpy(&vport->fc_sparam.nodeName, &vport->fc_nodename,
sizeof(struct lpfc_name));
- if (vport->fc_portname.u.wwn[0] == 0 || vport->phba->cfg_soft_wwpn)
+ /*
+ * If the port name has changed, then set the Param changes flag
+ * to unreg the login
+ */
+ if ((vport->fc_portname.u.wwn[0] != 0) &&
+ memcmp(&vport->fc_portname, &vport->fc_sparam.portName,
+ sizeof(struct lpfc_name)))
+ vport->vport_flag |= FAWWPN_PARAM_CHG;
+
+ if ((vport->fc_portname.u.wwn[0] == 0) ||
+ (vport->phba->cfg_soft_wwpn) ||
+ (vvvl == 1 && (cpu_to_be32(*fawwpn_key)) == FAPWWN_KEY_VENDOR) ||
+ (vport->vport_flag & FAWWPN_SET)) {
memcpy(&vport->fc_portname, &vport->fc_sparam.portName,
sizeof(struct lpfc_name));
+ vport->vport_flag &= ~FAWWPN_SET;
+ if (vvvl == 1 && (cpu_to_be32(*fawwpn_key)) ==
+ FAPWWN_KEY_VENDOR)
+ vport->vport_flag |= FAWWPN_SET;
+ }
else
memcpy(&vport->fc_sparam.portName, &vport->fc_portname,
sizeof(struct lpfc_name));
@@ -4518,9 +4537,15 @@ lpfc_sli4_async_fc_evt(struct lpfc_hba *phba, struct lpfc_acqe_fc_la *acqe_fc)
/* Parse and translate link attention fields */
la = (struct lpfc_mbx_read_top *)&pmb->u.mb.un.varReadTop;
la->eventTag = acqe_fc->event_tag;
- bf_set(lpfc_mbx_read_top_att_type, la,
- LPFC_FC_LA_TYPE_LINK_DOWN);
+ if (phba->sli4_hba.link_state.status ==
+ LPFC_FC_LA_TYPE_UNEXP_WWPN) {
+ bf_set(lpfc_mbx_read_top_att_type, la,
+ LPFC_FC_LA_TYPE_UNEXP_WWPN);
+ } else {
+ bf_set(lpfc_mbx_read_top_att_type, la,
+ LPFC_FC_LA_TYPE_LINK_DOWN);
+ }
/* Invoke the mailbox command callback function */
lpfc_mbx_cmpl_read_topology(phba, pmb);
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 18/22] Add Fabric assigned WWN support.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
Adding support for Fabric assigned WWPN and WWNN.
Firmware sends first FLOGI to fabric with vendor version changes.
On link up driver gets updated service parameter with FAWWN assigned port
name. Driver sends 2nd FLOGI with updated fawwpn and modifies the
vport->fc_portname in driver.
Note:
Soft wwpn will not be allowed when fawwpn is enabled.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc.h | 2 ++
drivers/scsi/lpfc/lpfc_attr.c | 8 ++++++++
drivers/scsi/lpfc/lpfc_els.c | 8 +++++---
drivers/scsi/lpfc/lpfc_hbadisc.c | 22 +++++++++++++++++-----
drivers/scsi/lpfc/lpfc_hw.h | 3 +++
drivers/scsi/lpfc/lpfc_hw4.h | 1 +
drivers/scsi/lpfc/lpfc_init.c | 31 ++++++++++++++++++++++++++++---
7 files changed, 64 insertions(+), 11 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc.h b/drivers/scsi/lpfc/lpfc.h
index 0fc1450..6d7840b 100644
--- a/drivers/scsi/lpfc/lpfc.h
+++ b/drivers/scsi/lpfc/lpfc.h
@@ -474,6 +474,8 @@ struct lpfc_vport {
unsigned long rcv_buffer_time_stamp;
uint32_t vport_flag;
#define STATIC_VPORT 1
+#define FAWWPN_SET 2
+#define FAWWPN_PARAM_CHG 4
uint16_t fdmi_num_disc;
uint32_t fdmi_hba_mask;
diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 4b32d02..e9d956d 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -2292,6 +2292,8 @@ lpfc_soft_wwn_enable_store(struct device *dev, struct device_attribute *attr,
struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
struct lpfc_hba *phba = vport->phba;
unsigned int cnt = count;
+ uint8_t vvvl = vport->fc_sparam.cmn.valid_vendor_ver_level;
+ u32 *fawwpn_key = (uint32_t *)&vport->fc_sparam.un.vendorVersion[0];
/*
* We're doing a simple sanity check for soft_wwpn setting.
@@ -2305,6 +2307,12 @@ lpfc_soft_wwn_enable_store(struct device *dev, struct device_attribute *attr,
* here. The intent is to protect against the random user or
* application that is just writing attributes.
*/
+ if (vvvl == 1 && (cpu_to_be32(*fawwpn_key)) == FAPWWN_KEY_VENDOR) {
+ lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
+ "0051 "LPFC_DRIVER_NAME" soft wwpn can not"
+ " be enabled: fawwpn is enabled\n");
+ return -EINVAL;
+ }
/* count may include a LF at end of string */
if (buf[cnt-1] == '\n')
diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index ffdcac0..35fc260 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -603,9 +603,11 @@ lpfc_check_clean_addr_bit(struct lpfc_vport *vport,
memcmp(&vport->fabric_portname, &sp->portName,
sizeof(struct lpfc_name)) ||
memcmp(&vport->fabric_nodename, &sp->nodeName,
- sizeof(struct lpfc_name)))
+ sizeof(struct lpfc_name)) ||
+ (vport->vport_flag & FAWWPN_PARAM_CHG)) {
fabric_param_changed = 1;
-
+ vport->vport_flag &= ~FAWWPN_PARAM_CHG;
+ }
/*
* Word 1 Bit 31 in common service parameter is overloaded.
* Word 1 Bit 31 in FLOGI request is multiple NPort request
@@ -8755,7 +8757,7 @@ lpfc_issue_els_fdisc(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
pcmd += sizeof(uint32_t); /* Node Name */
pcmd += sizeof(uint32_t); /* Node Name */
memcpy(pcmd, &vport->fc_nodename, 8);
-
+ memset(sp->un.vendorVersion, 0, sizeof(sp->un.vendorVersion));
lpfc_set_disctmo(vport);
phba->fc_stat.elsXmitFDISC++;
diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c
index d313dde..e824ec1 100644
--- a/drivers/scsi/lpfc/lpfc_hbadisc.c
+++ b/drivers/scsi/lpfc/lpfc_hbadisc.c
@@ -3002,6 +3002,7 @@ lpfc_mbx_cmpl_read_sparam(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
MAILBOX_t *mb = &pmb->u.mb;
struct lpfc_dmabuf *mp = (struct lpfc_dmabuf *) pmb->context1;
struct lpfc_vport *vport = pmb->vport;
+ struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
struct serv_parm *sp = &vport->fc_sparam;
uint32_t ed_tov;
@@ -3031,6 +3032,7 @@ lpfc_mbx_cmpl_read_sparam(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
}
lpfc_update_vport_wwn(vport);
+ fc_host_port_name(shost) = wwn_to_u64(vport->fc_portname.u.wwn);
if (vport->port_type == LPFC_PHYSICAL_PORT) {
memcpy(&phba->wwnn, &vport->fc_nodename, sizeof(phba->wwnn));
memcpy(&phba->wwpn, &vport->fc_portname, sizeof(phba->wwnn));
@@ -3309,6 +3311,7 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
struct lpfc_sli_ring *pring;
MAILBOX_t *mb = &pmb->u.mb;
struct lpfc_dmabuf *mp = (struct lpfc_dmabuf *) (pmb->context1);
+ uint8_t attn_type;
/* Unblock ELS traffic */
pring = lpfc_phba_elsring(phba);
@@ -3325,6 +3328,7 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
}
la = (struct lpfc_mbx_read_top *) &pmb->u.mb.un.varReadTop;
+ attn_type = bf_get(lpfc_mbx_read_top_att_type, la);
memcpy(&phba->alpa_map[0], mp->virt, 128);
@@ -3337,7 +3341,7 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
if (phba->fc_eventTag <= la->eventTag) {
phba->fc_stat.LinkMultiEvent++;
- if (bf_get(lpfc_mbx_read_top_att_type, la) == LPFC_ATT_LINK_UP)
+ if (attn_type == LPFC_ATT_LINK_UP)
if (phba->fc_eventTag != 0)
lpfc_linkdown(phba);
}
@@ -3353,7 +3357,7 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
}
phba->link_events++;
- if ((bf_get(lpfc_mbx_read_top_att_type, la) == LPFC_ATT_LINK_UP) &&
+ if ((attn_type == LPFC_ATT_LINK_UP) &&
!(phba->sli.sli_flag & LPFC_MENLO_MAINT)) {
phba->fc_stat.LinkUp++;
if (phba->link_flag & LS_LOOPBACK_MODE) {
@@ -3379,8 +3383,8 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
phba->wait_4_mlo_maint_flg);
}
lpfc_mbx_process_link_up(phba, la);
- } else if (bf_get(lpfc_mbx_read_top_att_type, la) ==
- LPFC_ATT_LINK_DOWN) {
+ } else if ((attn_type == LPFC_ATT_LINK_DOWN) ||
+ (attn_type == LPFC_ATT_UNEXP_WWPN)) {
phba->fc_stat.LinkDown++;
if (phba->link_flag & LS_LOOPBACK_MODE)
lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
@@ -3389,6 +3393,14 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
"Data: x%x x%x x%x\n",
la->eventTag, phba->fc_eventTag,
phba->pport->port_state, vport->fc_flag);
+ else if (attn_type == LPFC_ATT_UNEXP_WWPN)
+ lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
+ "1313 Link Down UNEXP WWPN Event x%x received "
+ "Data: x%x x%x x%x x%x x%x\n",
+ la->eventTag, phba->fc_eventTag,
+ phba->pport->port_state, vport->fc_flag,
+ bf_get(lpfc_mbx_read_top_mm, la),
+ bf_get(lpfc_mbx_read_top_fa, la));
else
lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
"1305 Link Down Event x%x received "
@@ -3400,7 +3412,7 @@ lpfc_mbx_cmpl_read_topology(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
lpfc_mbx_issue_link_down(phba);
}
if ((phba->sli.sli_flag & LPFC_MENLO_MAINT) &&
- ((bf_get(lpfc_mbx_read_top_att_type, la) == LPFC_ATT_LINK_UP))) {
+ (attn_type == LPFC_ATT_LINK_UP)) {
if (phba->link_state != LPFC_LINK_DOWN) {
phba->fc_stat.LinkDown++;
lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
diff --git a/drivers/scsi/lpfc/lpfc_hw.h b/drivers/scsi/lpfc/lpfc_hw.h
index 15ca214..26a5647 100644
--- a/drivers/scsi/lpfc/lpfc_hw.h
+++ b/drivers/scsi/lpfc/lpfc_hw.h
@@ -509,6 +509,8 @@ struct class_parms {
uint8_t word3Reserved2; /* Fc Word 3, bit 0: 7 */
};
+#define FAPWWN_KEY_VENDOR 0x42524344 /*valid vendor version fawwpn key*/
+
struct serv_parm { /* Structure is in Big Endian format */
struct csp cmn;
struct lpfc_name portName;
@@ -2885,6 +2887,7 @@ struct lpfc_mbx_read_top {
#define LPFC_ATT_RESERVED 0x00 /* Reserved - attType */
#define LPFC_ATT_LINK_UP 0x01 /* Link is up */
#define LPFC_ATT_LINK_DOWN 0x02 /* Link is down */
+#define LPFC_ATT_UNEXP_WWPN 0x06 /* Link is down Unexpected WWWPN */
uint32_t word3;
#define lpfc_mbx_read_top_alpa_granted_SHIFT 24
#define lpfc_mbx_read_top_alpa_granted_MASK 0x000000FF
diff --git a/drivers/scsi/lpfc/lpfc_hw4.h b/drivers/scsi/lpfc/lpfc_hw4.h
index 1527770..90499f9 100644
--- a/drivers/scsi/lpfc/lpfc_hw4.h
+++ b/drivers/scsi/lpfc/lpfc_hw4.h
@@ -3853,6 +3853,7 @@ struct lpfc_acqe_fc_la {
#define LPFC_FC_LA_TYPE_NO_HARD_ALPA 0x3
#define LPFC_FC_LA_TYPE_MDS_LINK_DOWN 0x4
#define LPFC_FC_LA_TYPE_MDS_LOOPBACK 0x5
+#define LPFC_FC_LA_TYPE_UNEXP_WWPN 0x6
#define lpfc_acqe_fc_la_port_type_SHIFT 6
#define lpfc_acqe_fc_la_port_type_MASK 0x00000003
#define lpfc_acqe_fc_la_port_type_WORD word0
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 61bf14c..8acacdf 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -335,6 +335,8 @@ lpfc_dump_wakeup_param_cmpl(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmboxq)
void
lpfc_update_vport_wwn(struct lpfc_vport *vport)
{
+ uint8_t vvvl = vport->fc_sparam.cmn.valid_vendor_ver_level;
+ u32 *fawwpn_key = (u32 *)&vport->fc_sparam.un.vendorVersion[0];
/* If the soft name exists then update it using the service params */
if (vport->phba->cfg_soft_wwnn)
u64_to_wwn(vport->phba->cfg_soft_wwnn,
@@ -354,9 +356,26 @@ lpfc_update_vport_wwn(struct lpfc_vport *vport)
memcpy(&vport->fc_sparam.nodeName, &vport->fc_nodename,
sizeof(struct lpfc_name));
- if (vport->fc_portname.u.wwn[0] == 0 || vport->phba->cfg_soft_wwpn)
+ /*
+ * If the port name has changed, then set the Param changes flag
+ * to unreg the login
+ */
+ if ((vport->fc_portname.u.wwn[0] != 0) &&
+ memcmp(&vport->fc_portname, &vport->fc_sparam.portName,
+ sizeof(struct lpfc_name)))
+ vport->vport_flag |= FAWWPN_PARAM_CHG;
+
+ if ((vport->fc_portname.u.wwn[0] == 0) ||
+ (vport->phba->cfg_soft_wwpn) ||
+ (vvvl == 1 && (cpu_to_be32(*fawwpn_key)) == FAPWWN_KEY_VENDOR) ||
+ (vport->vport_flag & FAWWPN_SET)) {
memcpy(&vport->fc_portname, &vport->fc_sparam.portName,
sizeof(struct lpfc_name));
+ vport->vport_flag &= ~FAWWPN_SET;
+ if (vvvl == 1 && (cpu_to_be32(*fawwpn_key)) ==
+ FAPWWN_KEY_VENDOR)
+ vport->vport_flag |= FAWWPN_SET;
+ }
else
memcpy(&vport->fc_sparam.portName, &vport->fc_portname,
sizeof(struct lpfc_name));
@@ -4518,9 +4537,15 @@ lpfc_sli4_async_fc_evt(struct lpfc_hba *phba, struct lpfc_acqe_fc_la *acqe_fc)
/* Parse and translate link attention fields */
la = (struct lpfc_mbx_read_top *)&pmb->u.mb.un.varReadTop;
la->eventTag = acqe_fc->event_tag;
- bf_set(lpfc_mbx_read_top_att_type, la,
- LPFC_FC_LA_TYPE_LINK_DOWN);
+ if (phba->sli4_hba.link_state.status ==
+ LPFC_FC_LA_TYPE_UNEXP_WWPN) {
+ bf_set(lpfc_mbx_read_top_att_type, la,
+ LPFC_FC_LA_TYPE_UNEXP_WWPN);
+ } else {
+ bf_set(lpfc_mbx_read_top_att_type, la,
+ LPFC_FC_LA_TYPE_LINK_DOWN);
+ }
/* Invoke the mailbox command callback function */
lpfc_mbx_cmpl_read_topology(phba, pmb);
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 18/22] Add Fabric assigned WWN support.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:49 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:49 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:37PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> Adding support for Fabric assigned WWPN and WWNN.
>
> Firmware sends first FLOGI to fabric with vendor version changes.
> On link up driver gets updated service parameter with FAWWN assigned port
> name. Driver sends 2nd FLOGI with updated fawwpn and modifies the
> vport->fc_portname in driver.
>
> Note:
> Soft wwpn will not be allowed when fawwpn is enabled.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Aparrt form the superflous parenthesis like in this hunk:
> @@ -354,9 +356,26 @@ lpfc_update_vport_wwn(struct lpfc_vport *vport)
> memcpy(&vport->fc_sparam.nodeName, &vport->fc_nodename,
> sizeof(struct lpfc_name));
>
> - if (vport->fc_portname.u.wwn[0] == 0 || vport->phba->cfg_soft_wwpn)
> + /*
> + * If the port name has changed, then set the Param changes flag
> + * to unreg the login
> + */
> + if ((vport->fc_portname.u.wwn[0] != 0) &&
> + memcmp(&vport->fc_portname, &vport->fc_sparam.portName,
> + sizeof(struct lpfc_name)))
> + vport->vport_flag |= FAWWPN_PARAM_CHG;
> +
> + if ((vport->fc_portname.u.wwn[0] == 0) ||
> + (vport->phba->cfg_soft_wwpn) ||
> + (vvvl == 1 && (cpu_to_be32(*fawwpn_key)) == FAPWWN_KEY_VENDOR) ||
> + (vport->vport_flag & FAWWPN_SET)) {
> memcpy(&vport->fc_portname, &vport->fc_sparam.portName,
> sizeof(struct lpfc_name));
> + vport->vport_flag &= ~FAWWPN_SET;
> + if (vvvl == 1 && (cpu_to_be32(*fawwpn_key)) ==
> + FAPWWN_KEY_VENDOR)
> + vport->vport_flag |= FAWWPN_SET;
> + }
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread* [PATCH-v1 18/22] Add Fabric assigned WWN support.
@ 2017-04-20 7:49 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:49 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:37PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> Adding support for Fabric assigned WWPN and WWNN.
>
> Firmware sends first FLOGI to fabric with vendor version changes.
> On link up driver gets updated service parameter with FAWWN assigned port
> name. Driver sends 2nd FLOGI with updated fawwpn and modifies the
> vport->fc_portname in driver.
>
> Note:
> Soft wwpn will not be allowed when fawwpn is enabled.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Aparrt form the superflous parenthesis like in this hunk:
> @@ -354,9 +356,26 @@ lpfc_update_vport_wwn(struct lpfc_vport *vport)
> memcpy(&vport->fc_sparam.nodeName, &vport->fc_nodename,
> sizeof(struct lpfc_name));
>
> - if (vport->fc_portname.u.wwn[0] == 0 || vport->phba->cfg_soft_wwpn)
> + /*
> + * If the port name has changed, then set the Param changes flag
> + * to unreg the login
> + */
> + if ((vport->fc_portname.u.wwn[0] != 0) &&
> + memcmp(&vport->fc_portname, &vport->fc_sparam.portName,
> + sizeof(struct lpfc_name)))
> + vport->vport_flag |= FAWWPN_PARAM_CHG;
> +
> + if ((vport->fc_portname.u.wwn[0] == 0) ||
> + (vport->phba->cfg_soft_wwpn) ||
> + (vvvl == 1 && (cpu_to_be32(*fawwpn_key)) == FAPWWN_KEY_VENDOR) ||
> + (vport->vport_flag & FAWWPN_SET)) {
> memcpy(&vport->fc_portname, &vport->fc_sparam.portName,
> sizeof(struct lpfc_name));
> + vport->vport_flag &= ~FAWWPN_SET;
> + if (vvvl == 1 && (cpu_to_be32(*fawwpn_key)) ==
> + FAPWWN_KEY_VENDOR)
> + vport->vport_flag |= FAWWPN_SET;
> + }
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 19/22] Fix implicit logo and RSCN handling for NVMET
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
NVMET didn't have any RSCN handling at all and
would not execute implicit LOGO when receiving a PLOGI
from an rport that NVMET had in state UNMAPPED.
Clean up the logic in lpfc_nlp_state_cleanup for
initiators (FCP and NVME). NVMET should not respond to
RSCN including allocating new ndlps so this code was
conditionalized when nvmet_support is true. The check
for NLP_RCV_PLOGI in lpfc_setup_disc_node was moved
below the check for nvmet_support to allow the NVMET
to recover initiator nodes correctly. The implicit
logo was introduced with lpfc_rcv_plogi when NVMET gets
a PLOGI on an ndlp in UNMAPPED state. The RSCN handling
was modified to not respond to an RSCN in NVMET. Instead
NVMET sends a GID_FT and determines if an NVMEP_INITIATOR
it has is UNMAPPED but no longer in the zone membership.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_ct.c | 68 +++++++++++++++++++++++++++++---------
drivers/scsi/lpfc/lpfc_disc.h | 1 +
drivers/scsi/lpfc/lpfc_els.c | 23 ++++++-------
drivers/scsi/lpfc/lpfc_hbadisc.c | 62 ++++++++++++++++++++++------------
drivers/scsi/lpfc/lpfc_nportdisc.c | 8 +++--
5 files changed, 110 insertions(+), 52 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_ct.c b/drivers/scsi/lpfc/lpfc_ct.c
index d3e9af9..1487406 100644
--- a/drivers/scsi/lpfc/lpfc_ct.c
+++ b/drivers/scsi/lpfc/lpfc_ct.c
@@ -537,19 +537,53 @@ lpfc_prep_node_fc4type(struct lpfc_vport *vport, uint32_t Did, uint8_t fc4_type)
}
}
+static void
+lpfc_ns_rsp_audit_did(struct lpfc_vport *vport, uint32_t Did, uint8_t fc4_type)
+{
+ struct lpfc_hba *phba = vport->phba;
+ struct lpfc_nodelist *ndlp = NULL;
+ struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
+
+ /*
+ * To conserve rpi's, filter out addresses for other
+ * vports on the same physical HBAs.
+ */
+ if (Did != vport->fc_myDID &&
+ (!lpfc_find_vport_by_did(phba, Did) ||
+ vport->cfg_peer_port_login)) {
+ if (!phba->nvmet_support) {
+ /* FCPI/NVMEI path. Process Did */
+ lpfc_prep_node_fc4type(vport, Did, fc4_type);
+ return;
+ }
+ /* NVMET path. NVMET only cares about NVMEI nodes. */
+ list_for_each_entry(ndlp, &vport->fc_nodes, nlp_listp) {
+ if (ndlp->nlp_type != NLP_NVME_INITIATOR ||
+ ndlp->nlp_state != NLP_STE_UNMAPPED_NODE)
+ continue;
+ spin_lock_irq(shost->host_lock);
+ if (ndlp->nlp_DID == Did)
+ ndlp->nlp_flag &= ~NLP_NVMET_RECOV;
+ else
+ ndlp->nlp_flag |= NLP_NVMET_RECOV;
+ spin_unlock_irq(shost->host_lock);
+ }
+ }
+}
+
static int
lpfc_ns_rsp(struct lpfc_vport *vport, struct lpfc_dmabuf *mp, uint8_t fc4_type,
uint32_t Size)
{
- struct lpfc_hba *phba = vport->phba;
struct lpfc_sli_ct_request *Response =
(struct lpfc_sli_ct_request *) mp->virt;
- struct lpfc_nodelist *ndlp = NULL;
struct lpfc_dmabuf *mlast, *next_mp;
uint32_t *ctptr = (uint32_t *) & Response->un.gid.PortType;
uint32_t Did, CTentry;
int Cnt;
struct list_head head;
+ struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
+ struct lpfc_nodelist *ndlp = NULL;
lpfc_set_disctmo(vport);
vport->num_disc_nodes = 0;
@@ -574,19 +608,7 @@ lpfc_ns_rsp(struct lpfc_vport *vport, struct lpfc_dmabuf *mp, uint8_t fc4_type,
/* Get next DID from NameServer List */
CTentry = *ctptr++;
Did = ((be32_to_cpu(CTentry)) & Mask_DID);
-
- ndlp = NULL;
-
- /*
- * Check for rscn processing or not
- * To conserve rpi's, filter out addresses for other
- * vports on the same physical HBAs.
- */
- if ((Did != vport->fc_myDID) &&
- ((lpfc_find_vport_by_did(phba, Did) == NULL) ||
- vport->cfg_peer_port_login))
- lpfc_prep_node_fc4type(vport, Did, fc4_type);
-
+ lpfc_ns_rsp_audit_did(vport, Did, fc4_type);
if (CTentry & (cpu_to_be32(SLI_CT_LAST_ENTRY)))
goto nsout1;
@@ -596,6 +618,22 @@ lpfc_ns_rsp(struct lpfc_vport *vport, struct lpfc_dmabuf *mp, uint8_t fc4_type,
}
+ /* All GID_FT entries processed. If the driver is running in
+ * in target mode, put impacted nodes into recovery and drop
+ * the RPI to flush outstanding IO.
+ */
+ if (vport->phba->nvmet_support) {
+ list_for_each_entry(ndlp, &vport->fc_nodes, nlp_listp) {
+ if (!(ndlp->nlp_flag & NLP_NVMET_RECOV))
+ continue;
+ lpfc_disc_state_machine(vport, ndlp, NULL,
+ NLP_EVT_DEVICE_RECOVERY);
+ spin_lock_irq(shost->host_lock);
+ ndlp->nlp_flag &= ~NLP_NVMET_RECOV;
+ spin_lock_irq(shost->host_lock);
+ }
+ }
+
nsout1:
list_del(&head);
return 0;
diff --git a/drivers/scsi/lpfc/lpfc_disc.h b/drivers/scsi/lpfc/lpfc_disc.h
index f4ff99d..9d5a379 100644
--- a/drivers/scsi/lpfc/lpfc_disc.h
+++ b/drivers/scsi/lpfc/lpfc_disc.h
@@ -157,6 +157,7 @@ struct lpfc_node_rrq {
#define NLP_LOGO_SND 0x00000100 /* sent LOGO request for this entry */
#define NLP_RNID_SND 0x00000400 /* sent RNID request for this entry */
#define NLP_ELS_SND_MASK 0x000007e0 /* sent ELS request for this entry */
+#define NLP_NVMET_RECOV 0x00001000 /* NVMET auditing node for recovery. */
#define NLP_DEFER_RM 0x00010000 /* Remove this ndlp if no longer used */
#define NLP_DELAY_TMO 0x00020000 /* delay timeout is running for node */
#define NLP_NPR_2B_DISC 0x00040000 /* node is included in num_disc_nodes */
diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index 35fc260..8d1c689 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -5861,8 +5861,11 @@ lpfc_rscn_recovery_check(struct lpfc_vport *vport)
(ndlp->nlp_state == NLP_STE_UNUSED_NODE) ||
!lpfc_rscn_payload_check(vport, ndlp->nlp_DID))
continue;
+
+ /* NVME Target mode does not do RSCN Recovery. */
if (vport->phba->nvmet_support)
continue;
+
lpfc_disc_state_machine(vport, ndlp, NULL,
NLP_EVT_DEVICE_RECOVERY);
lpfc_cancel_retry_delay_tmo(vport, ndlp);
@@ -6150,22 +6153,16 @@ lpfc_els_handle_rscn(struct lpfc_vport *vport)
ndlp = lpfc_findnode_did(vport, NameServer_DID);
if (ndlp && NLP_CHK_NODE_ACT(ndlp)
&& ndlp->nlp_state == NLP_STE_UNMAPPED_NODE) {
- /* Good ndlp, issue CT Request to NameServer */
+ /* Good ndlp, issue CT Request to NameServer. Need to
+ * know how many gidfts were issued. If none, then just
+ * flush the RSCN. Otherwise, the outstanding requests
+ * need to complete.
+ */
vport->gidft_inp = 0;
- if (lpfc_issue_gidft(vport) == 0)
- /* Wait for NameServer query cmpl before we can
- * continue
- */
+ if (lpfc_issue_gidft(vport) > 0)
return 1;
} else {
- /* If login to NameServer does not exist, issue one */
- /* Good status, issue PLOGI to NameServer */
- ndlp = lpfc_findnode_did(vport, NameServer_DID);
- if (ndlp && NLP_CHK_NODE_ACT(ndlp))
- /* Wait for NameServer login cmpl before we can
- continue */
- return 1;
-
+ /* Nameserver login in question. Revalidate. */
if (ndlp) {
ndlp = lpfc_enable_node(vport, ndlp,
NLP_STE_PLOGI_ISSUE);
diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c
index e824ec1..828fdee 100644
--- a/drivers/scsi/lpfc/lpfc_hbadisc.c
+++ b/drivers/scsi/lpfc/lpfc_hbadisc.c
@@ -4148,7 +4148,6 @@ lpfc_nlp_state_cleanup(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
int old_state, int new_state)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
- struct lpfc_hba *phba = vport->phba;
if (new_state == NLP_STE_UNMAPPED_NODE) {
ndlp->nlp_flag &= ~NLP_NODEV_REMOVE;
@@ -4167,14 +4166,14 @@ lpfc_nlp_state_cleanup(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
lpfc_unregister_remote_port(ndlp);
}
- /* Notify the NVME transport of this rport's loss */
- if (((phba->cfg_enable_fc4_type == LPFC_ENABLE_BOTH) ||
- (phba->cfg_enable_fc4_type == LPFC_ENABLE_NVME)) &&
- (vport->phba->nvmet_support == 0) &&
- ((ndlp->nlp_fc4_type & NLP_FC4_NVME) ||
- (ndlp->nlp_DID == Fabric_DID))) {
+ /* Notify the NVME transport of this rport's loss on the
+ * Initiator. For NVME Target, should upcall transport
+ * in the else clause when API available.
+ */
+ if (ndlp->nlp_fc4_type & NLP_FC4_NVME) {
vport->phba->nport_event_cnt++;
- lpfc_nvme_unregister_port(vport, ndlp);
+ if (vport->phba->nvmet_support == 0)
+ lpfc_nvme_unregister_port(vport, ndlp);
}
}
@@ -5128,6 +5127,8 @@ lpfc_setup_disc_node(struct lpfc_vport *vport, uint32_t did)
ndlp = lpfc_findnode_did(vport, did);
if (!ndlp) {
+ if (vport->phba->nvmet_support)
+ return NULL;
if ((vport->fc_flag & FC_RSCN_MODE) != 0 &&
lpfc_rscn_payload_check(vport, did) == 0)
return NULL;
@@ -5135,56 +5136,73 @@ lpfc_setup_disc_node(struct lpfc_vport *vport, uint32_t did)
if (!ndlp)
return NULL;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
- if (vport->phba->nvmet_support)
- return ndlp;
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
return ndlp;
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
+ if (vport->phba->nvmet_support)
+ return NULL;
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_NPR_NODE);
if (!ndlp)
return NULL;
- if (vport->phba->nvmet_support)
- return ndlp;
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
return ndlp;
}
+ /* The NVME Target does not want to actively manage an rport.
+ * The goal is to allow the target to reset its state and clear
+ * pending IO in preparation for the initiator to recover.
+ */
if ((vport->fc_flag & FC_RSCN_MODE) &&
!(vport->fc_flag & FC_NDISC_ACTIVE)) {
if (lpfc_rscn_payload_check(vport, did)) {
- /* If we've already received a PLOGI from this NPort
- * we don't need to try to discover it again.
- */
- if (ndlp->nlp_flag & NLP_RCV_PLOGI)
- return NULL;
/* Since this node is marked for discovery,
* delay timeout is not needed.
*/
lpfc_cancel_retry_delay_tmo(vport, ndlp);
+
+ /* NVME Target mode waits until rport is known to be
+ * impacted by the RSCN before it transitions. No
+ * active management - just go to NPR provided the
+ * node had a valid login.
+ */
if (vport->phba->nvmet_support)
return ndlp;
+
+ /* If we've already received a PLOGI from this NPort
+ * we don't need to try to discover it again.
+ */
+ if (ndlp->nlp_flag & NLP_RCV_PLOGI)
+ return NULL;
+
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
} else
ndlp = NULL;
} else {
- /* If we've already received a PLOGI from this NPort,
- * or we are already in the process of discovery on it,
- * we don't need to try to discover it again.
+ /* If the initiator received a PLOGI from this NPort or if the
+ * initiator is already in the process of discovery on it,
+ * there's no need to try to discover it again.
*/
if (ndlp->nlp_state == NLP_STE_ADISC_ISSUE ||
ndlp->nlp_state == NLP_STE_PLOGI_ISSUE ||
- ndlp->nlp_flag & NLP_RCV_PLOGI)
+ (!vport->phba->nvmet_support &&
+ ndlp->nlp_flag & NLP_RCV_PLOGI))
return NULL;
- lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
+
if (vport->phba->nvmet_support)
return ndlp;
+
+ /* Moving to NPR state clears unsolicited flags and
+ * allows for rediscovery
+ */
+ lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
+
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
diff --git a/drivers/scsi/lpfc/lpfc_nportdisc.c b/drivers/scsi/lpfc/lpfc_nportdisc.c
index 061626b..8777c2d 100644
--- a/drivers/scsi/lpfc/lpfc_nportdisc.c
+++ b/drivers/scsi/lpfc/lpfc_nportdisc.c
@@ -361,8 +361,12 @@ lpfc_rcv_plogi(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
case NLP_STE_PRLI_ISSUE:
case NLP_STE_UNMAPPED_NODE:
case NLP_STE_MAPPED_NODE:
- /* lpfc_plogi_confirm_nport skips fabric did, handle it here */
- if (!(ndlp->nlp_type & NLP_FABRIC)) {
+ /* For initiators, lpfc_plogi_confirm_nport skips fabric did.
+ * For target mode, execute implicit logo.
+ * Fabric nodes go into NPR.
+ */
+ if (!(ndlp->nlp_type & NLP_FABRIC) &&
+ !(phba->nvmet_support)) {
lpfc_els_rsp_acc(vport, ELS_CMD_PLOGI, cmdiocb,
ndlp, NULL);
return 1;
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 19/22] Fix implicit logo and RSCN handling for NVMET
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
NVMET didn't have any RSCN handling at all and
would not execute implicit LOGO when receiving a PLOGI
from an rport that NVMET had in state UNMAPPED.
Clean up the logic in lpfc_nlp_state_cleanup for
initiators (FCP and NVME). NVMET should not respond to
RSCN including allocating new ndlps so this code was
conditionalized when nvmet_support is true. The check
for NLP_RCV_PLOGI in lpfc_setup_disc_node was moved
below the check for nvmet_support to allow the NVMET
to recover initiator nodes correctly. The implicit
logo was introduced with lpfc_rcv_plogi when NVMET gets
a PLOGI on an ndlp in UNMAPPED state. The RSCN handling
was modified to not respond to an RSCN in NVMET. Instead
NVMET sends a GID_FT and determines if an NVMEP_INITIATOR
it has is UNMAPPED but no longer in the zone membership.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_ct.c | 68 +++++++++++++++++++++++++++++---------
drivers/scsi/lpfc/lpfc_disc.h | 1 +
drivers/scsi/lpfc/lpfc_els.c | 23 ++++++-------
drivers/scsi/lpfc/lpfc_hbadisc.c | 62 ++++++++++++++++++++++------------
drivers/scsi/lpfc/lpfc_nportdisc.c | 8 +++--
5 files changed, 110 insertions(+), 52 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_ct.c b/drivers/scsi/lpfc/lpfc_ct.c
index d3e9af9..1487406 100644
--- a/drivers/scsi/lpfc/lpfc_ct.c
+++ b/drivers/scsi/lpfc/lpfc_ct.c
@@ -537,19 +537,53 @@ lpfc_prep_node_fc4type(struct lpfc_vport *vport, uint32_t Did, uint8_t fc4_type)
}
}
+static void
+lpfc_ns_rsp_audit_did(struct lpfc_vport *vport, uint32_t Did, uint8_t fc4_type)
+{
+ struct lpfc_hba *phba = vport->phba;
+ struct lpfc_nodelist *ndlp = NULL;
+ struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
+
+ /*
+ * To conserve rpi's, filter out addresses for other
+ * vports on the same physical HBAs.
+ */
+ if (Did != vport->fc_myDID &&
+ (!lpfc_find_vport_by_did(phba, Did) ||
+ vport->cfg_peer_port_login)) {
+ if (!phba->nvmet_support) {
+ /* FCPI/NVMEI path. Process Did */
+ lpfc_prep_node_fc4type(vport, Did, fc4_type);
+ return;
+ }
+ /* NVMET path. NVMET only cares about NVMEI nodes. */
+ list_for_each_entry(ndlp, &vport->fc_nodes, nlp_listp) {
+ if (ndlp->nlp_type != NLP_NVME_INITIATOR ||
+ ndlp->nlp_state != NLP_STE_UNMAPPED_NODE)
+ continue;
+ spin_lock_irq(shost->host_lock);
+ if (ndlp->nlp_DID == Did)
+ ndlp->nlp_flag &= ~NLP_NVMET_RECOV;
+ else
+ ndlp->nlp_flag |= NLP_NVMET_RECOV;
+ spin_unlock_irq(shost->host_lock);
+ }
+ }
+}
+
static int
lpfc_ns_rsp(struct lpfc_vport *vport, struct lpfc_dmabuf *mp, uint8_t fc4_type,
uint32_t Size)
{
- struct lpfc_hba *phba = vport->phba;
struct lpfc_sli_ct_request *Response =
(struct lpfc_sli_ct_request *) mp->virt;
- struct lpfc_nodelist *ndlp = NULL;
struct lpfc_dmabuf *mlast, *next_mp;
uint32_t *ctptr = (uint32_t *) & Response->un.gid.PortType;
uint32_t Did, CTentry;
int Cnt;
struct list_head head;
+ struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
+ struct lpfc_nodelist *ndlp = NULL;
lpfc_set_disctmo(vport);
vport->num_disc_nodes = 0;
@@ -574,19 +608,7 @@ lpfc_ns_rsp(struct lpfc_vport *vport, struct lpfc_dmabuf *mp, uint8_t fc4_type,
/* Get next DID from NameServer List */
CTentry = *ctptr++;
Did = ((be32_to_cpu(CTentry)) & Mask_DID);
-
- ndlp = NULL;
-
- /*
- * Check for rscn processing or not
- * To conserve rpi's, filter out addresses for other
- * vports on the same physical HBAs.
- */
- if ((Did != vport->fc_myDID) &&
- ((lpfc_find_vport_by_did(phba, Did) == NULL) ||
- vport->cfg_peer_port_login))
- lpfc_prep_node_fc4type(vport, Did, fc4_type);
-
+ lpfc_ns_rsp_audit_did(vport, Did, fc4_type);
if (CTentry & (cpu_to_be32(SLI_CT_LAST_ENTRY)))
goto nsout1;
@@ -596,6 +618,22 @@ lpfc_ns_rsp(struct lpfc_vport *vport, struct lpfc_dmabuf *mp, uint8_t fc4_type,
}
+ /* All GID_FT entries processed. If the driver is running in
+ * in target mode, put impacted nodes into recovery and drop
+ * the RPI to flush outstanding IO.
+ */
+ if (vport->phba->nvmet_support) {
+ list_for_each_entry(ndlp, &vport->fc_nodes, nlp_listp) {
+ if (!(ndlp->nlp_flag & NLP_NVMET_RECOV))
+ continue;
+ lpfc_disc_state_machine(vport, ndlp, NULL,
+ NLP_EVT_DEVICE_RECOVERY);
+ spin_lock_irq(shost->host_lock);
+ ndlp->nlp_flag &= ~NLP_NVMET_RECOV;
+ spin_lock_irq(shost->host_lock);
+ }
+ }
+
nsout1:
list_del(&head);
return 0;
diff --git a/drivers/scsi/lpfc/lpfc_disc.h b/drivers/scsi/lpfc/lpfc_disc.h
index f4ff99d..9d5a379 100644
--- a/drivers/scsi/lpfc/lpfc_disc.h
+++ b/drivers/scsi/lpfc/lpfc_disc.h
@@ -157,6 +157,7 @@ struct lpfc_node_rrq {
#define NLP_LOGO_SND 0x00000100 /* sent LOGO request for this entry */
#define NLP_RNID_SND 0x00000400 /* sent RNID request for this entry */
#define NLP_ELS_SND_MASK 0x000007e0 /* sent ELS request for this entry */
+#define NLP_NVMET_RECOV 0x00001000 /* NVMET auditing node for recovery. */
#define NLP_DEFER_RM 0x00010000 /* Remove this ndlp if no longer used */
#define NLP_DELAY_TMO 0x00020000 /* delay timeout is running for node */
#define NLP_NPR_2B_DISC 0x00040000 /* node is included in num_disc_nodes */
diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index 35fc260..8d1c689 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -5861,8 +5861,11 @@ lpfc_rscn_recovery_check(struct lpfc_vport *vport)
(ndlp->nlp_state == NLP_STE_UNUSED_NODE) ||
!lpfc_rscn_payload_check(vport, ndlp->nlp_DID))
continue;
+
+ /* NVME Target mode does not do RSCN Recovery. */
if (vport->phba->nvmet_support)
continue;
+
lpfc_disc_state_machine(vport, ndlp, NULL,
NLP_EVT_DEVICE_RECOVERY);
lpfc_cancel_retry_delay_tmo(vport, ndlp);
@@ -6150,22 +6153,16 @@ lpfc_els_handle_rscn(struct lpfc_vport *vport)
ndlp = lpfc_findnode_did(vport, NameServer_DID);
if (ndlp && NLP_CHK_NODE_ACT(ndlp)
&& ndlp->nlp_state == NLP_STE_UNMAPPED_NODE) {
- /* Good ndlp, issue CT Request to NameServer */
+ /* Good ndlp, issue CT Request to NameServer. Need to
+ * know how many gidfts were issued. If none, then just
+ * flush the RSCN. Otherwise, the outstanding requests
+ * need to complete.
+ */
vport->gidft_inp = 0;
- if (lpfc_issue_gidft(vport) == 0)
- /* Wait for NameServer query cmpl before we can
- * continue
- */
+ if (lpfc_issue_gidft(vport) > 0)
return 1;
} else {
- /* If login to NameServer does not exist, issue one */
- /* Good status, issue PLOGI to NameServer */
- ndlp = lpfc_findnode_did(vport, NameServer_DID);
- if (ndlp && NLP_CHK_NODE_ACT(ndlp))
- /* Wait for NameServer login cmpl before we can
- continue */
- return 1;
-
+ /* Nameserver login in question. Revalidate. */
if (ndlp) {
ndlp = lpfc_enable_node(vport, ndlp,
NLP_STE_PLOGI_ISSUE);
diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c
index e824ec1..828fdee 100644
--- a/drivers/scsi/lpfc/lpfc_hbadisc.c
+++ b/drivers/scsi/lpfc/lpfc_hbadisc.c
@@ -4148,7 +4148,6 @@ lpfc_nlp_state_cleanup(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
int old_state, int new_state)
{
struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
- struct lpfc_hba *phba = vport->phba;
if (new_state == NLP_STE_UNMAPPED_NODE) {
ndlp->nlp_flag &= ~NLP_NODEV_REMOVE;
@@ -4167,14 +4166,14 @@ lpfc_nlp_state_cleanup(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
lpfc_unregister_remote_port(ndlp);
}
- /* Notify the NVME transport of this rport's loss */
- if (((phba->cfg_enable_fc4_type == LPFC_ENABLE_BOTH) ||
- (phba->cfg_enable_fc4_type == LPFC_ENABLE_NVME)) &&
- (vport->phba->nvmet_support == 0) &&
- ((ndlp->nlp_fc4_type & NLP_FC4_NVME) ||
- (ndlp->nlp_DID == Fabric_DID))) {
+ /* Notify the NVME transport of this rport's loss on the
+ * Initiator. For NVME Target, should upcall transport
+ * in the else clause when API available.
+ */
+ if (ndlp->nlp_fc4_type & NLP_FC4_NVME) {
vport->phba->nport_event_cnt++;
- lpfc_nvme_unregister_port(vport, ndlp);
+ if (vport->phba->nvmet_support == 0)
+ lpfc_nvme_unregister_port(vport, ndlp);
}
}
@@ -5128,6 +5127,8 @@ lpfc_setup_disc_node(struct lpfc_vport *vport, uint32_t did)
ndlp = lpfc_findnode_did(vport, did);
if (!ndlp) {
+ if (vport->phba->nvmet_support)
+ return NULL;
if ((vport->fc_flag & FC_RSCN_MODE) != 0 &&
lpfc_rscn_payload_check(vport, did) == 0)
return NULL;
@@ -5135,56 +5136,73 @@ lpfc_setup_disc_node(struct lpfc_vport *vport, uint32_t did)
if (!ndlp)
return NULL;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
- if (vport->phba->nvmet_support)
- return ndlp;
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
return ndlp;
} else if (!NLP_CHK_NODE_ACT(ndlp)) {
+ if (vport->phba->nvmet_support)
+ return NULL;
ndlp = lpfc_enable_node(vport, ndlp, NLP_STE_NPR_NODE);
if (!ndlp)
return NULL;
- if (vport->phba->nvmet_support)
- return ndlp;
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
return ndlp;
}
+ /* The NVME Target does not want to actively manage an rport.
+ * The goal is to allow the target to reset its state and clear
+ * pending IO in preparation for the initiator to recover.
+ */
if ((vport->fc_flag & FC_RSCN_MODE) &&
!(vport->fc_flag & FC_NDISC_ACTIVE)) {
if (lpfc_rscn_payload_check(vport, did)) {
- /* If we've already received a PLOGI from this NPort
- * we don't need to try to discover it again.
- */
- if (ndlp->nlp_flag & NLP_RCV_PLOGI)
- return NULL;
/* Since this node is marked for discovery,
* delay timeout is not needed.
*/
lpfc_cancel_retry_delay_tmo(vport, ndlp);
+
+ /* NVME Target mode waits until rport is known to be
+ * impacted by the RSCN before it transitions. No
+ * active management - just go to NPR provided the
+ * node had a valid login.
+ */
if (vport->phba->nvmet_support)
return ndlp;
+
+ /* If we've already received a PLOGI from this NPort
+ * we don't need to try to discover it again.
+ */
+ if (ndlp->nlp_flag & NLP_RCV_PLOGI)
+ return NULL;
+
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
} else
ndlp = NULL;
} else {
- /* If we've already received a PLOGI from this NPort,
- * or we are already in the process of discovery on it,
- * we don't need to try to discover it again.
+ /* If the initiator received a PLOGI from this NPort or if the
+ * initiator is already in the process of discovery on it,
+ * there's no need to try to discover it again.
*/
if (ndlp->nlp_state == NLP_STE_ADISC_ISSUE ||
ndlp->nlp_state == NLP_STE_PLOGI_ISSUE ||
- ndlp->nlp_flag & NLP_RCV_PLOGI)
+ (!vport->phba->nvmet_support &&
+ ndlp->nlp_flag & NLP_RCV_PLOGI))
return NULL;
- lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
+
if (vport->phba->nvmet_support)
return ndlp;
+
+ /* Moving to NPR state clears unsolicited flags and
+ * allows for rediscovery
+ */
+ lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
+
spin_lock_irq(shost->host_lock);
ndlp->nlp_flag |= NLP_NPR_2B_DISC;
spin_unlock_irq(shost->host_lock);
diff --git a/drivers/scsi/lpfc/lpfc_nportdisc.c b/drivers/scsi/lpfc/lpfc_nportdisc.c
index 061626b..8777c2d 100644
--- a/drivers/scsi/lpfc/lpfc_nportdisc.c
+++ b/drivers/scsi/lpfc/lpfc_nportdisc.c
@@ -361,8 +361,12 @@ lpfc_rcv_plogi(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
case NLP_STE_PRLI_ISSUE:
case NLP_STE_UNMAPPED_NODE:
case NLP_STE_MAPPED_NODE:
- /* lpfc_plogi_confirm_nport skips fabric did, handle it here */
- if (!(ndlp->nlp_type & NLP_FABRIC)) {
+ /* For initiators, lpfc_plogi_confirm_nport skips fabric did.
+ * For target mode, execute implicit logo.
+ * Fabric nodes go into NPR.
+ */
+ if (!(ndlp->nlp_type & NLP_FABRIC) &&
+ !(phba->nvmet_support)) {
lpfc_els_rsp_acc(vport, ELS_CMD_PLOGI, cmdiocb,
ndlp, NULL);
return 1;
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 19/22] Fix implicit logo and RSCN handling for NVMET
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 7:53 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:53 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:38PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> NVMET didn't have any RSCN handling at all and
> would not execute implicit LOGO when receiving a PLOGI
> from an rport that NVMET had in state UNMAPPED.
>
> Clean up the logic in lpfc_nlp_state_cleanup for
> initiators (FCP and NVME). NVMET should not respond to
> RSCN including allocating new ndlps so this code was
> conditionalized when nvmet_support is true. The check
> for NLP_RCV_PLOGI in lpfc_setup_disc_node was moved
> below the check for nvmet_support to allow the NVMET
> to recover initiator nodes correctly. The implicit
> logo was introduced with lpfc_rcv_plogi when NVMET gets
> a PLOGI on an ndlp in UNMAPPED state. The RSCN handling
> was modified to not respond to an RSCN in NVMET. Instead
> NVMET sends a GID_FT and determines if an NVMEP_INITIATOR
> it has is UNMAPPED but no longer in the zone membership.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 19/22] Fix implicit logo and RSCN handling for NVMET
@ 2017-04-20 7:53 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 7:53 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:38PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> NVMET didn't have any RSCN handling at all and
> would not execute implicit LOGO when receiving a PLOGI
> from an rport that NVMET had in state UNMAPPED.
>
> Clean up the logic in lpfc_nlp_state_cleanup for
> initiators (FCP and NVME). NVMET should not respond to
> RSCN including allocating new ndlps so this code was
> conditionalized when nvmet_support is true. The check
> for NLP_RCV_PLOGI in lpfc_setup_disc_node was moved
> below the check for nvmet_support to allow the NVMET
> to recover initiator nodes correctly. The implicit
> logo was introduced with lpfc_rcv_plogi when NVMET gets
> a PLOGI on an ndlp in UNMAPPED state. The RSCN handling
> was modified to not respond to an RSCN in NVMET. Instead
> NVMET sends a GID_FT and determines if an NVMEP_INITIATOR
> it has is UNMAPPED but no longer in the zone membership.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 20/22] Update ABORT processing for NVMET.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
The driver with nvme had this routine stubbed.
Right now XRI_ABORTED_CQE is not handled and the FC NVMET
Transport has a new API for the driver.
Missing code path, new NVME abort API
Update ABORT processing for NVMET
There are 3 new FC NVMET Transport API/ template routines for NVMET:
lpfc_nvmet_xmt_fcp_release
This NVMET template callback routine called to release context
associated with an IO This routine is ALWAYS called last, even
if the IO was aborted or completed in error.
lpfc_nvmet_xmt_fcp_abort
This NVMET template callback routine called to abort an exchange that
has an IO in progress
nvmet_fc_rcv_fcp_req
When the lpfc driver receives an ABTS, this NVME FC transport layer
callback routine is called. For this case there are 2 paths thru the
driver: the driver either has an outstanding exchange / context for the
XRI to be aborted or not. If not, a BA_RJT is issued otherwise a BA_ACC
NVMET Driver abort paths:
There are 2 paths for aborting an IO. The first one is we receive an IO and
decide not to process it because of lack of resources. An unsolicated ABTS
is immediately sent back to the initiator as a response.
lpfc_nvmet_unsol_fcp_buffer
lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE)
The second one is we sent the IO up to the NVMET transport layer to
process, and for some reason the NVME Transport layer decided to abort the
IO before it completes all its phases. For this case there are 2 paths
thru the driver:
the driver either has an outstanding TSEND/TRECEIVE/TRSP WQE or no
outstanding WQEs are present for the exchange / context.
lpfc_nvmet_xmt_fcp_abort
if (LPFC_NVMET_IO_INP)
lpfc_nvmet_sol_fcp_issue_abort (ABORT_WQE)
lpfc_nvmet_sol_fcp_abort_cmp
else
lpfc_nvmet_unsol_fcp_issue_abort
lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE)
lpfc_nvmet_unsol_fcp_abort_cmp
Context flags:
LPFC_NVMET_IOP - his flag signifies an IO is in progress on the exchange.
LPFC_NVMET_XBUSY - this flag indicates the IO completed but the firmware
is still busy with the corresponding exchange. The exchange should not be
reused until after a XRI_ABORTED_CQE is received for that exchange.
LPFC_NVMET_ABORT_OP - this flag signifies an ABORT_WQE was issued on the
exchange.
LPFC_NVMET_CTX_RLS - this flag signifies a context free was requested,
but we are deferring it due to an XBUSY or ABORT in progress.
A ctxlock is added to the context structure that is used whenever these
flags are set/read within the context of an IO.
The LPFC_NVMET_CTX_RLS flag is only set in the defer_relase routine when
the transport has resolved all IO associated with the buffer. The flag is
cleared when the CTX is associated with a new IO.
An exchange can has both an LPFC_NVMET_XBUSY and a LPFC_NVMET_ABORT_OP
condition active simultaneously. Both conditions must complete before the
exchange is freed.
When the abort callback (lpfc_nvmet_xmt_fcp_abort) is envoked:
If there is an outstanding IO, the driver will issue an ABORT_WQE. This
should result in 3 completions for the exchange:
1) IO cmpl with XB bit set
2) Abort WQE cmpl
3) XRI_ABORTED_CQE cmpl
For this scenerio, after completion #1, the NVMET Transport IO rsp
callback is called. After completion #2, no action is taken with respect
to the exchange / context. After completion #3, the exchange context is
free for re-use on another IO.
If there is no outstanding activity on the exchange, the driver will send a
ABTS to the Initiator. Upon completion of this WQE, the exchange / context
is freed for re-use on another IO.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_crtn.h | 7 +
drivers/scsi/lpfc/lpfc_debugfs.c | 55 ++++--
drivers/scsi/lpfc/lpfc_hw4.h | 3 +
drivers/scsi/lpfc/lpfc_init.c | 52 +++---
drivers/scsi/lpfc/lpfc_mbox.c | 7 +-
drivers/scsi/lpfc/lpfc_nvme.c | 45 +++--
drivers/scsi/lpfc/lpfc_nvmet.c | 349 +++++++++++++++++++++++++++++++--------
drivers/scsi/lpfc/lpfc_nvmet.h | 10 +-
drivers/scsi/lpfc/lpfc_sli.c | 7 +-
drivers/scsi/lpfc/lpfc_sli4.h | 2 +-
10 files changed, 412 insertions(+), 125 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_crtn.h b/drivers/scsi/lpfc/lpfc_crtn.h
index d859aff..24da922 100644
--- a/drivers/scsi/lpfc/lpfc_crtn.h
+++ b/drivers/scsi/lpfc/lpfc_crtn.h
@@ -24,6 +24,7 @@ typedef int (*node_filter)(struct lpfc_nodelist *, void *);
struct fc_rport;
struct fc_frame_header;
+struct lpfc_nvmet_rcv_ctx;
void lpfc_down_link(struct lpfc_hba *, LPFC_MBOXQ_t *);
void lpfc_sli_read_link_ste(struct lpfc_hba *);
void lpfc_dump_mem(struct lpfc_hba *, LPFC_MBOXQ_t *, uint16_t, uint16_t);
@@ -245,6 +246,10 @@ struct hbq_dmabuf *lpfc_sli4_rb_alloc(struct lpfc_hba *);
void lpfc_sli4_rb_free(struct lpfc_hba *, struct hbq_dmabuf *);
struct rqb_dmabuf *lpfc_sli4_nvmet_alloc(struct lpfc_hba *phba);
void lpfc_sli4_nvmet_free(struct lpfc_hba *phba, struct rqb_dmabuf *dmab);
+void lpfc_nvmet_rq_post(struct lpfc_hba *phba, struct lpfc_nvmet_rcv_ctx *ctxp,
+ struct lpfc_dmabuf *mp);
+int lpfc_nvmet_rcv_unsol_abort(struct lpfc_vport *vport,
+ struct fc_frame_header *fc_hdr);
void lpfc_sli4_build_dflt_fcf_record(struct lpfc_hba *, struct fcf_record *,
uint16_t);
int lpfc_sli4_rq_put(struct lpfc_queue *hq, struct lpfc_queue *dq,
@@ -302,6 +307,8 @@ int lpfc_sli_check_eratt(struct lpfc_hba *);
void lpfc_sli_handle_slow_ring_event(struct lpfc_hba *,
struct lpfc_sli_ring *, uint32_t);
void lpfc_sli4_handle_received_buffer(struct lpfc_hba *, struct hbq_dmabuf *);
+void lpfc_sli4_seq_abort_rsp(struct lpfc_vport *vport,
+ struct fc_frame_header *fc_hdr, bool aborted);
void lpfc_sli_def_mbox_cmpl(struct lpfc_hba *, LPFC_MBOXQ_t *);
void lpfc_sli4_unreg_rpi_cmpl_clr(struct lpfc_hba *, LPFC_MBOXQ_t *);
int lpfc_sli_issue_iocb(struct lpfc_hba *, uint32_t,
diff --git a/drivers/scsi/lpfc/lpfc_debugfs.c b/drivers/scsi/lpfc/lpfc_debugfs.c
index 55a8d8f..4d6cd7a 100644
--- a/drivers/scsi/lpfc/lpfc_debugfs.c
+++ b/drivers/scsi/lpfc/lpfc_debugfs.c
@@ -745,73 +745,102 @@ lpfc_debugfs_nvmestat_data(struct lpfc_vport *vport, char *buf, int size)
{
struct lpfc_hba *phba = vport->phba;
struct lpfc_nvmet_tgtport *tgtp;
+ struct lpfc_nvmet_rcv_ctx *ctxp, *next_ctxp;
int len = 0;
+ int cnt;
if (phba->nvmet_support) {
if (!phba->targetport)
return len;
tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"\nNVME Targetport Statistics\n");
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"LS: Rcv %08x Drop %08x Abort %08x\n",
atomic_read(&tgtp->rcv_ls_req_in),
atomic_read(&tgtp->rcv_ls_req_drop),
atomic_read(&tgtp->xmt_ls_abort));
if (atomic_read(&tgtp->rcv_ls_req_in) !=
atomic_read(&tgtp->rcv_ls_req_out)) {
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"Rcv LS: in %08x != out %08x\n",
atomic_read(&tgtp->rcv_ls_req_in),
atomic_read(&tgtp->rcv_ls_req_out));
}
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"LS: Xmt %08x Drop %08x Cmpl %08x Err %08x\n",
atomic_read(&tgtp->xmt_ls_rsp),
atomic_read(&tgtp->xmt_ls_drop),
atomic_read(&tgtp->xmt_ls_rsp_cmpl),
atomic_read(&tgtp->xmt_ls_rsp_error));
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"FCP: Rcv %08x Drop %08x\n",
atomic_read(&tgtp->rcv_fcp_cmd_in),
atomic_read(&tgtp->rcv_fcp_cmd_drop));
if (atomic_read(&tgtp->rcv_fcp_cmd_in) !=
atomic_read(&tgtp->rcv_fcp_cmd_out)) {
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"Rcv FCP: in %08x != out %08x\n",
atomic_read(&tgtp->rcv_fcp_cmd_in),
atomic_read(&tgtp->rcv_fcp_cmd_out));
}
- len += snprintf(buf+len, size-len,
- "FCP Rsp: read %08x readrsp %08x write %08x rsp %08x\n",
+ len += snprintf(buf + len, size - len,
+ "FCP Rsp: read %08x readrsp %08x "
+ "write %08x rsp %08x\n",
atomic_read(&tgtp->xmt_fcp_read),
atomic_read(&tgtp->xmt_fcp_read_rsp),
atomic_read(&tgtp->xmt_fcp_write),
atomic_read(&tgtp->xmt_fcp_rsp));
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"FCP Rsp: abort %08x drop %08x\n",
atomic_read(&tgtp->xmt_fcp_abort),
atomic_read(&tgtp->xmt_fcp_drop));
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"FCP Rsp Cmpl: %08x err %08x drop %08x\n",
atomic_read(&tgtp->xmt_fcp_rsp_cmpl),
atomic_read(&tgtp->xmt_fcp_rsp_error),
atomic_read(&tgtp->xmt_fcp_rsp_drop));
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"ABORT: Xmt %08x Err %08x Cmpl %08x",
atomic_read(&tgtp->xmt_abort_rsp),
atomic_read(&tgtp->xmt_abort_rsp_error),
atomic_read(&tgtp->xmt_abort_cmpl));
- len += snprintf(buf+len, size-len, "\n");
+ len += snprintf(buf + len, size - len, "\n");
+
+ cnt = 0;
+ spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ list_for_each_entry_safe(ctxp, next_ctxp,
+ &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
+ list) {
+ cnt++;
+ }
+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ if (cnt) {
+ len += snprintf(buf + len, size - len,
+ "ABORT: %d ctx entries\n", cnt);
+ spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ list_for_each_entry_safe(ctxp, next_ctxp,
+ &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
+ list) {
+ if (len >= (size - LPFC_DEBUG_OUT_LINE_SZ))
+ break;
+ len += snprintf(buf + len, size - len,
+ "Entry: oxid %x state %x "
+ "flag %x\n",
+ ctxp->oxid, ctxp->state,
+ ctxp->flag);
+ }
+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ }
} else {
if (!(phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME))
return len;
diff --git a/drivers/scsi/lpfc/lpfc_hw4.h b/drivers/scsi/lpfc/lpfc_hw4.h
index 90499f9..1d12f2b 100644
--- a/drivers/scsi/lpfc/lpfc_hw4.h
+++ b/drivers/scsi/lpfc/lpfc_hw4.h
@@ -2720,6 +2720,9 @@ struct lpfc_mbx_request_features {
#define lpfc_mbx_rq_ftr_rq_ifip_SHIFT 7
#define lpfc_mbx_rq_ftr_rq_ifip_MASK 0x00000001
#define lpfc_mbx_rq_ftr_rq_ifip_WORD word2
+#define lpfc_mbx_rq_ftr_rq_iaar_SHIFT 9
+#define lpfc_mbx_rq_ftr_rq_iaar_MASK 0x00000001
+#define lpfc_mbx_rq_ftr_rq_iaar_WORD word2
#define lpfc_mbx_rq_ftr_rq_perfh_SHIFT 11
#define lpfc_mbx_rq_ftr_rq_perfh_MASK 0x00000001
#define lpfc_mbx_rq_ftr_rq_perfh_WORD word2
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 8acacdf..e084a66 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -42,6 +42,10 @@
#include <scsi/scsi_device.h>
#include <scsi/scsi_host.h>
#include <scsi/scsi_transport_fc.h>
+#include <scsi/scsi_tcq.h>
+#include <scsi/fc/fc_fs.h>
+
+#include <linux/nvme-fc-driver.h>
#include "lpfc_hw4.h"
#include "lpfc_hw.h"
@@ -52,6 +56,7 @@
#include "lpfc.h"
#include "lpfc_scsi.h"
#include "lpfc_nvme.h"
+#include "lpfc_nvmet.h"
#include "lpfc_logmsg.h"
#include "lpfc_crtn.h"
#include "lpfc_vport.h"
@@ -1022,8 +1027,10 @@ static int
lpfc_hba_down_post_s4(struct lpfc_hba *phba)
{
struct lpfc_scsi_buf *psb, *psb_next;
+ struct lpfc_nvmet_rcv_ctx *ctxp, *ctxp_next;
LIST_HEAD(aborts);
LIST_HEAD(nvme_aborts);
+ LIST_HEAD(nvmet_aborts);
unsigned long iflag = 0;
struct lpfc_sglq *sglq_entry = NULL;
@@ -1046,16 +1053,10 @@ lpfc_hba_down_post_s4(struct lpfc_hba *phba)
list_for_each_entry(sglq_entry,
&phba->sli4_hba.lpfc_abts_els_sgl_list, list)
sglq_entry->state = SGL_FREED;
- list_for_each_entry(sglq_entry,
- &phba->sli4_hba.lpfc_abts_nvmet_sgl_list, list)
- sglq_entry->state = SGL_FREED;
list_splice_init(&phba->sli4_hba.lpfc_abts_els_sgl_list,
&phba->sli4_hba.lpfc_els_sgl_list);
- if (phba->sli4_hba.nvme_wq)
- list_splice_init(&phba->sli4_hba.lpfc_abts_nvmet_sgl_list,
- &phba->sli4_hba.lpfc_nvmet_sgl_list);
spin_unlock(&phba->sli4_hba.sgl_list_lock);
/* abts_scsi_buf_list_lock required because worker thread uses this
@@ -1072,6 +1073,8 @@ lpfc_hba_down_post_s4(struct lpfc_hba *phba)
spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
list_splice_init(&phba->sli4_hba.lpfc_abts_nvme_buf_list,
&nvme_aborts);
+ list_splice_init(&phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
+ &nvmet_aborts);
spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
}
@@ -1085,13 +1088,20 @@ lpfc_hba_down_post_s4(struct lpfc_hba *phba)
list_splice(&aborts, &phba->lpfc_scsi_buf_list_put);
spin_unlock_irqrestore(&phba->scsi_buf_list_put_lock, iflag);
- list_for_each_entry_safe(psb, psb_next, &nvme_aborts, list) {
- psb->pCmd = NULL;
- psb->status = IOSTAT_SUCCESS;
+ if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME) {
+ list_for_each_entry_safe(psb, psb_next, &nvme_aborts, list) {
+ psb->pCmd = NULL;
+ psb->status = IOSTAT_SUCCESS;
+ }
+ spin_lock_irqsave(&phba->nvme_buf_list_put_lock, iflag);
+ list_splice(&nvme_aborts, &phba->lpfc_nvme_buf_list_put);
+ spin_unlock_irqrestore(&phba->nvme_buf_list_put_lock, iflag);
+
+ list_for_each_entry_safe(ctxp, ctxp_next, &nvmet_aborts, list) {
+ ctxp->flag &= ~(LPFC_NVMET_XBUSY | LPFC_NVMET_ABORT_OP);
+ lpfc_nvmet_rq_post(phba, ctxp, &ctxp->rqb_buffer->hbuf);
+ }
}
- spin_lock_irqsave(&phba->nvme_buf_list_put_lock, iflag);
- list_splice(&nvme_aborts, &phba->lpfc_nvme_buf_list_put);
- spin_unlock_irqrestore(&phba->nvme_buf_list_put_lock, iflag);
lpfc_sli4_free_sp_events(phba);
return 0;
@@ -5812,6 +5822,7 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
/* Initialize the Abort nvme buffer list used by driver */
spin_lock_init(&phba->sli4_hba.abts_nvme_buf_list_lock);
INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_nvme_buf_list);
+ INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_nvmet_ctx_list);
/* Fast-path XRI aborted CQ Event work queue list */
INIT_LIST_HEAD(&phba->sli4_hba.sp_nvme_xri_aborted_work_queue);
}
@@ -6439,7 +6450,7 @@ lpfc_init_sgl_list(struct lpfc_hba *phba)
INIT_LIST_HEAD(&phba->sli4_hba.lpfc_els_sgl_list);
INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_els_sgl_list);
INIT_LIST_HEAD(&phba->sli4_hba.lpfc_nvmet_sgl_list);
- INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_nvmet_sgl_list);
+ INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_nvmet_ctx_list);
/* els xri-sgl book keeping */
phba->sli4_hba.els_xri_cnt = 0;
@@ -9966,17 +9977,19 @@ lpfc_sli4_xri_exchange_busy_wait(struct lpfc_hba *phba)
{
int wait_time = 0;
int nvme_xri_cmpl = 1;
+ int nvmet_xri_cmpl = 1;
int fcp_xri_cmpl = 1;
int els_xri_cmpl = list_empty(&phba->sli4_hba.lpfc_abts_els_sgl_list);
- int nvmet_xri_cmpl =
- list_empty(&phba->sli4_hba.lpfc_abts_nvmet_sgl_list);
if (phba->cfg_enable_fc4_type & LPFC_ENABLE_FCP)
fcp_xri_cmpl =
list_empty(&phba->sli4_hba.lpfc_abts_scsi_buf_list);
- if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME)
+ if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME) {
nvme_xri_cmpl =
list_empty(&phba->sli4_hba.lpfc_abts_nvme_buf_list);
+ nvmet_xri_cmpl =
+ list_empty(&phba->sli4_hba.lpfc_abts_nvmet_ctx_list);
+ }
while (!fcp_xri_cmpl || !els_xri_cmpl || !nvme_xri_cmpl ||
!nvmet_xri_cmpl) {
@@ -10002,9 +10015,12 @@ lpfc_sli4_xri_exchange_busy_wait(struct lpfc_hba *phba)
msleep(LPFC_XRI_EXCH_BUSY_WAIT_T1);
wait_time += LPFC_XRI_EXCH_BUSY_WAIT_T1;
}
- if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME)
+ if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME) {
nvme_xri_cmpl = list_empty(
&phba->sli4_hba.lpfc_abts_nvme_buf_list);
+ nvmet_xri_cmpl = list_empty(
+ &phba->sli4_hba.lpfc_abts_nvmet_ctx_list);
+ }
if (phba->cfg_enable_fc4_type & LPFC_ENABLE_FCP)
fcp_xri_cmpl = list_empty(
@@ -10013,8 +10029,6 @@ lpfc_sli4_xri_exchange_busy_wait(struct lpfc_hba *phba)
els_xri_cmpl =
list_empty(&phba->sli4_hba.lpfc_abts_els_sgl_list);
- nvmet_xri_cmpl =
- list_empty(&phba->sli4_hba.lpfc_abts_nvmet_sgl_list);
}
}
diff --git a/drivers/scsi/lpfc/lpfc_mbox.c b/drivers/scsi/lpfc/lpfc_mbox.c
index a928f51..ce25a18 100644
--- a/drivers/scsi/lpfc/lpfc_mbox.c
+++ b/drivers/scsi/lpfc/lpfc_mbox.c
@@ -2083,9 +2083,12 @@ lpfc_request_features(struct lpfc_hba *phba, struct lpfcMboxq *mboxq)
if (phba->max_vpi && phba->cfg_enable_npiv)
bf_set(lpfc_mbx_rq_ftr_rq_npiv, &mboxq->u.mqe.un.req_ftrs, 1);
- if (phba->nvmet_support)
+ if (phba->nvmet_support) {
bf_set(lpfc_mbx_rq_ftr_rq_mrqp, &mboxq->u.mqe.un.req_ftrs, 1);
-
+ /* iaab/iaar NOT set for now */
+ bf_set(lpfc_mbx_rq_ftr_rq_iaab, &mboxq->u.mqe.un.req_ftrs, 0);
+ bf_set(lpfc_mbx_rq_ftr_rq_iaar, &mboxq->u.mqe.un.req_ftrs, 0);
+ }
return;
}
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index 1a83ade..a439a50 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -867,15 +867,18 @@ lpfc_nvme_io_cmd_wqe_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pwqeIn,
break;
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_IOERR,
"6081 NVME Completion Protocol Error: "
- "status x%x result x%x placed x%x\n",
+ "xri %x status x%x result x%x "
+ "placed x%x\n",
+ lpfc_ncmd->cur_iocbq.sli4_xritag,
lpfc_ncmd->status, lpfc_ncmd->result,
wcqe->total_data_placed);
break;
default:
out_err:
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_IOERR,
- "6072 NVME Completion Error: "
+ "6072 NVME Completion Error: xri %x "
"status x%x result x%x placed x%x\n",
+ lpfc_ncmd->cur_iocbq.sli4_xritag,
lpfc_ncmd->status, lpfc_ncmd->result,
wcqe->total_data_placed);
nCmd->transferred_length = 0;
@@ -1428,7 +1431,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
phba = vport->phba;
/* Announce entry to new IO submit field. */
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
+ lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME_ABTS,
"6002 Abort Request to rport DID x%06x "
"for nvme_fc_req %p\n",
pnvme_rport->port_id,
@@ -1458,7 +1461,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
/* The remote node has to be ready to send an abort. */
if ((ndlp->nlp_state != NLP_STE_MAPPED_NODE) &&
!(ndlp->nlp_type & NLP_NVME_TARGET)) {
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_ABTS,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6048 rport %p, DID x%06x not ready for "
"IO. State x%x, Type x%x\n",
rport, pnvme_rport->port_id,
@@ -1473,7 +1476,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
/* driver queued commands are in process of being flushed */
if (phba->hba_flag & HBA_NVME_IOQ_FLUSH) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6139 Driver in reset cleanup - flushing "
"NVME Req now. hba_flag x%x\n",
phba->hba_flag);
@@ -1483,13 +1486,13 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
lpfc_nbuf = (struct lpfc_nvme_buf *)pnvme_fcreq->private;
if (!lpfc_nbuf) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6140 NVME IO req has no matching lpfc nvme "
"io buffer. Skipping abort req.\n");
return;
} else if (!lpfc_nbuf->nvmeCmd) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6141 lpfc NVME IO req has no nvme_fcreq "
"io buffer. Skipping abort req.\n");
return;
@@ -1505,7 +1508,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
*/
if (lpfc_nbuf->nvmeCmd != pnvme_fcreq) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6143 NVME req mismatch: "
"lpfc_nbuf %p nvmeCmd %p, "
"pnvme_fcreq %p. Skipping Abort xri x%x\n",
@@ -1517,7 +1520,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
/* Don't abort IOs no longer on the pending queue. */
if (!(nvmereq_wqe->iocb_flag & LPFC_IO_ON_TXCMPLQ)) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6142 NVME IO req %p not queued - skipping "
"abort req xri x%x\n",
pnvme_fcreq, nvmereq_wqe->sli4_xritag);
@@ -1531,7 +1534,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
/* Outstanding abort is in progress */
if (nvmereq_wqe->iocb_flag & LPFC_DRIVER_ABORTED) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6144 Outstanding NVME I/O Abort Request "
"still pending on nvme_fcreq %p, "
"lpfc_ncmd %p xri x%x\n",
@@ -1543,7 +1546,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
abts_buf = __lpfc_sli_get_iocbq(phba);
if (!abts_buf) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6136 No available abort wqes. Skipping "
"Abts req for nvme_fcreq %p xri x%x\n",
pnvme_fcreq, nvmereq_wqe->sli4_xritag);
@@ -1595,7 +1598,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
ret_val = lpfc_sli4_issue_wqe(phba, LPFC_FCP_RING, abts_buf);
spin_unlock_irqrestore(&phba->hbalock, flags);
if (ret_val == IOCB_ERROR) {
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6137 Failed abts issue_wqe with status x%x "
"for nvme_fcreq %p.\n",
ret_val, pnvme_fcreq);
@@ -1603,7 +1606,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
return;
}
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME_ABTS,
"6138 Transport Abort NVME Request Issued for "
"ox_id x%x on reqtag x%x\n",
nvmereq_wqe->sli4_xritag,
@@ -2107,6 +2110,12 @@ lpfc_release_nvme_buf(struct lpfc_hba *phba, struct lpfc_nvme_buf *lpfc_ncmd)
lpfc_ncmd->nonsg_phys = 0;
if (lpfc_ncmd->flags & LPFC_SBUF_XBUSY) {
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6310 XB release deferred for "
+ "ox_id x%x on reqtag x%x\n",
+ lpfc_ncmd->cur_iocbq.sli4_xritag,
+ lpfc_ncmd->cur_iocbq.iotag);
+
spin_lock_irqsave(&phba->sli4_hba.abts_nvme_buf_list_lock,
iflag);
lpfc_ncmd->nvmeCmd = NULL;
@@ -2548,6 +2557,12 @@ lpfc_sli4_nvme_xri_aborted(struct lpfc_hba *phba,
rxid, 1);
lpfc_sli4_abts_err_handler(phba, ndlp, axri);
}
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6311 XRI Aborted xri x%x tag x%x "
+ "released\n",
+ xri, lpfc_ncmd->cur_iocbq.iotag);
+
lpfc_release_nvme_buf(phba, lpfc_ncmd);
if (rrq_empty)
lpfc_worker_wake_up(phba);
@@ -2556,4 +2571,8 @@ lpfc_sli4_nvme_xri_aborted(struct lpfc_hba *phba,
}
spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
spin_unlock_irqrestore(&phba->hbalock, iflag);
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6312 XRI Aborted xri x%x not found\n", xri);
+
}
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index d0dda42..fe939e8 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -71,6 +71,26 @@ static int lpfc_nvmet_unsol_ls_issue_abort(struct lpfc_hba *,
struct lpfc_nvmet_rcv_ctx *,
uint32_t, uint16_t);
+void
+lpfc_nvmet_defer_release(struct lpfc_hba *phba, struct lpfc_nvmet_rcv_ctx *ctxp)
+{
+ unsigned long iflag;
+
+ lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
+ "6313 NVMET Defer ctx release xri x%x flg x%x\n",
+ ctxp->oxid, ctxp->flag);
+
+ spin_lock_irqsave(&phba->sli4_hba.abts_nvme_buf_list_lock, iflag);
+ if (ctxp->flag & LPFC_NVMET_CTX_RLS) {
+ spin_unlock_irqrestore(&phba->sli4_hba.abts_nvme_buf_list_lock,
+ iflag);
+ return;
+ }
+ ctxp->flag |= LPFC_NVMET_CTX_RLS;
+ list_add_tail(&ctxp->list, &phba->sli4_hba.lpfc_abts_nvmet_ctx_list);
+ spin_unlock_irqrestore(&phba->sli4_hba.abts_nvme_buf_list_lock, iflag);
+}
+
/**
* lpfc_nvmet_xmt_ls_rsp_cmp - Completion handler for LS Response
* @phba: Pointer to HBA context object.
@@ -139,6 +159,11 @@ lpfc_nvmet_rq_post(struct lpfc_hba *phba, struct lpfc_nvmet_rcv_ctx *ctxp,
struct lpfc_dmabuf *mp)
{
if (ctxp) {
+ if (ctxp->flag)
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6314 rq_post ctx xri x%x flag x%x\n",
+ ctxp->oxid, ctxp->flag);
+
if (ctxp->txrdy) {
pci_pool_free(phba->txrdy_payload_pool, ctxp->txrdy,
ctxp->txrdy_phys);
@@ -337,39 +362,55 @@ lpfc_nvmet_xmt_fcp_op_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
#endif
ctxp = cmdwqe->context2;
+ ctxp->flag &= ~LPFC_NVMET_IO_INP;
+
rsp = &ctxp->ctx.fcp_req;
op = rsp->op;
- ctxp->flag &= ~LPFC_NVMET_IO_INP;
status = bf_get(lpfc_wcqe_c_status, wcqe);
result = wcqe->parameter;
- if (!phba->targetport)
- goto out;
+ if (phba->targetport)
+ tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
+ else
+ tgtp = NULL;
lpfc_nvmeio_data(phba, "NVMET FCP CMPL: xri x%x op x%x status x%x\n",
ctxp->oxid, op, status);
- tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
if (status) {
rsp->fcp_error = NVME_SC_DATA_XFER_ERROR;
rsp->transferred_length = 0;
- atomic_inc(&tgtp->xmt_fcp_rsp_error);
+ if (tgtp)
+ atomic_inc(&tgtp->xmt_fcp_rsp_error);
+
+ /* pick up SLI4 exhange busy condition */
+ if (bf_get(lpfc_wcqe_c_xb, wcqe)) {
+ ctxp->flag |= LPFC_NVMET_XBUSY;
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6315 IO Cmpl XBUSY: xri x%x: %x/%x\n",
+ ctxp->oxid, status, result);
+ } else {
+ ctxp->flag &= ~LPFC_NVMET_XBUSY;
+ }
+
} else {
rsp->fcp_error = NVME_SC_SUCCESS;
if (op == NVMET_FCOP_RSP)
rsp->transferred_length = rsp->rsplen;
else
rsp->transferred_length = rsp->transfer_length;
- atomic_inc(&tgtp->xmt_fcp_rsp_cmpl);
+ if (tgtp)
+ atomic_inc(&tgtp->xmt_fcp_rsp_cmpl);
}
-out:
if ((op == NVMET_FCOP_READDATA_RSP) ||
(op == NVMET_FCOP_RSP)) {
/* Sanity check */
ctxp->state = LPFC_NVMET_STE_DONE;
ctxp->entry_cnt++;
+
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
if (phba->ktime_on) {
if (rsp->op == NVMET_FCOP_READDATA_RSP) {
@@ -542,10 +583,11 @@ lpfc_nvmet_xmt_fcp_op(struct nvmet_fc_target_port *tgtport,
#endif
/* Sanity check */
- if (ctxp->state == LPFC_NVMET_STE_ABORT) {
+ if ((ctxp->flag & LPFC_NVMET_ABTS_RCV) ||
+ (ctxp->state == LPFC_NVMET_STE_ABORT)) {
atomic_inc(&lpfc_nvmep->xmt_fcp_drop);
lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
- "6102 Bad state IO x%x aborted\n",
+ "6102 IO xri x%x aborted\n",
ctxp->oxid);
rc = -ENXIO;
goto aerr;
@@ -615,16 +657,27 @@ lpfc_nvmet_xmt_fcp_abort(struct nvmet_fc_target_port *tgtport,
struct lpfc_nvmet_rcv_ctx *ctxp =
container_of(req, struct lpfc_nvmet_rcv_ctx, ctx.fcp_req);
struct lpfc_hba *phba = ctxp->phba;
+ unsigned long flags;
lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
- "6103 Abort op: oxri x%x %d cnt %d\n",
- ctxp->oxid, ctxp->state, ctxp->entry_cnt);
+ "6103 Abort op: oxri x%x flg x%x cnt %d\n",
+ ctxp->oxid, ctxp->flag, ctxp->entry_cnt);
- lpfc_nvmeio_data(phba, "NVMET FCP ABRT: xri x%x state x%x cnt x%x\n",
- ctxp->oxid, ctxp->state, ctxp->entry_cnt);
+ lpfc_nvmeio_data(phba, "NVMET FCP ABRT: "
+ "xri x%x flg x%x cnt x%x\n",
+ ctxp->oxid, ctxp->flag, ctxp->entry_cnt);
atomic_inc(&lpfc_nvmep->xmt_fcp_abort);
ctxp->entry_cnt++;
+ spin_lock_irqsave(&ctxp->ctxlock, flags);
+
+ /* Since iaab/iaar are NOT set, we need to check
+ * if the firmware is in process of aborting IO
+ */
+ if (ctxp->flag & LPFC_NVMET_XBUSY) {
+ spin_unlock_irqrestore(&ctxp->ctxlock, flags);
+ return;
+ }
ctxp->flag |= LPFC_NVMET_ABORT_OP;
if (ctxp->flag & LPFC_NVMET_IO_INP)
lpfc_nvmet_sol_fcp_issue_abort(phba, ctxp, ctxp->sid,
@@ -632,13 +685,13 @@ lpfc_nvmet_xmt_fcp_abort(struct nvmet_fc_target_port *tgtport,
else
lpfc_nvmet_unsol_fcp_issue_abort(phba, ctxp, ctxp->sid,
ctxp->oxid);
+ spin_unlock_irqrestore(&ctxp->ctxlock, flags);
}
static void
lpfc_nvmet_xmt_fcp_release(struct nvmet_fc_target_port *tgtport,
struct nvmefc_tgt_fcp_req *rsp)
{
- struct lpfc_nvmet_tgtport *lpfc_nvmep = tgtport->private;
struct lpfc_nvmet_rcv_ctx *ctxp =
container_of(rsp, struct lpfc_nvmet_rcv_ctx, ctx.fcp_req);
struct lpfc_hba *phba = ctxp->phba;
@@ -646,27 +699,20 @@ lpfc_nvmet_xmt_fcp_release(struct nvmet_fc_target_port *tgtport,
bool aborting = false;
spin_lock_irqsave(&ctxp->ctxlock, flags);
- if (ctxp->flag & LPFC_NVMET_ABORT_OP) {
+ if ((ctxp->flag & LPFC_NVMET_ABORT_OP) ||
+ (ctxp->flag & LPFC_NVMET_XBUSY)) {
aborting = true;
- ctxp->flag |= LPFC_NVMET_CTX_RLS;
- }
- spin_unlock_irqrestore(&ctxp->ctxlock, flags);
-
- if (aborting)
/* let the abort path do the real release */
- return;
-
- /* Sanity check */
- if (ctxp->state != LPFC_NVMET_STE_DONE) {
- atomic_inc(&lpfc_nvmep->xmt_fcp_drop);
- lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
- "6117 Bad state IO x%x aborted\n",
- ctxp->oxid);
+ lpfc_nvmet_defer_release(phba, ctxp);
}
+ spin_unlock_irqrestore(&ctxp->ctxlock, flags);
lpfc_nvmeio_data(phba, "NVMET FCP FREE: xri x%x ste %d\n", ctxp->oxid,
ctxp->state, 0);
+ if (aborting)
+ return;
+
lpfc_nvmet_rq_post(phba, ctxp, &ctxp->rqb_buffer->hbuf);
}
@@ -801,7 +847,122 @@ void
lpfc_sli4_nvmet_xri_aborted(struct lpfc_hba *phba,
struct sli4_wcqe_xri_aborted *axri)
{
- /* TODO: work in progress */
+ uint16_t xri = bf_get(lpfc_wcqe_xa_xri, axri);
+ uint16_t rxid = bf_get(lpfc_wcqe_xa_remote_xid, axri);
+ struct lpfc_nvmet_rcv_ctx *ctxp, *next_ctxp;
+ struct lpfc_nodelist *ndlp;
+ unsigned long iflag = 0;
+ int rrq_empty = 0;
+ bool released = false;
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6317 XB aborted xri x%x rxid x%x\n", xri, rxid);
+
+ if (!(phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME))
+ return;
+ spin_lock_irqsave(&phba->hbalock, iflag);
+ spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ list_for_each_entry_safe(ctxp, next_ctxp,
+ &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
+ list) {
+ if (ctxp->rqb_buffer->sglq->sli4_xritag == xri) {
+ /* Check if we already received a free context call
+ * and we have completed processing an abort situation.
+ */
+ if ((ctxp->flag & LPFC_NVMET_CTX_RLS) &&
+ !(ctxp->flag & LPFC_NVMET_ABORT_OP)) {
+ list_del(&ctxp->list);
+ released = true;
+ }
+ ctxp->flag &= ~LPFC_NVMET_XBUSY;
+ spin_unlock(
+ &phba->sli4_hba.abts_nvme_buf_list_lock);
+
+ rrq_empty = list_empty(&phba->active_rrq_list);
+ spin_unlock_irqrestore(&phba->hbalock, iflag);
+ ndlp = lpfc_findnode_did(phba->pport, ctxp->sid);
+ if (ndlp && NLP_CHK_NODE_ACT(ndlp) &&
+ ((ndlp->nlp_state == NLP_STE_UNMAPPED_NODE) ||
+ (ndlp->nlp_state == NLP_STE_MAPPED_NODE))) {
+ lpfc_set_rrq_active(
+ phba, ndlp,
+ ctxp->rqb_buffer->sglq->sli4_lxritag,
+ rxid, 1);
+ lpfc_sli4_abts_err_handler(phba, ndlp, axri);
+ }
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6318 XB aborted %x flg x%x (%x)\n",
+ ctxp->oxid, ctxp->flag, released);
+ if (released)
+ lpfc_nvmet_rq_post(phba, ctxp,
+ &ctxp->rqb_buffer->hbuf);
+ if (rrq_empty)
+ lpfc_worker_wake_up(phba);
+ return;
+ }
+ }
+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ spin_unlock_irqrestore(&phba->hbalock, iflag);
+}
+
+int
+lpfc_nvmet_rcv_unsol_abort(struct lpfc_vport *vport,
+ struct fc_frame_header *fc_hdr)
+
+{
+#if (IS_ENABLED(CONFIG_NVME_TARGET_FC))
+ struct lpfc_hba *phba = vport->phba;
+ struct lpfc_nvmet_rcv_ctx *ctxp, *next_ctxp;
+ struct nvmefc_tgt_fcp_req *rsp;
+ uint16_t xri;
+ unsigned long iflag = 0;
+
+ xri = be16_to_cpu(fc_hdr->fh_ox_id);
+
+ spin_lock_irqsave(&phba->hbalock, iflag);
+ spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ list_for_each_entry_safe(ctxp, next_ctxp,
+ &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
+ list) {
+ if (ctxp->rqb_buffer->sglq->sli4_xritag == xri) {
+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ spin_unlock_irqrestore(&phba->hbalock, iflag);
+
+ spin_lock_irqsave(&ctxp->ctxlock, iflag);
+ ctxp->flag |= LPFC_NVMET_ABTS_RCV;
+ spin_unlock_irqrestore(&ctxp->ctxlock, iflag);
+
+ lpfc_nvmeio_data(phba,
+ "NVMET ABTS RCV: "
+ "xri x%x CPU %02x rjt %d\n",
+ xri, smp_processor_id(), 0);
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6319 NVMET Rcv ABTS:acc xri x%x\n",
+ xri);
+
+ rsp = &ctxp->ctx.fcp_req;
+ nvmet_fc_rcv_fcp_abort(phba->targetport, rsp);
+
+ /* Respond with BA_ACC accordingly */
+ lpfc_sli4_seq_abort_rsp(vport, fc_hdr, 1);
+ return 0;
+ }
+ }
+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ spin_unlock_irqrestore(&phba->hbalock, iflag);
+
+ lpfc_nvmeio_data(phba, "NVMET ABTS RCV: xri x%x CPU %02x rjt %d\n",
+ xri, smp_processor_id(), 1);
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6320 NVMET Rcv ABTS:rjt xri x%x\n", xri);
+
+ /* Respond with BA_RJT accordingly */
+ lpfc_sli4_seq_abort_rsp(vport, fc_hdr, 0);
+ return 0;
+#endif
}
void
@@ -1655,18 +1816,27 @@ lpfc_nvmet_sol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
atomic_inc(&tgtp->xmt_abort_cmpl);
- lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
- "6165 Abort cmpl: xri x%x WCQE: %08x %08x %08x %08x\n",
- ctxp->oxid, wcqe->word0, wcqe->total_data_placed,
- result, wcqe->word3);
-
ctxp->state = LPFC_NVMET_STE_DONE;
+
+ /* Check if we already received a free context call
+ * and we have completed processing an abort situation.
+ */
spin_lock_irqsave(&ctxp->ctxlock, flags);
- if (ctxp->flag & LPFC_NVMET_CTX_RLS)
+ if ((ctxp->flag & LPFC_NVMET_CTX_RLS) &&
+ !(ctxp->flag & LPFC_NVMET_XBUSY)) {
+ list_del(&ctxp->list);
released = true;
+ }
ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
spin_unlock_irqrestore(&ctxp->ctxlock, flags);
+ lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
+ "6165 ABORT cmpl: xri x%x flg x%x (%d) "
+ "WCQE: %08x %08x %08x %08x\n",
+ ctxp->oxid, ctxp->flag, released,
+ wcqe->word0, wcqe->total_data_placed,
+ result, wcqe->word3);
+
/*
* if transport has released ctx, then can reuse it. Otherwise,
* will be recycled by transport release call.
@@ -1677,10 +1847,15 @@ lpfc_nvmet_sol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
cmdwqe->context2 = NULL;
cmdwqe->context3 = NULL;
lpfc_sli_release_iocbq(phba, cmdwqe);
+
+ /* Since iaab/iaar are NOT set, there is no work left.
+ * For LPFC_NVMET_XBUSY, lpfc_sli4_nvmet_xri_aborted
+ * should have been called already.
+ */
}
/**
- * lpfc_nvmet_xmt_fcp_abort_cmp - Completion handler for ABTS
+ * lpfc_nvmet_unsol_fcp_abort_cmp - Completion handler for ABTS
* @phba: Pointer to HBA context object.
* @cmdwqe: Pointer to driver command WQE object.
* @wcqe: Pointer to driver response CQE object.
@@ -1690,8 +1865,8 @@ lpfc_nvmet_sol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
* The function frees memory resources used for the NVME commands.
**/
static void
-lpfc_nvmet_xmt_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
- struct lpfc_wcqe_complete *wcqe)
+lpfc_nvmet_unsol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
+ struct lpfc_wcqe_complete *wcqe)
{
struct lpfc_nvmet_rcv_ctx *ctxp;
struct lpfc_nvmet_tgtport *tgtp;
@@ -1706,35 +1881,54 @@ lpfc_nvmet_xmt_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
atomic_inc(&tgtp->xmt_abort_cmpl);
+ if (!ctxp) {
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6070 ABTS cmpl: WCQE: %08x %08x %08x %08x\n",
+ wcqe->word0, wcqe->total_data_placed,
+ result, wcqe->word3);
+ return;
+ }
+
+ /* Sanity check */
+ if (ctxp->state != LPFC_NVMET_STE_ABORT) {
+ lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
+ "6112 ABTS Wrong state:%d oxid x%x\n",
+ ctxp->state, ctxp->oxid);
+ }
+
+ /* Check if we already received a free context call
+ * and we have completed processing an abort situation.
+ */
+ ctxp->state = LPFC_NVMET_STE_DONE;
+ spin_lock_irqsave(&ctxp->ctxlock, flags);
+ if ((ctxp->flag & LPFC_NVMET_CTX_RLS) &&
+ !(ctxp->flag & LPFC_NVMET_XBUSY)) {
+ list_del(&ctxp->list);
+ released = true;
+ }
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
+ spin_unlock_irqrestore(&ctxp->ctxlock, flags);
+
lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
- "6070 Abort cmpl: ctx %p WCQE: %08x %08x %08x %08x\n",
- ctxp, wcqe->word0, wcqe->total_data_placed,
+ "6316 ABTS cmpl xri x%x flg x%x (%x) "
+ "WCQE: %08x %08x %08x %08x\n",
+ ctxp->oxid, ctxp->flag, released,
+ wcqe->word0, wcqe->total_data_placed,
result, wcqe->word3);
+ /*
+ * if transport has released ctx, then can reuse it. Otherwise,
+ * will be recycled by transport release call.
+ */
+ if (released)
+ lpfc_nvmet_rq_post(phba, ctxp, &ctxp->rqb_buffer->hbuf);
- if (ctxp) {
- /* Sanity check */
- if (ctxp->state != LPFC_NVMET_STE_ABORT) {
- lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
- "6112 ABORT Wrong state:%d oxid x%x\n",
- ctxp->state, ctxp->oxid);
- }
- ctxp->state = LPFC_NVMET_STE_DONE;
- spin_lock_irqsave(&ctxp->ctxlock, flags);
- if (ctxp->flag & LPFC_NVMET_CTX_RLS)
- released = true;
- ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
- spin_unlock_irqrestore(&ctxp->ctxlock, flags);
-
- /*
- * if transport has released ctx, then can reuse it. Otherwise,
- * will be recycled by transport release call.
- */
- if (released)
- lpfc_nvmet_rq_post(phba, ctxp, &ctxp->rqb_buffer->hbuf);
+ cmdwqe->context2 = NULL;
+ cmdwqe->context3 = NULL;
- cmdwqe->context2 = NULL;
- cmdwqe->context3 = NULL;
- }
+ /* Since iaab/iaar are NOT set, there is no work left.
+ * For LPFC_NVMET_XBUSY, lpfc_sli4_nvmet_xri_aborted
+ * should have been called already.
+ */
}
/**
@@ -1787,10 +1981,14 @@ lpfc_nvmet_unsol_issue_abort(struct lpfc_hba *phba,
struct lpfc_nodelist *ndlp;
lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
- "6067 Abort: sid %x xri x%x/x%x\n",
+ "6067 ABTS: sid %x xri x%x/x%x\n",
sid, xri, ctxp->wqeq->sli4_xritag);
tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
+ if (!ctxp->wqeq) {
+ ctxp->wqeq = ctxp->rqb_buffer->iocbq;
+ ctxp->wqeq->hba_wqidx = 0;
+ }
ndlp = lpfc_findnode_did(phba->pport, sid);
if (!ndlp || !NLP_CHK_NODE_ACT(ndlp) ||
@@ -1896,10 +2094,11 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
(ndlp->nlp_state != NLP_STE_MAPPED_NODE))) {
atomic_inc(&tgtp->xmt_abort_rsp_error);
lpfc_printf_log(phba, KERN_WARNING, LOG_NVME_ABTS,
- "6160 Drop ABTS - wrong NDLP state x%x.\n",
+ "6160 Drop ABORT - wrong NDLP state x%x.\n",
(ndlp) ? ndlp->nlp_state : NLP_STE_MAX_STATE);
/* No failure to an ABTS request. */
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
return 0;
}
@@ -1907,9 +2106,10 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
ctxp->abort_wqeq = lpfc_sli_get_iocbq(phba);
if (!ctxp->abort_wqeq) {
lpfc_printf_log(phba, KERN_WARNING, LOG_NVME_ABTS,
- "6161 Abort failed: No wqeqs: "
+ "6161 ABORT failed: No wqeqs: "
"xri: x%x\n", ctxp->oxid);
/* No failure to an ABTS request. */
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
return 0;
}
abts_wqeq = ctxp->abort_wqeq;
@@ -1917,8 +2117,8 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
ctxp->state = LPFC_NVMET_STE_ABORT;
/* Announce entry to new IO submit field. */
- lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
- "6162 Abort Request to rport DID x%06x "
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6162 ABORT Request to rport DID x%06x "
"for xri x%x x%x\n",
ctxp->sid, ctxp->oxid, ctxp->wqeq->sli4_xritag);
@@ -1934,6 +2134,7 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
"NVME Req now. hba_flag x%x oxid x%x\n",
phba->hba_flag, ctxp->oxid);
lpfc_sli_release_iocbq(phba, abts_wqeq);
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
return 0;
}
@@ -1945,6 +2146,7 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
"still pending on oxid x%x\n",
ctxp->oxid);
lpfc_sli_release_iocbq(phba, abts_wqeq);
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
return 0;
}
@@ -1992,9 +2194,10 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
if (rc == WQE_SUCCESS)
return 0;
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
lpfc_sli_release_iocbq(phba, abts_wqeq);
- lpfc_printf_log(phba, KERN_ERR, LOG_NVME,
- "6166 Failed abts issue_wqe with status x%x "
+ lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
+ "6166 Failed ABORT issue_wqe with status x%x "
"for oxid x%x.\n",
rc, ctxp->oxid);
return 1;
@@ -2023,8 +2226,8 @@ lpfc_nvmet_unsol_fcp_issue_abort(struct lpfc_hba *phba,
spin_lock_irqsave(&phba->hbalock, flags);
abts_wqeq = ctxp->wqeq;
- abts_wqeq->wqe_cmpl = lpfc_nvmet_xmt_fcp_abort_cmp;
- abts_wqeq->iocb_cmpl = 0;
+ abts_wqeq->wqe_cmpl = lpfc_nvmet_unsol_fcp_abort_cmp;
+ abts_wqeq->iocb_cmpl = NULL;
abts_wqeq->iocb_flag |= LPFC_IO_NVMET;
rc = lpfc_sli4_issue_wqe(phba, LPFC_FCP_RING, abts_wqeq);
spin_unlock_irqrestore(&phba->hbalock, flags);
@@ -2034,7 +2237,7 @@ lpfc_nvmet_unsol_fcp_issue_abort(struct lpfc_hba *phba,
}
aerr:
- lpfc_nvmet_rq_post(phba, ctxp, &ctxp->rqb_buffer->hbuf);
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
atomic_inc(&tgtp->xmt_abort_rsp_error);
lpfc_printf_log(phba, KERN_WARNING, LOG_NVME_ABTS,
"6135 Failed to Issue ABTS for oxid x%x. Status x%x\n",
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.h b/drivers/scsi/lpfc/lpfc_nvmet.h
index d8bac4c..128759f 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.h
+++ b/drivers/scsi/lpfc/lpfc_nvmet.h
@@ -75,6 +75,7 @@ struct lpfc_nvmet_rcv_ctx {
struct nvmefc_tgt_ls_req ls_req;
struct nvmefc_tgt_fcp_req fcp_req;
} ctx;
+ struct list_head list;
struct lpfc_hba *phba;
struct lpfc_iocbq *wqeq;
struct lpfc_iocbq *abort_wqeq;
@@ -96,10 +97,11 @@ struct lpfc_nvmet_rcv_ctx {
#define LPFC_NVMET_STE_RSP 4
#define LPFC_NVMET_STE_DONE 5
uint16_t flag;
-#define LPFC_NVMET_IO_INP 0x1
-#define LPFC_NVMET_ABORT_OP 0x2
-#define LPFC_NVMET_CTX_RLS 0x4
-
+#define LPFC_NVMET_IO_INP 0x1 /* IO is in progress on exchange */
+#define LPFC_NVMET_ABORT_OP 0x2 /* Abort WQE issued on exchange */
+#define LPFC_NVMET_XBUSY 0x4 /* XB bit set on IO cmpl */
+#define LPFC_NVMET_CTX_RLS 0x8 /* ctx free requested */
+#define LPFC_NVMET_ABTS_RCV 0x10 /* ABTS received on exchange */
struct rqb_dmabuf *rqb_buffer;
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index a6adc2b..cf19f49 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -16521,7 +16521,7 @@ lpfc_sli4_xri_inrange(struct lpfc_hba *phba,
* This function sends a basic response to a previous unsol sequence abort
* event after aborting the sequence handling.
**/
-static void
+void
lpfc_sli4_seq_abort_rsp(struct lpfc_vport *vport,
struct fc_frame_header *fc_hdr, bool aborted)
{
@@ -16697,6 +16697,11 @@ lpfc_sli4_handle_unsol_abort(struct lpfc_vport *vport,
}
lpfc_in_buf_free(phba, &dmabuf->dbuf);
+ if (phba->nvmet_support) {
+ lpfc_nvmet_rcv_unsol_abort(vport, &fc_hdr);
+ return;
+ }
+
/* Respond with BA_ACC or BA_RJT accordingly */
lpfc_sli4_seq_abort_rsp(vport, &fc_hdr, aborted);
}
diff --git a/drivers/scsi/lpfc/lpfc_sli4.h b/drivers/scsi/lpfc/lpfc_sli4.h
index 710458c..da46471 100644
--- a/drivers/scsi/lpfc/lpfc_sli4.h
+++ b/drivers/scsi/lpfc/lpfc_sli4.h
@@ -620,7 +620,7 @@ struct lpfc_sli4_hba {
struct list_head lpfc_els_sgl_list;
struct list_head lpfc_abts_els_sgl_list;
struct list_head lpfc_nvmet_sgl_list;
- struct list_head lpfc_abts_nvmet_sgl_list;
+ struct list_head lpfc_abts_nvmet_ctx_list;
struct list_head lpfc_abts_scsi_buf_list;
struct list_head lpfc_abts_nvme_buf_list;
struct lpfc_sglq **lpfc_sglq_active_list;
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 20/22] Update ABORT processing for NVMET.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
The driver with nvme had this routine stubbed.
Right now XRI_ABORTED_CQE is not handled and the FC NVMET
Transport has a new API for the driver.
Missing code path, new NVME abort API
Update ABORT processing for NVMET
There are 3 new FC NVMET Transport API/ template routines for NVMET:
lpfc_nvmet_xmt_fcp_release
This NVMET template callback routine called to release context
associated with an IO This routine is ALWAYS called last, even
if the IO was aborted or completed in error.
lpfc_nvmet_xmt_fcp_abort
This NVMET template callback routine called to abort an exchange that
has an IO in progress
nvmet_fc_rcv_fcp_req
When the lpfc driver receives an ABTS, this NVME FC transport layer
callback routine is called. For this case there are 2 paths thru the
driver: the driver either has an outstanding exchange / context for the
XRI to be aborted or not. If not, a BA_RJT is issued otherwise a BA_ACC
NVMET Driver abort paths:
There are 2 paths for aborting an IO. The first one is we receive an IO and
decide not to process it because of lack of resources. An unsolicated ABTS
is immediately sent back to the initiator as a response.
lpfc_nvmet_unsol_fcp_buffer
lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE)
The second one is we sent the IO up to the NVMET transport layer to
process, and for some reason the NVME Transport layer decided to abort the
IO before it completes all its phases. For this case there are 2 paths
thru the driver:
the driver either has an outstanding TSEND/TRECEIVE/TRSP WQE or no
outstanding WQEs are present for the exchange / context.
lpfc_nvmet_xmt_fcp_abort
if (LPFC_NVMET_IO_INP)
lpfc_nvmet_sol_fcp_issue_abort (ABORT_WQE)
lpfc_nvmet_sol_fcp_abort_cmp
else
lpfc_nvmet_unsol_fcp_issue_abort
lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE)
lpfc_nvmet_unsol_fcp_abort_cmp
Context flags:
LPFC_NVMET_IOP - his flag signifies an IO is in progress on the exchange.
LPFC_NVMET_XBUSY - this flag indicates the IO completed but the firmware
is still busy with the corresponding exchange. The exchange should not be
reused until after a XRI_ABORTED_CQE is received for that exchange.
LPFC_NVMET_ABORT_OP - this flag signifies an ABORT_WQE was issued on the
exchange.
LPFC_NVMET_CTX_RLS - this flag signifies a context free was requested,
but we are deferring it due to an XBUSY or ABORT in progress.
A ctxlock is added to the context structure that is used whenever these
flags are set/read within the context of an IO.
The LPFC_NVMET_CTX_RLS flag is only set in the defer_relase routine when
the transport has resolved all IO associated with the buffer. The flag is
cleared when the CTX is associated with a new IO.
An exchange can has both an LPFC_NVMET_XBUSY and a LPFC_NVMET_ABORT_OP
condition active simultaneously. Both conditions must complete before the
exchange is freed.
When the abort callback (lpfc_nvmet_xmt_fcp_abort) is envoked:
If there is an outstanding IO, the driver will issue an ABORT_WQE. This
should result in 3 completions for the exchange:
1) IO cmpl with XB bit set
2) Abort WQE cmpl
3) XRI_ABORTED_CQE cmpl
For this scenerio, after completion #1, the NVMET Transport IO rsp
callback is called. After completion #2, no action is taken with respect
to the exchange / context. After completion #3, the exchange context is
free for re-use on another IO.
If there is no outstanding activity on the exchange, the driver will send a
ABTS to the Initiator. Upon completion of this WQE, the exchange / context
is freed for re-use on another IO.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_crtn.h | 7 +
drivers/scsi/lpfc/lpfc_debugfs.c | 55 ++++--
drivers/scsi/lpfc/lpfc_hw4.h | 3 +
drivers/scsi/lpfc/lpfc_init.c | 52 +++---
drivers/scsi/lpfc/lpfc_mbox.c | 7 +-
drivers/scsi/lpfc/lpfc_nvme.c | 45 +++--
drivers/scsi/lpfc/lpfc_nvmet.c | 349 +++++++++++++++++++++++++++++++--------
drivers/scsi/lpfc/lpfc_nvmet.h | 10 +-
drivers/scsi/lpfc/lpfc_sli.c | 7 +-
drivers/scsi/lpfc/lpfc_sli4.h | 2 +-
10 files changed, 412 insertions(+), 125 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_crtn.h b/drivers/scsi/lpfc/lpfc_crtn.h
index d859aff..24da922 100644
--- a/drivers/scsi/lpfc/lpfc_crtn.h
+++ b/drivers/scsi/lpfc/lpfc_crtn.h
@@ -24,6 +24,7 @@ typedef int (*node_filter)(struct lpfc_nodelist *, void *);
struct fc_rport;
struct fc_frame_header;
+struct lpfc_nvmet_rcv_ctx;
void lpfc_down_link(struct lpfc_hba *, LPFC_MBOXQ_t *);
void lpfc_sli_read_link_ste(struct lpfc_hba *);
void lpfc_dump_mem(struct lpfc_hba *, LPFC_MBOXQ_t *, uint16_t, uint16_t);
@@ -245,6 +246,10 @@ struct hbq_dmabuf *lpfc_sli4_rb_alloc(struct lpfc_hba *);
void lpfc_sli4_rb_free(struct lpfc_hba *, struct hbq_dmabuf *);
struct rqb_dmabuf *lpfc_sli4_nvmet_alloc(struct lpfc_hba *phba);
void lpfc_sli4_nvmet_free(struct lpfc_hba *phba, struct rqb_dmabuf *dmab);
+void lpfc_nvmet_rq_post(struct lpfc_hba *phba, struct lpfc_nvmet_rcv_ctx *ctxp,
+ struct lpfc_dmabuf *mp);
+int lpfc_nvmet_rcv_unsol_abort(struct lpfc_vport *vport,
+ struct fc_frame_header *fc_hdr);
void lpfc_sli4_build_dflt_fcf_record(struct lpfc_hba *, struct fcf_record *,
uint16_t);
int lpfc_sli4_rq_put(struct lpfc_queue *hq, struct lpfc_queue *dq,
@@ -302,6 +307,8 @@ int lpfc_sli_check_eratt(struct lpfc_hba *);
void lpfc_sli_handle_slow_ring_event(struct lpfc_hba *,
struct lpfc_sli_ring *, uint32_t);
void lpfc_sli4_handle_received_buffer(struct lpfc_hba *, struct hbq_dmabuf *);
+void lpfc_sli4_seq_abort_rsp(struct lpfc_vport *vport,
+ struct fc_frame_header *fc_hdr, bool aborted);
void lpfc_sli_def_mbox_cmpl(struct lpfc_hba *, LPFC_MBOXQ_t *);
void lpfc_sli4_unreg_rpi_cmpl_clr(struct lpfc_hba *, LPFC_MBOXQ_t *);
int lpfc_sli_issue_iocb(struct lpfc_hba *, uint32_t,
diff --git a/drivers/scsi/lpfc/lpfc_debugfs.c b/drivers/scsi/lpfc/lpfc_debugfs.c
index 55a8d8f..4d6cd7a 100644
--- a/drivers/scsi/lpfc/lpfc_debugfs.c
+++ b/drivers/scsi/lpfc/lpfc_debugfs.c
@@ -745,73 +745,102 @@ lpfc_debugfs_nvmestat_data(struct lpfc_vport *vport, char *buf, int size)
{
struct lpfc_hba *phba = vport->phba;
struct lpfc_nvmet_tgtport *tgtp;
+ struct lpfc_nvmet_rcv_ctx *ctxp, *next_ctxp;
int len = 0;
+ int cnt;
if (phba->nvmet_support) {
if (!phba->targetport)
return len;
tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"\nNVME Targetport Statistics\n");
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"LS: Rcv %08x Drop %08x Abort %08x\n",
atomic_read(&tgtp->rcv_ls_req_in),
atomic_read(&tgtp->rcv_ls_req_drop),
atomic_read(&tgtp->xmt_ls_abort));
if (atomic_read(&tgtp->rcv_ls_req_in) !=
atomic_read(&tgtp->rcv_ls_req_out)) {
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"Rcv LS: in %08x != out %08x\n",
atomic_read(&tgtp->rcv_ls_req_in),
atomic_read(&tgtp->rcv_ls_req_out));
}
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"LS: Xmt %08x Drop %08x Cmpl %08x Err %08x\n",
atomic_read(&tgtp->xmt_ls_rsp),
atomic_read(&tgtp->xmt_ls_drop),
atomic_read(&tgtp->xmt_ls_rsp_cmpl),
atomic_read(&tgtp->xmt_ls_rsp_error));
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"FCP: Rcv %08x Drop %08x\n",
atomic_read(&tgtp->rcv_fcp_cmd_in),
atomic_read(&tgtp->rcv_fcp_cmd_drop));
if (atomic_read(&tgtp->rcv_fcp_cmd_in) !=
atomic_read(&tgtp->rcv_fcp_cmd_out)) {
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"Rcv FCP: in %08x != out %08x\n",
atomic_read(&tgtp->rcv_fcp_cmd_in),
atomic_read(&tgtp->rcv_fcp_cmd_out));
}
- len += snprintf(buf+len, size-len,
- "FCP Rsp: read %08x readrsp %08x write %08x rsp %08x\n",
+ len += snprintf(buf + len, size - len,
+ "FCP Rsp: read %08x readrsp %08x "
+ "write %08x rsp %08x\n",
atomic_read(&tgtp->xmt_fcp_read),
atomic_read(&tgtp->xmt_fcp_read_rsp),
atomic_read(&tgtp->xmt_fcp_write),
atomic_read(&tgtp->xmt_fcp_rsp));
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"FCP Rsp: abort %08x drop %08x\n",
atomic_read(&tgtp->xmt_fcp_abort),
atomic_read(&tgtp->xmt_fcp_drop));
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"FCP Rsp Cmpl: %08x err %08x drop %08x\n",
atomic_read(&tgtp->xmt_fcp_rsp_cmpl),
atomic_read(&tgtp->xmt_fcp_rsp_error),
atomic_read(&tgtp->xmt_fcp_rsp_drop));
- len += snprintf(buf+len, size-len,
+ len += snprintf(buf + len, size - len,
"ABORT: Xmt %08x Err %08x Cmpl %08x",
atomic_read(&tgtp->xmt_abort_rsp),
atomic_read(&tgtp->xmt_abort_rsp_error),
atomic_read(&tgtp->xmt_abort_cmpl));
- len += snprintf(buf+len, size-len, "\n");
+ len += snprintf(buf + len, size - len, "\n");
+
+ cnt = 0;
+ spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ list_for_each_entry_safe(ctxp, next_ctxp,
+ &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
+ list) {
+ cnt++;
+ }
+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ if (cnt) {
+ len += snprintf(buf + len, size - len,
+ "ABORT: %d ctx entries\n", cnt);
+ spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ list_for_each_entry_safe(ctxp, next_ctxp,
+ &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
+ list) {
+ if (len >= (size - LPFC_DEBUG_OUT_LINE_SZ))
+ break;
+ len += snprintf(buf + len, size - len,
+ "Entry: oxid %x state %x "
+ "flag %x\n",
+ ctxp->oxid, ctxp->state,
+ ctxp->flag);
+ }
+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ }
} else {
if (!(phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME))
return len;
diff --git a/drivers/scsi/lpfc/lpfc_hw4.h b/drivers/scsi/lpfc/lpfc_hw4.h
index 90499f9..1d12f2b 100644
--- a/drivers/scsi/lpfc/lpfc_hw4.h
+++ b/drivers/scsi/lpfc/lpfc_hw4.h
@@ -2720,6 +2720,9 @@ struct lpfc_mbx_request_features {
#define lpfc_mbx_rq_ftr_rq_ifip_SHIFT 7
#define lpfc_mbx_rq_ftr_rq_ifip_MASK 0x00000001
#define lpfc_mbx_rq_ftr_rq_ifip_WORD word2
+#define lpfc_mbx_rq_ftr_rq_iaar_SHIFT 9
+#define lpfc_mbx_rq_ftr_rq_iaar_MASK 0x00000001
+#define lpfc_mbx_rq_ftr_rq_iaar_WORD word2
#define lpfc_mbx_rq_ftr_rq_perfh_SHIFT 11
#define lpfc_mbx_rq_ftr_rq_perfh_MASK 0x00000001
#define lpfc_mbx_rq_ftr_rq_perfh_WORD word2
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 8acacdf..e084a66 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -42,6 +42,10 @@
#include <scsi/scsi_device.h>
#include <scsi/scsi_host.h>
#include <scsi/scsi_transport_fc.h>
+#include <scsi/scsi_tcq.h>
+#include <scsi/fc/fc_fs.h>
+
+#include <linux/nvme-fc-driver.h>
#include "lpfc_hw4.h"
#include "lpfc_hw.h"
@@ -52,6 +56,7 @@
#include "lpfc.h"
#include "lpfc_scsi.h"
#include "lpfc_nvme.h"
+#include "lpfc_nvmet.h"
#include "lpfc_logmsg.h"
#include "lpfc_crtn.h"
#include "lpfc_vport.h"
@@ -1022,8 +1027,10 @@ static int
lpfc_hba_down_post_s4(struct lpfc_hba *phba)
{
struct lpfc_scsi_buf *psb, *psb_next;
+ struct lpfc_nvmet_rcv_ctx *ctxp, *ctxp_next;
LIST_HEAD(aborts);
LIST_HEAD(nvme_aborts);
+ LIST_HEAD(nvmet_aborts);
unsigned long iflag = 0;
struct lpfc_sglq *sglq_entry = NULL;
@@ -1046,16 +1053,10 @@ lpfc_hba_down_post_s4(struct lpfc_hba *phba)
list_for_each_entry(sglq_entry,
&phba->sli4_hba.lpfc_abts_els_sgl_list, list)
sglq_entry->state = SGL_FREED;
- list_for_each_entry(sglq_entry,
- &phba->sli4_hba.lpfc_abts_nvmet_sgl_list, list)
- sglq_entry->state = SGL_FREED;
list_splice_init(&phba->sli4_hba.lpfc_abts_els_sgl_list,
&phba->sli4_hba.lpfc_els_sgl_list);
- if (phba->sli4_hba.nvme_wq)
- list_splice_init(&phba->sli4_hba.lpfc_abts_nvmet_sgl_list,
- &phba->sli4_hba.lpfc_nvmet_sgl_list);
spin_unlock(&phba->sli4_hba.sgl_list_lock);
/* abts_scsi_buf_list_lock required because worker thread uses this
@@ -1072,6 +1073,8 @@ lpfc_hba_down_post_s4(struct lpfc_hba *phba)
spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
list_splice_init(&phba->sli4_hba.lpfc_abts_nvme_buf_list,
&nvme_aborts);
+ list_splice_init(&phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
+ &nvmet_aborts);
spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
}
@@ -1085,13 +1088,20 @@ lpfc_hba_down_post_s4(struct lpfc_hba *phba)
list_splice(&aborts, &phba->lpfc_scsi_buf_list_put);
spin_unlock_irqrestore(&phba->scsi_buf_list_put_lock, iflag);
- list_for_each_entry_safe(psb, psb_next, &nvme_aborts, list) {
- psb->pCmd = NULL;
- psb->status = IOSTAT_SUCCESS;
+ if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME) {
+ list_for_each_entry_safe(psb, psb_next, &nvme_aborts, list) {
+ psb->pCmd = NULL;
+ psb->status = IOSTAT_SUCCESS;
+ }
+ spin_lock_irqsave(&phba->nvme_buf_list_put_lock, iflag);
+ list_splice(&nvme_aborts, &phba->lpfc_nvme_buf_list_put);
+ spin_unlock_irqrestore(&phba->nvme_buf_list_put_lock, iflag);
+
+ list_for_each_entry_safe(ctxp, ctxp_next, &nvmet_aborts, list) {
+ ctxp->flag &= ~(LPFC_NVMET_XBUSY | LPFC_NVMET_ABORT_OP);
+ lpfc_nvmet_rq_post(phba, ctxp, &ctxp->rqb_buffer->hbuf);
+ }
}
- spin_lock_irqsave(&phba->nvme_buf_list_put_lock, iflag);
- list_splice(&nvme_aborts, &phba->lpfc_nvme_buf_list_put);
- spin_unlock_irqrestore(&phba->nvme_buf_list_put_lock, iflag);
lpfc_sli4_free_sp_events(phba);
return 0;
@@ -5812,6 +5822,7 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
/* Initialize the Abort nvme buffer list used by driver */
spin_lock_init(&phba->sli4_hba.abts_nvme_buf_list_lock);
INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_nvme_buf_list);
+ INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_nvmet_ctx_list);
/* Fast-path XRI aborted CQ Event work queue list */
INIT_LIST_HEAD(&phba->sli4_hba.sp_nvme_xri_aborted_work_queue);
}
@@ -6439,7 +6450,7 @@ lpfc_init_sgl_list(struct lpfc_hba *phba)
INIT_LIST_HEAD(&phba->sli4_hba.lpfc_els_sgl_list);
INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_els_sgl_list);
INIT_LIST_HEAD(&phba->sli4_hba.lpfc_nvmet_sgl_list);
- INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_nvmet_sgl_list);
+ INIT_LIST_HEAD(&phba->sli4_hba.lpfc_abts_nvmet_ctx_list);
/* els xri-sgl book keeping */
phba->sli4_hba.els_xri_cnt = 0;
@@ -9966,17 +9977,19 @@ lpfc_sli4_xri_exchange_busy_wait(struct lpfc_hba *phba)
{
int wait_time = 0;
int nvme_xri_cmpl = 1;
+ int nvmet_xri_cmpl = 1;
int fcp_xri_cmpl = 1;
int els_xri_cmpl = list_empty(&phba->sli4_hba.lpfc_abts_els_sgl_list);
- int nvmet_xri_cmpl =
- list_empty(&phba->sli4_hba.lpfc_abts_nvmet_sgl_list);
if (phba->cfg_enable_fc4_type & LPFC_ENABLE_FCP)
fcp_xri_cmpl =
list_empty(&phba->sli4_hba.lpfc_abts_scsi_buf_list);
- if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME)
+ if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME) {
nvme_xri_cmpl =
list_empty(&phba->sli4_hba.lpfc_abts_nvme_buf_list);
+ nvmet_xri_cmpl =
+ list_empty(&phba->sli4_hba.lpfc_abts_nvmet_ctx_list);
+ }
while (!fcp_xri_cmpl || !els_xri_cmpl || !nvme_xri_cmpl ||
!nvmet_xri_cmpl) {
@@ -10002,9 +10015,12 @@ lpfc_sli4_xri_exchange_busy_wait(struct lpfc_hba *phba)
msleep(LPFC_XRI_EXCH_BUSY_WAIT_T1);
wait_time += LPFC_XRI_EXCH_BUSY_WAIT_T1;
}
- if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME)
+ if (phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME) {
nvme_xri_cmpl = list_empty(
&phba->sli4_hba.lpfc_abts_nvme_buf_list);
+ nvmet_xri_cmpl = list_empty(
+ &phba->sli4_hba.lpfc_abts_nvmet_ctx_list);
+ }
if (phba->cfg_enable_fc4_type & LPFC_ENABLE_FCP)
fcp_xri_cmpl = list_empty(
@@ -10013,8 +10029,6 @@ lpfc_sli4_xri_exchange_busy_wait(struct lpfc_hba *phba)
els_xri_cmpl =
list_empty(&phba->sli4_hba.lpfc_abts_els_sgl_list);
- nvmet_xri_cmpl =
- list_empty(&phba->sli4_hba.lpfc_abts_nvmet_sgl_list);
}
}
diff --git a/drivers/scsi/lpfc/lpfc_mbox.c b/drivers/scsi/lpfc/lpfc_mbox.c
index a928f51..ce25a18 100644
--- a/drivers/scsi/lpfc/lpfc_mbox.c
+++ b/drivers/scsi/lpfc/lpfc_mbox.c
@@ -2083,9 +2083,12 @@ lpfc_request_features(struct lpfc_hba *phba, struct lpfcMboxq *mboxq)
if (phba->max_vpi && phba->cfg_enable_npiv)
bf_set(lpfc_mbx_rq_ftr_rq_npiv, &mboxq->u.mqe.un.req_ftrs, 1);
- if (phba->nvmet_support)
+ if (phba->nvmet_support) {
bf_set(lpfc_mbx_rq_ftr_rq_mrqp, &mboxq->u.mqe.un.req_ftrs, 1);
-
+ /* iaab/iaar NOT set for now */
+ bf_set(lpfc_mbx_rq_ftr_rq_iaab, &mboxq->u.mqe.un.req_ftrs, 0);
+ bf_set(lpfc_mbx_rq_ftr_rq_iaar, &mboxq->u.mqe.un.req_ftrs, 0);
+ }
return;
}
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index 1a83ade..a439a50 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -867,15 +867,18 @@ lpfc_nvme_io_cmd_wqe_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pwqeIn,
break;
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_IOERR,
"6081 NVME Completion Protocol Error: "
- "status x%x result x%x placed x%x\n",
+ "xri %x status x%x result x%x "
+ "placed x%x\n",
+ lpfc_ncmd->cur_iocbq.sli4_xritag,
lpfc_ncmd->status, lpfc_ncmd->result,
wcqe->total_data_placed);
break;
default:
out_err:
lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_IOERR,
- "6072 NVME Completion Error: "
+ "6072 NVME Completion Error: xri %x "
"status x%x result x%x placed x%x\n",
+ lpfc_ncmd->cur_iocbq.sli4_xritag,
lpfc_ncmd->status, lpfc_ncmd->result,
wcqe->total_data_placed);
nCmd->transferred_length = 0;
@@ -1428,7 +1431,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
phba = vport->phba;
/* Announce entry to new IO submit field. */
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
+ lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME_ABTS,
"6002 Abort Request to rport DID x%06x "
"for nvme_fc_req %p\n",
pnvme_rport->port_id,
@@ -1458,7 +1461,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
/* The remote node has to be ready to send an abort. */
if ((ndlp->nlp_state != NLP_STE_MAPPED_NODE) &&
!(ndlp->nlp_type & NLP_NVME_TARGET)) {
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE | LOG_NVME_ABTS,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6048 rport %p, DID x%06x not ready for "
"IO. State x%x, Type x%x\n",
rport, pnvme_rport->port_id,
@@ -1473,7 +1476,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
/* driver queued commands are in process of being flushed */
if (phba->hba_flag & HBA_NVME_IOQ_FLUSH) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6139 Driver in reset cleanup - flushing "
"NVME Req now. hba_flag x%x\n",
phba->hba_flag);
@@ -1483,13 +1486,13 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
lpfc_nbuf = (struct lpfc_nvme_buf *)pnvme_fcreq->private;
if (!lpfc_nbuf) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6140 NVME IO req has no matching lpfc nvme "
"io buffer. Skipping abort req.\n");
return;
} else if (!lpfc_nbuf->nvmeCmd) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6141 lpfc NVME IO req has no nvme_fcreq "
"io buffer. Skipping abort req.\n");
return;
@@ -1505,7 +1508,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
*/
if (lpfc_nbuf->nvmeCmd != pnvme_fcreq) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6143 NVME req mismatch: "
"lpfc_nbuf %p nvmeCmd %p, "
"pnvme_fcreq %p. Skipping Abort xri x%x\n",
@@ -1517,7 +1520,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
/* Don't abort IOs no longer on the pending queue. */
if (!(nvmereq_wqe->iocb_flag & LPFC_IO_ON_TXCMPLQ)) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6142 NVME IO req %p not queued - skipping "
"abort req xri x%x\n",
pnvme_fcreq, nvmereq_wqe->sli4_xritag);
@@ -1531,7 +1534,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
/* Outstanding abort is in progress */
if (nvmereq_wqe->iocb_flag & LPFC_DRIVER_ABORTED) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6144 Outstanding NVME I/O Abort Request "
"still pending on nvme_fcreq %p, "
"lpfc_ncmd %p xri x%x\n",
@@ -1543,7 +1546,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
abts_buf = __lpfc_sli_get_iocbq(phba);
if (!abts_buf) {
spin_unlock_irqrestore(&phba->hbalock, flags);
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6136 No available abort wqes. Skipping "
"Abts req for nvme_fcreq %p xri x%x\n",
pnvme_fcreq, nvmereq_wqe->sli4_xritag);
@@ -1595,7 +1598,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
ret_val = lpfc_sli4_issue_wqe(phba, LPFC_FCP_RING, abts_buf);
spin_unlock_irqrestore(&phba->hbalock, flags);
if (ret_val == IOCB_ERROR) {
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME_ABTS,
"6137 Failed abts issue_wqe with status x%x "
"for nvme_fcreq %p.\n",
ret_val, pnvme_fcreq);
@@ -1603,7 +1606,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
return;
}
- lpfc_printf_vlog(vport, KERN_ERR, LOG_NVME,
+ lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME_ABTS,
"6138 Transport Abort NVME Request Issued for "
"ox_id x%x on reqtag x%x\n",
nvmereq_wqe->sli4_xritag,
@@ -2107,6 +2110,12 @@ lpfc_release_nvme_buf(struct lpfc_hba *phba, struct lpfc_nvme_buf *lpfc_ncmd)
lpfc_ncmd->nonsg_phys = 0;
if (lpfc_ncmd->flags & LPFC_SBUF_XBUSY) {
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6310 XB release deferred for "
+ "ox_id x%x on reqtag x%x\n",
+ lpfc_ncmd->cur_iocbq.sli4_xritag,
+ lpfc_ncmd->cur_iocbq.iotag);
+
spin_lock_irqsave(&phba->sli4_hba.abts_nvme_buf_list_lock,
iflag);
lpfc_ncmd->nvmeCmd = NULL;
@@ -2548,6 +2557,12 @@ lpfc_sli4_nvme_xri_aborted(struct lpfc_hba *phba,
rxid, 1);
lpfc_sli4_abts_err_handler(phba, ndlp, axri);
}
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6311 XRI Aborted xri x%x tag x%x "
+ "released\n",
+ xri, lpfc_ncmd->cur_iocbq.iotag);
+
lpfc_release_nvme_buf(phba, lpfc_ncmd);
if (rrq_empty)
lpfc_worker_wake_up(phba);
@@ -2556,4 +2571,8 @@ lpfc_sli4_nvme_xri_aborted(struct lpfc_hba *phba,
}
spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
spin_unlock_irqrestore(&phba->hbalock, iflag);
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6312 XRI Aborted xri x%x not found\n", xri);
+
}
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index d0dda42..fe939e8 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -71,6 +71,26 @@ static int lpfc_nvmet_unsol_ls_issue_abort(struct lpfc_hba *,
struct lpfc_nvmet_rcv_ctx *,
uint32_t, uint16_t);
+void
+lpfc_nvmet_defer_release(struct lpfc_hba *phba, struct lpfc_nvmet_rcv_ctx *ctxp)
+{
+ unsigned long iflag;
+
+ lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
+ "6313 NVMET Defer ctx release xri x%x flg x%x\n",
+ ctxp->oxid, ctxp->flag);
+
+ spin_lock_irqsave(&phba->sli4_hba.abts_nvme_buf_list_lock, iflag);
+ if (ctxp->flag & LPFC_NVMET_CTX_RLS) {
+ spin_unlock_irqrestore(&phba->sli4_hba.abts_nvme_buf_list_lock,
+ iflag);
+ return;
+ }
+ ctxp->flag |= LPFC_NVMET_CTX_RLS;
+ list_add_tail(&ctxp->list, &phba->sli4_hba.lpfc_abts_nvmet_ctx_list);
+ spin_unlock_irqrestore(&phba->sli4_hba.abts_nvme_buf_list_lock, iflag);
+}
+
/**
* lpfc_nvmet_xmt_ls_rsp_cmp - Completion handler for LS Response
* @phba: Pointer to HBA context object.
@@ -139,6 +159,11 @@ lpfc_nvmet_rq_post(struct lpfc_hba *phba, struct lpfc_nvmet_rcv_ctx *ctxp,
struct lpfc_dmabuf *mp)
{
if (ctxp) {
+ if (ctxp->flag)
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6314 rq_post ctx xri x%x flag x%x\n",
+ ctxp->oxid, ctxp->flag);
+
if (ctxp->txrdy) {
pci_pool_free(phba->txrdy_payload_pool, ctxp->txrdy,
ctxp->txrdy_phys);
@@ -337,39 +362,55 @@ lpfc_nvmet_xmt_fcp_op_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
#endif
ctxp = cmdwqe->context2;
+ ctxp->flag &= ~LPFC_NVMET_IO_INP;
+
rsp = &ctxp->ctx.fcp_req;
op = rsp->op;
- ctxp->flag &= ~LPFC_NVMET_IO_INP;
status = bf_get(lpfc_wcqe_c_status, wcqe);
result = wcqe->parameter;
- if (!phba->targetport)
- goto out;
+ if (phba->targetport)
+ tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
+ else
+ tgtp = NULL;
lpfc_nvmeio_data(phba, "NVMET FCP CMPL: xri x%x op x%x status x%x\n",
ctxp->oxid, op, status);
- tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
if (status) {
rsp->fcp_error = NVME_SC_DATA_XFER_ERROR;
rsp->transferred_length = 0;
- atomic_inc(&tgtp->xmt_fcp_rsp_error);
+ if (tgtp)
+ atomic_inc(&tgtp->xmt_fcp_rsp_error);
+
+ /* pick up SLI4 exhange busy condition */
+ if (bf_get(lpfc_wcqe_c_xb, wcqe)) {
+ ctxp->flag |= LPFC_NVMET_XBUSY;
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6315 IO Cmpl XBUSY: xri x%x: %x/%x\n",
+ ctxp->oxid, status, result);
+ } else {
+ ctxp->flag &= ~LPFC_NVMET_XBUSY;
+ }
+
} else {
rsp->fcp_error = NVME_SC_SUCCESS;
if (op == NVMET_FCOP_RSP)
rsp->transferred_length = rsp->rsplen;
else
rsp->transferred_length = rsp->transfer_length;
- atomic_inc(&tgtp->xmt_fcp_rsp_cmpl);
+ if (tgtp)
+ atomic_inc(&tgtp->xmt_fcp_rsp_cmpl);
}
-out:
if ((op == NVMET_FCOP_READDATA_RSP) ||
(op == NVMET_FCOP_RSP)) {
/* Sanity check */
ctxp->state = LPFC_NVMET_STE_DONE;
ctxp->entry_cnt++;
+
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
if (phba->ktime_on) {
if (rsp->op == NVMET_FCOP_READDATA_RSP) {
@@ -542,10 +583,11 @@ lpfc_nvmet_xmt_fcp_op(struct nvmet_fc_target_port *tgtport,
#endif
/* Sanity check */
- if (ctxp->state == LPFC_NVMET_STE_ABORT) {
+ if ((ctxp->flag & LPFC_NVMET_ABTS_RCV) ||
+ (ctxp->state == LPFC_NVMET_STE_ABORT)) {
atomic_inc(&lpfc_nvmep->xmt_fcp_drop);
lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
- "6102 Bad state IO x%x aborted\n",
+ "6102 IO xri x%x aborted\n",
ctxp->oxid);
rc = -ENXIO;
goto aerr;
@@ -615,16 +657,27 @@ lpfc_nvmet_xmt_fcp_abort(struct nvmet_fc_target_port *tgtport,
struct lpfc_nvmet_rcv_ctx *ctxp =
container_of(req, struct lpfc_nvmet_rcv_ctx, ctx.fcp_req);
struct lpfc_hba *phba = ctxp->phba;
+ unsigned long flags;
lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
- "6103 Abort op: oxri x%x %d cnt %d\n",
- ctxp->oxid, ctxp->state, ctxp->entry_cnt);
+ "6103 Abort op: oxri x%x flg x%x cnt %d\n",
+ ctxp->oxid, ctxp->flag, ctxp->entry_cnt);
- lpfc_nvmeio_data(phba, "NVMET FCP ABRT: xri x%x state x%x cnt x%x\n",
- ctxp->oxid, ctxp->state, ctxp->entry_cnt);
+ lpfc_nvmeio_data(phba, "NVMET FCP ABRT: "
+ "xri x%x flg x%x cnt x%x\n",
+ ctxp->oxid, ctxp->flag, ctxp->entry_cnt);
atomic_inc(&lpfc_nvmep->xmt_fcp_abort);
ctxp->entry_cnt++;
+ spin_lock_irqsave(&ctxp->ctxlock, flags);
+
+ /* Since iaab/iaar are NOT set, we need to check
+ * if the firmware is in process of aborting IO
+ */
+ if (ctxp->flag & LPFC_NVMET_XBUSY) {
+ spin_unlock_irqrestore(&ctxp->ctxlock, flags);
+ return;
+ }
ctxp->flag |= LPFC_NVMET_ABORT_OP;
if (ctxp->flag & LPFC_NVMET_IO_INP)
lpfc_nvmet_sol_fcp_issue_abort(phba, ctxp, ctxp->sid,
@@ -632,13 +685,13 @@ lpfc_nvmet_xmt_fcp_abort(struct nvmet_fc_target_port *tgtport,
else
lpfc_nvmet_unsol_fcp_issue_abort(phba, ctxp, ctxp->sid,
ctxp->oxid);
+ spin_unlock_irqrestore(&ctxp->ctxlock, flags);
}
static void
lpfc_nvmet_xmt_fcp_release(struct nvmet_fc_target_port *tgtport,
struct nvmefc_tgt_fcp_req *rsp)
{
- struct lpfc_nvmet_tgtport *lpfc_nvmep = tgtport->private;
struct lpfc_nvmet_rcv_ctx *ctxp =
container_of(rsp, struct lpfc_nvmet_rcv_ctx, ctx.fcp_req);
struct lpfc_hba *phba = ctxp->phba;
@@ -646,27 +699,20 @@ lpfc_nvmet_xmt_fcp_release(struct nvmet_fc_target_port *tgtport,
bool aborting = false;
spin_lock_irqsave(&ctxp->ctxlock, flags);
- if (ctxp->flag & LPFC_NVMET_ABORT_OP) {
+ if ((ctxp->flag & LPFC_NVMET_ABORT_OP) ||
+ (ctxp->flag & LPFC_NVMET_XBUSY)) {
aborting = true;
- ctxp->flag |= LPFC_NVMET_CTX_RLS;
- }
- spin_unlock_irqrestore(&ctxp->ctxlock, flags);
-
- if (aborting)
/* let the abort path do the real release */
- return;
-
- /* Sanity check */
- if (ctxp->state != LPFC_NVMET_STE_DONE) {
- atomic_inc(&lpfc_nvmep->xmt_fcp_drop);
- lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
- "6117 Bad state IO x%x aborted\n",
- ctxp->oxid);
+ lpfc_nvmet_defer_release(phba, ctxp);
}
+ spin_unlock_irqrestore(&ctxp->ctxlock, flags);
lpfc_nvmeio_data(phba, "NVMET FCP FREE: xri x%x ste %d\n", ctxp->oxid,
ctxp->state, 0);
+ if (aborting)
+ return;
+
lpfc_nvmet_rq_post(phba, ctxp, &ctxp->rqb_buffer->hbuf);
}
@@ -801,7 +847,122 @@ void
lpfc_sli4_nvmet_xri_aborted(struct lpfc_hba *phba,
struct sli4_wcqe_xri_aborted *axri)
{
- /* TODO: work in progress */
+ uint16_t xri = bf_get(lpfc_wcqe_xa_xri, axri);
+ uint16_t rxid = bf_get(lpfc_wcqe_xa_remote_xid, axri);
+ struct lpfc_nvmet_rcv_ctx *ctxp, *next_ctxp;
+ struct lpfc_nodelist *ndlp;
+ unsigned long iflag = 0;
+ int rrq_empty = 0;
+ bool released = false;
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6317 XB aborted xri x%x rxid x%x\n", xri, rxid);
+
+ if (!(phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME))
+ return;
+ spin_lock_irqsave(&phba->hbalock, iflag);
+ spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ list_for_each_entry_safe(ctxp, next_ctxp,
+ &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
+ list) {
+ if (ctxp->rqb_buffer->sglq->sli4_xritag == xri) {
+ /* Check if we already received a free context call
+ * and we have completed processing an abort situation.
+ */
+ if ((ctxp->flag & LPFC_NVMET_CTX_RLS) &&
+ !(ctxp->flag & LPFC_NVMET_ABORT_OP)) {
+ list_del(&ctxp->list);
+ released = true;
+ }
+ ctxp->flag &= ~LPFC_NVMET_XBUSY;
+ spin_unlock(
+ &phba->sli4_hba.abts_nvme_buf_list_lock);
+
+ rrq_empty = list_empty(&phba->active_rrq_list);
+ spin_unlock_irqrestore(&phba->hbalock, iflag);
+ ndlp = lpfc_findnode_did(phba->pport, ctxp->sid);
+ if (ndlp && NLP_CHK_NODE_ACT(ndlp) &&
+ ((ndlp->nlp_state == NLP_STE_UNMAPPED_NODE) ||
+ (ndlp->nlp_state == NLP_STE_MAPPED_NODE))) {
+ lpfc_set_rrq_active(
+ phba, ndlp,
+ ctxp->rqb_buffer->sglq->sli4_lxritag,
+ rxid, 1);
+ lpfc_sli4_abts_err_handler(phba, ndlp, axri);
+ }
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6318 XB aborted %x flg x%x (%x)\n",
+ ctxp->oxid, ctxp->flag, released);
+ if (released)
+ lpfc_nvmet_rq_post(phba, ctxp,
+ &ctxp->rqb_buffer->hbuf);
+ if (rrq_empty)
+ lpfc_worker_wake_up(phba);
+ return;
+ }
+ }
+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ spin_unlock_irqrestore(&phba->hbalock, iflag);
+}
+
+int
+lpfc_nvmet_rcv_unsol_abort(struct lpfc_vport *vport,
+ struct fc_frame_header *fc_hdr)
+
+{
+#if (IS_ENABLED(CONFIG_NVME_TARGET_FC))
+ struct lpfc_hba *phba = vport->phba;
+ struct lpfc_nvmet_rcv_ctx *ctxp, *next_ctxp;
+ struct nvmefc_tgt_fcp_req *rsp;
+ uint16_t xri;
+ unsigned long iflag = 0;
+
+ xri = be16_to_cpu(fc_hdr->fh_ox_id);
+
+ spin_lock_irqsave(&phba->hbalock, iflag);
+ spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ list_for_each_entry_safe(ctxp, next_ctxp,
+ &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
+ list) {
+ if (ctxp->rqb_buffer->sglq->sli4_xritag == xri) {
+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ spin_unlock_irqrestore(&phba->hbalock, iflag);
+
+ spin_lock_irqsave(&ctxp->ctxlock, iflag);
+ ctxp->flag |= LPFC_NVMET_ABTS_RCV;
+ spin_unlock_irqrestore(&ctxp->ctxlock, iflag);
+
+ lpfc_nvmeio_data(phba,
+ "NVMET ABTS RCV: "
+ "xri x%x CPU %02x rjt %d\n",
+ xri, smp_processor_id(), 0);
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6319 NVMET Rcv ABTS:acc xri x%x\n",
+ xri);
+
+ rsp = &ctxp->ctx.fcp_req;
+ nvmet_fc_rcv_fcp_abort(phba->targetport, rsp);
+
+ /* Respond with BA_ACC accordingly */
+ lpfc_sli4_seq_abort_rsp(vport, fc_hdr, 1);
+ return 0;
+ }
+ }
+ spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
+ spin_unlock_irqrestore(&phba->hbalock, iflag);
+
+ lpfc_nvmeio_data(phba, "NVMET ABTS RCV: xri x%x CPU %02x rjt %d\n",
+ xri, smp_processor_id(), 1);
+
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6320 NVMET Rcv ABTS:rjt xri x%x\n", xri);
+
+ /* Respond with BA_RJT accordingly */
+ lpfc_sli4_seq_abort_rsp(vport, fc_hdr, 0);
+ return 0;
+#endif
}
void
@@ -1655,18 +1816,27 @@ lpfc_nvmet_sol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
atomic_inc(&tgtp->xmt_abort_cmpl);
- lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
- "6165 Abort cmpl: xri x%x WCQE: %08x %08x %08x %08x\n",
- ctxp->oxid, wcqe->word0, wcqe->total_data_placed,
- result, wcqe->word3);
-
ctxp->state = LPFC_NVMET_STE_DONE;
+
+ /* Check if we already received a free context call
+ * and we have completed processing an abort situation.
+ */
spin_lock_irqsave(&ctxp->ctxlock, flags);
- if (ctxp->flag & LPFC_NVMET_CTX_RLS)
+ if ((ctxp->flag & LPFC_NVMET_CTX_RLS) &&
+ !(ctxp->flag & LPFC_NVMET_XBUSY)) {
+ list_del(&ctxp->list);
released = true;
+ }
ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
spin_unlock_irqrestore(&ctxp->ctxlock, flags);
+ lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
+ "6165 ABORT cmpl: xri x%x flg x%x (%d) "
+ "WCQE: %08x %08x %08x %08x\n",
+ ctxp->oxid, ctxp->flag, released,
+ wcqe->word0, wcqe->total_data_placed,
+ result, wcqe->word3);
+
/*
* if transport has released ctx, then can reuse it. Otherwise,
* will be recycled by transport release call.
@@ -1677,10 +1847,15 @@ lpfc_nvmet_sol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
cmdwqe->context2 = NULL;
cmdwqe->context3 = NULL;
lpfc_sli_release_iocbq(phba, cmdwqe);
+
+ /* Since iaab/iaar are NOT set, there is no work left.
+ * For LPFC_NVMET_XBUSY, lpfc_sli4_nvmet_xri_aborted
+ * should have been called already.
+ */
}
/**
- * lpfc_nvmet_xmt_fcp_abort_cmp - Completion handler for ABTS
+ * lpfc_nvmet_unsol_fcp_abort_cmp - Completion handler for ABTS
* @phba: Pointer to HBA context object.
* @cmdwqe: Pointer to driver command WQE object.
* @wcqe: Pointer to driver response CQE object.
@@ -1690,8 +1865,8 @@ lpfc_nvmet_sol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
* The function frees memory resources used for the NVME commands.
**/
static void
-lpfc_nvmet_xmt_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
- struct lpfc_wcqe_complete *wcqe)
+lpfc_nvmet_unsol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
+ struct lpfc_wcqe_complete *wcqe)
{
struct lpfc_nvmet_rcv_ctx *ctxp;
struct lpfc_nvmet_tgtport *tgtp;
@@ -1706,35 +1881,54 @@ lpfc_nvmet_xmt_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
atomic_inc(&tgtp->xmt_abort_cmpl);
+ if (!ctxp) {
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6070 ABTS cmpl: WCQE: %08x %08x %08x %08x\n",
+ wcqe->word0, wcqe->total_data_placed,
+ result, wcqe->word3);
+ return;
+ }
+
+ /* Sanity check */
+ if (ctxp->state != LPFC_NVMET_STE_ABORT) {
+ lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
+ "6112 ABTS Wrong state:%d oxid x%x\n",
+ ctxp->state, ctxp->oxid);
+ }
+
+ /* Check if we already received a free context call
+ * and we have completed processing an abort situation.
+ */
+ ctxp->state = LPFC_NVMET_STE_DONE;
+ spin_lock_irqsave(&ctxp->ctxlock, flags);
+ if ((ctxp->flag & LPFC_NVMET_CTX_RLS) &&
+ !(ctxp->flag & LPFC_NVMET_XBUSY)) {
+ list_del(&ctxp->list);
+ released = true;
+ }
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
+ spin_unlock_irqrestore(&ctxp->ctxlock, flags);
+
lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
- "6070 Abort cmpl: ctx %p WCQE: %08x %08x %08x %08x\n",
- ctxp, wcqe->word0, wcqe->total_data_placed,
+ "6316 ABTS cmpl xri x%x flg x%x (%x) "
+ "WCQE: %08x %08x %08x %08x\n",
+ ctxp->oxid, ctxp->flag, released,
+ wcqe->word0, wcqe->total_data_placed,
result, wcqe->word3);
+ /*
+ * if transport has released ctx, then can reuse it. Otherwise,
+ * will be recycled by transport release call.
+ */
+ if (released)
+ lpfc_nvmet_rq_post(phba, ctxp, &ctxp->rqb_buffer->hbuf);
- if (ctxp) {
- /* Sanity check */
- if (ctxp->state != LPFC_NVMET_STE_ABORT) {
- lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
- "6112 ABORT Wrong state:%d oxid x%x\n",
- ctxp->state, ctxp->oxid);
- }
- ctxp->state = LPFC_NVMET_STE_DONE;
- spin_lock_irqsave(&ctxp->ctxlock, flags);
- if (ctxp->flag & LPFC_NVMET_CTX_RLS)
- released = true;
- ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
- spin_unlock_irqrestore(&ctxp->ctxlock, flags);
-
- /*
- * if transport has released ctx, then can reuse it. Otherwise,
- * will be recycled by transport release call.
- */
- if (released)
- lpfc_nvmet_rq_post(phba, ctxp, &ctxp->rqb_buffer->hbuf);
+ cmdwqe->context2 = NULL;
+ cmdwqe->context3 = NULL;
- cmdwqe->context2 = NULL;
- cmdwqe->context3 = NULL;
- }
+ /* Since iaab/iaar are NOT set, there is no work left.
+ * For LPFC_NVMET_XBUSY, lpfc_sli4_nvmet_xri_aborted
+ * should have been called already.
+ */
}
/**
@@ -1787,10 +1981,14 @@ lpfc_nvmet_unsol_issue_abort(struct lpfc_hba *phba,
struct lpfc_nodelist *ndlp;
lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
- "6067 Abort: sid %x xri x%x/x%x\n",
+ "6067 ABTS: sid %x xri x%x/x%x\n",
sid, xri, ctxp->wqeq->sli4_xritag);
tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
+ if (!ctxp->wqeq) {
+ ctxp->wqeq = ctxp->rqb_buffer->iocbq;
+ ctxp->wqeq->hba_wqidx = 0;
+ }
ndlp = lpfc_findnode_did(phba->pport, sid);
if (!ndlp || !NLP_CHK_NODE_ACT(ndlp) ||
@@ -1896,10 +2094,11 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
(ndlp->nlp_state != NLP_STE_MAPPED_NODE))) {
atomic_inc(&tgtp->xmt_abort_rsp_error);
lpfc_printf_log(phba, KERN_WARNING, LOG_NVME_ABTS,
- "6160 Drop ABTS - wrong NDLP state x%x.\n",
+ "6160 Drop ABORT - wrong NDLP state x%x.\n",
(ndlp) ? ndlp->nlp_state : NLP_STE_MAX_STATE);
/* No failure to an ABTS request. */
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
return 0;
}
@@ -1907,9 +2106,10 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
ctxp->abort_wqeq = lpfc_sli_get_iocbq(phba);
if (!ctxp->abort_wqeq) {
lpfc_printf_log(phba, KERN_WARNING, LOG_NVME_ABTS,
- "6161 Abort failed: No wqeqs: "
+ "6161 ABORT failed: No wqeqs: "
"xri: x%x\n", ctxp->oxid);
/* No failure to an ABTS request. */
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
return 0;
}
abts_wqeq = ctxp->abort_wqeq;
@@ -1917,8 +2117,8 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
ctxp->state = LPFC_NVMET_STE_ABORT;
/* Announce entry to new IO submit field. */
- lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
- "6162 Abort Request to rport DID x%06x "
+ lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
+ "6162 ABORT Request to rport DID x%06x "
"for xri x%x x%x\n",
ctxp->sid, ctxp->oxid, ctxp->wqeq->sli4_xritag);
@@ -1934,6 +2134,7 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
"NVME Req now. hba_flag x%x oxid x%x\n",
phba->hba_flag, ctxp->oxid);
lpfc_sli_release_iocbq(phba, abts_wqeq);
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
return 0;
}
@@ -1945,6 +2146,7 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
"still pending on oxid x%x\n",
ctxp->oxid);
lpfc_sli_release_iocbq(phba, abts_wqeq);
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
return 0;
}
@@ -1992,9 +2194,10 @@ lpfc_nvmet_sol_fcp_issue_abort(struct lpfc_hba *phba,
if (rc == WQE_SUCCESS)
return 0;
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
lpfc_sli_release_iocbq(phba, abts_wqeq);
- lpfc_printf_log(phba, KERN_ERR, LOG_NVME,
- "6166 Failed abts issue_wqe with status x%x "
+ lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
+ "6166 Failed ABORT issue_wqe with status x%x "
"for oxid x%x.\n",
rc, ctxp->oxid);
return 1;
@@ -2023,8 +2226,8 @@ lpfc_nvmet_unsol_fcp_issue_abort(struct lpfc_hba *phba,
spin_lock_irqsave(&phba->hbalock, flags);
abts_wqeq = ctxp->wqeq;
- abts_wqeq->wqe_cmpl = lpfc_nvmet_xmt_fcp_abort_cmp;
- abts_wqeq->iocb_cmpl = 0;
+ abts_wqeq->wqe_cmpl = lpfc_nvmet_unsol_fcp_abort_cmp;
+ abts_wqeq->iocb_cmpl = NULL;
abts_wqeq->iocb_flag |= LPFC_IO_NVMET;
rc = lpfc_sli4_issue_wqe(phba, LPFC_FCP_RING, abts_wqeq);
spin_unlock_irqrestore(&phba->hbalock, flags);
@@ -2034,7 +2237,7 @@ lpfc_nvmet_unsol_fcp_issue_abort(struct lpfc_hba *phba,
}
aerr:
- lpfc_nvmet_rq_post(phba, ctxp, &ctxp->rqb_buffer->hbuf);
+ ctxp->flag &= ~LPFC_NVMET_ABORT_OP;
atomic_inc(&tgtp->xmt_abort_rsp_error);
lpfc_printf_log(phba, KERN_WARNING, LOG_NVME_ABTS,
"6135 Failed to Issue ABTS for oxid x%x. Status x%x\n",
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.h b/drivers/scsi/lpfc/lpfc_nvmet.h
index d8bac4c..128759f 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.h
+++ b/drivers/scsi/lpfc/lpfc_nvmet.h
@@ -75,6 +75,7 @@ struct lpfc_nvmet_rcv_ctx {
struct nvmefc_tgt_ls_req ls_req;
struct nvmefc_tgt_fcp_req fcp_req;
} ctx;
+ struct list_head list;
struct lpfc_hba *phba;
struct lpfc_iocbq *wqeq;
struct lpfc_iocbq *abort_wqeq;
@@ -96,10 +97,11 @@ struct lpfc_nvmet_rcv_ctx {
#define LPFC_NVMET_STE_RSP 4
#define LPFC_NVMET_STE_DONE 5
uint16_t flag;
-#define LPFC_NVMET_IO_INP 0x1
-#define LPFC_NVMET_ABORT_OP 0x2
-#define LPFC_NVMET_CTX_RLS 0x4
-
+#define LPFC_NVMET_IO_INP 0x1 /* IO is in progress on exchange */
+#define LPFC_NVMET_ABORT_OP 0x2 /* Abort WQE issued on exchange */
+#define LPFC_NVMET_XBUSY 0x4 /* XB bit set on IO cmpl */
+#define LPFC_NVMET_CTX_RLS 0x8 /* ctx free requested */
+#define LPFC_NVMET_ABTS_RCV 0x10 /* ABTS received on exchange */
struct rqb_dmabuf *rqb_buffer;
#ifdef CONFIG_SCSI_LPFC_DEBUG_FS
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index a6adc2b..cf19f49 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -16521,7 +16521,7 @@ lpfc_sli4_xri_inrange(struct lpfc_hba *phba,
* This function sends a basic response to a previous unsol sequence abort
* event after aborting the sequence handling.
**/
-static void
+void
lpfc_sli4_seq_abort_rsp(struct lpfc_vport *vport,
struct fc_frame_header *fc_hdr, bool aborted)
{
@@ -16697,6 +16697,11 @@ lpfc_sli4_handle_unsol_abort(struct lpfc_vport *vport,
}
lpfc_in_buf_free(phba, &dmabuf->dbuf);
+ if (phba->nvmet_support) {
+ lpfc_nvmet_rcv_unsol_abort(vport, &fc_hdr);
+ return;
+ }
+
/* Respond with BA_ACC or BA_RJT accordingly */
lpfc_sli4_seq_abort_rsp(vport, &fc_hdr, aborted);
}
diff --git a/drivers/scsi/lpfc/lpfc_sli4.h b/drivers/scsi/lpfc/lpfc_sli4.h
index 710458c..da46471 100644
--- a/drivers/scsi/lpfc/lpfc_sli4.h
+++ b/drivers/scsi/lpfc/lpfc_sli4.h
@@ -620,7 +620,7 @@ struct lpfc_sli4_hba {
struct list_head lpfc_els_sgl_list;
struct list_head lpfc_abts_els_sgl_list;
struct list_head lpfc_nvmet_sgl_list;
- struct list_head lpfc_abts_nvmet_sgl_list;
+ struct list_head lpfc_abts_nvmet_ctx_list;
struct list_head lpfc_abts_scsi_buf_list;
struct list_head lpfc_abts_nvme_buf_list;
struct lpfc_sglq **lpfc_sglq_active_list;
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 20/22] Update ABORT processing for NVMET.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 8:10 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 8:10 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:39PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> The driver with nvme had this routine stubbed.
>
> Right now XRI_ABORTED_CQE is not handled and the FC NVMET
> Transport has a new API for the driver.
>
> Missing code path, new NVME abort API
> Update ABORT processing for NVMET
>
> There are 3 new FC NVMET Transport API/ template routines for NVMET:
>
> lpfc_nvmet_xmt_fcp_release
> This NVMET template callback routine called to release context
> associated with an IO This routine is ALWAYS called last, even
> if the IO was aborted or completed in error.
>
> lpfc_nvmet_xmt_fcp_abort
> This NVMET template callback routine called to abort an exchange that
> has an IO in progress
>
> nvmet_fc_rcv_fcp_req
> When the lpfc driver receives an ABTS, this NVME FC transport layer
> callback routine is called. For this case there are 2 paths thru the
> driver: the driver either has an outstanding exchange / context for the
> XRI to be aborted or not. If not, a BA_RJT is issued otherwise a BA_ACC
>
> NVMET Driver abort paths:
>
> There are 2 paths for aborting an IO. The first one is we receive an IO and
> decide not to process it because of lack of resources. An unsolicated ABTS
> is immediately sent back to the initiator as a response.
> lpfc_nvmet_unsol_fcp_buffer
> lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE)
>
> The second one is we sent the IO up to the NVMET transport layer to
> process, and for some reason the NVME Transport layer decided to abort the
> IO before it completes all its phases. For this case there are 2 paths
> thru the driver:
> the driver either has an outstanding TSEND/TRECEIVE/TRSP WQE or no
> outstanding WQEs are present for the exchange / context.
> lpfc_nvmet_xmt_fcp_abort
> if (LPFC_NVMET_IO_INP)
> lpfc_nvmet_sol_fcp_issue_abort (ABORT_WQE)
> lpfc_nvmet_sol_fcp_abort_cmp
> else
> lpfc_nvmet_unsol_fcp_issue_abort
> lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE)
> lpfc_nvmet_unsol_fcp_abort_cmp
>
> Context flags:
> LPFC_NVMET_IOP - his flag signifies an IO is in progress on the exchange.
> LPFC_NVMET_XBUSY - this flag indicates the IO completed but the firmware
> is still busy with the corresponding exchange. The exchange should not be
> reused until after a XRI_ABORTED_CQE is received for that exchange.
> LPFC_NVMET_ABORT_OP - this flag signifies an ABORT_WQE was issued on the
> exchange.
> LPFC_NVMET_CTX_RLS - this flag signifies a context free was requested,
> but we are deferring it due to an XBUSY or ABORT in progress.
>
> A ctxlock is added to the context structure that is used whenever these
> flags are set/read within the context of an IO.
> The LPFC_NVMET_CTX_RLS flag is only set in the defer_relase routine when
> the transport has resolved all IO associated with the buffer. The flag is
> cleared when the CTX is associated with a new IO.
>
> An exchange can has both an LPFC_NVMET_XBUSY and a LPFC_NVMET_ABORT_OP
> condition active simultaneously. Both conditions must complete before the
> exchange is freed.
> When the abort callback (lpfc_nvmet_xmt_fcp_abort) is envoked:
> If there is an outstanding IO, the driver will issue an ABORT_WQE. This
> should result in 3 completions for the exchange:
> 1) IO cmpl with XB bit set
> 2) Abort WQE cmpl
> 3) XRI_ABORTED_CQE cmpl
> For this scenerio, after completion #1, the NVMET Transport IO rsp
> callback is called. After completion #2, no action is taken with respect
> to the exchange / context. After completion #3, the exchange context is
> free for re-use on another IO.
>
> If there is no outstanding activity on the exchange, the driver will send a
> ABTS to the Initiator. Upon completion of this WQE, the exchange / context
> is freed for re-use on another IO.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
[...]
> @@ -139,6 +159,11 @@ lpfc_nvmet_rq_post(struct lpfc_hba *phba, struct lpfc_nvmet_rcv_ctx *ctxp,
> struct lpfc_dmabuf *mp)
> {
> if (ctxp) {
> + if (ctxp->flag)
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6314 rq_post ctx xri x%x flag x%x\n",
> + ctxp->oxid, ctxp->flag);
> +
Indendation looks a bit odd here. Not sure if GCC-6's -Wmisleading-indent will
flag this.
> @@ -801,7 +847,122 @@ void
> lpfc_sli4_nvmet_xri_aborted(struct lpfc_hba *phba,
> struct sli4_wcqe_xri_aborted *axri)
> {
> - /* TODO: work in progress */
> + uint16_t xri = bf_get(lpfc_wcqe_xa_xri, axri);
> + uint16_t rxid = bf_get(lpfc_wcqe_xa_remote_xid, axri);
> + struct lpfc_nvmet_rcv_ctx *ctxp, *next_ctxp;
> + struct lpfc_nodelist *ndlp;
> + unsigned long iflag = 0;
> + int rrq_empty = 0;
> + bool released = false;
> +
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6317 XB aborted xri x%x rxid x%x\n", xri, rxid);
> +
> + if (!(phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME))
> + return;
> + spin_lock_irqsave(&phba->hbalock, iflag);
> + spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
> + list_for_each_entry_safe(ctxp, next_ctxp,
> + &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
> + list) {
> + if (ctxp->rqb_buffer->sglq->sli4_xritag == xri) {
if (ctxp->rqb_buffer->sglq->sli4_xritag != xri)
continue;
????
This saves the indent so the block below doesn't look so squezed and you don't
have to have a linebreak in spin_unlock() for instance.
> + /* Check if we already received a free context call
> + * and we have completed processing an abort situation.
> + */
> + if ((ctxp->flag & LPFC_NVMET_CTX_RLS) &&
> + !(ctxp->flag & LPFC_NVMET_ABORT_OP)) {
> + list_del(&ctxp->list);
> + released = true;
> + }
> + ctxp->flag &= ~LPFC_NVMET_XBUSY;
> + spin_unlock(
> + &phba->sli4_hba.abts_nvme_buf_list_lock);
> +
> + rrq_empty = list_empty(&phba->active_rrq_list);
> + spin_unlock_irqrestore(&phba->hbalock, iflag);
> + ndlp = lpfc_findnode_did(phba->pport, ctxp->sid);
> + if (ndlp && NLP_CHK_NODE_ACT(ndlp) &&
> + ((ndlp->nlp_state == NLP_STE_UNMAPPED_NODE) ||
> + (ndlp->nlp_state == NLP_STE_MAPPED_NODE))) {
> + lpfc_set_rrq_active(
> + phba, ndlp,
> + ctxp->rqb_buffer->sglq->sli4_lxritag,
> + rxid, 1);
> + lpfc_sli4_abts_err_handler(phba, ndlp, axri);
> + }
> +
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6318 XB aborted %x flg x%x (%x)\n",
> + ctxp->oxid, ctxp->flag, released);
> + if (released)
> + lpfc_nvmet_rq_post(phba, ctxp,
> + &ctxp->rqb_buffer->hbuf);
> + if (rrq_empty)
> + lpfc_worker_wake_up(phba);
> + return;
> + }
> + }
> + spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
> + spin_unlock_irqrestore(&phba->hbalock, iflag);
> +}
> +
> +int
> +lpfc_nvmet_rcv_unsol_abort(struct lpfc_vport *vport,
> + struct fc_frame_header *fc_hdr)
> +
> +{
> +#if (IS_ENABLED(CONFIG_NVME_TARGET_FC))
> + struct lpfc_hba *phba = vport->phba;
> + struct lpfc_nvmet_rcv_ctx *ctxp, *next_ctxp;
> + struct nvmefc_tgt_fcp_req *rsp;
> + uint16_t xri;
> + unsigned long iflag = 0;
> +
> + xri = be16_to_cpu(fc_hdr->fh_ox_id);
> +
> + spin_lock_irqsave(&phba->hbalock, iflag);
> + spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
> + list_for_each_entry_safe(ctxp, next_ctxp,
> + &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
> + list) {
Same here.
> + if (ctxp->rqb_buffer->sglq->sli4_xritag == xri) {
> + spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
> + spin_unlock_irqrestore(&phba->hbalock, iflag);
> +
> + spin_lock_irqsave(&ctxp->ctxlock, iflag);
> + ctxp->flag |= LPFC_NVMET_ABTS_RCV;
> + spin_unlock_irqrestore(&ctxp->ctxlock, iflag);
> +
> + lpfc_nvmeio_data(phba,
> + "NVMET ABTS RCV: "
> + "xri x%x CPU %02x rjt %d\n",
> + xri, smp_processor_id(), 0);
> +
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6319 NVMET Rcv ABTS:acc xri x%x\n",
> + xri);
> +
> + rsp = &ctxp->ctx.fcp_req;
> + nvmet_fc_rcv_fcp_abort(phba->targetport, rsp);
> +
> + /* Respond with BA_ACC accordingly */
> + lpfc_sli4_seq_abort_rsp(vport, fc_hdr, 1);
> + return 0;
> + }
> + }
> + spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
> + spin_unlock_irqrestore(&phba->hbalock, iflag);
> +
> + lpfc_nvmeio_data(phba, "NVMET ABTS RCV: xri x%x CPU %02x rjt %d\n",
> + xri, smp_processor_id(), 1);
> +
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6320 NVMET Rcv ABTS:rjt xri x%x\n", xri);
> +
> + /* Respond with BA_RJT accordingly */
> + lpfc_sli4_seq_abort_rsp(vport, fc_hdr, 0);
> + return 0;
> +#endif
> }
> @@ -1677,10 +1847,15 @@ lpfc_nvmet_sol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
> cmdwqe->context2 = NULL;
> cmdwqe->context3 = NULL;
> lpfc_sli_release_iocbq(phba, cmdwqe);
> +
> + /* Since iaab/iaar are NOT set, there is no work left.
> + * For LPFC_NVMET_XBUSY, lpfc_sli4_nvmet_xri_aborted
> + * should have been called already.
> + */
> }
>
> /**
> - * lpfc_nvmet_xmt_fcp_abort_cmp - Completion handler for ABTS
> + * lpfc_nvmet_unsol_fcp_abort_cmp - Completion handler for ABTS
> * @phba: Pointer to HBA context object.
> * @cmdwqe: Pointer to driver command WQE object.
> * @wcqe: Pointer to driver response CQE object.
> @@ -1690,8 +1865,8 @@ lpfc_nvmet_sol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
> * The function frees memory resources used for the NVME commands.
> **/
> static void
> -lpfc_nvmet_xmt_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
> - struct lpfc_wcqe_complete *wcqe)
> +lpfc_nvmet_unsol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
> + struct lpfc_wcqe_complete *wcqe)
> {
> struct lpfc_nvmet_rcv_ctx *ctxp;
> struct lpfc_nvmet_tgtport *tgtp;
> @@ -1706,35 +1881,54 @@ lpfc_nvmet_xmt_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
> tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
> atomic_inc(&tgtp->xmt_abort_cmpl);
>
> + if (!ctxp) {
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6070 ABTS cmpl: WCQE: %08x %08x %08x %08x\n",
> + wcqe->word0, wcqe->total_data_placed,
> + result, wcqe->word3);
> + return;
> + }
> +
> + /* Sanity check */
> + if (ctxp->state != LPFC_NVMET_STE_ABORT) {
> + lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
> + "6112 ABTS Wrong state:%d oxid x%x\n",
> + ctxp->state, ctxp->oxid);
OK, this is kinda odd as well. We do a sanity check on the state but don't
take any measures other than writing a log message? Maybe that's OK, but it
surely looks a bit odd to me when reviewing. If it's OK please amend the
comment above and explain this is intented behaviour.
Thanks,
Johannes
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread* [PATCH-v1 20/22] Update ABORT processing for NVMET.
@ 2017-04-20 8:10 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 8:10 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:39PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> The driver with nvme had this routine stubbed.
>
> Right now XRI_ABORTED_CQE is not handled and the FC NVMET
> Transport has a new API for the driver.
>
> Missing code path, new NVME abort API
> Update ABORT processing for NVMET
>
> There are 3 new FC NVMET Transport API/ template routines for NVMET:
>
> lpfc_nvmet_xmt_fcp_release
> This NVMET template callback routine called to release context
> associated with an IO This routine is ALWAYS called last, even
> if the IO was aborted or completed in error.
>
> lpfc_nvmet_xmt_fcp_abort
> This NVMET template callback routine called to abort an exchange that
> has an IO in progress
>
> nvmet_fc_rcv_fcp_req
> When the lpfc driver receives an ABTS, this NVME FC transport layer
> callback routine is called. For this case there are 2 paths thru the
> driver: the driver either has an outstanding exchange / context for the
> XRI to be aborted or not. If not, a BA_RJT is issued otherwise a BA_ACC
>
> NVMET Driver abort paths:
>
> There are 2 paths for aborting an IO. The first one is we receive an IO and
> decide not to process it because of lack of resources. An unsolicated ABTS
> is immediately sent back to the initiator as a response.
> lpfc_nvmet_unsol_fcp_buffer
> lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE)
>
> The second one is we sent the IO up to the NVMET transport layer to
> process, and for some reason the NVME Transport layer decided to abort the
> IO before it completes all its phases. For this case there are 2 paths
> thru the driver:
> the driver either has an outstanding TSEND/TRECEIVE/TRSP WQE or no
> outstanding WQEs are present for the exchange / context.
> lpfc_nvmet_xmt_fcp_abort
> if (LPFC_NVMET_IO_INP)
> lpfc_nvmet_sol_fcp_issue_abort (ABORT_WQE)
> lpfc_nvmet_sol_fcp_abort_cmp
> else
> lpfc_nvmet_unsol_fcp_issue_abort
> lpfc_nvmet_unsol_issue_abort (XMIT_SEQUENCE_WQE)
> lpfc_nvmet_unsol_fcp_abort_cmp
>
> Context flags:
> LPFC_NVMET_IOP - his flag signifies an IO is in progress on the exchange.
> LPFC_NVMET_XBUSY - this flag indicates the IO completed but the firmware
> is still busy with the corresponding exchange. The exchange should not be
> reused until after a XRI_ABORTED_CQE is received for that exchange.
> LPFC_NVMET_ABORT_OP - this flag signifies an ABORT_WQE was issued on the
> exchange.
> LPFC_NVMET_CTX_RLS - this flag signifies a context free was requested,
> but we are deferring it due to an XBUSY or ABORT in progress.
>
> A ctxlock is added to the context structure that is used whenever these
> flags are set/read within the context of an IO.
> The LPFC_NVMET_CTX_RLS flag is only set in the defer_relase routine when
> the transport has resolved all IO associated with the buffer. The flag is
> cleared when the CTX is associated with a new IO.
>
> An exchange can has both an LPFC_NVMET_XBUSY and a LPFC_NVMET_ABORT_OP
> condition active simultaneously. Both conditions must complete before the
> exchange is freed.
> When the abort callback (lpfc_nvmet_xmt_fcp_abort) is envoked:
> If there is an outstanding IO, the driver will issue an ABORT_WQE. This
> should result in 3 completions for the exchange:
> 1) IO cmpl with XB bit set
> 2) Abort WQE cmpl
> 3) XRI_ABORTED_CQE cmpl
> For this scenerio, after completion #1, the NVMET Transport IO rsp
> callback is called. After completion #2, no action is taken with respect
> to the exchange / context. After completion #3, the exchange context is
> free for re-use on another IO.
>
> If there is no outstanding activity on the exchange, the driver will send a
> ABTS to the Initiator. Upon completion of this WQE, the exchange / context
> is freed for re-use on another IO.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
[...]
> @@ -139,6 +159,11 @@ lpfc_nvmet_rq_post(struct lpfc_hba *phba, struct lpfc_nvmet_rcv_ctx *ctxp,
> struct lpfc_dmabuf *mp)
> {
> if (ctxp) {
> + if (ctxp->flag)
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6314 rq_post ctx xri x%x flag x%x\n",
> + ctxp->oxid, ctxp->flag);
> +
Indendation looks a bit odd here. Not sure if GCC-6's -Wmisleading-indent will
flag this.
> @@ -801,7 +847,122 @@ void
> lpfc_sli4_nvmet_xri_aborted(struct lpfc_hba *phba,
> struct sli4_wcqe_xri_aborted *axri)
> {
> - /* TODO: work in progress */
> + uint16_t xri = bf_get(lpfc_wcqe_xa_xri, axri);
> + uint16_t rxid = bf_get(lpfc_wcqe_xa_remote_xid, axri);
> + struct lpfc_nvmet_rcv_ctx *ctxp, *next_ctxp;
> + struct lpfc_nodelist *ndlp;
> + unsigned long iflag = 0;
> + int rrq_empty = 0;
> + bool released = false;
> +
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6317 XB aborted xri x%x rxid x%x\n", xri, rxid);
> +
> + if (!(phba->cfg_enable_fc4_type & LPFC_ENABLE_NVME))
> + return;
> + spin_lock_irqsave(&phba->hbalock, iflag);
> + spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
> + list_for_each_entry_safe(ctxp, next_ctxp,
> + &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
> + list) {
> + if (ctxp->rqb_buffer->sglq->sli4_xritag == xri) {
if (ctxp->rqb_buffer->sglq->sli4_xritag != xri)
continue;
????
This saves the indent so the block below doesn't look so squezed and you don't
have to have a linebreak in spin_unlock() for instance.
> + /* Check if we already received a free context call
> + * and we have completed processing an abort situation.
> + */
> + if ((ctxp->flag & LPFC_NVMET_CTX_RLS) &&
> + !(ctxp->flag & LPFC_NVMET_ABORT_OP)) {
> + list_del(&ctxp->list);
> + released = true;
> + }
> + ctxp->flag &= ~LPFC_NVMET_XBUSY;
> + spin_unlock(
> + &phba->sli4_hba.abts_nvme_buf_list_lock);
> +
> + rrq_empty = list_empty(&phba->active_rrq_list);
> + spin_unlock_irqrestore(&phba->hbalock, iflag);
> + ndlp = lpfc_findnode_did(phba->pport, ctxp->sid);
> + if (ndlp && NLP_CHK_NODE_ACT(ndlp) &&
> + ((ndlp->nlp_state == NLP_STE_UNMAPPED_NODE) ||
> + (ndlp->nlp_state == NLP_STE_MAPPED_NODE))) {
> + lpfc_set_rrq_active(
> + phba, ndlp,
> + ctxp->rqb_buffer->sglq->sli4_lxritag,
> + rxid, 1);
> + lpfc_sli4_abts_err_handler(phba, ndlp, axri);
> + }
> +
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6318 XB aborted %x flg x%x (%x)\n",
> + ctxp->oxid, ctxp->flag, released);
> + if (released)
> + lpfc_nvmet_rq_post(phba, ctxp,
> + &ctxp->rqb_buffer->hbuf);
> + if (rrq_empty)
> + lpfc_worker_wake_up(phba);
> + return;
> + }
> + }
> + spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
> + spin_unlock_irqrestore(&phba->hbalock, iflag);
> +}
> +
> +int
> +lpfc_nvmet_rcv_unsol_abort(struct lpfc_vport *vport,
> + struct fc_frame_header *fc_hdr)
> +
> +{
> +#if (IS_ENABLED(CONFIG_NVME_TARGET_FC))
> + struct lpfc_hba *phba = vport->phba;
> + struct lpfc_nvmet_rcv_ctx *ctxp, *next_ctxp;
> + struct nvmefc_tgt_fcp_req *rsp;
> + uint16_t xri;
> + unsigned long iflag = 0;
> +
> + xri = be16_to_cpu(fc_hdr->fh_ox_id);
> +
> + spin_lock_irqsave(&phba->hbalock, iflag);
> + spin_lock(&phba->sli4_hba.abts_nvme_buf_list_lock);
> + list_for_each_entry_safe(ctxp, next_ctxp,
> + &phba->sli4_hba.lpfc_abts_nvmet_ctx_list,
> + list) {
Same here.
> + if (ctxp->rqb_buffer->sglq->sli4_xritag == xri) {
> + spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
> + spin_unlock_irqrestore(&phba->hbalock, iflag);
> +
> + spin_lock_irqsave(&ctxp->ctxlock, iflag);
> + ctxp->flag |= LPFC_NVMET_ABTS_RCV;
> + spin_unlock_irqrestore(&ctxp->ctxlock, iflag);
> +
> + lpfc_nvmeio_data(phba,
> + "NVMET ABTS RCV: "
> + "xri x%x CPU %02x rjt %d\n",
> + xri, smp_processor_id(), 0);
> +
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6319 NVMET Rcv ABTS:acc xri x%x\n",
> + xri);
> +
> + rsp = &ctxp->ctx.fcp_req;
> + nvmet_fc_rcv_fcp_abort(phba->targetport, rsp);
> +
> + /* Respond with BA_ACC accordingly */
> + lpfc_sli4_seq_abort_rsp(vport, fc_hdr, 1);
> + return 0;
> + }
> + }
> + spin_unlock(&phba->sli4_hba.abts_nvme_buf_list_lock);
> + spin_unlock_irqrestore(&phba->hbalock, iflag);
> +
> + lpfc_nvmeio_data(phba, "NVMET ABTS RCV: xri x%x CPU %02x rjt %d\n",
> + xri, smp_processor_id(), 1);
> +
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6320 NVMET Rcv ABTS:rjt xri x%x\n", xri);
> +
> + /* Respond with BA_RJT accordingly */
> + lpfc_sli4_seq_abort_rsp(vport, fc_hdr, 0);
> + return 0;
> +#endif
> }
> @@ -1677,10 +1847,15 @@ lpfc_nvmet_sol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
> cmdwqe->context2 = NULL;
> cmdwqe->context3 = NULL;
> lpfc_sli_release_iocbq(phba, cmdwqe);
> +
> + /* Since iaab/iaar are NOT set, there is no work left.
> + * For LPFC_NVMET_XBUSY, lpfc_sli4_nvmet_xri_aborted
> + * should have been called already.
> + */
> }
>
> /**
> - * lpfc_nvmet_xmt_fcp_abort_cmp - Completion handler for ABTS
> + * lpfc_nvmet_unsol_fcp_abort_cmp - Completion handler for ABTS
> * @phba: Pointer to HBA context object.
> * @cmdwqe: Pointer to driver command WQE object.
> * @wcqe: Pointer to driver response CQE object.
> @@ -1690,8 +1865,8 @@ lpfc_nvmet_sol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
> * The function frees memory resources used for the NVME commands.
> **/
> static void
> -lpfc_nvmet_xmt_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
> - struct lpfc_wcqe_complete *wcqe)
> +lpfc_nvmet_unsol_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
> + struct lpfc_wcqe_complete *wcqe)
> {
> struct lpfc_nvmet_rcv_ctx *ctxp;
> struct lpfc_nvmet_tgtport *tgtp;
> @@ -1706,35 +1881,54 @@ lpfc_nvmet_xmt_fcp_abort_cmp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdwqe,
> tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
> atomic_inc(&tgtp->xmt_abort_cmpl);
>
> + if (!ctxp) {
> + lpfc_printf_log(phba, KERN_INFO, LOG_NVME_ABTS,
> + "6070 ABTS cmpl: WCQE: %08x %08x %08x %08x\n",
> + wcqe->word0, wcqe->total_data_placed,
> + result, wcqe->word3);
> + return;
> + }
> +
> + /* Sanity check */
> + if (ctxp->state != LPFC_NVMET_STE_ABORT) {
> + lpfc_printf_log(phba, KERN_ERR, LOG_NVME_ABTS,
> + "6112 ABTS Wrong state:%d oxid x%x\n",
> + ctxp->state, ctxp->oxid);
OK, this is kinda odd as well. We do a sanity check on the state but don't
take any measures other than writing a log message? Maybe that's OK, but it
surely looks a bit odd to me when reviewing. If it's OK please amend the
comment above and explain this is intented behaviour.
Thanks,
Johannes
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 21/22] Fix Express lane queue creation.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
The older sli4 adapters only supported the 64 byte WQE entry size.
The new adapter (fw) support both 64 and 128 byte WQE entry sizies.
The Express lane WQ was not being created with the 128 byte WQE sizes
when it was supported.
Not having the right WQE size created for the express lane work queue
caused the the firmware to overwrite the lun indentifier in the FCP header.
This patch correctly creates the express lane work queue with the
supported size.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_init.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index e084a66..7d3b199 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -12040,6 +12040,7 @@ int
lpfc_fof_queue_create(struct lpfc_hba *phba)
{
struct lpfc_queue *qdesc;
+ uint32_t wqesize;
/* Create FOF EQ */
qdesc = lpfc_sli4_queue_alloc(phba, phba->sli4_hba.eq_esize,
@@ -12060,8 +12061,11 @@ lpfc_fof_queue_create(struct lpfc_hba *phba)
phba->sli4_hba.oas_cq = qdesc;
/* Create OAS WQ */
- qdesc = lpfc_sli4_queue_alloc(phba, phba->sli4_hba.wq_esize,
+ wqesize = (phba->fcp_embed_io) ?
+ LPFC_WQE128_SIZE : phba->sli4_hba.wq_esize;
+ qdesc = lpfc_sli4_queue_alloc(phba, wqesize,
phba->sli4_hba.wq_ecount);
+
if (!qdesc)
goto out_error;
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* [PATCH-v1 21/22] Fix Express lane queue creation.
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
The older sli4 adapters only supported the 64 byte WQE entry size.
The new adapter (fw) support both 64 and 128 byte WQE entry sizies.
The Express lane WQ was not being created with the 128 byte WQE sizes
when it was supported.
Not having the right WQE size created for the express lane work queue
caused the the firmware to overwrite the lun indentifier in the FCP header.
This patch correctly creates the express lane work queue with the
supported size.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_init.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index e084a66..7d3b199 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -12040,6 +12040,7 @@ int
lpfc_fof_queue_create(struct lpfc_hba *phba)
{
struct lpfc_queue *qdesc;
+ uint32_t wqesize;
/* Create FOF EQ */
qdesc = lpfc_sli4_queue_alloc(phba, phba->sli4_hba.eq_esize,
@@ -12060,8 +12061,11 @@ lpfc_fof_queue_create(struct lpfc_hba *phba)
phba->sli4_hba.oas_cq = qdesc;
/* Create OAS WQ */
- qdesc = lpfc_sli4_queue_alloc(phba, phba->sli4_hba.wq_esize,
+ wqesize = (phba->fcp_embed_io) ?
+ LPFC_WQE128_SIZE : phba->sli4_hba.wq_esize;
+ qdesc = lpfc_sli4_queue_alloc(phba, wqesize,
phba->sli4_hba.wq_ecount);
+
if (!qdesc)
goto out_error;
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread* Re: [PATCH-v1 21/22] Fix Express lane queue creation.
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 8:11 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 8:11 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:40PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> The older sli4 adapters only supported the 64 byte WQE entry size.
> The new adapter (fw) support both 64 and 128 byte WQE entry sizies.
> The Express lane WQ was not being created with the 128 byte WQE sizes
> when it was supported.
>
> Not having the right WQE size created for the express lane work queue
> caused the the firmware to overwrite the lun indentifier in the FCP header.
>
> This patch correctly creates the express lane work queue with the
> supported size.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 21/22] Fix Express lane queue creation.
@ 2017-04-20 8:11 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 8:11 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:40PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> The older sli4 adapters only supported the 64 byte WQE entry size.
> The new adapter (fw) support both 64 and 128 byte WQE entry sizies.
> The Express lane WQ was not being created with the 128 byte WQE sizes
> when it was supported.
>
> Not having the right WQE size created for the express lane work queue
> caused the the firmware to overwrite the lun indentifier in the FCP header.
>
> This patch correctly creates the express lane work queue with the
> supported size.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 22/22] lpfc revison 11.2.0.12
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 4:46 ` jsmart2021
-1 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
To: linux-scsi, linux-nvme, sagi, martin.petersen
Cc: James Smart, Dick Kennedy, James Smart
From: James Smart <jsmart2021@gmail.com>
Update lpfc version to reflect this set of changes.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_version.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/scsi/lpfc/lpfc_version.h b/drivers/scsi/lpfc/lpfc_version.h
index d4e95e2..1c26dc6 100644
--- a/drivers/scsi/lpfc/lpfc_version.h
+++ b/drivers/scsi/lpfc/lpfc_version.h
@@ -20,7 +20,7 @@
* included with this package. *
*******************************************************************/
-#define LPFC_DRIVER_VERSION "11.2.0.10"
+#define LPFC_DRIVER_VERSION "11.2.0.12"
#define LPFC_DRIVER_NAME "lpfc"
/* Used for SLI 2/3 */
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* [PATCH-v1 22/22] lpfc revison 11.2.0.12
@ 2017-04-20 4:46 ` jsmart2021
0 siblings, 0 replies; 100+ messages in thread
From: jsmart2021 @ 2017-04-20 4:46 UTC (permalink / raw)
From: James Smart <jsmart2021@gmail.com>
Update lpfc version to reflect this set of changes.
Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
Signed-off-by: James Smart <james.smart at broadcom.com>
---
drivers/scsi/lpfc/lpfc_version.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/scsi/lpfc/lpfc_version.h b/drivers/scsi/lpfc/lpfc_version.h
index d4e95e2..1c26dc6 100644
--- a/drivers/scsi/lpfc/lpfc_version.h
+++ b/drivers/scsi/lpfc/lpfc_version.h
@@ -20,7 +20,7 @@
* included with this package. *
*******************************************************************/
-#define LPFC_DRIVER_VERSION "11.2.0.10"
+#define LPFC_DRIVER_VERSION "11.2.0.12"
#define LPFC_DRIVER_NAME "lpfc"
/* Used for SLI 2/3 */
--
2.1.0
^ permalink raw reply related [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 22/22] lpfc revison 11.2.0.12
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 8:11 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 8:11 UTC (permalink / raw)
To: jsmart2021
Cc: linux-scsi, linux-nvme, sagi, martin.petersen, Dick Kennedy, James Smart
On Wed, Apr 19, 2017 at 09:46:41PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> Update lpfc version to reflect this set of changes.
>
> Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
> Signed-off-by: James Smart <james.smart@broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 22/22] lpfc revison 11.2.0.12
@ 2017-04-20 8:11 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-20 8:11 UTC (permalink / raw)
On Wed, Apr 19, 2017@09:46:41PM -0700, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> Update lpfc version to reflect this set of changes.
>
> Signed-off-by: Dick Kennedy <dick.kennedy at broadcom.com>
> Signed-off-by: James Smart <james.smart at broadcom.com>
> ---
Reviewed-by: Johannes Thumshirn <jthumshirn at suse.de>
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 00/22] lpfc updates for 11.2.0.12
2017-04-20 4:46 ` jsmart2021
@ 2017-04-20 18:30 ` Jens Axboe
-1 siblings, 0 replies; 100+ messages in thread
From: Jens Axboe @ 2017-04-20 18:30 UTC (permalink / raw)
To: jsmart2021; +Cc: linux-scsi, linux-nvme, sagi, martin.petersen
On Wed, Apr 19 2017, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021@gmail.com>
>
> This patch set provides a number of bug fixes, code cleanups, and error
> handling, mostly in the nvme area of lpfc.
>
> There is one new feature in the series to add Fabric Assigned WWN support.
>
> The patches were cut against the linux-block tree, for-4.12/block branch,
> and are intended to be pulled in through that tree rather than the scsi trees.
>
> The patches are dependent on the FC nvme/nvmet patches from the following 2
> series:
> http://lists.infradead.org/pipermail/linux-nvme/2017-April/009250.html
> http://lists.infradead.org/pipermail/linux-nvme/2017-April/009256.html
James, if you can address the (minor) review comments from Johannes and
repost with the reviewed/tested-by collected, we can get this applied. I
almost applied it as-is, but we may as well just get the last little
bits fixed up.
Thanks!
--
Jens Axboe
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 00/22] lpfc updates for 11.2.0.12
@ 2017-04-20 18:30 ` Jens Axboe
0 siblings, 0 replies; 100+ messages in thread
From: Jens Axboe @ 2017-04-20 18:30 UTC (permalink / raw)
On Wed, Apr 19 2017, jsmart2021@gmail.com wrote:
> From: James Smart <jsmart2021 at gmail.com>
>
> This patch set provides a number of bug fixes, code cleanups, and error
> handling, mostly in the nvme area of lpfc.
>
> There is one new feature in the series to add Fabric Assigned WWN support.
>
> The patches were cut against the linux-block tree, for-4.12/block branch,
> and are intended to be pulled in through that tree rather than the scsi trees.
>
> The patches are dependent on the FC nvme/nvmet patches from the following 2
> series:
> http://lists.infradead.org/pipermail/linux-nvme/2017-April/009250.html
> http://lists.infradead.org/pipermail/linux-nvme/2017-April/009256.html
James, if you can address the (minor) review comments from Johannes and
repost with the reviewed/tested-by collected, we can get this applied. I
almost applied it as-is, but we may as well just get the last little
bits fixed up.
Thanks!
--
Jens Axboe
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 00/22] lpfc updates for 11.2.0.12
2017-04-20 18:30 ` Jens Axboe
@ 2017-04-20 18:54 ` James Smart
-1 siblings, 0 replies; 100+ messages in thread
From: James Smart @ 2017-04-20 18:54 UTC (permalink / raw)
To: Jens Axboe, jsmart2021; +Cc: linux-scsi, linux-nvme, sagi, martin.petersen
On 4/20/2017 11:30 AM, Jens Axboe wrote:
> On Wed, Apr 19 2017, jsmart2021@gmail.com wrote:
>> From: James Smart <jsmart2021@gmail.com>
>>
>> This patch set provides a number of bug fixes, code cleanups, and error
>> handling, mostly in the nvme area of lpfc.
>>
>> There is one new feature in the series to add Fabric Assigned WWN support.
>>
>> The patches were cut against the linux-block tree, for-4.12/block branch,
>> and are intended to be pulled in through that tree rather than the scsi trees.
>>
>> The patches are dependent on the FC nvme/nvmet patches from the following 2
>> series:
>> http://lists.infradead.org/pipermail/linux-nvme/2017-April/009250.html
>> http://lists.infradead.org/pipermail/linux-nvme/2017-April/009256.html
> James, if you can address the (minor) review comments from Johannes and
> repost with the reviewed/tested-by collected, we can get this applied. I
> almost applied it as-is, but we may as well just get the last little
> bits fixed up.
>
> Thanks!
>
yes - posting shortly.
-- james
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 00/22] lpfc updates for 11.2.0.12
@ 2017-04-20 18:54 ` James Smart
0 siblings, 0 replies; 100+ messages in thread
From: James Smart @ 2017-04-20 18:54 UTC (permalink / raw)
On 4/20/2017 11:30 AM, Jens Axboe wrote:
> On Wed, Apr 19 2017, jsmart2021@gmail.com wrote:
>> From: James Smart <jsmart2021 at gmail.com>
>>
>> This patch set provides a number of bug fixes, code cleanups, and error
>> handling, mostly in the nvme area of lpfc.
>>
>> There is one new feature in the series to add Fabric Assigned WWN support.
>>
>> The patches were cut against the linux-block tree, for-4.12/block branch,
>> and are intended to be pulled in through that tree rather than the scsi trees.
>>
>> The patches are dependent on the FC nvme/nvmet patches from the following 2
>> series:
>> http://lists.infradead.org/pipermail/linux-nvme/2017-April/009250.html
>> http://lists.infradead.org/pipermail/linux-nvme/2017-April/009256.html
> James, if you can address the (minor) review comments from Johannes and
> repost with the reviewed/tested-by collected, we can get this applied. I
> almost applied it as-is, but we may as well just get the last little
> bits fixed up.
>
> Thanks!
>
yes - posting shortly.
-- james
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 00/22] lpfc updates for 11.2.0.12
2017-04-20 18:30 ` Jens Axboe
@ 2017-04-21 7:00 ` Johannes Thumshirn
-1 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-21 7:00 UTC (permalink / raw)
To: Jens Axboe; +Cc: jsmart2021, linux-scsi, linux-nvme, sagi, martin.petersen
On Thu, Apr 20, 2017 at 12:30:48PM -0600, Jens Axboe wrote:
> On Wed, Apr 19 2017, jsmart2021@gmail.com wrote:
> > From: James Smart <jsmart2021@gmail.com>
> >
> > This patch set provides a number of bug fixes, code cleanups, and error
> > handling, mostly in the nvme area of lpfc.
> >
> > There is one new feature in the series to add Fabric Assigned WWN support.
> >
> > The patches were cut against the linux-block tree, for-4.12/block branch,
> > and are intended to be pulled in through that tree rather than the scsi trees.
> >
> > The patches are dependent on the FC nvme/nvmet patches from the following 2
> > series:
> > http://lists.infradead.org/pipermail/linux-nvme/2017-April/009250.html
> > http://lists.infradead.org/pipermail/linux-nvme/2017-April/009256.html
>
> James, if you can address the (minor) review comments from Johannes and
> repost with the reviewed/tested-by collected, we can get this applied. I
> almost applied it as-is, but we may as well just get the last little
> bits fixed up.
Hi Jens,
Just for clarification, shouldn't Martin take the patches? They're for
drivers/scsi/lpfc/ after all. I just want to avoid confusion and possible
merge conflicts.
Thanks,
Johannes
--
Johannes Thumshirn Storage
jthumshirn@suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 00/22] lpfc updates for 11.2.0.12
@ 2017-04-21 7:00 ` Johannes Thumshirn
0 siblings, 0 replies; 100+ messages in thread
From: Johannes Thumshirn @ 2017-04-21 7:00 UTC (permalink / raw)
On Thu, Apr 20, 2017@12:30:48PM -0600, Jens Axboe wrote:
> On Wed, Apr 19 2017, jsmart2021@gmail.com wrote:
> > From: James Smart <jsmart2021 at gmail.com>
> >
> > This patch set provides a number of bug fixes, code cleanups, and error
> > handling, mostly in the nvme area of lpfc.
> >
> > There is one new feature in the series to add Fabric Assigned WWN support.
> >
> > The patches were cut against the linux-block tree, for-4.12/block branch,
> > and are intended to be pulled in through that tree rather than the scsi trees.
> >
> > The patches are dependent on the FC nvme/nvmet patches from the following 2
> > series:
> > http://lists.infradead.org/pipermail/linux-nvme/2017-April/009250.html
> > http://lists.infradead.org/pipermail/linux-nvme/2017-April/009256.html
>
> James, if you can address the (minor) review comments from Johannes and
> repost with the reviewed/tested-by collected, we can get this applied. I
> almost applied it as-is, but we may as well just get the last little
> bits fixed up.
Hi Jens,
Just for clarification, shouldn't Martin take the patches? They're for
drivers/scsi/lpfc/ after all. I just want to avoid confusion and possible
merge conflicts.
Thanks,
Johannes
--
Johannes Thumshirn Storage
jthumshirn at suse.de +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 00/22] lpfc updates for 11.2.0.12
2017-04-21 7:00 ` Johannes Thumshirn
@ 2017-04-21 11:24 ` Martin K. Petersen
-1 siblings, 0 replies; 100+ messages in thread
From: Martin K. Petersen @ 2017-04-21 11:24 UTC (permalink / raw)
To: Johannes Thumshirn
Cc: Jens Axboe, jsmart2021, linux-scsi, linux-nvme, sagi, martin.petersen
Johannes Thumshirn <jthumshirn@suse.de> writes:
> Just for clarification, shouldn't Martin take the patches? They're for
> drivers/scsi/lpfc/ after all. I just want to avoid confusion and possible
> merge conflicts.
Yeah, at this point I think it's better we do a second stage merge
through 4.12/scsi-fixes. I'll set that up after Linus pulls from Jens.
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 00/22] lpfc updates for 11.2.0.12
@ 2017-04-21 11:24 ` Martin K. Petersen
0 siblings, 0 replies; 100+ messages in thread
From: Martin K. Petersen @ 2017-04-21 11:24 UTC (permalink / raw)
Johannes Thumshirn <jthumshirn at suse.de> writes:
> Just for clarification, shouldn't Martin take the patches? They're for
> drivers/scsi/lpfc/ after all. I just want to avoid confusion and possible
> merge conflicts.
Yeah, at this point I think it's better we do a second stage merge
through 4.12/scsi-fixes. I'll set that up after Linus pulls from Jens.
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply [flat|nested] 100+ messages in thread
* Re: [PATCH-v1 00/22] lpfc updates for 11.2.0.12
2017-04-21 11:24 ` Martin K. Petersen
@ 2017-04-21 16:41 ` James Smart
-1 siblings, 0 replies; 100+ messages in thread
From: James Smart @ 2017-04-21 16:41 UTC (permalink / raw)
To: Martin K. Petersen, Johannes Thumshirn
Cc: Jens Axboe, jsmart2021, linux-scsi, linux-nvme, sagi
On 4/21/2017 4:24 AM, Martin K. Petersen wrote:
> Johannes Thumshirn <jthumshirn@suse.de> writes:
>
>> Just for clarification, shouldn't Martin take the patches? They're for
>> drivers/scsi/lpfc/ after all. I just want to avoid confusion and possible
>> merge conflicts.
> Yeah, at this point I think it's better we do a second stage merge
> through 4.12/scsi-fixes. I'll set that up after Linus pulls from Jens.
>
no - most of the fixes are nvme related. There are at best 3 that are
"generic" or scsi specific. Given the nvme parts need the nvme-fc api
in the 2 dependencies... this won't be something that works well going
through scsi.
-- james
^ permalink raw reply [flat|nested] 100+ messages in thread
* [PATCH-v1 00/22] lpfc updates for 11.2.0.12
@ 2017-04-21 16:41 ` James Smart
0 siblings, 0 replies; 100+ messages in thread
From: James Smart @ 2017-04-21 16:41 UTC (permalink / raw)
On 4/21/2017 4:24 AM, Martin K. Petersen wrote:
> Johannes Thumshirn <jthumshirn at suse.de> writes:
>
>> Just for clarification, shouldn't Martin take the patches? They're for
>> drivers/scsi/lpfc/ after all. I just want to avoid confusion and possible
>> merge conflicts.
> Yeah, at this point I think it's better we do a second stage merge
> through 4.12/scsi-fixes. I'll set that up after Linus pulls from Jens.
>
no - most of the fixes are nvme related. There are at best 3 that are
"generic" or scsi specific. Given the nvme parts need the nvme-fc api
in the 2 dependencies... this won't be something that works well going
through scsi.
-- james
^ permalink raw reply [flat|nested] 100+ messages in thread