From: jsmart2021@gmail.com (James Smart)
Subject: [PATCH 2/2] lpfc: support nvmet_fc defer_rcv callback
Date: Tue, 1 Aug 2017 15:12:40 -0700 [thread overview]
Message-ID: <20170801221240.31723-3-jsmart2021@gmail.com> (raw)
In-Reply-To: <20170801221240.31723-1-jsmart2021@gmail.com>
Currently, calls to nvmet_fc_rcv_fcp_req() always copied the
FC-NVME cmd iu to a temporary buffer before returning, allowing
the driver to immediately repost the buffer to the hardware.
To address timing conditions on queue element structures vs async
command reception, the nvmet_fc transport occasionally may need to
hold on to the command iu buffer for a short period. In these cases,
the nvmet_fc_rcv_fcp_req() will return a special return code
(-EOVERFLOW). In these cases, the LLDD must delay until the new
defer_rcv lldd callback is called before recycling the buffer back
to the hw.
This patch adds support for the new nvmet_fc transport defer_rcv
callback and recognition of the new error code when passing commands
to the transport.
This patch is intended to enter the kernel through the nvme block
tree which pulls in the nvmet_fc api change at the same time. It is
not to be merged via the scsi trees without the latest nvme support
in it.
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 | 4 +++-
drivers/scsi/lpfc/lpfc_debugfs.c | 5 ++++-
drivers/scsi/lpfc/lpfc_nvmet.c | 30 ++++++++++++++++++++++++++++++
drivers/scsi/lpfc/lpfc_nvmet.h | 1 +
4 files changed, 38 insertions(+), 2 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 4ed48ed38e79..7ee1a94c0b33 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -205,8 +205,10 @@ lpfc_nvme_info_show(struct device *dev, struct device_attribute *attr,
atomic_read(&tgtp->xmt_ls_rsp_error));
len += snprintf(buf+len, PAGE_SIZE-len,
- "FCP: Rcv %08x Release %08x Drop %08x\n",
+ "FCP: Rcv %08x Defer %08x Release %08x "
+ "Drop %08x\n",
atomic_read(&tgtp->rcv_fcp_cmd_in),
+ atomic_read(&tgtp->rcv_fcp_cmd_defer),
atomic_read(&tgtp->xmt_fcp_release),
atomic_read(&tgtp->rcv_fcp_cmd_drop));
diff --git a/drivers/scsi/lpfc/lpfc_debugfs.c b/drivers/scsi/lpfc/lpfc_debugfs.c
index 5cc8b0f7d885..744f3f395b64 100644
--- a/drivers/scsi/lpfc/lpfc_debugfs.c
+++ b/drivers/scsi/lpfc/lpfc_debugfs.c
@@ -782,8 +782,11 @@ lpfc_debugfs_nvmestat_data(struct lpfc_vport *vport, char *buf, int size)
atomic_read(&tgtp->xmt_ls_rsp_error));
len += snprintf(buf + len, size - len,
- "FCP: Rcv %08x Drop %08x\n",
+ "FCP: Rcv %08x Defer %08x Release %08x "
+ "Drop %08x\n",
atomic_read(&tgtp->rcv_fcp_cmd_in),
+ atomic_read(&tgtp->rcv_fcp_cmd_defer),
+ atomic_read(&tgtp->xmt_fcp_release),
atomic_read(&tgtp->rcv_fcp_cmd_drop));
if (atomic_read(&tgtp->rcv_fcp_cmd_in) !=
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index fbeec344c6cc..a5dc1b83064a 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -841,12 +841,31 @@ lpfc_nvmet_xmt_fcp_release(struct nvmet_fc_target_port *tgtport,
lpfc_nvmet_ctxbuf_post(phba, ctxp->ctxbuf);
}
+static void
+lpfc_nvmet_defer_rcv(struct nvmet_fc_target_port *tgtport,
+ struct nvmefc_tgt_fcp_req *rsp)
+{
+ struct lpfc_nvmet_tgtport *tgtp;
+ struct lpfc_nvmet_rcv_ctx *ctxp =
+ container_of(rsp, struct lpfc_nvmet_rcv_ctx, ctx.fcp_req);
+ struct rqb_dmabuf *nvmebuf = ctxp->rqb_buffer;
+ struct lpfc_hba *phba = ctxp->phba;
+
+ lpfc_nvmeio_data(phba, "NVMET DEFERRCV: xri x%x sz %d CPU %02x\n",
+ ctxp->oxid, ctxp->size, smp_processor_id());
+
+ tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
+ atomic_inc(&tgtp->rcv_fcp_cmd_defer);
+ lpfc_rq_buf_free(phba, &nvmebuf->hbuf); /* repost */
+}
+
static struct nvmet_fc_target_template lpfc_tgttemplate = {
.targetport_delete = lpfc_nvmet_targetport_delete,
.xmt_ls_rsp = lpfc_nvmet_xmt_ls_rsp,
.fcp_op = lpfc_nvmet_xmt_fcp_op,
.fcp_abort = lpfc_nvmet_xmt_fcp_abort,
.fcp_req_release = lpfc_nvmet_xmt_fcp_release,
+ .defer_rcv = lpfc_nvmet_defer_rcv,
.max_hw_queues = 1,
.max_sgl_segments = LPFC_NVMET_DEFAULT_SEGS,
@@ -1504,6 +1523,17 @@ lpfc_nvmet_unsol_fcp_buffer(struct lpfc_hba *phba,
return;
}
+ /* Processing of FCP command is deferred */
+ if (rc == -EOVERFLOW) {
+ lpfc_nvmeio_data(phba,
+ "NVMET RCV BUSY: xri x%x sz %d from %06x\n",
+ oxid, size, sid);
+ /* defer reposting rcv buffer till .defer_rcv callback */
+ ctxp->rqb_buffer = (void *)nvmebuf;
+ atomic_inc(&tgtp->rcv_fcp_cmd_out);
+ return;
+ }
+
atomic_inc(&tgtp->rcv_fcp_cmd_drop);
lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
"6159 FCP Drop IO x%x: err x%x: x%x x%x x%x\n",
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.h b/drivers/scsi/lpfc/lpfc_nvmet.h
index e675ef17be08..48a76788b003 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.h
+++ b/drivers/scsi/lpfc/lpfc_nvmet.h
@@ -49,6 +49,7 @@ struct lpfc_nvmet_tgtport {
atomic_t rcv_fcp_cmd_in;
atomic_t rcv_fcp_cmd_out;
atomic_t rcv_fcp_cmd_drop;
+ atomic_t rcv_fcp_cmd_defer;
atomic_t xmt_fcp_release;
/* Stats counters - lpfc_nvmet_xmt_fcp_op */
--
2.13.1
WARNING: multiple messages have this Message-ID (diff)
From: James Smart <jsmart2021@gmail.com>
To: linux-nvme@lists.infradead.org
Cc: linux-scsi@vger.kernel.org, James Smart <jsmart2021@gmail.com>,
Dick Kennedy <dick.kennedy@broadcom.com>,
James Smart <james.smart@broadcom.com>
Subject: [PATCH 2/2] lpfc: support nvmet_fc defer_rcv callback
Date: Tue, 1 Aug 2017 15:12:40 -0700 [thread overview]
Message-ID: <20170801221240.31723-3-jsmart2021@gmail.com> (raw)
In-Reply-To: <20170801221240.31723-1-jsmart2021@gmail.com>
Currently, calls to nvmet_fc_rcv_fcp_req() always copied the
FC-NVME cmd iu to a temporary buffer before returning, allowing
the driver to immediately repost the buffer to the hardware.
To address timing conditions on queue element structures vs async
command reception, the nvmet_fc transport occasionally may need to
hold on to the command iu buffer for a short period. In these cases,
the nvmet_fc_rcv_fcp_req() will return a special return code
(-EOVERFLOW). In these cases, the LLDD must delay until the new
defer_rcv lldd callback is called before recycling the buffer back
to the hw.
This patch adds support for the new nvmet_fc transport defer_rcv
callback and recognition of the new error code when passing commands
to the transport.
This patch is intended to enter the kernel through the nvme block
tree which pulls in the nvmet_fc api change at the same time. It is
not to be merged via the scsi trees without the latest nvme support
in it.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
---
drivers/scsi/lpfc/lpfc_attr.c | 4 +++-
drivers/scsi/lpfc/lpfc_debugfs.c | 5 ++++-
drivers/scsi/lpfc/lpfc_nvmet.c | 30 ++++++++++++++++++++++++++++++
drivers/scsi/lpfc/lpfc_nvmet.h | 1 +
4 files changed, 38 insertions(+), 2 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 4ed48ed38e79..7ee1a94c0b33 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -205,8 +205,10 @@ lpfc_nvme_info_show(struct device *dev, struct device_attribute *attr,
atomic_read(&tgtp->xmt_ls_rsp_error));
len += snprintf(buf+len, PAGE_SIZE-len,
- "FCP: Rcv %08x Release %08x Drop %08x\n",
+ "FCP: Rcv %08x Defer %08x Release %08x "
+ "Drop %08x\n",
atomic_read(&tgtp->rcv_fcp_cmd_in),
+ atomic_read(&tgtp->rcv_fcp_cmd_defer),
atomic_read(&tgtp->xmt_fcp_release),
atomic_read(&tgtp->rcv_fcp_cmd_drop));
diff --git a/drivers/scsi/lpfc/lpfc_debugfs.c b/drivers/scsi/lpfc/lpfc_debugfs.c
index 5cc8b0f7d885..744f3f395b64 100644
--- a/drivers/scsi/lpfc/lpfc_debugfs.c
+++ b/drivers/scsi/lpfc/lpfc_debugfs.c
@@ -782,8 +782,11 @@ lpfc_debugfs_nvmestat_data(struct lpfc_vport *vport, char *buf, int size)
atomic_read(&tgtp->xmt_ls_rsp_error));
len += snprintf(buf + len, size - len,
- "FCP: Rcv %08x Drop %08x\n",
+ "FCP: Rcv %08x Defer %08x Release %08x "
+ "Drop %08x\n",
atomic_read(&tgtp->rcv_fcp_cmd_in),
+ atomic_read(&tgtp->rcv_fcp_cmd_defer),
+ atomic_read(&tgtp->xmt_fcp_release),
atomic_read(&tgtp->rcv_fcp_cmd_drop));
if (atomic_read(&tgtp->rcv_fcp_cmd_in) !=
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c
index fbeec344c6cc..a5dc1b83064a 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.c
+++ b/drivers/scsi/lpfc/lpfc_nvmet.c
@@ -841,12 +841,31 @@ lpfc_nvmet_xmt_fcp_release(struct nvmet_fc_target_port *tgtport,
lpfc_nvmet_ctxbuf_post(phba, ctxp->ctxbuf);
}
+static void
+lpfc_nvmet_defer_rcv(struct nvmet_fc_target_port *tgtport,
+ struct nvmefc_tgt_fcp_req *rsp)
+{
+ struct lpfc_nvmet_tgtport *tgtp;
+ struct lpfc_nvmet_rcv_ctx *ctxp =
+ container_of(rsp, struct lpfc_nvmet_rcv_ctx, ctx.fcp_req);
+ struct rqb_dmabuf *nvmebuf = ctxp->rqb_buffer;
+ struct lpfc_hba *phba = ctxp->phba;
+
+ lpfc_nvmeio_data(phba, "NVMET DEFERRCV: xri x%x sz %d CPU %02x\n",
+ ctxp->oxid, ctxp->size, smp_processor_id());
+
+ tgtp = (struct lpfc_nvmet_tgtport *)phba->targetport->private;
+ atomic_inc(&tgtp->rcv_fcp_cmd_defer);
+ lpfc_rq_buf_free(phba, &nvmebuf->hbuf); /* repost */
+}
+
static struct nvmet_fc_target_template lpfc_tgttemplate = {
.targetport_delete = lpfc_nvmet_targetport_delete,
.xmt_ls_rsp = lpfc_nvmet_xmt_ls_rsp,
.fcp_op = lpfc_nvmet_xmt_fcp_op,
.fcp_abort = lpfc_nvmet_xmt_fcp_abort,
.fcp_req_release = lpfc_nvmet_xmt_fcp_release,
+ .defer_rcv = lpfc_nvmet_defer_rcv,
.max_hw_queues = 1,
.max_sgl_segments = LPFC_NVMET_DEFAULT_SEGS,
@@ -1504,6 +1523,17 @@ lpfc_nvmet_unsol_fcp_buffer(struct lpfc_hba *phba,
return;
}
+ /* Processing of FCP command is deferred */
+ if (rc == -EOVERFLOW) {
+ lpfc_nvmeio_data(phba,
+ "NVMET RCV BUSY: xri x%x sz %d from %06x\n",
+ oxid, size, sid);
+ /* defer reposting rcv buffer till .defer_rcv callback */
+ ctxp->rqb_buffer = (void *)nvmebuf;
+ atomic_inc(&tgtp->rcv_fcp_cmd_out);
+ return;
+ }
+
atomic_inc(&tgtp->rcv_fcp_cmd_drop);
lpfc_printf_log(phba, KERN_ERR, LOG_NVME_IOERR,
"6159 FCP Drop IO x%x: err x%x: x%x x%x x%x\n",
diff --git a/drivers/scsi/lpfc/lpfc_nvmet.h b/drivers/scsi/lpfc/lpfc_nvmet.h
index e675ef17be08..48a76788b003 100644
--- a/drivers/scsi/lpfc/lpfc_nvmet.h
+++ b/drivers/scsi/lpfc/lpfc_nvmet.h
@@ -49,6 +49,7 @@ struct lpfc_nvmet_tgtport {
atomic_t rcv_fcp_cmd_in;
atomic_t rcv_fcp_cmd_out;
atomic_t rcv_fcp_cmd_drop;
+ atomic_t rcv_fcp_cmd_defer;
atomic_t xmt_fcp_release;
/* Stats counters - lpfc_nvmet_xmt_fcp_op */
--
2.13.1
next prev parent reply other threads:[~2017-08-01 22:12 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-01 22:12 [PATCH 0/2] nvmet_fc: work around overalloc of queue elements James Smart
2017-08-01 22:12 ` James Smart
2017-08-01 22:12 ` [PATCH 1/2] nvmet_fc: add defer_req callback for deferment of cmd buffer return James Smart
2017-08-01 22:12 ` James Smart
2017-08-02 8:16 ` Johannes Thumshirn
2017-08-02 8:16 ` Johannes Thumshirn
2017-08-02 14:29 ` James Smart
2017-08-02 14:29 ` James Smart
2017-08-10 9:04 ` Christoph Hellwig
2017-08-10 9:04 ` Christoph Hellwig
2017-08-10 9:06 ` Christoph Hellwig
2017-08-10 9:06 ` Christoph Hellwig
2017-08-16 7:53 ` Christoph Hellwig
2017-08-16 7:53 ` Christoph Hellwig
2017-08-01 22:12 ` James Smart [this message]
2017-08-01 22:12 ` [PATCH 2/2] lpfc: support nvmet_fc defer_rcv callback James Smart
2017-08-02 8:32 ` Johannes Thumshirn
2017-08-02 8:32 ` Johannes Thumshirn
2017-08-02 9:17 ` Johannes Thumshirn
2017-08-02 9:17 ` Johannes Thumshirn
2017-08-02 14:32 ` James Smart
2017-08-02 14:32 ` James Smart
2017-08-10 9:08 ` Christoph Hellwig
2017-08-10 9:08 ` Christoph Hellwig
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170801221240.31723-3-jsmart2021@gmail.com \
--to=jsmart2021@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.