Skip to contentSkip to conversation

nfsd: observability improvements

PATCH Superseded Jeff Layton 3 participants 27 messages Apr 9, 2025 – Apr 11, 2025

Subsystems (4)

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Jeff Layton

author maintainer
#0 lore raw

v2 cover nfsd: observability improvements

While troubleshooting a performance problem internally, it became evident that we needed tracepoints in nfsd_commit. The first patch adds that. While discussing that, Sargun pointed out some tracepoints he added using kprobes. Those are converted to static tracepoints here, and the legacy dprintk's removed.

Lastly, I've updated the svc_xprt_dequeue tracepoint to show how long the xprt sat on the queue before being serviced.

---
Changes in v2:
- Break tracepoints out into multiple patches
- Flesh out the tracepoints in these locations to display the same info
as legacy dprintks.
- have all the tracepoints SVC_XPRT_ENDPOINT_* info
- update svc_xprt_dequeue tracepoint to show how long xprt was on queue
- Link to v1: https://lore.kernel.org/r/[email protected]

---
Jeff Layton (12):
nfsd: add commit start/done tracepoints around nfsd_commit()
sunrpc: add info about xprt queue times to svc_xprt_dequeue tracepoint
sunrpc: move the SVC_RQST_EVENT_*() macros to common header
nfsd: add a tracepoint for nfsd_setattr
nfsd: add a tracepoint to nfsd_lookup_dentry
nfsd: add tracepoints around nfsd_create events
nfsd: add tracepoints for symlink events
nfsd: add tracepoints for hardlink events
nfsd: add tracepoints for unlink events
nfsd: add tracepoints to rename events
nfsd: add tracepoints for readdir events
nfsd: add tracepoint for getattr events

 fs/nfsd/nfs3proc.c              |  67 +++------
 fs/nfsd/nfs4proc.c              |  45 ++++++
 fs/nfsd/nfsproc.c               |  39 ++----
 fs/nfsd/trace.h                 | 298 ++++++++++++++++++++++++++++++++++++++++
 fs/nfsd/vfs.c                   |  10 +-
 include/linux/sunrpc/svc_xprt.h |   1 +
 include/trace/events/sunrpc.h   |  36 +----
 include/trace/misc/fs.h         |  21 +++
 include/trace/misc/sunrpc.h     |  23 ++++
 net/sunrpc/svc_xprt.c           |   1 +
 10 files changed, 433 insertions(+), 108 deletions(-)

---
base-commit: 71238ba71a67aab408cfe14b6a5ae3c9b83082f9
change-id: 20250303-nfsd-tracepoints-c82add9155a6

Best regards,

Signature · 1 line
Jeff Layton <[email protected]>
author maintainer
#1 lore raw

PATCH v2 1/12 nfsd: add commit start/done tracepoints around nfsd_commit()

Very useful for gauging how long the vfs_fsync_range() takes.

---
 fs/nfsd/trace.h | 2 ++
 fs/nfsd/vfs.c   | 3 +++
 2 files changed, 5 insertions(+)
fs/nfsd/trace.h +2 -0
index a7630e9f657776a9335ba68ad223641e3ed9121a..0d49fc064f7273f32c93732a993fd77bc0783f5d 100644
@@ -451,6 +451,8 @@ DEFINE_NFSD_IO_EVENT(write_start);
 DEFINE_NFSD_IO_EVENT(write_opened);
 DEFINE_NFSD_IO_EVENT(write_io_done);
 DEFINE_NFSD_IO_EVENT(write_done);
+DEFINE_NFSD_IO_EVENT(commit_start);
+DEFINE_NFSD_IO_EVENT(commit_done);
 
 DECLARE_EVENT_CLASS(nfsd_err_class,
 	TP_PROTO(struct svc_rqst *rqstp,
fs/nfsd/vfs.c +3 -0
index 86e199a758f6c69c508151fb14886f4b5da9c080..d1156a18a79579bf427fe5809dc93d06e241201e 100644
@@ -1366,6 +1366,8 @@ nfsd_commit(struct svc_rqst *rqstp, struct svc_fh *fhp, struct nfsd_file *nf,
 	loff_t			start, end;
 	struct nfsd_net		*nn;
 
+	trace_nfsd_commit_start(rqstp, fhp, offset, count);
+
 	/*
 	 * Convert the client-provided (offset, count) range to a
 	 * (start, end) range. If the client-provided range falls
@@ -1404,6 +1406,7 @@ nfsd_commit(struct svc_rqst *rqstp, struct svc_fh *fhp, struct nfsd_file *nf,
 	} else
 		nfsd_copy_write_verifier(verf, nn);
 
+	trace_nfsd_commit_done(rqstp, fhp, offset, count);
 	return err;
 }
 
Signature · 1 line
2.49.0
author maintainer
#2 lore raw

PATCH v2 2/12 sunrpc: add info about xprt queue times to svc_xprt_dequeue tracepoint

In reply to Jeff Layton

Currently, this tracepoint displays "wakeup-us", which is the time that the woken thread spent sleeping, before dequeueing the next xprt. Add a new statistic that shows how long the xprt sat on the queue before being serviced.

---
 include/linux/sunrpc/svc_xprt.h |  1 +
 include/trace/events/sunrpc.h   | 13 +++++++------
 net/sunrpc/svc_xprt.c           |  1 +
 3 files changed, 9 insertions(+), 6 deletions(-)
include/linux/sunrpc/svc_xprt.h +1 -0
index 72be609525796792274d5b8cb5ff37f73723fc23..369a89aea18618748607ee943247c327bf62c8d5 100644
@@ -53,6 +53,7 @@ struct svc_xprt {
 	struct svc_xprt_class	*xpt_class;
 	const struct svc_xprt_ops *xpt_ops;
 	struct kref		xpt_ref;
+	ktime_t			xpt_qtime;
 	struct list_head	xpt_list;
 	struct lwq_node		xpt_ready;
 	unsigned long		xpt_flags;
include/trace/events/sunrpc.h +7 -6
index 5d331383047b79b9f6dcd699c87287453c1a5f49..b5a0f0bc1a3b7cfd90ce0181a8a419db810988bb 100644
@@ -2040,19 +2040,20 @@ TRACE_EVENT(svc_xprt_dequeue,
 
 	TP_STRUCT__entry(
 		SVC_XPRT_ENDPOINT_FIELDS(rqst->rq_xprt)
-
 		__field(unsigned long, wakeup)
+		__field(unsigned long, qtime)
 	),
 
 	TP_fast_assign(
-		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		ktime_t ktime = ktime_get();
 
-		__entry->wakeup = ktime_to_us(ktime_sub(ktime_get(),
-							rqst->rq_qtime));
+		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		__entry->wakeup = ktime_to_us(ktime_sub(ktime, rqst->rq_qtime));
+		__entry->qtime = ktime_to_us(ktime_sub(ktime, rqst->rq_xprt->xpt_qtime));
 	),
 
-	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu",
-		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup)
+	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu qtime=%lu",
+		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup, __entry->qtime)
 );
 
 DECLARE_EVENT_CLASS(svc_xprt_event,
net/sunrpc/svc_xprt.c +1 -0
index ae25405d8bd22672a361d1fd3adfdcebb403f90f..32018557797b1f683d8b7259f5fccd029aebcd71 100644
@@ -488,6 +488,7 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 	pool = svc_pool_for_cpu(xprt->xpt_server);
 
 	percpu_counter_inc(&pool->sp_sockets_queued);
+	xprt->xpt_qtime = ktime_get();
 	lwq_enqueue(&xprt->xpt_ready, &pool->sp_xprts);
 
 	svc_pool_wake_idle_thread(pool);
Signature · 1 line
2.49.0
#3 lore raw

Re: [PATCH v2 02/12] sunrpc: add info about xprt queue times to svc_xprt_dequeue tracepoint

Quoted 4 lines · Jeff Layton

Currently, this tracepoint displays "wakeup-us", which is the time that the woken thread spent sleeping, before dequeueing the next xprt. Add a new statistic that shows how long the xprt sat on the queue before being serviced.

I don't understand the difference between "waiting on queue" and "sleeping". When are those two latency measurements not the same?

Quoted 63 lines
---
 include/linux/sunrpc/svc_xprt.h |  1 +
 include/trace/events/sunrpc.h   | 13 +++++++------
 net/sunrpc/svc_xprt.c           |  1 +
 3 files changed, 9 insertions(+), 6 deletions(-)
include/linux/sunrpc/svc_xprt.h +1 -0
index 72be609525796792274d5b8cb5ff37f73723fc23..369a89aea18618748607ee943247c327bf62c8d5 100644
@@ -53,6 +53,7 @@ struct svc_xprt {
 	struct svc_xprt_class	*xpt_class;
 	const struct svc_xprt_ops *xpt_ops;
 	struct kref		xpt_ref;
+	ktime_t			xpt_qtime;
 	struct list_head	xpt_list;
 	struct lwq_node		xpt_ready;
 	unsigned long		xpt_flags;
include/trace/events/sunrpc.h +7 -6
index 5d331383047b79b9f6dcd699c87287453c1a5f49..b5a0f0bc1a3b7cfd90ce0181a8a419db810988bb 100644
@@ -2040,19 +2040,20 @@ TRACE_EVENT(svc_xprt_dequeue,
 
 	TP_STRUCT__entry(
 		SVC_XPRT_ENDPOINT_FIELDS(rqst->rq_xprt)
-
 		__field(unsigned long, wakeup)
+		__field(unsigned long, qtime)
 	),
 
 	TP_fast_assign(
-		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		ktime_t ktime = ktime_get();
 
-		__entry->wakeup = ktime_to_us(ktime_sub(ktime_get(),
-							rqst->rq_qtime));
+		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		__entry->wakeup = ktime_to_us(ktime_sub(ktime, rqst->rq_qtime));
+		__entry->qtime = ktime_to_us(ktime_sub(ktime, rqst->rq_xprt->xpt_qtime));
 	),
 
-	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu",
-		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup)
+	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu qtime=%lu",
+		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup, __entry->qtime)
 );
 
 DECLARE_EVENT_CLASS(svc_xprt_event,
net/sunrpc/svc_xprt.c +1 -0
index ae25405d8bd22672a361d1fd3adfdcebb403f90f..32018557797b1f683d8b7259f5fccd029aebcd71 100644
@@ -488,6 +488,7 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 	pool = svc_pool_for_cpu(xprt->xpt_server);
 
 	percpu_counter_inc(&pool->sp_sockets_queued);
+	xprt->xpt_qtime = ktime_get();
 	lwq_enqueue(&xprt->xpt_ready, &pool->sp_xprts);
 
 	svc_pool_wake_idle_thread(pool);
Signature · 1 line
Chuck Lever
author maintainer
#4 lore raw

Re: [PATCH v2 02/12] sunrpc: add info about xprt queue times to svc_xprt_dequeue tracepoint

Quoted 9 lines · Chuck Lever
Quoted 4 lines · Jeff Layton

Currently, this tracepoint displays "wakeup-us", which is the time that the woken thread spent sleeping, before dequeueing the next xprt. Add a new statistic that shows how long the xprt sat on the queue before being serviced.

I don't understand the difference between "waiting on queue" and "sleeping". When are those two latency measurements not the same?

These are measuring two different things:

svc_rqst->rq_qtime represents the time between when thread on the sp_idle_threads list was woken. This patch adds svc_xprt->xpt_qtime, which represents the time that the svc_xprt was added to the lwq.

The first tells us how long the interval was between the thread being woken and the xprt being dequeued. The new statistic tells us how long between the xprt being enqueued and dequeued.

They could easily diverge if there were not enough threads available to service all of the queued xprts.

Quoted 66 lines
Quoted 64 lines
---
 include/linux/sunrpc/svc_xprt.h |  1 +
 include/trace/events/sunrpc.h   | 13 +++++++------
 net/sunrpc/svc_xprt.c           |  1 +
 3 files changed, 9 insertions(+), 6 deletions(-)
include/linux/sunrpc/svc_xprt.h +1 -0
index 72be609525796792274d5b8cb5ff37f73723fc23..369a89aea18618748607ee943247c327bf62c8d5 100644
@@ -53,6 +53,7 @@ struct svc_xprt {
 	struct svc_xprt_class	*xpt_class;
 	const struct svc_xprt_ops *xpt_ops;
 	struct kref		xpt_ref;
+	ktime_t			xpt_qtime;
 	struct list_head	xpt_list;
 	struct lwq_node		xpt_ready;
 	unsigned long		xpt_flags;
include/trace/events/sunrpc.h +7 -6
index 5d331383047b79b9f6dcd699c87287453c1a5f49..b5a0f0bc1a3b7cfd90ce0181a8a419db810988bb 100644
@@ -2040,19 +2040,20 @@ TRACE_EVENT(svc_xprt_dequeue,
 
 	TP_STRUCT__entry(
 		SVC_XPRT_ENDPOINT_FIELDS(rqst->rq_xprt)
-
 		__field(unsigned long, wakeup)
+		__field(unsigned long, qtime)
 	),
 
 	TP_fast_assign(
-		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		ktime_t ktime = ktime_get();
 
-		__entry->wakeup = ktime_to_us(ktime_sub(ktime_get(),
-							rqst->rq_qtime));
+		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		__entry->wakeup = ktime_to_us(ktime_sub(ktime, rqst->rq_qtime));
+		__entry->qtime = ktime_to_us(ktime_sub(ktime, rqst->rq_xprt->xpt_qtime));
 	),
 
-	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu",
-		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup)
+	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu qtime=%lu",
+		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup, __entry->qtime)
 );
 
 DECLARE_EVENT_CLASS(svc_xprt_event,
net/sunrpc/svc_xprt.c +1 -0
index ae25405d8bd22672a361d1fd3adfdcebb403f90f..32018557797b1f683d8b7259f5fccd029aebcd71 100644
@@ -488,6 +488,7 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 	pool = svc_pool_for_cpu(xprt->xpt_server);
 
 	percpu_counter_inc(&pool->sp_sockets_queued);
+	xprt->xpt_qtime = ktime_get();
 	lwq_enqueue(&xprt->xpt_ready, &pool->sp_xprts);
 
 	svc_pool_wake_idle_thread(pool);
Signature · 1 line
Jeff Layton <[email protected]>
author maintainer
#5 lore raw

Re: [PATCH v2 02/12] sunrpc: add info about xprt queue times to svc_xprt_dequeue tracepoint

Quoted 17 lines · Jeff Layton
Quoted 9 lines · Chuck Lever
Quoted 4 lines · Jeff Layton

Currently, this tracepoint displays "wakeup-us", which is the time that the woken thread spent sleeping, before dequeueing the next xprt. Add a new statistic that shows how long the xprt sat on the queue before being serviced.

I don't understand the difference between "waiting on queue" and "sleeping". When are those two latency measurements not the same?

These are measuring two different things:

svc_rqst->rq_qtime represents the time between when thread on the sp_idle_threads list was woken. This patch adds svc_xprt->xpt_qtime, which represents the time that the svc_xprt was added to the lwq.

Quoted 7 lines

The first tells us how long the interval was between the thread being woken and the xprt being dequeued. The new statistic tells us how long between the xprt being enqueued and dequeued.

They could easily diverge if there were not enough threads available to service all of the queued xprts.

Hi Chuck! If you're OK with my rationale above, I'd like to expedite merging this patch in particular.

The reason is that we have clients with the nfs_layout_flexfiles dataserver_timeo module parameter set for 6s. This helps them switch to an alternate mirror when a DS goes down, but we see a lot of RPC timeouts when this is set.

My theory is that the xprts are getting queued and it's taking a long time for a thread to pick it up. That should show up as a large value in the qtime field in this tracepoint if I'm correct.

Would you be amenable to that?

Quoted 67 lines
Quoted 66 lines
Quoted 64 lines
---
 include/linux/sunrpc/svc_xprt.h |  1 +
 include/trace/events/sunrpc.h   | 13 +++++++------
 net/sunrpc/svc_xprt.c           |  1 +
 3 files changed, 9 insertions(+), 6 deletions(-)
include/linux/sunrpc/svc_xprt.h +1 -0
index 72be609525796792274d5b8cb5ff37f73723fc23..369a89aea18618748607ee943247c327bf62c8d5 100644
@@ -53,6 +53,7 @@ struct svc_xprt {
 	struct svc_xprt_class	*xpt_class;
 	const struct svc_xprt_ops *xpt_ops;
 	struct kref		xpt_ref;
+	ktime_t			xpt_qtime;
 	struct list_head	xpt_list;
 	struct lwq_node		xpt_ready;
 	unsigned long		xpt_flags;
include/trace/events/sunrpc.h +7 -6
index 5d331383047b79b9f6dcd699c87287453c1a5f49..b5a0f0bc1a3b7cfd90ce0181a8a419db810988bb 100644
@@ -2040,19 +2040,20 @@ TRACE_EVENT(svc_xprt_dequeue,
 
 	TP_STRUCT__entry(
 		SVC_XPRT_ENDPOINT_FIELDS(rqst->rq_xprt)
-
 		__field(unsigned long, wakeup)
+		__field(unsigned long, qtime)
 	),
 
 	TP_fast_assign(
-		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		ktime_t ktime = ktime_get();
 
-		__entry->wakeup = ktime_to_us(ktime_sub(ktime_get(),
-							rqst->rq_qtime));
+		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		__entry->wakeup = ktime_to_us(ktime_sub(ktime, rqst->rq_qtime));
+		__entry->qtime = ktime_to_us(ktime_sub(ktime, rqst->rq_xprt->xpt_qtime));
 	),
 
-	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu",
-		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup)
+	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu qtime=%lu",
+		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup, __entry->qtime)
 );
 
 DECLARE_EVENT_CLASS(svc_xprt_event,
net/sunrpc/svc_xprt.c +1 -0
index ae25405d8bd22672a361d1fd3adfdcebb403f90f..32018557797b1f683d8b7259f5fccd029aebcd71 100644
@@ -488,6 +488,7 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 	pool = svc_pool_for_cpu(xprt->xpt_server);
 
 	percpu_counter_inc(&pool->sp_sockets_queued);
+	xprt->xpt_qtime = ktime_get();
 	lwq_enqueue(&xprt->xpt_ready, &pool->sp_xprts);
 
 	svc_pool_wake_idle_thread(pool);
Signature · 1 line
Jeff Layton <[email protected]>
#6 lore raw

Re: [PATCH v2 02/12] sunrpc: add info about xprt queue times to svc_xprt_dequeue tracepoint

Quoted 34 lines · Jeff Layton
Quoted 17 lines · Jeff Layton
Quoted 9 lines · Chuck Lever
Quoted 4 lines · Jeff Layton

Currently, this tracepoint displays "wakeup-us", which is the time that the woken thread spent sleeping, before dequeueing the next xprt. Add a new statistic that shows how long the xprt sat on the queue before being serviced.

I don't understand the difference between "waiting on queue" and "sleeping". When are those two latency measurements not the same?

These are measuring two different things:

svc_rqst->rq_qtime represents the time between when thread on the sp_idle_threads list was woken. This patch adds svc_xprt->xpt_qtime, which represents the time that the svc_xprt was added to the lwq.

Quoted 7 lines

The first tells us how long the interval was between the thread being woken and the xprt being dequeued. The new statistic tells us how long between the xprt being enqueued and dequeued.

They could easily diverge if there were not enough threads available to service all of the queued xprts.

Hi Chuck! If you're OK with my rationale above, I'd like to expedite merging this patch in particular.

The reason is that we have clients with the nfs_layout_flexfiles dataserver_timeo module parameter set for 6s. This helps them switch to an alternate mirror when a DS goes down, but we see a lot of RPC timeouts when this is set.

Quoted 5 lines

My theory is that the xprts are getting queued and it's taking a long time for a thread to pick it up. That should show up as a large value in the qtime field in this tracepoint if I'm correct.

Would you be amenable to that?

No objection, repost this one with the beefier rationale.

But it depends on what you mean by "expedite" -- v6.16 would be the next "normal" opportunity, since this change doesn't qualify as a bug fix.

Quoted 67 lines
Quoted 66 lines
Quoted 65 lines
Quoted 63 lines
---
 include/linux/sunrpc/svc_xprt.h |  1 +
 include/trace/events/sunrpc.h   | 13 +++++++------
 net/sunrpc/svc_xprt.c           |  1 +
 3 files changed, 9 insertions(+), 6 deletions(-)
include/linux/sunrpc/svc_xprt.h +1 -0
index 72be609525796792274d5b8cb5ff37f73723fc23..369a89aea18618748607ee943247c327bf62c8d5 100644
@@ -53,6 +53,7 @@ struct svc_xprt {
 	struct svc_xprt_class	*xpt_class;
 	const struct svc_xprt_ops *xpt_ops;
 	struct kref		xpt_ref;
+	ktime_t			xpt_qtime;
 	struct list_head	xpt_list;
 	struct lwq_node		xpt_ready;
 	unsigned long		xpt_flags;
include/trace/events/sunrpc.h +7 -6
index 5d331383047b79b9f6dcd699c87287453c1a5f49..b5a0f0bc1a3b7cfd90ce0181a8a419db810988bb 100644
@@ -2040,19 +2040,20 @@ TRACE_EVENT(svc_xprt_dequeue,
 
 	TP_STRUCT__entry(
 		SVC_XPRT_ENDPOINT_FIELDS(rqst->rq_xprt)
-
 		__field(unsigned long, wakeup)
+		__field(unsigned long, qtime)
 	),
 
 	TP_fast_assign(
-		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		ktime_t ktime = ktime_get();
 
-		__entry->wakeup = ktime_to_us(ktime_sub(ktime_get(),
-							rqst->rq_qtime));
+		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		__entry->wakeup = ktime_to_us(ktime_sub(ktime, rqst->rq_qtime));
+		__entry->qtime = ktime_to_us(ktime_sub(ktime, rqst->rq_xprt->xpt_qtime));
 	),
 
-	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu",
-		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup)
+	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu qtime=%lu",
+		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup, __entry->qtime)
 );
 
 DECLARE_EVENT_CLASS(svc_xprt_event,
net/sunrpc/svc_xprt.c +1 -0
index ae25405d8bd22672a361d1fd3adfdcebb403f90f..32018557797b1f683d8b7259f5fccd029aebcd71 100644
@@ -488,6 +488,7 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 	pool = svc_pool_for_cpu(xprt->xpt_server);
 
 	percpu_counter_inc(&pool->sp_sockets_queued);
+	xprt->xpt_qtime = ktime_get();
 	lwq_enqueue(&xprt->xpt_ready, &pool->sp_xprts);
 
 	svc_pool_wake_idle_thread(pool);
Signature · 1 line
Chuck Lever
author maintainer
#7 lore raw

Re: [PATCH v2 02/12] sunrpc: add info about xprt queue times to svc_xprt_dequeue tracepoint

Quoted 35 lines · Chuck Lever
Quoted 34 lines · Jeff Layton
Quoted 17 lines · Jeff Layton
Quoted 9 lines · Chuck Lever
Quoted 4 lines · Jeff Layton

Currently, this tracepoint displays "wakeup-us", which is the time that the woken thread spent sleeping, before dequeueing the next xprt. Add a new statistic that shows how long the xprt sat on the queue before being serviced.

I don't understand the difference between "waiting on queue" and "sleeping". When are those two latency measurements not the same?

These are measuring two different things:

svc_rqst->rq_qtime represents the time between when thread on the sp_idle_threads list was woken. This patch adds svc_xprt->xpt_qtime, which represents the time that the svc_xprt was added to the lwq.

Quoted 7 lines

The first tells us how long the interval was between the thread being woken and the xprt being dequeued. The new statistic tells us how long between the xprt being enqueued and dequeued.

They could easily diverge if there were not enough threads available to service all of the queued xprts.

Hi Chuck! If you're OK with my rationale above, I'd like to expedite merging this patch in particular.

The reason is that we have clients with the nfs_layout_flexfiles dataserver_timeo module parameter set for 6s. This helps them switch to an alternate mirror when a DS goes down, but we see a lot of RPC timeouts when this is set.

Quoted 8 lines
Quoted 5 lines

My theory is that the xprts are getting queued and it's taking a long time for a thread to pick it up. That should show up as a large value in the qtime field in this tracepoint if I'm correct.

Would you be amenable to that?

No objection, repost this one with the beefier rationale.

Will do.

Quoted 4 lines

But it depends on what you mean by "expedite" -- v6.16 would be the next "normal" opportunity, since this change doesn't qualify as a bug fix.

I was hoping for v6.15, as I don't want to wait months to figure this out. Eventually, I need this in a particular downstream distribution that is reticent to take patches that aren't already merged.

Quoted 70 lines
Quoted 68 lines
Quoted 67 lines
Quoted 66 lines
Quoted 64 lines
---
 include/linux/sunrpc/svc_xprt.h |  1 +
 include/trace/events/sunrpc.h   | 13 +++++++------
 net/sunrpc/svc_xprt.c           |  1 +
 3 files changed, 9 insertions(+), 6 deletions(-)
include/linux/sunrpc/svc_xprt.h +1 -0
index 72be609525796792274d5b8cb5ff37f73723fc23..369a89aea18618748607ee943247c327bf62c8d5 100644
@@ -53,6 +53,7 @@ struct svc_xprt {
 	struct svc_xprt_class	*xpt_class;
 	const struct svc_xprt_ops *xpt_ops;
 	struct kref		xpt_ref;
+	ktime_t			xpt_qtime;
 	struct list_head	xpt_list;
 	struct lwq_node		xpt_ready;
 	unsigned long		xpt_flags;
include/trace/events/sunrpc.h +7 -6
index 5d331383047b79b9f6dcd699c87287453c1a5f49..b5a0f0bc1a3b7cfd90ce0181a8a419db810988bb 100644
@@ -2040,19 +2040,20 @@ TRACE_EVENT(svc_xprt_dequeue,
 
 	TP_STRUCT__entry(
 		SVC_XPRT_ENDPOINT_FIELDS(rqst->rq_xprt)
-
 		__field(unsigned long, wakeup)
+		__field(unsigned long, qtime)
 	),
 
 	TP_fast_assign(
-		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		ktime_t ktime = ktime_get();
 
-		__entry->wakeup = ktime_to_us(ktime_sub(ktime_get(),
-							rqst->rq_qtime));
+		SVC_XPRT_ENDPOINT_ASSIGNMENTS(rqst->rq_xprt);
+		__entry->wakeup = ktime_to_us(ktime_sub(ktime, rqst->rq_qtime));
+		__entry->qtime = ktime_to_us(ktime_sub(ktime, rqst->rq_xprt->xpt_qtime));
 	),
 
-	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu",
-		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup)
+	TP_printk(SVC_XPRT_ENDPOINT_FORMAT " wakeup-us=%lu qtime=%lu",
+		SVC_XPRT_ENDPOINT_VARARGS, __entry->wakeup, __entry->qtime)
 );
 
 DECLARE_EVENT_CLASS(svc_xprt_event,
net/sunrpc/svc_xprt.c +1 -0
index ae25405d8bd22672a361d1fd3adfdcebb403f90f..32018557797b1f683d8b7259f5fccd029aebcd71 100644
@@ -488,6 +488,7 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 	pool = svc_pool_for_cpu(xprt->xpt_server);
 
 	percpu_counter_inc(&pool->sp_sockets_queued);
+	xprt->xpt_qtime = ktime_get();
 	lwq_enqueue(&xprt->xpt_ready, &pool->sp_xprts);
 
 	svc_pool_wake_idle_thread(pool);
Signature · 1 line
Jeff Layton <[email protected]>
author maintainer
#8 lore raw

PATCH v2 3/12 sunrpc: move the SVC_RQST_EVENT_*() macros to common header

In reply to Jeff Layton

We need this for some of the nfsd tracepoints, so move them to the common misc/sunrpc.h infrastructure.

---
 include/trace/events/sunrpc.h | 23 -----------------------
 include/trace/misc/sunrpc.h   | 23 +++++++++++++++++++++++
 2 files changed, 23 insertions(+), 23 deletions(-)
include/trace/events/sunrpc.h +0 -23
index b5a0f0bc1a3b7cfd90ce0181a8a419db810988bb..bd6a1e3631c291c55bde37cb73d2086d9b15c5de 100644
@@ -1716,29 +1716,6 @@ TRACE_DEFINE_ENUM(SVC_COMPLETE);
 		{ SVC_PENDING,	"SVC_PENDING" },	\
 		{ SVC_COMPLETE,	"SVC_COMPLETE" })
 
-#define SVC_RQST_ENDPOINT_FIELDS(r) \
-		__sockaddr(server, (r)->rq_xprt->xpt_locallen) \
-		__sockaddr(client, (r)->rq_xprt->xpt_remotelen) \
-		__field(unsigned int, netns_ino) \
-		__field(u32, xid)
-
-#define SVC_RQST_ENDPOINT_ASSIGNMENTS(r) \
-		do { \
-			struct svc_xprt *xprt = (r)->rq_xprt; \
-			__assign_sockaddr(server, &xprt->xpt_local, \
-					  xprt->xpt_locallen); \
-			__assign_sockaddr(client, &xprt->xpt_remote, \
-					  xprt->xpt_remotelen); \
-			__entry->netns_ino = xprt->xpt_net->ns.inum; \
-			__entry->xid = be32_to_cpu((r)->rq_xid); \
-		} while (0)
-
-#define SVC_RQST_ENDPOINT_FORMAT \
-		"xid=0x%08x server=%pISpc client=%pISpc"
-
-#define SVC_RQST_ENDPOINT_VARARGS \
-		__entry->xid, __get_sockaddr(server), __get_sockaddr(client)
-
 TRACE_EVENT_CONDITION(svc_authenticate,
 	TP_PROTO(
 		const struct svc_rqst *rqst,
include/trace/misc/sunrpc.h +23 -0
index 588557d07ea820116219cf8ac7b049976a7d89b1..c3c8feede26087bad9b4c8dd0019606ee9982259 100644
@@ -15,4 +15,27 @@
 #define SUNRPC_TRACE_TASK_SPECIFIER \
 	"task:" SUNRPC_TRACE_PID_SPECIFIER "@" SUNRPC_TRACE_CLID_SPECIFIER
 
+#define SVC_RQST_ENDPOINT_FIELDS(r) \
+		__sockaddr(server, (r)->rq_xprt->xpt_locallen) \
+		__sockaddr(client, (r)->rq_xprt->xpt_remotelen) \
+		__field(unsigned int, netns_ino) \
+		__field(u32, xid)
+
+#define SVC_RQST_ENDPOINT_ASSIGNMENTS(r) \
+		do { \
+			struct svc_xprt *xprt = (r)->rq_xprt; \
+			__assign_sockaddr(server, &xprt->xpt_local, \
+					  xprt->xpt_locallen); \
+			__assign_sockaddr(client, &xprt->xpt_remote, \
+					  xprt->xpt_remotelen); \
+			__entry->netns_ino = xprt->xpt_net->ns.inum; \
+			__entry->xid = be32_to_cpu((r)->rq_xid); \
+		} while (0)
+
+#define SVC_RQST_ENDPOINT_FORMAT \
+		"xid=0x%08x server=%pISpc client=%pISpc"
+
+#define SVC_RQST_ENDPOINT_VARARGS \
+		__entry->xid, __get_sockaddr(server), __get_sockaddr(client)
+
 #endif /* _TRACE_SUNRPC_BASE_H */
Signature · 1 line
2.49.0
author maintainer
#9 lore raw

PATCH v2 4/12 nfsd: add a tracepoint for nfsd_setattr

In reply to Jeff Layton

Turn Sargun's internal kprobe based implementation of this into a normal static tracepoint. Also, remove the dprintk's that got added recently with the fix for zero-length ACLs.

---
 fs/nfsd/trace.h         | 35 +++++++++++++++++++++++++++++++++++
 fs/nfsd/vfs.c           |  5 ++---
 include/trace/misc/fs.h | 21 +++++++++++++++++++++
 3 files changed, 58 insertions(+), 3 deletions(-)
fs/nfsd/trace.h +35 -0
index 0d49fc064f7273f32c93732a993fd77bc0783f5d..c496fed58e2eed15458f35a158fbfef39a972c55 100644
@@ -11,6 +11,7 @@
 #include <linux/tracepoint.h>
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/xprt.h>
+#include <trace/misc/fs.h>
 #include <trace/misc/nfs.h>
 #include <trace/misc/sunrpc.h>
 
@@ -2337,6 +2338,40 @@ DEFINE_EVENT(nfsd_copy_async_done_class,		\
 DEFINE_COPY_ASYNC_DONE_EVENT(done);
 DEFINE_COPY_ASYNC_DONE_EVENT(cancel);
 
+TRACE_EVENT(nfsd_setattr,
+	TP_PROTO(const struct svc_rqst *rqstp, const struct svc_fh *fhp,
+		 const struct iattr *iap, const struct timespec64 *guardtime),
+	TP_ARGS(rqstp, fhp, iap, guardtime),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__field(s64, gtime_tv_sec)
+		__field(u32, gtime_tv_nsec)
+		__field(unsigned int, ia_valid)
+		__field(loff_t, ia_size)
+		__field(uid_t, ia_uid)
+		__field(gid_t, ia_gid)
+		__field(umode_t, ia_mode)
+	),
+	TP_fast_assign(__entry->xid = be32_to_cpu(rqstp->rq_xid);
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__entry->gtime_tv_sec = guardtime ? guardtime->tv_sec : 0;
+		__entry->gtime_tv_nsec = guardtime ? guardtime->tv_nsec : 0;
+		__entry->ia_valid = iap->ia_valid;
+		__entry->ia_size = iap->ia_size;
+		__entry->ia_uid = __kuid_val(iap->ia_uid);
+		__entry->ia_gid = __kgid_val(iap->ia_gid);
+		__entry->ia_mode = iap->ia_mode;
+	),
+	TP_printk(
+		"xid=0x%08x fh_hash=0x%08x ia_valid=%s ia_size=%llu ia_mode=0%o ia_uid=%u ia_gid=%u guard_time=%lld.%u",
+		__entry->xid, __entry->fh_hash, show_ia_valid_flags(__entry->ia_valid),
+		__entry->ia_size, __entry->ia_mode, __entry->ia_uid, __entry->ia_gid,
+		__entry->gtime_tv_sec, __entry->gtime_tv_nsec
+	)
+)
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
fs/nfsd/vfs.c +2 -3
index d1156a18a79579bf427fe5809dc93d06e241201e..77ae22abc1a21ec587cf089b2a5f750464b5e985 100644
@@ -501,7 +501,8 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	bool		size_change = (iap->ia_valid & ATTR_SIZE);
 	int		retries;
 
-dprintk("nfsd_setattr pacl=%p valid=0x%x\n", attr->na_pacl, iap->ia_valid);
+	trace_nfsd_setattr(rqstp, fhp, iap, guardtime);
+
 	if (iap->ia_valid & ATTR_SIZE) {
 		accmode |= NFSD_MAY_WRITE|NFSD_MAY_OWNER_OVERRIDE;
 		ftype = S_IFREG;
@@ -597,7 +598,6 @@ dprintk("nfsd_setattr pacl=%p valid=0x%x\n", attr->na_pacl, iap->ia_valid);
 						NULL);
 	}
 	if (IS_ENABLED(CONFIG_FS_POSIX_ACL) && attr->na_pacl) {
-dprintk("at set_posix_acl\n");
 		/*
 		 * For any file system that is not ACL_SCOPE_FILE_OBJECT,
 		 * a_count == 0 MUST reply nfserr_inval.
@@ -612,7 +612,6 @@ dprintk("at set_posix_acl\n");
 							attr->na_pacl);
 		else
 			attr->na_paclerr = -EINVAL;
-dprintk("set_posix_acl=%d\n", attr->na_paclerr);
 	}
 out_fill_attrs:
 	/*
include/trace/misc/fs.h +21 -0
index 738b97f22f3651f2370830037a8f4bfdf9a42ad4..0406ebe2a80a499dfcadb7e63db4d9e4a84d4d64 100644
@@ -120,3 +120,24 @@
 		{ LOOKUP_BENEATH,	"BENEATH" }, \
 		{ LOOKUP_IN_ROOT,	"IN_ROOT" }, \
 		{ LOOKUP_CACHED,	"CACHED" })
+
+#define show_ia_valid_flags(flags)			\
+	__print_flags(flags, "|",			\
+		{ ATTR_MODE,		"MODE" },	\
+		{ ATTR_UID,		"UID" },	\
+		{ ATTR_GID,		"GID" },	\
+		{ ATTR_SIZE,		"SIZE" },	\
+		{ ATTR_ATIME,		"ATIME" },	\
+		{ ATTR_MTIME,		"MTIME" },	\
+		{ ATTR_CTIME,		"CTIME" },	\
+		{ ATTR_ATIME_SET,	"ATIME_SET" },	\
+		{ ATTR_MTIME_SET,	"MTIME_SET" },	\
+		{ ATTR_FORCE,		"FORCE" },	\
+		{ ATTR_KILL_SUID,	"KILL_SUID" },	\
+		{ ATTR_KILL_SGID,	"KILL_SGID" },	\
+		{ ATTR_FILE,		"FILE" },	\
+		{ ATTR_KILL_PRIV,	"KILL_PRIV" },	\
+		{ ATTR_OPEN,		"OPEN" },	\
+		{ ATTR_TIMES_SET,	"TIMES_SET" },	\
+		{ ATTR_TOUCH,		"TOUCH"},	\
+		{ ATTR_DELEG,		"DELEG"})
Signature · 1 line
2.49.0
#10 lore raw

Re: [PATCH v2 04/12] nfsd: add a tracepoint for nfsd_setattr

Quoted 70 lines · Jeff Layton

Turn Sargun's internal kprobe based implementation of this into a normal static tracepoint. Also, remove the dprintk's that got added recently with the fix for zero-length ACLs.

---
 fs/nfsd/trace.h         | 35 +++++++++++++++++++++++++++++++++++
 fs/nfsd/vfs.c           |  5 ++---
 include/trace/misc/fs.h | 21 +++++++++++++++++++++
 3 files changed, 58 insertions(+), 3 deletions(-)
fs/nfsd/trace.h +35 -0
index 0d49fc064f7273f32c93732a993fd77bc0783f5d..c496fed58e2eed15458f35a158fbfef39a972c55 100644
@@ -11,6 +11,7 @@
 #include <linux/tracepoint.h>
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/xprt.h>
+#include <trace/misc/fs.h>
 #include <trace/misc/nfs.h>
 #include <trace/misc/sunrpc.h>
 
@@ -2337,6 +2338,40 @@ DEFINE_EVENT(nfsd_copy_async_done_class,		\
 DEFINE_COPY_ASYNC_DONE_EVENT(done);
 DEFINE_COPY_ASYNC_DONE_EVENT(cancel);
 
+TRACE_EVENT(nfsd_setattr,
+	TP_PROTO(const struct svc_rqst *rqstp, const struct svc_fh *fhp,
+		 const struct iattr *iap, const struct timespec64 *guardtime),
+	TP_ARGS(rqstp, fhp, iap, guardtime),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__field(s64, gtime_tv_sec)
+		__field(u32, gtime_tv_nsec)
+		__field(unsigned int, ia_valid)
+		__field(loff_t, ia_size)
+		__field(uid_t, ia_uid)
+		__field(gid_t, ia_gid)
+		__field(umode_t, ia_mode)
+	),
+	TP_fast_assign(__entry->xid = be32_to_cpu(rqstp->rq_xid);
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__entry->gtime_tv_sec = guardtime ? guardtime->tv_sec : 0;
+		__entry->gtime_tv_nsec = guardtime ? guardtime->tv_nsec : 0;
+		__entry->ia_valid = iap->ia_valid;
+		__entry->ia_size = iap->ia_size;
+		__entry->ia_uid = __kuid_val(iap->ia_uid);
+		__entry->ia_gid = __kgid_val(iap->ia_gid);
+		__entry->ia_mode = iap->ia_mode;
+	),
+	TP_printk(
+		"xid=0x%08x fh_hash=0x%08x ia_valid=%s ia_size=%llu ia_mode=0%o ia_uid=%u ia_gid=%u guard_time=%lld.%u",
+		__entry->xid, __entry->fh_hash, show_ia_valid_flags(__entry->ia_valid),
+		__entry->ia_size, __entry->ia_mode, __entry->ia_uid, __entry->ia_gid,
+		__entry->gtime_tv_sec, __entry->gtime_tv_nsec
+	)
+)
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
fs/nfsd/vfs.c +0 -0
index d1156a18a79579bf427fe5809dc93d06e241201e..77ae22abc1a21ec587cf089b2a5f750464b5e985 100644

fs/nfsd/vfs.c nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, · line 504

Quoted 4 lines
fs/nfsd/vfs.c +0 -1
@@ -501,7 +501,8 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	bool		size_change = (iap->ia_valid & ATTR_SIZE);
 	int		retries;
 
-dprintk("nfsd_setattr pacl=%p valid=0x%x\n", attr->na_pacl, iap->ia_valid);

Nit:

These unindented dprintk call sites were introduced by Rick's POSIX ACL patch series, which is experimental. He's promised to remove them before he sends a final version of that series.

Your patch can leave them in place, as they will be gone soon by other means.

Quoted 51 lines
fs/nfsd/vfs.c +2 -2
+	trace_nfsd_setattr(rqstp, fhp, iap, guardtime);
+
 	if (iap->ia_valid & ATTR_SIZE) {
 		accmode |= NFSD_MAY_WRITE|NFSD_MAY_OWNER_OVERRIDE;
 		ftype = S_IFREG;
@@ -597,7 +598,6 @@ dprintk("nfsd_setattr pacl=%p valid=0x%x\n", attr->na_pacl, iap->ia_valid);
 						NULL);
 	}
 	if (IS_ENABLED(CONFIG_FS_POSIX_ACL) && attr->na_pacl) {
-dprintk("at set_posix_acl\n");
 		/*
 		 * For any file system that is not ACL_SCOPE_FILE_OBJECT,
 		 * a_count == 0 MUST reply nfserr_inval.
@@ -612,7 +612,6 @@ dprintk("at set_posix_acl\n");
 							attr->na_pacl);
 		else
 			attr->na_paclerr = -EINVAL;
-dprintk("set_posix_acl=%d\n", attr->na_paclerr);
 	}
 out_fill_attrs:
 	/*
include/trace/misc/fs.h +21 -0
index 738b97f22f3651f2370830037a8f4bfdf9a42ad4..0406ebe2a80a499dfcadb7e63db4d9e4a84d4d64 100644
@@ -120,3 +120,24 @@
 		{ LOOKUP_BENEATH,	"BENEATH" }, \
 		{ LOOKUP_IN_ROOT,	"IN_ROOT" }, \
 		{ LOOKUP_CACHED,	"CACHED" })
+
+#define show_ia_valid_flags(flags)			\
+	__print_flags(flags, "|",			\
+		{ ATTR_MODE,		"MODE" },	\
+		{ ATTR_UID,		"UID" },	\
+		{ ATTR_GID,		"GID" },	\
+		{ ATTR_SIZE,		"SIZE" },	\
+		{ ATTR_ATIME,		"ATIME" },	\
+		{ ATTR_MTIME,		"MTIME" },	\
+		{ ATTR_CTIME,		"CTIME" },	\
+		{ ATTR_ATIME_SET,	"ATIME_SET" },	\
+		{ ATTR_MTIME_SET,	"MTIME_SET" },	\
+		{ ATTR_FORCE,		"FORCE" },	\
+		{ ATTR_KILL_SUID,	"KILL_SUID" },	\
+		{ ATTR_KILL_SGID,	"KILL_SGID" },	\
+		{ ATTR_FILE,		"FILE" },	\
+		{ ATTR_KILL_PRIV,	"KILL_PRIV" },	\
+		{ ATTR_OPEN,		"OPEN" },	\
+		{ ATTR_TIMES_SET,	"TIMES_SET" },	\
+		{ ATTR_TOUCH,		"TOUCH"},	\
+		{ ATTR_DELEG,		"DELEG"})
Signature · 1 line
Chuck Lever
author maintainer
#11 lore raw

PATCH v2 5/12 nfsd: add a tracepoint to nfsd_lookup_dentry

In reply to Jeff Layton

...and drop the dprintk.

---
 fs/nfsd/trace.h | 19 +++++++++++++++++++
 fs/nfsd/vfs.c   |  2 +-
 2 files changed, 20 insertions(+), 1 deletion(-)
fs/nfsd/trace.h +19 -0
index c496fed58e2eed15458f35a158fbfef39a972c55..382849d7c321d6ded8213890c2e7075770aa716c 100644
@@ -2372,6 +2372,25 @@ TRACE_EVENT(nfsd_setattr,
 	)
 )
 
+TRACE_EVENT(nfsd_lookup_dentry,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 const char *name,
+		 unsigned int len),
+	TP_ARGS(rqstp, fhp, name, len),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__string_len(name, name, len)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__assign_str(name);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s",
+		  __entry->xid, __entry->fh_hash, __get_str(name))
+);
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
fs/nfsd/vfs.c +1 -1
index 77ae22abc1a21ec587cf089b2a5f750464b5e985..2b96806111d6a45b351707db6a93219cb91d45f5 100644
@@ -246,7 +246,7 @@ nfsd_lookup_dentry(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	struct dentry		*dentry;
 	int			host_err;
 
-	dprintk("nfsd: nfsd_lookup(fh %s, %.*s)\n", SVCFH_fmt(fhp), len,name);
+	trace_nfsd_lookup_dentry(rqstp, fhp, name, len);
 
 	dparent = fhp->fh_dentry;
 	exp = exp_get(fhp->fh_export);
Signature · 1 line
2.49.0
author maintainer
#12 lore raw

PATCH v2 6/12 nfsd: add tracepoints around nfsd_create events

In reply to Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c | 18 +++++-------------
 fs/nfsd/nfs4proc.c | 29 +++++++++++++++++++++++++++++
 fs/nfsd/nfsproc.c  |  6 +++---
 fs/nfsd/trace.h    | 39 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 76 insertions(+), 16 deletions(-)
fs/nfsd/nfs3proc.c +5 -13
index 372bdcf5e07a5c835da240ecebb02e3576eb2ca6..ea1280970ea11b2a82f0de88ad0422eef7063d6d 100644
@@ -14,6 +14,7 @@
 #include "xdr3.h"
 #include "vfs.h"
 #include "filecache.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -380,10 +381,7 @@ nfsd3_proc_create(struct svc_rqst *rqstp)
 	struct nfsd3_diropres *resp = rqstp->rq_resp;
 	svc_fh *dirfhp, *newfhp;
 
-	dprintk("nfsd: CREATE(3)   %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_create(rqstp, &argp->fh, S_IFREG, argp->name, argp->len);
 
 	dirfhp = fh_copy(&resp->dirfh, &argp->fh);
 	newfhp = fh_init(&resp->fh, NFS3_FHSIZE);
@@ -405,10 +403,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	argp->attrs.ia_valid &= ~ATTR_SIZE;
 	fh_copy(&resp->dirfh, &argp->fh);
@@ -471,13 +466,10 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp)
 	struct nfsd_attrs attrs = {
 		.na_iattr	= &argp->attrs,
 	};
-	int type;
+	int type = nfs3_ftypes[argp->ftype];
 	dev_t	rdev = 0;
 
-	dprintk("nfsd: MKNOD(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mknod(rqstp, &argp->fh, type, argp->name, argp->len);
 
 	fh_copy(&resp->dirfh, &argp->fh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
fs/nfsd/nfs4proc.c +29 -0
index 6e23d6103010197c0316b07c189fe12ec3033812..2c795103deaa4044596bd07d90db788169a32a0c 100644
@@ -250,6 +250,8 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	__be32 status;
 	int host_err;
 
+	trace_nfsd4_create_file(rqstp, fhp, S_IFREG, open->op_fname, open->op_fnamelen);
+
 	if (isdotent(open->op_fname, open->op_fnamelen))
 		return nfserr_exist;
 	if (!(iap->ia_valid & ATTR_MODE))
@@ -807,6 +809,29 @@ nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	return status;
 }
 
+static umode_t nfs_type_to_vfs_type(enum nfs_ftype4 nfstype)
+{
+	switch (nfstype) {
+	case NF4REG:
+		return S_IFREG;
+	case NF4DIR:
+		return S_IFDIR;
+	case NF4BLK:
+		return S_IFBLK;
+	case NF4CHR:
+		return S_IFCHR;
+	case NF4LNK:
+		return S_IFLNK;
+	case NF4SOCK:
+		return S_IFSOCK;
+	case NF4FIFO:
+		return S_IFIFO;
+	default:
+		break;
+	}
+	return 0;
+}
+
 static __be32
 nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	     union nfsd4_op_u *u)
@@ -822,6 +847,10 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	__be32 status;
 	dev_t rdev;
 
+	trace_nfsd4_create(rqstp, &cstate->current_fh,
+			   nfs_type_to_vfs_type(create->cr_type),
+			   create->cr_name, create->cr_namelen);
+
 	fh_init(&resfh, NFS4_FHSIZE);
 
 	status = fh_verify(rqstp, &cstate->current_fh, S_IFDIR, NFSD_MAY_NOP);
fs/nfsd/nfsproc.c +3 -3
index 6dda081eb24c00b834ab0965c3a35a12115bceb7..33d8cbf8785588d38d4ec5efd769c1d1d06c6a91 100644
@@ -10,6 +10,7 @@
 #include "cache.h"
 #include "xdr.h"
 #include "vfs.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -292,8 +293,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	int		hosterr;
 	dev_t		rdev = 0, wanted = new_decode_dev(attr->ia_size);
 
-	dprintk("nfsd: CREATE   %s %.*s\n",
-		SVCFH_fmt(dirfhp), argp->len, argp->name);
+	trace_nfsd_proc_create(rqstp, dirfhp, S_IFREG, argp->name, argp->len);
 
 	/* First verify the parent file handle */
 	resp->status = fh_verify(rqstp, dirfhp, S_IFDIR, NFSD_MAY_EXEC);
@@ -548,7 +548,7 @@ nfsd_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR    %s %.*s\n", SVCFH_fmt(&argp->fh), argp->len, argp->name);
+	trace_nfsd_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	if (resp->fh.fh_dentry) {
 		printk(KERN_WARNING
fs/nfsd/trace.h +39 -0
index 382849d7c321d6ded8213890c2e7075770aa716c..c6aff23a845f06c87e701d57ec577c2c5c5a743c 100644
@@ -2391,6 +2391,45 @@ TRACE_EVENT(nfsd_lookup_dentry,
 	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s",
 		  __entry->xid, __entry->fh_hash, __get_str(name))
 );
+
+DECLARE_EVENT_CLASS(nfsd_vfs_create_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 umode_t type,
+		 const char *name,
+		 unsigned int len),
+	TP_ARGS(rqstp, fhp, type, name, len),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__field(umode_t, type)
+		__string_len(name, name, len)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__entry->type = type;
+		__assign_str(name);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x type=%s name=%s",
+		  __entry->xid, __entry->fh_hash,
+		  show_fs_file_type(__entry->type), __get_str(name))
+);
+
+#define DEFINE_NFSD_VFS_CREATE_EVENT(__name)						\
+	DEFINE_EVENT(nfsd_vfs_create_class, __name,					\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,		\
+			      umode_t type, const char *name, unsigned int len),	\
+		     TP_ARGS(rqstp, fhp, type, name, len))
+
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mknod);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create_file);
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
2.49.0
#13 lore raw

Re: [PATCH v2 06/12] nfsd: add tracepoints around nfsd_create events

Quoted 92 lines · Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c | 18 +++++-------------
 fs/nfsd/nfs4proc.c | 29 +++++++++++++++++++++++++++++
 fs/nfsd/nfsproc.c  |  6 +++---
 fs/nfsd/trace.h    | 39 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 76 insertions(+), 16 deletions(-)
fs/nfsd/nfs3proc.c +5 -13
index 372bdcf5e07a5c835da240ecebb02e3576eb2ca6..ea1280970ea11b2a82f0de88ad0422eef7063d6d 100644
@@ -14,6 +14,7 @@
 #include "xdr3.h"
 #include "vfs.h"
 #include "filecache.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -380,10 +381,7 @@ nfsd3_proc_create(struct svc_rqst *rqstp)
 	struct nfsd3_diropres *resp = rqstp->rq_resp;
 	svc_fh *dirfhp, *newfhp;
 
-	dprintk("nfsd: CREATE(3)   %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_create(rqstp, &argp->fh, S_IFREG, argp->name, argp->len);
 
 	dirfhp = fh_copy(&resp->dirfh, &argp->fh);
 	newfhp = fh_init(&resp->fh, NFS3_FHSIZE);
@@ -405,10 +403,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	argp->attrs.ia_valid &= ~ATTR_SIZE;
 	fh_copy(&resp->dirfh, &argp->fh);
@@ -471,13 +466,10 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp)
 	struct nfsd_attrs attrs = {
 		.na_iattr	= &argp->attrs,
 	};
-	int type;
+	int type = nfs3_ftypes[argp->ftype];
 	dev_t	rdev = 0;
 
-	dprintk("nfsd: MKNOD(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mknod(rqstp, &argp->fh, type, argp->name, argp->len);
 
 	fh_copy(&resp->dirfh, &argp->fh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
fs/nfsd/nfs4proc.c +15 -0
index 6e23d6103010197c0316b07c189fe12ec3033812..2c795103deaa4044596bd07d90db788169a32a0c 100644
@@ -250,6 +250,8 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	__be32 status;
 	int host_err;
 
+	trace_nfsd4_create_file(rqstp, fhp, S_IFREG, open->op_fname, open->op_fnamelen);
+
 	if (isdotent(open->op_fname, open->op_fnamelen))
 		return nfserr_exist;
 	if (!(iap->ia_valid & ATTR_MODE))
@@ -807,6 +809,29 @@ nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	return status;
 }
 
+static umode_t nfs_type_to_vfs_type(enum nfs_ftype4 nfstype)
+{
+	switch (nfstype) {
+	case NF4REG:
+		return S_IFREG;
+	case NF4DIR:
+		return S_IFDIR;
+	case NF4BLK:
+		return S_IFBLK;
+	case NF4CHR:
+		return S_IFCHR;
+	case NF4LNK:
+		return S_IFLNK;

fs/nfsd/nfs4proc.c nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, · line 834

Quoted 10 lines
fs/nfsd/nfs4proc.c +10 -0
+	case NF4SOCK:
+		return S_IFSOCK;
+	case NF4FIFO:
+		return S_IFIFO;
+	default:
+		break;
+	}
+	return 0;
+}
+

Wondering what happens when trace points are disabled in the kernel build. Maybe this helper belongs in fs/nfsd/trace.h instead as a macro wrapper for __print_symbolic(). But see below.

Quoted 81 lines
fs/nfsd/nfs4proc.c +4 -0
 static __be32
 nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	     union nfsd4_op_u *u)
@@ -822,6 +847,10 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	__be32 status;
 	dev_t rdev;
 
+	trace_nfsd4_create(rqstp, &cstate->current_fh,
+			   nfs_type_to_vfs_type(create->cr_type),
+			   create->cr_name, create->cr_namelen);
+
 	fh_init(&resfh, NFS4_FHSIZE);
 
 	status = fh_verify(rqstp, &cstate->current_fh, S_IFDIR, NFSD_MAY_NOP);
fs/nfsd/nfsproc.c +3 -3
index 6dda081eb24c00b834ab0965c3a35a12115bceb7..33d8cbf8785588d38d4ec5efd769c1d1d06c6a91 100644
@@ -10,6 +10,7 @@
 #include "cache.h"
 #include "xdr.h"
 #include "vfs.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -292,8 +293,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	int		hosterr;
 	dev_t		rdev = 0, wanted = new_decode_dev(attr->ia_size);
 
-	dprintk("nfsd: CREATE   %s %.*s\n",
-		SVCFH_fmt(dirfhp), argp->len, argp->name);
+	trace_nfsd_proc_create(rqstp, dirfhp, S_IFREG, argp->name, argp->len);
 
 	/* First verify the parent file handle */
 	resp->status = fh_verify(rqstp, dirfhp, S_IFDIR, NFSD_MAY_EXEC);
@@ -548,7 +548,7 @@ nfsd_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR    %s %.*s\n", SVCFH_fmt(&argp->fh), argp->len, argp->name);
+	trace_nfsd_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	if (resp->fh.fh_dentry) {
 		printk(KERN_WARNING
fs/nfsd/trace.h +28 -0
index 382849d7c321d6ded8213890c2e7075770aa716c..c6aff23a845f06c87e701d57ec577c2c5c5a743c 100644
@@ -2391,6 +2391,45 @@ TRACE_EVENT(nfsd_lookup_dentry,
 	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s",
 		  __entry->xid, __entry->fh_hash, __get_str(name))
 );
+
+DECLARE_EVENT_CLASS(nfsd_vfs_create_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 umode_t type,
+		 const char *name,
+		 unsigned int len),
+	TP_ARGS(rqstp, fhp, type, name, len),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__field(umode_t, type)
+		__string_len(name, name, len)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__entry->type = type;
+		__assign_str(name);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x type=%s name=%s",
+		  __entry->xid, __entry->fh_hash,
+		  show_fs_file_type(__entry->type), __get_str(name))
+);
+
+#define DEFINE_NFSD_VFS_CREATE_EVENT(__name)						\
+	DEFINE_EVENT(nfsd_vfs_create_class, __name,					\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,		\

fs/nfsd/trace.h TRACE_EVENT(nfsd_lookup_dentry, · line 2431

Quoted 10 lines
fs/nfsd/trace.h +10 -0
+			      umode_t type, const char *name, unsigned int len),	\
+		     TP_ARGS(rqstp, fhp, type, name, len))
+
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mknod);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create_file);

I think we would be better off with one or two new trace points in nfsd_create() and nfsd_create_setattr() instead of all of these...

Unless I've missed what you are trying to observe...?

Quoted 5 lines
fs/nfsd/trace.h +1 -0
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
Chuck Lever
author maintainer
#14 lore raw

Re: [PATCH v2 06/12] nfsd: add tracepoints around nfsd_create events

Quoted 99 lines · Chuck Lever
Quoted 98 lines · Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c | 18 +++++-------------
 fs/nfsd/nfs4proc.c | 29 +++++++++++++++++++++++++++++
 fs/nfsd/nfsproc.c  |  6 +++---
 fs/nfsd/trace.h    | 39 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 76 insertions(+), 16 deletions(-)
fs/nfsd/nfs3proc.c +5 -13
index 372bdcf5e07a5c835da240ecebb02e3576eb2ca6..ea1280970ea11b2a82f0de88ad0422eef7063d6d 100644
@@ -14,6 +14,7 @@
 #include "xdr3.h"
 #include "vfs.h"
 #include "filecache.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -380,10 +381,7 @@ nfsd3_proc_create(struct svc_rqst *rqstp)
 	struct nfsd3_diropres *resp = rqstp->rq_resp;
 	svc_fh *dirfhp, *newfhp;
 
-	dprintk("nfsd: CREATE(3)   %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_create(rqstp, &argp->fh, S_IFREG, argp->name, argp->len);
 
 	dirfhp = fh_copy(&resp->dirfh, &argp->fh);
 	newfhp = fh_init(&resp->fh, NFS3_FHSIZE);
@@ -405,10 +403,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	argp->attrs.ia_valid &= ~ATTR_SIZE;
 	fh_copy(&resp->dirfh, &argp->fh);
@@ -471,13 +466,10 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp)
 	struct nfsd_attrs attrs = {
 		.na_iattr	= &argp->attrs,
 	};
-	int type;
+	int type = nfs3_ftypes[argp->ftype];
 	dev_t	rdev = 0;
 
-	dprintk("nfsd: MKNOD(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mknod(rqstp, &argp->fh, type, argp->name, argp->len);
 
 	fh_copy(&resp->dirfh, &argp->fh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
fs/nfsd/nfs4proc.c +21 -0
index 6e23d6103010197c0316b07c189fe12ec3033812..2c795103deaa4044596bd07d90db788169a32a0c 100644
@@ -250,6 +250,8 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	__be32 status;
 	int host_err;
 
+	trace_nfsd4_create_file(rqstp, fhp, S_IFREG, open->op_fname, open->op_fnamelen);
+
 	if (isdotent(open->op_fname, open->op_fnamelen))
 		return nfserr_exist;
 	if (!(iap->ia_valid & ATTR_MODE))
@@ -807,6 +809,29 @@ nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	return status;
 }
 
+static umode_t nfs_type_to_vfs_type(enum nfs_ftype4 nfstype)
+{
+	switch (nfstype) {
+	case NF4REG:
+		return S_IFREG;
+	case NF4DIR:
+		return S_IFDIR;
+	case NF4BLK:
+		return S_IFBLK;
+	case NF4CHR:
+		return S_IFCHR;
+	case NF4LNK:
+		return S_IFLNK;
+	case NF4SOCK:
+		return S_IFSOCK;
+	case NF4FIFO:
+		return S_IFIFO;
+	default:
+		break;
Quoted 9 lines

fs/nfsd/nfs4proc.c nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, · line 834

Quoted 4 lines
fs/nfsd/nfs4proc.c +4 -0
+	}
+	return 0;
+}
+

Wondering what happens when trace points are disabled in the kernel build. Maybe this helper belongs in fs/nfsd/trace.h instead as a macro wrapper for __print_symbolic(). But see below.

If tracepoints are disabled, then the only caller of this static function would go away, so it should get optimized out. I don't see how you'd make this a wrapper around __print_symbolic(), since the point is to pass in a NFS version-independent constant that the tracepoint class can use as a type.

Quoted 96 lines
Quoted 82 lines
fs/nfsd/nfs4proc.c +4 -0
 static __be32
 nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	     union nfsd4_op_u *u)
@@ -822,6 +847,10 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	__be32 status;
 	dev_t rdev;
 
+	trace_nfsd4_create(rqstp, &cstate->current_fh,
+			   nfs_type_to_vfs_type(create->cr_type),
+			   create->cr_name, create->cr_namelen);
+
 	fh_init(&resfh, NFS4_FHSIZE);
 
 	status = fh_verify(rqstp, &cstate->current_fh, S_IFDIR, NFSD_MAY_NOP);
fs/nfsd/nfsproc.c +3 -3
index 6dda081eb24c00b834ab0965c3a35a12115bceb7..33d8cbf8785588d38d4ec5efd769c1d1d06c6a91 100644
@@ -10,6 +10,7 @@
 #include "cache.h"
 #include "xdr.h"
 #include "vfs.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -292,8 +293,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	int		hosterr;
 	dev_t		rdev = 0, wanted = new_decode_dev(attr->ia_size);
 
-	dprintk("nfsd: CREATE   %s %.*s\n",
-		SVCFH_fmt(dirfhp), argp->len, argp->name);
+	trace_nfsd_proc_create(rqstp, dirfhp, S_IFREG, argp->name, argp->len);
 
 	/* First verify the parent file handle */
 	resp->status = fh_verify(rqstp, dirfhp, S_IFDIR, NFSD_MAY_EXEC);
@@ -548,7 +548,7 @@ nfsd_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR    %s %.*s\n", SVCFH_fmt(&argp->fh), argp->len, argp->name);
+	trace_nfsd_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	if (resp->fh.fh_dentry) {
 		printk(KERN_WARNING
fs/nfsd/trace.h +28 -0
index 382849d7c321d6ded8213890c2e7075770aa716c..c6aff23a845f06c87e701d57ec577c2c5c5a743c 100644
@@ -2391,6 +2391,45 @@ TRACE_EVENT(nfsd_lookup_dentry,
 	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s",
 		  __entry->xid, __entry->fh_hash, __get_str(name))
 );
+
+DECLARE_EVENT_CLASS(nfsd_vfs_create_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 umode_t type,
+		 const char *name,
+		 unsigned int len),
+	TP_ARGS(rqstp, fhp, type, name, len),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__field(umode_t, type)
+		__string_len(name, name, len)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__entry->type = type;
+		__assign_str(name);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x type=%s name=%s",
+		  __entry->xid, __entry->fh_hash,
+		  show_fs_file_type(__entry->type), __get_str(name))
+);
+
+#define DEFINE_NFSD_VFS_CREATE_EVENT(__name)						\
+	DEFINE_EVENT(nfsd_vfs_create_class, __name,					\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,		\

fs/nfsd/trace.h TRACE_EVENT(nfsd_lookup_dentry, · line 2431

Quoted 10 lines
fs/nfsd/trace.h +10 -0
+			      umode_t type, const char *name, unsigned int len),	\
+		     TP_ARGS(rqstp, fhp, type, name, len))
+
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mknod);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create_file);

I think we would be better off with one or two new trace points in nfsd_create() and nfsd_create_setattr() instead of all of these...

Quoted 2 lines

Unless I've missed what you are trying to observe...?

I'll look into doing it that way.

Quoted 8 lines
Quoted 6 lines
fs/nfsd/trace.h +1 -0
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
Jeff Layton <[email protected]>
#15 lore raw

Re: [PATCH v2 06/12] nfsd: add tracepoints around nfsd_create events

Quoted 110 lines · Jeff Layton
Quoted 99 lines · Chuck Lever
Quoted 98 lines · Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c | 18 +++++-------------
 fs/nfsd/nfs4proc.c | 29 +++++++++++++++++++++++++++++
 fs/nfsd/nfsproc.c  |  6 +++---
 fs/nfsd/trace.h    | 39 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 76 insertions(+), 16 deletions(-)
fs/nfsd/nfs3proc.c +5 -13
index 372bdcf5e07a5c835da240ecebb02e3576eb2ca6..ea1280970ea11b2a82f0de88ad0422eef7063d6d 100644
@@ -14,6 +14,7 @@
 #include "xdr3.h"
 #include "vfs.h"
 #include "filecache.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -380,10 +381,7 @@ nfsd3_proc_create(struct svc_rqst *rqstp)
 	struct nfsd3_diropres *resp = rqstp->rq_resp;
 	svc_fh *dirfhp, *newfhp;
 
-	dprintk("nfsd: CREATE(3)   %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_create(rqstp, &argp->fh, S_IFREG, argp->name, argp->len);
 
 	dirfhp = fh_copy(&resp->dirfh, &argp->fh);
 	newfhp = fh_init(&resp->fh, NFS3_FHSIZE);
@@ -405,10 +403,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	argp->attrs.ia_valid &= ~ATTR_SIZE;
 	fh_copy(&resp->dirfh, &argp->fh);
@@ -471,13 +466,10 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp)
 	struct nfsd_attrs attrs = {
 		.na_iattr	= &argp->attrs,
 	};
-	int type;
+	int type = nfs3_ftypes[argp->ftype];
 	dev_t	rdev = 0;
 
-	dprintk("nfsd: MKNOD(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mknod(rqstp, &argp->fh, type, argp->name, argp->len);
 
 	fh_copy(&resp->dirfh, &argp->fh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
fs/nfsd/nfs4proc.c +21 -0
index 6e23d6103010197c0316b07c189fe12ec3033812..2c795103deaa4044596bd07d90db788169a32a0c 100644
@@ -250,6 +250,8 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	__be32 status;
 	int host_err;
 
+	trace_nfsd4_create_file(rqstp, fhp, S_IFREG, open->op_fname, open->op_fnamelen);
+
 	if (isdotent(open->op_fname, open->op_fnamelen))
 		return nfserr_exist;
 	if (!(iap->ia_valid & ATTR_MODE))
@@ -807,6 +809,29 @@ nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	return status;
 }
 
+static umode_t nfs_type_to_vfs_type(enum nfs_ftype4 nfstype)
+{
+	switch (nfstype) {
+	case NF4REG:
+		return S_IFREG;
+	case NF4DIR:
+		return S_IFDIR;
+	case NF4BLK:
+		return S_IFBLK;
+	case NF4CHR:
+		return S_IFCHR;
+	case NF4LNK:
+		return S_IFLNK;
+	case NF4SOCK:
+		return S_IFSOCK;
+	case NF4FIFO:
+		return S_IFIFO;
+	default:
+		break;
Quoted 9 lines

fs/nfsd/nfs4proc.c nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, · line 834

Quoted 4 lines
fs/nfsd/nfs4proc.c +4 -0
+	}
+	return 0;
+}
+

Wondering what happens when trace points are disabled in the kernel build. Maybe this helper belongs in fs/nfsd/trace.h instead as a macro wrapper for __print_symbolic(). But see below.

Quoted 2 lines

If tracepoints are disabled, then the only caller of this static function would go away, so it should get optimized out.

AIUI, the compiler will complain in that case. If we need to keep this function here, then this is a legitimate use for "inline".

Quoted 115 lines

I don't see how
you'd make this a wrapper around __print_symbolic(), since the point is to pass in a NFS version-independent constant that the tracepoint class can use as a type.

Quoted 96 lines
Quoted 82 lines
fs/nfsd/nfs4proc.c +4 -0
 static __be32
 nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	     union nfsd4_op_u *u)
@@ -822,6 +847,10 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	__be32 status;
 	dev_t rdev;
 
+	trace_nfsd4_create(rqstp, &cstate->current_fh,
+			   nfs_type_to_vfs_type(create->cr_type),
+			   create->cr_name, create->cr_namelen);
+
 	fh_init(&resfh, NFS4_FHSIZE);
 
 	status = fh_verify(rqstp, &cstate->current_fh, S_IFDIR, NFSD_MAY_NOP);
fs/nfsd/nfsproc.c +3 -3
index 6dda081eb24c00b834ab0965c3a35a12115bceb7..33d8cbf8785588d38d4ec5efd769c1d1d06c6a91 100644
@@ -10,6 +10,7 @@
 #include "cache.h"
 #include "xdr.h"
 #include "vfs.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -292,8 +293,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	int		hosterr;
 	dev_t		rdev = 0, wanted = new_decode_dev(attr->ia_size);
 
-	dprintk("nfsd: CREATE   %s %.*s\n",
-		SVCFH_fmt(dirfhp), argp->len, argp->name);
+	trace_nfsd_proc_create(rqstp, dirfhp, S_IFREG, argp->name, argp->len);
 
 	/* First verify the parent file handle */
 	resp->status = fh_verify(rqstp, dirfhp, S_IFDIR, NFSD_MAY_EXEC);
@@ -548,7 +548,7 @@ nfsd_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR    %s %.*s\n", SVCFH_fmt(&argp->fh), argp->len, argp->name);
+	trace_nfsd_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	if (resp->fh.fh_dentry) {
 		printk(KERN_WARNING
fs/nfsd/trace.h +28 -0
index 382849d7c321d6ded8213890c2e7075770aa716c..c6aff23a845f06c87e701d57ec577c2c5c5a743c 100644
@@ -2391,6 +2391,45 @@ TRACE_EVENT(nfsd_lookup_dentry,
 	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s",
 		  __entry->xid, __entry->fh_hash, __get_str(name))
 );
+
+DECLARE_EVENT_CLASS(nfsd_vfs_create_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 umode_t type,
+		 const char *name,
+		 unsigned int len),
+	TP_ARGS(rqstp, fhp, type, name, len),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__field(umode_t, type)
+		__string_len(name, name, len)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__entry->type = type;
+		__assign_str(name);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x type=%s name=%s",
+		  __entry->xid, __entry->fh_hash,
+		  show_fs_file_type(__entry->type), __get_str(name))
+);
+
+#define DEFINE_NFSD_VFS_CREATE_EVENT(__name)						\
+	DEFINE_EVENT(nfsd_vfs_create_class, __name,					\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,		\

fs/nfsd/trace.h TRACE_EVENT(nfsd_lookup_dentry, · line 2431

Quoted 10 lines
fs/nfsd/trace.h +10 -0
+			      umode_t type, const char *name, unsigned int len),	\
+		     TP_ARGS(rqstp, fhp, type, name, len))
+
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mknod);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create_file);

I think we would be better off with one or two new trace points in nfsd_create() and nfsd_create_setattr() instead of all of these...

Quoted 2 lines

Unless I've missed what you are trying to observe...?

I'll look into doing it that way.

Quoted 8 lines
Quoted 6 lines
fs/nfsd/trace.h +1 -0
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
Chuck Lever
#16 lore raw

Re: [PATCH v2 06/12] nfsd: add tracepoints around nfsd_create events

Quoted 111 lines · Chuck Lever
Quoted 110 lines · Jeff Layton
Quoted 108 lines · Chuck Lever
Quoted 92 lines · Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c | 18 +++++-------------
 fs/nfsd/nfs4proc.c | 29 +++++++++++++++++++++++++++++
 fs/nfsd/nfsproc.c  |  6 +++---
 fs/nfsd/trace.h    | 39 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 76 insertions(+), 16 deletions(-)
fs/nfsd/nfs3proc.c +5 -13
index 372bdcf5e07a5c835da240ecebb02e3576eb2ca6..ea1280970ea11b2a82f0de88ad0422eef7063d6d 100644
@@ -14,6 +14,7 @@
 #include "xdr3.h"
 #include "vfs.h"
 #include "filecache.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -380,10 +381,7 @@ nfsd3_proc_create(struct svc_rqst *rqstp)
 	struct nfsd3_diropres *resp = rqstp->rq_resp;
 	svc_fh *dirfhp, *newfhp;
 
-	dprintk("nfsd: CREATE(3)   %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_create(rqstp, &argp->fh, S_IFREG, argp->name, argp->len);
 
 	dirfhp = fh_copy(&resp->dirfh, &argp->fh);
 	newfhp = fh_init(&resp->fh, NFS3_FHSIZE);
@@ -405,10 +403,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	argp->attrs.ia_valid &= ~ATTR_SIZE;
 	fh_copy(&resp->dirfh, &argp->fh);
@@ -471,13 +466,10 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp)
 	struct nfsd_attrs attrs = {
 		.na_iattr	= &argp->attrs,
 	};
-	int type;
+	int type = nfs3_ftypes[argp->ftype];
 	dev_t	rdev = 0;
 
-	dprintk("nfsd: MKNOD(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mknod(rqstp, &argp->fh, type, argp->name, argp->len);
 
 	fh_copy(&resp->dirfh, &argp->fh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
fs/nfsd/nfs4proc.c +15 -0
index 6e23d6103010197c0316b07c189fe12ec3033812..2c795103deaa4044596bd07d90db788169a32a0c 100644
@@ -250,6 +250,8 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	__be32 status;
 	int host_err;
 
+	trace_nfsd4_create_file(rqstp, fhp, S_IFREG, open->op_fname, open->op_fnamelen);
+
 	if (isdotent(open->op_fname, open->op_fnamelen))
 		return nfserr_exist;
 	if (!(iap->ia_valid & ATTR_MODE))
@@ -807,6 +809,29 @@ nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	return status;
 }
 
+static umode_t nfs_type_to_vfs_type(enum nfs_ftype4 nfstype)
+{
+	switch (nfstype) {
+	case NF4REG:
+		return S_IFREG;
+	case NF4DIR:
+		return S_IFDIR;
+	case NF4BLK:
+		return S_IFBLK;
+	case NF4CHR:
+		return S_IFCHR;
+	case NF4LNK:
+		return S_IFLNK;

fs/nfsd/nfs4proc.c nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, · line 834

Quoted 10 lines
fs/nfsd/nfs4proc.c +10 -0
+	case NF4SOCK:
+		return S_IFSOCK;
+	case NF4FIFO:
+		return S_IFIFO;
+	default:
+		break;
+	}
+	return 0;
+}
+

Wondering what happens when trace points are disabled in the kernel build. Maybe this helper belongs in fs/nfsd/trace.h instead as a macro wrapper for __print_symbolic(). But see below.

Quoted 5 lines
Quoted 2 lines

If tracepoints are disabled, then the only caller of this static function would go away, so it should get optimized out.

AIUI, the compiler will complain in that case. If we need to keep this function here, then this is a legitimate use for "inline".

or maybe_unused.

Quoted 117 lines
Quoted 115 lines

I don't see how
you'd make this a wrapper around __print_symbolic(), since the point is to pass in a NFS version-independent constant that the tracepoint class can use as a type.

Quoted 96 lines
Quoted 82 lines
fs/nfsd/nfs4proc.c +4 -0
 static __be32
 nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	     union nfsd4_op_u *u)
@@ -822,6 +847,10 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	__be32 status;
 	dev_t rdev;
 
+	trace_nfsd4_create(rqstp, &cstate->current_fh,
+			   nfs_type_to_vfs_type(create->cr_type),
+			   create->cr_name, create->cr_namelen);
+
 	fh_init(&resfh, NFS4_FHSIZE);
 
 	status = fh_verify(rqstp, &cstate->current_fh, S_IFDIR, NFSD_MAY_NOP);
fs/nfsd/nfsproc.c +3 -3
index 6dda081eb24c00b834ab0965c3a35a12115bceb7..33d8cbf8785588d38d4ec5efd769c1d1d06c6a91 100644
@@ -10,6 +10,7 @@
 #include "cache.h"
 #include "xdr.h"
 #include "vfs.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -292,8 +293,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	int		hosterr;
 	dev_t		rdev = 0, wanted = new_decode_dev(attr->ia_size);
 
-	dprintk("nfsd: CREATE   %s %.*s\n",
-		SVCFH_fmt(dirfhp), argp->len, argp->name);
+	trace_nfsd_proc_create(rqstp, dirfhp, S_IFREG, argp->name, argp->len);
 
 	/* First verify the parent file handle */
 	resp->status = fh_verify(rqstp, dirfhp, S_IFDIR, NFSD_MAY_EXEC);
@@ -548,7 +548,7 @@ nfsd_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR    %s %.*s\n", SVCFH_fmt(&argp->fh), argp->len, argp->name);
+	trace_nfsd_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	if (resp->fh.fh_dentry) {
 		printk(KERN_WARNING
fs/nfsd/trace.h +28 -0
index 382849d7c321d6ded8213890c2e7075770aa716c..c6aff23a845f06c87e701d57ec577c2c5c5a743c 100644
@@ -2391,6 +2391,45 @@ TRACE_EVENT(nfsd_lookup_dentry,
 	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s",
 		  __entry->xid, __entry->fh_hash, __get_str(name))
 );
+
+DECLARE_EVENT_CLASS(nfsd_vfs_create_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 umode_t type,
+		 const char *name,
+		 unsigned int len),
+	TP_ARGS(rqstp, fhp, type, name, len),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__field(umode_t, type)
+		__string_len(name, name, len)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__entry->type = type;
+		__assign_str(name);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x type=%s name=%s",
+		  __entry->xid, __entry->fh_hash,
+		  show_fs_file_type(__entry->type), __get_str(name))
+);
+
+#define DEFINE_NFSD_VFS_CREATE_EVENT(__name)						\
+	DEFINE_EVENT(nfsd_vfs_create_class, __name,					\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,		\

fs/nfsd/trace.h TRACE_EVENT(nfsd_lookup_dentry, · line 2431

Quoted 10 lines
fs/nfsd/trace.h +10 -0
+			      umode_t type, const char *name, unsigned int len),	\
+		     TP_ARGS(rqstp, fhp, type, name, len))
+
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mknod);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create_file);

I think we would be better off with one or two new trace points in nfsd_create() and nfsd_create_setattr() instead of all of these...

Quoted 2 lines

Unless I've missed what you are trying to observe...?

I'll look into doing it that way.

Quoted 8 lines
Quoted 6 lines
fs/nfsd/trace.h +1 -0
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
Chuck Lever
author maintainer
#17 lore raw

Re: [PATCH v2 06/12] nfsd: add tracepoints around nfsd_create events

In reply to Chuck Lever
Quoted 204 lines · Chuck Lever
Quoted 92 lines · Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c | 18 +++++-------------
 fs/nfsd/nfs4proc.c | 29 +++++++++++++++++++++++++++++
 fs/nfsd/nfsproc.c  |  6 +++---
 fs/nfsd/trace.h    | 39 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 76 insertions(+), 16 deletions(-)
fs/nfsd/nfs3proc.c +5 -13
index 372bdcf5e07a5c835da240ecebb02e3576eb2ca6..ea1280970ea11b2a82f0de88ad0422eef7063d6d 100644
@@ -14,6 +14,7 @@
 #include "xdr3.h"
 #include "vfs.h"
 #include "filecache.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -380,10 +381,7 @@ nfsd3_proc_create(struct svc_rqst *rqstp)
 	struct nfsd3_diropres *resp = rqstp->rq_resp;
 	svc_fh *dirfhp, *newfhp;
 
-	dprintk("nfsd: CREATE(3)   %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_create(rqstp, &argp->fh, S_IFREG, argp->name, argp->len);
 
 	dirfhp = fh_copy(&resp->dirfh, &argp->fh);
 	newfhp = fh_init(&resp->fh, NFS3_FHSIZE);
@@ -405,10 +403,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	argp->attrs.ia_valid &= ~ATTR_SIZE;
 	fh_copy(&resp->dirfh, &argp->fh);
@@ -471,13 +466,10 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp)
 	struct nfsd_attrs attrs = {
 		.na_iattr	= &argp->attrs,
 	};
-	int type;
+	int type = nfs3_ftypes[argp->ftype];
 	dev_t	rdev = 0;
 
-	dprintk("nfsd: MKNOD(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mknod(rqstp, &argp->fh, type, argp->name, argp->len);
 
 	fh_copy(&resp->dirfh, &argp->fh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
fs/nfsd/nfs4proc.c +15 -0
index 6e23d6103010197c0316b07c189fe12ec3033812..2c795103deaa4044596bd07d90db788169a32a0c 100644
@@ -250,6 +250,8 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	__be32 status;
 	int host_err;
 
+	trace_nfsd4_create_file(rqstp, fhp, S_IFREG, open->op_fname, open->op_fnamelen);
+
 	if (isdotent(open->op_fname, open->op_fnamelen))
 		return nfserr_exist;
 	if (!(iap->ia_valid & ATTR_MODE))
@@ -807,6 +809,29 @@ nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	return status;
 }
 
+static umode_t nfs_type_to_vfs_type(enum nfs_ftype4 nfstype)
+{
+	switch (nfstype) {
+	case NF4REG:
+		return S_IFREG;
+	case NF4DIR:
+		return S_IFDIR;
+	case NF4BLK:
+		return S_IFBLK;
+	case NF4CHR:
+		return S_IFCHR;
+	case NF4LNK:
+		return S_IFLNK;

fs/nfsd/nfs4proc.c nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, · line 834

Quoted 10 lines
fs/nfsd/nfs4proc.c +10 -0
+	case NF4SOCK:
+		return S_IFSOCK;
+	case NF4FIFO:
+		return S_IFIFO;
+	default:
+		break;
+	}
+	return 0;
+}
+

Wondering what happens when trace points are disabled in the kernel build. Maybe this helper belongs in fs/nfsd/trace.h instead as a macro wrapper for __print_symbolic(). But see below.

Quoted 81 lines
fs/nfsd/nfs4proc.c +4 -0
 static __be32
 nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	     union nfsd4_op_u *u)
@@ -822,6 +847,10 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	__be32 status;
 	dev_t rdev;
 
+	trace_nfsd4_create(rqstp, &cstate->current_fh,
+			   nfs_type_to_vfs_type(create->cr_type),
+			   create->cr_name, create->cr_namelen);
+
 	fh_init(&resfh, NFS4_FHSIZE);
 
 	status = fh_verify(rqstp, &cstate->current_fh, S_IFDIR, NFSD_MAY_NOP);
fs/nfsd/nfsproc.c +3 -3
index 6dda081eb24c00b834ab0965c3a35a12115bceb7..33d8cbf8785588d38d4ec5efd769c1d1d06c6a91 100644
@@ -10,6 +10,7 @@
 #include "cache.h"
 #include "xdr.h"
 #include "vfs.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -292,8 +293,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	int		hosterr;
 	dev_t		rdev = 0, wanted = new_decode_dev(attr->ia_size);
 
-	dprintk("nfsd: CREATE   %s %.*s\n",
-		SVCFH_fmt(dirfhp), argp->len, argp->name);
+	trace_nfsd_proc_create(rqstp, dirfhp, S_IFREG, argp->name, argp->len);
 
 	/* First verify the parent file handle */
 	resp->status = fh_verify(rqstp, dirfhp, S_IFDIR, NFSD_MAY_EXEC);
@@ -548,7 +548,7 @@ nfsd_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR    %s %.*s\n", SVCFH_fmt(&argp->fh), argp->len, argp->name);
+	trace_nfsd_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	if (resp->fh.fh_dentry) {
 		printk(KERN_WARNING
fs/nfsd/trace.h +28 -0
index 382849d7c321d6ded8213890c2e7075770aa716c..c6aff23a845f06c87e701d57ec577c2c5c5a743c 100644
@@ -2391,6 +2391,45 @@ TRACE_EVENT(nfsd_lookup_dentry,
 	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s",
 		  __entry->xid, __entry->fh_hash, __get_str(name))
 );
+
+DECLARE_EVENT_CLASS(nfsd_vfs_create_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 umode_t type,
+		 const char *name,
+		 unsigned int len),
+	TP_ARGS(rqstp, fhp, type, name, len),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__field(umode_t, type)
+		__string_len(name, name, len)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__entry->type = type;
+		__assign_str(name);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x type=%s name=%s",
+		  __entry->xid, __entry->fh_hash,
+		  show_fs_file_type(__entry->type), __get_str(name))
+);
+
+#define DEFINE_NFSD_VFS_CREATE_EVENT(__name)						\
+	DEFINE_EVENT(nfsd_vfs_create_class, __name,					\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,		\

fs/nfsd/trace.h TRACE_EVENT(nfsd_lookup_dentry, · line 2431

Quoted 10 lines
fs/nfsd/trace.h +10 -0
+			      umode_t type, const char *name, unsigned int len),	\
+		     TP_ARGS(rqstp, fhp, type, name, len))
+
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mknod);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create_file);

I think we would be better off with one or two new trace points in nfsd_create() and nfsd_create_setattr() instead of all of these...

Quoted 2 lines

Unless I've missed what you are trying to observe...?

Now I remember why I did it this way. Doing it the way you suggest makes this all a bit messy:

Most of the callers create files via nfsd_create(), but nfsd3_create_file calls vfs_create() directly. It does call nfsd_create_setattr(), but that's really not the right place for this either, since it's doing a setattr and not the create itself. Notably, it isn't passed the filename.

Note too that burying this tracepoint down in nfs_create() also means that error conditions can happen before the tracepoint that may mean that it won't fire. So, if you're watching the traces to see when CREATE or MKDIR calls occur, you may miss some of them.

How do you want to proceed?

Signature · 1 line
Jeff Layton <[email protected]>
#18 lore raw

Re: [PATCH v2 06/12] nfsd: add tracepoints around nfsd_create events

Quoted 217 lines · Jeff Layton
Quoted 204 lines · Chuck Lever
Quoted 92 lines · Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c | 18 +++++-------------
 fs/nfsd/nfs4proc.c | 29 +++++++++++++++++++++++++++++
 fs/nfsd/nfsproc.c  |  6 +++---
 fs/nfsd/trace.h    | 39 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 76 insertions(+), 16 deletions(-)
fs/nfsd/nfs3proc.c +5 -13
index 372bdcf5e07a5c835da240ecebb02e3576eb2ca6..ea1280970ea11b2a82f0de88ad0422eef7063d6d 100644
@@ -14,6 +14,7 @@
 #include "xdr3.h"
 #include "vfs.h"
 #include "filecache.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -380,10 +381,7 @@ nfsd3_proc_create(struct svc_rqst *rqstp)
 	struct nfsd3_diropres *resp = rqstp->rq_resp;
 	svc_fh *dirfhp, *newfhp;
 
-	dprintk("nfsd: CREATE(3)   %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_create(rqstp, &argp->fh, S_IFREG, argp->name, argp->len);
 
 	dirfhp = fh_copy(&resp->dirfh, &argp->fh);
 	newfhp = fh_init(&resp->fh, NFS3_FHSIZE);
@@ -405,10 +403,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	argp->attrs.ia_valid &= ~ATTR_SIZE;
 	fh_copy(&resp->dirfh, &argp->fh);
@@ -471,13 +466,10 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp)
 	struct nfsd_attrs attrs = {
 		.na_iattr	= &argp->attrs,
 	};
-	int type;
+	int type = nfs3_ftypes[argp->ftype];
 	dev_t	rdev = 0;
 
-	dprintk("nfsd: MKNOD(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_mknod(rqstp, &argp->fh, type, argp->name, argp->len);
 
 	fh_copy(&resp->dirfh, &argp->fh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
fs/nfsd/nfs4proc.c +15 -0
index 6e23d6103010197c0316b07c189fe12ec3033812..2c795103deaa4044596bd07d90db788169a32a0c 100644
@@ -250,6 +250,8 @@ nfsd4_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	__be32 status;
 	int host_err;
 
+	trace_nfsd4_create_file(rqstp, fhp, S_IFREG, open->op_fname, open->op_fnamelen);
+
 	if (isdotent(open->op_fname, open->op_fnamelen))
 		return nfserr_exist;
 	if (!(iap->ia_valid & ATTR_MODE))
@@ -807,6 +809,29 @@ nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	return status;
 }
 
+static umode_t nfs_type_to_vfs_type(enum nfs_ftype4 nfstype)
+{
+	switch (nfstype) {
+	case NF4REG:
+		return S_IFREG;
+	case NF4DIR:
+		return S_IFDIR;
+	case NF4BLK:
+		return S_IFBLK;
+	case NF4CHR:
+		return S_IFCHR;
+	case NF4LNK:
+		return S_IFLNK;

fs/nfsd/nfs4proc.c nfsd4_commit(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, · line 834

Quoted 10 lines
fs/nfsd/nfs4proc.c +10 -0
+	case NF4SOCK:
+		return S_IFSOCK;
+	case NF4FIFO:
+		return S_IFIFO;
+	default:
+		break;
+	}
+	return 0;
+}
+

Wondering what happens when trace points are disabled in the kernel build. Maybe this helper belongs in fs/nfsd/trace.h instead as a macro wrapper for __print_symbolic(). But see below.

Quoted 81 lines
fs/nfsd/nfs4proc.c +4 -0
 static __be32
 nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	     union nfsd4_op_u *u)
@@ -822,6 +847,10 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	__be32 status;
 	dev_t rdev;
 
+	trace_nfsd4_create(rqstp, &cstate->current_fh,
+			   nfs_type_to_vfs_type(create->cr_type),
+			   create->cr_name, create->cr_namelen);
+
 	fh_init(&resfh, NFS4_FHSIZE);
 
 	status = fh_verify(rqstp, &cstate->current_fh, S_IFDIR, NFSD_MAY_NOP);
fs/nfsd/nfsproc.c +3 -3
index 6dda081eb24c00b834ab0965c3a35a12115bceb7..33d8cbf8785588d38d4ec5efd769c1d1d06c6a91 100644
@@ -10,6 +10,7 @@
 #include "cache.h"
 #include "xdr.h"
 #include "vfs.h"
+#include "trace.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
@@ -292,8 +293,7 @@ nfsd_proc_create(struct svc_rqst *rqstp)
 	int		hosterr;
 	dev_t		rdev = 0, wanted = new_decode_dev(attr->ia_size);
 
-	dprintk("nfsd: CREATE   %s %.*s\n",
-		SVCFH_fmt(dirfhp), argp->len, argp->name);
+	trace_nfsd_proc_create(rqstp, dirfhp, S_IFREG, argp->name, argp->len);
 
 	/* First verify the parent file handle */
 	resp->status = fh_verify(rqstp, dirfhp, S_IFDIR, NFSD_MAY_EXEC);
@@ -548,7 +548,7 @@ nfsd_proc_mkdir(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
-	dprintk("nfsd: MKDIR    %s %.*s\n", SVCFH_fmt(&argp->fh), argp->len, argp->name);
+	trace_nfsd_proc_mkdir(rqstp, &argp->fh, S_IFDIR, argp->name, argp->len);
 
 	if (resp->fh.fh_dentry) {
 		printk(KERN_WARNING
fs/nfsd/trace.h +28 -0
index 382849d7c321d6ded8213890c2e7075770aa716c..c6aff23a845f06c87e701d57ec577c2c5c5a743c 100644
@@ -2391,6 +2391,45 @@ TRACE_EVENT(nfsd_lookup_dentry,
 	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s",
 		  __entry->xid, __entry->fh_hash, __get_str(name))
 );
+
+DECLARE_EVENT_CLASS(nfsd_vfs_create_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 umode_t type,
+		 const char *name,
+		 unsigned int len),
+	TP_ARGS(rqstp, fhp, type, name, len),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__field(umode_t, type)
+		__string_len(name, name, len)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__entry->type = type;
+		__assign_str(name);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x type=%s name=%s",
+		  __entry->xid, __entry->fh_hash,
+		  show_fs_file_type(__entry->type), __get_str(name))
+);
+
+#define DEFINE_NFSD_VFS_CREATE_EVENT(__name)						\
+	DEFINE_EVENT(nfsd_vfs_create_class, __name,					\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,		\

fs/nfsd/trace.h TRACE_EVENT(nfsd_lookup_dentry, · line 2431

Quoted 10 lines
fs/nfsd/trace.h +10 -0
+			      umode_t type, const char *name, unsigned int len),	\
+		     TP_ARGS(rqstp, fhp, type, name, len))
+
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mkdir);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mknod);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create);
+DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create_file);

I think we would be better off with one or two new trace points in nfsd_create() and nfsd_create_setattr() instead of all of these...

Quoted 2 lines

Unless I've missed what you are trying to observe...?

Now I remember why I did it this way. Doing it the way you suggest makes this all a bit messy:

Most of the callers create files via nfsd_create(), but nfsd3_create_file calls vfs_create() directly. It does call nfsd_create_setattr(), but that's really not the right place for this either, since it's doing a setattr and not the create itself. Notably, it isn't passed the filename.

Quoted 6 lines

Note too that burying this tracepoint down in nfs_create() also means that error conditions can happen before the tracepoint that may mean that it won't fire. So, if you're watching the traces to see when CREATE or MKDIR calls occur, you may miss some of them.

How do you want to proceed?

You can see incoming RPC messages via the svc_process trace point. If you want to see the arguments for the operations too, you enable the nfsd_vfs_ tracepoints. It is reasonable to add trace points in the error flows if you believe those are important to surface.

For CREATE, add the new nfsd_vfs_create trace point in nfsd_create() and also in nfsd3_create_file().

How does that sound?

Signature · 1 line
Chuck Lever
author maintainer
#19 lore raw

PATCH v2 7/12 nfsd: add tracepoints for symlink events

In reply to Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c |  8 +++-----
 fs/nfsd/nfs4proc.c |  3 +++
 fs/nfsd/nfsproc.c  |  7 +++----
 fs/nfsd/trace.h    | 35 +++++++++++++++++++++++++++++++++++
 4 files changed, 44 insertions(+), 9 deletions(-)
fs/nfsd/nfs3proc.c +3 -5
index ea1280970ea11b2a82f0de88ad0422eef7063d6d..587fc92597e7c77d078e871b8d12684c6b5efa2d 100644
@@ -423,6 +423,9 @@ nfsd3_proc_symlink(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
+	trace_nfsd3_proc_symlink(rqstp, &argp->ffh, argp->fname, argp->flen,
+				 argp->tname, argp->tlen);
+
 	if (argp->tlen == 0) {
 		resp->status = nfserr_inval;
 		goto out;
@@ -440,11 +443,6 @@ nfsd3_proc_symlink(struct svc_rqst *rqstp)
 		goto out;
 	}
 
-	dprintk("nfsd: SYMLINK(3)  %s %.*s -> %.*s\n",
-				SVCFH_fmt(&argp->ffh),
-				argp->flen, argp->fname,
-				argp->tlen, argp->tname);
-
 	fh_copy(&resp->dirfh, &argp->ffh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
 	resp->status = nfsd_symlink(rqstp, &resp->dirfh, argp->fname,
fs/nfsd/nfs4proc.c +3 -0
index 2c795103deaa4044596bd07d90db788169a32a0c..e22596a2e311861be1e4f595d77547be04634ce7 100644
@@ -873,6 +873,9 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	current->fs->umask = create->cr_umask;
 	switch (create->cr_type) {
 	case NF4LNK:
+		trace_nfsd4_symlink(rqstp, &cstate->current_fh,
+				    create->cr_name, create->cr_namelen,
+				    create->cr_data, create->cr_datalen);
 		status = nfsd_symlink(rqstp, &cstate->current_fh,
 				      create->cr_name, create->cr_namelen,
 				      create->cr_data, &attrs, &resfh);
fs/nfsd/nfsproc.c +3 -4
index 33d8cbf8785588d38d4ec5efd769c1d1d06c6a91..0674ed6b978f6caa1325a9271f2fde9b3ef60945 100644
@@ -506,6 +506,9 @@ nfsd_proc_symlink(struct svc_rqst *rqstp)
 	};
 	struct svc_fh	newfh;
 
+	trace_nfsd_proc_symlink(rqstp, &argp->ffh, argp->fname, argp->flen,
+				argp->tname, argp->tlen);
+
 	if (argp->tlen > NFS_MAXPATHLEN) {
 		resp->status = nfserr_nametoolong;
 		goto out;
@@ -519,10 +522,6 @@ nfsd_proc_symlink(struct svc_rqst *rqstp)
 		goto out;
 	}
 
-	dprintk("nfsd: SYMLINK  %s %.*s -> %.*s\n",
-		SVCFH_fmt(&argp->ffh), argp->flen, argp->fname,
-		argp->tlen, argp->tname);
-
 	fh_init(&newfh, NFS_FHSIZE);
 	resp->status = nfsd_symlink(rqstp, &argp->ffh, argp->fname, argp->flen,
 				    argp->tname, &attrs, &newfh);
fs/nfsd/trace.h +35 -0
index c6aff23a845f06c87e701d57ec577c2c5c5a743c..850dbf1240b234b67dd7d75d6903c0f49dc01261 100644
@@ -2430,6 +2430,41 @@ DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mknod);
 DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create);
 DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create_file);
 
+DECLARE_EVENT_CLASS(nfsd_vfs_symlink_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 const char *name,
+		 unsigned int namelen,
+		 const char *tgt,
+		 unsigned int tgtlen),
+	TP_ARGS(rqstp, fhp, name, namelen, tgt, tgtlen),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__string_len(name, name, namelen)
+		__string_len(tgt, tgt, tgtlen)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__assign_str(name);
+		__assign_str(tgt);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s target=%s",
+		  __entry->xid, __entry->fh_hash,
+		  __get_str(name), __get_str(tgt))
+);
+
+#define DEFINE_NFSD_VFS_SYMLINK_EVENT(__name)					\
+	DEFINE_EVENT(nfsd_vfs_symlink_class, __name,				\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,	\
+			      const char *name, unsigned int namelen,		\
+			      const char *tgt, unsigned int tgtlen),		\
+		     TP_ARGS(rqstp, fhp, name, namelen, tgt, tgtlen))
+
+DEFINE_NFSD_VFS_SYMLINK_EVENT(nfsd_proc_symlink);
+DEFINE_NFSD_VFS_SYMLINK_EVENT(nfsd3_proc_symlink);
+DEFINE_NFSD_VFS_SYMLINK_EVENT(nfsd4_symlink);
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
2.49.0
#20 lore raw

Re: [PATCH v2 07/12] nfsd: add tracepoints for symlink events

Quoted 108 lines · Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c |  8 +++-----
 fs/nfsd/nfs4proc.c |  3 +++
 fs/nfsd/nfsproc.c  |  7 +++----
 fs/nfsd/trace.h    | 35 +++++++++++++++++++++++++++++++++++
 4 files changed, 44 insertions(+), 9 deletions(-)
fs/nfsd/nfs3proc.c +3 -5
index ea1280970ea11b2a82f0de88ad0422eef7063d6d..587fc92597e7c77d078e871b8d12684c6b5efa2d 100644
@@ -423,6 +423,9 @@ nfsd3_proc_symlink(struct svc_rqst *rqstp)
 		.na_iattr	= &argp->attrs,
 	};
 
+	trace_nfsd3_proc_symlink(rqstp, &argp->ffh, argp->fname, argp->flen,
+				 argp->tname, argp->tlen);
+
 	if (argp->tlen == 0) {
 		resp->status = nfserr_inval;
 		goto out;
@@ -440,11 +443,6 @@ nfsd3_proc_symlink(struct svc_rqst *rqstp)
 		goto out;
 	}
 
-	dprintk("nfsd: SYMLINK(3)  %s %.*s -> %.*s\n",
-				SVCFH_fmt(&argp->ffh),
-				argp->flen, argp->fname,
-				argp->tlen, argp->tname);
-
 	fh_copy(&resp->dirfh, &argp->ffh);
 	fh_init(&resp->fh, NFS3_FHSIZE);
 	resp->status = nfsd_symlink(rqstp, &resp->dirfh, argp->fname,
fs/nfsd/nfs4proc.c +3 -0
index 2c795103deaa4044596bd07d90db788169a32a0c..e22596a2e311861be1e4f595d77547be04634ce7 100644
@@ -873,6 +873,9 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	current->fs->umask = create->cr_umask;
 	switch (create->cr_type) {
 	case NF4LNK:
+		trace_nfsd4_symlink(rqstp, &cstate->current_fh,
+				    create->cr_name, create->cr_namelen,
+				    create->cr_data, create->cr_datalen);
 		status = nfsd_symlink(rqstp, &cstate->current_fh,
 				      create->cr_name, create->cr_namelen,
 				      create->cr_data, &attrs, &resfh);
fs/nfsd/nfsproc.c +3 -4
index 33d8cbf8785588d38d4ec5efd769c1d1d06c6a91..0674ed6b978f6caa1325a9271f2fde9b3ef60945 100644
@@ -506,6 +506,9 @@ nfsd_proc_symlink(struct svc_rqst *rqstp)
 	};
 	struct svc_fh	newfh;
 
+	trace_nfsd_proc_symlink(rqstp, &argp->ffh, argp->fname, argp->flen,
+				argp->tname, argp->tlen);
+
 	if (argp->tlen > NFS_MAXPATHLEN) {
 		resp->status = nfserr_nametoolong;
 		goto out;
@@ -519,10 +522,6 @@ nfsd_proc_symlink(struct svc_rqst *rqstp)
 		goto out;
 	}
 
-	dprintk("nfsd: SYMLINK  %s %.*s -> %.*s\n",
-		SVCFH_fmt(&argp->ffh), argp->flen, argp->fname,
-		argp->tlen, argp->tname);
-
 	fh_init(&newfh, NFS_FHSIZE);
 	resp->status = nfsd_symlink(rqstp, &argp->ffh, argp->fname, argp->flen,
 				    argp->tname, &attrs, &newfh);
fs/nfsd/trace.h +25 -0
index c6aff23a845f06c87e701d57ec577c2c5c5a743c..850dbf1240b234b67dd7d75d6903c0f49dc01261 100644
@@ -2430,6 +2430,41 @@ DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mknod);
 DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create);
 DEFINE_NFSD_VFS_CREATE_EVENT(nfsd4_create_file);
 
+DECLARE_EVENT_CLASS(nfsd_vfs_symlink_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 const char *name,
+		 unsigned int namelen,
+		 const char *tgt,
+		 unsigned int tgtlen),
+	TP_ARGS(rqstp, fhp, name, namelen, tgt, tgtlen),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__string_len(name, name, namelen)
+		__string_len(tgt, tgt, tgtlen)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__assign_str(name);
+		__assign_str(tgt);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s target=%s",
+		  __entry->xid, __entry->fh_hash,
+		  __get_str(name), __get_str(tgt))
+);
+

fs/nfsd/trace.h DEFINE_NFSD_VFS_CREATE_EVENT(nfsd3_proc_mknod); · line 2467

Quoted 10 lines
fs/nfsd/trace.h +10 -0
+#define DEFINE_NFSD_VFS_SYMLINK_EVENT(__name)					\
+	DEFINE_EVENT(nfsd_vfs_symlink_class, __name,				\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,	\
+			      const char *name, unsigned int namelen,		\
+			      const char *tgt, unsigned int tgtlen),		\
+		     TP_ARGS(rqstp, fhp, name, namelen, tgt, tgtlen))
+
+DEFINE_NFSD_VFS_SYMLINK_EVENT(nfsd_proc_symlink);
+DEFINE_NFSD_VFS_SYMLINK_EVENT(nfsd3_proc_symlink);
+DEFINE_NFSD_VFS_SYMLINK_EVENT(nfsd4_symlink);

Likewise, maybe one tracepoint in nfsd_symlink() would be better? This comment also applies to the remaining patches in this series.

If you're trying to capture the NFS version too, perhaps SVC_RQST_ENDPOINT can be expanded to record the RPC program and version.

Also, name these new tracepoints "nfsd_vfs_yada" and then you can use a glob to enable the VFS tracepoints in one go, like so:

  # trace-cmd record -e nfsd:nfsd_vfs_\*
Quoted 4 lines
fs/nfsd/trace.h +0 -0
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
Chuck Lever
author maintainer
#21 lore raw

PATCH v2 8/12 nfsd: add tracepoints for hardlink events

In reply to Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c |  7 +------
 fs/nfsd/nfs4proc.c |  3 +++
 fs/nfsd/nfsproc.c  |  7 +------
 fs/nfsd/trace.h    | 35 +++++++++++++++++++++++++++++++++++
 4 files changed, 40 insertions(+), 12 deletions(-)
fs/nfsd/nfs3proc.c +1 -6
index 587fc92597e7c77d078e871b8d12684c6b5efa2d..97aaf98d0af7dc565b21937ecca4852dd9253221 100644
@@ -564,12 +564,7 @@ nfsd3_proc_link(struct svc_rqst *rqstp)
 	struct nfsd3_linkargs *argp = rqstp->rq_argp;
 	struct nfsd3_linkres  *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: LINK(3)     %s ->\n",
-				SVCFH_fmt(&argp->ffh));
-	dprintk("nfsd:   -> %s %.*s\n",
-				SVCFH_fmt(&argp->tfh),
-				argp->tlen,
-				argp->tname);
+	trace_nfsd_proc_link(rqstp, &argp->ffh, &argp->tfh, argp->tname, argp->tlen);
 
 	fh_copy(&resp->fh,  &argp->ffh);
 	fh_copy(&resp->tfh, &argp->tfh);
fs/nfsd/nfs4proc.c +3 -0
index e22596a2e311861be1e4f595d77547be04634ce7..7dffae2f16d9fa8dea043b7bf300eaca52c0aa7c 100644
@@ -977,6 +977,9 @@ nfsd4_link(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	struct nfsd4_link *link = &u->link;
 	__be32 status;
 
+	trace_nfsd4_link(rqstp, &cstate->save_fh, &cstate->current_fh,
+			 link->li_name, link->li_namelen);
+
 	status = nfsd_link(rqstp, &cstate->current_fh,
 			   link->li_name, link->li_namelen, &cstate->save_fh);
 	if (!status)
fs/nfsd/nfsproc.c +1 -6
index 0674ed6b978f6caa1325a9271f2fde9b3ef60945..b40b5ab1d3b17dd8974fcaeda3ac7c26baee67cf 100644
@@ -481,12 +481,7 @@ nfsd_proc_link(struct svc_rqst *rqstp)
 	struct nfsd_linkargs *argp = rqstp->rq_argp;
 	struct nfsd_stat *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: LINK     %s ->\n",
-		SVCFH_fmt(&argp->ffh));
-	dprintk("nfsd:    %s %.*s\n",
-		SVCFH_fmt(&argp->tfh),
-		argp->tlen,
-		argp->tname);
+	trace_nfsd_proc_link(rqstp, &argp->ffh, &argp->tfh, argp->tname, argp->tlen);
 
 	resp->status = nfsd_link(rqstp, &argp->tfh, argp->tname, argp->tlen,
 				 &argp->ffh);
fs/nfsd/trace.h +35 -0
index 850dbf1240b234b67dd7d75d6903c0f49dc01261..9ff919a08f424bfe023cf91244fe08effbdf993e 100644
@@ -2465,6 +2465,41 @@ DECLARE_EVENT_CLASS(nfsd_vfs_symlink_class,
 DEFINE_NFSD_VFS_SYMLINK_EVENT(nfsd_proc_symlink);
 DEFINE_NFSD_VFS_SYMLINK_EVENT(nfsd3_proc_symlink);
 DEFINE_NFSD_VFS_SYMLINK_EVENT(nfsd4_symlink);
+
+DECLARE_EVENT_CLASS(nfsd_vfs_link_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *sfhp,
+		 struct svc_fh *tfhp,
+		 const char *name,
+		 unsigned int namelen),
+	TP_ARGS(rqstp, sfhp, tfhp, name, namelen),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, sfh_hash)
+		__field(u32, tfh_hash)
+		__string_len(name, name, namelen)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->sfh_hash = knfsd_fh_hash(&sfhp->fh_handle);
+		__entry->tfh_hash = knfsd_fh_hash(&tfhp->fh_handle);
+		__assign_str(name);
+	),
+	TP_printk("xid=0x%08x src_fh=0x%08x tgt_fh=0x%08x name=%s",
+		  __entry->xid, __entry->sfh_hash, __entry->tfh_hash,
+		  __get_str(name))
+);
+
+#define DEFINE_NFSD_VFS_LINK_EVENT(__name)					\
+	DEFINE_EVENT(nfsd_vfs_link_class, __name,				\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *sfhp,	\
+			      struct svc_fh *tfhp, const char *name,		\
+			      unsigned int namelen),				\
+		     TP_ARGS(rqstp, sfhp, tfhp, name, namelen))
+
+DEFINE_NFSD_VFS_LINK_EVENT(nfsd_proc_link);
+DEFINE_NFSD_VFS_LINK_EVENT(nfsd3_proc_link);
+DEFINE_NFSD_VFS_LINK_EVENT(nfsd4_link);
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
2.49.0
author maintainer
#22 lore raw

PATCH v2 9/12 nfsd: add tracepoints for unlink events

In reply to Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c | 10 ++--------
 fs/nfsd/nfs4proc.c |  2 ++
 fs/nfsd/nfsproc.c  |  5 ++---
 fs/nfsd/trace.h    | 34 ++++++++++++++++++++++++++++++++++
 4 files changed, 40 insertions(+), 11 deletions(-)
fs/nfsd/nfs3proc.c +2 -8
index 97aaf98d0af7dc565b21937ecca4852dd9253221..8893bf5e0b1d15b24e9c2c71fa1a8a09586a03d3 100644
@@ -501,10 +501,7 @@ nfsd3_proc_remove(struct svc_rqst *rqstp)
 	struct nfsd3_diropargs *argp = rqstp->rq_argp;
 	struct nfsd3_attrstat *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: REMOVE(3)   %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_remove(rqstp, &argp->fh, argp->name, argp->len);
 
 	/* Unlink. -S_IFDIR means file must not be a directory */
 	fh_copy(&resp->fh, &argp->fh);
@@ -523,10 +520,7 @@ nfsd3_proc_rmdir(struct svc_rqst *rqstp)
 	struct nfsd3_diropargs *argp = rqstp->rq_argp;
 	struct nfsd3_attrstat *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: RMDIR(3)    %s %.*s\n",
-				SVCFH_fmt(&argp->fh),
-				argp->len,
-				argp->name);
+	trace_nfsd3_proc_rmdir(rqstp, &argp->fh, argp->name, argp->len);
 
 	fh_copy(&resp->fh, &argp->fh);
 	resp->status = nfsd_unlink(rqstp, &resp->fh, S_IFDIR,
fs/nfsd/nfs4proc.c +2 -0
index 7dffae2f16d9fa8dea043b7bf300eaca52c0aa7c..8524e78201e22984517e93cd9a2834190266c633 100644
@@ -1114,6 +1114,8 @@ nfsd4_remove(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	struct nfsd4_remove *remove = &u->remove;
 	__be32 status;
 
+	trace_nfsd4_remove(rqstp, &cstate->current_fh, remove->rm_name, remove->rm_namelen);
+
 	if (opens_in_grace(SVC_NET(rqstp)))
 		return nfserr_grace;
 	status = nfsd_unlink(rqstp, &cstate->current_fh, 0,
fs/nfsd/nfsproc.c +2 -3
index b40b5ab1d3b17dd8974fcaeda3ac7c26baee67cf..55656bb0264c31c10419ed41240c91ba66493106 100644
@@ -445,8 +445,7 @@ nfsd_proc_remove(struct svc_rqst *rqstp)
 	struct nfsd_diropargs *argp = rqstp->rq_argp;
 	struct nfsd_stat *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: REMOVE   %s %.*s\n", SVCFH_fmt(&argp->fh),
-		argp->len, argp->name);
+	trace_nfsd_proc_remove(rqstp, &argp->fh, argp->name, argp->len);
 
 	/* Unlink. -SIFDIR means file must not be a directory */
 	resp->status = nfsd_unlink(rqstp, &argp->fh, -S_IFDIR,
@@ -572,7 +571,7 @@ nfsd_proc_rmdir(struct svc_rqst *rqstp)
 	struct nfsd_diropargs *argp = rqstp->rq_argp;
 	struct nfsd_stat *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: RMDIR    %s %.*s\n", SVCFH_fmt(&argp->fh), argp->len, argp->name);
+	trace_nfsd_proc_rmdir(rqstp, &argp->fh, argp->name, argp->len);
 
 	resp->status = nfsd_unlink(rqstp, &argp->fh, S_IFDIR,
 				   argp->name, argp->len);
fs/nfsd/trace.h +34 -0
index 9ff919a08f424bfe023cf91244fe08effbdf993e..dd984917bd0a741ac545c06631ab2a7de8af5158 100644
@@ -2500,6 +2500,40 @@ DECLARE_EVENT_CLASS(nfsd_vfs_link_class,
 DEFINE_NFSD_VFS_LINK_EVENT(nfsd_proc_link);
 DEFINE_NFSD_VFS_LINK_EVENT(nfsd3_proc_link);
 DEFINE_NFSD_VFS_LINK_EVENT(nfsd4_link);
+
+DECLARE_EVENT_CLASS(nfsd_vfs_unlink_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 const char *name,
+		 unsigned int len),
+	TP_ARGS(rqstp, fhp, name, len),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__string_len(name, name, len)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__assign_str(name);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x name=%s",
+		  __entry->xid, __entry->fh_hash,
+		  __get_str(name))
+);
+
+#define DEFINE_NFSD_VFS_UNLINK_EVENT(__name)						\
+	DEFINE_EVENT(nfsd_vfs_unlink_class, __name,					\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,		\
+			      const char *name, unsigned int len),			\
+		     TP_ARGS(rqstp, fhp, name, len))
+
+DEFINE_NFSD_VFS_UNLINK_EVENT(nfsd_proc_remove);
+DEFINE_NFSD_VFS_UNLINK_EVENT(nfsd_proc_rmdir);
+DEFINE_NFSD_VFS_UNLINK_EVENT(nfsd3_proc_remove);
+DEFINE_NFSD_VFS_UNLINK_EVENT(nfsd3_proc_rmdir);
+DEFINE_NFSD_VFS_UNLINK_EVENT(nfsd4_remove);
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
2.49.0
author maintainer
#23 lore raw

PATCH v2 10/12 nfsd: add tracepoints to rename events

In reply to Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c | 10 ++--------
 fs/nfsd/nfs4proc.c |  4 ++++
 fs/nfsd/nfsproc.c  |  6 ++----
 fs/nfsd/trace.h    | 40 ++++++++++++++++++++++++++++++++++++++++
 4 files changed, 48 insertions(+), 12 deletions(-)
fs/nfsd/nfs3proc.c +2 -8
index 8893bf5e0b1d15b24e9c2c71fa1a8a09586a03d3..4fd3c2284eb96c1d712639675140412b84eadb2f 100644
@@ -535,14 +535,8 @@ nfsd3_proc_rename(struct svc_rqst *rqstp)
 	struct nfsd3_renameargs *argp = rqstp->rq_argp;
 	struct nfsd3_renameres *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: RENAME(3)   %s %.*s ->\n",
-				SVCFH_fmt(&argp->ffh),
-				argp->flen,
-				argp->fname);
-	dprintk("nfsd: -> %s %.*s\n",
-				SVCFH_fmt(&argp->tfh),
-				argp->tlen,
-				argp->tname);
+	trace_nfsd3_proc_rename(rqstp, &argp->ffh, &argp->tfh, argp->fname,
+				argp->flen, argp->tname, argp->tlen);
 
 	fh_copy(&resp->ffh, &argp->ffh);
 	fh_copy(&resp->tfh, &argp->tfh);
fs/nfsd/nfs4proc.c +4 -0
index 8524e78201e22984517e93cd9a2834190266c633..7e6c80e0482a997d4085c87dae88d10c2f06b77b 100644
@@ -1132,6 +1132,10 @@ nfsd4_rename(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	struct nfsd4_rename *rename = &u->rename;
 	__be32 status;
 
+	trace_nfsd4_rename(rqstp, &cstate->save_fh, &cstate->current_fh,
+			   rename->rn_sname, rename->rn_snamelen,
+			   rename->rn_tname, rename->rn_tnamelen);
+
 	if (opens_in_grace(SVC_NET(rqstp)))
 		return nfserr_grace;
 	status = nfsd_rename(rqstp, &cstate->save_fh, rename->rn_sname,
fs/nfsd/nfsproc.c +2 -4
index 55656bb0264c31c10419ed41240c91ba66493106..d99e1bff2f8a99e477e3cf21eb7058cfe40a7cb4 100644
@@ -461,10 +461,8 @@ nfsd_proc_rename(struct svc_rqst *rqstp)
 	struct nfsd_renameargs *argp = rqstp->rq_argp;
 	struct nfsd_stat *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: RENAME   %s %.*s -> \n",
-		SVCFH_fmt(&argp->ffh), argp->flen, argp->fname);
-	dprintk("nfsd:        ->  %s %.*s\n",
-		SVCFH_fmt(&argp->tfh), argp->tlen, argp->tname);
+	trace_nfsd_proc_rename(rqstp, &argp->ffh, &argp->tfh, argp->fname,
+			       argp->flen, argp->tname, argp->tlen);
 
 	resp->status = nfsd_rename(rqstp, &argp->ffh, argp->fname, argp->flen,
 				   &argp->tfh, argp->tname, argp->tlen);
fs/nfsd/trace.h +40 -0
index dd984917bd0a741ac545c06631ab2a7de8af5158..7bf3ee4acd9862171cae5caefced3507f4897e90 100644
@@ -2534,6 +2534,46 @@ DEFINE_NFSD_VFS_UNLINK_EVENT(nfsd3_proc_remove);
 DEFINE_NFSD_VFS_UNLINK_EVENT(nfsd3_proc_rmdir);
 DEFINE_NFSD_VFS_UNLINK_EVENT(nfsd4_remove);
 
+DECLARE_EVENT_CLASS(nfsd_vfs_rename_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *sfhp,
+		 struct svc_fh *tfhp,
+		 const char *name,
+		 unsigned int namelen,
+		 const char *tgt,
+		 unsigned int tgtlen),
+	TP_ARGS(rqstp, sfhp, tfhp, name, namelen, tgt, tgtlen),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, sfh_hash)
+		__field(u32, tfh_hash)
+		__string_len(name, name, namelen)
+		__string_len(tgt, tgt, tgtlen)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->sfh_hash = knfsd_fh_hash(&sfhp->fh_handle);
+		__entry->tfh_hash = knfsd_fh_hash(&tfhp->fh_handle);
+		__assign_str(name);
+		__assign_str(tgt);
+	),
+	TP_printk("xid=0x%08x sfh_hash=0x%08x tfh_hash=0x%08x name=%s target=%s",
+		  __entry->xid, __entry->sfh_hash, __entry->tfh_hash,
+		  __get_str(name), __get_str(tgt))
+);
+
+#define DEFINE_NFSD_VFS_RENAME_EVENT(__name)					\
+	DEFINE_EVENT(nfsd_vfs_rename_class, __name,				\
+		     TP_PROTO(struct svc_rqst *rqstp,				\
+			      struct svc_fh *sfhp, struct svc_fh *tfhp,		\
+			      const char *name, unsigned int namelen,		\
+			      const char *tgt, unsigned int tgtlen),		\
+		     TP_ARGS(rqstp, sfhp, tfhp, name, namelen, tgt, tgtlen))
+
+DEFINE_NFSD_VFS_RENAME_EVENT(nfsd_proc_rename);
+DEFINE_NFSD_VFS_RENAME_EVENT(nfsd3_proc_rename);
+DEFINE_NFSD_VFS_RENAME_EVENT(nfsd4_rename);
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
2.49.0
author maintainer
#24 lore raw

PATCH v2 11/12 nfsd: add tracepoints for readdir events

In reply to Jeff Layton

...and remove the legacy dprintks.

---
 fs/nfsd/nfs3proc.c |  8 ++------
 fs/nfsd/nfs4proc.c |  2 ++
 fs/nfsd/nfsproc.c  |  4 +---
 fs/nfsd/trace.h    | 32 ++++++++++++++++++++++++++++++++
 4 files changed, 37 insertions(+), 9 deletions(-)
fs/nfsd/nfs3proc.c +2 -6
index 4fd3c2284eb96c1d712639675140412b84eadb2f..b9ed55a4cdd740bc0aa84b7cbc0e37906d55d666 100644
@@ -592,9 +592,7 @@ nfsd3_proc_readdir(struct svc_rqst *rqstp)
 	struct nfsd3_readdirres  *resp = rqstp->rq_resp;
 	loff_t		offset;
 
-	dprintk("nfsd: READDIR(3)  %s %d bytes at %d\n",
-				SVCFH_fmt(&argp->fh),
-				argp->count, (u32) argp->cookie);
+	trace_nfsd3_proc_readdir(rqstp, &argp->fh, argp->cookie, argp->count);
 
 	nfsd3_init_dirlist_pages(rqstp, resp, argp->count);
 
@@ -626,9 +624,7 @@ nfsd3_proc_readdirplus(struct svc_rqst *rqstp)
 	struct nfsd3_readdirres  *resp = rqstp->rq_resp;
 	loff_t	offset;
 
-	dprintk("nfsd: READDIR+(3) %s %d bytes at %d\n",
-				SVCFH_fmt(&argp->fh),
-				argp->count, (u32) argp->cookie);
+	trace_nfsd3_proc_readdirplus(rqstp, &argp->fh, argp->cookie, argp->count);
 
 	nfsd3_init_dirlist_pages(rqstp, resp, argp->count);
 
fs/nfsd/nfs4proc.c +2 -0
index 7e6c80e0482a997d4085c87dae88d10c2f06b77b..8dd1233693dc82febe300f6f2714059c718909bc 100644
@@ -1080,6 +1080,8 @@ nfsd4_readdir(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	u64 cookie = readdir->rd_cookie;
 	static const nfs4_verifier zeroverf;
 
+	trace_nfsd4_readdir(rqstp, &cstate->current_fh, cookie, readdir->rd_dircount);
+
 	/* no need to check permission - this will be done in nfsd_readdir() */
 
 	if (readdir->rd_bmval[1] & NFSD_WRITEONLY_ATTRS_WORD1)
fs/nfsd/nfsproc.c +1 -3
index d99e1bff2f8a99e477e3cf21eb7058cfe40a7cb4..ce3f1ca636f79687e65077effcc0588639d9366d 100644
@@ -606,9 +606,7 @@ nfsd_proc_readdir(struct svc_rqst *rqstp)
 	struct nfsd_readdirres *resp = rqstp->rq_resp;
 	loff_t		offset;
 
-	dprintk("nfsd: READDIR  %s %d bytes at %d\n",
-		SVCFH_fmt(&argp->fh),		
-		argp->count, argp->cookie);
+	trace_nfsd_proc_readdir(rqstp, &argp->fh, argp->cookie, argp->count);
 
 	nfsd_init_dirlist_pages(rqstp, resp, argp->count);
 
fs/nfsd/trace.h +32 -0
index 7bf3ee4acd9862171cae5caefced3507f4897e90..1d48a37dd33fb4c6e338534d576bcc8fd1a8f54d 100644
@@ -2574,6 +2574,38 @@ DEFINE_NFSD_VFS_RENAME_EVENT(nfsd_proc_rename);
 DEFINE_NFSD_VFS_RENAME_EVENT(nfsd3_proc_rename);
 DEFINE_NFSD_VFS_RENAME_EVENT(nfsd4_rename);
 
+DECLARE_EVENT_CLASS(nfsd_vfs_readdir_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp,
+		 u64 cookie, u32 count),
+	TP_ARGS(rqstp, fhp, cookie, count),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+		__field(u64, cookie)
+		__field(u32, count)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+		__entry->cookie = cookie;
+		__entry->count = count;
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x cookie=0x%llx count=%u",
+		  __entry->xid, __entry->fh_hash, __entry->cookie, __entry->count)
+);
+
+#define DEFINE_NFSD_VFS_READDIR_EVENT(__name)					\
+	DEFINE_EVENT(nfsd_vfs_readdir_class, __name,				\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp,	\
+			      u64 cookie, u32 count),				\
+		     TP_ARGS(rqstp, fhp, cookie, count))
+
+DEFINE_NFSD_VFS_READDIR_EVENT(nfsd_proc_readdir);
+DEFINE_NFSD_VFS_READDIR_EVENT(nfsd3_proc_readdir);
+DEFINE_NFSD_VFS_READDIR_EVENT(nfsd3_proc_readdirplus);
+DEFINE_NFSD_VFS_READDIR_EVENT(nfsd4_readdir);
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
2.49.0
author maintainer
#25 lore raw

PATCH v2 12/12 nfsd: add tracepoint for getattr events

In reply to Jeff Layton

...including STATFS on v2 and FSSTAT on v3. Also, remove the dprintks.

---
 fs/nfsd/nfs3proc.c |  6 ++----
 fs/nfsd/nfs4proc.c |  2 ++
 fs/nfsd/nfsproc.c  |  4 ++--
 fs/nfsd/trace.h    | 27 +++++++++++++++++++++++++++
 4 files changed, 33 insertions(+), 6 deletions(-)
fs/nfsd/nfs3proc.c +2 -4
index b9ed55a4cdd740bc0aa84b7cbc0e37906d55d666..d2c6902ca9e2496d90ddf13e96b119e415c23cfc 100644
@@ -70,8 +70,7 @@ nfsd3_proc_getattr(struct svc_rqst *rqstp)
 	struct nfsd_fhandle *argp = rqstp->rq_argp;
 	struct nfsd3_attrstat *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: GETATTR(3)  %s\n",
-		SVCFH_fmt(&argp->fh));
+	trace_nfsd3_proc_getattr(rqstp, &argp->fh);
 
 	fh_copy(&resp->fh, &argp->fh);
 	resp->status = fh_verify(rqstp, &resp->fh, 0,
@@ -665,8 +664,7 @@ nfsd3_proc_fsstat(struct svc_rqst *rqstp)
 	struct nfsd_fhandle *argp = rqstp->rq_argp;
 	struct nfsd3_fsstatres *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: FSSTAT(3)   %s\n",
-				SVCFH_fmt(&argp->fh));
+	trace_nfsd3_proc_fsstat(rqstp, &argp->fh);
 
 	resp->status = nfsd_statfs(rqstp, &argp->fh, &resp->stats, 0);
 	fh_put(&argp->fh);
fs/nfsd/nfs4proc.c +2 -0
index 8dd1233693dc82febe300f6f2714059c718909bc..1b2cf5a2d8265659edaa7177bc778f74c1a532a3 100644
@@ -955,6 +955,8 @@ nfsd4_getattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	struct nfsd4_getattr *getattr = &u->getattr;
 	__be32 status;
 
+	trace_nfsd4_getattr(rqstp, &cstate->current_fh);
+
 	status = fh_verify(rqstp, &cstate->current_fh, 0, NFSD_MAY_NOP);
 	if (status)
 		return status;
fs/nfsd/nfsproc.c +2 -2
index ce3f1ca636f79687e65077effcc0588639d9366d..35b9f79f077adce4499c7f52244d69da8f5090c4 100644
@@ -55,7 +55,7 @@ nfsd_proc_getattr(struct svc_rqst *rqstp)
 	struct nfsd_fhandle *argp = rqstp->rq_argp;
 	struct nfsd_attrstat *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: GETATTR  %s\n", SVCFH_fmt(&argp->fh));
+	trace_nfsd_proc_getattr(rqstp, &argp->fh);
 
 	fh_copy(&resp->fh, &argp->fh);
 	resp->status = fh_verify(rqstp, &resp->fh, 0,
@@ -631,7 +631,7 @@ nfsd_proc_statfs(struct svc_rqst *rqstp)
 	struct nfsd_fhandle *argp = rqstp->rq_argp;
 	struct nfsd_statfsres *resp = rqstp->rq_resp;
 
-	dprintk("nfsd: STATFS   %s\n", SVCFH_fmt(&argp->fh));
+	trace_nfsd_proc_statfs(rqstp, &argp->fh);
 
 	resp->status = nfsd_statfs(rqstp, &argp->fh, &resp->stats,
 				   NFSD_MAY_BYPASS_GSS_ON_ROOT);
fs/nfsd/trace.h +27 -0
index 1d48a37dd33fb4c6e338534d576bcc8fd1a8f54d..ffe16a7ab1f24db999763bcc220e31cf8035d412 100644
@@ -2606,6 +2606,33 @@ DEFINE_NFSD_VFS_READDIR_EVENT(nfsd3_proc_readdir);
 DEFINE_NFSD_VFS_READDIR_EVENT(nfsd3_proc_readdirplus);
 DEFINE_NFSD_VFS_READDIR_EVENT(nfsd4_readdir);
 
+DECLARE_EVENT_CLASS(nfsd_vfs_getattr_class,
+	TP_PROTO(struct svc_rqst *rqstp,
+		 struct svc_fh *fhp),
+	TP_ARGS(rqstp, fhp),
+	TP_STRUCT__entry(
+		SVC_RQST_ENDPOINT_FIELDS(rqstp)
+		__field(u32, fh_hash)
+	),
+	TP_fast_assign(
+		SVC_RQST_ENDPOINT_ASSIGNMENTS(rqstp);
+		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
+	),
+	TP_printk("xid=0x%08x fh_hash=0x%08x",
+		  __entry->xid, __entry->fh_hash)
+);
+
+#define DEFINE_NFSD_VFS_GETATTR_EVENT(__name)					\
+	DEFINE_EVENT(nfsd_vfs_getattr_class, __name,				\
+		     TP_PROTO(struct svc_rqst *rqstp, struct svc_fh *fhp),	\
+		     TP_ARGS(rqstp, fhp))
+
+DEFINE_NFSD_VFS_GETATTR_EVENT(nfsd_proc_getattr);
+DEFINE_NFSD_VFS_GETATTR_EVENT(nfsd_proc_statfs);
+DEFINE_NFSD_VFS_GETATTR_EVENT(nfsd3_proc_getattr);
+DEFINE_NFSD_VFS_GETATTR_EVENT(nfsd3_proc_fsstat);
+DEFINE_NFSD_VFS_GETATTR_EVENT(nfsd4_getattr);
+
 #endif /* _NFSD_TRACE_H */
 
 #undef TRACE_INCLUDE_PATH
Signature · 1 line
2.49.0
#26 lore raw

Re: (subset) [PATCH v2 00/12] nfsd: observability improvements

In reply to Jeff Layton

From: Chuck Lever <[email protected]>

Quoted 10 lines · Jeff Layton

While troubleshooting a performance problem internally, it became evident that we needed tracepoints in nfsd_commit. The first patch adds that. While discussing that, Sargun pointed out some tracepoints he added using kprobes. Those are converted to static tracepoints here, and the legacy dprintk's removed.

Lastly, I've updated the svc_xprt_dequeue tracepoint to show how long the xprt sat on the queue before being serviced.

[...]

Applied to nfsd-testing, thanks!

[01/12] nfsd: add commit start/done tracepoints around nfsd_commit()

        commit: f2a3825118e6177f4de5ee9996248ced2918b08b

--
Chuck Lever