2023-03-10 15:44:59

by Jeff Layton

Subject: [PATCH] lockd: fix tracepoint status name definitions when CONFIG_LOCKD_V4 isn't defined

There is probably a better way to do this that doesn't repeat anything.

Reported-by: kernel test robot <[email protected]>
Link: https://lore.kernel.org/oe-kbuild-all/[email protected]/
Signed-off-by: Jeff Layton <[email protected]>
---
fs/lockd/trace.h | 10 ++++++++++
1 file changed, 10 insertions(+)

Chuck, I'm fine if you want to just fold this into the original
tracepoint patch too. Also, let me know if you see a way to express this
better.

diff --git a/fs/lockd/trace.h b/fs/lockd/trace.h
index 3e84e3efaf22..11f1381b566c 100644
--- a/fs/lockd/trace.h
+++ b/fs/lockd/trace.h
@@ -10,6 +10,7 @@
#include <linux/nfs.h>
#include <linux/lockd/lockd.h>

+#ifdef CONFIG_LOCKD_V4
#define show_nlm_status(val) \
__print_symbolic(val, \
{ NLM_LCK_GRANTED, "LCK_GRANTED" }, \
@@ -22,6 +23,15 @@
{ NLM_STALE_FH, "STALE_FH" }, \
{ NLM_FBIG, "FBIG" }, \
{ NLM_FAILED, "FAILED" })
+#else
+#define show_nlm_status(val) \
+ __print_symbolic(val, \
+ { NLM_LCK_GRANTED, "LCK_GRANTED" }, \
+ { NLM_LCK_DENIED, "LCK_DENIED" }, \
+ { NLM_LCK_DENIED_NOLOCKS, "LCK_DENIED_NOLOCKS" }, \
+ { NLM_LCK_BLOCKED, "LCK_BLOCKED" }, \
+ { NLM_LCK_DENIED_GRACE_PERIOD, "LCK_DENIED_GRACE_PERIOD" })
+#endif

DECLARE_EVENT_CLASS(nlmclnt_lock_event,
TP_PROTO(
--
2.39.2

2023-03-10 15:50:12

by Chuck Lever III

Subject: Re: [PATCH] lockd: fix tracepoint status name definitions when CONFIG_LOCKD_V4 isn't defined

> On Mar 10, 2023, at 10:32 AM, Jeff Layton <[email protected]> wrote:
>
> There is probably a better way to do this that doesn't repeat anything.
>
> Reported-by: kernel test robot <[email protected]>
> Link: https://lore.kernel.org/oe-kbuild-all/[email protected]/
> Signed-off-by: Jeff Layton <[email protected]>
> ---
> fs/lockd/trace.h | 10 ++++++++++
> 1 file changed, 10 insertions(+)
>
> Chuck, I'm fine if you want to just fold this into the original
> tracepoint patch too. Also, let me know if you see a way to express this
> better.

Interesting, I have the lockd tracepoint patch in nfsd-next, and
haven't heard a peep from the bots.

I'll fold something in.


> diff --git a/fs/lockd/trace.h b/fs/lockd/trace.h
> index 3e84e3efaf22..11f1381b566c 100644
> --- a/fs/lockd/trace.h
> +++ b/fs/lockd/trace.h
> @@ -10,6 +10,7 @@
> #include <linux/nfs.h>
> #include <linux/lockd/lockd.h>

Btw, here I think TRACE_DEFINE_ENUMs will be needed. I can confirm
that, and add if necessary.


> +#ifdef CONFIG_LOCKD_V4
> #define show_nlm_status(val) \
> __print_symbolic(val, \
> { NLM_LCK_GRANTED, "LCK_GRANTED" }, \
> @@ -22,6 +23,15 @@
> { NLM_STALE_FH, "STALE_FH" }, \
> { NLM_FBIG, "FBIG" }, \
> { NLM_FAILED, "FAILED" })
> +#else
> +#define show_nlm_status(val) \
> + __print_symbolic(val, \
> + { NLM_LCK_GRANTED, "LCK_GRANTED" }, \
> + { NLM_LCK_DENIED, "LCK_DENIED" }, \
> + { NLM_LCK_DENIED_NOLOCKS, "LCK_DENIED_NOLOCKS" }, \
> + { NLM_LCK_BLOCKED, "LCK_BLOCKED" }, \
> + { NLM_LCK_DENIED_GRACE_PERIOD, "LCK_DENIED_GRACE_PERIOD" })
> +#endif
>
> DECLARE_EVENT_CLASS(nlmclnt_lock_event,
> TP_PROTO(
> --
> 2.39.2
>

--
Chuck Lever