mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Gabriele Monaco <gmonaco@redhat.com>
To: wen.yang@linux.dev
Cc: Steven Rostedt <rostedt@goodmis.org>,
	linux-trace-kernel@vger.kernel.org,
		linux-kernel@vger.kernel.org
Subject: Re: [PATCH v3 4/9] rv/ha: fix ha_invariant_passed_ns silent bypass of invariant check
Date: Mon, 15 Jun 2026 12:12:26 +0200	[thread overview]
Message-ID: <4442008c90e6957d9d6916a3b30afd1401b2f760.camel@redhat.com> (raw)
In-Reply-To: <812b7b8e8979b4ab00ac7727e3fea578799f2a8b.1780847473.git.wen.yang@linux.dev>

On Mon, 2026-06-08 at 00:13 +0800, wen.yang@linux.dev wrote:
> From: Wen Yang <wen.yang@linux.dev>
> 
> The function is documented as "prepare the invariant and return the
> time
> since reset", but on the first call (env_store == U64_MAX) it exits
> early without calling ha_set_invariant_ns():
> 
>   if (ha_monitor_env_invalid(ha_mon, env))  /* env_store == U64_MAX
> */
>       return 0;   /* ha_set_invariant_ns skipped, env_store stays
> U64_MAX */
>   ...
>   ha_set_invariant_ns(ha_mon, env, expire - passed, time_ns);
> 
> This leaves env_store == U64_MAX, so ha_check_invariant_ns() always
> passes on the first activation regardless of elapsed time:
> 
>   return READ_ONCE(ha_mon->env_store[env]) >= time_ns;  /* U64_MAX >=
> any */
> 
> Fix: establish the guard before converting to the invariant:
> 
>   if (ha_monitor_env_invalid(ha_mon, env))
>       ha_reset_clk_ns(ha_mon, env, time_ns); /* guard: env_store =
> time_ns */
>   passed = ha_get_env(ha_mon, env, time_ns);
>   ha_set_invariant_ns(ha_mon, env, expire - passed, time_ns);
>                                  /* invariant: env_store = time_ns +
> expire */
> 
> Apply the same fix to ha_invariant_passed_jiffy().
> 
> Signed-off-by: Wen Yang <wen.yang@linux.dev>

This is neat and probably works just fine in your case. Anyway I'm
planning on a more seamless solution not involving invalid states at
all. Initialisation would include a reset, which should work better on
monitors doing da_handle_start(), so not running the first event.
That's, by the way, pretty much what you wanted by doing reset() on the
(virtual) init transition.

We first needs Nam's simplification though [1]. We can synchronise
during this development cycle, you probably won't really need to bother
for now.

Thanks,
Gabriele

[1] -
https://lore.kernel.org/lkml/08188c28f274da63a3f8549add3086a92aef45e5.1780908661.git.namcao@linutronix.de

> ---
>  include/rv/ha_monitor.h | 17 +++++++++++++----
>  1 file changed, 13 insertions(+), 4 deletions(-)
> 
> diff --git a/include/rv/ha_monitor.h b/include/rv/ha_monitor.h
> index 28d3c74cabfc..e5860900a337 100644
> --- a/include/rv/ha_monitor.h
> +++ b/include/rv/ha_monitor.h
> @@ -365,16 +365,22 @@ static inline bool ha_check_invariant_ns(struct
> ha_monitor *ha_mon,
>  }
>  /*
>   * ha_invariant_passed_ns - prepare the invariant and return the
> time since reset
> + *
> + * If the env has not been initialised yet (first entry into a state
> with an
> + * invariant), anchor the guard clock at the current time so that
> the full
> + * budget is available from this point.  This preserves the
> documented
> + * guard→invariant ordering: ha_set_invariant_ns() is always
> preceded by a
> + * valid guard representation in env_store.
>   */
>  static inline u64 ha_invariant_passed_ns(struct ha_monitor *ha_mon,
> enum envs env,
>  				   u64 expire, u64 time_ns)
>  {
> -	u64 passed = 0;
> +	u64 passed;
>  
>  	if (env < 0 || env >= ENV_MAX_STORED)
>  		return 0;
>  	if (ha_monitor_env_invalid(ha_mon, env))
> -		return 0;
> +		ha_reset_clk_ns(ha_mon, env, time_ns);
>  	passed = ha_get_env(ha_mon, env, time_ns);
>  	ha_set_invariant_ns(ha_mon, env, expire - passed, time_ns);
>  	return passed;
> @@ -404,16 +410,19 @@ static inline bool
> ha_check_invariant_jiffy(struct ha_monitor *ha_mon,
>  }
>  /*
>   * ha_invariant_passed_jiffy - prepare the invariant and return the
> time since reset
> + *
> + * Same first-use semantics as ha_invariant_passed_ns(): anchor the
> guard clock
> + * now if the env has not been initialised.
>   */
>  static inline u64 ha_invariant_passed_jiffy(struct ha_monitor
> *ha_mon, enum envs env,
>  				      u64 expire, u64 time_ns)
>  {
> -	u64 passed = 0;
> +	u64 passed;
>  
>  	if (env < 0 || env >= ENV_MAX_STORED)
>  		return 0;
>  	if (ha_monitor_env_invalid(ha_mon, env))
> -		return 0;
> +		ha_reset_clk_jiffy(ha_mon, env);
>  	passed = ha_get_env(ha_mon, env, time_ns);
>  	ha_set_invariant_jiffy(ha_mon, env, expire - passed);
>  	return passed;


  reply	other threads:[~2026-06-15 10:12 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-06-07 16:13 [PATCH v3 0/9] rv/tlob: Add task latency over budget RV monitor wen.yang
2026-06-07 16:13 ` [PATCH v3 1/9] rv/da: introduce DA_MON_ALLOCATION_STRATEGY wen.yang
2026-06-15  9:56   ` Gabriele Monaco
2026-06-07 16:13 ` [PATCH v3 2/9] rv: add generic uprobe infrastructure for RV monitors wen.yang
2026-06-16  9:49   ` Gabriele Monaco
2026-06-28 16:47     ` Wen Yang
2026-06-30  8:48       ` Gabriele Monaco
2026-06-30 18:44         ` Wen Yang
2026-07-01 11:57           ` Gabriele Monaco
2026-06-07 16:13 ` [PATCH v3 3/9] rv/tlob: add tlob model DOT file wen.yang
2026-06-07 16:13 ` [PATCH v3 4/9] rv/ha: fix ha_invariant_passed_ns silent bypass of invariant check wen.yang
2026-06-15 10:12   ` Gabriele Monaco [this message]
2026-06-07 16:13 ` [PATCH v3 5/9] rv/ha: make da_monitor_reset_hook and EVENT_NONE_LBL overridable wen.yang
2026-06-15 10:16   ` Gabriele Monaco
2026-06-07 16:13 ` [PATCH v3 6/9] rv/tlob: add tlob hybrid automaton monitor wen.yang
2026-06-15 15:24   ` Gabriele Monaco
2026-06-07 16:13 ` [PATCH v3 7/9] rv/tlob: add KUnit tests for the tlob monitor wen.yang
2026-06-17  7:49   ` Gabriele Monaco
2026-06-07 16:13 ` [PATCH v3 8/9] selftests/verification: fix verificationtest-ktap for out-of-tree execution wen.yang
2026-06-16 11:14   ` Gabriele Monaco
2026-06-07 16:13 ` [PATCH v3 9/9] selftests/verification: add tlob selftests wen.yang
2026-06-16 14:58   ` Gabriele Monaco
2026-06-17 15:09   ` Gabriele Monaco
2026-06-22  9:26   ` Gabriele Monaco
2026-06-13 16:00 ` [PATCH v3 0/9] rv/tlob: Add task latency over budget RV monitor Wen Yang

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4442008c90e6957d9d6916a3b30afd1401b2f760.camel@redhat.com \
    --to=gmonaco@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-trace-kernel@vger.kernel.org \
    --cc=rostedt@goodmis.org \
    --cc=wen.yang@linux.dev \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®