Re: [PATCH v2] mm/page_alloc: trace PCP refills and PCP zone lock usage

[Steven Rostedt <rostedt@goodmis.org>]

On Wed, 29 Apr 2026 11:31:53 +0800 (CST)
SUVONOV BUNYOD <b.suvonov@sjtu.edu.cn> wrote:

> Thanks for reviewing Steven,
> 
> >Why this change? It makes it much harder to understand.
> >
> >The above is not a normal macro. Ignore any checkpatch warnings about it.
> >The proper way to do the TP_STRUCT__entry() is to make it just like a struct:
> >
> >struct {
> >	unsigned long		pfn;
> >	unsigned int		order;
> >	int			migratetype;
> >};
> >
> >Thus, the macro should be:
> >
> >	TP_STRUCT__entry(
> >		__field(	unsigned long,	pfn		)
> >		__field(	unsigned int,	order		)
> >		__field(	int,		migratetype	)
> >		),  
> 
> 
> Yeah sorry for the formatting issue, will fix in v3. Any other concerns?
> What do you think about the introduction of those tracepoints themselves?
>

It's a basic tracepoint and nothing unusual about it. I only watch over how
tracepoints are created and some use cases and make sure they are done
properly. But the introduction of tracepoints in other subsystems are up to
the maintainers of those subsystems. They are the ones that know what is
useful or not.

In other words, it's up to the MM subsystem maintainers to decide.

-- Steve