Re: [PATCH v5 03/13] coresight: Refactor etm4_config_timestamp_event()

[Mike Leach]

Hi Leo,

On Thu, 20 Nov 2025 at 14:25, Leo Yan <leo.yan@xxxxxxx> wrote:
>
> On Thu, Nov 20, 2025 at 02:18:21PM +0000, Coresight ML wrote:
>
> [...]
>
> > Maybe define a general macro but with extra checking:
> >
> >   #define TRCCNTCTLRn_RLDEVENT_MASK  GENMASK(15, 8)
> >
> >   #define ETM4_RS_SEL_EVENT(paired, sel) ({  \
> >       if (paired)                            \
> >           assert(!(sel & ~GENMASK(3, 0)));   \
> >       else                                   \
> >           assert(!(sel & ~GENMASK(4, 0)));   \
> >       FIELD_PREP(TRCCNTCTLRn_RLDEVENT_MASK,  \
> >                  ((paird << 7) | sel));      \
> >   })
>
> It'd be better to use BUILD_BUG_ON() instead of assert().

That is a decent option - except I would have ETM4_RS_SEL_EVENT(mask,
paired, sel) - so it can be used for every register that has the form
<REGNAME>_<EVENTFIEILD>_MASK

It is slightly vulnerable though if someone passes something that is
not 1/0 for paired - which is why I preferred the two separate macros.

Mike

-- 
Mike Leach
Principal Engineer, ARM Ltd.
Manchester Design Centre. UK