2020-04-14 13:52:33

by Shengjiu Wang

[permalink] [raw]
Subject: Re: [PATCH v7 5/7] ASoC: fsl_asrc: Move common definition to fsl_asrc_common

On Tue, Apr 14, 2020 at 10:09 AM Nicolin Chen <[email protected]> wrote:
>
> On Tue, Apr 14, 2020 at 08:43:07AM +0800, Shengjiu Wang wrote:
> > There is a new ASRC included in i.MX serial platform, there
> > are some common definition can be shared with each other.
> > So move the common definition to a separate header file.
> >
> > And add fsl_asrc_pair_priv and fsl_asrc_priv for
> > the variable specific for the module, which can be used
> > internally.
> >
> > Signed-off-by: Shengjiu Wang <[email protected]>
>
> > diff --git a/sound/soc/fsl/fsl_asrc.c b/sound/soc/fsl/fsl_asrc.c
> > +static size_t fsl_asrc_get_pair_priv_size(void)
> > +{
> > + return sizeof(struct fsl_asrc_pair_priv);
> > +}
>
> Perhaps we haven't understood completely each other's point.
>
> Yet, would the following change work?

Should work, almost same, or do you want me to use variable to
replace function pointer?

best regards
wang shengjiu


2020-04-14 13:59:59

by Nicolin Chen

[permalink] [raw]
Subject: Re: [PATCH v7 5/7] ASoC: fsl_asrc: Move common definition to fsl_asrc_common

On Tue, Apr 14, 2020 at 10:21:29AM +0800, Shengjiu Wang wrote:
> On Tue, Apr 14, 2020 at 10:09 AM Nicolin Chen <[email protected]> wrote:
> >
> > On Tue, Apr 14, 2020 at 08:43:07AM +0800, Shengjiu Wang wrote:
> > > There is a new ASRC included in i.MX serial platform, there
> > > are some common definition can be shared with each other.
> > > So move the common definition to a separate header file.
> > >
> > > And add fsl_asrc_pair_priv and fsl_asrc_priv for
> > > the variable specific for the module, which can be used
> > > internally.
> > >
> > > Signed-off-by: Shengjiu Wang <[email protected]>
> >
> > > diff --git a/sound/soc/fsl/fsl_asrc.c b/sound/soc/fsl/fsl_asrc.c
> > > +static size_t fsl_asrc_get_pair_priv_size(void)
> > > +{
> > > + return sizeof(struct fsl_asrc_pair_priv);
> > > +}
> >
> > Perhaps we haven't understood completely each other's point.
> >
> > Yet, would the following change work?
>
> Should work, almost same

Function call involving branch instruction may fail CPU's branch
prediction and hurt performance, depending on the CPU arch. If a
variable simply does the job, we should avoid doing that.

> or do you want me to use variable to replace function pointer?

Yes. And please add my ack once you change it as the reset LGTM:

Acked-by: Nicolin Chen <[email protected]>