On Sat, Apr 09, 2022 at 01:47:21AM +0300, Almahallawy, Khaled wrote:
> On Fri, 2022-04-08 at 20:21 +0300, Imre Deak wrote:
> > Factor out from drm_dp_dpcd_read() a function to probe a DPCD address
> > with a 1-byte read access. This will be needed by the next patch
> > doing a
> > read from an LTTPR address, which must happen without the preceding
> > wake-up read in drm_dp_dpcd_read().
> >
> > v2: Add a probe function instead of exporting drm_dp_dpcd_access().
> > (Jani)
> >
> > Cc: Jani Nikula <jani.nikula@xxxxxxxxxxxxxxx>
> > Cc: dri-devel@xxxxxxxxxxxxxxxxxxxxx
> > Signed-off-by: Imre Deak <imre.deak@xxxxxxxxx>
> > ---
> > drivers/gpu/drm/dp/drm_dp.c | 28 +++++++++++++++++++++++++---
> > include/drm/dp/drm_dp_helper.h | 1 +
> > 2 files changed, 26 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/dp/drm_dp.c
> > b/drivers/gpu/drm/dp/drm_dp.c
> > index 580016a1b9eb7..b58e30132768d 100644
> > --- a/drivers/gpu/drm/dp/drm_dp.c
> > +++ b/drivers/gpu/drm/dp/drm_dp.c
> > @@ -527,6 +527,29 @@ static int drm_dp_dpcd_access(struct drm_dp_aux
> > *aux, u8 request,
> > return ret;
> > }
> >
> > +/**
> > + * drm_dp_dpcd_probe() - probe a given DPCD address with a 1-byte read access
> > + * @aux: DisplayPort AUX channel (SST)
> > + * @offset: address of the register to probe
> > + *
> > + * Probe the provided DPCD address by reading 1 byte from it. The function can
> > + * be used to trigger some side-effect the read access has, like waking up the
> > + * sink, without the need for the read-out value.
> > + *
> > + * Returns 0 if the read access suceeded, or a negative error code on failure.
> > + */
> > +int drm_dp_dpcd_probe(struct drm_dp_aux *aux, unsigned int offset)
> > +{
> > + u8 buffer;
> > + int ret;
> > +
> > + ret = drm_dp_dpcd_access(aux, DP_AUX_NATIVE_READ, offset, &buffer, 1);
> > + WARN_ON(ret == 0);
> > +
>
> Could you add "drm_dp_dump_access" similar to drm_dp_dpcd_read/write
> in order for this aux tranaction appears in the log?
I considered that, however I'd find the log before each actual read
transfer too verbose.
However logging in case of a failure could be moved here.
>
> Thanks
> Khaled
> > + return ret < 0 ? ret : 0;
> > +}
> > +EXPORT_SYMBOL(drm_dp_dpcd_probe);
> > +
> > /**
> > * drm_dp_dpcd_read() - read a series of bytes from the DPCD
> > * @aux: DisplayPort AUX channel (SST or MST)
> > @@ -559,9 +582,8 @@ ssize_t drm_dp_dpcd_read(struct drm_dp_aux *aux,
> > unsigned int offset,
> > * monitor doesn't power down exactly after the throw away
> > read.
> > */
> > if (!aux->is_remote) {
> > - ret = drm_dp_dpcd_access(aux, DP_AUX_NATIVE_READ,
> > DP_DPCD_REV,
> > - buffer, 1);
> > - if (ret != 1)
> > + ret = drm_dp_dpcd_probe(aux, DP_DPCD_REV);
> > + if (ret < 0)
> > goto out;
> > }
> >
> > diff --git a/include/drm/dp/drm_dp_helper.h
> > b/include/drm/dp/drm_dp_helper.h
> > index 1eccd97419436..91af98e6617c6 100644
> > --- a/include/drm/dp/drm_dp_helper.h
> > +++ b/include/drm/dp/drm_dp_helper.h
> > @@ -2053,6 +2053,7 @@ struct drm_dp_aux {
> > bool is_remote;
> > };
> >
> > +int drm_dp_dpcd_probe(struct drm_dp_aux *aux, unsigned int offset);
> > ssize_t drm_dp_dpcd_read(struct drm_dp_aux *aux, unsigned int
> > offset,
> > void *buffer, size_t size);
> > ssize_t drm_dp_dpcd_write(struct drm_dp_aux *aux, unsigned int
> > offset,
