| Message ID | 20220408193536.RFC.4.Icea616f57331fbaa3d48c529f300c9a8ebd37fb5@changeid (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | drm/dp: Improvements for DP AUX channel | expand |
On 09/04/2022 05:36, Douglas Anderson wrote: > Let's add support for being able to read the HPD pin even if it's > hooked directly to the controller. This will allow us to get more > accurate delays also lets us take away the waiting in the AUX transfer > functions of the eDP controller drivers. > > Signed-off-by: Douglas Anderson <dianders@chromium.org> > --- > > drivers/gpu/drm/panel/panel-edp.c | 37 ++++++++++++++++++++++++++----- > 1 file changed, 31 insertions(+), 6 deletions(-) > > diff --git a/drivers/gpu/drm/panel/panel-edp.c b/drivers/gpu/drm/panel/panel-edp.c > index 1732b4f56e38..4a143eb9544b 100644 > --- a/drivers/gpu/drm/panel/panel-edp.c > +++ b/drivers/gpu/drm/panel/panel-edp.c > @@ -417,6 +417,19 @@ static int panel_edp_get_hpd_gpio(struct device *dev, struct panel_edp *p) > return 0; > } > > +static bool panel_edp_can_read_hpd(struct panel_edp *p) > +{ > + return !p->no_hpd && (p->hpd_gpio || (p->aux && p->aux->is_hpd_asserted)); > +} > + > +static bool panel_edp_read_hpd(struct panel_edp *p) > +{ > + if (p->hpd_gpio) > + return gpiod_get_value_cansleep(p->hpd_gpio); > + > + return p->aux->is_hpd_asserted(p->aux); > +} > + > static int panel_edp_prepare_once(struct panel_edp *p) > { > struct device *dev = p->base.dev; > @@ -441,13 +454,21 @@ static int panel_edp_prepare_once(struct panel_edp *p) > if (delay) > msleep(delay); > > - if (p->hpd_gpio) { > + if (panel_edp_can_read_hpd(p)) { > if (p->desc->delay.hpd_absent) > hpd_wait_us = p->desc->delay.hpd_absent * 1000UL; > else > hpd_wait_us = 2000000; > > - err = readx_poll_timeout(gpiod_get_value_cansleep, p->hpd_gpio, > + /* > + * Extra max delay, mostly to account for ps8640. ps8640 > + * is crazy and the bridge chip driver itself has over 200 ms > + * of delay if it needs to do the pm_runtime resume of the > + * bridge chip to read the HPD. > + */ > + hpd_wait_us += 3000000; I think this should come in a separate commit and ideally this should be configurable somehow. Other hosts wouldn't need such 'additional' delay. With this change removed: Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> > + > + err = readx_poll_timeout(panel_edp_read_hpd, p, > hpd_asserted, hpd_asserted, > 1000, hpd_wait_us); > if (hpd_asserted < 0) > @@ -532,18 +553,22 @@ static int panel_edp_enable(struct drm_panel *panel) > /* > * If there is a "prepare_to_enable" delay then that's supposed to be > * the delay from HPD going high until we can turn the backlight on. > - * However, we can only count this if HPD is handled by the panel > - * driver, not if it goes to a dedicated pin on the controller. > + * However, we can only count this if HPD is readable by the panel > + * driver. > + * > * If we aren't handling the HPD pin ourselves then the best we > * can do is assume that HPD went high immediately before we were > - * called (and link training took zero time). > + * called (and link training took zero time). Note that "no-hpd" > + * actually counts as handling HPD ourselves since we're doing the > + * worst case delay (in prepare) ourselves. > * > * NOTE: if we ever end up in this "if" statement then we're > * guaranteed that the panel_edp_wait() call below will do no delay. > * It already handles that case, though, so we don't need any special > * code for it. > */ > - if (p->desc->delay.prepare_to_enable && !p->hpd_gpio && !p->no_hpd) > + if (p->desc->delay.prepare_to_enable && > + !panel_edp_can_read_hpd(p) && !p->no_hpd) > delay = max(delay, p->desc->delay.prepare_to_enable); > > if (delay)
Hi, On Thu, Apr 14, 2022 at 5:51 PM Dmitry Baryshkov <dmitry.baryshkov@linaro.org> wrote: > > On 09/04/2022 05:36, Douglas Anderson wrote: > > Let's add support for being able to read the HPD pin even if it's > > hooked directly to the controller. This will allow us to get more > > accurate delays also lets us take away the waiting in the AUX transfer > > functions of the eDP controller drivers. > > > > Signed-off-by: Douglas Anderson <dianders@chromium.org> > > --- > > > > drivers/gpu/drm/panel/panel-edp.c | 37 ++++++++++++++++++++++++++----- > > 1 file changed, 31 insertions(+), 6 deletions(-) > > > > diff --git a/drivers/gpu/drm/panel/panel-edp.c b/drivers/gpu/drm/panel/panel-edp.c > > index 1732b4f56e38..4a143eb9544b 100644 > > --- a/drivers/gpu/drm/panel/panel-edp.c > > +++ b/drivers/gpu/drm/panel/panel-edp.c > > @@ -417,6 +417,19 @@ static int panel_edp_get_hpd_gpio(struct device *dev, struct panel_edp *p) > > return 0; > > } > > > > +static bool panel_edp_can_read_hpd(struct panel_edp *p) > > +{ > > + return !p->no_hpd && (p->hpd_gpio || (p->aux && p->aux->is_hpd_asserted)); > > +} > > + > > +static bool panel_edp_read_hpd(struct panel_edp *p) > > +{ > > + if (p->hpd_gpio) > > + return gpiod_get_value_cansleep(p->hpd_gpio); > > + > > + return p->aux->is_hpd_asserted(p->aux); > > +} > > + > > static int panel_edp_prepare_once(struct panel_edp *p) > > { > > struct device *dev = p->base.dev; > > @@ -441,13 +454,21 @@ static int panel_edp_prepare_once(struct panel_edp *p) > > if (delay) > > msleep(delay); > > > > - if (p->hpd_gpio) { > > + if (panel_edp_can_read_hpd(p)) { > > if (p->desc->delay.hpd_absent) > > hpd_wait_us = p->desc->delay.hpd_absent * 1000UL; > > else > > hpd_wait_us = 2000000; > > > > - err = readx_poll_timeout(gpiod_get_value_cansleep, p->hpd_gpio, > > + /* > > + * Extra max delay, mostly to account for ps8640. ps8640 > > + * is crazy and the bridge chip driver itself has over 200 ms > > + * of delay if it needs to do the pm_runtime resume of the > > + * bridge chip to read the HPD. > > + */ > > + hpd_wait_us += 3000000; > > I think this should come in a separate commit and ideally this should be > configurable somehow. Other hosts wouldn't need such 'additional' delay. > > With this change removed: > > Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> What would you think about changing the API slightly? Instead of is_hpd_asserted(), we change it to wait_hpd_asserted() and it takes a timeout in microseconds. If you pass 0 for the timeout the function is defined to behave the same as is_hpd_asserted() today--AKA a single poll of the line. -Doug
On Sat, 16 Apr 2022 at 00:17, Doug Anderson <dianders@chromium.org> wrote: > > Hi, > > On Thu, Apr 14, 2022 at 5:51 PM Dmitry Baryshkov > <dmitry.baryshkov@linaro.org> wrote: > > > > On 09/04/2022 05:36, Douglas Anderson wrote: > > > Let's add support for being able to read the HPD pin even if it's > > > hooked directly to the controller. This will allow us to get more > > > accurate delays also lets us take away the waiting in the AUX transfer > > > functions of the eDP controller drivers. > > > > > > Signed-off-by: Douglas Anderson <dianders@chromium.org> > > > --- > > > > > > drivers/gpu/drm/panel/panel-edp.c | 37 ++++++++++++++++++++++++++----- > > > 1 file changed, 31 insertions(+), 6 deletions(-) > > > > > > diff --git a/drivers/gpu/drm/panel/panel-edp.c b/drivers/gpu/drm/panel/panel-edp.c > > > index 1732b4f56e38..4a143eb9544b 100644 > > > --- a/drivers/gpu/drm/panel/panel-edp.c > > > +++ b/drivers/gpu/drm/panel/panel-edp.c > > > @@ -417,6 +417,19 @@ static int panel_edp_get_hpd_gpio(struct device *dev, struct panel_edp *p) > > > return 0; > > > } > > > > > > +static bool panel_edp_can_read_hpd(struct panel_edp *p) > > > +{ > > > + return !p->no_hpd && (p->hpd_gpio || (p->aux && p->aux->is_hpd_asserted)); > > > +} > > > + > > > +static bool panel_edp_read_hpd(struct panel_edp *p) > > > +{ > > > + if (p->hpd_gpio) > > > + return gpiod_get_value_cansleep(p->hpd_gpio); > > > + > > > + return p->aux->is_hpd_asserted(p->aux); > > > +} > > > + > > > static int panel_edp_prepare_once(struct panel_edp *p) > > > { > > > struct device *dev = p->base.dev; > > > @@ -441,13 +454,21 @@ static int panel_edp_prepare_once(struct panel_edp *p) > > > if (delay) > > > msleep(delay); > > > > > > - if (p->hpd_gpio) { > > > + if (panel_edp_can_read_hpd(p)) { > > > if (p->desc->delay.hpd_absent) > > > hpd_wait_us = p->desc->delay.hpd_absent * 1000UL; > > > else > > > hpd_wait_us = 2000000; > > > > > > - err = readx_poll_timeout(gpiod_get_value_cansleep, p->hpd_gpio, > > > + /* > > > + * Extra max delay, mostly to account for ps8640. ps8640 > > > + * is crazy and the bridge chip driver itself has over 200 ms > > > + * of delay if it needs to do the pm_runtime resume of the > > > + * bridge chip to read the HPD. > > > + */ > > > + hpd_wait_us += 3000000; > > > > I think this should come in a separate commit and ideally this should be > > configurable somehow. Other hosts wouldn't need such 'additional' delay. > > > > With this change removed: > > > > Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> > > What would you think about changing the API slightly? Instead of > is_hpd_asserted(), we change it to wait_hpd_asserted() and it takes a > timeout in microseconds. If you pass 0 for the timeout the function is > defined to behave the same as is_hpd_asserted() today--AKA a single > poll of the line. This might work. Can you check it, please? BTW: are these changes dependent on the first part of the patchset? It might be worth splitting the patchset into two parts.
Hi, On Fri, Apr 15, 2022 at 3:12 PM Dmitry Baryshkov <dmitry.baryshkov@linaro.org> wrote: > > On Sat, 16 Apr 2022 at 00:17, Doug Anderson <dianders@chromium.org> wrote: > > > > Hi, > > > > On Thu, Apr 14, 2022 at 5:51 PM Dmitry Baryshkov > > <dmitry.baryshkov@linaro.org> wrote: > > > > > > On 09/04/2022 05:36, Douglas Anderson wrote: > > > > Let's add support for being able to read the HPD pin even if it's > > > > hooked directly to the controller. This will allow us to get more > > > > accurate delays also lets us take away the waiting in the AUX transfer > > > > functions of the eDP controller drivers. > > > > > > > > Signed-off-by: Douglas Anderson <dianders@chromium.org> > > > > --- > > > > > > > > drivers/gpu/drm/panel/panel-edp.c | 37 ++++++++++++++++++++++++++----- > > > > 1 file changed, 31 insertions(+), 6 deletions(-) > > > > > > > > diff --git a/drivers/gpu/drm/panel/panel-edp.c b/drivers/gpu/drm/panel/panel-edp.c > > > > index 1732b4f56e38..4a143eb9544b 100644 > > > > --- a/drivers/gpu/drm/panel/panel-edp.c > > > > +++ b/drivers/gpu/drm/panel/panel-edp.c > > > > @@ -417,6 +417,19 @@ static int panel_edp_get_hpd_gpio(struct device *dev, struct panel_edp *p) > > > > return 0; > > > > } > > > > > > > > +static bool panel_edp_can_read_hpd(struct panel_edp *p) > > > > +{ > > > > + return !p->no_hpd && (p->hpd_gpio || (p->aux && p->aux->is_hpd_asserted)); > > > > +} > > > > + > > > > +static bool panel_edp_read_hpd(struct panel_edp *p) > > > > +{ > > > > + if (p->hpd_gpio) > > > > + return gpiod_get_value_cansleep(p->hpd_gpio); > > > > + > > > > + return p->aux->is_hpd_asserted(p->aux); > > > > +} > > > > + > > > > static int panel_edp_prepare_once(struct panel_edp *p) > > > > { > > > > struct device *dev = p->base.dev; > > > > @@ -441,13 +454,21 @@ static int panel_edp_prepare_once(struct panel_edp *p) > > > > if (delay) > > > > msleep(delay); > > > > > > > > - if (p->hpd_gpio) { > > > > + if (panel_edp_can_read_hpd(p)) { > > > > if (p->desc->delay.hpd_absent) > > > > hpd_wait_us = p->desc->delay.hpd_absent * 1000UL; > > > > else > > > > hpd_wait_us = 2000000; > > > > > > > > - err = readx_poll_timeout(gpiod_get_value_cansleep, p->hpd_gpio, > > > > + /* > > > > + * Extra max delay, mostly to account for ps8640. ps8640 > > > > + * is crazy and the bridge chip driver itself has over 200 ms > > > > + * of delay if it needs to do the pm_runtime resume of the > > > > + * bridge chip to read the HPD. > > > > + */ > > > > + hpd_wait_us += 3000000; > > > > > > I think this should come in a separate commit and ideally this should be > > > configurable somehow. Other hosts wouldn't need such 'additional' delay. > > > > > > With this change removed: > > > > > > Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> > > > > What would you think about changing the API slightly? Instead of > > is_hpd_asserted(), we change it to wait_hpd_asserted() and it takes a > > timeout in microseconds. If you pass 0 for the timeout the function is > > defined to behave the same as is_hpd_asserted() today--AKA a single > > poll of the line. > > This might work. Can you check it, please? Cool. I'll spin with this. Hopefully early next week unless my inbox blows up. ...or my main PC's SSD like happened this week. ;-) > BTW: are these changes dependent on the first part of the patchset? It > might be worth splitting the patchset into two parts. Definitely not. As per the cover letter, this is two series jammed into one. I'm happy to split them up. The 2nd half seems much less controversial. -Doug
On 16/04/2022 03:12, Doug Anderson wrote: > Hi, > > On Fri, Apr 15, 2022 at 3:12 PM Dmitry Baryshkov > <dmitry.baryshkov@linaro.org> wrote: >> >> On Sat, 16 Apr 2022 at 00:17, Doug Anderson <dianders@chromium.org> wrote: >>> >>> Hi, >>> >>> On Thu, Apr 14, 2022 at 5:51 PM Dmitry Baryshkov >>> <dmitry.baryshkov@linaro.org> wrote: >>>> >>>> On 09/04/2022 05:36, Douglas Anderson wrote: >>>>> Let's add support for being able to read the HPD pin even if it's >>>>> hooked directly to the controller. This will allow us to get more >>>>> accurate delays also lets us take away the waiting in the AUX transfer >>>>> functions of the eDP controller drivers. >>>>> >>>>> Signed-off-by: Douglas Anderson <dianders@chromium.org> >>>>> --- >>>>> >>>>> drivers/gpu/drm/panel/panel-edp.c | 37 ++++++++++++++++++++++++++----- >>>>> 1 file changed, 31 insertions(+), 6 deletions(-) >>>>> >>>>> diff --git a/drivers/gpu/drm/panel/panel-edp.c b/drivers/gpu/drm/panel/panel-edp.c >>>>> index 1732b4f56e38..4a143eb9544b 100644 >>>>> --- a/drivers/gpu/drm/panel/panel-edp.c >>>>> +++ b/drivers/gpu/drm/panel/panel-edp.c >>>>> @@ -417,6 +417,19 @@ static int panel_edp_get_hpd_gpio(struct device *dev, struct panel_edp *p) >>>>> return 0; >>>>> } >>>>> >>>>> +static bool panel_edp_can_read_hpd(struct panel_edp *p) >>>>> +{ >>>>> + return !p->no_hpd && (p->hpd_gpio || (p->aux && p->aux->is_hpd_asserted)); >>>>> +} >>>>> + >>>>> +static bool panel_edp_read_hpd(struct panel_edp *p) >>>>> +{ >>>>> + if (p->hpd_gpio) >>>>> + return gpiod_get_value_cansleep(p->hpd_gpio); >>>>> + >>>>> + return p->aux->is_hpd_asserted(p->aux); >>>>> +} >>>>> + >>>>> static int panel_edp_prepare_once(struct panel_edp *p) >>>>> { >>>>> struct device *dev = p->base.dev; >>>>> @@ -441,13 +454,21 @@ static int panel_edp_prepare_once(struct panel_edp *p) >>>>> if (delay) >>>>> msleep(delay); >>>>> >>>>> - if (p->hpd_gpio) { >>>>> + if (panel_edp_can_read_hpd(p)) { >>>>> if (p->desc->delay.hpd_absent) >>>>> hpd_wait_us = p->desc->delay.hpd_absent * 1000UL; >>>>> else >>>>> hpd_wait_us = 2000000; >>>>> >>>>> - err = readx_poll_timeout(gpiod_get_value_cansleep, p->hpd_gpio, >>>>> + /* >>>>> + * Extra max delay, mostly to account for ps8640. ps8640 >>>>> + * is crazy and the bridge chip driver itself has over 200 ms >>>>> + * of delay if it needs to do the pm_runtime resume of the >>>>> + * bridge chip to read the HPD. >>>>> + */ >>>>> + hpd_wait_us += 3000000; >>>> >>>> I think this should come in a separate commit and ideally this should be >>>> configurable somehow. Other hosts wouldn't need such 'additional' delay. >>>> >>>> With this change removed: >>>> >>>> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> >>> >>> What would you think about changing the API slightly? Instead of >>> is_hpd_asserted(), we change it to wait_hpd_asserted() and it takes a >>> timeout in microseconds. If you pass 0 for the timeout the function is >>> defined to behave the same as is_hpd_asserted() today--AKA a single >>> poll of the line. >> >> This might work. Can you check it, please? > > Cool. I'll spin with this. Hopefully early next week unless my inbox > blows up. ...or my main PC's SSD like happened this week. ;-) > > >> BTW: are these changes dependent on the first part of the patchset? It >> might be worth splitting the patchset into two parts. > > Definitely not. As per the cover letter, this is two series jammed > into one. I'm happy to split them up. The 2nd half seems much less > controversial. Great, let's get it in then. As you have time.
Hi, On Fri, Apr 15, 2022 at 5:14 PM Dmitry Baryshkov <dmitry.baryshkov@linaro.org> wrote: > > On 16/04/2022 03:12, Doug Anderson wrote: > > Hi, > > > > On Fri, Apr 15, 2022 at 3:12 PM Dmitry Baryshkov > > <dmitry.baryshkov@linaro.org> wrote: > >> > >> On Sat, 16 Apr 2022 at 00:17, Doug Anderson <dianders@chromium.org> wrote: > >>> > >>> Hi, > >>> > >>> On Thu, Apr 14, 2022 at 5:51 PM Dmitry Baryshkov > >>> <dmitry.baryshkov@linaro.org> wrote: > >>>> > >>>> On 09/04/2022 05:36, Douglas Anderson wrote: > >>>>> Let's add support for being able to read the HPD pin even if it's > >>>>> hooked directly to the controller. This will allow us to get more > >>>>> accurate delays also lets us take away the waiting in the AUX transfer > >>>>> functions of the eDP controller drivers. > >>>>> > >>>>> Signed-off-by: Douglas Anderson <dianders@chromium.org> > >>>>> --- > >>>>> > >>>>> drivers/gpu/drm/panel/panel-edp.c | 37 ++++++++++++++++++++++++++----- > >>>>> 1 file changed, 31 insertions(+), 6 deletions(-) > >>>>> > >>>>> diff --git a/drivers/gpu/drm/panel/panel-edp.c b/drivers/gpu/drm/panel/panel-edp.c > >>>>> index 1732b4f56e38..4a143eb9544b 100644 > >>>>> --- a/drivers/gpu/drm/panel/panel-edp.c > >>>>> +++ b/drivers/gpu/drm/panel/panel-edp.c > >>>>> @@ -417,6 +417,19 @@ static int panel_edp_get_hpd_gpio(struct device *dev, struct panel_edp *p) > >>>>> return 0; > >>>>> } > >>>>> > >>>>> +static bool panel_edp_can_read_hpd(struct panel_edp *p) > >>>>> +{ > >>>>> + return !p->no_hpd && (p->hpd_gpio || (p->aux && p->aux->is_hpd_asserted)); > >>>>> +} > >>>>> + > >>>>> +static bool panel_edp_read_hpd(struct panel_edp *p) > >>>>> +{ > >>>>> + if (p->hpd_gpio) > >>>>> + return gpiod_get_value_cansleep(p->hpd_gpio); > >>>>> + > >>>>> + return p->aux->is_hpd_asserted(p->aux); > >>>>> +} > >>>>> + > >>>>> static int panel_edp_prepare_once(struct panel_edp *p) > >>>>> { > >>>>> struct device *dev = p->base.dev; > >>>>> @@ -441,13 +454,21 @@ static int panel_edp_prepare_once(struct panel_edp *p) > >>>>> if (delay) > >>>>> msleep(delay); > >>>>> > >>>>> - if (p->hpd_gpio) { > >>>>> + if (panel_edp_can_read_hpd(p)) { > >>>>> if (p->desc->delay.hpd_absent) > >>>>> hpd_wait_us = p->desc->delay.hpd_absent * 1000UL; > >>>>> else > >>>>> hpd_wait_us = 2000000; > >>>>> > >>>>> - err = readx_poll_timeout(gpiod_get_value_cansleep, p->hpd_gpio, > >>>>> + /* > >>>>> + * Extra max delay, mostly to account for ps8640. ps8640 > >>>>> + * is crazy and the bridge chip driver itself has over 200 ms > >>>>> + * of delay if it needs to do the pm_runtime resume of the > >>>>> + * bridge chip to read the HPD. > >>>>> + */ > >>>>> + hpd_wait_us += 3000000; > >>>> > >>>> I think this should come in a separate commit and ideally this should be > >>>> configurable somehow. Other hosts wouldn't need such 'additional' delay. > >>>> > >>>> With this change removed: > >>>> > >>>> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> > >>> > >>> What would you think about changing the API slightly? Instead of > >>> is_hpd_asserted(), we change it to wait_hpd_asserted() and it takes a > >>> timeout in microseconds. If you pass 0 for the timeout the function is > >>> defined to behave the same as is_hpd_asserted() today--AKA a single > >>> poll of the line. > >> > >> This might work. Can you check it, please? > > > > Cool. I'll spin with this. Hopefully early next week unless my inbox > > blows up. ...or my main PC's SSD like happened this week. ;-) > > > > > >> BTW: are these changes dependent on the first part of the patchset? It > >> might be worth splitting the patchset into two parts. > > > > Definitely not. As per the cover letter, this is two series jammed > > into one. I'm happy to split them up. The 2nd half seems much less > > controversial. > > Great, let's get it in then. As you have time. Breadcrumbs: I've posted this as: https://lore.kernel.org/r/20220418171757.2282651-1-dianders@chromium.org -Doug
diff --git a/drivers/gpu/drm/panel/panel-edp.c b/drivers/gpu/drm/panel/panel-edp.c index 1732b4f56e38..4a143eb9544b 100644 --- a/drivers/gpu/drm/panel/panel-edp.c +++ b/drivers/gpu/drm/panel/panel-edp.c @@ -417,6 +417,19 @@ static int panel_edp_get_hpd_gpio(struct device *dev, struct panel_edp *p) return 0; } +static bool panel_edp_can_read_hpd(struct panel_edp *p) +{ + return !p->no_hpd && (p->hpd_gpio || (p->aux && p->aux->is_hpd_asserted)); +} + +static bool panel_edp_read_hpd(struct panel_edp *p) +{ + if (p->hpd_gpio) + return gpiod_get_value_cansleep(p->hpd_gpio); + + return p->aux->is_hpd_asserted(p->aux); +} + static int panel_edp_prepare_once(struct panel_edp *p) { struct device *dev = p->base.dev; @@ -441,13 +454,21 @@ static int panel_edp_prepare_once(struct panel_edp *p) if (delay) msleep(delay); - if (p->hpd_gpio) { + if (panel_edp_can_read_hpd(p)) { if (p->desc->delay.hpd_absent) hpd_wait_us = p->desc->delay.hpd_absent * 1000UL; else hpd_wait_us = 2000000; - err = readx_poll_timeout(gpiod_get_value_cansleep, p->hpd_gpio, + /* + * Extra max delay, mostly to account for ps8640. ps8640 + * is crazy and the bridge chip driver itself has over 200 ms + * of delay if it needs to do the pm_runtime resume of the + * bridge chip to read the HPD. + */ + hpd_wait_us += 3000000; + + err = readx_poll_timeout(panel_edp_read_hpd, p, hpd_asserted, hpd_asserted, 1000, hpd_wait_us); if (hpd_asserted < 0) @@ -532,18 +553,22 @@ static int panel_edp_enable(struct drm_panel *panel) /* * If there is a "prepare_to_enable" delay then that's supposed to be * the delay from HPD going high until we can turn the backlight on. - * However, we can only count this if HPD is handled by the panel - * driver, not if it goes to a dedicated pin on the controller. + * However, we can only count this if HPD is readable by the panel + * driver. + * * If we aren't handling the HPD pin ourselves then the best we * can do is assume that HPD went high immediately before we were - * called (and link training took zero time). + * called (and link training took zero time). Note that "no-hpd" + * actually counts as handling HPD ourselves since we're doing the + * worst case delay (in prepare) ourselves. * * NOTE: if we ever end up in this "if" statement then we're * guaranteed that the panel_edp_wait() call below will do no delay. * It already handles that case, though, so we don't need any special * code for it. */ - if (p->desc->delay.prepare_to_enable && !p->hpd_gpio && !p->no_hpd) + if (p->desc->delay.prepare_to_enable && + !panel_edp_can_read_hpd(p) && !p->no_hpd) delay = max(delay, p->desc->delay.prepare_to_enable); if (delay)
Let's add support for being able to read the HPD pin even if it's hooked directly to the controller. This will allow us to get more accurate delays also lets us take away the waiting in the AUX transfer functions of the eDP controller drivers. Signed-off-by: Douglas Anderson <dianders@chromium.org> --- drivers/gpu/drm/panel/panel-edp.c | 37 ++++++++++++++++++++++++++----- 1 file changed, 31 insertions(+), 6 deletions(-)