| Message ID | a8d8a93ddbac69e5dddd9c89a147f8cf2ad043a4.1495006350.git-series.maxime.ripard@free-electrons.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, May 17, 2017 at 3:40 PM, Maxime Ripard <maxime.ripard@free-electrons.com> wrote: > It appears that the total vertical resolution needs to be doubled when > we're not in interlaced. Make sure that is the case. > > Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com> > --- > drivers/gpu/drm/sun4i/sun4i_tcon.c | 25 +++++++++++++++++++++---- > 1 file changed, 21 insertions(+), 4 deletions(-) > > diff --git a/drivers/gpu/drm/sun4i/sun4i_tcon.c b/drivers/gpu/drm/sun4i/sun4i_tcon.c > index 62e254aedb57..a0f9a8a516c7 100644 > --- a/drivers/gpu/drm/sun4i/sun4i_tcon.c > +++ b/drivers/gpu/drm/sun4i/sun4i_tcon.c > @@ -153,7 +153,7 @@ static int sun4i_tcon_get_clk_delay(struct drm_display_mode *mode, > void sun4i_tcon0_mode_set(struct sun4i_tcon *tcon, > struct drm_display_mode *mode) > { > - unsigned int bp, hsync, vsync; > + unsigned int bp, hsync, vsync, vtotal; > u8 clk_delay; > u32 val = 0; > > @@ -192,9 +192,26 @@ void sun4i_tcon0_mode_set(struct sun4i_tcon *tcon, > DRM_DEBUG_DRIVER("Setting vertical total %d, backporch %d\n", > mode->crtc_vtotal, bp); > > + /* > + * The vertical resolution needs to be doubled in all > + * cases. We could use crtc_vtotal and always multiply by two, > + * but that leads to a rounding error in interlace when vtotal > + * is odd. > + * > + * This happens with TV's PAL for example, where vtotal will > + * be 625, crtc_vtotal 312, and thus crtc_vtotal * 2 will be > + * 624, which apparently confuses the hardware. > + * > + * To work around this, we will always use vtotal, and > + * multiply by two only if we're not in interlace. > + */ > + vtotal = mode->vtotal; > + if (!(mode->flags & DRM_MODE_FLAG_INTERLACE)) > + vtotal = vtotal * 2; > + > /* Set vertical display timings */ > regmap_write(tcon->regs, SUN4I_TCON0_BASIC2_REG, > - SUN4I_TCON0_BASIC2_V_TOTAL(mode->crtc_vtotal * 2) | > + SUN4I_TCON0_BASIC2_V_TOTAL(vtotal) | > SUN4I_TCON0_BASIC2_V_BACKPORCH(bp)); > > /* Set Hsync and Vsync length */ > @@ -279,9 +296,9 @@ void sun4i_tcon1_mode_set(struct sun4i_tcon *tcon, > /* Set vertical display timings */ > bp = mode->crtc_vtotal - mode->crtc_vsync_start; > DRM_DEBUG_DRIVER("Setting vertical total %d, backporch %d\n", > - mode->vtotal, bp); > + mode->crtc_vtotal, bp); You should duplicate the logic from channel 0 to channel 1. In fact, interlaced modes happen predominantly on channel 1, for composite and HDMI modes. ChenYu > regmap_write(tcon->regs, SUN4I_TCON1_BASIC4_REG, > - SUN4I_TCON1_BASIC4_V_TOTAL(mode->vtotal) | > + SUN4I_TCON1_BASIC4_V_TOTAL(mode->crtc_vtotal * 2) | > SUN4I_TCON1_BASIC4_V_BACKPORCH(bp)); > > /* Set Hsync and Vsync length */ > -- > git-series 0.9.1
On Wed, May 17, 2017 at 05:50:45PM +0800, Chen-Yu Tsai wrote: > On Wed, May 17, 2017 at 3:40 PM, Maxime Ripard > <maxime.ripard@free-electrons.com> wrote: > > It appears that the total vertical resolution needs to be doubled when > > we're not in interlaced. Make sure that is the case. > > > > Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com> > > --- > > drivers/gpu/drm/sun4i/sun4i_tcon.c | 25 +++++++++++++++++++++---- > > 1 file changed, 21 insertions(+), 4 deletions(-) > > > > diff --git a/drivers/gpu/drm/sun4i/sun4i_tcon.c b/drivers/gpu/drm/sun4i/sun4i_tcon.c > > index 62e254aedb57..a0f9a8a516c7 100644 > > --- a/drivers/gpu/drm/sun4i/sun4i_tcon.c > > +++ b/drivers/gpu/drm/sun4i/sun4i_tcon.c > > @@ -153,7 +153,7 @@ static int sun4i_tcon_get_clk_delay(struct drm_display_mode *mode, > > void sun4i_tcon0_mode_set(struct sun4i_tcon *tcon, > > struct drm_display_mode *mode) > > { > > - unsigned int bp, hsync, vsync; > > + unsigned int bp, hsync, vsync, vtotal; > > u8 clk_delay; > > u32 val = 0; > > > > @@ -192,9 +192,26 @@ void sun4i_tcon0_mode_set(struct sun4i_tcon *tcon, > > DRM_DEBUG_DRIVER("Setting vertical total %d, backporch %d\n", > > mode->crtc_vtotal, bp); > > > > + /* > > + * The vertical resolution needs to be doubled in all > > + * cases. We could use crtc_vtotal and always multiply by two, > > + * but that leads to a rounding error in interlace when vtotal > > + * is odd. > > + * > > + * This happens with TV's PAL for example, where vtotal will > > + * be 625, crtc_vtotal 312, and thus crtc_vtotal * 2 will be > > + * 624, which apparently confuses the hardware. > > + * > > + * To work around this, we will always use vtotal, and > > + * multiply by two only if we're not in interlace. > > + */ > > + vtotal = mode->vtotal; > > + if (!(mode->flags & DRM_MODE_FLAG_INTERLACE)) > > + vtotal = vtotal * 2; > > + > > /* Set vertical display timings */ > > regmap_write(tcon->regs, SUN4I_TCON0_BASIC2_REG, > > - SUN4I_TCON0_BASIC2_V_TOTAL(mode->crtc_vtotal * 2) | > > + SUN4I_TCON0_BASIC2_V_TOTAL(vtotal) | > > SUN4I_TCON0_BASIC2_V_BACKPORCH(bp)); > > > > /* Set Hsync and Vsync length */ > > @@ -279,9 +296,9 @@ void sun4i_tcon1_mode_set(struct sun4i_tcon *tcon, > > /* Set vertical display timings */ > > bp = mode->crtc_vtotal - mode->crtc_vsync_start; > > DRM_DEBUG_DRIVER("Setting vertical total %d, backporch %d\n", > > - mode->vtotal, bp); > > + mode->crtc_vtotal, bp); > > You should duplicate the logic from channel 0 to channel 1. > In fact, interlaced modes happen predominantly on channel 1, > for composite and HDMI modes. Gaah, you're right, this was obviously meant for the channel 1.. I'll resend. Maxime
diff --git a/drivers/gpu/drm/sun4i/sun4i_tcon.c b/drivers/gpu/drm/sun4i/sun4i_tcon.c index 62e254aedb57..a0f9a8a516c7 100644 --- a/drivers/gpu/drm/sun4i/sun4i_tcon.c +++ b/drivers/gpu/drm/sun4i/sun4i_tcon.c @@ -153,7 +153,7 @@ static int sun4i_tcon_get_clk_delay(struct drm_display_mode *mode, void sun4i_tcon0_mode_set(struct sun4i_tcon *tcon, struct drm_display_mode *mode) { - unsigned int bp, hsync, vsync; + unsigned int bp, hsync, vsync, vtotal; u8 clk_delay; u32 val = 0; @@ -192,9 +192,26 @@ void sun4i_tcon0_mode_set(struct sun4i_tcon *tcon, DRM_DEBUG_DRIVER("Setting vertical total %d, backporch %d\n", mode->crtc_vtotal, bp); + /* + * The vertical resolution needs to be doubled in all + * cases. We could use crtc_vtotal and always multiply by two, + * but that leads to a rounding error in interlace when vtotal + * is odd. + * + * This happens with TV's PAL for example, where vtotal will + * be 625, crtc_vtotal 312, and thus crtc_vtotal * 2 will be + * 624, which apparently confuses the hardware. + * + * To work around this, we will always use vtotal, and + * multiply by two only if we're not in interlace. + */ + vtotal = mode->vtotal; + if (!(mode->flags & DRM_MODE_FLAG_INTERLACE)) + vtotal = vtotal * 2; + /* Set vertical display timings */ regmap_write(tcon->regs, SUN4I_TCON0_BASIC2_REG, - SUN4I_TCON0_BASIC2_V_TOTAL(mode->crtc_vtotal * 2) | + SUN4I_TCON0_BASIC2_V_TOTAL(vtotal) | SUN4I_TCON0_BASIC2_V_BACKPORCH(bp)); /* Set Hsync and Vsync length */ @@ -279,9 +296,9 @@ void sun4i_tcon1_mode_set(struct sun4i_tcon *tcon, /* Set vertical display timings */ bp = mode->crtc_vtotal - mode->crtc_vsync_start; DRM_DEBUG_DRIVER("Setting vertical total %d, backporch %d\n", - mode->vtotal, bp); + mode->crtc_vtotal, bp); regmap_write(tcon->regs, SUN4I_TCON1_BASIC4_REG, - SUN4I_TCON1_BASIC4_V_TOTAL(mode->vtotal) | + SUN4I_TCON1_BASIC4_V_TOTAL(mode->crtc_vtotal * 2) | SUN4I_TCON1_BASIC4_V_BACKPORCH(bp)); /* Set Hsync and Vsync length */
It appears that the total vertical resolution needs to be doubled when we're not in interlaced. Make sure that is the case. Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com> --- drivers/gpu/drm/sun4i/sun4i_tcon.c | 25 +++++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-)