| Message ID | 86zhm782g5.fsf@baylibre.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | drm: meson: global clean-up (use proper macros, update comments ...) | expand |
Julien Masson <jmasson@baylibre.com> writes: > This patch series aims to clean-up differents parts of the drm meson > code source. > > Couple macros have been defined and used to set several registers > instead of using magic constants. > > I also took the opportunity to: > - add/remove/update comments > - remove useless code > - minor fix/improvment Nice set of cleanups, thanks! I especially like the extra in-code comments. Could you also add to the cover-letter how this was tested, and on what platforms so we know it's not going to introduce any regressions. Thanks, Kevin
On 25/06/2019 01:24, Kevin Hilman wrote: > Julien Masson <jmasson@baylibre.com> writes: > >> This patch series aims to clean-up differents parts of the drm meson >> code source. >> >> Couple macros have been defined and used to set several registers >> instead of using magic constants. >> >> I also took the opportunity to: >> - add/remove/update comments >> - remove useless code >> - minor fix/improvment > > Nice set of cleanups, thanks! I especially like the extra in-code > comments. > > Could you also add to the cover-letter how this was tested, and on what > platforms so we know it's not going to introduce any regressions. > > Thanks, > > Kevin > Apart the wrong magic value in patch 4 that I'll fix while applying, Reviewed-by: Neil Armstrong <narmstrong@baylibre.com> I'll run a few tests on all the supported SoC versions: - GXBB - GXL - GXM - G12A/G12B and push to drm-misc-next. Neil
Hi Julien, On 08/08/2019 16:12, Neil Armstrong wrote: > On 25/06/2019 01:24, Kevin Hilman wrote: >> Julien Masson <jmasson@baylibre.com> writes: >> >>> This patch series aims to clean-up differents parts of the drm meson >>> code source. >>> >>> Couple macros have been defined and used to set several registers >>> instead of using magic constants. >>> >>> I also took the opportunity to: >>> - add/remove/update comments >>> - remove useless code >>> - minor fix/improvment >> >> Nice set of cleanups, thanks! I especially like the extra in-code >> comments. >> >> Could you also add to the cover-letter how this was tested, and on what >> platforms so we know it's not going to introduce any regressions. >> >> Thanks, >> >> Kevin >> > > Apart the wrong magic value in patch 4 that I'll fix while applying, > Reviewed-by: Neil Armstrong <narmstrong@baylibre.com> > > I'll run a few tests on all the supported SoC versions: > - GXBB > - GXL > - GXM > - G12A/G12B Tested on : - 1024x768x60 DVI Dell monitor - 3840x2160x60 HDMI2.0 Samsung TV Tested : - Simple DMT mode - Advanced HDMI2.0 4k60 mode with SCDC negotiation and scrambling - OSD layer and scaling - Video layer and scaling SoCs: - GXBB: meson-gxbb-odroidc2 modetest -M meson -s 32:1024x768-60 -P 33@37:1024x768@AR24 -P 35@37:1024x768@NV12 modetest -M meson -s 32:1024x768-60 -P 33@37:1024x768@AR24 -P 35@37:1024x768*2@NV12 modetest -M meson -s 32:1024x768-60 -P 33@37:512x384*2@AR24 -P 35@37:2048x1536*0.5@NV12 modetest -M meson -s 32:3840x2160-60 -P 33@37:3840x2160@AR24 -P 35@37:3840x2160@NV12 modetest -M meson -s 32:3840x2160-60 -P 33@37:1920x1080*2@AR24 -P 35@37:3840x2160*0.8@NV12 No visual issues - GXL: meson-gxl-s905x-libretech-cc modetest -M meson -s 34:1024x768-60 -P 35@39:1024x768@AR24 -P 37@39:1024x768@NV12 modetest -M meson -s 34:1024x768-60 -P 35@39:1024x768@AR24 -P 37@39:1024x768*2@NV12 modetest -M meson -s 34:1024x768-60 -P 35@39:512x384*2@AR24 -P 37@39:2048x1536*0.5@NV12 modetest -M meson -s 34:3840x2160-60 -P 35@39:3840x2160@AR24 -P 37@39:3840x2160@NV12 modetest -M meson -s 34:3840x2160-60 -P 35@39:1920x1080*2@AR24 -P 37@39:3840x2160*0.8@NV12 No visual issues - GXM: meson-gxm-khadas-vim2 modetest -M meson -s 32:1024x768-60 -P 33@37:1024x768@AR24 -P 35@37:1024x768@NV12 modetest -M meson -s 32:1024x768-60 -P 33@37:1024x768@AR24 -P 35@37:1024x768*2@NV12 modetest -M meson -s 32:1024x768-60 -P 33@37:512x384*2@AR24 -P 35@37:2048x1536*0.5@NV12 modetest -M meson -s 32:3840x2160-60 -P 33@37:3840x2160@AR24 -P 35@37:3840x2160@NV12 modetest -M meson -s 32:3840x2160-60 -P 33@37:1920x1080*2@AR24 -P 35@37:3840x2160*0.8@NV12 With the following fix on patch 4: ========><==================================== --- a/drivers/gpu/drm/meson/meson_vpp.c +++ b/drivers/gpu/drm/meson/meson_vpp.c @@ -98,7 +98,7 @@ void meson_vpp_init(struct meson_drm *priv) priv->io_base + _REG(VIU_MISC_CTRL1)); writel_relaxed(VPP_PPS_DUMMY_DATA_MODE, priv->io_base + _REG(VPP_DOLBY_CTRL)); - writel_relaxed(0x108080, + writel_relaxed(0x1020080, priv->io_base + _REG(VPP_DUMMY_DATA1)); } else if (meson_vpu_is_compatible(priv, "amlogic,meson-g12a-vpu")) writel_relaxed(0xf, priv->io_base + _REG(DOLBY_PATH_CTRL)); ========><==================================== No visual issues - G12A/G12B: meson-g12a-sei510 Issue found in patch 5, the following fix will be applied while applying onto drm-misc-next : ========><==================================== --- a/drivers/gpu/drm/meson/meson_viu.c +++ b/drivers/gpu/drm/meson/meson_viu.c @@ -407,7 +407,7 @@ void meson_viu_init(struct meson_drm *priv) VIU_OSD_BLEND_HOLD_LINES(4), priv->io_base + _REG(VIU_OSD_BLEND_CTRL)); - writel_relaxed(VIU_OSD1_POSTBLD_SRC_OSD1, + writel_relaxed(OSD_BLEND_PATH_SEL_ENABLE, priv->io_base + _REG(OSD1_BLEND_SRC_CTRL)); writel_relaxed(OSD_BLEND_PATH_SEL_ENABLE, priv->io_base + _REG(OSD2_BLEND_SRC_CTRL)); ========><==================================== modetest -M meson -s 34:800x600 -P 35@39:800x600@AR24 -P 37@39:800x600@NV12 modetest -M meson -s 34:800x600 -P 35@39:800x600@AR24 -P 37@39:800x600*2@NV12 modetest -M meson -s 34:800x600 -P 35@39:400x300*2@AR24 -P 37@39:1600x600*0.5@NV12 modetest -M meson -s 34:3840x2160-60 -P 35@39:3840x2160@AR24 -P 37@39:3840x2160@NV12 modetest -M meson -s 34:3840x2160-60 -P 35@39:1920x1080*2@AR24 -P 37@39:3840x2160*0.8@NV12 No visual issues > > and push to drm-misc-next. > > Neil > Applying to drm-misc-next with the fixes Neil