From mboxrd@z Thu Jan 1 00:00:00 1970 From: maxi.jourdan@wanadoo.fr (Maxime Jourdan) Date: Fri, 10 Aug 2018 08:35:58 +0200 Subject: [PATCH 4/4] drm/meson: convert to the new canvas module In-Reply-To: <5b6cc316.1c69fb81.682d3.1216@mx.google.com> References: <20180801185128.23440-1-maxi.jourdan@wanadoo.fr> <20180801185128.23440-5-maxi.jourdan@wanadoo.fr> <5b6cc316.1c69fb81.682d3.1216@mx.google.com> Message-ID: To: linus-amlogic@lists.infradead.org List-Id: linus-amlogic.lists.infradead.org 2018-08-10 0:41 GMT+02:00 Rob Herring : > Hi, this is an automated email from Rob's (experimental) review bot. I > found a couple of common problems with your patch. Please see below. > > On Wed, 1 Aug 2018 20:51:28 +0200, Maxime Jourdan wrote: >> This removes the meson_canvas files within the meson/drm layer >> and makes use of the new canvas module that is referenced in the dts. >> >> Canvases can be used by different IPs and modules, and it is as such >> preferable to rely on a module that can safely dispatch canvases on >> demand. >> >> Signed-off-by: Maxime Jourdan > > The preferred subject prefix is "dt-bindings: : ...". > >> --- >> .../bindings/display/amlogic,meson-vpu.txt | 9 +-- >> arch/arm64/boot/dts/amlogic/meson-gx.dtsi | 7 +- >> drivers/gpu/drm/meson/Kconfig | 1 + >> drivers/gpu/drm/meson/Makefile | 2 +- >> drivers/gpu/drm/meson/meson_canvas.c | 70 ------------------- >> drivers/gpu/drm/meson/meson_canvas.h | 42 ----------- >> drivers/gpu/drm/meson/meson_crtc.c | 5 +- >> drivers/gpu/drm/meson/meson_drv.c | 35 ++++++---- >> drivers/gpu/drm/meson/meson_drv.h | 5 +- >> drivers/gpu/drm/meson/meson_plane.c | 3 +- >> drivers/gpu/drm/meson/meson_viu.c | 1 - >> 11 files changed, 39 insertions(+), 141 deletions(-) >> delete mode 100644 drivers/gpu/drm/meson/meson_canvas.c >> delete mode 100644 drivers/gpu/drm/meson/meson_canvas.h >> > > DT bindings (including binding headers) should be a separate patch. See > Documentation/devicetree/bindings/submitting-patches.txt. > Hi, What's the standard procedure here ? The reason I kept devicetree+drm changes together is because I didn't want to have floating commits that would break the drm module. Should I split the changes anyway ? Maxime