On Thu, Apr 9, 2020 at 12:47 AM Enric Balletbo i Serra <enric.balletbo@xxxxxxxxxxxxx> wrote: > > Hi Chun-Kuang > > Many thanks to let us know this, I completely missed. > > On 8/4/20 18:09, Chun-Kuang Hu wrote: > > Hi, Hsin-Yi: > > > > Enric Balletbo Serra <eballetbo@xxxxxxxxx> 於 2020年4月8日 週三 下午11:48寫道： > >> > >> Missatge de Hsin-Yi Wang <hsinyi@xxxxxxxxxxxx> del dia dl., 18 de nov. > >> 2019 a les 11:43: > >>> > >>> In order to use GCE function, we need add some informations > >>> into display node (mboxes, mediatek,gce-client-reg, mediatek,gce-events). > >>> > >>> Signed-off-by: Hsin-Yi Wang <hsinyi@xxxxxxxxxxxx> > >> > >> With the fix provided by Hsin-Yi [1] and after the mmsys patches > >> landed I get GCE function working on mt8173, so > >> > >> Tested-by: Enric Balletbo i Serra <enric.balletbo@xxxxxxxxxxxxx> > >> > >> [1] https://lore.kernel.org/lkml/20200406051131.225748-1-hsinyi@xxxxxxxxxxxx/ > >> > >> Thanks, > >> Enric > >> > >>> --- > >>> - This is based on series "support gce on mt8183 platform" > >>> https://patchwork.kernel.org/cover/11208309/ > >>> - gce setting in 8183: > >>> https://patchwork.kernel.org/patch/11127105/ > >>> --- > >>> arch/arm64/boot/dts/mediatek/mt8173.dtsi | 16 ++++++++++++++++ > >>> 1 file changed, 16 insertions(+) > >>> > >>> diff --git a/arch/arm64/boot/dts/mediatek/mt8173.dtsi b/arch/arm64/boot/dts/mediatek/mt8173.dtsi > >>> index 15f1842f6df3..e84ec3f95d81 100644 > >>> --- a/arch/arm64/boot/dts/mediatek/mt8173.dtsi > >>> +++ b/arch/arm64/boot/dts/mediatek/mt8173.dtsi > >>> @@ -911,6 +911,11 @@ mmsys: clock-controller@14000000 { > >>> assigned-clocks = <&topckgen CLK_TOP_MM_SEL>; > >>> assigned-clock-rates = <400000000>; > >>> #clock-cells = <1>; > >>> + mboxes = <&gce 0 CMDQ_THR_PRIO_HIGHEST 1>, > >>> + <&gce 1 CMDQ_THR_PRIO_HIGHEST 1>; > > > > Because of patch [1], atomic_exec parameter should be removed. > > > > [1] https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20200408&id=19d8e335d58a961a7ed377e5e2e63664b92c0b9c > > > > Many thanks to point this. Hsin-Yi, as the mt8173 device-tree will not land in > current merge window (Matthias correct me if I am wrong) maybe you could send a > new version of the device-tree including these changes? > > > >>> + mediatek,gce-client-reg = <&gce SUBSYS_1400XXXX 0 0x1000>; > >>> + mediatek,gce-events = <CMDQ_EVENT_MUTEX0_STREAM_EOF>, > >>> + <CMDQ_EVENT_MUTEX1_STREAM_EOF>; > > > > Because of patch [2], these events should be moved to mutex node. > > > > [2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.6&id=60fa8c13ab1a33b8b958efb1510ec2fd8a064bcc > > > > Hsin-Yi, note this will also impact your fixes patch [1]. > > [1] https://lore.kernel.org/lkml/20200406051131.225748-1-hsinyi@xxxxxxxxxxxx/ > > Thanks, > Enric > > > Regards, > > Chun-Kuang. > > Thanks for pointing this out. v2: https://lore.kernel.org/lkml/20200409055012.199320-1-hsinyi@xxxxxxxxxxxx/
