* [PATCH v3 0/1] drm/amd/display: add DCN support for ARM64 @ 2022-10-27 19:52 Ao Zhong 2022-10-27 19:52 ` [PATCH v3 1/1] " Ao Zhong 0 siblings, 1 reply; 8+ messages in thread From: Ao Zhong @ 2022-10-27 19:52 UTC (permalink / raw) To: Harry Wentland, Leo Li, Rodrigo Siqueira, Arnd Bergmann, Nathan Chancellor, Alex Deucher, Stephen Rothwell, Christian König Cc: amd-gfx, linux-arm-kernel, Ao Zhong Hello everyone, Recently I got a SBSA ARM64 workstation, and try to use it as my daily drive machine after installing an AMD RX6400 graphics card. Since the newer AMD GPUs require DCN (Display Core Next) support to work properly, DCN is not supported on ARM64 platforms. Because some code in DCN needs to use FPU, and aarch64 has no soft floating-point toolchain. Displaycore developers have started isolating code that needs to use FPU, and most of the code has been isolated. In the process of trying to enable DCN support for the ARM64 platform, I found that there are two remaining code blocks that are not isolated. Due to -mgeneral-regs-only is enabled by default to compile the kernel on the ARM64 platform, these two code blocks will cause the compilation to fail after enabling DCN support. I have submitted two patches to isolate the remaining two code blocks that need to use the FPU. Links: https://patchwork.freedesktop.org/patch/508813/ Links: https://patchwork.freedesktop.org/patch/508816/ After merging these two patches, we can enable DCN support for ARM64 platform. I removed -mgeneral-regs-only CFLAG in the dml/Makefile for the code that needs to use hardware FPU, and add a control mechanism in amdgpu_dm/dc_fpu.c for ARM Neon. Change Since V2: - Drop CONFIG_DRM_AMD_DC_DCN check when initializing the dml_rcflags for ARM64 - Drop CONFIG_ARM64 and CONFIG_DRM_AMD_DC_DCN check for the per-file cflags list entirely Change Since V1: - replace $(dml_rcflags_arm64) with $(dml_rcflags) in dml/Makefile Ao Zhong (1): drm/amd/display: add DCN support for ARM64 drivers/gpu/drm/amd/display/Kconfig | 2 +- .../gpu/drm/amd/display/amdgpu_dm/dc_fpu.c | 6 ++++++ drivers/gpu/drm/amd/display/dc/dml/Makefile | 20 +++++++++++++++---- 3 files changed, 23 insertions(+), 5 deletions(-) -- 2.37.4 _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH v3 1/1] drm/amd/display: add DCN support for ARM64 2022-10-27 19:52 [PATCH v3 0/1] drm/amd/display: add DCN support for ARM64 Ao Zhong @ 2022-10-27 19:52 ` Ao Zhong 2022-10-27 20:46 ` Arnd Bergmann 2022-10-28 15:09 ` Nathan Chancellor 0 siblings, 2 replies; 8+ messages in thread From: Ao Zhong @ 2022-10-27 19:52 UTC (permalink / raw) To: Harry Wentland, Leo Li, Rodrigo Siqueira, Arnd Bergmann, Nathan Chancellor, Alex Deucher, Stephen Rothwell, Christian König Cc: amd-gfx, linux-arm-kernel, Ao Zhong After moving all FPU code to the DML folder, we can enable DCN support for the ARM64 platform. Remove the -mgeneral-regs-only CFLAG from the code in the DML folder that needs to use hardware FPU, and add a control mechanism for ARM Neon. Signed-off-by: Ao Zhong <hacc1225@gmail.com> --- drivers/gpu/drm/amd/display/Kconfig | 2 +- .../gpu/drm/amd/display/amdgpu_dm/dc_fpu.c | 6 ++++++ drivers/gpu/drm/amd/display/dc/dml/Makefile | 20 +++++++++++++++---- 3 files changed, 23 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig index 0142affcdaa3..a7f1c4e51719 100644 --- a/drivers/gpu/drm/amd/display/Kconfig +++ b/drivers/gpu/drm/amd/display/Kconfig @@ -6,7 +6,7 @@ config DRM_AMD_DC bool "AMD DC - Enable new display engine" default y select SND_HDA_COMPONENT if SND_HDA_CORE - select DRM_AMD_DC_DCN if (X86 || PPC64) + select DRM_AMD_DC_DCN if (X86 || PPC64 || (ARM64 && KERNEL_MODE_NEON)) help Choose this option if you want to use the new display engine support for AMDGPU. This adds required support for Vega and diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c index ab0c6d191038..1743ca0a3641 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c @@ -31,6 +31,8 @@ #elif defined(CONFIG_PPC64) #include <asm/switch_to.h> #include <asm/cputable.h> +#elif defined(CONFIG_ARM64) +#include <asm/neon.h> #endif /** @@ -99,6 +101,8 @@ void dc_fpu_begin(const char *function_name, const int line) preempt_disable(); enable_kernel_fp(); } +#elif defined(CONFIG_ARM64) + kernel_neon_begin(); #endif } @@ -136,6 +140,8 @@ void dc_fpu_end(const char *function_name, const int line) disable_kernel_fp(); preempt_enable(); } +#elif defined(CONFIG_ARM64) + kernel_neon_end(); #endif } diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile index d0c6cf61c676..d4e93bed1c8e 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile @@ -33,6 +33,10 @@ ifdef CONFIG_PPC64 dml_ccflags := -mhard-float -maltivec endif +ifdef CONFIG_ARM64 +dml_rcflags := -mgeneral-regs-only +endif + ifdef CONFIG_CC_IS_GCC ifeq ($(call cc-ifversion, -lt, 0701, y), y) IS_OLD_GCC = 1 @@ -55,8 +59,6 @@ frame_warn_flag := -Wframe-larger-than=2048 endif CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_ccflags) - -ifdef CONFIG_DRM_AMD_DC_DCN CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn10/dcn10_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn20/dcn20_fpu.o := $(dml_ccflags) @@ -88,7 +90,6 @@ CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calcs.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_auto.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_ccflags) -Wno-tautological-compare CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_rcflags) -CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn2x/dcn2x.o := $(dml_rcflags) CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_mode_vba_20.o := $(dml_rcflags) CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_rq_dlg_calc_20.o := $(dml_rcflags) CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_mode_vba_20v2.o := $(dml_rcflags) @@ -105,7 +106,18 @@ CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_util_32.o := $(dml_rcf CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_rcflags) CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_rcflags) CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dsc/rc_calc_fpu.o := $(dml_rcflags) -endif +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn10/dcn10_fpu.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/dcn20_fpu.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/display_mode_vba_314.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/display_rq_dlg_calc_314.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/dcn314_fpu.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn30/dcn30_fpu.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn32/dcn32_fpu.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn321/dcn321_fpu.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn31/dcn31_fpu.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn302/dcn302_fpu.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn303/dcn303_fpu.o := $(dml_rcflags) +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_rcflags) CFLAGS_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/display_rq_dlg_helpers.o := $(dml_ccflags) CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_rcflags) -- 2.37.4 _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH v3 1/1] drm/amd/display: add DCN support for ARM64 2022-10-27 19:52 ` [PATCH v3 1/1] " Ao Zhong @ 2022-10-27 20:46 ` Arnd Bergmann 2022-10-28 15:09 ` Nathan Chancellor 1 sibling, 0 replies; 8+ messages in thread From: Arnd Bergmann @ 2022-10-27 20:46 UTC (permalink / raw) To: Ao Zhong, Harry Wentland, Leo Li, Rodrigo Siqueira, Nathan Chancellor, Alex Deucher, Stephen Rothwell, Christian König Cc: amd-gfx, linux-arm-kernel On Thu, Oct 27, 2022, at 21:52, Ao Zhong wrote: > After moving all FPU code to the DML folder, we can enable DCN support > for the ARM64 platform. Remove the -mgeneral-regs-only CFLAG from the > code in the DML folder that needs to use hardware FPU, and add a control > mechanism for ARM Neon. > > Signed-off-by: Ao Zhong <hacc1225@gmail.com> Looks good to me, Acked-by: Arnd Bergmann <arnd@arndb.de> _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH v3 1/1] drm/amd/display: add DCN support for ARM64 2022-10-27 19:52 ` [PATCH v3 1/1] " Ao Zhong 2022-10-27 20:46 ` Arnd Bergmann @ 2022-10-28 15:09 ` Nathan Chancellor 2022-10-28 15:28 ` Ao Zhong 2022-10-28 15:35 ` Rodrigo Siqueira Jordao 1 sibling, 2 replies; 8+ messages in thread From: Nathan Chancellor @ 2022-10-28 15:09 UTC (permalink / raw) To: Ao Zhong Cc: Harry Wentland, Leo Li, Rodrigo Siqueira, Arnd Bergmann, Alex Deucher, Stephen Rothwell, Christian König, amd-gfx, linux-arm-kernel Hi Ao, On Thu, Oct 27, 2022 at 09:52:29PM +0200, Ao Zhong wrote: > After moving all FPU code to the DML folder, we can enable DCN support > for the ARM64 platform. Remove the -mgeneral-regs-only CFLAG from the > code in the DML folder that needs to use hardware FPU, and add a control > mechanism for ARM Neon. > > Signed-off-by: Ao Zhong <hacc1225@gmail.com> > --- > drivers/gpu/drm/amd/display/Kconfig | 2 +- > .../gpu/drm/amd/display/amdgpu_dm/dc_fpu.c | 6 ++++++ > drivers/gpu/drm/amd/display/dc/dml/Makefile | 20 +++++++++++++++---- > 3 files changed, 23 insertions(+), 5 deletions(-) > > diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig > index 0142affcdaa3..a7f1c4e51719 100644 > --- a/drivers/gpu/drm/amd/display/Kconfig > +++ b/drivers/gpu/drm/amd/display/Kconfig > @@ -6,7 +6,7 @@ config DRM_AMD_DC > bool "AMD DC - Enable new display engine" > default y > select SND_HDA_COMPONENT if SND_HDA_CORE > - select DRM_AMD_DC_DCN if (X86 || PPC64) > + select DRM_AMD_DC_DCN if (X86 || PPC64 || (ARM64 && KERNEL_MODE_NEON)) > help > Choose this option if you want to use the new display engine > support for AMDGPU. This adds required support for Vega and > diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c > index ab0c6d191038..1743ca0a3641 100644 > --- a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c > +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c > @@ -31,6 +31,8 @@ > #elif defined(CONFIG_PPC64) > #include <asm/switch_to.h> > #include <asm/cputable.h> > +#elif defined(CONFIG_ARM64) > +#include <asm/neon.h> > #endif > > /** > @@ -99,6 +101,8 @@ void dc_fpu_begin(const char *function_name, const int line) > preempt_disable(); > enable_kernel_fp(); > } > +#elif defined(CONFIG_ARM64) > + kernel_neon_begin(); > #endif > } > > @@ -136,6 +140,8 @@ void dc_fpu_end(const char *function_name, const int line) > disable_kernel_fp(); > preempt_enable(); > } > +#elif defined(CONFIG_ARM64) > + kernel_neon_end(); > #endif > } > > diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile > index d0c6cf61c676..d4e93bed1c8e 100644 > --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile > +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile > @@ -33,6 +33,10 @@ ifdef CONFIG_PPC64 > dml_ccflags := -mhard-float -maltivec > endif > > +ifdef CONFIG_ARM64 > +dml_rcflags := -mgeneral-regs-only > +endif > + > ifdef CONFIG_CC_IS_GCC > ifeq ($(call cc-ifversion, -lt, 0701, y), y) > IS_OLD_GCC = 1 > @@ -55,8 +59,6 @@ frame_warn_flag := -Wframe-larger-than=2048 > endif > > CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_ccflags) > - > -ifdef CONFIG_DRM_AMD_DC_DCN > CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_ccflags) > CFLAGS_$(AMDDALPATH)/dc/dml/dcn10/dcn10_fpu.o := $(dml_ccflags) > CFLAGS_$(AMDDALPATH)/dc/dml/dcn20/dcn20_fpu.o := $(dml_ccflags) > @@ -88,7 +90,6 @@ CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calcs.o := $(dml_ccflags) > CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_auto.o := $(dml_ccflags) > CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_ccflags) -Wno-tautological-compare > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_rcflags) > -CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn2x/dcn2x.o := $(dml_rcflags) > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_mode_vba_20.o := $(dml_rcflags) > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_rq_dlg_calc_20.o := $(dml_rcflags) > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_mode_vba_20v2.o := $(dml_rcflags) > @@ -105,7 +106,18 @@ CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_util_32.o := $(dml_rcf > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_rcflags) > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_rcflags) > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dsc/rc_calc_fpu.o := $(dml_rcflags) > -endif > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn10/dcn10_fpu.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/dcn20_fpu.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/display_mode_vba_314.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/display_rq_dlg_calc_314.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/dcn314_fpu.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn30/dcn30_fpu.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn32/dcn32_fpu.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn321/dcn321_fpu.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn31/dcn31_fpu.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn302/dcn302_fpu.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn303/dcn303_fpu.o := $(dml_rcflags) > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_rcflags) > CFLAGS_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_ccflags) > CFLAGS_$(AMDDALPATH)/dc/dml/display_rq_dlg_helpers.o := $(dml_ccflags) > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_rcflags) > -- > 2.37.4 > Enabling this code for arm64 reveals the following warnings when building allmodconfig with clang. A very recent change in LLVM added the variable and spill information so that will not be seen with older versions but I also saw the warnings with clang 15.0.0 from Fedora. drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn30/display_mode_vba_30.c:1852:13: error: stack frame size (2112) exceeds limit (2048) in 'DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation' [-Werror,-Wframe-larger-than] static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation( ^ 1152/2112 (54.55%) spills, 960/2112 (45.45%) variables drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn30/display_mode_vba_30.c:3542:6: error: stack frame size (2240) exceeds limit (2048) in 'dml30_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] void dml30_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) ^ 1451/2240 (64.78%) spills, 789/2240 (35.22%) variables drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn31/display_mode_vba_31.c:3791:6: error: stack frame size (2736) exceeds limit (2048) in 'dml31_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] void dml31_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) ^ 1342/2736 (49.05%) spills, 1394/2736 (50.95%) variables drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn314/display_mode_vba_314.c:3890:6: error: stack frame size (2720) exceeds limit (2048) in 'dml314_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] void dml314_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) ^ 1325/2720 (48.71%) spills, 1395/2720 (51.29%) variables drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/display_mode_vba_32.c:59:13: error: stack frame size (2208) exceeds limit (2048) in 'DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation' [-Werror,-Wframe-larger-than] static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation( ^ 1397/2208 (63.27%) spills, 811/2208 (36.73%) variables drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/display_mode_vba_32.c:1681:6: error: stack frame size (2496) exceeds limit (2048) in 'dml32_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) ^ 1504/2496 (60.26%) spills, 992/2496 (39.74%) variables I can try to take a look at these next week, as I have prior obligations today, unless others wanted to help me out ;) Additionally, I see the following errors with GCC 12.2.1 from Fedora when building allmodconfig. Seems like some $(dml_rcflags) might be missing. drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c: In function ‘dcn10_resource_construct_fp’: drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1313:52: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types 1313 | dcn_soc->dram_clock_change_latency = 23; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~ drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1317:45: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types 1317 | dc->dcn_soc->urgent_latency = 3; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1319:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types 1319 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 41.60f; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~ drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1329:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types 1329 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 19.2f; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~ drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1330:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types 1330 | dc->dcn_soc->fabric_and_dram_bandwidth_vnom0p8 = 17.066f; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~ drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1331:65: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types 1331 | dc->dcn_soc->fabric_and_dram_bandwidth_vmid0p72 = 14.933f; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~ drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1332:65: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types 1332 | dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 = 12.8f; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~ drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1334:72: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types 1334 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 20.80f; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~ make[6]: *** [scripts/Makefile.build:250: drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.o] Error 1 drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c: In function ‘dcn32_populate_dml_pipes_from_context’: drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c:1921:70: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types 1921 | pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_luma = 0; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c:1922:72: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types 1922 | pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_chroma = 0; | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ make[6]: *** [scripts/Makefile.build:250: drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.o] Error 1 Cheers, Nathan _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH v3 1/1] drm/amd/display: add DCN support for ARM64 2022-10-28 15:09 ` Nathan Chancellor @ 2022-10-28 15:28 ` Ao Zhong 2022-10-28 15:35 ` Rodrigo Siqueira Jordao 1 sibling, 0 replies; 8+ messages in thread From: Ao Zhong @ 2022-10-28 15:28 UTC (permalink / raw) To: Nathan Chancellor Cc: Harry Wentland, Leo Li, Rodrigo Siqueira, Arnd Bergmann, Alex Deucher, Stephen Rothwell, Christian König, amd-gfx, linux-arm-kernel Am 28.10.22 um 17:09 schrieb Nathan Chancellor: > Additionally, I see the following errors with GCC 12.2.1 from Fedora > when building allmodconfig. Seems like some $(dml_rcflags) might be > missing. > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c: In function ‘dcn10_resource_construct_fp’: > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1313:52: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1313 | dcn_soc->dram_clock_change_latency = 23; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1317:45: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1317 | dc->dcn_soc->urgent_latency = 3; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1319:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1319 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 41.60f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1329:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1329 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 19.2f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1330:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1330 | dc->dcn_soc->fabric_and_dram_bandwidth_vnom0p8 = 17.066f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1331:65: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1331 | dc->dcn_soc->fabric_and_dram_bandwidth_vmid0p72 = 14.933f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1332:65: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1332 | dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 = 12.8f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1334:72: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1334 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 20.80f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~ > make[6]: *** [scripts/Makefile.build:250: drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.o] Error 1 > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c: In function ‘dcn32_populate_dml_pipes_from_context’: > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c:1921:70: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1921 | pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_luma = 0; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c:1922:72: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1922 | pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_chroma = 0; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ > make[6]: *** [scripts/Makefile.build:250: drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.o] Error 1 Hi Nathan, it seems like you haven't merged my patch to isolate the remaining FPU code. Links: https://patchwork.freedesktop.org/patch/508813/ Links: https://patchwork.freedesktop.org/patch/508816/ Rodrigo says both patches have been applied, but I haven't seen them in the amd-staging-drm-next branch. This error occurs because the remaining FPU code in dcn10_resource.c has not been isolated. Ao _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH v3 1/1] drm/amd/display: add DCN support for ARM64 2022-10-28 15:09 ` Nathan Chancellor 2022-10-28 15:28 ` Ao Zhong @ 2022-10-28 15:35 ` Rodrigo Siqueira Jordao 2022-10-28 15:48 ` Nathan Chancellor 1 sibling, 1 reply; 8+ messages in thread From: Rodrigo Siqueira Jordao @ 2022-10-28 15:35 UTC (permalink / raw) To: Nathan Chancellor, Ao Zhong Cc: Harry Wentland, Leo Li, Arnd Bergmann, Alex Deucher, Stephen Rothwell, Christian König, amd-gfx, linux-arm-kernel On 2022-10-28 11:09, Nathan Chancellor wrote: > Hi Ao, > > On Thu, Oct 27, 2022 at 09:52:29PM +0200, Ao Zhong wrote: >> After moving all FPU code to the DML folder, we can enable DCN support >> for the ARM64 platform. Remove the -mgeneral-regs-only CFLAG from the >> code in the DML folder that needs to use hardware FPU, and add a control >> mechanism for ARM Neon. >> >> Signed-off-by: Ao Zhong <hacc1225@gmail.com> >> --- >> drivers/gpu/drm/amd/display/Kconfig | 2 +- >> .../gpu/drm/amd/display/amdgpu_dm/dc_fpu.c | 6 ++++++ >> drivers/gpu/drm/amd/display/dc/dml/Makefile | 20 +++++++++++++++---- >> 3 files changed, 23 insertions(+), 5 deletions(-) >> >> diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig >> index 0142affcdaa3..a7f1c4e51719 100644 >> --- a/drivers/gpu/drm/amd/display/Kconfig >> +++ b/drivers/gpu/drm/amd/display/Kconfig >> @@ -6,7 +6,7 @@ config DRM_AMD_DC >> bool "AMD DC - Enable new display engine" >> default y >> select SND_HDA_COMPONENT if SND_HDA_CORE >> - select DRM_AMD_DC_DCN if (X86 || PPC64) >> + select DRM_AMD_DC_DCN if (X86 || PPC64 || (ARM64 && KERNEL_MODE_NEON)) >> help >> Choose this option if you want to use the new display engine >> support for AMDGPU. This adds required support for Vega and >> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c >> index ab0c6d191038..1743ca0a3641 100644 >> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c >> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c >> @@ -31,6 +31,8 @@ >> #elif defined(CONFIG_PPC64) >> #include <asm/switch_to.h> >> #include <asm/cputable.h> >> +#elif defined(CONFIG_ARM64) >> +#include <asm/neon.h> >> #endif >> >> /** >> @@ -99,6 +101,8 @@ void dc_fpu_begin(const char *function_name, const int line) >> preempt_disable(); >> enable_kernel_fp(); >> } >> +#elif defined(CONFIG_ARM64) >> + kernel_neon_begin(); >> #endif >> } >> >> @@ -136,6 +140,8 @@ void dc_fpu_end(const char *function_name, const int line) >> disable_kernel_fp(); >> preempt_enable(); >> } >> +#elif defined(CONFIG_ARM64) >> + kernel_neon_end(); >> #endif >> } >> >> diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile >> index d0c6cf61c676..d4e93bed1c8e 100644 >> --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile >> +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile >> @@ -33,6 +33,10 @@ ifdef CONFIG_PPC64 >> dml_ccflags := -mhard-float -maltivec >> endif >> >> +ifdef CONFIG_ARM64 >> +dml_rcflags := -mgeneral-regs-only >> +endif >> + >> ifdef CONFIG_CC_IS_GCC >> ifeq ($(call cc-ifversion, -lt, 0701, y), y) >> IS_OLD_GCC = 1 >> @@ -55,8 +59,6 @@ frame_warn_flag := -Wframe-larger-than=2048 >> endif >> >> CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_ccflags) >> - >> -ifdef CONFIG_DRM_AMD_DC_DCN >> CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_ccflags) >> CFLAGS_$(AMDDALPATH)/dc/dml/dcn10/dcn10_fpu.o := $(dml_ccflags) >> CFLAGS_$(AMDDALPATH)/dc/dml/dcn20/dcn20_fpu.o := $(dml_ccflags) >> @@ -88,7 +90,6 @@ CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calcs.o := $(dml_ccflags) >> CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_auto.o := $(dml_ccflags) >> CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_ccflags) -Wno-tautological-compare >> CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_rcflags) >> -CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn2x/dcn2x.o := $(dml_rcflags) >> CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_mode_vba_20.o := $(dml_rcflags) >> CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_rq_dlg_calc_20.o := $(dml_rcflags) >> CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_mode_vba_20v2.o := $(dml_rcflags) >> @@ -105,7 +106,18 @@ CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_util_32.o := $(dml_rcf >> CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_rcflags) >> CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_rcflags) >> CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dsc/rc_calc_fpu.o := $(dml_rcflags) >> -endif >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn10/dcn10_fpu.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/dcn20_fpu.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/display_mode_vba_314.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/display_rq_dlg_calc_314.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/dcn314_fpu.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn30/dcn30_fpu.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn32/dcn32_fpu.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn321/dcn321_fpu.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn31/dcn31_fpu.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn302/dcn302_fpu.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn303/dcn303_fpu.o := $(dml_rcflags) >> +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_rcflags) >> CFLAGS_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_ccflags) >> CFLAGS_$(AMDDALPATH)/dc/dml/display_rq_dlg_helpers.o := $(dml_ccflags) >> CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_rcflags) >> -- >> 2.37.4 >> > > Enabling this code for arm64 reveals the following warnings when > building allmodconfig with clang. A very recent change in LLVM added the > variable and spill information so that will not be seen with older > versions but I also saw the warnings with clang 15.0.0 from Fedora. > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn30/display_mode_vba_30.c:1852:13: error: stack frame size (2112) exceeds limit (2048) in 'DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation' [-Werror,-Wframe-larger-than] > static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation( > ^ > 1152/2112 (54.55%) spills, 960/2112 (45.45%) variables > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn30/display_mode_vba_30.c:3542:6: error: stack frame size (2240) exceeds limit (2048) in 'dml30_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > void dml30_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > ^ > 1451/2240 (64.78%) spills, 789/2240 (35.22%) variables > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn31/display_mode_vba_31.c:3791:6: error: stack frame size (2736) exceeds limit (2048) in 'dml31_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > void dml31_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > ^ > 1342/2736 (49.05%) spills, 1394/2736 (50.95%) variables > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn314/display_mode_vba_314.c:3890:6: error: stack frame size (2720) exceeds limit (2048) in 'dml314_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > void dml314_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > ^ > 1325/2720 (48.71%) spills, 1395/2720 (51.29%) variables > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/display_mode_vba_32.c:59:13: error: stack frame size (2208) exceeds limit (2048) in 'DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation' [-Werror,-Wframe-larger-than] > static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation( > ^ > 1397/2208 (63.27%) spills, 811/2208 (36.73%) variables > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/display_mode_vba_32.c:1681:6: error: stack frame size (2496) exceeds limit (2048) in 'dml32_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > ^ > 1504/2496 (60.26%) spills, 992/2496 (39.74%) variables > > I can try to take a look at these next week, as I have prior obligations > today, unless others wanted to help me out ;) Hi Nathan, Thanks a lot for checking this. About those stack frame issues, I guess they are generic, right? I mean, I don't see why we only see it if we enable arm64. > Additionally, I see the following errors with GCC 12.2.1 from Fedora > when building allmodconfig. Seems like some $(dml_rcflags) might be > missing. > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c: In function ‘dcn10_resource_construct_fp’: > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1313:52: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1313 | dcn_soc->dram_clock_change_latency = 23; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1317:45: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1317 | dc->dcn_soc->urgent_latency = 3; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1319:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1319 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 41.60f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1329:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1329 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 19.2f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1330:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1330 | dc->dcn_soc->fabric_and_dram_bandwidth_vnom0p8 = 17.066f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1331:65: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1331 | dc->dcn_soc->fabric_and_dram_bandwidth_vmid0p72 = 14.933f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1332:65: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1332 | dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 = 12.8f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1334:72: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1334 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 20.80f; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~ > make[6]: *** [scripts/Makefile.build:250: drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.o] Error 1 > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c: In function ‘dcn32_populate_dml_pipes_from_context’: > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c:1921:70: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1921 | pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_luma = 0; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c:1922:72: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > 1922 | pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_chroma = 0; > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ > make[6]: *** [scripts/Makefile.build:250: drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.o] Error 1 Could you check if you have this commit in your branch? drm/amd/display: move remaining FPU code to dml folder iirc, I merged it around Tue: https://gitlab.freedesktop.org/agd5f/linux/-/commits/amd-staging-drm-next Thanks Siqueira > > Cheers, > Nathan _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH v3 1/1] drm/amd/display: add DCN support for ARM64 2022-10-28 15:35 ` Rodrigo Siqueira Jordao @ 2022-10-28 15:48 ` Nathan Chancellor 2022-10-31 18:47 ` Nathan Chancellor 0 siblings, 1 reply; 8+ messages in thread From: Nathan Chancellor @ 2022-10-28 15:48 UTC (permalink / raw) To: Rodrigo Siqueira Jordao Cc: Ao Zhong, Harry Wentland, Leo Li, Arnd Bergmann, Alex Deucher, Stephen Rothwell, Christian König, amd-gfx, linux-arm-kernel Hi Rodrigo, On Fri, Oct 28, 2022 at 11:35:32AM -0400, Rodrigo Siqueira Jordao wrote: > > > On 2022-10-28 11:09, Nathan Chancellor wrote: > > Hi Ao, > > > > On Thu, Oct 27, 2022 at 09:52:29PM +0200, Ao Zhong wrote: > > > After moving all FPU code to the DML folder, we can enable DCN support > > > for the ARM64 platform. Remove the -mgeneral-regs-only CFLAG from the > > > code in the DML folder that needs to use hardware FPU, and add a control > > > mechanism for ARM Neon. > > > > > > Signed-off-by: Ao Zhong <hacc1225@gmail.com> > > > --- > > > drivers/gpu/drm/amd/display/Kconfig | 2 +- > > > .../gpu/drm/amd/display/amdgpu_dm/dc_fpu.c | 6 ++++++ > > > drivers/gpu/drm/amd/display/dc/dml/Makefile | 20 +++++++++++++++---- > > > 3 files changed, 23 insertions(+), 5 deletions(-) > > > > > > diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig > > > index 0142affcdaa3..a7f1c4e51719 100644 > > > --- a/drivers/gpu/drm/amd/display/Kconfig > > > +++ b/drivers/gpu/drm/amd/display/Kconfig > > > @@ -6,7 +6,7 @@ config DRM_AMD_DC > > > bool "AMD DC - Enable new display engine" > > > default y > > > select SND_HDA_COMPONENT if SND_HDA_CORE > > > - select DRM_AMD_DC_DCN if (X86 || PPC64) > > > + select DRM_AMD_DC_DCN if (X86 || PPC64 || (ARM64 && KERNEL_MODE_NEON)) > > > help > > > Choose this option if you want to use the new display engine > > > support for AMDGPU. This adds required support for Vega and > > > diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c > > > index ab0c6d191038..1743ca0a3641 100644 > > > --- a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c > > > +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c > > > @@ -31,6 +31,8 @@ > > > #elif defined(CONFIG_PPC64) > > > #include <asm/switch_to.h> > > > #include <asm/cputable.h> > > > +#elif defined(CONFIG_ARM64) > > > +#include <asm/neon.h> > > > #endif > > > /** > > > @@ -99,6 +101,8 @@ void dc_fpu_begin(const char *function_name, const int line) > > > preempt_disable(); > > > enable_kernel_fp(); > > > } > > > +#elif defined(CONFIG_ARM64) > > > + kernel_neon_begin(); > > > #endif > > > } > > > @@ -136,6 +140,8 @@ void dc_fpu_end(const char *function_name, const int line) > > > disable_kernel_fp(); > > > preempt_enable(); > > > } > > > +#elif defined(CONFIG_ARM64) > > > + kernel_neon_end(); > > > #endif > > > } > > > diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile > > > index d0c6cf61c676..d4e93bed1c8e 100644 > > > --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile > > > +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile > > > @@ -33,6 +33,10 @@ ifdef CONFIG_PPC64 > > > dml_ccflags := -mhard-float -maltivec > > > endif > > > +ifdef CONFIG_ARM64 > > > +dml_rcflags := -mgeneral-regs-only > > > +endif > > > + > > > ifdef CONFIG_CC_IS_GCC > > > ifeq ($(call cc-ifversion, -lt, 0701, y), y) > > > IS_OLD_GCC = 1 > > > @@ -55,8 +59,6 @@ frame_warn_flag := -Wframe-larger-than=2048 > > > endif > > > CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_ccflags) > > > - > > > -ifdef CONFIG_DRM_AMD_DC_DCN > > > CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_ccflags) > > > CFLAGS_$(AMDDALPATH)/dc/dml/dcn10/dcn10_fpu.o := $(dml_ccflags) > > > CFLAGS_$(AMDDALPATH)/dc/dml/dcn20/dcn20_fpu.o := $(dml_ccflags) > > > @@ -88,7 +90,6 @@ CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calcs.o := $(dml_ccflags) > > > CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_auto.o := $(dml_ccflags) > > > CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_ccflags) -Wno-tautological-compare > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_rcflags) > > > -CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn2x/dcn2x.o := $(dml_rcflags) > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_mode_vba_20.o := $(dml_rcflags) > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_rq_dlg_calc_20.o := $(dml_rcflags) > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_mode_vba_20v2.o := $(dml_rcflags) > > > @@ -105,7 +106,18 @@ CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_util_32.o := $(dml_rcf > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_rcflags) > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_rcflags) > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dsc/rc_calc_fpu.o := $(dml_rcflags) > > > -endif > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn10/dcn10_fpu.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/dcn20_fpu.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/display_mode_vba_314.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/display_rq_dlg_calc_314.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/dcn314_fpu.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn30/dcn30_fpu.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn32/dcn32_fpu.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn321/dcn321_fpu.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn31/dcn31_fpu.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn302/dcn302_fpu.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn303/dcn303_fpu.o := $(dml_rcflags) > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_rcflags) > > > CFLAGS_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_ccflags) > > > CFLAGS_$(AMDDALPATH)/dc/dml/display_rq_dlg_helpers.o := $(dml_ccflags) > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_rcflags) > > > -- > > > 2.37.4 > > > > > > > Enabling this code for arm64 reveals the following warnings when > > building allmodconfig with clang. A very recent change in LLVM added the > > variable and spill information so that will not be seen with older > > versions but I also saw the warnings with clang 15.0.0 from Fedora. > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn30/display_mode_vba_30.c:1852:13: error: stack frame size (2112) exceeds limit (2048) in 'DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation' [-Werror,-Wframe-larger-than] > > static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation( > > ^ > > 1152/2112 (54.55%) spills, 960/2112 (45.45%) variables > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn30/display_mode_vba_30.c:3542:6: error: stack frame size (2240) exceeds limit (2048) in 'dml30_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > > void dml30_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > > ^ > > 1451/2240 (64.78%) spills, 789/2240 (35.22%) variables > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn31/display_mode_vba_31.c:3791:6: error: stack frame size (2736) exceeds limit (2048) in 'dml31_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > > void dml31_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > > ^ > > 1342/2736 (49.05%) spills, 1394/2736 (50.95%) variables > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn314/display_mode_vba_314.c:3890:6: error: stack frame size (2720) exceeds limit (2048) in 'dml314_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > > void dml314_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > > ^ > > 1325/2720 (48.71%) spills, 1395/2720 (51.29%) variables > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/display_mode_vba_32.c:59:13: error: stack frame size (2208) exceeds limit (2048) in 'DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation' [-Werror,-Wframe-larger-than] > > static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation( > > ^ > > 1397/2208 (63.27%) spills, 811/2208 (36.73%) variables > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/display_mode_vba_32.c:1681:6: error: stack frame size (2496) exceeds limit (2048) in 'dml32_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > > void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > > ^ > > 1504/2496 (60.26%) spills, 992/2496 (39.74%) variables > > > > I can try to take a look at these next week, as I have prior obligations > > today, unless others wanted to help me out ;) > > Hi Nathan, > > Thanks a lot for checking this. > > About those stack frame issues, I guess they are generic, right? I mean, I > don't see why we only see it if we enable arm64. I do not see these warnings on x86_64 as I had to tackle them for 6.0. powerpc allmodconfig is already broken with clang for other reasons so I am not sure if these warnings are there; I can try to test that. I am not sure why things are going sideways for arm64 with clang here but I would really like to keep allmodconfig building without any issues, as that is a good target for catching new warnings due to -Werror. > > Additionally, I see the following errors with GCC 12.2.1 from Fedora > > when building allmodconfig. Seems like some $(dml_rcflags) might be > > missing. > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c: In function ‘dcn10_resource_construct_fp’: > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1313:52: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > > 1313 | dcn_soc->dram_clock_change_latency = 23; > > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~ > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1317:45: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > > 1317 | dc->dcn_soc->urgent_latency = 3; > > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1319:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > > 1319 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 41.60f; > > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~ > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1329:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > > 1329 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 19.2f; > > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~ > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1330:64: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > > 1330 | dc->dcn_soc->fabric_and_dram_bandwidth_vnom0p8 = 17.066f; > > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~ > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1331:65: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > > 1331 | dc->dcn_soc->fabric_and_dram_bandwidth_vmid0p72 = 14.933f; > > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~ > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1332:65: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > > 1332 | dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 = 12.8f; > > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~ > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.c:1334:72: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > > 1334 | dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = 20.80f; > > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~ > > make[6]: *** [scripts/Makefile.build:250: drivers/gpu/drm/amd/amdgpu/../display/dc/dcn10/dcn10_resource.o] Error 1 > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c: In function ‘dcn32_populate_dml_pipes_from_context’: > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c:1921:70: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > > 1921 | pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_luma = 0; > > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ > > drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.c:1922:72: error: ‘-mgeneral-regs-only’ is incompatible with the use of floating-point types > > 1922 | pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_chroma = 0; > > | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~ > > make[6]: *** [scripts/Makefile.build:250: drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource.o] Error 1 > > Could you check if you have this commit in your branch? > > drm/amd/display: move remaining FPU code to dml folder > > iirc, I merged it around Tue: > https://gitlab.freedesktop.org/agd5f/linux/-/commits/amd-staging-drm-next Ah sorry, I had to use the drm-next branch because amd-staging-drm-next is based on 5.18, meaning it does not have support for a recent pahole flag that is on by default, which breaks all of my builds. That support was enabled in 6.0 if I remember correctly; drm-next is based on 6.1-rc1 so everything works fine there. You can just ignore this error since it was already fixed. Cheers, Nathan _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH v3 1/1] drm/amd/display: add DCN support for ARM64 2022-10-28 15:48 ` Nathan Chancellor @ 2022-10-31 18:47 ` Nathan Chancellor 0 siblings, 0 replies; 8+ messages in thread From: Nathan Chancellor @ 2022-10-31 18:47 UTC (permalink / raw) To: Rodrigo Siqueira Jordao, Ao Zhong Cc: Harry Wentland, Leo Li, Arnd Bergmann, Alex Deucher, Stephen Rothwell, Christian König, amd-gfx, linux-arm-kernel Hi Rodrigo and Ao, On Fri, Oct 28, 2022 at 08:48:26AM -0700, Nathan Chancellor wrote: > On Fri, Oct 28, 2022 at 11:35:32AM -0400, Rodrigo Siqueira Jordao wrote: > > > > > > On 2022-10-28 11:09, Nathan Chancellor wrote: > > > Hi Ao, > > > > > > On Thu, Oct 27, 2022 at 09:52:29PM +0200, Ao Zhong wrote: > > > > After moving all FPU code to the DML folder, we can enable DCN support > > > > for the ARM64 platform. Remove the -mgeneral-regs-only CFLAG from the > > > > code in the DML folder that needs to use hardware FPU, and add a control > > > > mechanism for ARM Neon. > > > > > > > > Signed-off-by: Ao Zhong <hacc1225@gmail.com> > > > > --- > > > > drivers/gpu/drm/amd/display/Kconfig | 2 +- > > > > .../gpu/drm/amd/display/amdgpu_dm/dc_fpu.c | 6 ++++++ > > > > drivers/gpu/drm/amd/display/dc/dml/Makefile | 20 +++++++++++++++---- > > > > 3 files changed, 23 insertions(+), 5 deletions(-) > > > > > > > > diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig > > > > index 0142affcdaa3..a7f1c4e51719 100644 > > > > --- a/drivers/gpu/drm/amd/display/Kconfig > > > > +++ b/drivers/gpu/drm/amd/display/Kconfig > > > > @@ -6,7 +6,7 @@ config DRM_AMD_DC > > > > bool "AMD DC - Enable new display engine" > > > > default y > > > > select SND_HDA_COMPONENT if SND_HDA_CORE > > > > - select DRM_AMD_DC_DCN if (X86 || PPC64) > > > > + select DRM_AMD_DC_DCN if (X86 || PPC64 || (ARM64 && KERNEL_MODE_NEON)) > > > > help > > > > Choose this option if you want to use the new display engine > > > > support for AMDGPU. This adds required support for Vega and > > > > diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c > > > > index ab0c6d191038..1743ca0a3641 100644 > > > > --- a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c > > > > +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c > > > > @@ -31,6 +31,8 @@ > > > > #elif defined(CONFIG_PPC64) > > > > #include <asm/switch_to.h> > > > > #include <asm/cputable.h> > > > > +#elif defined(CONFIG_ARM64) > > > > +#include <asm/neon.h> > > > > #endif > > > > /** > > > > @@ -99,6 +101,8 @@ void dc_fpu_begin(const char *function_name, const int line) > > > > preempt_disable(); > > > > enable_kernel_fp(); > > > > } > > > > +#elif defined(CONFIG_ARM64) > > > > + kernel_neon_begin(); > > > > #endif > > > > } > > > > @@ -136,6 +140,8 @@ void dc_fpu_end(const char *function_name, const int line) > > > > disable_kernel_fp(); > > > > preempt_enable(); > > > > } > > > > +#elif defined(CONFIG_ARM64) > > > > + kernel_neon_end(); > > > > #endif > > > > } > > > > diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile > > > > index d0c6cf61c676..d4e93bed1c8e 100644 > > > > --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile > > > > +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile > > > > @@ -33,6 +33,10 @@ ifdef CONFIG_PPC64 > > > > dml_ccflags := -mhard-float -maltivec > > > > endif > > > > +ifdef CONFIG_ARM64 > > > > +dml_rcflags := -mgeneral-regs-only > > > > +endif > > > > + > > > > ifdef CONFIG_CC_IS_GCC > > > > ifeq ($(call cc-ifversion, -lt, 0701, y), y) > > > > IS_OLD_GCC = 1 > > > > @@ -55,8 +59,6 @@ frame_warn_flag := -Wframe-larger-than=2048 > > > > endif > > > > CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_ccflags) > > > > - > > > > -ifdef CONFIG_DRM_AMD_DC_DCN > > > > CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_ccflags) > > > > CFLAGS_$(AMDDALPATH)/dc/dml/dcn10/dcn10_fpu.o := $(dml_ccflags) > > > > CFLAGS_$(AMDDALPATH)/dc/dml/dcn20/dcn20_fpu.o := $(dml_ccflags) > > > > @@ -88,7 +90,6 @@ CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calcs.o := $(dml_ccflags) > > > > CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_auto.o := $(dml_ccflags) > > > > CFLAGS_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_ccflags) -Wno-tautological-compare > > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_vba.o := $(dml_rcflags) > > > > -CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn2x/dcn2x.o := $(dml_rcflags) > > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_mode_vba_20.o := $(dml_rcflags) > > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_rq_dlg_calc_20.o := $(dml_rcflags) > > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/display_mode_vba_20v2.o := $(dml_rcflags) > > > > @@ -105,7 +106,18 @@ CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_util_32.o := $(dml_rcf > > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_rcflags) > > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_rcflags) > > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dsc/rc_calc_fpu.o := $(dml_rcflags) > > > > -endif > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn10/dcn10_fpu.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn20/dcn20_fpu.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/display_mode_vba_314.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/display_rq_dlg_calc_314.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn314/dcn314_fpu.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn30/dcn30_fpu.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn32/dcn32_fpu.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn321/dcn321_fpu.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn31/dcn31_fpu.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn302/dcn302_fpu.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn303/dcn303_fpu.o := $(dml_rcflags) > > > > +CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_rcflags) > > > > CFLAGS_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_ccflags) > > > > CFLAGS_$(AMDDALPATH)/dc/dml/display_rq_dlg_helpers.o := $(dml_ccflags) > > > > CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dml1_display_rq_dlg_calc.o := $(dml_rcflags) > > > > -- > > > > 2.37.4 > > > > > > > > > > Enabling this code for arm64 reveals the following warnings when > > > building allmodconfig with clang. A very recent change in LLVM added the > > > variable and spill information so that will not be seen with older > > > versions but I also saw the warnings with clang 15.0.0 from Fedora. > > > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn30/display_mode_vba_30.c:1852:13: error: stack frame size (2112) exceeds limit (2048) in 'DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation' [-Werror,-Wframe-larger-than] > > > static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation( > > > ^ > > > 1152/2112 (54.55%) spills, 960/2112 (45.45%) variables > > > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn30/display_mode_vba_30.c:3542:6: error: stack frame size (2240) exceeds limit (2048) in 'dml30_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > > > void dml30_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > > > ^ > > > 1451/2240 (64.78%) spills, 789/2240 (35.22%) variables > > > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn31/display_mode_vba_31.c:3791:6: error: stack frame size (2736) exceeds limit (2048) in 'dml31_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > > > void dml31_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > > > ^ > > > 1342/2736 (49.05%) spills, 1394/2736 (50.95%) variables > > > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn314/display_mode_vba_314.c:3890:6: error: stack frame size (2720) exceeds limit (2048) in 'dml314_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > > > void dml314_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > > > ^ > > > 1325/2720 (48.71%) spills, 1395/2720 (51.29%) variables > > > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/display_mode_vba_32.c:59:13: error: stack frame size (2208) exceeds limit (2048) in 'DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation' [-Werror,-Wframe-larger-than] > > > static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation( > > > ^ > > > 1397/2208 (63.27%) spills, 811/2208 (36.73%) variables > > > > > > drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/display_mode_vba_32.c:1681:6: error: stack frame size (2496) exceeds limit (2048) in 'dml32_ModeSupportAndSystemConfigurationFull' [-Werror,-Wframe-larger-than] > > > void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) > > > ^ > > > 1504/2496 (60.26%) spills, 992/2496 (39.74%) variables > > > > > > I can try to take a look at these next week, as I have prior obligations > > > today, unless others wanted to help me out ;) > > > > Hi Nathan, > > > > Thanks a lot for checking this. > > > > About those stack frame issues, I guess they are generic, right? I mean, I > > don't see why we only see it if we enable arm64. > > I do not see these warnings on x86_64 as I had to tackle them for 6.0. > powerpc allmodconfig is already broken with clang for other reasons so I > am not sure if these warnings are there; I can try to test that. I am > not sure why things are going sideways for arm64 with clang here but I > would really like to keep allmodconfig building without any issues, as > that is a good target for catching new warnings due to -Werror. I was thinking about this over the weekend and I do not think it is fair to hold up this change for these warnings, as I am not sure when I will be able to investigate them fully. However, I still do not want to see our builds broken over this. Would you consider adding the following diff and sending a v4? This would allow others to look into and fix the warnings if they have a vested interest in making this code compile with clang on arm64, while not impacting people who use clang and do not care about this code or people who just use GCC. Thoughts? Cheers, Nathan diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig index a7f1c4e51719..843a55a6a3ac 100644 --- a/drivers/gpu/drm/amd/display/Kconfig +++ b/drivers/gpu/drm/amd/display/Kconfig @@ -6,7 +6,8 @@ config DRM_AMD_DC bool "AMD DC - Enable new display engine" default y select SND_HDA_COMPONENT if SND_HDA_CORE - select DRM_AMD_DC_DCN if (X86 || PPC64 || (ARM64 && KERNEL_MODE_NEON)) + # !CC_IS_CLANG: https://github.com/ClangBuiltLinux/linux/issues/1752 + select DRM_AMD_DC_DCN if (X86 || PPC64 || (ARM64 && KERNEL_MODE_NEON && !CC_IS_CLANG)) help Choose this option if you want to use the new display engine support for AMDGPU. This adds required support for Vega and _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply related [flat|nested] 8+ messages in thread
end of thread, other threads:[~2022-10-31 18:49 UTC | newest] Thread overview: 8+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2022-10-27 19:52 [PATCH v3 0/1] drm/amd/display: add DCN support for ARM64 Ao Zhong 2022-10-27 19:52 ` [PATCH v3 1/1] " Ao Zhong 2022-10-27 20:46 ` Arnd Bergmann 2022-10-28 15:09 ` Nathan Chancellor 2022-10-28 15:28 ` Ao Zhong 2022-10-28 15:35 ` Rodrigo Siqueira Jordao 2022-10-28 15:48 ` Nathan Chancellor 2022-10-31 18:47 ` Nathan Chancellor
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox