Skip to contentSkip to conversation

vulkan-samples: Disable on 32bit ARM

PATCH In review Khem Raj 2 participants 3 messages Sep 10, 2024

lore
Share & tools
lore

Khem Raj

author
#0 lore raw

PATCH 1/1 vulkan-samples: Disable on 32bit ARM

vulkan-samples depend on atsc module and this module only supports x86_64/arm64 as it seems, Neon support check does not do a good job of detecting 64bit arm + neon to enable it and it gets enabled for 32bit arm and it uses 64bit neon ISA ending in build errors e.g.

astc/Source/astcenc_vecmathlib_neon_4.h:420:16: error: 'vaddvq_u32' was not declared in this scope; did you mean 'vaddq_u32'?

if upstream fixes the build for arm we can re-enable it

---
 meta/recipes-graphics/vulkan/vulkan-samples_git.bb | 2 ++
 1 file changed, 2 insertions(+)
meta/recipes-graphics/vulkan/vulkan-samples_git.bb +2 -0
index 93d4cdfdd9e..e1a2d13cdff 100644
@@ -37,3 +37,5 @@ EXTRA_OECMAKE += "-DCMAKE_DISABLE_PRECOMPILE_HEADERS=ON"
 EXTRA_OECMAKE += "-DVKB_WSI_SELECTION=D2D"
 
 COMPATIBLE_HOST:x86 = "null"
+COMPATIBLE_HOST:arm = "null"
+COMPATIBLE_HOST:armv7ve = "null"
#1 lore raw
Quoted 3 lines · Khem Raj via lists.openembedded.org

vulkan-samples depend on atsc module and this module only supports x86_64/arm64 as it seems,

I’m wondering if we should reverse this logic and just list the platforms that vulkan-samples actually works on.

For example, it also fails on 32-bit ppc:

vulkan-samples/git/git/components/filesystem/src/std_filesystem.cpp:52:13: error: narrowing conversion of 'size' from 'long long unsigned int' to 'size_t' {aka 'unsigned int'} [-Werror=narrowing]

Ross

author
#2 lore raw
Quoted 8 lines · Ross Burton
Quoted 3 lines · Khem Raj via lists.openembedded.org

vulkan-samples depend on atsc module and this module only supports x86_64/arm64 as it seems,

I’m wondering if we should reverse this logic and just list the platforms that vulkan-samples actually works on.

Yeah that would be arm64 and x86_64 AFAICT

Quoted 9 lines

For example, it also fails on 32-bit ppc:

vulkan-samples/git/git/components/filesystem/src/std_filesystem.cpp:52:13:
error: narrowing conversion of 'size' from 'long long unsigned int' to 'size_t' {aka 'unsigned int'} [-Werror=narrowing]

Ross