cannot convert 'float32x4_t' toBuild Error 'uint32x4_t' and 'uint16x8_t' to 'uint32x4_t' on engines/ags/lib/allegro/surface_neon.cpp and surgace_neon.h

[Kelvfimer]

Hello

---

Description

---

Starting commit 1cf3c7832a236e93b682f5c04fe4daf0332ad2aa I'm having several build errors stating:

engines/ags/lib/allegro/surface_neon.cpp: In function 'uint32x4_t AGS3::blendTintSpriteSIMD(uint32x4_t, uint32x4_t, uint32x4_t, bool)':

engines/ags/lib/allegro/surface_neon.cpp:255:111: error: cannot convert 'float32x4_t' to 'uint32x4_t'

engines/ags/lib/allegro/surface_neon.cpp:22:
/home/kelv/EmuELEC/build.EmuELEC-Amlogic-ng.aarch64-4/toolchain/lib/gcc/aarch64-libreelec-linux-gnu/12.2.0/include/arm_neon.h:10605:23: note: initializing argument 1 of 'uint32x4_t vceqq_u32(uint32x4_t, uint32x4_t)'
10605 | vceqq_u32 (uint32x4_t a, uint32x4_t b)

| ~

engines/ags/lib/allegro/surface_neon.cpp:256:111: error: cannot convert 'float32x4_t' to 'uint32x4_t'

256 | float32x4_t hgfactors = vcvtq_f32_u32(vandq_u32(vandq_u32(vceqq_f32(ssg, smaxes), vmvnq_u32(vceqq_u32(ssg, ssr))), vmovq_n_u32(1)));

lib/gcc/aarch64-libreelec-linux-gnu/12.2.0/include/arm_neon.h:1134:23: note: initializing argument 1 of 'uint32x4_t vandq_u32(uint32x4_t, uint32x4_t)'

1134 | vandq_u32 (uint32x4_t a, uint32x4_t b)

| ~

/home/kelv/EmuELEC/build.EmuELEC-Amlogic-ng.aarch64-4/build/scummvmsa-9f4a32f88be3256aa28909deba345b8a1097727b/engines/ags/lib/allegro/surface_neon.cpp:386:88: error: cannot convert 'uint16x8_t' to 'uint32x4_t'

386 | ch2 = vandq_u32(rgbBlendSIMD2Bpp(srcCols, destCols, alphas), vmvnq_u32(mask));

surface_neon.cpp:387:34: error: cannot convert 'uint16x8_t' to 'uint32x4_t'

---

Platform

---

Emuelec 4.6 linux 4.9 odroid n2+

---

Expected behaviour

---

To be able to build it again.

[Kelvfimer]

Tested on commit b174006c5bfa9b4bd2b87eda377ab5b07d9d31b5, the issue is already FIXED.

thx you so much!!!

[Kelvfimer]

Fixed on latest commit tested b174006c5bfa9b4bd2b87eda377ab5b07d9d31b5