What do you want to achieve?
I would like to leverage Neon instructions on Cortex-A76 (ARMv8.2-A architecture) running LVGL 9.3.
What have you tried so far?
I tried to enable Neon support in the configuration file:
#define LV_USE_DRAW_SW_ASM LV_DRAW_SW_ASM_NEON
However, the compilation fails due to unknown mnemonics, for example:
lvgl/src/draw/sw/blend/neon/lv_blend_neon.S:702: Error: unknown mnemonic 'vbif' -- 'vbif D_B,S_565_L,TMP_D1'
It seems lv_blend_neon.S
provides assembly code for the ARMv7 architecture:
(...)
#if LV_USE_DRAW_SW_ASM == LV_DRAW_SW_ASM_NEON
.text
.fpu neon
.arch armv7a
(...)
Do we have/plan to have the assembly for the 64-bit architecture (such as ARMv8.2-A)?
Environment
- Compiler: aarch64-none-linux-gnu-gcc (12.2.1)
- MCU/MPU/Board: Cortex-A76 based SoC
- LVGL version: 9.3.0