Re: [FFmpeg-devel] [PATCH 3/3] avcodec/hevc: Add wasm simd128 idct

2024-11-14 Thread Zhao Zhili
> Le 13 novembre 2024 16:16:36 GMT+02:00, Zhao Zhili > a écrit : > >From: Zhao Zhili >> > > > >We don't use intrinsics normally for performance reasons. However, > >WASM isn't machine instruction.

Re: [FFmpeg-devel] [PATCH 3/3] avcodec/hevc: Add wasm simd128 idct

2024-11-13 Thread Rémi Denis-Courmont
Le 13 novembre 2024 16:16:36 GMT+02:00, Zhao Zhili a écrit : >From: Zhao Zhili > >We don't use intrinsics normally for performance reasons. However, >WASM isn't machine instruction. I don't think that's relevant to the matter. What may be relevant is that WASM is an infinite register ISA, so

[FFmpeg-devel] [PATCH 3/3] avcodec/hevc: Add wasm simd128 idct

2024-11-13 Thread Zhao Zhili
From: Zhao Zhili We don't use intrinsics normally for performance reasons. However, WASM isn't machine instruction. I suspect that in this scenario, handwritten assembly may not bring much performance compared with intrinsics. And I don't want to dig into the WASM ABI. So this patch is implemente