Change Details

Optimize mesh normal calculation. - Remove the intermediate `lnors_weighted` array, accumulate directly into the normal array using a lock for thread safety. - Remove single threaded iteration over loops. The performance difference depends a lot on use case and threads available. Rough measurements for `BKE_mesh_calc_normals_poly`: - 2.92x faster with ~25 million quads. - 2.7x faster with ~2 million quads. - 1.3x slower with 512k quads. - 1.6x slower with 256k quads. - 2.26x slower with 128k quads. - 1.5x slower with 16k quads. - 1.15x slower with 8k quads. Tested using 32 cores (64 threads). ---- Observations: - In my tests the spin-lock was almost never waiting, roughly ~0.01% of additions. - The overhead of locking was negligible (replacing with `add_v3_v3` didn't make a significant difference to performance). - Changing optimization flags didn't make a significant difference (`-O2`, `-O3`, `-Ofast`, both GCC and CLANG gave comparable results). ---- Posting this for review as this reverts rBd130c66db436b1fccbbde040839bc4cb5ddaacd2, The only significant things I can see that are different in this patch compared to the code before rBd130c66db436b1fccbbde040839bc4cb5ddaacd2 was applied are. - Lock the entire vector before adding (instead of 3x `atomic_add_and_fetch_fl` calls per vertex). - Vertex accumulation runs in parallel. ---- Thanks to @easythrees for helping investigate this patch.

Optimize mesh normal calculation. - Remove the intermediate `lnors_weighted` array, accumulate directly into the normal array using a lock for thread safety. - Remove single threaded iteration over loops. - Final normalization is now done in-line, using an array of remaining loops to detect the last vertex accessed. The performance difference depends a lot on use case and threads available. Rough measurements for `BKE_mesh_calc_normals_poly`: - 2.92x faster with ~25 million quads. - 2.7x faster with ~2 million quads. - 1.3x slower with 512k quads. - 1.6x slower with 256k quads. - 2.26x slower with 128k quads. - 1.5x slower with 16k quads. - 1.15x slower with 8k quads. Tested using 32 cores (64 threads). ---- Observations: - In my tests the spin-lock was almost never waiting, roughly ~0.01% of additions. - The overhead of locking was negligible (replacing with `add_v3_v3` didn't make a significant difference to performance). - Changing optimization flags didn't make a significant difference (`-O2`, `-O3`, `-Ofast`, both GCC and CLANG gave comparable results). ---- Posting this for review as this reverts rBd130c66db436b1fccbbde040839bc4cb5ddaacd2, The only significant things I can see that are different in this patch compared to the code before rBd130c66db436b1fccbbde040839bc4cb5ddaacd2 was applied are. - Lock the entire vector before adding (instead of 3x `atomic_add_and_fetch_fl` calls per vertex). - Vertex accumulation runs in parallel. ---- Thanks to @easythrees for helping investigate this patch.

Optimize mesh normal calculation. - Remove the intermediate `lnors_weighted` array, accumulate directly into the normal array using a lock for thread safety. - Remove single threaded iteration over loops. - Final normalization is now done in-line, using an array of remaining loops to detect the last vertex accessed. The performance difference depends a lot on use case and threads available. Rough measurements for `BKE_mesh_calc_normals_poly`: - 2.92x faster with ~25 million quads. - 2.7x faster with ~2 million quads. - 1.3x slower with 512k quads. - 1.6x slower with 256k quads. - 2.26x slower with 128k quads. - 1.5x slower with 16k quads. - 1.15x slower with 8k quads. Tested using 32 cores (64 threads). ---- Observations: - In my tests the spin-lock was almost never waiting, roughly ~0.01% of additions. - The overhead of locking was negligible (replacing with `add_v3_v3` didn't make a significant difference to performance). - Changing optimization flags didn't make a significant difference (`-O2`, `-O3`, `-Ofast`, both GCC and CLANG gave comparable results). ---- Posting this for review as this reverts rBd130c66db436b1fccbbde040839bc4cb5ddaacd2, The only significant things I can see that are different in this patch compared to the code before rBd130c66db436b1fccbbde040839bc4cb5ddaacd2 was applied are. - Lock the entire vector before adding (instead of 3x `atomic_add_and_fetch_fl` calls per vertex). - Vertex accumulation runs in parallel. ---- Thanks to @easythrees for helping investigate this patch.