diff options
| -rw-r--r-- | main.c | 31 |
1 files changed, 23 insertions, 8 deletions
@@ -78,7 +78,8 @@ void RenderFullscren(void) { DWORD64 end = __rdtsc(); DWORD64 diff = end - start; - avg_fs = ((avg_fs * frame_count) + diff) / (frame_count + 1); + avg_fs = (avg_fs + diff) / 2; + // avg_fs = ((avg_fs * frame_count) + diff) / (frame_count + 1); char buffer[1024]; wsprintfA(buffer, "fullscreen scan %u cycles per triangle\n", diff); @@ -121,8 +122,8 @@ void RenderTriangle(void) { DWORD64 end = __rdtsc(); DWORD64 diff = end - start; - // avg = (avg + diff) / 2; - avg = ((avg * frame_count) + diff) / (frame_count + 1); + avg = (avg + diff) / 2; + // avg = ((avg * frame_count) + diff) / (frame_count + 1); char buffer[1024]; wsprintfA(buffer, "%u cycles per triangle\n", diff); @@ -268,8 +269,8 @@ void RenderTriangleSIMD(void) { DWORD64 end = __rdtsc(); DWORD64 diff = end - start; - // avg_simd = (avg_simd + diff) / 2; - avg_simd = ((avg_simd * frame_count) + diff) / (frame_count + 1); + avg_simd = (avg_simd + diff) / 2; + // avg_simd = ((avg_simd * frame_count) + diff) / (frame_count + 1); char buffer[1024]; wsprintfA(buffer, "SIMD %u cycles per triangle\n", diff); @@ -309,6 +310,7 @@ void RenderTriangleSIMDSimplified(void) { __m128i m_one = _mm_set1_epi32(-1); __m128i two = _mm_set1_epi32(2); + __m128i mask = _mm_set_epi32(0x000000ff, 0x000000ff, 0x000000ff, 0x000000ff); __m128i vx = _mm_set_epi32(min_x + 1, min_x, min_x + 1, min_x); __m128i vy = _mm_set_epi32(min_y + 1, min_y + 1, min_y, min_y); @@ -337,6 +339,9 @@ void RenderTriangleSIMDSimplified(void) { __m128i diffy2 = _mm_mullo_epi32(v1v2x, two); __m128i diffy3 = _mm_mullo_epi32(v2v0x, two); + UINT32 area = (t.y[0] - t.y[1]) * t.x[2] + (t.x[1] - t.x[0]) * t.y[2] + (t.x[0] * t.y[1] - t.y[0] * t.x[1]); + __m128i v_area = _mm_set1_epi32(0x0fffffff / area); + for (UINT32 y = min_y; y < max_y + 1; y += 2) { UINT32 row_1 = y * g_width; UINT32 row_2 = (y + 1) * g_width; @@ -349,8 +354,17 @@ void RenderTriangleSIMDSimplified(void) { __m128i or = _mm_or_si128(dt1_or_dt2, dt3); __m128i gtz = _mm_cmpgt_epi32(or, m_one); - pixels[(row_1 + x) / 2] = _mm_extract_epi64(gtz, 0); - pixels[(row_2 + x) / 2] = _mm_extract_epi64(gtz, 1); + __m128i b1 = _mm_srli_epi32(_mm_mullo_epi32(dt1, v_area), 20); + __m128i b2 = _mm_srli_epi32(_mm_mullo_epi32(dt2, v_area), 20); + __m128i b3 = _mm_srli_epi32(_mm_mullo_epi32(dt3, v_area), 20); + + __m128i masked_b1 = _mm_and_si128(b1, mask); + __m128i masked_b2 = _mm_slli_epi32(_mm_and_si128(b2, mask), 8); + __m128i masked_b3 = _mm_slli_epi32(_mm_and_si128(b3, mask), 16); + __m128i color = _mm_and_si128(_mm_or_si128(_mm_or_si128(masked_b3, masked_b2), masked_b1), gtz); + + pixels[(row_1 + x) / 2] = _mm_extract_epi64(color, 0); + pixels[(row_2 + x) / 2] = _mm_extract_epi64(color, 1); dt1 = _mm_add_epi32(dt1, diffx1); dt2 = _mm_add_epi32(dt2, diffx2); @@ -365,7 +379,8 @@ void RenderTriangleSIMDSimplified(void) { DWORD64 end = __rdtsc(); DWORD64 diff = end - start; - avg_simd_simpl = ((avg_simd_simpl * frame_count) + diff) / (frame_count + 1); + // avg_simd_simpl = ((avg_simd_simpl * frame_count) + diff) / (frame_count + 1); + avg_simd_simpl = (avg_simd_simpl + diff) / 2; char buffer[1024]; wsprintfA(buffer, "SIMD simplified %u cycles per triangle\n", diff); |