From 0a0d2c944372590721f5accc74d03789c1fa6098 Mon Sep 17 00:00:00 2001 From: Ian Romanick Date: Thu, 6 Nov 2014 22:51:45 -0800 Subject: [PATCH] mesa: Micro-optimize _mesa_is_valid_prim_mode You would not believe the mess GCC 4.8.3 generated for the old switch-statement. On Bay Trail-D using Fedora 20 compile flags (-m64 -O2 -mtune=generic for 64-bit and -m32 -march=i686 -mtune=atom for 32-bit), affects Gl32Batch7: 32-bit: Difference at 95.0% confidence -0.37374% +/- 0.184057% (n=40) 64-bit: Difference at 95.0% confidence 0.966722% +/- 0.338442% (n=40) The regression on 32-bit is odd. Callgrind says the caller, _mesa_is_valid_prim_mode is faster. Before it says 2,293,760 cycles, and after it says 917,504. Signed-off-by: Ian Romanick Reviewed-by: Kenneth Graunke --- src/mesa/main/api_validate.c | 30 ++++++++++++------------------ 1 file changed, 12 insertions(+), 18 deletions(-) diff --git a/src/mesa/main/api_validate.c b/src/mesa/main/api_validate.c index b882f0e837d..9c2e29e6472 100644 --- a/src/mesa/main/api_validate.c +++ b/src/mesa/main/api_validate.c @@ -113,27 +113,21 @@ check_valid_to_render(struct gl_context *ctx, const char *function) bool _mesa_is_valid_prim_mode(struct gl_context *ctx, GLenum mode) { - switch (mode) { - case GL_POINTS: - case GL_LINES: - case GL_LINE_LOOP: - case GL_LINE_STRIP: - case GL_TRIANGLES: - case GL_TRIANGLE_STRIP: - case GL_TRIANGLE_FAN: + /* The overwhelmingly common case is (mode <= GL_TRIANGLE_FAN). Test that + * first and exit. You would think that a switch-statement would be the + * right approach, but at least GCC 4.7.2 generates some pretty dire code + * for the common case. + */ + if (likely(mode <= GL_TRIANGLE_FAN)) return true; - case GL_QUADS: - case GL_QUAD_STRIP: - case GL_POLYGON: + + if (mode <= GL_POLYGON) return (ctx->API == API_OPENGL_COMPAT); - case GL_LINES_ADJACENCY: - case GL_LINE_STRIP_ADJACENCY: - case GL_TRIANGLES_ADJACENCY: - case GL_TRIANGLE_STRIP_ADJACENCY: + + if (mode <= GL_TRIANGLE_STRIP_ADJACENCY) return _mesa_has_geometry_shaders(ctx); - default: - return false; - } + + return false; } -- 2.30.2