From: Juan A. Suarez Romero Date: Wed, 8 Jan 2020 10:48:29 +0000 (+0000) Subject: nir/lower_double_ops: add note for lowering mod X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=784c454607be3e8dc395de429d9b99521d5ef8a8;p=mesa.git nir/lower_double_ops: add note for lowering mod Add a note to clarify that while Vulkan allows mod(x,y) to be in [0, y] range, OpenGL does not allow it, so the lowering ensures the result is always in [0, y) range, as this lowering is shared by the Vulkan and OpenGL implementation. Reviewed-by: Elie Tournier Tested-by: Marge Bot Part-of: --- diff --git a/src/compiler/nir/nir_lower_double_ops.c b/src/compiler/nir/nir_lower_double_ops.c index 322ae49bacd..e0617211d45 100644 --- a/src/compiler/nir/nir_lower_double_ops.c +++ b/src/compiler/nir/nir_lower_double_ops.c @@ -427,6 +427,10 @@ lower_mod(nir_builder *b, nir_ssa_def *src0, nir_ssa_def *src1) * If the division is lowered, it could add some rounding errors that make * floor() to return the quotient minus one when x = N * y. If this is the * case, we return zero because mod(x, y) output value is [0, y). + * + * Worth to note that Vulkan allows the output value to be in range [0, y], + * so mod(x, y) could return y; but as OpenGL does not allow this, we add + * the extra instruction to ensure the value is always in [0, y). */ nir_ssa_def *floor = nir_ffloor(b, nir_fdiv(b, src0, src1)); nir_ssa_def *mod = nir_fsub(b, src0, nir_fmul(b, src1, floor));