erfc(1.960/sqrt(2.0)) was completely wrong. It expected 0.95, but the actual result should be ~0.05.
Also updated: erfc(2.576/sqrt(2.0)) for more precision in expected result Tested with clover/libclc on PITCAIRN and Beignet on haswell. This prevents regressions if/when the piglit ULP calculation is ever fixed. Signed-off-by: Aaron Watry <[email protected]> --- generated_tests/gen_cl_math_builtins.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/generated_tests/gen_cl_math_builtins.py b/generated_tests/gen_cl_math_builtins.py index 29b4e1a..a14d97b 100644 --- a/generated_tests/gen_cl_math_builtins.py +++ b/generated_tests/gen_cl_math_builtins.py @@ -252,8 +252,8 @@ tests = { 'arg_types' : [F, F], 'function_type': 'ttt', 'values' : [ - [1.0, 0.95, 0.009995, 1.994999, 0.0, 0.0, 2.0], # Result - [0.0, 1.960/sqrt(2.0), 2.576/sqrt(2.0), -2.807/sqrt(2.0), 11.1, float("inf"), float("-inf")] # Arg0 + [1.0, 0.049995790296440745, 0.0099950646314, 1.994999, 0.0, 0.0, 2.0], # Result + [0.0, 1.960/sqrt(2.0), 2.576/sqrt(2.0), -2.807/sqrt(2.0), 11.1, float("inf"), float("-inf")] # Arg0 ], 'tolerance' : 16 }, -- 2.9.3 _______________________________________________ Piglit mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/piglit
