On Thu, 12 Dec 2019, Michael Niedermayer wrote:
On Wed, Dec 11, 2019 at 11:21:27AM +0200, Martin Storsjö wrote:
As the values generated by av_bmg_get can be arbitrarily large
(only the stddev is specified), we can't use a fixed tolerance.
Calculate a dynamic tolerance (like in float_dsp from 38f966b2222db),
based on the individual steps of the calculation.
This fixes running this test with certain seeds, when built with
clang for mingw/x86_32.
---
tests/checkasm/af_afir.c | 14 +++++++++++++-
1 file changed, 13 insertions(+), 1 deletion(-)
diff --git a/tests/checkasm/af_afir.c b/tests/checkasm/af_afir.c
index e3fb76e8e0..e791f88b97 100644
--- a/tests/checkasm/af_afir.c
+++ b/tests/checkasm/af_afir.c
@@ -53,7 +53,19 @@ static void test_fcmul_add(const float *src0, const float
*src1, const float *sr
call_ref(cdst, src1, src2, LEN);
call_new(odst, src1, src2, LEN);
for (i = 0; i <= LEN*2; i++) {
- if (!float_near_abs_eps(cdst[i], odst[i], 6.2e-05)) {
+ int idx = i & ~1;
+ float cre = src2[idx];
+ float cim = src2[idx + 1];
+ float tre = src1[idx];
+ float tim = src1[idx + 1];
+ double t = fabs(src0[i]) +
+ fabs(tre) + fabs(tim) + fabs(cre) + fabs(tim) +
Is it intended to add fabs(tim) twice ?
Oops, that's a typo, it was meant to be tre+tim+cre+cim.
// Martin
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
