Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

mad single/double round implementation mode is unspecified in PRM 1.0 due to a specs bug #13

Open
pjaaskel opened this issue Jan 14, 2016 · 0 comments
Assignees

Comments

@pjaaskel
Copy link

Due to accidentally leaving out the runtime query for asking whether mad is implemented with a single or double roundings (https://app.devzing.com/hsafoundation/bugzilla/show_bug.cgi?id=540) as specified in PRM (http://www.hsafoundation.com/html/HSA_Library.htm#PRM/Topics/05_Arithmetic/floating_point_optimization.htm) from the specs version 1.0, I think the conformance test should not include tests which produce different results based on the implementation.

It currently assumes single rounding implementation which produces some differences with my double rounding implementation.

At least prm/core/arithmetic/intopt/mad/mad_ftz_f32_00028 in the Base suite produces different results in [3544] and [3940] which I think is due to this.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants