Does recent behavioral-model change cause p4c tests to fail? #889
Comments
antoninbas
added a commit
that referenced
this issue
May 11, 2020
A recent fix for a separate issue (#888) exposed a separate bug: the type of expressions for which the top-level operator was "access_field" was not deduced properly (set to UNKNOWN instead of DATA), which caused bmv2 to assert when building the Action object. The same issue existed for "access_union_header", which is an operator introduced more recently (and possibly not used by p4c yet). Fixes #889
antoninbas
added a commit
that referenced
this issue
May 11, 2020
A recent fix for a separate issue (#888) exposed a separate bug: the type of expressions for which the top-level operator was "access_field" was not deduced properly (set to UNKNOWN instead of DATA), which caused bmv2 to assert when building the Action object. The same issue existed for "access_union_header", which is an operator introduced more recently (and possibly not used by p4c yet). Fixes #889
antoninbas
added a commit
that referenced
this issue
May 11, 2020
A recent fix for a separate issue (#888) exposed a separate bug: the type of expressions for which the top-level operator was "access_field" was not deduced properly (set to UNKNOWN instead of DATA), which caused bmv2 to assert when building the Action object. The same issue existed for "access_union_header", which is an operator introduced more recently (and possibly not used by p4c yet). Fixes #889
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
See some details on this p4c comment: p4lang/p4c#2370 (comment)
It is not yet clear to me whether the recent behavioral-model commit has an error, or perhaps p4c is producing bad JSON for the p4c test programs with
runtime-indexin their names, and this behavioral-model change exposes that.The text was updated successfully, but these errors were encountered: