I can do that if you prefer. I just followed this approach for consistency because it was the one followed for cfloat test.
El dl., 3 de nov. 2025, 16:01, Jonathan Wakely <[email protected]> va escriure: > > > On Monday, 3 November 2025, Xavier Bonaventura <[email protected]> > wrote: > >> Hi, sorry but I screwed up my previous mail. This is my first patch and >> the first time I submit a patch to a mailing list. >> >> Now that I know how to add an introduction before sending a patch, let >> me explain a bit. >> When taking a look to the test suite of the standard library I found >> what I believe it was a missing test case. I do not think this is >> intentional. Because I believe the change is trivial (famous last words) >> I just provide the patch. >> I run the tests for 18_support for unix and they were passing. >> I read quite a lot of documents about contributing, but I might have >> miss some part. If that is the case let me know and sorry for that. >> > > Thanks, the patch looks correct. > > But we could also just add those lines to the existing values.cc guarded > by #if __cpluscplus >= 201103 > > I don't think there's any advantage to testing once with the long long > constants and again with them. One test is enough. > >
