On Jan 2, 2019, at 5:39 AM, Rainer Orth <r...@cebitec.uni-bielefeld.de> wrote: > > It's on the brink, I'd say. The effective-keyword form is shorter and > more descriptive and can be extended to more targets way more easily. > Checking 2018-12 testresults, I find failures of the ident-*.c tests on > hppa2.0w-hp-hpux11.11 and powerpc-ibm-aix7.2.0.0. If those were to be > handled as well, I'd strongly prefer the effective-keyword variant > rather than adding more and more targets to the individual tests.
For a completely different thought, we can compare to a fine grain feature database that dejagnu or the running program can probe and adjust for... #ifdef __has_feature(directive_ident) #endif then, this is directly populated from the port's #define, and the duplication is gone. I think I like this direction. Let maintenance, faster port time, higher quality port, and testing less noisy in general.
