> > Okay for trunk? I would like to check that one in to avoid the noise > > in the regression reports. > > I've tested this version in a few trees.
Thanks Thomas. > That's because of effective-target 'struct_musttail' for '-m32' > reporting: > > struct_musttail1494739.cc: In function 'foo bar()': > struct_musttail1494739.cc:5:88: error: cannot tail-call: return value > used after call > > (I'm just mentioning the latter "regressions" in case those are > unexpected.) I believe that's because these test cases are handled by the GIMPLE level tail call handling in tree-tailcall (which avoids any target restrictions), while the TCL test checks for the generic case using an extern (so hits target restrictions). While this could probably be distinguished in the test case probing I don't think it's worth it. Some of this is just for the frontend, which is architecture independent enough. -Andi
