Paul Eggert wrote: > > To me, such a compiler bug in a configure test does not warrant a workaround > > in Gnulib — especially since it's scheduled to be fixed in the next compiler > > release. > > Ah, I merely copied the simple pattern I see used in many other modules > to catch similar infloops. Also, I was worried that if ICX has the bug, > other Clang variants will also have it.
Thanks for dealing with this! (while I was lazy...) > But if you think the > recently-installed Gnulib patch unnecessary, feel free to revert it. Your patch is perfect. Bruno
