>> > Create still another version of it, you mean. Well sure. It's a one line change as opposed to mucking with how structures are supposed to be laid out.
>>> I am testing Nick's patch on ppc32 Darwin and plan to check in if it >>> looks like that was not broken. >> >> I would very much suggest that you not check in the 32-bit linux >> part, >> as it >> isn't the correct fix. > > This community is less obsessed than gcc'ers about finding "the > correct" fix, which speeds things up tremendously. In this case > ppc32 Linux didn't work before and does now; that's enough of a gain > to check something in IMO. If you want to improve it, you can. > In this case, Nick already asked me to check in a fix for it so I'll do that when I get a chance. I'll disagree that checking in something known to be wrong when there's an easy fix speeds anything up at all. > That said, it's true TARGET_ALIGN_MAC68K should be always false on > Linux, and the macro could be simplified. It's not even just that, it's a change to how the structure alignment is supposed to be. The two platforms are different in this way. -eric _______________________________________________ llvm-commits mailing list llvm-commits@cs.uiuc.edu http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
