On Fri, Feb 03, 2017 at 09:04:06PM -0600, William Harrington wrote: > On Fri, 3 Feb 2017 21:48:18 +0000 > Ken Moffat <[email protected]> wrote: > > > 'm not sure if gcc-6.3 is weirder than I expected, or if I'm slowly > > spreading my own weirdness into the things I touch ;-) > > GNU Tools build -g -O2 by default. With gcc, the later flags take precedence. > So if using CFLAGS and the package overrides CFLAGS so you get -O3 -g -O2 -g > then -O2 takes precedence. It is discussed clearly in the optimization > documentation of GCC. So with what I see, is that the CFLAGS are taking place > after the defaults which is what I'm concerned about. CFLAGS should be used > after the defaults to override them as -g -O and -s are very dependent of > where they are in the command line regarding FLAG variables. > > Sincerely, > William Harrington
Sure, I understand that later -O variants override the earlier, but invoking g++ with -O2 works and -O3 fails in the belief it needs c++-11. That's what I think is weird, some extra optimization has broken it. Meanwhile, upstream told me that compiling without optimization (his words) for production is a bad idea, and that the codebase will move from -ansi to -std=++-11 soon : I suspect that by that time I may well have stopped caring about this package - personally, I hold it to be self-evident that -O2 is usually sufficient optimization. Latex can be useful in some places (for me, as someone who is not a mathematician, particularly using xelatex to avoid Knuth's ugly fonts and to cope with most languages), but I'm losing patience with building the parts not shipped in the main source. ĸen (yes, I'm pissed off by upstream when we have enough potential problems in our own builds - does it show ?) -- `I shall take my mountains', said Lu-Tze. `The climate will be good for them.' -- Small Gods -- http://lists.linuxfromscratch.org/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
