| 1 |
A couple of CFLAGS that were never stable in any version of gcc: |
| 2 |
|
| 3 |
-freduce-all-givs |
| 4 |
-mfpmath=sse,387 |
| 5 |
|
| 6 |
The first flag will break things such as gcc, glibc, the kernel, |
| 7 |
binutils, bash and other nontrivial packages with every version of gcc |
| 8 |
I've ever tried (which is most of them). The flag has been denoted an |
| 9 |
experiment gone wrong and is no longer in CVS HEAD. The latter flag |
| 10 |
doesn't work properly and generates bad (sometimes segfaulting) code |
| 11 |
wherever SSE-related inline assembly is involved, and is slower than |
| 12 |
either -mfpmath=sse or -mfpmath=387 in 100% of cases according to those |
| 13 |
upstream. |
| 14 |
|
| 15 |
But yeah, -freduce-all-givs is a far, FAR more dangerous global flag |
| 16 |
than -ffast-math, which can in theory give you a working (if not |
| 17 |
supported) system when set globally. |
| 18 |
|
| 19 |
Also, perhaps highlight -fPIC in those cases where it shouldn't be used? |
| 20 |
Is that possible? Ehm. I don't mean that. Is that feasible? :-P |
| 21 |
|
| 22 |
Rob |
| 23 |
|
| 24 |
Ciaran McCreesh wrote: |
| 25 |
> I've got some better vim syntax highlighting rules for make.conf lined |
| 26 |
> up to go into the next app-vim/gentoo-syntax. One of the things that |
| 27 |
> this lets me do is highlighting of (some kinds of) mistakes. What're the |
| 28 |
> most common screwups that people see? My current list is: |
| 29 |
> |
| 30 |
> * CFLAGS: -0[123s] instead of -O[123s], -ffast-math globally |
| 31 |
> |
| 32 |
> * USE: +foo flags |
| 33 |
> |
| 34 |
> * ACCEPT_KEYWORDS: random duff entries (syntax is easy enough for me to |
| 35 |
> validate the whole thing) |
| 36 |
> |
| 37 |
> * CHOST: there're certain settings that will cause glibc to fall over |
| 38 |
> and die that people who are used to debian/sparc will sometimes use |
| 39 |
> |
| 40 |
> * MAKEOPTS: I'm kinda tempted to highlight "-j N" (where n is integer) |
| 41 |
> |
Archives