News for hackers. Stuff that matters.
Re:GCC x86 Performance Hints (2012) by jcranmer on Friday August 21, 2020 @10:40AM (#24235259)
> So that means that you tell the compiler to assume that SSE is available on the target. By default GCC outputs code that's compatible with the baseline, which is perfectly reasonable IMO. Same reason why -march=native is also not the default, it is assumed that you may want to ship your binaries to other computers.

If your x86 computer was made after the US invaded Iraq, it supports SSE2 instructions.

Re:GCC x86 Performance Hints (2012) by simias on Friday August 21, 2020 @11:52AM (#24235884)
And I'm currently writing code for a 32bit CPU that was first introduced in 2000. It's not x86 so SSE is irrelevant, but my point is that GCC is routinely used to build millions if not billions of lines of code, they can't just YOLO-deprecate things as if it were a javascript framework.
Re:GCC x86 Performance Hints (2012) by cozzyd on Friday August 21, 2020 @04:29PM (#24238571)
It's also possible someone may prefer the 80-bit intermediate precision of the x87 registers. I guess.
Re:GCC x86 Performance Hints (2012) by jcranmer on Friday August 21, 2020 @06:01PM (#24239331)
long double in either x87 or SSE mode will force the use of 80-bit fp values.