Oh GCC I am so disappointhttps://godbolt.org/g/wZgDtA
-
-
Replying to @mattgodbolt
That makes so little difference! (and it's so difficult to avoid given the mess the ISA is)
1 reply 0 retweets 1 like -
Replying to @volatile_void
Fair point :) The referenced library seems to think it worthwhile writing the code longhand (which I was riling against in an internal discussion at work)
1 reply 0 retweets 0 likes -
Replying to @mattgodbolt
Having thought more about it, GCC's version use a 32x32->64 multiplication, while Clang's uses 64x64->128, right? If so, I'm almost sure GCC's version is faster on some amd64 processors.
1 reply 0 retweets 1 like -
Replying to @volatile_void
Yeah I was just thinking that too! Thanks for taking a closer look! :D
2 replies 0 retweets 0 likes
Yes, GCC's version looks clearly better to me. But it should do it consistently for the version written with *. 32-bit operand values (cast up to 64-bit type) should not generate a 64x64 mul, only a 32x32 one.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.