On 2022-03-07 at 23:32:42, Taylor Blau wrote: > On Mon, Mar 07, 2022 at 11:25:52PM +0000, brian m. carlson wrote: > > In the block SHA-1 code, we have special assembly code for i386 and > > amd64 to perform rotations with assembly. This is supposed to help pick > > the correct rotation operation depending on which rotation is smaller, > > which can help some systems perform slightly better, since any circular > > rotation can be specified as either a rotate left or a rotate right. > > However, this isn't needed, so we should remove it. > > At -O1 or higher (at least on GCC) this optimization is indeed > performed. Here's a Godbolt example that shows this: > > https://godbolt.org/z/9zMP93hr1 > > so I agree that this code isn't helping us at all. And in the > meantime... Thanks for providing a link. I also did a similar test there with slightly different code (and unfortunately closed the window before saving the link) but it demonstrated the same thing: that the compiler can optimize this case adequately. My (substantially) past experience with testing GCC in this case has shown the same thing. > > The downside of using this code, however, is that it uses a GCC > > extension, which makes the compiler complain when using -pedantic unless > > it's prefixed with __extension__. We could fix that, but since it's > > not needed, let's just remove it. We haven't noticed this because > > almost everyone uses the SHA1DC code instead, but it still shows up for > > some people. > > ...it makes it impossible to compile git if you have > `BLK_SHA1=YesPlease` and `DEVELOPER=1` in your environment. So I am > happy to see this go. > > On another note: missing Signed-off-by? I'll send an otherwise unchanged v2 with that in a second. -- brian m. carlson (he/him or they/them) Toronto, Ontario, CA
Attachment:
signature.asc
Description: PGP signature
