Me gusta!

Accepted subject to the changes I have proposed.

This is above my paygrade as far as numerical programming goes, but let's just hope it's fine.

No-brainer.
Looks good to me! Approved for commit.

I control the upstream, so I can adjust things to better suit ports (read: avoiding a local patch). FreeBSD ports is the first class target of this software, and currently I'm not worried about supporting any other Operating System or build environment (this is squarely for my work with the Framework Laptop 12 system integration -- largely for supporting the media function keys on the Framework keyboards and providing immediate responsive feedback to the user when a key is pressed).

This one may need a Makefile patch to remove -I/usr/local/include and -L/usr/local/lib. Check if it then compiles without USES=localbase; after all, it seems to be using pkg-config to find its dependencies. I recommend USES=localbase:ldflags instead of USES=localbase.

Thank you for your submission.

You are allowed to make changes, and yes we're working on this together.
It's just that I already spent time adapting the code to our conventions, so you changing it all back seems strange.

In D59462#1365088, @kib wrote:

In D59462#1365021, @fuz wrote:

Oh yeah, using the archlevel framework sounds like a good idea! I didn't even think about that, as it's specific to libc. Other libm functions may profit in the future.

Do you have any notes about the updated patch?

BTW, I was not able to come up with a C expression that can be used to initialize the function table, so I resorted to filling it in assembler same as is done for strings.

Welcome back!

Restore the copyright header and please have a look at the other things, too.

Oh yeah, using the archlevel framework sounds like a good idea! I didn't even think about that, as it's specific to libc. Other libm functions may profit in the future.

There could be some performance issues related to the SSE/AVX transition state machine, but checking the Intel Software Optimization Manual, it seems like executing 128 bit AVX instructions seems unproblematic in any steady-state scenario.

Love it!

I've rewritten c8rtomb() using Björn Höhrmann's algorithm. I'll send you an e-mail with the details shortly.
Right now I am somewhat stumped about the stupidity of our locale implementation and will have to think about it some more.

CC @bdrewery who's e-mail address was changed to his project address in this patch set.

• (user@host.tld) -> Aq Mt user@host.tld

• @freebsd.org -> @FreeBSD.org

discard hunks pertaining contrib/ as per @ziraee's advice.

I'll have some time this weekend (hackathon) to look at this change set and try to find a better implementation.

Well it was previously a platform-independent definition in _stdint.h, so I don't see how it would become platform-dependent in the future. That said, it can always be changed if needed.

If the size is the same on all platforms, it should not be in the platform-specific _types.h files.

I trust you that you have checked that it is.

Looks good.
Approved for commit.

Good job!