Daniel Lemire
Daniel Lemire
Fantastic. These are very impressive numbers. > What does one need to do/which files to touch/what tools are there, to add a new architecture to simdutf? Firstly, we have an...
@camel-cdr I am getting the impression that we'll have at least one code reviewer (@clausecker). :-)
> I'm not sure what the correct way to check for it is, but [hwprobe](https://www.kernel.org/doc/html/v6.7-rc8/arch/riscv/hwprobe.html) looks like it would work for checking V support. So we are assuming that RISC-V...
@camel-cdr You referred to icelake in your issue, and the icelake kernel is different, in part because it benefits from compress instructions (VBMI2).
Your understanding is correct. We proceed in this manner because we want to make the library available in a single-header form, without making assumptions about the build system.
Not only would it be welcome, but we would be happy to help.
Punycode is generally applied on labels and labels are limited to 63 bytes...
> will enable the linker to remove any unused sections in the binary that consumes simdutf statically. This may benefit the end-user of libsimdutf.a more than it does libsimdutf.a itself....
That's a great project. It will take some time, but given that every system with SVE has NEON, one could start with the NEON kernel (called arm64) and create a...
This PR increases the length of the data dependency chain, so it is at risk of leading to lower instructions per cycle performance. Let us try it out. Build the...