perf: use trunc(dval(&u)) instead of (Long)(dval(&u)) #22676
Conversation
…per-digit dependency chain in float register
|
This file is old and complicated and previous changees were rejected. IMHO it would make more sense to replace this with a known correct and fast alternative like https://github.com/kolemannix/ffc.h |
|
I'm sure replacing the whole logic would be a good idea, but I don't see why the file being old should lead to changes being rejected. It's a rather minimal change with decent performance gains (especially on aarch64) for a common operation and doesn't make replacing the logic entirely in the future any harder. |
On the contrary, Arnaud fixed some lock contention issues in this file. I don't see a problem. |
|
When I tried cleaning up this file a few years ago I was told no as updating it would be difficult (#5622). But I don't really care so whatever. |
… to keep per-digit dependency chain in float register
aarch64:
x86_64 (sse4.1):
LLM disclosure: optimisation found entirely by Fable, I verified, wrote the comment and benchmarked.