Threadwell
Search
Search
Dark mode
Light mode
Reader mode
Explorer
Recent Notes
DemetriSpanos
Oct 06, 2026
archive
twitter
demetrispanos
NOTimothyLottes
Oct 06, 2026
archive
twitter
notimothylottes
bonzajplc
Oct 06, 2026
archive
twitter
bonzajplc
fiigii
11 items with this tag.
Aug 29, 2026
fiigii
archive
twitter
fiigii
Aug 25, 2026
If you’ve done serious compiler optimization work on a product with a sufficiently large user base, you’ve almost certainly seen bug reports like this:
archive
twitter
fiigii
Aug 26, 2026
@fiigii In our case, it's also often just blocked by the memory model requirements.
archive
twitter
fiigii
Aug 26, 2026
@fiigii E-graphs seem to be one way to address this, allowing the compiler to explore different optimisation choices and then distilling the best outcome out of them.
archive
twitter
fiigii
Aug 26, 2026
@fiigii more language-tier facilities to drive it? stuff like [[likely]] and etc.
archive
twitter
fiigii
Aug 26, 2026
@fiigii Don't worry it exists, they call it PGO, we had it in Visual C++ 2005.
archive
twitter
fiigii
Aug 26, 2026
@fiigii llvm is flexible, you can have custom passess, and it has, I don't know, 2000 knobs, what you can tweak the "heuristics" with as you wish, good luck
archive
twitter
fiigii
Aug 26, 2026
@fiigii very true!
archive
twitter
fiigii
Aug 26, 2026
@fiigii > This needs to change.
archive
twitter
fiigii
Aug 26, 2026
"Why should I care about the performance of everyone else’s programs?" Because then they will be forced to do likewise.
archive
twitter
fiigii
Aug 26, 2026
@fiigii I think inline assembly should be the right answer, but compilers don’t have a really great support of it, apart from Odin right now
archive
twitter
fiigii