Daniel Xu
Daniel Xu
The kernel seems to dump most of their pointer-y printf extensions under `%p` prefix: https://www.kernel.org/doc/html/latest/core-api/printk-formats.html#pointer-types Perhaps we can try to follow that? A lot of kernel devs use bpftrace so...
@chengshuyi good point. Improving printf() to use percpu map memory would be a nice change. I don't know off the top of my head how much work it'd be, though....
Closing due to inactivity.
We have a script that generates .deb's: https://github.com/facebookincubator/below/blob/main/scripts/build_deb.sh Is that sufficient for you?
See pointer in https://github.com/bpftrace/bpftrace/issues/3331#issuecomment-2610133171
I suspect this would be a good candidate for exposing as a helper in kernel. There’s a lot of clever tricks for string searching and they can be accelerated with...
Yeah here's the optimized `strstr()` in the kernel. Despite naming, I believe it's also used for x86-64: https://github.com/torvalds/linux/blob/master/arch/x86/lib/strstr_32.c Should be an easy win (as well as being an easy kernel...
> > looks like maps represent all integers as i64. is that why maps don't have this issue? do we have any desire to unify the behavior? > > Good...
(First posted to discord) Here's the script to merge weights: [merge_tinker_adapter_to_hf_model.py](https://github.com/user-attachments/files/23468925/merge_tinker_adapter_to_hf_model.py) At some point we'll offer an API for this.
I think what @JakeHillion has laid out makes a lot of sense to me.