[RFC] Interest in additional endianness documentation

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hello all!

Thank you to everyone in the community for building/working on such a
great tool! I am helping build a userspace implementation of eBPF and
following Dave's standardization process closely.

I was investigating how endianness affects the order of the components
of instructions in memory. In particular, the order of the src/dest
fields. In the IETF mailing list archives I found Jose's excellent
pointer to the implementation-defined nature of the order of bit
fields (https://mailarchive.ietf.org/arch/msg/bpf/I7abyqaiXj_DYuoLln5joMw90k0/)
and spent the better part of today digging through gcc's source code
to find the *actual* logic that controls the order (yes, I believe
that I ultimately found it!).

Nevertheless, given that there is a reliance on the behavior of gcc
and clang, would you be interested in having additional documentation
about it? I would love to write it but don't want to waste your time.
I was thinking:

1. Adding documentation about field order to include/uapi/linux/bpf.h
where the insn struct is defined.
2. Adding documentation to the design QA (Documentation/bpf/bpf_design_QA.rst)
3. Anything else that others might want.

I apologize for the long email and I just want to be helpful. If this
seems worthwhile, please let me know. Otherwise, I will defer and not
bother you all further.

Thanks!
Will




[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux