Re: FW: ebpf-docs: draft of ISA doc updates in progress

[Christoph Hellwig <hch@infradead.org>]

On Wed, Sep 14, 2022 at 02:22:51PM +0800, Shung-Hsi Yu wrote:
> As discussed in yesterday's session, there's no graceful abortion on
> division by zero, instead, the BPF verifier in Linux prevents division by
> zero from happening. Here a few additional notes:

Hmm, I thought Alexei pointed out a while ago that divide by zero is
now defined to return 0 following.  Ok, reading further along I think
that is what you describe with the pseudo-code below. 

> While BPF ISA only supports direct call BPF_CALL[1], technically there is an
> opcode 0x8d (BPF_JUMP | BPF_CALL | BPF_X) that has the indirect call
> semantic, and Clang emit such indirect call instruction if user attempt to
> compile with -O0.
> 
> I think it worth mentioning in this document for better clarity, perhaps
> simply saying that indirect call is not part of BPF ISA is enough.

Which brings up another question:  Do we need a list of opcodes
that someone else defined somewhere that are not considered valid
eBPF?  Or how do we get clang and gcc to stop producing invalid
eBPF might be the better question.