[rep_lodsb]

So you can have bit arrays of any length in memory, rather than just 32 bits in a register.

[cmovq]

That makes sense. LLVM could probably do better here by using the memory operand version:

https://godbolt.org/z/jeqbaPsMz

[jxors]

The memory operand version tends to be as slow or slower than the manual implementation, so LLVM is right to avoid it.

[cmovq]

Right, it has much worse throughput:

Memory: https://uica.uops.info/tmp/f022a3c0a70e4ae5ab3588ebe65fd2a5_...

Register: https://uica.uops.info/tmp/e80e60e0c4914955b11dc6590711c1b8_...

[ack_complete]

Don't think the memory operand version would work here. If I understand the x86 architectural manual description, the 32-bit operand form interprets the bit offset as signed. A 64-bit operand could work around that but then run into issues with over-read due to fetching 64 bits of data.