FuelLabs
Add newer control flow opcodes
@dmihal Can you test if this works for you? If you have a unit test, please push to the branch, or alternatively share it here so that i can add it.
Looking into this now, some code is missing, but I'm looking at adding it now.
Some notes:
- Bigint math opcodes like
WQCMare also missing - Could we have some CI in the future that automatically checks that new opcodes are available in assembly?
These should be added to https://github.com/FuelLabs/sway/blob/master/test/src/e2e_vm_tests/test_programs/should_fail/asm_disallowed_opcodes/src/main.sw should they not?
These should be added to https://github.com/FuelLabs/sway/blob/master/test/src/e2e_vm_tests/test_programs/should_fail/asm_disallowed_opcodes/src/main.sw should they not?
This is a good point. But it looks like @dmihal is actually using these in assembly.
Well it looks like we have a test that enforced the rule that you're not allowed to use control flow opcodes in asm blocks, which I guess is why we didn't bother to implement these opcodes previously.
Now I'm open to the possibility of loosening that rule given asm blocks are explicitly unsafe, but first I would like to understand why we forbade that inclusion in the first place. The reason for it may be important.
Yes, we need these opcodes for our basic version of LDC
I imagine that control flow opcodes were blocked previously because they essentially allow you to jump "out" of the assembly block. However given that we currently have a real need to use them, and as @IGI-111 mentioned asm is already unsafe, I'd recommend us loosening this rule
I'm closing this in favor of https://github.com/FuelLabs/sway/issues/5563 since it looks like we can do without jumps using that intrinsic