I’m still really enjoying this series. But I noticed that there are some divergences between the code in the post and the code in your repo. It threw me off a bit when I copy-and-pasted the new tests but missed a change in (I think it was) Compile_entry.
Uh oh! Do you know specifically what caught you? I am trying to avoid differences, except for some small things like function attributes. I want to fix whatever you ran into.
I went back and looked again… looks like I lost sync on the instruction encoding blog. I updated a couple of the Emit_xxx functions but not all of them. Then when I did, my tests were looking for the wrong bytes.
Oh, okay! Yeah, the instruction encoding thing was one huge overhaul. The follow-up post also changes every rbp reference to rsp. In any case, thank you for bringing this up.
Is your follow-along implementation public? I would love to link to it!
Please let me know what feedback you have!
I’m still really enjoying this series. But I noticed that there are some divergences between the code in the post and the code in your repo. It threw me off a bit when I copy-and-pasted the new tests but missed a change in (I think it was)
Compile_entry.Uh oh! Do you know specifically what caught you? I am trying to avoid differences, except for some small things like function attributes. I want to fix whatever you ran into.
I went back and looked again… looks like I lost sync on the instruction encoding blog. I updated a couple of the Emit_xxx functions but not all of them. Then when I did, my tests were looking for the wrong bytes.
Chalk it up to user error. :-)
Oh, okay! Yeah, the instruction encoding thing was one huge overhaul. The follow-up post also changes every
rbpreference torsp. In any case, thank you for bringing this up.Is your follow-along implementation public? I would love to link to it!
That was it! The
rbptorspchange.I’ve made my repo public now. https://github.com/mtnygard/complisp
Sweet!