Autosync - Refreshed

[FlyGoat]

This work is based on #1803.

Note that only MIPS is included in this PR. As that's the arch which I most familiar.

I'll work on RISC-V, sparc and others after got this portion merged.

[FlyGoat]

\cc @XVilka @kabeor

[kabeor]

@FlyGoat I will create a new branch for this feature tomorrow and plz rebase after that

[XVilka]

Were there any changes required on the TableGen backend side?

• https://github.com/capstone-engine/llvm-capstone/tree/auto-sync-update/llvm/utils/TableGen/Capstone

Since this code is autogenerated, I suspect it might

[kabeor]

plz rebase to https://github.com/capstone-engine/capstone/tree/autosync branch

[FlyGoat]

plz rebase to https://github.com/capstone-engine/capstone/tree/autosync branch

Done

Were there any changes required on the TableGen backend side?

No actually I started with files generated by Phosphorus15.

[FlyGoat]

Any update?

[kabeor]

@FlyGoat hi,as I mentioned in #1803 (comment), plz "do not add any .inc file in this PR, we should PR that alone".
Besides plz create a new pr to rebase llvm to released version first, then merge this pr.

[kabeor]

I think it's better if we can auto generate .inc files and releasing by github action

[XVilka]

@FlyGoat how is it going? Do you need any help?

[FlyGoat]

Ah my fault, I thought I had updated PR but actually not. Will push tomorrow.

Sorry.

[kabeor]

@FlyGoat Any update?

[FlyGoat]

@kabeor sorry for the latency, PR sent on llvm-capstone repo.

[FlyGoat]

Btw it seems impossible to sepreate code change and auto generated table change without breaking build between them.

[XVilka]

@FlyGoat I think then splitting the change into two sets of commits - one without autogenerated files that breaks the build and the second with updated autogenerated files that fix the build. @kabeor what do you think about this approach? It will make the review easier as well.

[kabeor]

@XVilka yeah，agree with that

[XVilka]

@FlyGoat how is it going?

[Rot127]

Just as notification: #1803 (comment)

[XVilka]

Probably could be closed in favor of #1949
@kabeor