Rebase the change and fixes some format issues.

Resolve clang-tidy warnings.

I am commandeering this work as original author stopped working this project.

Are there any more comments?

In D85678#2211912, @MaskRay wrote:

In D85678#2211909, @haowei wrote:

In D85678#2211865, @MaskRay wrote:

This change moves ELFObjHandler and other elfabi code to a new IFS library so it can be used by both llvm-elfabi and llvm-ifs tool for reading and writing ELF objects. Previously I move it to TextAPI but it causes cyclic dependencies.

I can't find any explanation why the new file hierarchy works better and why 'IFS' instead of a more specific name (.e.g InterfaceStub) is picked.

In my first try I put it in TextAPI library and it causes a dependency loop as Object library is depends on TextAPI. Otherwise I think TextAPI is a better place for these files.

I think the sole reason that Object depends on TextAPI is due to D66159. Have you tried contacting the author about thinking whether we should continue organizing the two libraries this way?

I am OK with InterfaceStub name. It explains the purpose clearer.

If you look at existing library names. Abbreviation is not common, so IFS should probably be expanded.

In D85678#2211865, @MaskRay wrote:

This change moves ELFObjHandler and other elfabi code to a new IFS library so it can be used by both llvm-elfabi and llvm-ifs tool for reading and writing ELF objects. Previously I move it to TextAPI but it causes cyclic dependencies.

I can't find any explanation why the new file hierarchy works better and why 'IFS' instead of a more specific name (.e.g InterfaceStub) is picked.

In D85678#2211863, @MaskRay wrote:

I would like to improve llvm-ifs tool on ELF object handling. Right now it uses yaml2obj which is not robust.

How is it not robust?

correct formating.

Add 1 line fix for the constraint on allocated handle in function allocateSingleHandle at line 647

Superseded by D35968

Thanks for reviewing this patch. I have modified the checker according NoQ's suggestions and refactored some long functions.

In D33263#756900, @zaks.anna wrote:

Looks good. Thank you!
Do you have commit access or should we commit?