Make a module out of debuginfo

[nrc]

Closes#20780

r? @michaelwoerister

I'm sure this could be done better with deeper knowledge of debuginfo, but this seems like a good start.

[michaelwoerister]

Cool! I'll take a look at this over the weekend.

[michaelwoerister]

I've taken a look at this and overall I think this is the right direction to go in. However, I would factor some things a little differently:

Metadata creation is the most complicated thing in this module. Especially for type metadata many things interconnect and coupling is rather tight. However, only the type_metadata() function really needs to be exposed to the rest of the debuginfo module, while the specialized implementations called by type_metadata() are more or less implementation details, including the TypeMap and handling of recursive types. This PR splits these things up into various files though (mod.rs, adt.rs, and metadata.rs). I would merge adt.rs into metadata.rs (maybe make adt.rs a private sub-module of metadata.rs) and move the TypeMap, UniqueTypeId, RecursiveTypeDescription, and create_and_register_recursive_type_forward_declaration into this new metadata.rs file. This way, we should be able to keep most things in metadata.rs private. Only type_metadata(), file_metadata(), scope_metadata(), and compile_unit_metadata() would be public.

Some nits:

• I think create.rs doesn't quite carry its weight. I'd get rid of it and move is_node_local_to_unit() and create_DIArray() to utils.rs, and declare_local() to mod.rs.
• I'd rename types.rs to naming.rs or type_naming.rs or something to that effect because it only deals with naming.
• I'd move create_scope_map() from utils.rs to its own file since this is a rather important part of creating debuginfo (although it is just one big function).
• Source location handling could also be moved from mod.rs into its own module but that's not a high priority.

Apart from the above, this looks good to me. I like that the GDB stuff get's its own module and that mod.rs gets condensed down to the few things that are actually needed by other parts of the compiler.

[bors]

☔ The latest upstream changes (presumably #23606) made this pull request unmergeable. Please resolve the merge conflicts.

[nrc]

@michaelwoerister ping for r? All changes made as suggested.

[michaelwoerister]

@bors r+ 7bfb5ed

[michaelwoerister]

Thanks!

[bors]

⌛ Testing commit 7bfb5ed with merge b2998a7...

[bors]

💔 Test failed - auto-mac-32-opt

[alexcrichton]

@bors: retry

On Wed, Apr 29, 2015 at 2:52 PM, bors notifications@github.com wrote:

[image: 💔] Test failed - auto-mac-32-opt
http://buildbot.rust-lang.org/builders/auto-mac-32-opt/builds/4760

—
Reply to this email directly or view it on GitHub
#24762 (comment).