[cfe-dev] [llvm-dev] Sharing MemoryBuffers between front ends and LLVM

Hal Finkel via cfe-dev cfe-dev at lists.llvm.org
Thu Mar 16 12:03:54 PDT 2017


On 03/16/2017 12:29 PM, Roger Ferrer Ibanez via llvm-dev wrote:
> Hi all,
>
> I'm implementing interleaved source in assembly output. Early reviews raised the concern that the current implementation will be opening files (using a llvm::MemoryBuffer) that are likely to be in the memory of the front end (commonly clang but I think we want this to be front end agnostic).
>
> I'm now exploring ideas to avoid reopening files and let LLVM reuse the files the FE had to open.

I think this is a good idea.

>
> I am assuming that the front end will use llvm::MemoryBuffer (e.g.: clang does indirectly through clang::SourceManager).
>
> So for buffers related to named files (including stdin, which does not have name and is handled in a special way) we could have in the LLVM context a MemoryBufferRegistry. The idea is to add new creators of MemoryBuffer (the ones that work on named files and stdin) that can be passed a reference to that llvm::MemoryBufferRegistry. MemoryBuffer objects would register/deregister themselves at creation/destruction. This registry can then be used as a cache of already opened files from which retrieve a reference to the MemoryBuffer itself using the file path. These new interfaces would be opt-in for all users of MemoryBuffer.
>
> Back to my case,  the new AsmPrinterHandler could now use the MemoryBufferRegistry of the LLVM context. If there is none or the memory buffer associated to a file path has been already deregistered (or was never registered e.g. because we are using a .ll file directly), it would open the file as usual, otherwise it would reuse the registered MemoryBuffer.
>
> I see a few downsides of this approach, though.
>
> It overlaps a bit with the existing SourceManager in clang which already does some caching work through the clang::ContentCache class. At first the cache seems hard to abstract away as it uses clang::FileEntry and looks pretty tailored for clang needs.

I don't see why this is hard to abstract? As you describe it, it sounds 
like we just need to define some callback that can be implemented by the 
FE. For example, we could add the following to LLVMContext:

typedef ErrorOr<MemoryBufferRef> (*FileBufferCallbackTy)(StringRef 
FileName, void *Context);

void setFileBufferCallback(FileBufferCallbackTy Callback, void *Context 
= nullptr);
FileBufferCallbackTy getFileBufferCallback() const;

Using this, we can create an additional abstraction in LLVM that will 
open a given file and return a memory buffer, asking the FE first if 
there's a callback, or otherwise just opening the file directly.

>
> Also, assuming that the front end is using a MemoryBuffer may be a too strong requirement, in particular for FE's that are mostly unaware of LLVM except for a final LLVM codegen pass. This would mean that the files would be reopened even if they are already in the memory of the FE.

Given that you can construct a MemoryBuffer from a StringRef, and a 
StringRef can be constructed around an arbitrary buffer, if we wanted to 
create a wrapper around this interface usable by a less-LLVM-aware FE, 
that seems possible. I don't know how important this is right now.

>
> Finally the file path may not even be a good identifier to reuse MemoryBuffer objects.

It might not be, but I think that we can make clear that the keys used 
are intended to correspond to the file names from the debug info (or 
module name).

Thanks again,
Hal

>
> Thoughts?
>
> Thank you,
> Roger
>
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev

-- 
Hal Finkel
Lead, Compiler Technology and Programming Languages
Leadership Computing Facility
Argonne National Laboratory




More information about the cfe-dev mailing list