Displaying 6 results from an estimated 6 matches for "cxx_devirtu".
2018 Mar 29
0
[cfe-dev] RFC: Devirtualization v2
...produce IR that gets linked together (e.g. clang vs. swift).
>
> How about this approach:
> - Instead of taking a meaningless !{} argument, invariant.group takes a string argument which identifies a metadata-dependent optimization. In your case, it would be something like !"clang.cxx_devirtualization".
> - Functions have a "supported optimizations" list which declares all the metadata-reliant optimizations they promise to have correct metadata for. So e.g. clang++ would list "clang.cxx_devirtualization" on every single function it compiled, regardless of...
2018 Mar 30
2
[cfe-dev] RFC: Devirtualization v2
...together (e.g. clang vs. swift).
>>
>> How about this approach:
>> - Instead of taking a meaningless !{} argument, invariant.group takes a
>> string argument which identifies a metadata-dependent optimization. In
>> your case, it would be something like !"clang.cxx_devirtualization".
>> - Functions have a "supported optimizations" list which declares all
>> the metadata-reliant optimizations they promise to have correct metadata
>> for. So e.g. clang++ would list "clang.cxx_devirtualization" on every
>> single funct...
2018 Mar 29
2
[cfe-dev] RFC: Devirtualization v2
...IR that gets linked together (e.g. clang vs. swift).
>
> How about this approach:
> - Instead of taking a meaningless !{} argument, invariant.group takes a
> string argument which identifies a metadata-dependent optimization. In
> your case, it would be something like !"clang.cxx_devirtualization".
> - Functions have a "supported optimizations" list which declares all the
> metadata-reliant optimizations they promise to have correct metadata for.
> So e.g. clang++ would list "clang.cxx_devirtualization" on every single
> function it compiled,...
2018 Mar 28
0
[cfe-dev] RFC: Devirtualization v2
...frontends that produce IR that gets linked together (e.g. clang vs. swift).
How about this approach:
- Instead of taking a meaningless !{} argument, invariant.group takes a string argument which identifies a metadata-dependent optimization. In your case, it would be something like !"clang.cxx_devirtualization".
- Functions have a "supported optimizations" list which declares all the metadata-reliant optimizations they promise to have correct metadata for. So e.g. clang++ would list "clang.cxx_devirtualization" on every single function it compiled, regardless of wheth...
2018 Mar 30
0
[cfe-dev] RFC: Devirtualization v2
...that gets linked together (e.g. clang vs. swift).
>>
>> How about this approach:
>> - Instead of taking a meaningless !{} argument, invariant.group takes a string argument which identifies a metadata-dependent optimization. In your case, it would be something like !"clang.cxx_devirtualization".
>> - Functions have a "supported optimizations" list which declares all the metadata-reliant optimizations they promise to have correct metadata for. So e.g. clang++ would list "clang.cxx_devirtualization" on every single function it compiled, regardless...
2018 Mar 19
4
RFC: Devirtualization v2
Hi folks,
here is a link to the proposal that we've been working on lately:
https://docs.google.com/document/d/16GVtCpzK8sIHNc2qZz6RN8amICNBt
vjWUod2SujZVEo/edit?usp=sharing
But for the record, I also paste it here. Feedback will be really
appreciated!