- User Since
- Jan 8 2015, 1:53 PM (455 w, 1 d)
Thu, Sep 7
Wed, Sep 6
LGTM for NVPTX side.
Tue, Sep 5
Aug 29 2023
Aug 28 2023
This patch breaks breaks two tests:
Aug 24 2023
For some context about why it's needed see https://github.com/compiler-explorer/compiler-explorer/pull/5373#issuecomment-1687127788
The short version is that currently CUDA compilation is broken w/ clang with unpatched libstdc++. Ubuntu and Debian patch libstdc++ to avoid the problem, but this should be handled by clang.
Aug 23 2023
Aug 22 2023
Aug 21 2023
Aug 18 2023
Fixed test RUN lines
@fhahn who else should take a look at the patch?
Same reproducer but for CUDA: https://godbolt.org/z/WhjTMffnx
Aug 17 2023
Aug 14 2023
Aug 11 2023
Aug 10 2023
I think we're commingling two different things here.
- representing a value as a string
- printing something out for debugging purposes.
Aug 9 2023
Aug 7 2023
Do we really need it at all?
Sorry about the silence -- I was away for the last couple of weeks.
Jul 21 2023
Jul 20 2023
It's unfortunate that NVIDIA decided to break existing naming convention. :-/
Jul 18 2023
@rsmith Richard, PTAL. This needs your language lawyering expertise.
We should probably update documentation that C++ standard version for CUDA/HIP compilation now matches C++ default instead of previously used c++14.
Jul 17 2023
Jul 13 2023
Jul 12 2023
Can anyone push this?
Jul 11 2023
Jul 10 2023
Ah, never mind. We're only dealing with shared variables here. The patch is fine.
This may be potentially problematic for CUDA as we may need to refer to the global variable from the host side. E.g. via cudaMemcpyFromSymbol/cudaMemcpyToSymbol.
Looks good in general.
Looking at CUDA headers, it appears that changing only compiler-generated-glue may be insufficient. A lot of other CUDA API calls need to be changed to _ptsz variant and for that we need to have CUDA_API_PER_THREAD_DEFAULT_STREAM defined.
Jun 30 2023
clang-formatted whole file.
Addressed comments. Reformatted the changes.
Jun 29 2023
LGTM in general, with a nit.
LGTM in general with a minor suggestion.
Jun 28 2023
Jun 23 2023
Jun 22 2023
Added more tests for vector loads w/ small alignment.
Fixed an error in lowering extending vector loads.
Preserve behavior for lowering integer vectors.
Jun 21 2023
Jun 20 2023
My understanding of nuances here is not sufficient for a meaningful review. @nhaehnle's LGTM works for me.
Jun 16 2023
Just a FYI -- in the past, running too many tests in parallel caused some tests to hang. I've empirically settled on - j4 on cuda bots. Things may have improved in recent CUDA and driver versions, but I didn't push parallel tests further.
Jun 15 2023
Jun 14 2023
Jun 13 2023
Apparently this change causes a hang in https://github.com/llvm/llvm-project/blob/cdcbef1b14582f5b5b296bec750f60a1629ecc32/llvm/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp#L3410
It's unfortunate that we don't seem to have a way to lower unreachable via normal lowering mechanisms. AFAICT we don't even have a selection DAG node type for it.
Jun 12 2023
LGTM in principle.