Summary:
The very first version of the `clang-linker-wrapper` used `--` as a
separator for the host and device arguments. I moved away from this
towards a commandline parsing implementation years ago but never got
around to officially removing this.
Summary:
The linker wrapper's job is to sort various embedded inputs into a list
of files that participate in a single link job. So far, this has been
completely 1-to-1, that is, each input file participates in exactly one
link job. However, support for AMD's target-id requires that one input
file may participate in multiple link jobs. For example, if given a
`gfx90a` static library and a `gfx90a:xnack+` object file input, we
should link the gfx90a` target into the `gfx90a:xnack+` job. These are
considered separate CPUs that can be mutually linked more or less.
This patch adds the necessary logic to make this happen. It primarily
reworks the logic to copy relevant input files into a separate list. So,
it moves construction of the final list of link jobs into the extraction
phase. We also need to copy the files in the case that it is needed more
than once, as the entire workflow expects ownership of said file.
Summary:
The 'ockl' bitcode library from the ROCm device library contains several
implementations of functions like `printf` and `malloc`. We currently do
not depend on these in the OpenMP toolchain, so we shouldn't be linking
them. The primary motivation behind this change is the library rewriting
calls to `printf` and pulling in other unused 'hostcall' routines.
The linker wrapper runs LTO internally, so it needs to know the
optimization level the user requested, if any. Previously this was only
done in `-foffload-lto` mode as we were assuming that this would enble
LTO. However, AMDGPU always performs LTO, and it's possible to run clang
on object files to link without passing this flag. So we should just
respect it always.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D158298
An upcoming change D156363 will change -### to exit with code 1 if
hasErrorOccurred. Some AMDGPU tests would exit with code 0 due to not
found GPU installation.
There are static checks on the target identifiers allowed in a single
TU. Previously theses checks were only applied to HIP even though they
should be the same for OpenMP targeting AMDGPU. Simply enable these
checks for OpenMP.
Reviewed By: JonChesterfield, yaxunl
Differential Revision: https://reviews.llvm.org/D152965
AMDGPU sometimes uses a novel formatting for their offloading
architecture called the target id. This merges the attributes and the
architecture name into a single string. Previously, we were passing this
as the canonical architecture name. This caused the linker wrapper to
fail to find relevant libraries and then pass an incalid CPU name. This
patch changes the handling in the offload packager to handle the
canonical architecture and then extract the features.
Reviewed By: yaxunl
Differential Revision: https://reviews.llvm.org/D150998
The AMDGPU target can only emit LLVM-IR, so we can always rely on LTO to
link the static version of the runtime optimally. Using the static
library only has a few advantages. Namely, it avoids several known bugs
and allows us to optimize out more functions. This is legal since the
changes in D142486 and D142484
Depends on D142486 D142484
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D142491
Currently, the behaviour of `-save-temps` changes the generated output
when offloading to AMDGPU. This is because we only have a single phase
and it contains the `-disable-llvm-passes` flags which results in
unoptimized bitcode. We need to make sure we generate another phase that
produces both the optimized and unoptimized bitcode. There used to be a
check that turned these phases into a no-op. But I believe it is more
correct to not generate them this way in the first place. Doing this
requires a bit of a hack, replacing an already generated phase action,
but it should be fine.
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D141440
Summary:
There's some static checks on the library, we can't do offloading with
`libgomp` for OpenMP. This patch specifies the library for the tests to
avoid this breaking tests.
Previously when using the `clang-offload-packager` we did not pass the
active offloading kinds. Then in Clang when we attempted to detect when
there was host-offloading action that needed to be embedded in the host
we did not find it. This patch adds the active offloading kinds so we
know when there is input to be embedded.
Reviewed By: tra
Differential Revision: https://reviews.llvm.org/D134189
Previously, we linked in the ROCm device libraries which provide math
and other utility functions late. This is not stricly correct as this
library contains several flags that are only set per-TU, such as fast
math or denormalization. This patch changes this to pass the bitcode
libraries per-TU using the same method we use for the CUDA libraries.
This has the advantage that we correctly propagate attributes making
this implementation more correct. Additionally, many annoying unused
functions were not being fully removed during LTO. This lead to
erroneous warning messages and remarks on unused functions.
I am not sure if not finding these libraries should be a hard error. let
me know if it should be demoted to a warning saying that some device
utilities will not work without them.
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D133726
The changes in D130020 removed all support for the old method of
compiling OpenMP offloading programs. This means that
`-fopenmp-new-driver` has no effect and `-fno-openmp-new-driver` does
not work. This patch removes the use and documentation of this flag.
Note that the `--offload-new-driver` flag still exists for using the new
driver optionally with CUDA and HIP.
Reviewed By: tra
Differential Revision: https://reviews.llvm.org/D133367
Recently OpenMP has transitioned to using the "new" driver which
primarily merges the device and host linking phases into a single
wrapper that handles both at the same time. This replaced a few tools
that were only used for OpenMP offloading, such as the
`clang-offload-wrapper` and `clang-nvlink-wrapper`. The new driver
carries some marked benefits compared to the old driver that is now
being deprecated. Things like device-side LTO, static library
support, and more compatible tooling. As such, we should be able to
completely deprecate the old driver, at least for OpenMP. The old driver
support will still exist for CUDA and HIP, although both of these can
currently be compiled on Linux with `--offload-new-driver` to use the new
method.
Note that this does not deprecate the `clang-offload-bundler`, although
it is unused by OpenMP now, it is still used by the HIP toolchain both
as their device binary format and object format.
When I proposed deprecating this code I heard some vendors voice
concernes about needing to update their code in their fork. They should
be able to just revert this commit if it lands.
Reviewed By: jdoerfert, MaskRay, ye-luo
Differential Revision: https://reviews.llvm.org/D130020
Previously an opt-in flag `-fopenmp-new-driver` was used to enable the
new offloading driver. After passing tests for a few months it should be
sufficiently mature to flip the switch and make it the default. The new
offloading driver is now enabled if there is OpenMP and OpenMP
offloading present and the new `-fno-openmp-new-driver` is not present.
The new offloading driver has three main benefits over the old method:
- Static library support
- Device-side LTO
- Unified clang driver stages
Depends on D122683
Differential Revision: https://reviews.llvm.org/D122831
This patch passes in the AMDPGU math libraries to the linker wrapper.
The wrapper already handles linking OpenMP bitcode libraries via the
`--target-library` option. This should be sufficient to link in math
libraries for the accompanying architecture.
Fixes#53526.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D119841
This patch passes in the AMDPGU math libraries to the linker wrapper.
The wrapper already handles linking OpenMP bitcode libraries via the
`--target-library` option. This should be sufficient to link in math
libraries for the accompanying architecture.
Fixes#53526.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D119841
Summary:
The name of the AMDGPU device library was changes. Previously it was
called 'libomptarget-amdgcn'. This patch changes fixes the tests to use
the new name of the library and adds a new flag with the same name.
This patch completely removes the old OpenMP device runtime. Previously,
the old runtime had the prefix `libomptarget-new-` and the old runtime
was simply called `libomptarget-`. This patch makes the formerly new
runtime the only runtime available. The entire project has been deleted,
and all references to the `libomptarget-new` runtime has been replaced
with `libomptarget-`.
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D118934
This patch changes the `-fopenmp-target-new-runtime` option which controls if
the new or old device runtime is used to be true by default. Disabling this to
use the old runtime now requires using `-fno-openmp-target-new-runtime`.
Reviewed By: JonChesterfield, tianshilei1992, gregrodgers, ronlieb
Differential Revision: https://reviews.llvm.org/D114890
Given D109057, change test runner to use the libomptarget-x-bc-path
argument instead of the LIBRARY_PATH environment variable to find the device
library.
Also drop the use of LIBRARY_PATH environment variable as it is far
too easy to pull in the device library from an unrelated toolchain by accident
with the current setup. No loss in flexibility to developers as the clang
commandline used here is still available.
Reviewed By: jdoerfert, tianshilei1992
Differential Revision: https://reviews.llvm.org/D109061
This fixes the 'unused linker option: -lm' warning when compiling
program with -c.
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D107952
Math libraries are linked only when -lm is specified. This is because
host system could be missing rocm-device-libs.
Reviewed By: JonChesterfield, yaxunl
Differential Revision: https://reviews.llvm.org/D105981
Previously clang would print a binary blob into the bundled file
for amdgcn. With this patch, it will instead print textual IR as
expected.
Reviewed By: JonChesterfield, ronlieb
Differential Revision: https://reviews.llvm.org/D102065
Change-Id: I10c0127ab7357787769fdf9a2edd4b3071e790a1
Previously clang would print a binary blob into the bundled file
for amdgcn. With this patch, it will instead print textual IR as
expected.
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D102065
The offload action is used in four different ways as explained
in Driver.cpp:4495. When -c is present, the final phase will be
assemble (linker when -c is not present). However, this phase
is skipped according to D96769 for amdgcn. So, offload action
arrives into following situation,
compile (device) ---> offload ---> offload
without -c the chain looks like,
compile (device) ---> offload ---> linker (device)
---> offload
The former situation creates an unhandled case which causes
problem. The solution presented in this patch delays the D96769
logic until job creation time. This keeps the offload action
in the 1 of the 4 specified situations.
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D101901
Remove emit-llvm-bc from addClangTargetOptions as it conflicts with -E for save-temps.
AMDGCN does not yet support linking object files so backend and assemble actions are
skipped, leaving LLVM IR as the output format.
Reviewed By: JonChesterfield, ronlieb
Differential Revision: https://reviews.llvm.org/D96769
This patch uses the existing logic of CUDA for searching libomptarget
and extracts it to a common method.
Reviewed By: JonChesterfield, tianshilei1992
Differential Revision: https://reviews.llvm.org/D96248
This patch adds AMDGPUOpenMPToolChain for supporting OpenMP
offloading to AMD GPU's.
Originally authored by Greg Rodgers
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D94961
