Add support for extensible dialects, which are dialects that can be
extended at runtime with new operations and types.
These operations and types cannot at the moment implement traits
or interfaces.
Differential Revision: https://reviews.llvm.org/D104554
A data transfer statement must have REC= in its control list
if (and only if) the unit was opened with ACCESS='DIRECT'.
The runtime wasn't catching this error, but was just silently
advancing to the next record as if the access were sequential.
Differential Revision: https://reviews.llvm.org/D120838
Advancement to new output lines was taking fixed-sized direct-access
and internal character array element lengths into account, but not
RECL= settings from OPEN statements.
Differential Revision: https://reviews.llvm.org/D120837
Until Zfinx is supported in CodeGen we need to convert all Zfinx
register classes to GPR.
Remove the zfinx-types.ll test which didn't test anything meaningful
since -mattr=zfinx isn't implemented completely in llc.
Follow up to D93298.
Clang is crashing on the following statement
char var[9];
__asm__ ("" : "=r" (var) : "0" (var));
This is similar to existing test: crbug_999160_regtest
The issue happens when EmitAsmStmt is trying to convert input to match
output type length. However, that is not guaranteed to be successful all the
time and if the statement itself is invalid like having an array type in
the example, we should give a regular error message here instead of
using assert().
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D120596
This patch enables PLT analysis for aarch64. It is used by the static
relocations in order to provide final symbol address of PLT entry for some
instructions like ADRP.
Vladislav Khmelevsky,
Advanced Software Technology Lab, Huawei
Differential Revision: https://reviews.llvm.org/D118088
Name resolution was properly probing the table of unrestricted
specific intrinsics to find "abs", but failing to capture the
result type and save it in the created symbol table entry.
Differential Revision: https://reviews.llvm.org/D120749
This patch enables the lowering of various allocatable assignements
for character type and numeric types.
This patch is part of the upstreaming effort from fir-dev branch.
Depends on D120819
Reviewed By: PeteSteinfeld, schweitz
Differential Revision: https://reviews.llvm.org/D120820
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
Co-authored-by: Jean Perier <jperier@nvidia.com>
The upstream project ships CMake rules for building vanilla gtest/gmock which conflict with the names chosen by LLVM. Since LLVM's build rules here are quite specific to LLVM, prefixing them to avoid collision is the right thing (i.e. there does not appear to be a path to letting someone *replace* LLVM's googletest with one they bring, so co-existence should be the goal).
This allows LLVM to be included with testing enabled within projects that themselves have a dependency on an official gtest release.
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D120789
Root issue which triggered the revert was fixed in 689bab. No changes in the reapplied patch.
Original commit message follows:
SLP currently schedules all instructions within a scheduling window which stretches from the first instr
uction potentially vectorized to the last. This window can include a very large number of unrelated instruct
ions which are not being considered for vectorization. This change switches the code to only schedule the su
b-graph consisting of the instructions being vectorized and their transitive users.
This has the effect of greatly reducing the amount of work performed in large basic blocks, and thus greatly improves compile time on degenerate examples. To understand the effects, I added some statistics (not planned for upstream contribution). Here's an illustration from my motivating example:
Before this patch:
704357 SLP - Number of calcDeps actions
699021 SLP - Number of schedule calls
5598 SLP - Number of ReSchedule actions
59 SLP - Number of ReScheduleOnFail actions
10084 SLP - Number of schedule resets
8523 SLP - Number of vector instructions generated
After this patch:
102895 SLP - Number of calcDeps actions
161916 SLP - Number of schedule calls
5637 SLP - Number of ReSchedule actions
55 SLP - Number of ReScheduleOnFail actions
10083 SLP - Number of schedule resets
8403 SLP - Number of vector instructions generated
I do want to highlight that there is a small difference in number of generated vector instructions. This example is hitting the bailout due to maximum window size, and the change in scheduling is slightly perturbing when and how we hit it. This can be seen in the RescheduleOnFail counter change. Given that, I think we can safely ignore.
The downside of this change can be seen in the large test diff. We group all vectorizable instructions together at the bottom of the scheduling region. This means that vector instructions can move quite far from their original point in code. While maybe undesirable, I don't see this as being a major problem as this pass is not intended to be a general scheduling pass.
For context, it's worth noting that the pre-scheduling that SLP does while building the vector tree is exactly the sub-graph scheduling implemented by this patch.
Differential Revision: https://reviews.llvm.org/D118538
While a collection of allocas are technically vectorizeable - by forming a wider alloca - this was not a transform SLP actually knows how to do. Instead, we were forming a bundle with missing dependencies, and then relying on the scheduling code to preserve program order if multiple instructions were scheduleable at once. I haven't been able to write a test case, but I'm 99% sure this was wrong in some edge case.
The unknown op case was flowing down the shufflevector path. This did result in some splat handling being lost with this change, but the same lack of splat handling is visible in a whole bunch of simple examples for the gather path. I didn't consider this interesting to fix given how narrow the splat of allocas case is.
Summary:
This patch correctly handles the `--sysroot=` option when passed to the
linker wrapper. This allows users to correctly find libraries that may
contain offloading code if using this option.
LLVM defines several default datalayouts for integer and floating point types that are not being considered when importing into MLIR. This patch remedies this.
Reviewed By: ftynse
Differential Revision: https://reviews.llvm.org/D120832
The verify call was taking 50% of the compile time in our internal LLVM
fork when trying to unroll many loops.
Differential Revision: https://reviews.llvm.org/D113028
Instead of emitting 0 > Hi, emit Hi < 0. If Hi needs to be expanded again
this will allow the special case for sign bit tests in ExpandIntOp_SETCC
to trigger.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D120761
This miscompile was introduced in D119527.
This was a special pattern for rotate+bswap on RV32. It doesn't
work for RV64 since the rotate needs to be half the bitwidth. The
equivalent pattern for RV64 is ROTR ((GREV x, 56), 32) so match
that instead.
This could be generalized further as noted in the new FIXME.
Reviewed By: Chenbing.Zheng
Differential Revision: https://reviews.llvm.org/D120686
https://reviews.llvm.org/D120423 replaced the use of stacksave/restore with memref.alloca_scope, but kept the save/restore at the same location. This PR places the allocation scope within the wsloop, thus keeping the same allocation scope as the original scf.parallel (e.g. no longer over stack allocating).
Reviewed By: ftynse
Differential Revision: https://reviews.llvm.org/D120772
In Streaming SVE mode full NEON is not available, even though this is
implied from armv8-a. LLVM previously inferred that NEON needed to be
disabled when setting +streaming-sve, but there is no need to infer
this from +streaming-sve, because we can explicitly disable NEON using
LLVM's attribute mechanism. This is specifically relevant because
+streaming-sve is not a user-facing feature, but rather an LLVM internal
feature.
Reviewed By: paulwalker-arm
Differential Revision: https://reviews.llvm.org/D120809
This patch enables the lowering of basic modules and functions/subroutines
in modules.
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: PeteSteinfeld
Differential Revision: https://reviews.llvm.org/D120819
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
Co-authored-by: Jean Perier <jperier@nvidia.com>
This patch adds the lowering of the `inquire` statement.
This patch is part of the upstreaming effort from fir-dev branch.
Depends on D120822
Reviewed By: schweitz
Differential Revision: https://reviews.llvm.org/D120823
Co-authored-by: Jean Perier <jperier@nvidia.com>
This patches adds lowering for couple of basic io statements such as `flush`,
`endfile`, `backspace` and `rewind`
This patch is part of the upstreaming effort from fir-dev branch.
Depends on D120821
Reviewed By: schweitz
Differential Revision: https://reviews.llvm.org/D120822
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
Co-authored-by: Jean Perier <jperier@nvidia.com>
This patch adds the lowering of open and close statements
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: schweitz
Differential Revision: https://reviews.llvm.org/D120821
Co-authored-by: Jean Perier <jperier@nvidia.com>
Also, changes how the CSR loop is indexed, which should avoid bugs like the one fixed by rG4a57bb5a3b74bdad9b0518009a7d7ac7ca2ac650
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D120668
Instead of relying on underlying instructions, this patch updates
VPScalarIVStepsRecipe to only store the required type information.
This removes access to unrelated information, as well as avoiding issues
with the same underlying instruction being shared by multiple recipes.
This change should only change the debug output and not cause any
codegen changes, hence NFCI.
SIInstrInfo::FoldImmediate tried to delete move-immediate instructions
after folding them into their only use. This did not work because it was
checking hasOneNonDBGUse after doing the fold, at which point there
should be no uses. This seems to have no effect on codegen, it just
means less stuff for DCE to clean up later.
Differential Revision: https://reviews.llvm.org/D120815
This is an alternative to D120330, which disables MachineSink for
functions with irreducible cycles entirely. This avoids both the
correctness problem, and ensures we don't perform non-profitable
sinks into cycles. At the same time, it may also disable
profitable sinks in the same function. This can be made more
precise by using MachineCycleInfo in the future.
Fixes https://github.com/llvm/llvm-project/issues/53990.
Differential Revision: https://reviews.llvm.org/D120800
This is a recommit without changes. I originally reverted this
due to a significant code-size regression on tramp3d-v4, however
further investigation showed that in the tramp3d-v4 case this
change enables additional optimizations (in particular more
jump threading), which happens to reduce the size of a function
just enough to be eligible for inlining at hot callsites, which
results in the code size increase. As such, this was just bad
luck.
-----
This one-use limitation is artificial, we do not increase
instruction count if we perform the fold with multiple uses. The
motivating case is shown in @sub_eq_zero_select, where the one-use
limitation causes us to miss a subsequent select fold.
I believe the backend is pretty good about reusing flag-producing
subs for cmps with same operands, so I think doing this is fine.
Differential Revision: https://reviews.llvm.org/D120337
Currently we only check for splat shuffles, this extends it to see if the source operand is a splat across the demanded elts based upon the shuffle mask
Krzysztof Parzyszek