Summary:
Currently the RISC-V backend does not realign the stack. This can be an issue even for the RV32I/RV64I ABIs (where the stack is 16-byte aligned), though is rare. It will be much more comment with RV32E (though the alignment requirements for common data types remain under-documented...).
This patch adds minimal support for stack realignment. It should cope with large realignments. It will error out if the stack needs realignment and variable sized objects are present.
It feels like a lot of the code like getFrameIndexReference and determineFrameLayout could be refactored somehow, as right now it feels fiddly and brittle. We also seem to allocate a lot more memory than GCC does for equivalent C code.
Reviewers: asb
Reviewed By: asb
Subscribers: wwei, jrtc27, s.egerton, MaskRay, Jim, lenary, hiraditya, rbar, johnrusso, simoncook, apazos, sabuasal, niosHD, kito-cheng, shiva0217, zzheng, edward-jones, rogfer01, MartinMosbeck, brucehoult, the_o, rkruppe, PkmX, jocewei, psnobl, benna, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D62007
llvm-svn: 368300
This patch attempts to peek through vectors based on the demanded bits/elt of a particular ISD::EXTRACT_VECTOR_ELT node, allowing us to avoid dependencies on ops that have no impact on the extract.
In particular this helps remove some unnecessary scalar->vector->scalar patterns.
The wasm shift patterns are annoying - @tlively has indicated that the wasm vector shift codegen are to be refactored in the near-term and isn't considered a major issue.
Differential Revision: https://reviews.llvm.org/D65887
llvm-svn: 368276
G_JUMP_TABLE and G_BRJT appear from translation of switch statement.
Select these two instructions for MIPS32, both pic and non-pic.
Differential Revision: https://reviews.llvm.org/D65861
llvm-svn: 368274
VLDRH needs to have an alignment of at least 2, including the
widening/narrowing versions. This tightens up the ISel patterns for it and
alters allowsMisalignedMemoryAccesses so that unaligned accesses are expanded
through the stack. It also fixed some incorrect shift amounts, which seemed to
be passing a multiple not a shift.
Differential Revision: https://reviews.llvm.org/D65580
llvm-svn: 368256
This adjusts the load/store tests for better testing of alignments. It also
adds some extra alignment 1 tests, useful for future commits.
llvm-svn: 368255
If we're splitting the 512-bit vector anyway and we have zero/sign bits, then we might as well use pack instructions to concat and truncate at once.
Differential Revision: https://reviews.llvm.org/D65904
llvm-svn: 368210
This reverts commit fbc563e2cb6c5f1d0200b390513506b6aca4d2e9 "Create
unique, but identically-named ELF sections for explicitly-sectioned
functions and globals when using -function-sections and
-data-sections."
Reason for revert: sections are created with potentially wrong
attributes.
llvm-svn: 368204
The assert that caused this to be reverted should be fixed now.
Original commit message:
This patch changes our defualt legalization behavior for 16, 32, and
64 bit vectors with i8/i16/i32/i64 scalar types from promotion to
widening. For example, v8i8 will now be widened to v16i8 instead of
promoted to v8i16. This keeps the elements widths the same and pads
with undef elements. We believe this is a better legalization strategy.
But it carries some issues due to the fragmented vector ISA. For
example, i8 shifts and multiplies get widened and then later have
to be promoted/split into vXi16 vectors.
This has the potential to cause regressions so we wanted to get
it in early in the 10.0 cycle so we have plenty of time to
address them.
Next steps will be to merge tests that explicitly test the command
line option. And then we can remove the option and its associated
code.
llvm-svn: 368183
Prevent the LoadStoreOptimizer from pairing any load/store instructions with
instructions from the prologue/epilogue if the CFI information has encoded the
operations as separate instructions. This would otherwise lead to a mismatch
of the actual prologue size from the size as recorded in the Windows CFI.
Reviewers: efriedma, mstorsjo, ssijaric
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D65817
llvm-svn: 368164
In particular this helps the SSE vector shift cvttps2dq+add+shl pattern by avoiding the need for zeros in shuffle style extensions to vXi32 types as we'll be shifting out those bits anyway
llvm-svn: 368155
Currently we check whether LR is stored/loaded to/from inbetween the
loop decrement and loop end pseudo instructions. There's two problems
here:
- It relies on all load/store instructions being labelled as such in
tablegen.
- Actually any use of loop decrement is troublesome because the value
doesn't exist!
So we need to check for any read/write of LR that occurs between the
two instructions and revert if we find anything.
Differential Revision: https://reviews.llvm.org/D65792
llvm-svn: 368130
This reverts commit 3de33245d2c992c9e0af60372043540b60f3a810.
This commit broke the MSan buildbots. See
https://reviews.llvm.org/rL367901 for more information.
llvm-svn: 368107
This reverts commit 3f572c7b8405f36993ec8a226dcddd57283a7c1e.
The MSan sanitizer buildbot was broken by rL367901. This commit
(rL368079) depends on the broken commit that need to be reverted, and
thus itself is being reverted.
See https://reviews.llvm.org/rL367901 for more information.
llvm-svn: 368106
If we're after type legalization we should only be trying to turn
v2i64 into v2i32. So bitcast to v4i32, shuffle the even elements
together. Then use X86ISD::CVTSI2P. The alternative is to leave
the v2i64 type alone and let it scalarized. Hopefully keeping
it packed is better.
Fixes PR42905.
llvm-svn: 368091
This flag is now the default behavior so we no longer need to
set it in tests.
Some redundant tests have been removed after verifying we have
an equivalent test that didn't use the flag.
llvm-svn: 368079
https://reviews.llvm.org/D65698
This adds a KnownBits analysis pass for GISel. This was done as a
pass (compared to static functions) so that we can add other features
such as caching queries(within a pass and across passes) in the future.
This patch only adds the basic pass boiler plate, and implements a lazy
non caching knownbits implementation (ported from SelectionDAG). I've
also hooked up the AArch64PreLegalizerCombiner pass to use this - there
should be no compile time regression as the analysis is lazy.
llvm-svn: 368065
D62198 introduced an option to relax the checks for
hasOnlyUniformBranches. This commit turns the option on by default, for
better code generation in some cases in AMDGPU.
Differential Revision: https://reviews.llvm.org/D63198
Change-Id: I9cbff002a1e74d3b7eb96b4192dc8129936d537d
llvm-svn: 368042
This patch changes the DAG legalizer to respect the operation actions
set by the target for strict floating-point operations. (Currently, the
legalizer will usually fall back to mutate to the non-strict action
(which is assumed to be legal), and only skip mutation if the strict
operation is marked legal.)
With this patch, if whenever a strict operation is marked as Legal or
Custom, it is passed to the target as usual. Only if it is marked as
Expand will the legalizer attempt to mutate to the non-strict operation.
Note that this will now fail if the non-strict operation is itself
marked as Custom -- the target will have to provide a Custom definition
for the strict operation then as well.
Reviewed By: hfinkel
Differential Revision: https://reviews.llvm.org/D65226
llvm-svn: 368012
Summary:
When the WebAssembly backend encounters a return type that doesn't
fit within i32, SelectionDAG performs sret demotion, adding an
additional argument to the start of the function that contains
a pointer to an sret buffer to use instead. However, this conflicts
with the emscripten sjlj lowering pass. There we translate calls like:
```
call {i32, i32} @foo()
```
into (in pseudo-llvm)
```
%addr = @foo
call {i32, i32} @__invoke_{i32,i32}(%addr)
```
i.e. we perform an indirect call through an extra function.
However, the sret transform now transforms this into
the equivalent of
```
%addr = @foo
%sret = alloca {i32, i32}
call {i32, i32} @__invoke_{i32,i32}(%sret, %addr)
```
(while simultaneously translation the implementation of @foo as well).
Unfortunately, this doesn't work out. The __invoke_ ABI expected
the function address to be the first argument, causing crashes.
There is several possible ways to fix this:
1. Implementing the sret rewrite at the IR level as well and performing
it as part of lowering to __invoke
2. Fixing the wasm backend to recognize that __invoke has a special ABI
3. A change to the binaryen/emscripten ABI to recognize this situation
This revision implements the middle option, teaching the backend to
treat __invoke_ functions specially in sret lowering. This is achieved
by
1) Introducing a new CallingConv ID for invoke functions
2) When this CallingConv ID is seen in the backend and the first argument
is marked as sret (a function pointer would never be marked as sret),
swapping the first two arguments.
Reviewed By: tlively, aheejin
Differential Revision: https://reviews.llvm.org/D65463
llvm-svn: 367935
FastISel already does this since the initial arm64 port was upstreamed, so
it seems there are no issues with doing this at -O0 for very small memcpys.
Gives a 0.2% geomean code size improvement on CTMark.
Differential Revision: https://reviews.llvm.org/D65758
llvm-svn: 367919
This patch changes our defualt legalization behavior for 16, 32, and
64 bit vectors with i8/i16/i32/i64 scalar types from promotion to
widening. For example, v8i8 will now be widened to v16i8 instead of
promoted to v8i16. This keeps the elements widths the same and pads
with undef elements. We believe this is a better legalization strategy.
But it carries some issues due to the fragmented vector ISA. For
example, i8 shifts and multiplies get widened and then later have
to be promoted/split into vXi16 vectors.
This has the potential to cause regressions so we wanted to get
it in early in the 10.0 cycle so we have plenty of time to
address them.
Next steps will be to merge tests that explicitly test the command
line option. And then we can remove the option and its associated
code.
llvm-svn: 367901
Patch D56593 by @courbet results in calls to `bcmp()` in some cases, should
the target support the it. Unless `TTI::MemCmpExpansionOptions()`
is overridden by the target.
In a proprietary benchmark we see a performance drop of about 12% on PNG
compression before this patch, though it passes all tests.
This patch mirrors X86 for AArch64 and initializes
`TTI::MemCmpExpansionOptions()` to then expand calls to `bcmp()` when
appropriate. No tuning of the parameters was performed, but, at this point,
it's enough to recover the performance drop above.
This problem also exists on ARM. Once a consensus is reached for AArch64, we
can work to fix ARM as well.
Authors:
- Evandro Menezes (@evandro) <e.menezes@samsung.com>
- Brian Rzycki (@brzycki) <b.rzycki@samsung.com>
Differential revision: https://reviews.llvm.org/D64805
llvm-svn: 367898
Summary:
The Arm Neoverse N1 Software Optimization Guide [1], Section "4.8 Branch
instruction alignment" states:
"Consider aligning subroutine entry points and branch targets to 32B
boundaries, within the bounds of the code-density requirements of the
program."
This patch sets the preferred function alignment on Neoverse N1 to 2^4=16B.
This was already the case in some of the latest Cortex-A CPUs. Benchmarking
in previous Cortex-A CPUs suggested that 16B alignment is already better
than the default. See commit d04ee305.
The reason we don't set it to 32B right now (as the optimisation guide
suggests) is that this will impact code size and perhaps the instruction
cache performance. Therefore we need benchmark numbers first.
I have also added testing for A75 and A76 that we were missing.
[1] https://developer.arm.com/docs/swog309707/latest
Reviewers: fhahn, greened, samparker, dmgreen
Reviewed By: dmgreen
Subscribers: dmgreen, javed.absar, kristof.beyls, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D65654
llvm-svn: 367894
Summary:
This is a follow up to r367237. MachineFunction::getMachineMemOperand()
adds the offset parameter to the existing offset instead of resetting it.
So we need to reset the offset to the correct value after calling this
function.
Reviewers: arsenm
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D65557
llvm-svn: 367881
This was switching to use a format store for a non-format store for
f16 types. Also fixes i16/f16 stores on targets without legal f16.
The corresponding loads also need to be fixed.
llvm-svn: 367872
Without context we assume SGPR. Allowing VGPR constants theoretically
helps avoid a copy. This seems to not actually work now, and the
choice isn't based on the use bank.
llvm-svn: 367871
Summary:
This patch adds initial support for the SVE calling convention such that
SVE types can be passed as arguments and return values to/from a
subroutine.
The SVE AAPCS states [1]:
z0-z7 are used to pass scalable vector arguments to a subroutine,
and to return scalable vector results from a function. If a
subroutine takes arguments in scalable vector or predicate
registers, or if it is a function that returns results in such
registers, it must ensure that the entire contents of z8-z23 are
preserved across the call. In other cases it need only preserve the
low 64 bits of z8-z15, as described in §5.1.2.
p0-p3 are used to pass scalable predicate arguments to a subroutine
and to return scalable predicate results from a function. If a
subroutine takes arguments in scalable vector or predicate
registers, or if it is a function that returns results in these
registers, it must ensure that p4-p15 are preserved across the call.
In other cases it need not preserve any scalable predicate register
contents.
SVE predicate and data registers are passed indirectly (i.e. spilled to the
stack and pass the address) if they exceed the registers used for argument
passing defined by the PCS referenced above. Until SVE stack support is merged
we can't spill SVE registers to the stack, so currently an llvm_unreachable is
used where we will eventually handle this.
[1] https://static.docs.arm.com/100986/0000/100986_0000.pdf
Reviewed By: ostannard
Differential Revision: https://reviews.llvm.org/D65448
llvm-svn: 367859
