Commit Graph

533 Commits

Author SHA1 Message Date
Justin Hibbits d52990c71b Introduce codegen for the Signal Processing Engine
Summary:
The Signal Processing Engine (SPE) is found on NXP/Freescale e500v1,
e500v2, and several e200 cores.  This adds support targeting the e500v2,
as this is more common than the e500v1, and is in SoCs still on the
market.

This patch is very intrusive because the SPE is binary incompatible with
the traditional FPU.  After discussing with others, the cleanest
solution was to make both SPE and FPU features on top of a base PowerPC
subset, so all FPU instructions are now wrapped with HasFPU predicates.

Supported by this are:
* Code generation following the SPE ABI at the LLVM IR level (calling
conventions)
* Single- and Double-precision math at the level supported by the APU.

Still to do:
* Vector operations
* SPE intrinsics

As this changes the Callee-saved register list order, one test, which
tests the precise generated code, was updated to account for the new
register order.

Reviewed by: nemanjai
Differential Revision: https://reviews.llvm.org/D44830

llvm-svn: 337347
2018-07-18 04:25:10 +00:00
Stefan Pintilie 3d76326d24 [Power9] Add __float128 support for compare operations
Added handling for the select f128.

Differential Revision: https://reviews.llvm.org/D48294

llvm-svn: 336548
2018-07-09 13:36:14 +00:00
QingShan Zhang 9f0fe9a3f8 If the arch is P9, we will select the DFLOADf32/DFLOADf64 pseudo instruction when we are loading a floating,
and expand it post RA basing on the register pressure. However, we miss to do the add-imm peephole for these pseudo instruction.

Differential Revision: https://reviews.llvm.org/D47568
Reviewed By: Nemanjai

llvm-svn: 335024
2018-06-19 06:54:51 +00:00
Hiroshi Inoue 863fb7a4b8 [NFC] fix formatting
llvm-svn: 334263
2018-06-08 04:00:54 +00:00
Hiroshi Inoue 01ef4c2c64 [PowerPC] avoid unprofitable Repl32 flag in BitPermutationSelector
BitPermutationSelector sets Repl32 flag for bit groups which can be (potentially) benefit from 32-bit rotate-and-mask instructions with bit replication, i.e. rlwinm/rlwimi copies lower 32 bits into upper 32 bits on 64-bit PowerPC before rotation.
However, enforcing 32-bit instruction sometimes results in redundant generated code.
For example, the following simple code is compiled into rotldi + rlwimi while it can be compiled into only rldimi instruction if Repl32 flag is not set on the bit group for (a & 0xFFFFFFFF).

uint64_t func(uint64_t a, uint64_t b) {
	return (a & 0xFFFFFFFF) | (b << 32) ;
}

To avoid such problem, this patch checks the potential benefit of Repl32 flag before setting it. If a bit group does not require rotation (i.e. RLAmt == 0) and won't be merged into another group, we do not benefit from Repl32 flag on this group.

Differential Revision: https://reviews.llvm.org/D47867

llvm-svn: 334195
2018-06-07 13:21:14 +00:00
Hiroshi Inoue b557846083 [PowerPC] fix trivial typos in comment, NFC
llvm-svn: 334191
2018-06-07 12:49:12 +00:00
Hiroshi Inoue 955655f558 [PowerPC] reduce rotate in BitPermutationSelector
BitPermutationSelector builds the output value by repeating rotate-and-mask instructions with input registers.
Here, we may avoid one rotate instruction if we start building from an input register that does not require rotation.

For example of the test case bitfieldinsert.ll, it first rotates left r4 by 8 bits and then inserts some bits from r5 without rotation.
This can be executed by one rlwimi instruction, which rotates r4 by 8 bits and inserts its bits into r5.

This patch adds a check for rotation amounts in the comparator used in sorting to process the input without rotation first.

Differential Revision: https://reviews.llvm.org/D47765

llvm-svn: 334011
2018-06-05 11:58:01 +00:00
Hiroshi Inoue 9796b47df1 [NFC] Zero initialize local variables
This patch makes local variables zero initialized to avoid broken values in debug output.

llvm-svn: 333754
2018-06-01 14:23:15 +00:00
Lei Huang 716103f1cd [PowerPC] Fix the incorrect iterator inside peephole
Instruction selection can insert nodes into the underlying list after the root
node so iterating will thereby miss it. We should NOT assume that, the root node
is the last element in the DAG nodelist.

Patch by: steven.zhang (Qing Shan Zhang)

Differential Revision: https://reviews.llvm.org/D47437

llvm-svn: 333415
2018-05-29 13:38:56 +00:00
Nicola Zaghen d34e60ca85 Rename DEBUG macro to LLVM_DEBUG.
The DEBUG() macro is very generic so it might clash with other projects.
The renaming was done as follows:
- git grep -l 'DEBUG' | xargs sed -i 's/\bDEBUG\s\?(/LLVM_DEBUG(/g'
- git diff -U0 master | ../clang/tools/clang-format/clang-format-diff.py -i -p1 -style LLVM
- Manual change to APInt
- Manually chage DOCS as regex doesn't match it.

In the transition period the DEBUG() macro is still present and aliased
to the LLVM_DEBUG() one.

Differential Revision: https://reviews.llvm.org/D43624

llvm-svn: 332240
2018-05-14 12:53:11 +00:00
Hiroshi Inoue 33486787cb [PowerPC] fix incorrect vectorization of abs() on POWER9
Vectorized loops with abs() returns incorrect results on POWER9. This patch fixes it.
For example the following code returns negative result if input values are negative though it sums up the absolute value of the inputs.

int vpx_satd_c(const int16_t *coeff, int length) {
  int satd = 0;
  for (int i = 0; i < length; ++i) satd += abs(coeff[i]);
  return satd;
}

This problem causes test failures for libvpx.
For vector absolute and vector absolute difference on POWER9, LLVM generates VABSDUW (Vector Absolute Difference Unsigned Word) instruction or variants.
Since these instructions are for unsigned integers, we need adjustment for signed integers.
For abs(sub(a, b)), we generate VABSDUW(a+0x80000000, b+0x80000000). Otherwise, abs(sub(-1, 0)) returns 0xFFFFFFFF(=-1) instead of 1. For abs(a), we generate VABSDUW(a+0x80000000, 0x80000000).

Differential Revision: https://reviews.llvm.org/D45522

llvm-svn: 330497
2018-04-21 09:32:17 +00:00
Mandeep Singh Grang 327fd5e47c [PowerPC] Change std::sort to llvm::sort in response to r327219
Summary:
r327219 added wrappers to std::sort which randomly shuffle the container before sorting.
This will help in uncovering non-determinism caused due to undefined sorting
order of objects having the same key.

To make use of that infrastructure we need to invoke llvm::sort instead of std::sort.

Note: This patch is one of a series of patches to replace *all* std::sort to llvm::sort.
Refer the comments section in D44363 for a list of all the required patches.

Reviewers: hfinkel, RKSimon

Reviewed By: RKSimon

Subscribers: nemanjai, kbarton, llvm-commits

Differential Revision: https://reviews.llvm.org/D44870

llvm-svn: 329535
2018-04-08 16:45:04 +00:00
Hiroshi Inoue a2eefb6d9a [PowerPC] allow D-form VSX load/store when accessing FrameIndex without offset
VSX D-form load/store instructions of POWER9 require the offset be a multiple of 16 and a helper`isOffsetMultipleOf` is used to check this.
So far, the helper handles FrameIndex + offset case, but not handling FrameIndex without offset case. Due to this, we are missing opportunities to exploit D-form instructions when accessing an object or array allocated on stack.
For example, x-form store (stxvx) is used for int a[4] = {0}; instead of d-form store (stxv). For larger arrays, D-form instruction is not used when accessing the first 16-byte. Using D-form instructions reduces register pressure as well as instructions.

Differential Revision: https://reviews.llvm.org/D45079

llvm-svn: 329377
2018-04-06 05:41:16 +00:00
Craig Topper 2fa1436206 [IR][CodeGen] Remove dependency on EVT from IR/Function.cpp. Move EVT to CodeGen layer.
Currently EVT is in the IR layer only because of Function.cpp needing a very small piece of the functionality of EVT::getEVTString(). The rest of EVT is used in codegen making CodeGen a better place for it.

The previous code converted a Type* to EVT and then called getEVTString. This was only expected to handle the primitive types from Type*. Since there only a few primitive types, we can just print them as strings directly.

Differential Revision: https://reviews.llvm.org/D45017

llvm-svn: 328806
2018-03-29 17:21:10 +00:00
Strahinja Petrovic 06cf6a6490 [PowerPC] Secure PLT support
This patch supports secure PLT mode for PowerPC 32 architecture.

Differential Revision: https://reviews.llvm.org/D42112

llvm-svn: 328617
2018-03-27 11:23:53 +00:00
David Blaikie 36a0f226b1 Fix layering by moving ValueTypes.h from CodeGen to IR
ValueTypes.h is implemented in IR already.

llvm-svn: 328397
2018-03-23 23:58:31 +00:00
David Blaikie 13e77db2df Fix layering of MachineValueType.h by moving it from CodeGen to Support
This is used by llvm tblgen as well as by LLVM Targets, so the only
common place is Support for now. (maybe we need another target for these
sorts of things - but for now I'm at least making them correct & we can
make them better if/when people have strong feelings)

llvm-svn: 328395
2018-03-23 23:58:25 +00:00
Zaara Syeda 1110c4d336 [PowerPC] Optimize TLS initial-exec sequence to use X-Form loads/stores
This patch adds new load/store instructions for integer scalar types
which can be used for X-Form when fed by add with an @tls relocation.

Differential Revision: https://reviews.llvm.org/D43315

llvm-svn: 327635
2018-03-15 15:34:41 +00:00
Nemanja Ivanovic bcc82c9a78 [PowerPC] Disable shrink-wrapping when getting PC address through the LR
The instruction sequence used to get the address of the PC into a GPR requires
that we clobber the link register. Doing so without having first saved it in
the prologue leaves the function unable to return. Currently, this sequence is
emitted into the entry block. To ensure the prologue is inserted before this
sequence, disable shrink-wrapping.

This fixes PR33547.

Differential Revision: https://reviews.llvm.org/D43677

llvm-svn: 325972
2018-02-23 23:08:34 +00:00
Matthias Braun f1caa2833f MachineFunction: Return reference from getFunction(); NFC
The Function can never be nullptr so we can return a reference.

llvm-svn: 320884
2017-12-15 22:22:58 +00:00
Nemanja Ivanovic 1794cdc481 Fix code causing fallthrough warnings in the PPC back end.
llvm-svn: 320806
2017-12-15 11:47:48 +00:00
Francis Visoiu Mistrih 5df3bbf3e6 [CodeGen] Print global addresses as @foo in both MIR and debug output
Work towards the unification of MIR and debug output by printing
`@foo` instead of `<ga:@foo>`.

Also print target flags in the MIR format since most of them are used on
global address operands.

Only debug syntax is affected.

llvm-svn: 320682
2017-12-14 10:03:09 +00:00
Nemanja Ivanovic b0783cccb7 [PowerPC] Follow-up to r318436 to get the missed CSE opportunities
The last of the three patches that https://reviews.llvm.org/D40348 was
broken up into.
Canonicalize the materialization of constants so that they are more likely
to be CSE'd regardless of the bit-width of the use. If a constant can be
materialized using PPC::LI, materialize it the same way always.
For example:
  li 4, -1
  li 4, 255
  li 4, 65535
are equivalent if the uses only use the low byte. Canonicalize it to the
first form.

Differential Revision: https://reviews.llvm.org/D40348

llvm-svn: 320473
2017-12-12 12:09:34 +00:00
Nemanja Ivanovic 4364513cb2 Follow-up to r319434 to turn the pass on by default
Now that the patch has gone through the buildbot cycle,
turn it on by default.

llvm-svn: 319535
2017-12-01 12:02:59 +00:00
Nemanja Ivanovic db7e77047c [PowerPC] Recommit r314244 with refactoring and off by default
This re-commits everything that was pulled in r314244. The transformation
is off by default (patch to enable it to follow). The code is refactored
to have a single entry-point and provide fine-grained control over patterns
that it selects. This patch also fixes the bugs in the original code.

Everything that failed with the original patch has been re-tested with this
patch (with the transformation turned on). So the patch to turn this on is
soon to follow.

Differential Revision: https://reviews.llvm.org/D38575

llvm-svn: 319434
2017-11-30 13:39:10 +00:00
Francis Visoiu Mistrih 9d7bb0cb40 [CodeGen] Print register names in lowercase in both MIR and debug output
As part of the unification of the debug format and the MIR format,
always print registers as lowercase.

* Only debug printing is affected. It now follows MIR.

Differential Revision: https://reviews.llvm.org/D40417

llvm-svn: 319187
2017-11-28 17:15:09 +00:00
David Blaikie b3bde2ea50 Fix a bunch more layering of CodeGen headers that are in Target
All these headers already depend on CodeGen headers so moving them into
CodeGen fixes the layering (since CodeGen depends on Target, not the
other way around).

llvm-svn: 318490
2017-11-17 01:07:10 +00:00
David Blaikie 3f833edc7c Target/TargetInstrInfo.h -> CodeGen/TargetInstrInfo.h to match layering
This header includes CodeGen headers, and is not, itself, included by
any Target headers, so move it into CodeGen to match the layering of its
implementation.

llvm-svn: 317647
2017-11-08 01:01:31 +00:00
Benjamin Kramer 16610028ea Remove unused variables. No functionality change.
llvm-svn: 315185
2017-10-08 19:11:02 +00:00
Hiroshi Inoue dcedd66b00 [PowerPC] support ZERO_EXTEND in tryBitPermutation
This patch add a support of ISD::ZERO_EXTEND in PPCDAGToDAGISel::tryBitPermutation to increase the opportunity to use rotate-and-mask by reordering ZEXT and ANDI.
Since tryBitPermutation stops analyzing nodes if it hits a ZEXT node while traversing SDNodes, we want to avoid ZEXT between two nodes that can be folded into a rotate-and-mask instruction.

For example, we allow these nodes

      t9: i32 = add t7, Constant:i32<1>
    t11: i32 = and t9, Constant:i32<255>
  t12: i64 = zero_extend t11
t14: i64 = shl t12, Constant:i64<2>

to be folded into a rotate-and-mask instruction.
Such case often happens in array accesses with logical AND operation in the index, e.g. array[i & 0xFF];

Differential Revision: https://reviews.llvm.org/D37514

llvm-svn: 314655
2017-10-02 09:24:00 +00:00
Nemanja Ivanovic e22ebeab1a [PowerPC] Reverting sequence of patches for elimination of comparison instructions
In the past while, I've committed a number of patches in the PowerPC back end
aimed at eliminating comparison instructions. However, this causes some failures
in proprietary source and these issues are not observed in SPEC or any open
source packages I've been able to run.
As a result, I'm pulling the entire series and will refactor it to:
- Have a single entry point for easy control
- Have fine-grained control over which patterns we transform

A side-effect of this is that test cases for these patches (and modified by
them) are XFAIL-ed. This is a temporary measure as it is counter-productive
to remove/modify these test cases and then have to modify them again when
the refactored patch is recommitted.
The failure will be investigated in parallel to the refactoring effort and
the recommit will either have a fix for it or will leave this transformation
off by default until the problem is resolved.

llvm-svn: 314244
2017-09-26 20:42:47 +00:00
Nemanja Ivanovic f7bc9ce378 [PowerPC] Eliminate compares - add i64 sext/zext handling for SETLT/SETGT
As mentioned in https://reviews.llvm.org/D33718, this simply adds another
pattern to the compare elimination sequence and is committed without a
differential review.

llvm-svn: 314106
2017-09-25 14:05:46 +00:00
Nemanja Ivanovic f894ce35d0 [PowerPC] Eliminate compares - add i64 sext/zext handling for SETLE/SETGE
As mentioned in https://reviews.llvm.org/D33718, this simply adds another
pattern to the compare elimination sequence and is committed without a
differential review.

llvm-svn: 314073
2017-09-24 05:48:11 +00:00
Nemanja Ivanovic 35db4f956a [PowerPC] Eliminate compares - add i32 sext/zext handling for SETULT/SETUGT
As mentioned in https://reviews.llvm.org/D33718, this simply adds another
pattern to the compare elimination sequence and is committed without a
differential revision.

llvm-svn: 314062
2017-09-23 12:53:03 +00:00
Nemanja Ivanovic c4980799ab [PowerPC] Eliminate compares - add i32 sext/zext handling for SETULE/SETUGE
As mentioned in https://reviews.llvm.org/D33718, this simply adds another
pattern to the compare elimination sequence and is committed without a
differential revision.

llvm-svn: 314060
2017-09-23 09:50:12 +00:00
Nemanja Ivanovic 41c4a109d8 [PowerPC] Eliminate compares - add i32 sext/zext handling for SETLT/SETGT
As mentioned in https://reviews.llvm.org/D33718, this simply adds another
pattern to the compare elimination sequence and is committed without a
differential revision.

llvm-svn: 314055
2017-09-23 04:41:34 +00:00
Nemanja Ivanovic cea42b7fff Remove the default clause from a fully-covering switch
to appease bots that use a compiler that warns about this
and use -Werror.

llvm-svn: 313980
2017-09-22 12:26:00 +00:00
Nemanja Ivanovic d6f93f5143 Recommit r310809 with a fix for the spill problem
This patch re-commits the patch that was pulled out due to a
problem it caused, but with a fix for the problem. The fix
was reviewed separately by Eric Christopher and Hal Finkel.

Differential Revision: https://reviews.llvm.org/D38054

llvm-svn: 313978
2017-09-22 11:50:25 +00:00
Kyle Butt 8c0314c3ed PPC: Don't select lxv/stxv for insufficiently aligned stack slots.
The lxv/stxv instructions require an offset that is 0 % 16. Previously we were
selecting lxv/stxv for loads and stores to the stack where the offset from the
slot was a multiple of 16, but the stack slot was not 16 or more byte aligned.
When the frame gets lowered these transform to r(1|31) + slot + offset.
If slot is not aligned, slot + offset may not be 0 % 16.
Now we require 16 byte or more alignment for select lxv/stxv to stack slots.

Includes a testcase that shows both sufficiently and insufficiently aligned
stack slots.

llvm-svn: 312843
2017-09-09 00:37:56 +00:00
Hiroshi Inoue cc555bd0ac [PowerPC] better instruction selection for OR (XOR) with a 32-bit immediate
- recommitting after fixing a test failure on MacOS

On PPC64, OR (XOR) with a 32-bit immediate can be done with only two instructions, i.e. ori + oris.
But the current LLVM generates three or four instructions for this purpose (and also it clobbers one GPR).

This patch makes PPC backend generate ori + oris (xori + xoris) for OR (XOR) with a 32-bit immediate.

e.g. (x | 0xFFFFFFFF) should be

	ori 3, 3, 65535
	oris 3, 3, 65535

but LLVM generates without this patch

	li 4, 0
	oris 4, 4, 65535
	ori 4, 4, 65535
	or 3, 3, 4

Differential Revision: https://reviews.llvm.org/D34757

llvm-svn: 311538
2017-08-23 08:55:18 +00:00
Hiroshi Inoue dbb285ca51 Revert rL311526: [PowerPC] better instruction selection for OR (XOR) with a 32-bit immediate
This reverts commit rL311526 due to failures in some buildbot.

llvm-svn: 311530
2017-08-23 06:38:05 +00:00
Hiroshi Inoue c4449df1b0 [PowerPC] better instruction selection for OR (XOR) with a 32-bit immediate
On PPC64, OR (XOR) with a 32-bit immediate can be done with only two instructions, i.e. ori + oris.
But the current LLVM generates three or four instructions for this purpose (and also it clobbers one GPR).

This patch makes PPC backend generate ori + oris (xori + xoris) for OR (XOR) with a 32-bit immediate.

e.g. (x | 0xFFFFFFFF) should be

	ori 3, 3, 65535
	oris 3, 3, 65535

but LLVM generates without this patch

	li 4, 0
	oris 4, 4, 65535
	ori 4, 4, 65535
	or 3, 3, 4

Differential Revision: https://reviews.llvm.org/D34757

llvm-svn: 311526
2017-08-23 05:15:15 +00:00
Chandler Carruth fe6b509f83 [PowerPC] Revert r310346 (and followups r310356 & r310424) which
introduce a miscompile bug.

There appears to be a bug where the generated code to extract the sign
bit doesn't work correctly for 32-bit inputs. I've replied to the
original commit pointing out the problem. I think I see by inspection
(and reading the manual for PPC) how to fix this, but I can't be 100%
confident and I also don't know what the best way to test this is.
Currently it seems nearly impossible to get the backend to hit this code
path, but the patch autohr is likely in a better position to craft such
test cases than I am, and based on where the bug is it should be easily
done.

Original commit message for r310346:
"""
[PowerPC] Eliminate compares - add i32 sext/zext handling for SETLE/SETGE

Adds handling for SETLE/SETGE comparisons on i32 values. Furthermore, it
adds the handling for the special case where RHS == 0.

Differential Revision: https://reviews.llvm.org/D34048
"""

llvm-svn: 310809
2017-08-14 03:41:00 +00:00
Nemanja Ivanovic 3f98fb4a3f My commit r310346 introduced some valid warnings. This cleans them up.
llvm-svn: 310424
2017-08-08 22:17:31 +00:00
Nemanja Ivanovic bed7136eee Appease compilers that have the -Wcovered-switch-default switch.
llvm-svn: 310356
2017-08-08 12:41:56 +00:00
Nemanja Ivanovic 809fbfa6a1 [PowerPC] Eliminate compares - add i32 sext/zext handling for SETLE/SETGE
Adds handling for SETLE/SETGE comparisons on i32 values. Furthermore, it adds
the handling for the special case where RHS == 0.

Differential Revision: https://reviews.llvm.org/D34048

llvm-svn: 310346
2017-08-08 11:20:44 +00:00
Hiroshi Inoue 5703fe37ab [PowerPC] Change method names; NFC
Changed method names based on the discussion in https://reviews.llvm.org/D34986:
getInt64 -> selectI64Imm,
getInt64Count -> selectI64ImmInstrCount.

llvm-svn: 309541
2017-07-31 06:27:09 +00:00
Nemanja Ivanovic 864c953773 [PowerPC] - Recommit r304907 now that the issue has been fixed
This is just a recommit since the issue that the commit exposed is now
resolved.

llvm-svn: 308995
2017-07-25 17:54:51 +00:00
Nemanja Ivanovic 3c7e276d24 [PowerPC] Ensure displacements for DQ-Form instructions are multiples of 16
As outlined in the PR, we didn't ensure that displacements for DQ-Form
instructions are multiples of 16. Since the instruction encoding encodes
a quad-word displacement, a sub-16 byte displacement is meaningless and
ends up being encoded incorrectly.

Fixes https://bugs.llvm.org/show_bug.cgi?id=33671.

Differential Revision: https://reviews.llvm.org/D35007

llvm-svn: 307934
2017-07-13 18:17:10 +00:00
Hiroshi Inoue f55ee1b476 [PowerPC] avoid redundant analysis while lowering an immediate; NFC
This patch reduces compilation time by avoiding redundant analysis while selecting instructions to create an immediate.
If the instruction count required to create the input number without rotate is 2, we do not need further analysis to find a shorter instruction sequence with rotate; rotate + load constant cannot be done by 1 instruction (i.e. getInt64CountDirectnever return 0).
This patch should not change functionality.

Differential Revision: https://reviews.llvm.org/D34986

llvm-svn: 307623
2017-07-11 05:28:26 +00:00