| Commit message (Collapse) | Author | Age | Files | Lines |
| ... | |
| |
|
|
|
|
| |
This also fixes some of the ReadAfterLd issues due to InstRW.
llvm-svn: 330544
|
| |
|
|
|
|
| |
instrw overrides.
llvm-svn: 330542
|
| |
|
|
|
|
| |
latency.
llvm-svn: 330541
|
| |
|
|
|
|
| |
This matches the other FENCE instructions.
llvm-svn: 330533
|
| |
|
|
|
|
| |
OpSizeFixed.
llvm-svn: 330532
|
| |
|
|
|
|
|
|
|
|
| |
'data32' in 16-bit mode. Hack the asm parser to convert 'data32' to 'data16' in 16-bit mode.
Improve the error messages to match GNU assembler.
This also allows us to remove the hack from the disassembler table building.
llvm-svn: 330531
|
| |
|
|
|
|
| |
scheduler models.
llvm-svn: 330527
|
| |
|
|
| |
llvm-svn: 330525
|
| |
|
|
|
|
| |
models.
llvm-svn: 330523
|
| |
|
|
|
|
| |
VPERM2I128/VINSERTI128
llvm-svn: 330522
|
| |
|
|
|
|
| |
pack/unpack instruction instrw overrides from scheduler models.
llvm-svn: 330521
|
| |
|
|
|
|
|
|
| |
PMULHW/PMULHUW.
Ultimately I want to use this to remove the intrinsics for these instructions.
llvm-svn: 330520
|
| |
|
|
| |
llvm-svn: 330517
|
| |
|
|
|
|
| |
overrides.
llvm-svn: 330514
|
| |
|
|
|
|
| |
D45629
llvm-svn: 330513
|
| |
|
|
|
|
|
|
| |
from scheduler models.
The required the default skylake schedules to be updated - these were being completely overriden by the InstRW and the existing values not used at all.
llvm-svn: 330510
|
| |
|
|
|
|
| |
scheduler models.
llvm-svn: 330508
|
| |
|
|
| |
llvm-svn: 330505
|
| |
|
|
| |
llvm-svn: 330503
|
| |
|
|
| |
llvm-svn: 330501
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
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
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Split the fp and integer vector logical instruction scheduler classes - older CPUs especially often handled these on different pipes.
This unearthed a couple of things that are also handled in this patch:
(1) We were tagging avx512 fp logic ops as WriteFAdd, probably because of the lack of WriteFLogic
(2) SandyBridge had integer logic ops only using Port5, when afaict they can use Ports015.
(3) Cleaned up x86 FCHS/FABS scheduling as they are typically treated as fp logic ops.
Differential Revision: https://reviews.llvm.org/D45629
llvm-svn: 330480
|
| |
|
|
| |
llvm-svn: 330472
|
| |
|
|
|
|
|
|
|
|
| |
There was some unfortunate interaction between VSPLAT and BITCAST
related to the selection of constant vectors (coming from selecting
shuffles). Introduce VSPLATW that always splats a 32-bit word, and
can have arbitrary result type (to avoid BITCASTs of VSPLAT).
Clean up the previous selection of BITCAST/VSPLAT.
llvm-svn: 330471
|
| |
|
|
|
|
|
| |
Fixed slots have negative values, and TRI::stackSlot2Index and
TRI::index2StackSlot do not handle negative numbers.
llvm-svn: 330468
|
| |
|
|
| |
llvm-svn: 330465
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Three new instructions:
umonitor - Sets up a linear address range to be
monitored by hardware and activates the monitor.
The address range should be a writeback memory
caching type.
umwait - A hint that allows the processor to
stop instruction execution and enter an
implementation-dependent optimized state
until occurrence of a class of events.
tpause - Directs the processor to enter an
implementation-dependent optimized state
until the TSC reaches the value in EDX:EAX.
Also modifying the description of the mfence
instruction, as the rep prefix (0xF3) was allowed
before, which would conflict with umonitor during
disassembly.
Before:
$ echo 0xf3,0x0f,0xae,0xf0 | llvm-mc -disassemble
.text
mfence
After:
$ echo 0xf3,0x0f,0xae,0xf0 | llvm-mc -disassemble
.text
umonitor %rax
Reviewers: craig.topper, zvi
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D45253
llvm-svn: 330462
|
| |
|
|
|
|
|
|
|
|
| |
First off, this is more correct than having the B. Second off, this was making
a bot upset. This fixes that.
Update the test to include -verify-machineinstrs as well to prevent stuff like
this slipping by non debug/assert builds in the future.
llvm-svn: 330459
|
| |
|
|
| |
llvm-svn: 330428
|
| |
|
|
|
|
| |
Same as other cacheline instructions
llvm-svn: 330424
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
instructions.
This is patch [4/4] in a series to add assembler/disassembler support for
SVE's contiguous LD1 (scalar+scalar) instructions:
- Patch [1/4]: https://reviews.llvm.org/D45687
- Patch [2/4]: https://reviews.llvm.org/D45688
- Patch [3/4]: https://reviews.llvm.org/D45689
- Patch [4/4]: https://reviews.llvm.org/D45690
Reviewers: fhahn, rengolin, javed.absar, huntergr, SjoerdMeijer, t.p.northover, echristo, evandro
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D45690
llvm-svn: 330423
|
| |
|
|
|
|
|
|
|
|
| |
Diagnostic:
'index must be multiple of 3 in range [-32, 28]'
Must be:
'index must be multiple of 4 in range [-32, 28]'
llvm-svn: 330407
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
This is patch [3/4] in a series to add assembler/disassembler support for
SVE's contiguous LD1 (scalar+scalar) instructions:
- Patch [1/4]: https://reviews.llvm.org/D45687
- Patch [2/4]: https://reviews.llvm.org/D45688
- Patch [3/4]: https://reviews.llvm.org/D45689
- Patch [4/4]: https://reviews.llvm.org/D45690
Reviewers: fhahn, rengolin, javed.absar, huntergr, SjoerdMeijer, t.p.northover, echristo, evandro
Reviewed By: SjoerdMeijer
Subscribers: tschuett, kristof.beyls, llvm-commits
Differential Revision: https://reviews.llvm.org/D45689
llvm-svn: 330406
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
Reading Atmel's AT697E errata document this does not seem like a valid
workaround. While the text only mentions SDIV, it says that the ICC flags
can be wrong, and those are only generated by SDIVcc. Verification on
hardware shows that simply replacing SDIV with SDIVcc does not avoid
the bug with negative operands.
This reverts r283727.
Reviewers: lero_chris, jyknight
Reviewed By: jyknight
Subscribers: fedor.sergeev, jrtc27, llvm-commits
Differential Revision: https://reviews.llvm.org/D45813
llvm-svn: 330397
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Using `clr reg`/`mov %g0, reg`/`or %g0, %g0, reg` to zero a register
looks much better than `sethi 0, reg`.
Reviewers: jyknight, venkatra
Reviewed By: jyknight
Subscribers: eraman, fedor.sergeev, jrtc27, llvm-commits
Differential Revision: https://reviews.llvm.org/D45810
llvm-svn: 330396
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
In some cases the shift/extend needs to be explicitly parsed together
with the register, rather than as a separate operand. This is needed
for addressing modes where the instruction as a whole dictates the
scaling/extend, rather than specific bits in the instruction.
By parsing them as a single operand, we avoid the need to pass an
extra operand in all CodeGen patterns (because all operands need to
have an associated value), and we avoid the need to update TableGen to
accept operands that have no associated bits in the instruction.
An added benefit of parsing them together is that the assembler
can give a sensible diagnostic if the scaling is not correct.
This is patch [2/4] in a series to add assembler/disassembler support for
SVE's contiguous LD1 (scalar+scalar) instructions:
- Patch [1/4]: https://reviews.llvm.org/D45687
- Patch [2/4]: https://reviews.llvm.org/D45688
- Patch [3/4]: https://reviews.llvm.org/D45689
- Patch [4/4]: https://reviews.llvm.org/D45690
Reviewers: fhahn, rengolin, javed.absar, huntergr, SjoerdMeijer, t.p.northover, echristo, evandro
Reviewed By: fhahn, SjoerdMeijer
Subscribers: kristof.beyls, llvm-commits
Differential Revision: https://reviews.llvm.org/D45688
llvm-svn: 330394
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
This fixes a case where the argument to a sendmsg intrinsic
ends up in a VGPR, for whatever reason.
The underlying performance issue is that a multiplication that
can be an s_mul_i32 is instead needlessly generated as
v_mul_u32_u24, but this is not addressed by this patch.
Change-Id: I61fd4034314d5acdf6074632c30b65364dfa7328
Reviewers: arsenm, rampitec
Subscribers: kzhuravl, wdng, yaxunl, dstuttard, tpr, t-tye, llvm-commits
Differential Revision: https://reviews.llvm.org/D45826
llvm-svn: 330393
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
If a 64-bit register is used as an operand in inline assembly together
with a memory reference, the memory addressing will be wrong. The
addressing will be a single reg, instead of reg+reg or reg+imm. This
will generate a bad offset value or an exception in printMemOperand().
For example:
```
long long int val = 5;
long long int mem;
__asm__ volatile ("std %1, %0":"=m"(mem):"r"(val));
```
becomes:
```
std %i0, [%i2+589833]
```
The problem is that SelectInlineAsmMemoryOperand() is never called for
the memory references if one of the operands is a 64-bit register.
By calling SelectInlineAsmMemoryOperands() in tryInlineAsm() the Sparc
version of SelectInlineAsmMemoryOperand() gets called for each memory
reference.
Reviewers: jyknight, venkatra
Reviewed By: jyknight
Subscribers: eraman, fedor.sergeev, jrtc27, llvm-commits
Differential Revision: https://reviews.llvm.org/D45761
llvm-svn: 330392
|
| |
|
|
|
|
| |
Differential Revision: https://reviews.llvm.org/D45790
llvm-svn: 330365
|
| |
|
|
|
|
| |
Silvermont and Goldmont have the same issue on popcnt as Sandy Bridge, Haswell, Broadwell, and Skylake. Believe it is fixed in Goldmont Plus.
llvm-svn: 330358
|
| |
|
|
|
|
| |
Luckily this was on instructions not supported on Silvermont....
llvm-svn: 330351
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Intel CPUs.
The XCHG16rr/XCHG32rr/XCHG64rr instructions should be 3 uops just like XCHG8rr. I believe they're just implemented as 3 move uops with a temporary register.
XADD is probably 2 moves and an add also using a temporary register.
Change the latency for both from 2 cycles to 3 cycles. Only 2 of the uops are serialized in their execution, the move into the temporary and the move out of the temporary. The move from one GPR to the other should be able to go in parallel with this if there are ALU resources available.
llvm-svn: 330349
|
| |
|
|
|
|
| |
There's a lot more but I'd prefer focussing on removing unnecessary InstRWs first.
llvm-svn: 330347
|
| |
|
|
| |
llvm-svn: 330344
|
| |
|
|
|
|
|
|
| |
We should also check that the "bottom" basic block of a loopis a successor of the "header" basic block, otherwise we don't propagate the information correctly when the CFG is complex. This fixes an important rendering problem with Wolfsentein 2, because of one vector-memory wait was missing.
Differential Revision: https://reviews.llvm.org/D43831
llvm-svn: 330337
|
| |
|
|
| |
llvm-svn: 330333
|
| |
|
|
|
|
| |
These are already handled identically by WriteALU.
llvm-svn: 330332
|
| |
|
|
| |
llvm-svn: 330330
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
instructions
These instructions lacked the correct predicates, were not marked
as loads and stores and lacked the proper instruction mapping information.
In the case of microMIPS sw(l|r)e (EVA) these instructions were using the load
EVA description.
Reviewers: abeserminji, smaksimovic, atanasyan
Differential Revision: https://reviews.llvm.org/D45626
llvm-svn: 330326
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
This is the patch that lowers x86 intrinsics to native IR
in order to enable optimizations. The patch also includes folding
of previously missing saturation patterns so that IR emits the same
machine instructions as the intrinsics.
Patch by tkrupa
Differential Revision: https://reviews.llvm.org/D44785
llvm-svn: 330322
|