Similar to what was recently done to ParseATTOperand. Make
ParseIntelOperand directly responsible for adding to the operand
vector instead of returning the operand. Return a bool for error.
Remove ErrorOperand since it is no longer used.
findAllocaForValue uses AllocaForValue to cache resolved values.
The function is used only to resolve arguments of lifetime
intrinsic which usually are not fare for allocas. So result reuse
is likely unnoticeable.
In followup patches I'd like to replace the function with
GetUnderlyingObjects.
Depends on D84616.
Differential Revision: https://reviews.llvm.org/D84617
Fix for the issue raised in https://github.com/rust-lang/rust/issues/74632.
The current heuristic for inserting LFENCEs uses a quadratic-time algorithm. This can apparently cause substantial compilation slowdowns for building Rust projects, where functions > 5000 LoC are apparently common.
The updated heuristic in this patch implements a linear-time algorithm. On a set of benchmarks, the slowdown factor for the generated code was comparable (2.55x geo mean for the quadratic-time heuristic, vs. 2.58x for the linear-time heuristic). Both heuristics offer the same security properties, namely, mitigating LVI.
This patch also includes some formatting fixes.
Differential Revision: https://reviews.llvm.org/D84471
After the recent change to the tuning settings for pentium4 to improve our default 32-bit behavior, I've decided to see about implementing -mtune support. This way we could have a default architecture CPU of "pentium4" or "x86-64" and a default tuning cpu of "generic". And we could change our "pentium4" tuning settings back to what they were before.
As a step to supporting this, this patch separates all of the features lists for the CPUs into 2 lists. I'm using the Proc class and a new ProcModel class to concat the 2 lists before passing to the target independent ProcessorModel. Future work to truly support mtune would change ProcessorModel to take 2 lists separately. I've diffed the X86GenSubtargetInfo.inc file before and after this patch to ensure that the final feature list for the CPUs isn't changed.
Differential Revision: https://reviews.llvm.org/D84879
Avoid recursively calling copyPhysReg for AGPR handling. This was
dropping the necessary super register implicit defs to avoid liveness
verifier errors.
LLVM selection dag assumes "switch" indices are pointer sized, which causes problems for our 32-bit br_table. The new function ensures 32-bit operands don't get unnecessarily extended, and 64-bit operands get truncated.
Note that the changes to the existing test test exactly that: the addition of -NEXT in 2 places ensures no extension is inserted (which the test previously ignored) and that the wrap is present (previously omitted in wasm64 mode).
Differential Revision: https://reviews.llvm.org/D84705
We are using undef on the indirect move source subreg and then
using implicit super-reg. This creates a problem in RA when
Greedy decides to split the register. It reassigns the implicit
super-reg but does not bother to change undef source because
it is really does not matter. The fix is to stop lying to RA and
drop undef flag.
This has also hit a problem in SIFoldOperands as it can fold
immediate into an indirect move since there is no undef flag
anymore. That results in multiple test failures, so added the
check for this case.
Differential Revision: https://reviews.llvm.org/D84899
Get rid of all fixmes and base heuristic on `num-clustered-dwords`. The main intuition behind this is as
follows. The existing heuristic roughly summarizes as below:
* Assume, all the mem ops instructions participating in the clustering process, loads/stores same num bytes
* If num bytes loaded by each mem op is 4 bytes, then cluster at max 5 mem ops, that is at max 20 bytes
* If num bytes loaded by each mem op is 8 bytes, then cluster at max 3 mem ops, that is at max 24 bytes
* If num bytes loaded by each mem op is 16 bytes, then cluster at max 2 mem ops, that is at max 32 bytes
So, we need to make sure that the new heuristic do not completey deviate away from the above one, and it
properly handles both the sub-word loads and the wide loads.
Reviewed By: arsenm, rampitec
Differential Revision: https://reviews.llvm.org/D84354
We parse .arch so that some `.arch i386; .code32` code can assemble. It seems
that X86AsmParser does not do a good job tracking what features are needed to
assemble instructions. GNU as's x86 port supports a very wide range of .arch
operands. Ignore the operand for now.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D84900
The operand to these instructions is both input and output.
These are not yet emitted by the compiler and the assembler already
works fine, so can't test in this patch. But D75044 will use XPACI
and provide test coverage for this patch as well.
Differential Revision: https://reviews.llvm.org/D84298
Summary:
This patch implements -ffunction-sections on AIX.
This patch focuses on assembly generation.
Follow-on patch needs to handle:
1. -ffunction-sections implication for jump table.
2. Object file generation path and associated testing.
Differential Revision: https://reviews.llvm.org/D83875
As long as we can extract the lowest 128-bit subvector from the pre-truncated source vector, then we don't care what size it is.
The next stage will be to support non-zero extraction indices, as long as its still coming from the lowest 128-bit subvector.
When building code at -O0 We weren't falling back to DAG ISel correctly
when encountering alloca instructions with scalable vector types. This
is because the alloca has no operands that are scalable. I've fixed this by
adding a check in AArch64ISelLowering::fallBackToDAGISel for alloca
instructions with scalable types.
Differential Revision: https://reviews.llvm.org/D84746
Continue the change made to ParseATTOperand to take the vector by
reference. Let ParseMemOperand add its memory operand to the
vector and just return true/false to indicate error.
A '*' after the segment is equivalent to a '*' before the segment register. To make the AsmMatcher table work we need to place the '*' token into the operand vector before the full memory operand. To accomplish this I've modified some portions of operand parsing to expose the operand vector to ParseATTOperand so that the token can be pushed to the vector after parsing the segment register and before creating the memory operand using that segment register.
Fixes PR46879
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D84895
Summary:
Some instructions have set the wrong [RM] flag, this patch is to fix it.
Instructions x(v|s)r(d|s)pi[zmp]? and fri[npzm] use fixed rounding
directions without referencing current rounding mode.
Also, the SETRNDi, SETRND, BCLRn, MTFSFI, MTFSB0, MTFSB1, MTFSFb,
MTFSFI, MTFSFI_rec, MTFSF, MTFSF_rec should also fix the RM flag.
Reviewed By: jsji
Differential Revision: https://reviews.llvm.org/D81360
I still think it's highly questionable that we have two intrinsics
with identical behavior and only vary by the name of the libcall used
if it happens to be lowered that way, but try to reduce the feature
delta between SDAG and GlobalISel for recently added intrinsics. I'm
not sure which opcode should be considered the canonical one, but
lower roundeven back to round.
Instead of never accepting v8f32/v4f64 FHADD/FHSUB if the input shuffle masks cross lanes, perform the matching and determine if the post shuffle mask simplifies to a 'whole lane shuffle' mask - in which case we are guaranteed to cheaply perform this as a VPERM2F128 shuffle.
MFMA instructions shall not be scheduled back to back
to avoid MAI SIMD stall. Tell post-RA schedule we would
prefer some other instruction instead.
Differential Revision: https://reviews.llvm.org/D84883
Adds frontend and backend options to enable and disable the
PowerPC paired vector memory operations added in ISA 3.1.
Instructions using these options will be added in subsequent patches.
Differential Revision: https://reviews.llvm.org/D83722
In future, we'd like to use the perfect-shuffle mechanism to deal with these
shuffle permutations. For now, this improves performance by avoiding the
super-expensive const-pool load + tbl instruction.
Differential Revision: https://reviews.llvm.org/D84866
Port the wide immediate case from AArch64DAGToDAGISel::SelectAddrModeXRO.
If we have a wide immediate which can't be represented in an add, we can end up
with code like this:
```
mov x0, imm
add x1, base, x0
ldr x2, [x1, 0]
```
If we use the [base, xN] addressing mode instead, we can produce this:
```
mov x0, imm
ldr x2, [base, x0]
```
This saves 0.4% code size on 7zip at -O3, and gives a geomean code size
improvement of 0.1% on CTMark.
Differential Revision: https://reviews.llvm.org/D84784
I never completed the work on the patches referenced by
f8bf7d7f42f28fa18144091022236208e199f331, but this was intended to
avoid folding immediate writes into m0 which the coalescer doesn't
understand very well. Relax this to allow simple SGPR immediates to
fold directly into VGPR copies. This pattern shows up routinely in
current GlobalISel code since nothing is smart enough to emit VGPR
constants yet.
When it was first created, CFGSort only made sure BBs in each
`MachineLoop` are sorted together. After we added exception support,
CFGSort now also sorts BBs in each `WebAssemblyException`, which
represents a `catch` block, together, and
`Region` class was introduced to be a thin wrapper for both
`MachineLoop` and `WebAssemblyException`.
But how we compute those loops and exceptions is different.
`MachineLoopInfo` is constructed using the standard loop computation
algorithm in LLVM; the definition of loop is "a set of BBs that are
dominated by a loop header and have a path back to the loop header". So
even if some BBs are semantically contained by a loop in the original
program, or in other words dominated by a loop header, if they don't
have a path back to the loop header, they are not considered a part of
the loop. For example, if a BB is dominated by a loop header but
contains `call abort()` or `rethrow`, it wouldn't have a path back to
the header, so it is not included in the loop.
But `WebAssemblyException` is wasm-specific data structure, and its
algorithm is simple: a `WebAssemblyException` consists of an EH pad and
all BBs dominated by the EH pad. So this scenario is possible: (This is
also the situation in the newly added test in cfg-stackify-eh.ll)
```
Loop L: header, A, ehpad, latch
Exception E: ehpad, latch, B
```
(B contains `abort()`, so it does not have a path back to the loop
header, so it is not included in L.)
And it is sorted in this order:
```
header
A
ehpad
latch
B
```
And when CFGStackify places `end_loop` or `end_try` markers, it
previously used `WebAssembly::getBottom()`, which returns the latest BB
in the sorted order, and placed the marker there. So in this case the
marker placements will be like this:
```
loop
header
try
A
catch
ehpad
latch
end_loop <-- misplaced!
B
end_try
```
in which nesting between the loop and the exception is not correct.
`end_loop` marker has to be placed after `B`, and also after `end_try`.
Maybe the fundamental way to solve this problem is to come up with our
own algorithm for computing loop region too, in which we include all BBs
dominated by a loop header in a loop. But this takes a lot more effort.
The only thing we need to fix is actually, `getBottom()`. If we make it
return the right BB, which means in case of a loop, the latest BB of the
loop itself and all exceptions contained in there, we are good.
This renames `Region` and `RegionInfo` to `SortRegion` and
`SortRegionInfo` and extracts them into their own file. And add
`getBottom` to `SortRegionInfo` class, from which it can access
`WebAssemblyExceptionInfo`, so that it can compute a correct bottom
block for loops.
Reviewed By: dschuff
Differential Revision: https://reviews.llvm.org/D84724
I've been looking at missed vectorizations in one codebase.
One particular thing that stands out is that some of the loops
reach vectorizer in a rather mangled form, with weird PHI's,
and some of the loops aren't even in a rotated form.
After taking a more detailed look, that happened because
the loop's headers were too big by then. It is evident that
SimplifyCFG's common code hoisting transform is at fault there,
because the pattern it handles is precisely the unrotated
loop basic block structure.
Surprizingly, `SimplifyCFGOpt::HoistThenElseCodeToIf()` is enabled
by default, and is always run, unlike it's friend, common code sinking
transform, `SinkCommonCodeFromPredecessors()`, which is not enabled
by default and is only run once very late in the pipeline.
I'm proposing to harmonize this, and disable common code hoisting
until //late// in pipeline. Definition of //late// may vary,
here currently i've picked the same one as for code sinking,
but i suppose we could enable it as soon as right after
loop rotation happens.
Experimentation shows that this does indeed unsurprizingly help,
more loops got rotated, although other issues remain elsewhere.
Now, this undoubtedly seriously shakes phase ordering.
This will undoubtedly be a mixed bag in terms of both compile- and
run- time performance, codesize. Since we no longer aggressively
hoist+deduplicate common code, we don't pay the price of said hoisting
(which wasn't big). That may allow more loops to be rotated,
so we pay that price. That, in turn, that may enable all the transforms
that require canonical (rotated) loop form, including but not limited to
vectorization, so we pay that too. And in general, no deduplication means
more [duplicate] instructions going through the optimizations. But there's still
late hoisting, some of them will be caught late.
As per benchmarks i've run {F12360204}, this is mostly within the noise,
there are some small improvements, some small regressions.
One big regression i saw i fixed in rG8d487668d09fb0e4e54f36207f07c1480ffabbfd, but i'm sure
this will expose many more pre-existing missed optimizations, as usual :S
llvm-compile-time-tracker.com thoughts on this:
http://llvm-compile-time-tracker.com/compare.php?from=e40315d2b4ed1e38962a8f33ff151693ed4ada63&to=c8289c0ecbf235da9fb0e3bc052e3c0d6bff5cf9&stat=instructions
* this does regress compile-time by +0.5% geomean (unsurprizingly)
* size impact varies; for ThinLTO it's actually an improvement
The largest fallout appears to be in GVN's load partial redundancy
elimination, it spends *much* more time in
`MemoryDependenceResults::getNonLocalPointerDependency()`.
Non-local `MemoryDependenceResults` is widely-known to be, uh, costly.
There does not appear to be a proper solution to this issue,
other than silencing the compile-time performance regression
by tuning cut-off thresholds in `MemoryDependenceResults`,
at the cost of potentially regressing run-time performance.
D84609 attempts to move in that direction, but the path is unclear
and is going to take some time.
If we look at stats before/after diffs, some excerpts:
* RawSpeed (the target) {F12360200}
* -14 (-73.68%) loops not rotated due to the header size (yay)
* -272 (-0.67%) `"Number of live out of a loop variables"` - good for vectorizer
* -3937 (-64.19%) common instructions hoisted
* +561 (+0.06%) x86 asm instructions
* -2 basic blocks
* +2418 (+0.11%) IR instructions
* vanilla test-suite + RawSpeed + darktable {F12360201}
* -36396 (-65.29%) common instructions hoisted
* +1676 (+0.02%) x86 asm instructions
* +662 (+0.06%) basic blocks
* +4395 (+0.04%) IR instructions
It is likely to be sub-optimal for when optimizing for code size,
so one might want to change tune pipeline by enabling sinking/hoisting
when optimizing for size.
Reviewed By: mkazantsev
Differential Revision: https://reviews.llvm.org/D84108
Summary:
PPC only supports the instruction selection for v16i8, v8i16, v4i32,
v2i64, v4f32 and v2f64 for ISD::SETCC, don't support the v1i128, so
v1i128 for ISD::SETCC will crash.
This patch is to set v1i128 to expand to avoid crash.
Reviewed By: steven.zhang
Differential Revision: https://reviews.llvm.org/D84238
expressions.
Also "fix" the longstanding bug where the computed size depends on the
order of the visitation. We could try to predict the allocation order
used by legalization, but it would never be 100% perfect. Until we
start fixing the addresses somehow (or have a more reliable allocation
scheme later), just try to compute the size based on the worst case
padding.
This patch uses the feature added in D79162 to fix the cost of a
sext/zext of a masked load, or a trunc for a masked store.
Previously, those were considered cheap or even free, but it's
not the case as we cannot split the load in the same way we would for
normal loads.
This updates the costs to better reflect reality, and adds a test for it
in test/Analysis/CostModel/ARM/cast.ll.
It also adds a vectorizer test that showcases the improvement: in some
cases, the vectorizer will now choose a smaller VF when
tail-predication is enabled, which results in better codegen. (Because
if it were to use a higher VF in those cases, the code we see above
would be generated, and the vmovs would block tail-predication later in
the process, resulting in very poor codegen overall)
Original Patch by Pierre van Houtryve
Differential Revision: https://reviews.llvm.org/D79163
Currently, getCastInstrCost has limited information about the cast it's
rating, often just the opcode and types. Sometimes there is a context
instruction as well, but it isn't trustworthy: for instance, when the
vectorizer is rating a plan, it calls getCastInstrCost with the old
instructions when, in fact, it's trying to evaluate the cost of the
instruction post-vectorization. Thus, the current system can get the
cost of certain casts incorrect as the correct cost can vary greatly
based on the context in which it's used.
For example, if the vectorizer queries getCastInstrCost to evaluate the
cost of a sext(load) with tail predication enabled, getCastInstrCost
will think it's free most of the time, but it's not always free. On ARM
MVE, a VLD2 group cannot be extended like a normal VLDR can. Similar
situations can come up with how masked loads can be extended when being
split.
To fix that, this path adds a new parameter to getCastInstrCost to give
it a hint about the context of the cast. It adds a CastContextHint enum
which contains the type of the load/store being created by the
vectorizer - one for each of the types it can produce.
Original patch by Pierre van Houtryve
Differential Revision: https://reviews.llvm.org/D79162
Optimize some specific immediates selection by materializing them with sub/mvn
instructions as opposed to loading them from the constant pool.
Patch by Ben Shi, powerman1st@163.com.
Differential Revision: https://reviews.llvm.org/D83745
If the mask input to getV4X86ShuffleImm8 only refers to a single source element (+ undefs) then canonicalize to a full broadcast.
getV4X86ShuffleImm8 defaults to inline values for undefs, which can be useful for shuffle widening/narrowing but does leave SimplifyDemanded* calls thinking the shuffle depends on unnecessary elements.
I'm still investigating what we should do more generally to avoid these undemanded elements, but broadcast cases was a simpler win.
