* Replace getUserCost with getInstructionCost, covering all cost kinds.
* Remove getInstructionLatency, it's not implemented by any backends, and we should fold the functionality into getUserCost (now getInstructionCost) to make it easier for targets to handle the cost kinds with their existing cost callbacks.
Original Patch by @samparker (Sam Parker)
Differential Revision: https://reviews.llvm.org/D79483
It would be better for CodeMetrics to use hasOneLiveUse while analyzing
static and called once callsites, since inline cost now uses
hasOneLiveUse instead of hasOneUse to avoid overpessimization on dead
constant cases (since this patch https://reviews.llvm.org/D109294).
This change has no noticeable influence now, but it helps improve the
accuracy of cost models of passes that use CodeMetrics.
Reviewed By: fhahn, nikic
Differential Revision: https://reviews.llvm.org/D130461
Per the documentation in Support/InstructionCost.h, the purpose of an invalid cost is so that clients can change behavior on impossible to cost inputs. CodeMetrics was instead asserting that invalid costs never occurred.
On a target with an incomplete cost model - e.g. RISCV - this means that transformations would crash on (falsely) invalid constructs - e.g. scalable vectors. While we certainly should improve the cost model - and I plan to do so in the near future - we also shouldn't be crashing. This violates the explicitly stated purpose of an invalid InstructionCost.
I updated all of the "easy" consumers where bailouts were locally obvious. I plan to follow up with loop unroll in a following change.
Differential Revision: https://reviews.llvm.org/D127131
As discussed in D112016, our current requirement of speculatability
for ephemeral is overly strict: What we really care about is that
the instruction will be DCEd once the assume is dropped. For that
it is sufficient that the instruction is side-effect free and not
a terminator.
In particular, this allows non-dereferenceable loads to be ephemeral
values.
Differential Revision: https://reviews.llvm.org/D112179
This reverts commit b7d870eae7 and the
subsequent fix "[Polly] Fix build after AssumptionCache change (D96168)"
(commit e6810cab09).
It caused indeterminism in the output, such that e.g. the
polly-x86_64-linux buildbot failed accasionally.
PR49043 exposed a problem when it comes to RAUW llvm.assumes. While
D96106 would fix it for GVNSink, it seems a more general concern. To
avoid future problems this patch moves away from the vector of weak
reference model used in the assumption cache. Instead, we track the
llvm.assume calls with a callback handle which will remove itself from
the cache if the call is deleted.
Fixes PR49043.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D96168
83daa49758 made loop-rotate more conservative in the presence of
function calls in the prepare-for-lto stage. The code did not properly
account for calls that are no actual function calls, like calls to
intrinsics. This patch updates the code to ensure only calls that are
lowered to actual calls are considered inline candidates.
D84108 exposed a bad interaction between inlining and loop-rotation
during regular LTO, which is causing notable regressions in at least
CINT2006/473.astar.
The problem boils down to: we now rotate a loop just before the vectorizer
which requires duplicating a function call in the preheader when compiling
the individual files ('prepare for LTO'). But this then prevents further
inlining of the function during LTO.
This patch tries to resolve this issue by making LoopRotate more
conservative with respect to rotating loops that have inline-able calls
during the 'prepare for LTO' stage.
I think this change intuitively improves the current situation in
general. Loop-rotate tries hard to avoid creating headers that are 'too
big'. At the moment, it assumes all inlining already happened and the
cost of duplicating a call is equal to just doing the call. But with LTO,
inlining also happens during full LTO and it is possible that a previously
duplicated call is actually a huge function which gets inlined
during LTO.
From the perspective of LV, not much should change overall. Most loops
calling user-provided functions won't get vectorized to start with
(unless we can infer that the function does not touch memory, has no
other side effects). If we do not inline the 'inline-able' call during
the LTO stage, we merely delayed loop-rotation & vectorization. If we
inline during LTO, chances should be very high that the inlined code is
itself vectorizable or the user call was not vectorizable to start with.
There could of course be scenarios where we inline a sufficiently large
function with code not profitable to vectorize, which would have be
vectorized earlier (by scalarzing the call). But even in that case,
there probably is no big performance impact, because it should be mostly
down to the cost-model to reject vectorization in that case. And then
the version with scalarized calls should also not be beneficial. In a way,
LV should have strictly more information after inlining and make more
accurate decisions (barring cost-model issues).
There is of course plenty of room for things to go wrong unexpectedly,
so we need to keep a close look at actual performance and address any
follow-up issues.
I took a look at the impact on statistics for
MultiSource/SPEC2000/SPEC2006. There are a few benchmarks with fewer
loops rotated, but no change to the number of loops vectorized.
Reviewed By: sanwou01
Differential Revision: https://reviews.llvm.org/D94232
There are several different types of cost that TTI tries to provide
explicit information for: throughput, latency, code size along with
a vague 'intersection of code-size cost and execution cost'.
The vectorizer is a keen user of RecipThroughput and there's at least
'getInstructionThroughput' and 'getArithmeticInstrCost' designed to
help with this cost. The latency cost has a single use and a single
implementation. The intersection cost appears to cover most of the
rest of the API.
getUserCost is explicitly called from within TTI when the user has
been explicit in wanting the code size (also only one use) as well
as a few passes which are concerned with a mixture of size and/or
a relative cost. In many cases these costs are closely related, such
as when multiple instructions are required, but one evident diverging
cost in this function is for div/rem.
This patch adds an argument so that the cost required is explicit,
so that we can make the important distinction when necessary.
Differential Revision: https://reviews.llvm.org/D78635
Remove SmallPtrSet include, replace with forward declaration and include SmallPtrSet.h in CodeMetrics.cpp directly.
Remove unused llvm::DataLayout/Instruction forward declarations.
to reflect the new license.
We understand that people may be surprised that we're moving the header
entirely to discuss the new license. We checked this carefully with the
Foundation's lawyer and we believe this is the correct approach.
Essentially, all code in the project is now made available by the LLVM
project under our new license, so you will see that the license headers
include that license only. Some of our contributors have contributed
code under our old license, and accordingly, we have retained a copy of
our old license notice in the top-level files in each project and
repository.
llvm-svn: 351636
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
I did this a long time ago with a janky python script, but now
clang-format has built-in support for this. I fed clang-format every
line with a #include and let it re-sort things according to the precise
LLVM rules for include ordering baked into clang-format these days.
I've reverted a number of files where the results of sorting includes
isn't healthy. Either places where we have legacy code relying on
particular include ordering (where possible, I'll fix these separately)
or where we have particular formatting around #include lines that
I didn't want to disturb in this patch.
This patch is *entirely* mechanical. If you get merge conflicts or
anything, just ignore the changes in this patch and run clang-format
over your #include lines in the files.
Sorry for any noise here, but it is important to keep these things
stable. I was seeing an increasing number of patches with irrelevant
re-ordering of #include lines because clang-format was used. This patch
at least isolates that churn, makes it easy to skip when resolving
conflicts, and gets us to a clean baseline (again).
llvm-svn: 304787
After r289755, the AssumptionCache is no longer needed. Variables affected by
assumptions are now found by using the new operand-bundle-based scheme. This
new scheme is more computationally efficient, and also we need much less
code...
llvm-svn: 289756
There was an efficiency problem with how we processed @llvm.assume in
ValueTracking (and other places). The AssumptionCache tracked all of the
assumptions in a given function. In order to find assumptions relevant to
computing known bits, etc. we searched every assumption in the function. For
ValueTracking, that means that we did O(#assumes * #values) work in InstCombine
and other passes (with a constant factor that can be quite large because we'd
repeat this search at every level of recursion of the analysis).
Several of us discussed this situation at the last developers' meeting, and
this implements the discussed solution: Make the values that an assume might
affect operands of the assume itself. To avoid exposing this detail to
frontends and passes that need not worry about it, I've used the new
operand-bundle feature to add these extra call "operands" in a way that does
not affect the intrinsic's signature. I think this solution is relatively
clean. InstCombine adds these extra operands based on what ValueTracking, LVI,
etc. will need and then those passes need only search the users of the values
under consideration. This should fix the computational-complexity problem.
At this point, no passes depend on the AssumptionCache, and so I'll remove
that as a follow-up change.
Differential Revision: https://reviews.llvm.org/D27259
llvm-svn: 289755
number of assume intrinsics.
The classical way to have a cache-friendly vector style container when
we need queue semantics for BFS instead of stack semantics for DFS is to
use an ever-growing vector and an index. Erasing from the front requires
O(size) work, and unless we expect the worklist to grow *very* large,
its probably cheaper to just grow and race down the list.
But that makes it more bad that we're putting the assume intrinsics in
this at all. We end up looking at the (by definition empty) use list to
see if they're ephemeral (when we've already put them in that set), etc.
Instead, directly populate the worklist with the operands when we mark
the assume intrinsics as ephemeral. Also, test the visited set *before*
putting things into the worklist so we don't accumulate the same value
in the list 100s of times.
It would be nice to use a set-vector for this but I think its useful to
test the set earlier to avoid repeatedly querying whether the same
instruction is safe to speculate.
Hopefully with these changes the number of values pushed onto the
worklist is smaller, and we avoid quadratic work by letting it grow as
necessary.
Differential Revision: https://reviews.llvm.org/D23396
llvm-svn: 279099
Remove implicit ilist iterator conversions from LLVMAnalysis.
I came across something really scary in `llvm::isKnownNotFullPoison()`
which relied on `Instruction::getNextNode()` being completely broken
(not surprising, but scary nevertheless). This function is documented
(and coded to) return `nullptr` when it gets to the sentinel, but with
an `ilist_half_node` as a sentinel, the sentinel check looks into some
other memory and we don't recognize we've hit the end.
Rooting out these scary cases is the reason I'm removing the implicit
conversions before doing anything else with `ilist`; I'm not at all
surprised that clients rely on badness.
I found another scary case -- this time, not relying on badness, just
bad (but I guess getting lucky so far) -- in
`ObjectSizeOffsetEvaluator::compute_()`. Here, we save out the
insertion point, do some things, and then restore it. Previously, we
let the iterator auto-convert to `Instruction*`, and then set it back
using the `Instruction*` version:
Instruction *PrevInsertPoint = Builder.GetInsertPoint();
/* Logic that may change insert point */
if (PrevInsertPoint)
Builder.SetInsertPoint(PrevInsertPoint);
The check for `PrevInsertPoint` doesn't protect correctly against bad
accesses. If the insertion point has been set to the end of a basic
block (i.e., `SetInsertPoint(SomeBB)`), then `GetInsertPoint()` returns
an iterator pointing at the list sentinel. The version of
`SetInsertPoint()` that's getting called will then call
`PrevInsertPoint->getParent()`, which explodes horribly. The only
reason this hasn't blown up is that it's fairly unlikely the builder is
adding to the end of the block; usually, we're adding instructions
somewhere before the terminator.
llvm-svn: 249925
This introduces the basic functionality to support "token types".
The motivation stems from the need to perform operations on a Value
whose provenance cannot be obscured.
There are several applications for such a type but my immediate
motivation stems from WinEH. Our personality routine enforces a
single-entry - single-exit regime for cleanups. After several rounds of
optimizations, we may be left with a terminator whose "cleanup-entry
block" is not entirely clear because control flow has merged two
cleanups together. We have experimented with using labels as operands
inside of instructions which are not terminators to indicate where we
came from but found that LLVM does not expect such exotic uses of
BasicBlocks.
Instead, we can use this new type to clearly associate the "entry point"
and "exit point" of our cleanup. This is done by having the cleanuppad
yield a Token and consuming it at the cleanupret.
The token type makes it impossible to obscure or otherwise hide the
Value, making it trivial to track the relationship between the two
points.
What is the burden to the optimizer? Well, it turns out we have already
paid down this cost by accepting that there are certain calls that we
are not permitted to duplicate, optimizations have to watch out for
such instructions anyway. There are additional places in the optimizer
that we will probably have to update but early examination has given me
the impression that this will not be heroic.
Differential Revision: http://reviews.llvm.org/D11861
llvm-svn: 245029
a cache of assumptions for a single function, and an immutable pass that
manages those caches.
The motivation for this change is two fold. Immutable analyses are
really hacks around the current pass manager design and don't exist in
the new design. This is usually OK, but it requires that the core logic
of an immutable pass be reasonably partitioned off from the pass logic.
This change does precisely that. As a consequence it also paves the way
for the *many* utility functions that deal in the assumptions to live in
both pass manager worlds by creating an separate non-pass object with
its own independent API that they all rely on. Now, the only bits of the
system that deal with the actual pass mechanics are those that actually
need to deal with the pass mechanics.
Once this separation is made, several simplifications become pretty
obvious in the assumption cache itself. Rather than using a set and
callback value handles, it can just be a vector of weak value handles.
The callers can easily skip the handles that are null, and eventually we
can wrap all of this up behind a filter iterator.
For now, this adds boiler plate to the various passes, but this kind of
boiler plate will end up making it possible to port these passes to the
new pass manager, and so it will end up factored away pretty reasonably.
llvm-svn: 225131
This is to be consistent with StringSet and ultimately with the standard
library's associative container insert function.
This lead to updating SmallSet::insert to return pair<iterator, bool>,
and then to update SmallPtrSet::insert to return pair<iterator, bool>,
and then to update all the existing users of those functions...
llvm-svn: 222334
We need to make sure that we visit all operands of an instruction before moving
deeper in the operand graph. We had been pushing operands onto the back of the work
set, and popping them off the back as well, meaning that we might visit an
instruction before visiting all of its uses that sit in between it and the call
to @llvm.assume.
To provide an explicit example, given the following:
%q0 = extractelement <4 x float> %rd, i32 0
%q1 = extractelement <4 x float> %rd, i32 1
%q2 = extractelement <4 x float> %rd, i32 2
%q3 = extractelement <4 x float> %rd, i32 3
%q4 = fadd float %q0, %q1
%q5 = fadd float %q2, %q3
%q6 = fadd float %q4, %q5
%qi = fcmp olt float %q6, %q5
call void @llvm.assume(i1 %qi)
%q5 is used by both %qi and %q6. When we visit %qi, it will be marked as
ephemeral, and we'll queue %q6 and %q5. %q6 will be marked as ephemeral and
we'll queue %q4 and %q5. Under the old system, we'd then visit %q4, which
would become ephemeral, %q1 and then %q0, which would become ephemeral as
well, and now we have a problem. We'd visit %rd, but it would not be marked as
ephemeral because we've not yet visited %q2 and %q3 (because we've not yet
visited %q5).
This will be covered by a test case in a follow-up commit that enables
ephemeral-value awareness in the SLP vectorizer.
llvm-svn: 219815
This adds a set of utility functions for collecting 'ephemeral' values. These
are LLVM IR values that are used only by @llvm.assume intrinsics (directly or
indirectly), and thus will be removed prior to code generation, implying that
they should be considered free for certain purposes (like inlining). The
inliner's cost analysis, and a few other passes, have been updated to account
for ephemeral values using the provided functionality.
This functionality is important for the usability of @llvm.assume, because it
limits the "non-local" side-effects of adding llvm.assume on inlining, loop
unrolling, etc. (these are hints, and do not generate code, so they should not
directly contribute to estimates of execution cost).
llvm-svn: 217335
The "noduplicate" attribute of call instructions is sometimes queried directly
and sometimes through the cannotDuplicate() predicate. This patch streamlines
all queries to use the cannotDuplicate() predicate. It also adds this predicate
to InvokeInst, to mirror what CallInst has.
llvm-svn: 204049
generic function calls and intrinsics. This is somewhat overlapping with
an existing intrinsic cost method, but that one seems targetted at
vector intrinsics. I'll merge them or separate their names and use cases
in a separate commit.
This sinks the test of 'callIsSmall' down into TTI where targets can
control it. The whole thing feels very hack-ish to me though. I've left
a FIXME comment about the fundamental design problem this presents. It
isn't yet clear to me what the users of this function *really* care
about. I'll have to do more analysis to figure that out. Putting this
here at least provides it access to proper analysis pass tools and other
such. It also allows us to more cleanly implement the baseline cost
interfaces in TTI.
With this commit, it is now theoretically possible to simplify much of
the inline cost analysis's handling of calls by calling through to this
interface. That conversion will have to happen in subsequent commits as
it requires more extensive restructuring of the inline cost analysis.
The CodeMetrics class is now really only in the business of running over
a block of code and aggregating the metrics on that block of code, with
the actual cost evaluation done entirely in terms of TTI.
llvm-svn: 173148
is free. The whole CodeMetrics API should probably be reworked more, but
this is enough to allow deleting the duplicate code there for computing
whether an instruction is free.
All of the passes using this have been updated to pull in TTI and hand
it to the CodeMetrics stuff. Further, a dead CodeMetrics API
(analyzeFunction) is nuked for lack of users.
llvm-svn: 173036
into their new header subdirectory: include/llvm/IR. This matches the
directory structure of lib, and begins to correct a long standing point
of file layout clutter in LLVM.
There are still more header files to move here, but I wanted to handle
them in separate commits to make tracking what files make sense at each
layer easier.
The only really questionable files here are the target intrinsic
tablegen files. But that's a battle I'd rather not fight today.
I've updated both CMake and Makefile build systems (I think, and my
tests think, but I may have missed something).
I've also re-sorted the includes throughout the project. I'll be
committing updates to Clang, DragonEgg, and Polly momentarily.
llvm-svn: 171366
directly.
This is in preparation for removing the use of the 'Attribute' class as a
collection of attributes. That will shift to the AttributeSet class instead.
llvm-svn: 171253
Similarly inlining of the function is inhibited, if that would duplicate the call (in particular inlining is still allowed when there is only one callsite and the function has internal linkage).
llvm-svn: 170704
Sooooo many of these had incorrect or strange main module includes.
I have manually inspected all of these, and fixed the main module
include to be the nearest plausible thing I could find. If you own or
care about any of these source files, I encourage you to take some time
and check that these edits were sensible. I can't have broken anything
(I strictly added headers, and reordered them, never removed), but they
may not be the headers you'd really like to identify as containing the
API being implemented.
Many forward declarations and missing includes were added to a header
files to allow them to parse cleanly when included first. The main
module rule does in fact have its merits. =]
llvm-svn: 169131
r165941: Resubmit the changes to llvm core to update the functions to
support different pointer sizes on a per address space basis.
Despite this commit log, this change primarily changed stuff outside of
VMCore, and those changes do not carry any tests for correctness (or
even plausibility), and we have consistently found questionable or flat
out incorrect cases in these changes. Most of them are probably correct,
but we need to devise a system that makes it more clear when we have
handled the address space concerns correctly, and ideally each pass that
gets updated would receive an accompanying test case that exercises that
pass specificaly w.r.t. alternate address spaces.
However, from this commit, I have retained the new C API entry points.
Those were an orthogonal change that probably should have been split
apart, but they seem entirely good.
In several places the changes were very obvious cleanups with no actual
multiple address space code added; these I have not reverted when
I spotted them.
In a few other places there were merge conflicts due to a cleaner
solution being implemented later, often not using address spaces at all.
In those cases, I've preserved the new code which isn't address space
dependent.
This is part of my ongoing effort to clean out the partial address space
code which carries high risk and low test coverage, and not likely to be
finished before the 3.2 release looms closer. Duncan and I would both
like to see the above issues addressed before we return to these
changes.
llvm-svn: 167222
We use the enums to query whether an Attributes object has that attribute. The
opaque layer is responsible for knowing where that specific attribute is stored.
llvm-svn: 165488
Simon Pilgrim