Commit Graph

39 Commits

Author SHA1 Message Date
Shinji Okumura e21a22a7a8 [Attributor] fix AANoUndef initialization
Currently, `AANoUndefImpl::initialize` mistakenly always indicates optimistic fixpoint for function returned position.
 This is because an associated value is `Function` in the case, and `isGuaranteedNotToBeUndefOrPoison` returns true for Function.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D86361
2020-08-22 05:06:14 +09:00
Johannes Doerfert 3edea15f9a [Attributor] Simplify comparison against constant null pointer
Comparison against null is a common pattern that usually is followed by
error handling code and the likes. We now use AANonNull to simplify
these comparisons optimistically in order to make more code dead early
on.

Reviewed By: uenoku

Differential Revision: https://reviews.llvm.org/D86145
2020-08-20 01:44:58 -05:00
Johannes Doerfert d01ad217ba [Attributor][FIX] Do not use cyclic arguments for `nonnull`
`AADereferenceable::getAssumedDereferenceableBytes()` is actually
deducing `dereferenceable_or_null`. We should not use that information
to deduce `nonnull`, since it doesn't imply `nonnull`.
2020-08-20 01:44:58 -05:00
Shinji Okumura 5e361e2aa4 [Attributor] Deduce noundef attribute
This patch introduces a new abstract attribute `AANoUndef` which corresponds to `noundef` IR attribute and deduce them.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D85184
2020-08-18 18:05:54 +09:00
Johannes Doerfert b27bdf955a [Attributor][FIX] Handle function pointers properly in AANonNull
Before we tired to create a dominator tree for a declaration when we
wanted to determine if the function pointer is `nonnull`. We now avoid
looking at global values if `Value::getPointerDereferenceableBytes` not
already determined `nonnull`.
2020-08-17 23:36:35 -05:00
Shinji Okumura 5f55a8193c [Attributor] Implement AAPotentialValues
This patch provides an implementation of `AAPotentialValues`.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D85632
2020-08-14 20:51:14 +09:00
Johannes Doerfert 3a033921ed [Attributor][NFC] Reformat tests after D85099
Reviewed By: sstefan1

Differential Revision: https://reviews.llvm.org/D85700
2020-08-12 01:04:19 -05:00
Arthur Eubanks 4a10029d7e [NewPM][Attributor] Pin tests with -attributor to legacy PM
All these tests already explicitly test against both legacy PM and NPM.

$ sed -i 's/ -attributor / -attributor -enable-new-pm=0 /g' $(rg --path-separator // -l -- -passes=)
$ sed -i 's/ -attributor-cgscc / -attributor-cgscc -enable-new-pm=0 /g' $(rg --path-separator // -l -- -passes=)

Now all tests in Transforms/Attributor/ pass under NPM.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D84813
2020-07-29 09:02:30 -07:00
sstefan1 e3d646c699 [Attributor][NFC] applying update_test_checks with --check-attributes
Summary:
All tests are updated, except wrapper.ll since it is not working nicely
with newly created functions.

Reviewers: jdoerfert, uenoku, baziotis, homerdin

Subscribers: arphaman, jfb, kuter, bbn, okura, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D84130
2020-07-20 08:17:34 +02:00
Nikita Popov f89f7da999 [IR] Convert null-pointer-is-valid into an enum attribute
The "null-pointer-is-valid" attribute needs to be checked by many
pointer-related combines. To make the check more efficient, convert
it from a string into an enum attribute.

In the future, this attribute may be replaced with data layout
properties.

Differential Revision: https://reviews.llvm.org/D78862
2020-05-15 19:41:07 +02:00
Johannes Doerfert c115a78f0d [Attributor] Make AAIsDead dependences optional to prevent top state
We should never give up on AAIsDead as it guards other AAs from
unreachable code (in which SSA properties are meaningless). We did
however use required dependences on some queries in AAIsDead which
caused us to invalidate AAIsDead if the queried AA got invalidated.
We now use optional dependences instead. The bug that exposed this is
added to the liveness.ll test and other test changes show the impact.

Bug report by @sdmitriev.
2020-05-11 15:32:47 -05:00
Johannes Doerfert 713ee3aa77 [Attributor] Use "simplify to constant" in genericValueTraversal
As we replace values with constants interprocedurally, we also need to
do this "look-through" step during the generic value traversal or we
would derive properties from replaced values. While this is often not
problematic, it is when we use the "kind" of a value for reasoning,
e.g., accesses to arguments allow `argmemonly`.
2020-05-10 19:06:11 -05:00
Johannes Doerfert 31c03b9223 [Attributor] Use existing helpers to determine IR facts
We now use getPointerDereferenceableBytes to determine `nonnull` and
`dereferenceable` facts from the IR. We also use getPointerAlignment in
AAAlign for the same reason. The latter can interfere with callbacks so
we do restrict it to non-function-pointers for now.
2020-05-10 19:06:10 -05:00
Johannes Doerfert edf0391491 [Attributor][FIX] Record dependences for assumed dead abstract attributes
In a recent patch we introduced a problem with abstract attributes that
were assumed dead at some point. Since `Attributor::updateAA` was
introduced in 95e0d28b71, we did not
remember the dependence on the liveness AA when an abstract attribute
was assumed dead and therefore not updated.

Explicit reproducer added in liveness.ll.

---

Single run of the Attributor module and then CGSCC pass (oldPM)
for SPASS/clause.c (~10k LLVM-IR loc):

Before:
```
calls to allocation functions: 509242 (345483/s)
temporary memory allocations: 98666 (66937/s)
peak heap memory consumption: 18.60MB
peak RSS (including heaptrack overhead): 103.29MB
total memory leaked: 269.10KB
```

After:
```
calls to allocation functions: 529332 (355494/s)
temporary memory allocations: 102107 (68574/s)
peak heap memory consumption: 19.40MB
peak RSS (including heaptrack overhead): 102.79MB
total memory leaked: 269.10KB
```

Difference:
```
calls to allocation functions: 20090 (1339333/s)
temporary memory allocations: 3441 (229400/s)
peak heap memory consumption: 801.45KB
peak RSS (including heaptrack overhead): 0B
total memory leaked: 0B
```
2020-05-07 17:00:50 -05:00
Johannes Doerfert 95e0d28b71 [Attributor] Remember only necessary dependences
Before we eagerly put dependences into the QueryMap as soon as we
encountered them (via `Attributor::getAAFor<>` or
`Attributor::recordDependence`). Now we will wait to see if the
dependence is useful, that is if the target is not already in a fixpoint
state at the end of the update. If so, there is no need to record the
dependence at all.

Due to the abstraction via `Attributor::updateAA` we will now also treat
the very first update (during attribute creation) as we do subsequent
updates.

Finally this resolves the problematic usage of QueriedNonFixAA.

---

Single run of the Attributor module and then CGSCC pass (oldPM)
for SPASS/clause.c (~10k LLVM-IR loc):

Before:
```
calls to allocation functions: 554675 (389245/s)
temporary memory allocations: 101574 (71280/s)
peak heap memory consumption: 28.46MB
peak RSS (including heaptrack overhead): 116.26MB
total memory leaked: 269.10KB
```

After:
```
calls to allocation functions: 512465 (345559/s)
temporary memory allocations: 98832 (66643/s)
peak heap memory consumption: 22.54MB
peak RSS (including heaptrack overhead): 106.58MB
total memory leaked: 269.10KB
```

Difference:
```
calls to allocation functions: -42210 (-727758/s)
temporary memory allocations: -2742 (-47275/s)
peak heap memory consumption: -5.92MB
peak RSS (including heaptrack overhead): 0B
total memory leaked: 0B
```
2020-05-03 22:01:51 -05:00
Johannes Doerfert 2f97b8b891 [Attributor][NFC] Proactively ask for `nocapure` on call site arguments
This minimizes test noise later on and is in line with other attributes
we derive proactively.
2020-05-03 21:38:06 -05:00
Tyker 42431da895 [AssumeBundles] Use assume bundles in isKnownNonZero
Summary: Use nonnull and dereferenceable from an assume bundle in isKnownNonZero

Reviewers: jdoerfert, nikic, lebedev.ri, reames, fhahn, sstefan1

Reviewed By: jdoerfert

Subscribers: fhahn, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D76149
2020-04-24 20:41:51 +02:00
Johannes Doerfert 10ff24d853 [Attributor][NFC] Remove and update old check lines 2020-04-24 01:08:23 -05:00
Johannes Doerfert e2b53a4c05 [Attributor][NFC] Remove obsolete option from tests
Since D76871 it is sufficient to run `opt -atributor` or
`-attributor-cgscc`.
2020-04-21 15:22:10 -05:00
Johannes Doerfert 0741dec27b [Attributor][FIX] Handle droppable uses when replacing values
Since we use the fact that some uses are droppable in the Attributor we
need to handle them explicitly when we replace uses. As an example, an
assumed dead value can have live droppable users. In those we cannot
replace the value simply by an undef. Instead, we either drop the uses
(via `dropDroppableUses`) or keep them as they are. In this patch we do
both, depending on the situation. For values that are dead but not
necessarily removed we keep droppable uses around because they contain
information we might be able to use later. For values that are removed
we drop droppable uses explicitly to avoid replacement with undef.
2020-04-16 00:56:08 -05:00
Johannes Doerfert 3ca54f4595 [Attributor] Unify testing (=updates,prefixes,run configurations,...)
When the Attributor was created the test update scripts were not well
suited to deal with the challenges of IR attribute checking. This
partially improved.

Since then we also added three additional configurations that need
testing; in total we now have the following four:
{ TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) }

Finally, the number of developers and tests grew rapidly (partially due
to the addition of ArgumentPromotion and IPConstantProp tests), which
resulted in tests only being run in some configurations, different
prefixes being used, and different "styles" of checks being used.

Due to the above reasons I believed we needed to take another look at
the test update scripts. While we started to use them, via UTC_ARGS:
--enable/disable, the other problems remained. To improve the testing
situation for *all* configurations, to simplify future updates to the
test, and to help identify subtle effects of future changes, we now use
the test update scripts for (almost) all Attributor tests.

An exhaustive prefix list minimizes the number of check lines and makes
it easy to identify and compare configurations.

Tests have been adjusted in the process but we tried to keep their
intend unchanged.

Reviewed By: sstefan1

Differential Revision: https://reviews.llvm.org/D76588
2020-04-15 19:59:51 -05:00
Florian Hahn 0c22cb0fd7 Temporarily revert "[Attributor] Unify testing (=updates,...)"
This patch reverts the 2 patches below, as on most systems the disabled
tests actually pass and that causes most bots to be red, including
  http://green.lab.llvm.org/green/job/clang-stage1-RA/8541/
  http://green.lab.llvm.org/green/job/clang-stage1-cmake-RA-expensive/15646/
  http://lab.llvm.org:8011/builders/clang-ppc64le-linux-lnt/builds/23690
  http://lab.llvm.org:8011/builders/sanitizer-x86_64-linux-bootstrap/builds/16751

* [Attributor] Disable three tests until the SCC update bug was fixed
  commit 2ae1a76c27.

* [Attributor] Unify testing (=updates,prefixes,run configurations,...)
  2bcf5793e1.
2020-04-09 11:11:50 +01:00
Johannes Doerfert 2bcf5793e1 [Attributor] Unify testing (=updates,prefixes,run configurations,...)
When the Attributor was created the test update scripts were not well
suited to deal with the challenges of IR attribute checking. This
partially improved.

Since then we also added three additional configurations that need
testing; in total we now have the following four:
{ TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) }

Finally, the number of developers and tests grew rapidly (partially due
to the addition of ArgumentPromotion and IPConstantProp tests), which
resulted in tests only being run in some configurations, different
prefixes being used, and different "styles" of checks being used.

Due to the above reasons I believed we needed to take another look at
the test update scripts. While we started to use them, via UTC_ARGS:
--enable/disable, the other problems remained. To improve the testing
situation for *all* configurations, to simplify future updates to the
test, and to help identify subtle effects of future changes, we now use
the test update scripts for (almost) all Attributor tests.

An exhaustive prefix list minimizes the number of check lines and makes
it easy to identify and compare configurations.

Tests have been adjusted in the process but we tried to keep their
intend unchanged.

Reviewed By: sstefan1

Differential Revision: https://reviews.llvm.org/D76588
2020-04-08 22:52:46 -05:00
Johannes Doerfert 5699d08b79 [Attributor] Use knowledge retained in llvm.assume (operand bundles)
This patch integrates operand bundle llvm.assumes [0] with the
Attributor. Most IRAttributes will now look at uses of the associated
value and if there are llvm.assume operand bundle uses with the right
tag we will check if they are in the must-be-executed-context (around
the context instruction). Droppable users, which is currently only
llvm::assume, are handled special in some places now as well.

[0] http://lists.llvm.org/pipermail/llvm-dev/2019-December/137632.html

Reviewed By: uenoku

Differential Revision: https://reviews.llvm.org/D74888
2020-03-24 15:33:40 -05:00
Pankaj Gode bf990530ae [Attributor] Improve noalias preservation using reachability
Resolution for below fixme:
(ii) Check whether the value is captured in the scope using AANoCapture.
FIXME: This is conservative though, it is better to look at CFG and
             check only uses possibly executed before this callsite.

Propagates caller argument's noalias attribute to callee.

Reviewed by: jdoerfert, uenoku

Reviewers: jdoerfert, sstefan1, uenoku

Subscribers: uenoku, sstefan1, hiraditya, llvm-commits

Differential Revision: https://reviews.llvm.org/D71617
2020-03-13 21:09:08 +05:30
Hideto Ueno bdcbdb4848 [Attributor] Deduction based on path exploration
This patch introduces the propagation of known information based on path exploration.
For example,
```
int u(int c, int *p){
  if(c) {
     return *p;
  } else {
     return *p + 1;
  }
}
```
An argument `p` is dereferenced whatever c's value is.

For an instruction `CtxI`, we accumulate branch instructions in the must-be-executed-context of `CtxI` and then, we take the conjunction of the successors' known state.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D65593
2020-03-09 14:29:26 +09:00
Hideto Ueno 2c0edbf19c [Attributor] Use AssumptionCache in AANonNullFloating::initialize 2020-02-25 13:00:03 +09:00
Johannes Doerfert b6e5808b21 [Attributor][NFC] Prepare some tests to be used with update test script 2020-02-20 02:44:05 -06:00
Hideto Ueno e253cdda35 [MustExecute] Add backward exploration for must-be-executed-context
Summary:
As mentioned in D71974, it is useful for must-be-executed-context to explore CFG backwardly.
This patch is ported from parts of D64975. We use a dominator tree to find the previous context if
a dominator tree is available.

Reviewers: jdoerfert, hfinkel, baziotis, sstefan1

Reviewed By: jdoerfert

Subscribers: hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D74817
2020-02-20 14:49:30 +09:00
Johannes Doerfert b2c76002ca [Attributor] Ignore uses if a value is simplified
If we have a replacement for a value, via AAValueSimplify, the original
value will lose all its uses. Thus, as long as a value is simplified we
can skip the uses in checkForAllUses, given that these uses are
transitive uses for the simplified version and will therefore affect the
simplified version as necessary.

Since this allowed us to remove calls without side-effects and a known
return value, we need to make sure not to eliminate `musttail` calls.
Those we keep around, or later remove the entire `musttail` call chain.
2020-02-12 17:36:38 -06:00
Johannes Doerfert 86509e8c3b [Attributor] Use assumed information to determine side-effects
We relied on wouldInstructionBeTriviallyDead before but that functions
does not take assumed information, especially for calls, into account.
The replacement, AAIsDead::isAssumeSideEffectFree, does.

This change makes AAIsDeadCallSiteReturn more complex as we can have
a dead call or only dead users.

The test have been modified to include a side effect where there was
none in order to keep the coverage.

Reviewed By: sstefan1

Differential Revision: https://reviews.llvm.org/D73311
2020-02-12 17:36:38 -06:00
Johannes Doerfert 4c62a35860 [Attributor][FIX] Remove duplicate, half-broken functionality
The changeXXXAfterManifest functions are better suited to deal with
changes so we should prefer them. These functions also recursively
delete dead instructions which is why we see test changes.
2020-02-11 11:02:59 -06:00
Johannes Doerfert 53992c7bf7 [Attributor] Modularize AANoAliasCallSiteArgument to simplify extensions
This patch modularizes the way we check for no-alias call site arguments
by putting the existing logic into helper functions. The reasoning was
not changed but special cases for readonly/readnone were added.
2020-01-28 23:39:29 -06:00
Roman Lebedev a6492e2271
[IR] Value::getPointerAlignment(): handle pointer constants
Summary:
New `@test13` in `Attributor/align.ll` is the main motivation - `null` pointer
really does not limit our alignment knowledge, in fact it is fully aligned
since it has no bits set.

Here we don't special-case `null` pointer because it is somewhat controversial
to add one more place where we enforce that `null` pointer is zero,
but instead we do the more general thing of trying to perform constant-fold
of pointer constant to an integer, and perform alignment inferrment on that.

Reviewers: jdoerfert, gchatelet, courbet, sstefan1

Reviewed By: jdoerfert

Subscribers: hiraditya, arphaman, jfb, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D73131
2020-01-22 01:32:46 +03:00
Johannes Doerfert 412a0101a9 [Attributor][FIX] Do not derive nonnull and dereferenceable w/o access
An inbounds GEP results in poison if the value is not "inbounds", not in
UB. We accidentally derived nonnull and dereferenceable from these
inbounds GEPs even in the absence of accesses that would make the poison
to UB.
2020-01-03 10:43:40 -06:00
Johannes Doerfert 751336340d [Attributor] Function signature rewrite infrastructure
As part of the Attributor manifest we want to change the signature of
functions. This patch introduces a fairly generic interface to do so.
As a first, very simple, use case, we remove unused arguments. A second
use case, pointer privatization, will be committed with this patch as
well.

A lot of the code and ideas are taken from argument promotion and we
run all argument promotion tests through this framework as well.

Reviewed By: uenoku

Differential Revision: https://reviews.llvm.org/D68765
2019-12-31 02:31:33 -06:00
Johannes Doerfert 2888019871 [Attributor] Annotate the memory behavior of call site arguments
Especially for callbacks, annotating the call site arguments is
important. Doing so exposed a too strong dependence of AAMemoryBehavior
on AANoCapture since we handle the case of potentially captured pointers
explicitly.

The changes to the tests are all mechanical.
2019-12-31 01:33:21 -06:00
Johannes Doerfert a7d992c0f2 [ValueTracking] Allow context-sensitive nullness check for non-pointers
Summary:
Same as D60846 and D69571 but with a fix for the problem encountered
after them. Both times it was a missing context adjustment in the
handling of PHI nodes.

The reproducers created from the bugs that caused the old commits to be
reverted are included.

Reviewers: nikic, nlopes, mkazantsev, spatel, dlrobertson, uabelho, hakzsam, hans

Subscribers: hiraditya, bollu, asbirlea, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D71181
2019-12-09 15:15:52 -06:00
Hideto Ueno 96552036e3 [Attributor] Copy or port test cases related to Attributor to` Attributor` test folder
Summary:
This patch moves the test cases related to Attributor to `Transforms/Attributor` folder.
We have used `Transforms/FunctionAttrs` as the primary folder for Attributor test but we need to change testing way now.

For the test cases which I think functionattrs doesn't infer anything something like (willreturn, nosync, value-simplify, h2s ..etc), I moved them with the command `git mv`.

For the test cases in which functoinattrs and attributor are tested, I copied the test to the folder and remove the check only used by functoinattrs.

Reviewers: jdoerfert, sstefan1

Reviewed By: jdoerfert

Subscribers: jfb, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D70843
2019-12-02 15:36:29 +00:00