llvm-svn: 277921

Not actually used yet...

llvm-svn: 277919

llvm-svn: 277916

A parameter was documented with the wrong name.
No functionality change is intended.

llvm-svn: 277915

Reasoning about a select in terms of a min or max allows us to derive a
tigher bound on the result.

llvm-svn: 277914

No functional change is intended.

llvm-svn: 277913

The current approach isn't a long-term viable pattern. Given the set of
architectures A, vendors V, operating systems O, and environments E, it
does |A| * |V| * |O| * |E| * 4! tests. As LLVM grows, this test keeps
getting slower, despite my working very hard to make it get some
"optimizations" even in -O0 builds in order to lower the constant
factors. Fundamentally, we're doing an unreasonable amount of work.i

Looking at the specific thing being tested -- the goal seems very
clearly to be testing the *permutations*, not the *combinations*. The
combinations are driving up the complexity much more than anything else.

Instead, test every possible value for a given triple entry in every
permutation of *some* triple. This really seems to cover the core goal
of the test. Every single possible triple component is tested in every
position. But because we keep the rest of the triple constant, it does
so in a dramatically more scalable amount of time. With this model we do
(|A| + |V| + |O| + |E|) * 4! tests.

For me on a debug build, this goes from running for 19 seconds to 19
milliseconds, or a 1000x improvement. This makes a world of difference
for the critical path of 'ninja check-llvm' and other extremely common
workflows.

Thanks to Renato, Dean, and David for the helpful review comments and
helping me refine the explanation of the change.

Differential Revision: https://reviews.llvm.org/D23156

llvm-svn: 277912

Reviewers: majnemer

Subscribers: mcrosier, llvm-commits

Differential Revision: https://reviews.llvm.org/D23231

llvm-svn: 277910

GVN-Hoist appears to miscompile llvm-testsuite
SingleSource/Benchmarks/Misc/fbench.c at the moment.

I filed http://llvm.org/PR28880

This reverts commit r277786.

llvm-svn: 277909

Summary:
This is the 4c patch of the coroutine series. CoroElide pass now checks if PostSplit coro.begin
is referenced by coro.subfn.addr intrinsics. If so replace coro.subfn.addrs with an appropriate coroutine
subfunction associated with that coro.begin.

Documentation and overview is here: http://llvm.org/docs/Coroutines.html.

Upstreaming sequence (rough plan)
1.Add documentation. (https://reviews.llvm.org/D22603)
2.Add coroutine intrinsics. (https://reviews.llvm.org/D22659)
3.Add empty coroutine passes. (https://reviews.llvm.org/D22847)
4.Add coroutine devirtualization + tests.
ab) Lower coro.resume and coro.destroy (https://reviews.llvm.org/D22998)
c) Do devirtualization <= we are here
5.Add CGSCC restart trigger + tests.
6.Add coroutine heap elision + tests.
7.Add the rest of the logic (split into more patches)

Reviewers: majnemer

Subscribers: mehdi_amini, llvm-commits

Differential Revision: https://reviews.llvm.org/D23229

llvm-svn: 277908

It breaks ExecutionEngine/OrcLazy/weak-function.ll on most bots.

Script:
--
...
--
Exit Code: 1

Command Output (stderr):
--
Could not find main function.

llvm-svn: 277907

The following function currently relies on tail-merging for if
conversion to succeed. The common tail of cond_true and cond_false is
extracted, and this then forms a diamond pattern that can be
successfully if converted.

If this block does not get extracted, either because tail-merging is
disabled or the threshold is higher, we should still recognize this
pattern and if-convert it.
define i32 @t2(i32 %a, i32 %b) nounwind {
entry:
	%tmp1434 = icmp eq i32 %a, %b		; <i1> [#uses=1]
	br i1 %tmp1434, label %bb17, label %bb.outer

bb.outer:		; preds = %cond_false, %entry
	%b_addr.021.0.ph = phi i32 [ %b, %entry ], [ %tmp10, %cond_false ]
	%a_addr.026.0.ph = phi i32 [ %a, %entry ], [ %a_addr.026.0, %cond_false ]
	br label %bb

bb:		; preds = %cond_true, %bb.outer
	%indvar = phi i32 [ 0, %bb.outer ], [ %indvar.next, %cond_true ]
	%tmp. = sub i32 0, %b_addr.021.0.ph
	%tmp.40 = mul i32 %indvar, %tmp.
	%a_addr.026.0 = add i32 %tmp.40, %a_addr.026.0.ph
	%tmp3 = icmp sgt i32 %a_addr.026.0, %b_addr.021.0.ph
	br i1 %tmp3, label %cond_true, label %cond_false

cond_true:		; preds = %bb
	%tmp7 = sub i32 %a_addr.026.0, %b_addr.021.0.ph
	%tmp1437 = icmp eq i32 %tmp7, %b_addr.021.0.ph
	%indvar.next = add i32 %indvar, 1
	br i1 %tmp1437, label %bb17, label %bb

cond_false:		; preds = %bb
	%tmp10 = sub i32 %b_addr.021.0.ph, %a_addr.026.0
	%tmp14 = icmp eq i32 %a_addr.026.0, %tmp10
	br i1 %tmp14, label %bb17, label %bb.outer

bb17:		; preds = %cond_false, %cond_true, %entry
	%a_addr.026.1 = phi i32 [ %a, %entry ], [ %tmp7, %cond_true ], [ %a_addr.026.0, %cond_false ]
	ret i32 %a_addr.026.1
}

Without tail-merging or diamond-tail if conversion:
LBB1_1:                                 @ %bb
                                        @ =>This Inner Loop Header: Depth=1
        cmp     r0, r1
        ble     LBB1_3
@ BB#2:                                 @ %cond_true
                                        @   in Loop: Header=BB1_1 Depth=1
        subs    r0, r0, r1
        cmp     r1, r0
        it      ne
        cmpne   r0, r1
        bgt     LBB1_4
LBB1_3:                                 @ %cond_false
                                        @   in Loop: Header=BB1_1 Depth=1
        subs    r1, r1, r0
        cmp     r1, r0
        bne     LBB1_1
LBB1_4:                                 @ %bb17
        bx      lr

With diamond-tail if conversion, but without tail-merging:
@ BB#0:                                 @ %entry
        cmp     r0, r1
        it      eq
        bxeq    lr
LBB1_1:                                 @ %bb
                                        @ =>This Inner Loop Header: Depth=1
        cmp     r0, r1
        ite     le
        suble   r1, r1, r0
        subgt   r0, r0, r1
        cmp     r1, r0
        bne     LBB1_1
@ BB#2:                                 @ %bb17
        bx      lr

llvm-svn: 277905

ScanInstructions is now 2 functions:
AnalyzeBranches and ScanInstructions. ScanInstructions also now takes a
pair of arguments delimiting the instructions to be scanned. This will
be used for forked diamond support to re-scan only a portion of the
block.

llvm-svn: 277904

llvm-svn: 277903

Skipping debug instructions occurrs repeatedly, factor it out.

llvm-svn: 277902

This reverts commit r277877.
Try to appease clang-x64-ninja-win7 buildbot.

llvm-svn: 277901

This adds partial support for weak functions to the CompileOnDemandLayer by
modifying the addLogicalModule method to check for existing stub definitions
before building a new stub for a weak function. This scheme is sufficient to
support ODR definitions, but fails for general weak definitions if strong
definition is encountered after the first weak definition. (A more extensive
refactor will be required to fully support weak symbols).

This patch does *not* add weak symbol support to RuntimeDyld: I hope to add
that in the near future.

llvm-svn: 277896

llvm-svn: 277893

llvm-svn: 277892

Fixes PR28764.  Right now there is no way to test this, but (as
mentioned on the PR) with Michael Zolotukhin's yet to be checked in
LoopSimplify verfier, 8 of the llvm-lit tests for IRCE crash.

llvm-svn: 277891

This resubmits a3770391c5fb64108d565e12f61dd77ce71b5b4f,
which was reverted due to breakages on non-Windows machines.

Due to differences in template instantiation rules on Microsoft
and non-Microsoft platforms, a member access restriction was
triggering on non-Microsoft compilers.  Previously, a friend
declaration for std::vector<> had been introduced into the
DebugMap class to make the member access restriction pass,
but the introduction of support for SmallVector<> meant that
an additional friend declaration would need to be added.

This didn't really make a lot of sense since the user of the
macro is probably only using one type (SmallVector<>, vector<>,
etc) and we could in theory add support for even more types
to this macro in the future (e.g. std::deque), so rather than
add another friend declaration, I just made the type being
referenced a public nested typedef instead of a private nested
typedef.

llvm-svn: 277888

Note that this fold really belongs in InstSimplify.
Refactoring here anyway as an intermediate step because
there's a planned addition to this function in D23134.

Differential Revision: https://reviews.llvm.org/D23223

llvm-svn: 277883

This breaks building dsymutil, causing my local build and many bots to
fail.

This reverts r277870.

llvm-svn: 277881

Summary:
Originally the plan was to use the custom worklist to do some block popping,
and because we don't actually need a visited set. The custom one we have
here is slightly broken, and it's not worth fixing vs using depth_first_iterator since we aren't going to go the route we originally
were.

Fixes PR28874
Reviewers: george.burgess.iv

Subscribers: llvm-commits, gberry

Differential Revision: https://reviews.llvm.org/D23187

llvm-svn: 277880

It was breaking the -Werror build.

llvm-svn: 277878

This fixes PR28825. The problem was that we only checked if a value from
a created inner loop is used in the outer loop, and fixed LCSSA for
them. But we missed to fixup LCSSA for values used in exits of the outer
loop.

llvm-svn: 277877

llvm-svn: 277876

This is running opt without specifying a triple, which isn't correct.

llvm-svn: 277875

llvm-svn: 277873

Summary:
Rewrite domination verifier to handle local domination as well.
This catches a bug Geoff Berry noticed.

Reviewers: george.burgess.iv

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D23184

llvm-svn: 277872

Until now, our use case for the visitor has been to take a stream of bytes
representing a type stream, deserialize the records in sequence, and do
something with them, where "something" is determined by how the user
implements a particular set of callbacks on an abstract class.

For actually writing PDBs, however, we want to do the reverse. We have
some kind of description of the list of records in their in-memory format,
and we want to process each one. Perhaps by serializing them to a byte
stream, or perhaps by converting them from one description format (Yaml)
to another (in-memory representation).

This was difficult in the current model because deserialization and
invoking the callbacks were tightly coupled.

With this patch we change this so that TypeDeserializer is itself an
implementation of the particular set of callbacks. This decouples
deserialization from the iteration over a list of records and invocation
of the callbacks.  TypeDeserializer is initialized with another
implementation of the callback interface, so that upon deserialization it
can pass the deserialized record through to the next set of callbacks. In
a sense this is like an implementation of the Decorator design pattern,
where the Deserializer is a decorator.

This will be useful for writing Pdbs from yaml, where we have a
description of the type records in Yaml format. In this case, the visitor
implementation would have each visitation callback method implemented in
such a way as to extract the proper set of fields from the Yaml, and it
could maintain state that builds up a list of these records. Finally at
the end we can pass this information through to another set of callbacks
which serializes them into a byte stream.

Reviewed By: majnemer, ruiu, rnk
Differential Revision: https://reviews.llvm.org/D23177

llvm-svn: 277871

Currently YAML sequences require std::vectors. All of the methods that the
YAML parser accesses though are present in SmallVector, so there's no
reason we can't support SmallVector inherently. This patch does that.

Reviewed By: majnemer
Differential Revision: https://reviews.llvm.org/D23213

llvm-svn: 277870

Summary:
This is the setting of the Vulkan closed source driver.

It decreases the max wave count from 10 to 8.

26010 shaders in 14650 tests
Totals:
VGPRS: 829593 -> 808440 (-2.55 %)
Spilled SGPRs: 81878 -> 42226 (-48.43 %)
Spilled VGPRs: 367 -> 358 (-2.45 %)
Scratch VGPRs: 1764 -> 1748 (-0.91 %) dwords per thread
Code Size: 36677864 -> 35923932 (-2.06 %) bytes

There is a massive decrease in SGPR spilling in general and -7.4% spilled
VGPRs for DiRT Showdown (= SGPRs spilled to scratch?)

Reviewers: arsenm, tstellarAMD, nhaehnle

Subscribers: arsenm, llvm-commits, kzhuravl

Differential Revision: https://reviews.llvm.org/D23034

llvm-svn: 277867

Summary: Thumb2 supports encoding immediates with specific patterns into mov.w by splatting the low 8 bits into other bytes.

I'm resubmitting this patch. The test case in the original commit
r277610 does not specify triple, so builds with differnt default triple
will have different output.

This patch fixed trile as thumb-darwin-apple.

Reviewers: john.brawn, jmolloy, bruno

Subscribers: jmolloy, aemerson, rengolin, samparker, llvm-commits

Differential Revision: https://reviews.llvm.org/D23090

llvm-svn: 277865

llvm-svn: 277864

llvm-svn: 277863

llvm-svn: 277861

Replace hot-callsite based heuristic to use its own threshold parameter instead of share inline-hint parameter

Summary: Hot callsites should have higher threshold than inline hints. This patch uses separate threshold parameter for hot callsites.

Reviewers: davidxl, eraman

Subscribers: Prazek, llvm-commits

Differential Revision: https://reviews.llvm.org/D22368

llvm-svn: 277860

Differential Revision: https://reviews.llvm.org/D23185

llvm-svn: 277859

The analysis manager was made not optional and turned into a
reference instead of a pointer in r272978. Some comments were
still refering to the previous behavior.

llvm-svn: 277857