Commit Graph

345 Commits

Author SHA1 Message Date
Quentin Colombet
8934fe31f3 [RegAllocGreedy] Record missed hint for late recoloring.
In https://reviews.llvm.org/D25347, Geoff noticed that we still have
useless copy that we can eliminate after register allocation. At the
time the allocation is chosen for those copies, they are not useless
but, because of changes in the surrounding code, later on they might
become useless.
The Greedy allocator already has a mechanism to deal with such cases
with a late recoloring. However, we missed to record the some of the
missed hints.

This commit fixes that.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@287070 91177308-0d34-0410-b5e6-96231b3b80d8
2016-11-16 01:07:12 +00:00
Tom Stellard
107362cf5f RegAllocGreedy: Properly initialize this pass, so that -run-pass will work
Reviewers: qcolombet, MatzeB

Subscribers: wdng, llvm-commits

Differential Revision: https://reviews.llvm.org/D26572

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@286895 91177308-0d34-0410-b5e6-96231b3b80d8
2016-11-14 21:50:13 +00:00
Wei Mi
85bacc4129 [RegAllocGreedy] Another fix about NewVRegs for last chance recoloring after r281783.
About when we should move a vreg from CurrentNewVRegs to NewVRegs,
if the vreg in CurrentNewVRegs was added into RecoloringCandidate and was
evicted, it shouldn't be added to NewVRegs because its physical register
will be restored at the end of tryLastChanceRecoloring after the recoloring
failed. If the vreg in CurrentNewVRegs was not in RecoloringCandidate, i.e.
it was evicted in selectOrSplitImpl inside tryRecoloringCandidates, its
physical register will not be restored even if the recoloring failed. In
that case, we need to add the vreg to NewVRegs.

Same as r281783, the problem was seen on out-of-tree target and we didn't
have a test case that reproduce the problem with in-tree targets.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@286259 91177308-0d34-0410-b5e6-96231b3b80d8
2016-11-08 18:19:36 +00:00
Quentin Colombet
154f790710 [RAGreedy] Empty live-ranges always succeed in last chance recoloring.
Relax the constraint for empty live-ranges while doing last chance
recoloring. Indeed, those live-ranges do not need an actual color to be
fond for the recoloring to work.
Empty live-range may happen as a result of splitting/spilling.

Unfortunately no test case for in-tree targets.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@284152 91177308-0d34-0410-b5e6-96231b3b80d8
2016-10-13 19:27:48 +00:00
Dylan McKay
25d9f1156f [RegAllocGreedy] Attempt to split unspillable live intervals
Summary:
Previously, when allocating unspillable live ranges, we would never
attempt to split. We would always bail out and try last ditch graph
recoloring.

This patch changes this by attempting to split all live intervals before
performing recoloring.

This fixes LLVM bug PR14879.

I can't add test cases for any backends other than AVR because none of
them have small enough register classes to trigger the bug.

Reviewers: qcolombet

Subscribers: MatzeB

Differential Revision: https://reviews.llvm.org/D25070

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@283838 91177308-0d34-0410-b5e6-96231b3b80d8
2016-10-11 01:04:36 +00:00
Mehdi Amini
3ffe113e11 Turn cl::values() (for enum) from a vararg function to using C++ variadic template
The core of the change is supposed to be NFC, however it also fixes
what I believe was an undefined behavior when calling:

 va_start(ValueArgs, Desc);

with Desc being a StringRef.

Differential Revision: https://reviews.llvm.org/D25342

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@283671 91177308-0d34-0410-b5e6-96231b3b80d8
2016-10-08 19:41:06 +00:00
Mehdi Amini
67f335d992 Use StringRef in Pass/PassManager APIs (NFC)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@283004 91177308-0d34-0410-b5e6-96231b3b80d8
2016-10-01 02:56:57 +00:00
Dylan McKay
eabad28062 Revert "[RegAllocGreedy] Attempt to split unspillable live intervals"
It was accidentally committed.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@282855 91177308-0d34-0410-b5e6-96231b3b80d8
2016-09-30 14:05:15 +00:00
Dylan McKay
ea9ff22768 [RegAllocGreedy] Attempt to split unspillable live intervals
Summary:
Previously, when allocating unspillable live ranges, we would never
attempt to split. We would always bail out and try last ditch graph
recoloring.

This patch changes this by attempting to split all live intervals before
performing recoloring.

This fixes LLVM bug PR14879.

I can't add test cases for any backends other than AVR because none of
them have small enough register classes to trigger the bug.

Reviewers: qcolombet

Subscribers: MatzeB

Differential Revision: https://reviews.llvm.org/D25070

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@282852 91177308-0d34-0410-b5e6-96231b3b80d8
2016-09-30 13:59:20 +00:00
Quentin Colombet
005bfb8238 [RegAllocGreedy] Fix the list of NewVRegs for last chance recoloring.
When trying to recolor a register we may split live-ranges in the
process. When we create new live-ranges we will have to process them,
but when we move a register from Assign to Split, the allocation is not
changed until the whole recoloring session is successful.
Therefore, only push the live-ranges that changed from Assign to
Split when the recoloring is successful.

Same as the previous commit, I was not able to produce a test case that
reproduce the problem with in-tree targets.

Note: The bug has been here since the recoloring scheme has been added
back in r200883 (Feb 2014).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@281783 91177308-0d34-0410-b5e6-96231b3b80d8
2016-09-16 22:00:50 +00:00
Quentin Colombet
12bac3e6a2 [RegAllocGreedy] Fix an assertion and condition when last chance recoloring is used.
When last chance recoloring is used, the list of NewVRegs may not be
empty when calling selectOrSplitImpl. Indeed, another coloring may have
taken place with splitting/spilling in the same recoloring session.

Relax an assertion to take this into account and adapt a condition to
act as if the NewVRegs were local to this selectOrSplitImpl instance.

Unfortunately I am unable to produce a test case for this, I was only
able to reproduce the conditions on an out-of-tree target.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@281782 91177308-0d34-0410-b5e6-96231b3b80d8
2016-09-16 22:00:42 +00:00
Matthias Braun
db9ce2fda6 MachineFunction: Introduce NoPHIs property
I want to compute the SSA property of .mir files automatically in
upcoming patches. The problem with this is that some inputs will be
reported as static single assignment with some passes claiming not to
support SSA form.  In reality though those passes do not support PHI
instructions => Track the presence of PHI instructions separate from the
SSA property.

Differential Revision: https://reviews.llvm.org/D22719

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@279573 91177308-0d34-0410-b5e6-96231b3b80d8
2016-08-23 21:19:49 +00:00
Wei Mi
4eae2788ea Allow dead insts to be kept in DeadRemat only when they are rematerializable.
Because isReallyTriviallyReMaterializableGeneric puts many limits on
rematerializable instructions, this fix can prevent instructions with
tied virtual operands and instructions with virtual register uses from
being kept in DeadRemat, so as to workaround the live interval consistency
problem for the dummy instructions kept in DeadRemat.

But we still need to fix the live interval consistency problem. This patch
is just a short time relieve. PR28464 has been filed as a reminder.

Differential Revision: http://reviews.llvm.org/D19486


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@274928 91177308-0d34-0410-b5e6-96231b3b80d8
2016-07-08 21:08:09 +00:00
Wei Mi
815b02e9ce Recommit r265547, and r265610,r265639,r265657 on top of it, plus
two fixes with one about error verify-regalloc reported, and
another about live range update of phi after rematerialization.

r265547:
Replace analyzeSiblingValues with new algorithm to fix its compile
time issue. The patch is to solve PR17409 and its duplicates.

analyzeSiblingValues is a N x N complexity algorithm where N is
the number of siblings generated by reg splitting. Although it
causes siginificant compile time issue when N is large, it is also
important for performance since it removes redundent spills and
enables rematerialization.

To solve the compile time issue, the patch removes analyzeSiblingValues
and replaces it with lower cost alternatives containing two parts. The
first part creates a new spill hoisting method in postOptimization of
register allocation. It does spill hoisting at once after all the spills
are generated instead of inside every instance of selectOrSplit. The
second part queries the define expr of the original register for
rematerializaiton and keep it always available during register allocation
even if it is already dead. It deletes those dead instructions only in
postOptimization. With the two parts in the patch, it can remove
analyzeSiblingValues without sacrificing performance.

Patches on top of r265547:
r265610 "Fix the compare-clang diff error introduced by r265547."
r265639 "Fix the sanitizer bootstrap error in r265547."
r265657 "InlineSpiller.cpp: Escap \@ in r265547. [-Wdocumentation]"

Differential Revision: http://reviews.llvm.org/D15302
Differential Revision: http://reviews.llvm.org/D18934
Differential Revision: http://reviews.llvm.org/D18935
Differential Revision: http://reviews.llvm.org/D18936


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@266162 91177308-0d34-0410-b5e6-96231b3b80d8
2016-04-13 03:08:27 +00:00
Hans Wennborg
d8b70fb72f Revert r265547 "Recommit r265309 after fixed an invalid memory reference bug happened"
It caused PR27275: "ARM: Bad machine code: Using an undefined physical register"

Also reverting the following commits that were landed on top:
r265610 "Fix the compare-clang diff error introduced by r265547."
r265639 "Fix the sanitizer bootstrap error in r265547."
r265657 "InlineSpiller.cpp: Escap \@ in r265547. [-Wdocumentation]"

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@265790 91177308-0d34-0410-b5e6-96231b3b80d8
2016-04-08 15:17:43 +00:00
Wei Mi
1145b58255 Recommit r265309 after fixed an invalid memory reference bug happened
when DenseMap growed and moved memory. I verified it fixed the bootstrap
problem on x86_64-linux-gnu but I cannot verify whether it fixes
the bootstrap error on clang-ppc64be-linux. I will watch the build-bot
result closely.

Replace analyzeSiblingValues with new algorithm to fix its compile
time issue. The patch is to solve PR17409 and its duplicates.

analyzeSiblingValues is a N x N complexity algorithm where N is
the number of siblings generated by reg splitting. Although it
causes siginificant compile time issue when N is large, it is also
important for performance since it removes redundent spills and
enables rematerialization.

To solve the compile time issue, the patch removes analyzeSiblingValues
and replaces it with lower cost alternatives containing two parts. The
first part creates a new spill hoisting method in postOptimization of
register allocation. It does spill hoisting at once after all the spills
are generated instead of inside every instance of selectOrSplit. The
second part queries the define expr of the original register for
rematerializaiton and keep it always available during register allocation
even if it is already dead. It deletes those dead instructions only in
postOptimization. With the two parts in the patch, it can remove
analyzeSiblingValues without sacrificing performance.

Differential Revision: http://reviews.llvm.org/D15302


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@265547 91177308-0d34-0410-b5e6-96231b3b80d8
2016-04-06 15:41:07 +00:00
Wei Mi
e19a68b607 Revert r265309 and r265312 because they caused some errors I need to investigate.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@265317 91177308-0d34-0410-b5e6-96231b3b80d8
2016-04-04 17:45:03 +00:00
Wei Mi
d31cb9b5c5 Replace analyzeSiblingValues with new algorithm to fix its compile
time issue. The patch is to solve PR17409 and its duplicates.

analyzeSiblingValues is a N x N complexity algorithm where N is
the number of siblings generated by reg splitting. Although it
causes siginificant compile time issue when N is large, it is also
important for performance since it removes redundent spills and
enables rematerialization.

To solve the compile time issue, the patch removes analyzeSiblingValues
and replaces it with lower cost alternatives containing two parts. The
first part creates a new spill hoisting method in postOptimization of
register allocation. It does spill hoisting at once after all the spills
are generated instead of inside every instance of selectOrSplit. The
second part queries the define expr of the original register for
rematerializaiton and keep it always available during register allocation
even if it is already dead. It deletes those dead instructions only in
postOptimization. With the two parts in the patch, it can remove
analyzeSiblingValues without sacrificing performance.

Differential Revision: http://reviews.llvm.org/D15302



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@265309 91177308-0d34-0410-b5e6-96231b3b80d8
2016-04-04 16:42:40 +00:00
Derek Schuff
b65f550d9a Add MachineVerifier check for AllVRegsAllocated MachineFunctionProperty
Summary:
Check that any function that has the property set is free of virtual
register operands.

Also, it is actually VirtRegMap (and not the register allocators) that
acutally remove the VReg operands (except for RegAllocFast).

Reviewers: qcolombet

Subscribers: MatzeB, llvm-commits, qcolombet

Differential Revision: http://reviews.llvm.org/D18535

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@264755 91177308-0d34-0410-b5e6-96231b3b80d8
2016-03-29 17:40:22 +00:00
Derek Schuff
fadd113c9b Introduce MachineFunctionProperties and the AllVRegsAllocated property
MachineFunctionProperties represents a set of properties that a MachineFunction
can have at particular points in time. Existing examples of this idea are
MachineRegisterInfo::isSSA() and MachineRegisterInfo::tracksLiveness() which
will eventually be switched to use this mechanism.
This change introduces the AllVRegsAllocated property; i.e. the property that
all virtual registers have been allocated and there are no VReg operands
left.

With this mechanism, passes can declare that they require a particular property
to be set, or that they set or clear properties by implementing e.g.
MachineFunctionPass::getRequiredProperties(). The MachineFunctionPass base class
verifies that the requirements are met, and handles the setting and clearing
based on the delcarations. Passes can also directly query and update the current
properties of the MF if they want to have conditional behavior.

This change annotates the target-independent post-regalloc passes; future
changes will also annotate target-specific ones.

Reviewers: qcolombet, hfinkel

Differential Revision: http://reviews.llvm.org/D18421

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@264593 91177308-0d34-0410-b5e6-96231b3b80d8
2016-03-28 17:05:30 +00:00
Richard Trieu
1b96cbe208 Remove uses of builtin comma operator.
Cleanup for upcoming Clang warning -Wcomma.  No functionality change intended.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@261270 91177308-0d34-0410-b5e6-96231b3b80d8
2016-02-18 22:09:30 +00:00
Chandler Carruth
9146833fa3 [PM/AA] Rebuild LLVM's alias analysis infrastructure in a way compatible
with the new pass manager, and no longer relying on analysis groups.

This builds essentially a ground-up new AA infrastructure stack for
LLVM. The core ideas are the same that are used throughout the new pass
manager: type erased polymorphism and direct composition. The design is
as follows:

- FunctionAAResults is a type-erasing alias analysis results aggregation
  interface to walk a single query across a range of results from
  different alias analyses. Currently this is function-specific as we
  always assume that aliasing queries are *within* a function.

- AAResultBase is a CRTP utility providing stub implementations of
  various parts of the alias analysis result concept, notably in several
  cases in terms of other more general parts of the interface. This can
  be used to implement only a narrow part of the interface rather than
  the entire interface. This isn't really ideal, this logic should be
  hoisted into FunctionAAResults as currently it will cause
  a significant amount of redundant work, but it faithfully models the
  behavior of the prior infrastructure.

- All the alias analysis passes are ported to be wrapper passes for the
  legacy PM and new-style analysis passes for the new PM with a shared
  result object. In some cases (most notably CFL), this is an extremely
  naive approach that we should revisit when we can specialize for the
  new pass manager.

- BasicAA has been restructured to reflect that it is much more
  fundamentally a function analysis because it uses dominator trees and
  loop info that need to be constructed for each function.

All of the references to getting alias analysis results have been
updated to use the new aggregation interface. All the preservation and
other pass management code has been updated accordingly.

The way the FunctionAAResultsWrapperPass works is to detect the
available alias analyses when run, and add them to the results object.
This means that we should be able to continue to respect when various
passes are added to the pipeline, for example adding CFL or adding TBAA
passes should just cause their results to be available and to get folded
into this. The exception to this rule is BasicAA which really needs to
be a function pass due to using dominator trees and loop info. As
a consequence, the FunctionAAResultsWrapperPass directly depends on
BasicAA and always includes it in the aggregation.

This has significant implications for preserving analyses. Generally,
most passes shouldn't bother preserving FunctionAAResultsWrapperPass
because rebuilding the results just updates the set of known AA passes.
The exception to this rule are LoopPass instances which need to preserve
all the function analyses that the loop pass manager will end up
needing. This means preserving both BasicAAWrapperPass and the
aggregating FunctionAAResultsWrapperPass.

Now, when preserving an alias analysis, you do so by directly preserving
that analysis. This is only necessary for non-immutable-pass-provided
alias analyses though, and there are only three of interest: BasicAA,
GlobalsAA (formerly GlobalsModRef), and SCEVAA. Usually BasicAA is
preserved when needed because it (like DominatorTree and LoopInfo) is
marked as a CFG-only pass. I've expanded GlobalsAA into the preserved
set everywhere we previously were preserving all of AliasAnalysis, and
I've added SCEVAA in the intersection of that with where we preserve
SCEV itself.

One significant challenge to all of this is that the CGSCC passes were
actually using the alias analysis implementations by taking advantage of
a pretty amazing set of loop holes in the old pass manager's analysis
management code which allowed analysis groups to slide through in many
cases. Moving away from analysis groups makes this problem much more
obvious. To fix it, I've leveraged the flexibility the design of the new
PM components provides to just directly construct the relevant alias
analyses for the relevant functions in the IPO passes that need them.
This is a bit hacky, but should go away with the new pass manager, and
is already in many ways cleaner than the prior state.

Another significant challenge is that various facilities of the old
alias analysis infrastructure just don't fit any more. The most
significant of these is the alias analysis 'counter' pass. That pass
relied on the ability to snoop on AA queries at different points in the
analysis group chain. Instead, I'm planning to build printing
functionality directly into the aggregation layer. I've not included
that in this patch merely to keep it smaller.

Note that all of this needs a nearly complete rewrite of the AA
documentation. I'm planning to do that, but I'd like to make sure the
new design settles, and to flesh out a bit more of what it looks like in
the new pass manager first.

Differential Revision: http://reviews.llvm.org/D12080

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@247167 91177308-0d34-0410-b5e6-96231b3b80d8
2015-09-09 17:55:00 +00:00
Robert Lougher
0d87d635d7 Trace copies when checking for rematerializability in spill weight calculation
PR24139 contains an analysis of poor register allocation. One of the findings
was that when calculating the spill weight, a rematerializable interval once
split is no longer rematerializable. This is because the isRematerializable
check in CalcSpillWeights.cpp does not follow the copies introduced by live
range splitting (after splitting, the live interval register definition is a
copy which is not rematerializable).

Reviewers: qcolombet

Differential Revision: http://reviews.llvm.org/D11686


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@244439 91177308-0d34-0410-b5e6-96231b3b80d8
2015-08-10 11:59:44 +00:00
Quentin Colombet
3df507cbfa [RAGreedy] Add an experimental deferred spilling feature.
The idea of deferred spilling is to delay the insertion of spill code until the
very end of the allocation. A "candidate" to spill variable might not required
to be spilled because of other evictions that happened after this decision was
taken. The spirit is similar to the optimistic coloring strategy implemented in
Preston and Briggs graph coloring algorithm.

For now, this feature is highly experimental. Although correct, it would require
much more modification to properly model the effect of spilling.

Anyway, this early patch helps prototyping this feature.

Note: The test case cannot unfortunately be reduced and is probably fragile.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@242585 91177308-0d34-0410-b5e6-96231b3b80d8
2015-07-17 23:04:06 +00:00
Matthias Braun
2aa5727755 TargetRegisterInfo: Provide a way to check assigned registers in getRegAllocationHints()
Pass a const reference to LiveRegMatrix to getRegAllocationHints()
because some targets can prodive better hints if they can test whether a
physreg has been used for register allocation yet.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@242340 91177308-0d34-0410-b5e6-96231b3b80d8
2015-07-15 22:16:00 +00:00
Matthias Braun
65a4053af8 RAGreedy: Keep track of allocated PhysRegs internally
Do not use MachineRegisterInfo::setPhysRegUsed()/isPhysRegUsed()
anymore. This bitset changes function-global state and is set by the
VirtRegRewriter anyway.
Simply use a bitvector private to RAGreedy.

Differential Revision: http://reviews.llvm.org/D10910

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@242169 91177308-0d34-0410-b5e6-96231b3b80d8
2015-07-14 17:38:17 +00:00
Matthias Braun
8e4eaabdb8 RegAllocGreedy: Allow target to specify register class ordering.
Specify an allocation order with a register class. This is used by register
allocators with a greedy heuristic. This is usefull as it is sometimes
beneficial to color more constrained classes first.

Differential Revision: http://reviews.llvm.org/D8626

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@233743 91177308-0d34-0410-b5e6-96231b3b80d8
2015-03-31 19:57:53 +00:00
Matthias Braun
3f1ec42ec7 RegAllocGreedy: Improve live interval order in ReverseLocal mode
When allocating live intervals in linear order and all of them are local
to a single basic block you get an optimal coloring. This is also true
if you reverse the order, but it is not true if you sort live ranges
beginnings in reverse order, change to sort live range endings in
reverse order. Take the following live ranges for example:

   |---| |--------|
|----------| |-------|

They get colored suboptimally with 3 registers if you sort the live range
starting points in reverse order (but optimally with live range begins in order,
or live range ends in reverse order).

Apparently the previous strategy was intentional because of allocation
time considerations. I am having a hard time replicating these effects,
while I see substantial improvements in allocation quality with this
change.

No testcase as none of the (in tree) targets use reverse order mode.

Differential Revision: http://reviews.llvm.org/D8625

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@233742 91177308-0d34-0410-b5e6-96231b3b80d8
2015-03-31 19:57:49 +00:00
Eric Christopher
4ec858ec4b Have TargetRegisterInfo::getLargestLegalSuperClass take a
MachineFunction argument so that it can look up the subtarget
rather than using a cached one in some Targets.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@231888 91177308-0d34-0410-b5e6-96231b3b80d8
2015-03-10 23:46:01 +00:00
Quentin Colombet
9d60e0ff0a [RegAllocGreedy] Introduce a late pass to repair broken hints.
A broken hint is a copy where both ends are assigned different colors. When a
variable gets evicted in the neighborhood of such copies, it is likely we can
reconcile some of them.


** Context **

Copies are inserted during the register allocation via splitting. These split
points are required to relax the constraints on the allocation problem. When
such a point is inserted, both ends of the copy would not share the same color
with respect to the current allocation problem. When variables get evicted,
the allocation problem becomes different and some split point may not be
required anymore. However, the related variables may already have been colored.

This usually shows up in the assembly with pattern like this:
def A
...
save A to B
def A
use A
restore A from B
...
use B

Whereas we could simply have done:
def B
...
def A
use A
...
use B


** Proposed Solution **

A variable having a broken hint is marked for late recoloring if and only if
selecting a register for it evict another variable. Indeed, if no eviction
happens this is pointless to look for recoloring opportunities as it means the
situation was the same as the initial allocation problem where we had to break
the hint.

Finally, when everything has been allocated, we look for recoloring
opportunities for all the identified candidates.
The recoloring is performed very late to rely on accurate copy cost (all
involved variables are allocated).
The recoloring is simple unlike the last change recoloring. It propagates the
color of the broken hint to all its copy-related variables. If the color is
available for them, the recoloring uses it, otherwise it gives up on that hint
even if a more complex coloring would have worked.

The recoloring happens only if it is profitable. The profitability is evaluated
using the expected frequency of the copies of the currently recolored variable
with a) its current color and b) with the target color. If a) is greater or
equal than b), then it is profitable and the recoloring happen.


** Example **

Consider the following example:
BB1:
  a =
  b =
BB2:
  ...
   = b
   = a
Let us assume b gets split:
BB1:
  a =
  b =
BB2:
  c = b
  ...
  d = c
  = d
  = a
Because of how the allocation work, b, c, and d may be assigned different
colors. Now, if a gets evicted to make room for c, assuming b and d were
assigned to something different than a.
We end up with:
BB1:
  a =
  st a, SpillSlot
  b =
BB2:
  c = b
  ...
  d = c
  = d
  e = ld SpillSlot
  = e
This is likely that we can assign the same register for b, c, and d,
getting rid of 2 copies.


** Performances **

Both ARM64 and x86_64 show performance improvements of up to 3% for the
llvm-testsuite + externals with Os and O3. There are a few regressions too that
comes from the (in)accuracy of the block frequency estimate.

<rdar://problem/18312047>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@225422 91177308-0d34-0410-b5e6-96231b3b80d8
2015-01-08 01:16:39 +00:00
Craig Topper
a5babc8a31 Move register class name strings to a single array in MCRegisterInfo to reduce static table size and number of relocation entries.
Indices into the table are stored in each MCRegisterClass instead of a pointer. A new method, getRegClassName, is added to MCRegisterInfo and TargetRegisterInfo to lookup the string in the table.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@222118 91177308-0d34-0410-b5e6-96231b3b80d8
2014-11-17 05:50:14 +00:00
Arnaud A. de Grandmaison
8025a39d11 [PBQP] Tweak spill costs and coalescing benefits
This patch improves how the different costs (register, interference, spill
and coalescing) relates together. The assumption is now that:
 - coalescing (or any other "side effect" of reg alloc) is negative, and
   instead of being derived from a spill cost, they use the block
   frequency info.
 - spill costs are in the [MinSpillCost:+inf( range
 - register or interference costs are in [0.0:MinSpillCost( or +inf

The current MinSpillCost is set to 10.0, which is a random value high
enough that the current constraint builders do not need to worry about
when settings costs. It would however be worth adding a normalization
step for register and interference costs as the last step in the
constraint builder chain to ensure they are not greater than SpillMinCost
(unless this has some sense for some architectures). This would work well
with the current builder pipeline, where all costs are tweaked relatively
to each others, but could grow above MinSpillCost if the pipeline is
deep enough.

The current heuristic is tuned to depend rather on the number of uses of
a live interval rather than a density of uses, as used by the greedy
allocator. This heuristic provides a few percent improvement on a number
of benchmarks (eembc, spec, ...) and will definitely need to change once
spill placement is implemented: the current spill placement is really
ineficient, so making the cost proportionnal to the number of use is a
clear win.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@221292 91177308-0d34-0410-b5e6-96231b3b80d8
2014-11-04 20:51:24 +00:00
Eric Christopher
1dd55ba94e Grab the subtarget and subtarget dependent variables off of
MachineFunction rather than TargetMachine.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@219671 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-14 07:22:00 +00:00
Renato Golin
b157cb7afd Revert 202433 - Provide a target override for the latest regalloc heuristic
That commit was introduced in order to help investigate a problem in ARM
codegen breaking from commit 202304 (Add a limit to the heuristic that register
allocates instructions in local order). Recent analisys indicated that the
problem no longer exists, so I'm reverting this change.

See PR18996.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@218981 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-03 12:20:53 +00:00
Craig Topper
3512034554 Simplify creation of a bunch of ArrayRefs by using None, makeArrayRef or just letting them be implicitly created.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@216525 91177308-0d34-0410-b5e6-96231b3b80d8
2014-08-27 05:25:25 +00:00
Eric Christopher
9f85dccfc6 Remove the TargetMachine forwards for TargetSubtargetInfo based
information and update all callers. No functional change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@214781 91177308-0d34-0410-b5e6-96231b3b80d8
2014-08-04 21:25:23 +00:00
David Blaikie
ec31a302b7 Remove uses of the redundant ".reset(nullptr)" of unique_ptr, in favor of ".reset()"
It's also possible to just write "= nullptr", but there's some question
of whether that's as readable, so I leave it up to authors to pick which
they prefer for now. If we want to discuss standardizing on one or the
other, we can do that at some point in the future.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@213438 91177308-0d34-0410-b5e6-96231b3b80d8
2014-07-19 01:05:11 +00:00
Quentin Colombet
5599fde88e [RegAllocGreedy] Provide a subtarget hook to disable the local reassignment
heuristic.
By default, no functionality change.
This is a follow-up of r212099.

This hook provides a finer grain to control the optimization.

<rdar://problem/17444599>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212204 91177308-0d34-0410-b5e6-96231b3b80d8
2014-07-02 18:32:04 +00:00
Quentin Colombet
230bb1b426 [RegAllocGreedy] Provide a flag to disable the local reassignment heuristic.
By default, no functionality change.

Before evicting a local variable, this heuristic tries to find another (set of)
local(s) that can be reassigned to a free color.

In some extreme cases (large basic blocks with tons of local variables), the
compilation time is dominated by the local interference checks that this
heuristic must perform, with no code gen gain.
E.g., the motivating example takes 4 minutes to compile with this heuristic, 12
seconds without.

Improving the situation will likely require to make drastic changes to the
register allocator and/or the interference check framework.

For now, provide this flag to better understand the impact of that heuristic.

<rdar://problem/17444599>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212099 91177308-0d34-0410-b5e6-96231b3b80d8
2014-07-01 14:08:37 +00:00
Chandler Carruth
8677f2ff9a [Modules] Remove potential ODR violations by sinking the DEBUG_TYPE
define below all header includes in the lib/CodeGen/... tree. While the
current modules implementation doesn't check for this kind of ODR
violation yet, it is likely to grow support for it in the future. It
also removes one layer of macro pollution across all the included
headers.

Other sub-trees will follow.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206837 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-22 02:02:50 +00:00
Craig Topper
4ba844388c [C++11] More 'nullptr' conversion. In some cases just using a boolean check instead of comparing to nullptr.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206142 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-14 00:51:57 +00:00
Quentin Colombet
010311b104 [RegAllocGreedy][Last Chance Recoloring] Change the name of the exhaustive search option.
fexhaustive-register-search => exhaustive-register-search
'f' is a Clang thing!

This is related to PR18747.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206075 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-11 21:51:09 +00:00
Quentin Colombet
92a892e8f9 [RegAllocGreedy][Last Chance Recoloring] Addition of
-fexhaustive-register-search option to allow an exhaustive search during last
chance recoloring.

This is related to PR18747

Patch by MAYUR PANDEY <mayur.p@samsung.com>. 


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@206072 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-11 21:39:44 +00:00
Duncan P. N. Exon Smith
861e4db905 RegAlloc: Account for a variable entry block frequency
Until r197284, the entry frequency was constant -- i.e., set to 2^14.
Although current ToT still has a constant entry frequency, since r197284
that has been an implementation detail (which is soon going to change).

  - r204690 made the wrong assumption for the CSRCost metric.  Adjust
    callee-saved register cost based on entry frequency.

  - r185393 made the wrong assumption (although it was valid at the
    time).  Update SpillPlacement.cpp::Threshold to be relative to the
    entry frequency.

Since ToT still has 2^14 entry frequency, this should have no observable
functionality change.

<rdar://problem/14292693>

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@205789 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-08 19:18:56 +00:00
Quentin Colombet
3b2b5dfa9b [RegAllocGreedy][Last Chance Recoloring] Emit diagnostics when last chance
recoloring cut-offs are encountered and register allocation failed.

This is related to PR18747

Patch by MAYUR PANDEY <mayur.p@samsung.com>.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@205601 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-04 02:05:21 +00:00
Quentin Colombet
cc99615837 Revert r205599, the commit was not intended to have so many changes
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@205600 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-04 02:02:49 +00:00
Quentin Colombet
c65a77b92d [RegAllocGreedy][Last Chance Recoloring] Emit diagnostics when last chance
recoloring cut-offs are hit.

This is related to PR18747.

Patch by MAYUR PANDEY <mayur.p@samsung.com>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@205599 91177308-0d34-0410-b5e6-96231b3b80d8
2014-04-04 01:58:57 +00:00
Manman Ren
d9524d66cd Provide a target override for the cost of using a callee-saved register
for the first time.

Thanks Andy for the discussion.
rdar://16162005


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@204979 91177308-0d34-0410-b5e6-96231b3b80d8
2014-03-27 23:10:04 +00:00
Manman Ren
b5e6ddca56 Register Allocator: refactoring and add comments.
No functionality change. Thanks Andy for reviewing.

rdar://16162005


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@204962 91177308-0d34-0410-b5e6-96231b3b80d8
2014-03-27 21:21:57 +00:00
Manman Ren
3d7d0bc71a Add comments. Addressing review comments from Evan on r204690.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@204864 91177308-0d34-0410-b5e6-96231b3b80d8
2014-03-26 22:14:09 +00:00
Manman Ren
692d1830f3 Register Allocator: check other options before using a CSR for the first time.
When register allocator's stage is RS_Spill, we choose spill over using the CSR
for the first time, if the spill cost is lower than CSRCost. 
When register allocator's stage is < RS_Split, we choose pre-splitting over
using the CSR for the first time, if the cost of splitting is lower than
CSRCost.

CSRCost is set with command-line option "regalloc-csr-first-time-cost". The
default value is 0 to generate the same codes as before this commit.

With a value of 15 (1 << 14 is the entry frequency), I measured performance
gain of 3% on 253.perlbmk and 1.7% on 197.parser, with instrumented PGO,
on an arm device.

rdar://16162005


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@204690 91177308-0d34-0410-b5e6-96231b3b80d8
2014-03-25 00:16:25 +00:00
Manman Ren
66124f9e89 Register Allocator: refactoring (no functionality change).
Factor out two functions calculateRegionSplitCost and doRegionSplit
from tryRegionSplit. These two functions will be used in coming patches.

rdar://16162005


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@204684 91177308-0d34-0410-b5e6-96231b3b80d8
2014-03-24 23:23:42 +00:00
Craig Topper
9f998de891 [C++11] Add 'override' keyword to virtual methods that override their base class.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@203220 91177308-0d34-0410-b5e6-96231b3b80d8
2014-03-07 09:26:03 +00:00
Ahmed Charles
f4ccd11075 Replace OwningPtr<T> with std::unique_ptr<T>.
This compiles with no changes to clang/lld/lldb with MSVC and includes
overloads to various functions which are used by those projects and llvm
which have OwningPtr's as parameters. This should allow out of tree
projects some time to move. There are also no changes to libs/Target,
which should help out of tree targets have time to move, if necessary.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@203083 91177308-0d34-0410-b5e6-96231b3b80d8
2014-03-06 05:51:42 +00:00
Benjamin Kramer
9efaf2f2da [C++11] Use std::tie to simplify compare operators.
No functionality change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@202751 91177308-0d34-0410-b5e6-96231b3b80d8
2014-03-03 19:58:30 +00:00
Alp Toker
18510b7e43 [C++11] Expand and eliminate the LLVM_ENUM_INT_TYPE() macro
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@202607 91177308-0d34-0410-b5e6-96231b3b80d8
2014-03-02 03:20:38 +00:00
Andrew Trick
4c34f71b81 Provide a target override for the latest regalloc heuristic.
This is a temporary workaround for native arm linux builds:
PR18996: Changing regalloc order breaks "lencod" on native arm linux builds.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@202433 91177308-0d34-0410-b5e6-96231b3b80d8
2014-02-27 21:37:33 +00:00
Andrew Trick
eaf8a32859 Add a limit to the heuristic that register allocates instructions in local order.
This handles pathological cases in which we see 2x increase in spill
code for large blocks (~50k instructions). I don't have a unit test
for this behavior.

Fixes rdar://16072279.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@202304 91177308-0d34-0410-b5e6-96231b3b80d8
2014-02-26 22:07:26 +00:00
Manman Ren
d225511ea7 Remove outdated comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@202186 91177308-0d34-0410-b5e6-96231b3b80d8
2014-02-25 19:47:15 +00:00
Alp Toker
bf930d5c1f Fix typos
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@202107 91177308-0d34-0410-b5e6-96231b3b80d8
2014-02-25 04:21:15 +00:00
Quentin Colombet
126099ea23 [RegAlloc] Fix the assertion in the last chance recoloring to match the
condition at the call site.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@201296 91177308-0d34-0410-b5e6-96231b3b80d8
2014-02-13 05:17:37 +00:00
Quentin Colombet
1a10a51431 [RegAlloc] Add a last chance recoloring mechanism when everything else failed to
find a register.

The idea is to choose a color for the variable that cannot be allocated and
recolor its interferences around. Unlike the current register allocation scheme,
it is allowed to change the color of an already assigned (but maybe not
splittable or spillable) live interval while propagating this change to its
neighbors.
In other word, there are two things that may help finding an available color:
- Already assigned variables (RS_Done) can be recolored to different color.
- The recoloring allows to catch solutions that needs to touch more that just
  the neighbors of the current allocated variable.

E.g.,
vA can use {R1, R2    }
vB can use {    R2, R3}
vC can use {R1        }
Where vA, vB, and vC cannot be split anymore (they are reloads for instance) and
they all interfere.

vA is assigned R1
vB is assigned R2
vC tries to evict vA but vA is already done.
=> Regular register allocation heuristic fails.

Last chance recoloring kicks in:
vC does as if vA was evicted => vC uses R1.
vC is marked as fixed.
vA needs to find a color.
None are available.
vA cannot evict vC: vC is a fixed virtual register now.
vA does as if vB was evicted => vA uses R2.
vB needs to find a color.
R3 is available.
Recoloring => vC = R1, vA = R2, vB = R3.

<rdar://problem/15947839>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@200883 91177308-0d34-0410-b5e6-96231b3b80d8
2014-02-05 22:13:59 +00:00
NAKAMURA Takumi
1a3b25343f RegAllocGreedy.cpp: Use more simple value as Hysteresis, to suppress -mfpmath-dependent behavior.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@200738 91177308-0d34-0410-b5e6-96231b3b80d8
2014-02-04 06:29:38 +00:00
Quentin Colombet
fb57392a8d [RegAlloc] Make tryInstructionSplit less aggressive.
The greedy register allocator tries to split a live-range around each
instruction where it is used or defined to relax the constraints on the entire
live-range (this is a last chance split before falling back to spill).
The goal is to have a big live-range that is unconstrained (i.e., that can use
the largest legal register class) and several small local live-range that carry
the constraints implied by each instruction.
E.g.,
Let csti be the constraints on operation i.

V1=
op1 V1(cst1)
op2 V1(cst2)

V1 live-range is constrained on the intersection of cst1 and cst2.

tryInstructionSplit relaxes those constraints by aggressively splitting each
def/use point:
V1=
V2 = V1
V3 = V2
op1 V3(cst1)
V4 = V2
op2 V4(cst2)

Because of how the coalescer infrastructure works, each new variable (V3, V4)
that is alive at the same time as V1 (or its copy, here V2) interfere with V1.
Thus, we end up with an uncoalescable copy for each split point.

To make tryInstructionSplit less aggressive, we check if the split point
actually relaxes the constraints on the whole live-range. If it does not, we do
not insert it.
Indeed, it will not help the global allocation problem:
- V1 will have the same constraints.
- V1 will have the same interference + possibly the newly added split variable
  VS.
- VS will produce an uncoalesceable copy if alive at the same time as V1.

<rdar://problem/15570057>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@198369 91177308-0d34-0410-b5e6-96231b3b80d8
2014-01-02 22:47:22 +00:00
Michael Gottesman
523823b897 [block-freq] Rename getEntryFrequency() -> getEntryFreq() to match getBlockFreq() in all *BlockFrequencyInfo*.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@197304 91177308-0d34-0410-b5e6-96231b3b80d8
2013-12-14 02:37:38 +00:00
Michael Gottesman
c1b4dfd345 [block-freq] Update MachineBlockPlacement and RegAllocGreedy to use the new MachineBlockFrequencyInfo methods.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@197290 91177308-0d34-0410-b5e6-96231b3b80d8
2013-12-14 00:25:45 +00:00
Andrew Trick
2e3f799f01 Add TargetRegisterInfo::reverseLocalAssignment hook.
This hook reverses the order of assignment for local live ranges. This
will generally allocate shorter local live ranges first. For targets with
many registers, this could reduce regalloc compile time by a large
factor. It should still achieve optimal coloring; however, it can change
register eviction decisions. It is disabled by default for two reasons:
(1) Top-down allocation is simpler and easier to debug for targets that
don't benefit from reversing the order.
(2) Bottom-up allocation could result in poor evicition decisions on some
targets affecting the performance of compiled code.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@197001 91177308-0d34-0410-b5e6-96231b3b80d8
2013-12-11 03:40:15 +00:00
Aditya Nandakumar
226e3eab9b Check hint registers for interference only once before evictions
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@196536 91177308-0d34-0410-b5e6-96231b3b80d8
2013-12-05 21:18:40 +00:00
Andrew Trick
4dc7377b20 Reverse the order of eviction checks for possible compile time savings. No functionality.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@195969 91177308-0d34-0410-b5e6-96231b3b80d8
2013-11-29 23:49:38 +00:00
Andrew Trick
09f1b252af DEBUG shouldEvict decisions
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@195490 91177308-0d34-0410-b5e6-96231b3b80d8
2013-11-22 19:07:42 +00:00
Andrew Trick
833a9e9395 Minor cleanup. EvictionCost ctor was confusing relative to the other costs floating around in the code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@195489 91177308-0d34-0410-b5e6-96231b3b80d8
2013-11-22 19:07:38 +00:00
Aditya Nandakumar
3da011f561 Fixed an extra for(typo) in the comments
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@195171 91177308-0d34-0410-b5e6-96231b3b80d8
2013-11-19 23:51:32 +00:00
Aaron Ballman
eb36024720 Replacing HUGE_VALF with llvm::huge_valf in order to work around a warning triggered in MSVC 12.
Patch reviewed by Reid Kleckner and Jim Grosbach.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@194533 91177308-0d34-0410-b5e6-96231b3b80d8
2013-11-13 00:15:44 +00:00
Arnaud A. de Grandmaison
095f994ba6 CalcSpillWeights: give a better describing name to calculateSpillWeights
Besides, this relates it more obviously to the VirtRegAuxInfo::calculateSpillWeightAndHint.

No functionnal change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@194404 91177308-0d34-0410-b5e6-96231b3b80d8
2013-11-11 19:04:45 +00:00
Arnaud A. de Grandmaison
a77da0579b CalculateSpillWeights does not need to be a pass
Based on discussions with Lang Hames and Jakob Stoklund Olesen at the hacker's lab, and in the light of upcoming work on the PBQP register allocator, it was though that CalcSpillWeights does not need to be a pass. This change will enable to customize / tune the spill weight computation depending on the allocator.

Update the documentation style while there.

No functionnal change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@194356 91177308-0d34-0410-b5e6-96231b3b80d8
2013-11-10 17:46:31 +00:00
Arnaud A. de Grandmaison
d241fa7a61 Revert "CalculateSpillWeights does not need to be a pass"
Temporarily revert my previous commit until I understand why it breaks 3 target tests.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@194272 91177308-0d34-0410-b5e6-96231b3b80d8
2013-11-08 18:19:19 +00:00
Arnaud A. de Grandmaison
663fcde3d3 CalculateSpillWeights does not need to be a pass
Based on discussions with Lang Hames and Jakob Stoklund Olesen at the hacker's lab, and in the light of upcoming work on the PBQP register allocator, it was though that CalcSpillWeights does not need to be a pass. This change will enable to customize / tune the spill weight computation depending on the allocator.

Update the documentation style while there.

No functionnal change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@194269 91177308-0d34-0410-b5e6-96231b3b80d8
2013-11-08 17:56:29 +00:00
Matthias Braun
4f3b5e8c92 Represent RegUnit liveness with LiveRange instance
Previously LiveInterval has been used, but having a spill weight and
register number is unnecessary for a register unit.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@192397 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-10 21:29:02 +00:00
Reid Kleckner
73f615b0bd Explicitly request unsigned enum types when desired
This fixes repeated -Wmicrosoft warnings when self-hosting clang on
Windows, and gets us real unsigned enum types with MSVC.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@192227 91177308-0d34-0410-b5e6-96231b3b80d8
2013-10-08 20:15:11 +00:00
Eli Friedman
ae43dac300 Fix unused variables.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@190448 91177308-0d34-0410-b5e6-96231b3b80d8
2013-09-10 23:18:14 +00:00
Mark Lacey
1feb5854ae Track new virtual registers by register number.
Track new virtual registers by register number, rather than by the live
interval created for them. This is the first step in separating the
creation of new virtual registers and new live intervals.  Eventually
live intervals will be created and populated on demand after the virtual
registers have been created and used in instructions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188434 91177308-0d34-0410-b5e6-96231b3b80d8
2013-08-14 23:50:04 +00:00
Andrew Trick
c0173e6f9f Down-scale slot index distance to save bits.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187438 91177308-0d34-0410-b5e6-96231b3b80d8
2013-07-30 19:59:19 +00:00
Andrew Trick
bef4c3e069 RegAllocGreedy comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187141 91177308-0d34-0410-b5e6-96231b3b80d8
2013-07-25 18:35:22 +00:00
Andrew Trick
8adae96fd9 Evict local live ranges if they can be reassigned.
The previous change to local live range allocation also suppressed
eviction of local ranges. In rare cases, this could result in more
expensive register choices. This commit actually revives a feature
that I added long ago: check if live ranges can be reassigned before
eviction. But now it only happens in rare cases of evicting a local
live range because another local live range wants a cheaper register.

The benefit is improved code size for some benchmarks on x86 and armv7.

I measured no significant compile time increase and performance
changes are noise.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187140 91177308-0d34-0410-b5e6-96231b3b80d8
2013-07-25 18:35:19 +00:00
Andrew Trick
6ea2b9608a Allocate local registers in order for optimal coloring.
Also avoid locals evicting locals just because they want a cheaper register.

Problem: MI Sched knows exactly how many registers we have and assumes
they can be colored. In cases where we have large blocks, usually from
unrolled loops, greedy coloring fails. This is a source of
"regressions" from the MI Scheduler on x86. I noticed this issue on
x86 where we have long chains of two-address defs in the same live
range. It's easy to see this in matrix multiplication benchmarks like
IRSmk and even the unit test misched-matmul.ll.

A fundamental difference between the LLVM register allocator and
conventional graph coloring is that in our model a live range can't
discover its neighbors, it can only verify its neighbors. That's why
we initially went for greedy coloring and added eviction to deal with
the hard cases. However, for singly defined and two-address live
ranges, we can optimally color without visiting neighbors simply by
processing the live ranges in instruction order.

Other beneficial side effects:

It is much easier to understand and debug regalloc for large blocks
when the live ranges are allocated in order. Yes, global allocation is
still very confusing, but it's nice to be able to comprehend what
happened locally.

Heuristics could be added to bias register assignment based on
instruction locality (think late register pairing, banks...).

Intuituvely this will make some test cases that are on the threshold
of register pressure more stable.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187139 91177308-0d34-0410-b5e6-96231b3b80d8
2013-07-25 18:35:14 +00:00
Andrew Trick
5dca613978 Dump LIS before regalloc. MI sched changes them.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187107 91177308-0d34-0410-b5e6-96231b3b80d8
2013-07-25 07:26:26 +00:00
Jakob Stoklund Olesen
03ef600623 Remove floats from live range splitting costs.
These floats all represented block frequencies anyway, so just use the
BlockFrequency class directly.

Some floating point computations remain in tryLocalSplit(). They are
estimating spill weights which are still floats.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@186435 91177308-0d34-0410-b5e6-96231b3b80d8
2013-07-16 18:26:18 +00:00
Benjamin Kramer
4eed756153 Switch spill weights from a basic loop depth estimation to BlockFrequencyInfo.
The main advantages here are way better heuristics, taking into account not
just loop depth but also __builtin_expect and other static heuristics and will
eventually learn how to use profile info. Most of the work in this patch is
pushing the MachineBlockFrequencyInfo analysis into the right places.

This is good for a 5% speedup on zlib's deflate (x86_64), there were some very
unfortunate spilling decisions in its hottest loop in longest_match(). Other
benchmarks I tried were mostly neutral.

This changes register allocation in subtle ways, update the tests for it.
2012-02-20-MachineCPBug.ll was deleted as it's very fragile and the instruction
it looked for was gone already (but the FileCheck pattern picked up unrelated
stuff).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184105 91177308-0d34-0410-b5e6-96231b3b80d8
2013-06-17 19:00:36 +00:00
David Blaikie
453f4f0130 Use only explicit bool conversion operators
BitVector/SmallBitVector::reference::operator bool remain implicit since
they model more exactly a bool, rather than something else that can be
boolean tested.

The most common (non-buggy) case are where such objects are used as
return expressions in bool-returning functions or as boolean function
arguments. In those cases I've used (& added if necessary) a named
function to provide the equivalent (or sometimes negative, depending on
convenient wording) test.

One behavior change (YAMLParser) was made, though no test case is
included as I'm not sure how to reach that code path. Essentially any
comparison of llvm::yaml::document_iterators would be invalid if neither
iterator was at the end.

This helped uncover a couple of bugs in Clang - test cases provided for
those in a separate commit along with similar changes to `operator bool`
instances in Clang.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@181868 91177308-0d34-0410-b5e6-96231b3b80d8
2013-05-15 07:36:59 +00:00
Andy Gibbs
200241e4de Replace uses of the deprecated std::auto_ptr with OwningPtr.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@179373 91177308-0d34-0410-b5e6-96231b3b80d8
2013-04-12 10:56:28 +00:00
Bill Wendling
69d37fa635 Remove unused #includes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@176467 91177308-0d34-0410-b5e6-96231b3b80d8
2013-03-05 01:00:45 +00:00
Jakob Stoklund Olesen
6d6132986d Limit the search space in RAGreedy::tryEvict().
When tryEvict() is looking for a cheaper register in the allocation
order, skip the tail of too expensive registers when possible.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172281 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-12 00:57:44 +00:00
Jakob Stoklund Olesen
f7999fe1cb Speed up the AllocationOrder class a bit.
Allow the central functions to be inlined, and use the argumentless
isHint() function when possible.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@169319 91177308-0d34-0410-b5e6-96231b3b80d8
2012-12-04 22:25:16 +00:00
Jakob Stoklund Olesen
fc6374439e Add VirtRegMap::hasKnownPreference().
Virtual registers with a known preferred register are prioritized by
RAGreedy. This function makes the condition explicit without depending
on getRegAllocPref().

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@169179 91177308-0d34-0410-b5e6-96231b3b80d8
2012-12-03 23:23:50 +00:00
Chandler Carruth
d04a8d4b33 Use the new script to sort the includes of every file under lib.
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. =]

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@169131 91177308-0d34-0410-b5e6-96231b3b80d8
2012-12-03 16:50:05 +00:00
Jakob Stoklund Olesen
1ead68d769 Make the LiveRegMatrix analysis available to targets.
No functional change, just moved header files.

Targets can inject custom passes between register allocation and
rewriting. This makes it possible to tweak the register allocation
before rewriting, using the full global interference checking available
from LiveRegMatrix.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@168806 91177308-0d34-0410-b5e6-96231b3b80d8
2012-11-28 19:13:06 +00:00
Evan Cheng
bb36a43872 Fix a significant recent(?) regression. StackSlotColoring no longer did anything
because LiveStackAnalysis was not preserved by VirtRegWriter. This caused
big stack usage regression in some cases.

rdar://12340383


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164408 91177308-0d34-0410-b5e6-96231b3b80d8
2012-09-21 20:04:28 +00:00
Dmitri Gribenko
67c8978617 Fix a couple of Doxygen comment issues pointed out by -Wdocumentation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@163721 91177308-0d34-0410-b5e6-96231b3b80d8
2012-09-12 16:59:47 +00:00
David Blaikie
986d76d7b3 Tidy up a few more uses of MF.getFunction()->getName().
Based on CR feedback from r162301 and Craig Topper's refactoring in r162347
here are a few other places that could use the same API (& in one instance drop
a Function.h dependency).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@162367 91177308-0d34-0410-b5e6-96231b3b80d8
2012-08-22 17:18:53 +00:00
Craig Topper
96601ca332 Add a getName function to MachineFunction. Use it in places that previously did getFunction()->getName(). Remove includes of Function.h that are no longer needed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@162347 91177308-0d34-0410-b5e6-96231b3b80d8
2012-08-22 06:07:19 +00:00