Summary: There are places in MachineBlockPlacement where a worklist is filled in pretty much identical way. The code is duplicated. This refactor it so that the same code is used in both scenarii.
Reviewers: chandlerc, majnemer, rafael, MatzeB, escha, silvas
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D18077
llvm-svn: 263495
If we have a loop with a rarely taken path, we will prune that from the blocks which get added as part of the loop chain. The problem is that we weren't then recognizing the loop chain as schedulable when considering the preheader when forming the function chain. We'd then fall to various non-predecessors before finally scheduling the loop chain (as if the CFG was unnatural.) The net result was that there could be lots of garbage between a loop preheader and the loop, even though we could have directly fallen into the loop. It also meant we separated hot code with regions of colder code.
The particular reason for the rejection of the loop chain was that we were scanning predecessor of the header, seeing the backedge, believing that was a globally more important predecessor (true), but forgetting to account for the fact the backedge precessor was already part of the existing loop chain (oops!.
Differential Revision: http://reviews.llvm.org/D17830
llvm-svn: 262547
Summary: This option is being added for testing purposes.
Reviewers: mcrosier
Subscribers: mcrosier, llvm-commits
Differential Revision: http://reviews.llvm.org/D16410
llvm-svn: 258409
(This is the second attempt to submit this patch. The first caused two assertion
failures and was reverted. See https://llvm.org/bugs/show_bug.cgi?id=25687)
The patch in http://reviews.llvm.org/D13745 is broken into four parts:
1. New interfaces without functional changes (http://reviews.llvm.org/D13908).
2. Use new interfaces in SelectionDAG, while in other passes treat probabilities
as weights (http://reviews.llvm.org/D14361).
3. Use new interfaces in all other passes.
4. Remove old interfaces.
This patch is 3+4 above. In this patch, MBB won't provide weight-based
interfaces any more, which are totally replaced by probability-based ones.
The interface addSuccessor() is redesigned so that the default probability is
unknown. We allow unknown probabilities but don't allow using it together
with known probabilities in successor list. That is to say, we either have a
list of successors with all known probabilities, or all unknown
probabilities. In the latter case, we assume each successor has 1/N
probability where N is the number of successors. An assertion checks if the
user is attempting to add a successor with the disallowed mixed use as stated
above. This can help us catch many misuses.
All uses of weight-based interfaces are now updated to use probability-based
ones.
Differential revision: http://reviews.llvm.org/D14973
llvm-svn: 254377
and the follow-up r254356: "Fix a bug in MachineBlockPlacement that may cause assertion failure during BranchProbability construction."
Asserts were firing in Chromium builds. See PR25687.
llvm-svn: 254366
The patch in http://reviews.llvm.org/D13745 is broken into four parts:
1. New interfaces without functional changes (http://reviews.llvm.org/D13908).
2. Use new interfaces in SelectionDAG, while in other passes treat probabilities
as weights (http://reviews.llvm.org/D14361).
3. Use new interfaces in all other passes.
4. Remove old interfaces.
This patch is 3+4 above. In this patch, MBB won't provide weight-based
interfaces any more, which are totally replaced by probability-based ones.
The interface addSuccessor() is redesigned so that the default probability is
unknown. We allow unknown probabilities but don't allow using it together
with known probabilities in successor list. That is to say, we either have a
list of successors with all known probabilities, or all unknown
probabilities. In the latter case, we assume each successor has 1/N
probability where N is the number of successors. An assertion checks if the
user is attempting to add a successor with the disallowed mixed use as stated
above. This can help us catch many misuses.
All uses of weight-based interfaces are now updated to use probability-based
ones.
Differential revision: http://reviews.llvm.org/D14973
llvm-svn: 254348
When looking for the best successor from the outer loop for a block
belonging to an inner loop, the edge probability computation can be
improved so that edges in the inner loop are ignored. For example,
suppose we are building chains for the non-loop part of the following
code, and looking for B1's best successor. Assume the true body is very
hot, then B3 should be the best candidate. However, because of the
existence of the back edge from B1 to B0, the probability from B1 to B3
can be very small, preventing B3 to be its successor. In this patch, when
computing the probability of the edge from B1 to B3, the weight on the
back edge B1->B0 is ignored, so that B1->B3 will have 100% probability.
if (...)
do {
B0;
... // some branches
B1;
} while(...);
else
B2;
B3;
Differential revision: http://reviews.llvm.org/D10825
llvm-svn: 253414
In the current BB placement algorithm, a loop chain always contains all loop blocks. This has a drawback that cold blocks in the loop may be inserted on a hot function path, hence increasing branch cost and also reducing icache locality.
Consider a simple example shown below:
A
|
B⇆C
|
D
When B->C is quite cold, the best BB-layout should be A,B,D,C. But the current implementation produces A,C,B,D.
This patch filters those cold blocks off from the loop chain by comparing the ratio:
LoopBBFreq / LoopFreq
to 20%: if it is less than 20%, we don't include this BB to the loop chain. Here LoopFreq is the frequency of the loop when we reduce the loop into a single node. In general we have more cold blocks when the loop has few iterations. And vice versa.
Differential revision: http://reviews.llvm.org/D11662
llvm-svn: 251833
Currently, in MachineBlockPlacement pass the loop is rotated to let the best exit to be the last BB in the loop chain, to maximize the fall-through from the loop to outside. With profile data, we can determine the cost in terms of missed fall through opportunities when rotating a loop chain and select the best rotation. Basically, there are three kinds of cost to consider for each rotation:
1. The possibly missed fall through edge (if it exists) from BB out of the loop to the loop header.
2. The possibly missed fall through edges (if they exist) from the loop exits to BB out of the loop.
3. The missed fall through edge (if it exists) from the last BB to the first BB in the loop chain.
Therefore, the cost for a given rotation is the sum of costs listed above. We select the best rotation with the smallest cost. This is only for PGO mode when we have more precise edge frequencies.
Differential revision: http://reviews.llvm.org/D10717
llvm-svn: 250754
Remove implicit ilist iterator conversions from MachineBasicBlock.cpp.
I've also added an overload of `splice()` that takes a pointer, since
it's a natural API. This is similar to the overloads I added for
`remove()` and `erase()` in r249867.
llvm-svn: 249883
We can now run 32-bit programs with empty catch bodies. The next step
is to change PEI so that we get funclet prologues and epilogues.
llvm-svn: 246235
1. Create a utility function normalizeEdgeWeights() in MachineBranchProbabilityInfo that normalizes a list of edge weights so that the sum of then can fit in uint32_t.
2. Provide an interface in MachineBasicBlock to normalize its successors' weights.
3. Add a flag in MachineBasicBlock that tracks whether its successors' weights are normalized.
4. Provide an overload of getSumForBlock that accepts a non-const pointer to a MBB so that it can force normalizing this MBB's successors' weights.
5. Update several uses of getSumForBlock() by eliminating the once needed weight scale.
Differential Revision: http://reviews.llvm.org/D11442
llvm-svn: 244154
Create wrapper methods in the Function class for the OptimizeForSize and MinSize
attributes. We want to hide the logic of "or'ing" them together when optimizing
just for size (-Os).
Currently, we are not consistent about this and rely on a front-end to always set
OptimizeForSize (-Os) if MinSize (-Oz) is on. Thus, there are 18 FIXME changes here
that should be added as follow-on patches with regression tests.
This patch is NFC-intended: it just replaces existing direct accesses of the attributes
by the equivalent wrapper call.
Differential Revision: http://reviews.llvm.org/D11734
llvm-svn: 243994
The patch is generated using this command:
tools/clang/tools/extra/clang-tidy/tool/run-clang-tidy.py -fix \
-checks=-*,llvm-namespace-comment -header-filter='llvm/.*|clang/.*' \
llvm/lib/
Thanks to Eugene Kosov for the original patch!
llvm-svn: 240137
With the option -outline-optional-branches, LLVM will place optional
branches out of line (more details on r231230).
With this patch, this is not done for short optional branches. A short
optional branch is a branch containing a single block with an
instruction count below a certain threshold (defaulting to 3). Still
everything is guarded under -outline-optional-branches).
Outlining a short branch can't significantly improve code locality. It
can however decrease performance because of the additional jmp and in
cases where the optional branch is hot. This fixes a compile time
regression I have observed in a benchmark.
Review: http://reviews.llvm.org/D8108
llvm-svn: 232802
already been added and the inconsistency made choosing names and
changing code more annoying. Plus, wow are they better for this code!
llvm-svn: 231347
result reasonable.
This code predated clang-format and so there was a reasonable amount of
crufty formatting that had accumulated. This should ensure that neither
myself nor others end up with formatting-only changes sneaking into
other fixes.
llvm-svn: 231341
just arbitrarily interleaving unrelated control flows once they get
moved "out-of-line" (both outside of natural CFG ordering and with
diamonds that cannot be fully laid out by chaining fallthrough edges).
This easy solution doesn't work in practice, and it isn't just a small
bug. It looks like a very different strategy will be required. I'm
working on that now, and it'll again go behind some flag so that
everyone can experiment and make sure it is working well for them.
llvm-svn: 231332
a flag for now.
First off, thanks to Daniel Jasper for really pointing out the issue
here. It's been here forever (at least, I think it was there when
I first wrote this code) without getting really noticed or fixed.
The key problem is what happens when two reasonably common patterns
happen at the same time: we outline multiple cold regions of code, and
those regions in turn have diamonds or other CFGs for which we can't
just topologically lay them out. Consider some C code that looks like:
if (a1()) { if (b1()) c1(); else d1(); f1(); }
if (a2()) { if (b2()) c2(); else d2(); f2(); }
done();
Now consider the case where a1() and a2() are unlikely to be true. In
that case, we might lay out the first part of the function like:
a1, a2, done;
And then we will be out of successors in which to build the chain. We go
to find the best block to continue the chain with, which is perfectly
reasonable here, and find "b1" let's say. Laying out successors gets us
to:
a1, a2, done; b1, c1;
At this point, we will refuse to lay out the successor to c1 (f1)
because there are still un-placed predecessors of f1 and we want to try
to preserve the CFG structure. So we go get the next best block, d1.
... wait for it ...
Except that the next best block *isn't* d1. It is b2! d1 is waaay down
inside these conditionals. It is much less important than b2. Except
that this is exactly what we didn't want. If we keep going we get the
entire set of the rest of the CFG *interleaved*!!!
a1, a2, done; b1, c1; b2, c2; d1, f1; d2, f2;
So we clearly need a better strategy here. =] My current favorite
strategy is to actually try to place the block whose predecessor is
closest. This very simply ensures that we unwind these kinds of CFGs the
way that is natural and fitting, and should minimize the number of cache
lines instructions are spread across.
It also happens to be *dead simple*. It's like the datastructure was
specifically set up for this use case or something. We only push blocks
onto the work list when the last predecessor for them is placed into the
chain. So the back of the worklist *is* the nearest next block.
Unfortunately, a change like this is going to cause *soooo* many
benchmarks to swing wildly. So for now I'm adding this under a flag so
that we and others can validate that this is fixing the problems
described, that it seems possible to enable, and hopefully that it fixes
more of our problems long term.
llvm-svn: 231238
In a CFG with the edges A->B->C and A->C, B is an optional branch.
LLVM's default behavior is to lay the blocks out naturally, i.e. A, B,
C, in order to improve code locality and fallthroughs. However, if a
function contains many of those optional branches only a few of which
are taken, this leads to a lot of unnecessary icache misses. Moving B
out of line can work around this.
Review: http://reviews.llvm.org/D7719
llvm-svn: 231230
No functional changes intended.
(I plan on doing some modifications to this function and would like to
have as few unrelated changes as possible in the patch)
llvm-svn: 229649
Canonicalize access to function attributes to use the simpler API.
getAttributes().getAttribute(AttributeSet::FunctionIndex, Kind)
=> getFnAttribute(Kind)
getAttributes().hasAttribute(AttributeSet::FunctionIndex, Kind)
=> hasFnAttribute(Kind)
Also, add `Function::getFnStackAlignment()`, and canonicalize:
getAttributes().getStackAlignment(AttributeSet::FunctionIndex)
=> getFnStackAlignment()
llvm-svn: 229208
Some benchmarks have shown that this could lead to a potential
performance benefit, and so adding some flags to try to help measure the
difference.
A possible explanation. In diamond-shaped CFGs (A followed by either
B or C both followed by D), putting B and C both in between A and
D leads to the code being less dense than it could be. Always either
B or C have to be skipped increasing the chance of cache misses etc.
Moving either B or C to after D might be beneficial on average.
In the long run, but we should probably do a better job of analyzing the
basic block and branch probabilities to move the correct one of B or
C to after D. But even if we don't use this in the long run, it is
a good baseline for benchmarking.
Original patch authored by Daniel Jasper with test tweaks and a second
flag added by me.
Differential Revision: http://reviews.llvm.org/D6969
llvm-svn: 226034
The existing code provided for specifying a global loop alignment preference.
However, the preferred loop alignment might depend on the loop itself. For
recent POWER cores, loops between 5 and 8 instructions should have 32-byte
alignment (while the others are better with 16-byte alignment) so that the
entire loop will fit in one i-cache line.
To support this, getPrefLoopAlignment has been made virtual, and can be
provided with an optional MachineLoop* so the target can inspect the loop
before answering the query. The default behavior, as before, is to return the
value set with setPrefLoopAlignment. MachineBlockPlacement now queries the
target for each loop instead of only once per function. There should be no
functional change for other targets.
llvm-svn: 225117
This is to be consistent with StringSet and ultimately with the standard
library's associative container insert function.
This lead to updating SmallSet::insert to return pair<iterator, bool>,
and then to update SmallPtrSet::insert to return pair<iterator, bool>,
and then to update all the existing users of those functions...
llvm-svn: 222334
shorter/easier and have the DAG use that to do the same lookup. This
can be used in the future for TargetMachine based caching lookups from
the MachineFunction easily.
Update the MIPS subtarget switching machinery to update this pointer
at the same time it runs.
llvm-svn: 214838
string_ostream is a safe and efficient string builder that combines opaque
stack storage with a built-in ostream interface.
small_string_ostream<bytes> additionally permits an explicit stack storage size
other than the default 128 bytes to be provided. Beyond that, storage is
transferred to the heap.
This convenient class can be used in most places an
std::string+raw_string_ostream pair or SmallString<>+raw_svector_ostream pair
would previously have been used, in order to guarantee consistent access
without byte truncation.
The patch also converts much of LLVM to use the new facility. These changes
include several probable bug fixes for truncated output, a programming error
that's no longer possible with the new interface.
llvm-svn: 211749
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.
llvm-svn: 206837
The motivation is to mark dump methods as used in debug builds so that they can
be called from lldb, but to not do so in release builds so that they can be
dead-stripped.
There's lots of potential follow-up work suggested in the thread
"Should dump methods be LLVM_ATTRIBUTE_USED only in debug builds?" on cfe-dev,
but everyone seems to agreen on this subset.
Macro name chosen by fair coin toss.
llvm-svn: 198456
We now only allow breaking source order if the exit block frequency is
significantly higher than the other exit block. The actual bias is
currently under a flag so the best cut-off can be found; the flag
defaults to the old behavior. The idea is to get some benchmark coverage
over different values for the flag and pick the best one.
When we require the new frequency to be at least 20% higher than the old
frequency I see a 5% speedup on zlib's deflate when compressing a random
file on x86_64/westmere. Hal reported a small speedup on Fhourstones on
a BG/Q and no regressions in the test suite.
The test case is the full long_match function from zlib's deflate. I was
reluctant to add it for previous tweaks to branch probabilities because
it's large and potentially fragile, but changed my mind since it's an
important use case and more likely to break with all the current work
going into the PGO infrastructure.
Differential Revision: http://llvm-reviews.chandlerc.com/D2202
llvm-svn: 195265
(4.58s vs 3.2s on an oldish Mac Tower).
The corresponding src is excerpted bellow. The lopp accounts for about 90% of execution time.
--------------------
cat -n test-suite/MultiSource/Benchmarks/Olden/em3d/make_graph.c
90
91 for (k=0; k<j; k++)
92 if (other_node == cur_node->to_nodes[k]) break;
The defective layout is sketched bellow, where the two branches need to swap.
------------------------------------------------------------------------
L:
...
if (cond) goto out-of-loop
goto L
While this code sequence is defective, I don't understand why it incurs 1/3 of
execution time. CPU-event-profiling indicates the poor laoyout dose not increase
in br-misprediction; it dosen't increase stall cycle at all, and it dosen't
prevent the CPU detect the loop (i.e. Loop-Stream-Detector seems to be working fine
as well)...
The root cause of the problem is that the layout pass calls AnalyzeBranch()
with basic-block which is not updated to reflect its current layout.
rdar://13966341
llvm-svn: 183174
When debugging performance regressions we often ask ourselves if the regression
that we see is due to poor isel/sched/ra or due to some micro-architetural
problem. When comparing two code sequences one good way to rule out front-end
bottlenecks (and other the issues) is to force code alignment. This pass adds
a flag that forces the alignment of all of the basic blocks in the program.
llvm-svn: 179353
This fixes some of the cycles between libCodeGen and libSelectionDAG. It's still
a complete mess but as long as the edges consist of virtual call it doesn't
cause breakage. BasicTTI did static calls and thus broke some build
configurations.
llvm-svn: 172246
directly.
This is in preparation for removing the use of the 'Attribute' class as a
collection of attributes. That will shift to the AttributeSet class instead.
llvm-svn: 171253
Sooooo many of these had incorrect or strange main module includes.
I have manually inspected all of these, and fixed the main module
include to be the nearest plausible thing I could find. If you own or
care about any of these source files, I encourage you to take some time
and check that these edits were sensible. I can't have broken anything
(I strictly added headers, and reordered them, never removed), but they
may not be the headers you'd really like to identify as containing the
API being implemented.
Many forward declarations and missing includes were added to a header
files to allow them to parse cleanly when included first. The main
module rule does in fact have its merits. =]
llvm-svn: 169131
We use the enums to query whether an Attributes object has that attribute. The
opaque layer is responsible for knowing where that specific attribute is stored.
llvm-svn: 165488
Previously, MBP essentially aligned every branch target it could. This
bloats code quite a bit, especially non-looping code which has no real
reason to prefer aligned branch targets so heavily.
As Andy said in review, it's still a bit odd to do this without a real
cost model, but this at least has much more plausible heuristics.
Fixes PR13265.
llvm-svn: 161409
We branch to the successor with higher edge weight first.
Convert from
je LBB4_8 --> to outer loop
jmp LBB4_14 --> to inner loop
to
jne LBB4_14
jmp LBB4_8
PR12750
rdar: 11393714
llvm-svn: 161018
rotation. When there is a loop backedge which is an unconditional
branch, we will end up with a branch somewhere no matter what. Try
placing this backedge in a fallthrough position above the loop header as
that will definitely remove at least one branch from the loop iteration,
where whole loop rotation may not.
I haven't seen any benchmarks where this is important but loop-blocks.ll
tests for it, and so this will be covered when I flip the default.
llvm-svn: 154812
laid out in a form with a fallthrough into the header and a fallthrough
out of the bottom. In that case, leave the loop alone because any
rotation will introduce unnecessary branches. If either side looks like
it will require an explicit branch, then the rotation won't add any, do
it to ensure the branch occurs outside of the loop (if possible) and
maximize the benefit of the fallthrough in the bottom.
llvm-svn: 154806
This is a complex change that resulted from a great deal of
experimentation with several different benchmarks. The one which proved
the most useful is included as a test case, but I don't know that it
captures all of the relevant changes, as I didn't have specific
regression tests for each, they were more the result of reasoning about
what the old algorithm would possibly do wrong. I'm also failing at the
moment to craft more targeted regression tests for these changes, if
anyone has ideas, it would be welcome.
The first big thing broken with the old algorithm is the idea that we
can take a basic block which has a loop-exiting successor and a looping
successor and use the looping successor as the layout top in order to
get that particular block to be the bottom of the loop after layout.
This happens to work in many cases, but not in all.
The second big thing broken was that we didn't try to select the exit
which fell into the nearest enclosing loop (to which we exit at all). As
a consequence, even if the rotation worked perfectly, it would result in
one of two bad layouts. Either the bottom of the loop would get
fallthrough, skipping across a nearer enclosing loop and thereby making
it discontiguous, or it would be forced to take an explicit jump over
the nearest enclosing loop to earch its successor. The point of the
rotation is to get fallthrough, so we need it to fallthrough to the
nearest loop it can.
The fix to the first issue is to actually layout the loop from the loop
header, and then rotate the loop such that the correct exiting edge can
be a fallthrough edge. This is actually much easier than I anticipated
because we can handle all the hard parts of finding a viable rotation
before we do the layout. We just store that, and then rotate after
layout is finished. No inner loops get split across the post-rotation
backedge because we check for them when selecting the rotation.
That fix exposed a latent problem with our exitting block selection --
we should allow the backedge to point into the middle of some inner-loop
chain as there is no real penalty to it, the whole point is that it
*won't* be a fallthrough edge. This may have blocked the rotation at all
in some cases, I have no idea and no test case as I've never seen it in
practice, it was just noticed by inspection.
Finally, all of these fixes, and studying the loops they produce,
highlighted another problem: in rotating loops like this, we sometimes
fail to align the destination of these backwards jumping edges. Fix this
by actually walking the backwards edges rather than relying on loopinfo.
This fixes regressions on heapsort if block placement is enabled as well
as lots of other cases where the previous logic would introduce an
abundance of unnecessary branches into the execution.
llvm-svn: 154783
the loop header has a non-loop predecessor which has been pre-fused into
its chain due to unanalyzable branches. In this case, rotating the
header into the body of the loop in order to place a loop exit at the
bottom of the loop is a Very Bad Idea as it makes the loop
non-contiguous.
I'm working on a good test case for this, but it's a bit annoynig to
craft. I should get one shortly, but I'm submitting this now so I can
begin the (lengthy) performance analysis process. An initial run of LNT
looks really, really good, but there is too much noise there for me to
trust it much.
llvm-svn: 154395
where a chain outside of the loop block-set ended up in the worklist for
scheduling as part of the contiguous loop. However, asserting the first
block in the chain is in the loop-set isn't a valid check -- we may be
forced to drag a chain into the worklist due to one block in the chain
being part of the loop even though the first block is *not* in the loop.
This occurs when we have been forced to form a chain early due to
un-analyzable branches.
No test case here as I have no idea how to even begin reducing one, and
it will be hopelessly fragile. We have to somehow end up with a loop
header of an inner loop which is a successor of a basic block with an
unanalyzable pair of branch instructions. Ow. Self-host triggers it so
it is unlikely it will regress.
This at least gets block placement back to passing selfhost and the test
suite. There are still a lot of slowdown that I don't like coming out of
block placement, although there are now also a lot of speedups. =[ I'm
seeing swings in both directions up to 10%. I'm going to try to find
time to dig into this and see if we can turn this on for 3.1 as it does
a really good job of cleaning up after some loops that degraded with the
inliner changes.
llvm-svn: 154287
Moving toward a uniform style of pass definition to allow easier target configuration.
Globally declare Pass ID.
Globally declare pass initializer.
Use INITIALIZE_PASS consistently.
Add a call to the initializer from CodeGen.cpp.
Remove redundant "createPass" functions and "getPassName" methods.
While cleaning up declarations, cleaned up comments (sorry for large diff).
llvm-svn: 150100
fallthrough) in cases where we might fail to rotate an exit to an outer
loop onto the end of the loop chain.
Having *some* rotation, but not performing this rotation, is the primary
fix of thep performance regression with -enable-block-placement for
Olden/em3d (a whopping 30% regression). Still working on reducing the
test case that actually exercises this and the new rotation strategy out
of this code, but I want to check if this regresses other test cases
first as that may indicate it isn't the correct fix.
llvm-svn: 145195
was centered around the premise of laying out a loop in a chain, and
then rotating that chain. This is good for preserving contiguous layout,
but bad for actually making sane rotations. In order to keep it safe,
I had to essentially make it impossible to rotate deeply nested loops.
The information needed to correctly reason about a deeply nested loop is
actually available -- *before* we layout the loop. We know the inner
loops are already fused into chains, etc. We lose information the moment
we actually lay out the loop.
The solution was the other alternative for this algorithm I discussed
with Benjamin and some others: rather than rotating the loop
after-the-fact, try to pick a profitable starting block for the loop's
layout, and then use our existing layout logic. I was worried about the
complexity of this "pick" step, but it turns out such complexity is
needed to handle all the important cases I keep teasing out of benchmarks.
This is, I'm afraid, a bit of a work-in-progress. It is still
misbehaving on some likely important cases I'm investigating in Olden.
It also isn't really tested. I'm going to try to craft some interesting
nested-loop test cases, but it's likely to be extremely time consuming
and I don't want to go there until I'm sure I'm testing the correct
behavior. Sadly I can't come up with a way of getting simple, fine
grained test cases for this logic. We need complex loop structures to
even trigger much of it.
llvm-svn: 145183
heavily on AnalyzeBranch. That routine doesn't behave as we want given
that rotation occurs mid-way through re-ordering the function. Instead
merely check that there are not unanalyzable branching constructs
present, and then reason about the CFG via successor lists. This
actually simplifies my mental model for all of this as well.
The concrete result is that we now will rotate more loop chains. I've
added a test case from Olden highlighting the effect. There is still
a bit more to do here though in order to regain all of the performance
in Olden.
llvm-svn: 145179
pass. This is designed to achieve one of the important optimizations
that the old code placement pass did, but more simply.
This is a somewhat rough and *very* conservative version of the
transform. We could get a lot fancier here if there are profitable cases
to do so. In particular, this only looks for a single pattern, it
insists that the loop backedge being rotated away is the last backedge
in the chain, and it doesn't provide any means of doing better in-loop
placement due to the rotation. However, it appears that it will handle
the important loops I am finding in the LLVM test suite.
llvm-svn: 145158
need lots of fanciness around retaining a reference to a Chain's slot in
the BlockToChain map, but that's all gone now. We can just go directly
to allocating the new chain (which will update the mapping for us) and
using it.
Somewhat gross mechanically generated test case replicates the issue
Duncan spotted when actually testing this out.
llvm-svn: 145120
conflicts, we should only be adding the first block of the chain to the
list, lest we try to merge into the middle of that chain. Most of the
places we were doing this we already happened to be looking at the first
block, but there is no reason to assume that, and in some cases it was
clearly wrong.
I've added a couple of tests here. One already worked, but I like having
an explicit test for it. The other is reduced from a test case Duncan
reduced for me and used to crash. Now it is handled correctly.
llvm-svn: 145119
further. This invariant just wasn't going to work in the face of
unanalyzable branches; we need to be resillient to the phenomenon of
chains poking into a loop and poking out of a loop. In fact, we already
were, we just needed to not assert on it.
This was found during a bootstrap with block placement turned on.
llvm-svn: 145100
reversed in the function's original ordering, and we happened to
encounter it while handling an outer unnatural CFG structure.
Thanks to the test case reduced from GCC's source by Benjamin Kramer.
This may also fix a crasher in gzip that Duncan reduced for me, but
I haven't yet gotten to testing that one.
llvm-svn: 145094
properly account for the *global* probability of the edge being taken.
This manifested as a very large number of unconditional branches to
blocks being merged against the CFG even though they weren't
particularly hot within the CFG.
The fix is to check whether the edge being merged is both locally hot
relative to other successors for the source block, and globally hot
compared to other (unmerged) predecessors of the destination block.
This introduces a new crasher on GCC single-source, but it's currently
behind a flag, and Ben has offered to work on the reduction. =]
llvm-svn: 145010
formation phase and into the initial walk of the basic blocks. We
essentially pre-merge all blocks where unanalyzable fallthrough exists,
as we won't be able to update the terminators effectively after any
reorderings. This is quite a bit more principled as there may be CFGs
where the second half of the unanalyzable pair has some analyzable
predecessor that gets placed first. Then it may get placed next,
implicitly breaking the unanalyzable branch even though we never even
looked at the part that isn't analyzable. I've included a test case that
triggers this (thanks Benjamin yet again!), and I'm hoping to synthesize
some more general ones as I dig into related issues.
Also, to make this new scheme work we have to be able to handle branches
into the middle of a chain, so add this check. We always fallback on the
incoming ordering.
Finally, this starts to really underscore a known limitation of the
current implementation -- we don't consider broken predecessors when
merging successors. This can caused major missed opportunities, and is
something I'm planning on looking at next (modulo more bug reports).
llvm-svn: 144994
block sequence when recovering from unanalyzable control flow
constructs, *always* use the function sequence. I'm not sure why I ever
went down the path of trying to use the loop sequence, it is
fundamentally not the correct sequence to use. We're trying to preserve
the incoming layout in the cases of unreasonable control flow, and that
is only encoded at the function level. We already have a filter to
select *exactly* the sub-set of blocks within the function that we're
trying to form into a chain.
The resulting code layout is also significantly better because of this.
In several places we were ending up with completely unreasonable control
flow constructs due to the ordering chosen by the loop structure for its
internal storage. This change removes a completely wasteful vector of
basic blocks, saving memory allocation in the common case even though it
costs us CPU in the fairly rare case of unnatural loops. Finally, it
fixes the latest crasher reduced out of GCC's single source. Thanks
again to Benjamin Kramer for the reduction, my bugpoint skills failed at
it.
llvm-svn: 144627
cleans up all the chains allocated during the processing of each
function so that for very large inputs we don't just grow memory usage
without bound.
llvm-svn: 144533
tests when I forcibly enabled block placement.
It is apparantly possible for an unanalyzable block to fallthrough to
a non-loop block. I don't actually beleive this is correct, I believe
that 'canFallThrough' is returning true needlessly for the code
construct, and I've left a bit of a FIXME on the verification code to
try to track down why this is coming up.
Anyways, removing the assert doesn't degrade the correctness of the algorithm.
llvm-svn: 144532
this pass. We're leaving already merged blocks on the worklist, and
scanning them again and again only to determine each time through that
indeed they aren't viable. We can instead remove them once we're going
to have to scan the worklist. This is the easy way to implement removing
them. If this remains on the profile (as I somewhat suspect it will), we
can get a lot more clever here, as the worklist's order is essentially
irrelevant. We can use swapping and fold the two loops to reduce
overhead even when there are many blocks on the worklist but only a few
of them are removed.
llvm-svn: 144531
time it is queried to compute the probability of a single successor.
This makes computing the probability of every successor of a block in
sequence... really really slow. ;] This switches to a linear walk of the
successors rather than a quadratic one. One of several quadratic
behaviors slowing this pass down.
I'm not really thrilled with moving the sum code into the public
interface of MBPI, but I don't (at the moment) have ideas for a better
interface. My direction I'm thinking in for a better interface is to
have MBPI actually retain much more state and make *all* of these
queries cheap. That's a lot of work, and would require invasive changes.
Until then, this seems like the least bad (ie, least quadratic)
solution. Suggestions welcome.
llvm-svn: 144530
get loop info structures associated with them, and so we need some way
to make forward progress selecting and placing basic blocks. The
technique used here is pretty brutal -- it just scans the list of blocks
looking for the first unplaced candidate. It keeps placing blocks like
this until the CFG becomes tractable.
The cost is somewhat unfortunate, it requires allocating a vector of all
basic block pointers eagerly. I have some ideas about how to simplify
and optimize this, but I'm trying to get the logic correct first.
Thanks to Benjamin Kramer for the reduced test case out of GCC. Sadly
there are other bugs that GCC is tickling that I'm reducing and working
on now.
llvm-svn: 144516
when we fail to place all the blocks of a loop. Currently this is
happening for unnatural loops, and this logic helps more immediately
point to the problem.
llvm-svn: 144504
branches that also may involve fallthrough. In the case of blocks with
no fallthrough, we can still re-order the blocks profitably. For example
instruction decoding will in some cases continue past an indirect jump,
making laying out its most likely successor there profitable.
Note, no test case. I don't know how to write a test case that exercises
this logic, but it matches the described desired semantics in
discussions with Jakob and others. If anyone has a nice example of IR
that will trigger this, that would be lovely.
Also note, there are still assertion failures in real world code with
this. I'm digging into those next, now that I know this isn't the cause.
llvm-svn: 144499
second algorithm, but only loosely. It is more heavily based on the last
discussion I had with Andy. It continues to walk from the inner-most
loop outward, but there is a key difference. With this algorithm we
ensure that as we visit each loop, the entire loop is merged into
a single chain. At the end, the entire function is treated as a "loop",
and merged into a single chain. This chain forms the desired sequence of
blocks within the function. Switching to a single algorithm removes my
biggest problem with the previous approaches -- they had different
behavior depending on which system triggered the layout. Now there is
exactly one algorithm and one basis for the decision making.
The other key difference is how the chain is formed. This is based
heavily on the idea Andy mentioned of keeping a worklist of blocks that
are viable layout successors based on the CFG. Having this set allows us
to consistently select the best layout successor for each block. It is
expensive though.
The code here remains very rough. There is a lot that needs to be done
to clean up the code, and to make the runtime cost of this pass much
lower. Very much WIP, but this was a giant chunk of code and I'd rather
folks see it sooner than later. Everything remains behind a flag of
course.
I've added a couple of tests to exercise the issues that this iteration
was motivated by: loop structure preservation. I've also fixed one test
that was exhibiting the broken behavior of the previous version.
llvm-svn: 144495
the mailing list. Suggestions for other statistics to collect would be
awesome. =]
Currently these are implemented as a separate pass guarded by a separate
flag. I'm not thrilled by that, but I wanted to be able to collect the
statistics for the old code placement as well as the new in order to
have a point of comparison. I'm planning on folding them into the single
pass if / when there is only one pass of interest.
llvm-svn: 143537
discussions with Andy. Fundamentally, the previous algorithm is both
counter productive on several fronts and prioritizing things which
aren't necessarily the most important: static branch prediction.
The new algorithm uses the existing loop CFG structure information to
walk through the CFG itself to layout blocks. It coalesces adjacent
blocks within the loop where the CFG allows based on the most likely
path taken. Finally, it topologically orders the block chains that have
been formed. This allows it to choose a (mostly) topologically valid
ordering which still priorizes fallthrough within the structural
constraints.
As a final twist in the algorithm, it does violate the CFG when it
discovers a "hot" edge, that is an edge that is more than 4x hotter than
the competing edges in the CFG. These are forcibly merged into
a fallthrough chain.
Future transformations that need te be added are rotation of loop exit
conditions to be fallthrough, and better isolation of cold block chains.
I'm also planning on adding statistics to model how well the algorithm
does at laying out blocks based on the probabilities it receives.
The old tests mostly still pass, and I have some new tests to add, but
the nested loops are still behaving very strangely. This almost seems
like working-as-intended as it rotated the exit branch to be
fallthrough, but I'm not convinced this is actually the best layout. It
is well supported by the probabilities for loops we currently get, but
those are pretty broken for nested loops, so this may change later.
llvm-svn: 142743
it's a bit more plausible to use this instead of CodePlacementOpt. The
code for this was shamelessly stolen from CodePlacementOpt, and then
trimmed down a bit. There doesn't seem to be much utility in returning
true/false from this pass as we may or may not have rewritten all of the
blocks. Also, the statistic of counting how many loops were aligned
doesn't seem terribly important so I removed it. If folks would like it
to be included, I'm happy to add it back.
This was probably the most egregious of the missing features, and now
I'm going to start gathering some performance numbers and looking at
specific loop structures that have different layout between the two.
Test is updated to include both basic loop alignment and nested loop
alignment.
llvm-svn: 142645
block frequency analyses. This differs substantially from the existing
block-placement pass in LLVM:
1) It operates on the Machine-IR in the CodeGen layer. This exposes much
more (and more precise) information and opportunities. Also, the
results are more stable due to fewer transforms ocurring after the
pass runs.
2) It uses the generalized probability and frequency analyses. These can
model static heuristics, code annotation derived heuristics as well
as eventual profile loading. By basing the optimization on the
analysis interface it can work from any (or a combination) of these
inputs.
3) It uses a more aggressive algorithm, both building chains from tho
bottom up to maximize benefit, and using an SCC-based walk to layout
chains of blocks in a profitable ordering without O(N^2) iterations
which the old pass involves.
The pass is currently gated behind a flag, and not enabled by default
because it still needs to grow some important features. Most notably, it
needs to support loop aligning and careful layout of loop structures
much as done by hand currently in CodePlacementOpt. Once it supports
these, and has sufficient testing and quality tuning, it should replace
both of these passes.
Thanks to Nick Lewycky and Richard Smith for help authoring & debugging
this, and to Jakob, Andy, Eric, Jim, and probably a few others I'm
forgetting for reviewing and answering all my questions. Writing
a backend pass is *sooo* much better now than it used to be. =D
llvm-svn: 142641
Haicheng Wu