rather than a bitfield, a great suggestion by Chris during code review.
There is still quite a bit of cruft in the interface, but that requires
sorting out some awkward uses of the cost inside the actual inliner.
No functionality changed intended here.
llvm-svn: 153853
Post-RA scheduling gives a significant performance improvement on
the embedded cores, so turn it on. Using full anti-dep. breaking is
important for FP-intensive blocks, so turn it on (just on the
embedded cores for now; this should also be good on the 970s because
post-ra scheduling is all that we have for now, but that should have
more testing first).
llvm-svn: 153843
This adds a full itinerary for IBM's PPC64 A2 embedded core. These
cores form the basis for the CPUs in the new IBM BG/Q supercomputer.
llvm-svn: 153842
This also avoids emitting the information twice, which led to code bloat. On i386-linux-Release+Asserts
with all targets built this change shaves a whopping 1.3 MB off clang. The number is probably exaggerated
by recent inliner changes but the methods were already enormous with the old inline cost computation.
The DWARF reg -> LLVM reg mapping doesn't seem to have holes in it, so it could be a simple lookup table.
I didn't implement that optimization yet to avoid potentially changing functionality.
There is still some duplication both in tablegen and the generated code that should be cleaned up eventually.
llvm-svn: 153837
As a side note, I really dislike array_pod_sort... Do we really still
care about any STL implementations that get this so wrong? Does libc++?
llvm-svn: 153834
a single missing character. Somehow, this had gone untested. I've added
tests for returns-twice logic specifically with the always-inliner that
would have caught this, and fixed the bug.
Thanks to Matt for the careful review and spotting this!!! =D
llvm-svn: 153832
Loads and stores can have different pipeline behavior, especially on
embedded chips. This change allows those differences to be expressed.
Except for the 440 scheduler, there are no functionality changes.
On the 440, the latency adjustment is only by one cycle, and so this
probably does not affect much. Nevertheless, it will make a larger
difference in the future and this removes a FIXME from the 440 itin.
llvm-svn: 153821
This is the CodeGen equivalent of r153747. I tested that there is not noticeable
performance difference with any combination of -O0/-O2 /-g when compiling
gcc as a single compilation unit.
llvm-svn: 153817
Dynamic linking on PPC64 has had problems since we had to move the top-down
hazard-detection logic post-ra. For dynamic linking to work there needs to be
a nop placed after every call. It turns out that it is really hard to guarantee
that nothing will be placed in between the call (bl) and the nop during post-ra
scheduling. Previous attempts at fixing this by placing logic inside the
hazard detector only partially worked.
This is now fixed in a different way: call+nop codegen-only instructions. As far
as CodeGen is concerned the pair is now a single instruction and cannot be split.
This solution works much better than previous attempts.
The scoreboard hazard detector is also renamed to be more generic, there is currently
no cpu-specific logic in it.
llvm-svn: 153816
the very high overhead of the complex inline cost analysis when all it
wants to do is detect three patterns which must not be inlined. Comment
the code, clean it up, and leave some hints about possible performance
improvements if this ever shows up on a profile.
Moving this off of the (now more expensive) inline cost analysis is
particularly important because we have to run this inliner even at -O0.
llvm-svn: 153814
interfaces. These methods were used in the old inline cost system where
there was a persistent cache that had to be updated, invalidated, and
cleared. We're now doing more direct computations that don't require
this intricate dance. Even if we resume some level of caching, it would
almost certainly have a simpler and more narrow interface than this.
llvm-svn: 153813
on a per-callsite walk of the called function's instructions, in
breadth-first order over the potentially reachable set of basic blocks.
This is a major shift in how inline cost analysis works to improve the
accuracy and rationality of inlining decisions. A brief outline of the
algorithm this moves to:
- Build a simplification mapping based on the callsite arguments to the
function arguments.
- Push the entry block onto a worklist of potentially-live basic blocks.
- Pop the first block off of the *front* of the worklist (for
breadth-first ordering) and walk its instructions using a custom
InstVisitor.
- For each instruction's operands, re-map them based on the
simplification mappings available for the given callsite.
- Compute any simplification possible of the instruction after
re-mapping, and store that back int othe simplification mapping.
- Compute any bonuses, costs, or other impacts of the instruction on the
cost metric.
- When the terminator is reached, replace any conditional value in the
terminator with any simplifications from the mapping we have, and add
any successors which are not proven to be dead from these
simplifications to the worklist.
- Pop the next block off of the front of the worklist, and repeat.
- As soon as the cost of inlining exceeds the threshold for the
callsite, stop analyzing the function in order to bound cost.
The primary goal of this algorithm is to perfectly handle dead code
paths. We do not want any code in trivially dead code paths to impact
inlining decisions. The previous metric was *extremely* flawed here, and
would always subtract the average cost of two successors of
a conditional branch when it was proven to become an unconditional
branch at the callsite. There was no handling of wildly different costs
between the two successors, which would cause inlining when the path
actually taken was too large, and no inlining when the path actually
taken was trivially simple. There was also no handling of the code
*path*, only the immediate successors. These problems vanish completely
now. See the added regression tests for the shiny new features -- we
skip recursive function calls, SROA-killing instructions, and high cost
complex CFG structures when dead at the callsite being analyzed.
Switching to this algorithm required refactoring the inline cost
interface to accept the actual threshold rather than simply returning
a single cost. The resulting interface is pretty bad, and I'm planning
to do lots of interface cleanup after this patch.
Several other refactorings fell out of this, but I've tried to minimize
them for this patch. =/ There is still more cleanup that can be done
here. Please point out anything that you see in review.
I've worked really hard to try to mirror at least the spirit of all of
the previous heuristics in the new model. It's not clear that they are
all correct any more, but I wanted to minimize the change in this single
patch, it's already a bit ridiculous. One heuristic that is *not* yet
mirrored is to allow inlining of functions with a dynamic alloca *if*
the caller has a dynamic alloca. I will add this back, but I think the
most reasonable way requires changes to the inliner itself rather than
just the cost metric, and so I've deferred this for a subsequent patch.
The test case is XFAIL-ed until then.
As mentioned in the review mail, this seems to make Clang run about 1%
to 2% faster in -O0, but makes its binary size grow by just under 4%.
I've looked into the 4% growth, and it can be fixed, but requires
changes to other parts of the inliner.
llvm-svn: 153812
visitor will now visit a CallInst and an InvokeInst with
instruction-specific visitors, then visit a generic CallSite visitor,
then delegate back to the Instruction visitor and the TerminatorInst
visitors depending on whether a call or an invoke originally. This will
be used in the soon-to-land inline cost rewrite.
llvm-svn: 153811
The powi intrinsic requires special handling because it always takes a single
integer power regardless of the result type. As a result, we can vectorize
only if the powers are equal. Fixes PR12364.
llvm-svn: 153797
First small step toward modeling multi-register multi-pressure. In the
future, register units can also be used to model liveness and
aliasing.
llvm-svn: 153794
ARMConstantIslandPass still has bugs where jump table compression can
cause constant pool entries to go out of range.
Add a safety margin of 2 bytes when placing constant islands, but use
the real max displacement for verification.
<rdar://problem/11156595>
llvm-svn: 153789
Use an explicit comparator instead of the default.
The sets are sorted, but not using the default comparator. Hopefully,
this will unbreak the Linux builders.
llvm-svn: 153772
When an immediate is both a value [t2_]so_imm and a [t2_]so_imm_neg,
we want to use the non-negated form to make sure we prefer the normal
encoding, not the aliased encoding via the negation of, e.g., 'cmp.w'.
llvm-svn: 153770
TableGen emits lists of sub-registers, super-registers, and overlaps. Put
them all in a single table and use a SequenceToOffsetTable to share
suffixes.
llvm-svn: 153761
This is similar to the StringToOffsetTable we use to produce string
tables, but it can be used for other sequences than strings, and it
eliminates entries for suffixes.
llvm-svn: 153760
For 'adds r2, r2, #56' outside of an IT block, the 16-bit encoding T2
can be used for this syntax. Prefer the narrow encoding when possible.
rdar://11156277
llvm-svn: 153759
1. The main works will made in the RuntimeDyLdImpl with uses the ObjectFile class. RuntimeDyLdMachO and RuntimeDyLdELF now only parses relocations and resolve it. This is allows to make improvements of the RuntimeDyLd more easily. In addition the support for COFF can be easily added.
2. Added ARM relocations to RuntimeDyLdELF.
3. Added support for stub functions for the ARM, allowing to do a long branch.
4. Added support for external functions that are not loaded from the object files, but can be loaded from external libraries. Now MCJIT can correctly execute the code containing the printf, putc, and etc.
5. The sections emitted instead functions, thanks Jim Grosbach. MemoryManager.startFunctionBody() and MemoryManager.endFunctionBody() have been removed.
6. MCJITMemoryManager.allocateDataSection() and MCJITMemoryManager. allocateCodeSection() used JMM->allocateSpace() instead of JMM->allocateCodeSection() and JMM->allocateDataSection(), because I got an error: "Cannot allocate an allocated block!" with object file contains more than one code or data sections.
llvm-svn: 153754
here but it has no other uses, then we have a problem. E.g.,
int foo (const int *x) {
char a[*x];
return 0;
}
If we assign 'a' a vreg and fast isel later on has to use the selection
DAG isel, it will want to copy the value to the vreg. However, there are
no uses, which goes counter to what selection DAG isel expects.
<rdar://problem/11134152>
llvm-svn: 153705
This pass splits basic blocks to insert constant islands, and it
doesn't recompute the live-in lists. No later passes depend on accurate
liveness information.
This fixes PR12410 where the machine code verifier was complaining.
llvm-svn: 153700
We are sometimes allocatinog from the DPair register class which
contains odd-even pairs in addition to the Q registers.
Place the Q registers first in the DPair allocation order as they can be
copied with a single instruction. The odd-even pairs should only be
allocated as a last resort.
llvm-svn: 153699
ARM recently gained DPair, DTriple, and DQuad register classes.
Update copyPhysReg() to handle copies in these register classes.
No test case, it is difficult to make the register allocator emit the
odd copies reliably. The missing DPair copy caused a failure on
partialsums in the nightly test suite.
<rdar://problem/11147997>
llvm-svn: 153686
CodeGenPrepare sinks compare instructions down to their uses to prevent
live flags and predicate registers across basic blocks.
PRE of a compare instruction prevents that, forcing the i1 compare
result into a general purpose register. That is usually more expensive
than the redundant compare PRE was trying to eliminate in the first
place.
llvm-svn: 153657
Module-level ASM may contain definitions of functions and globals. However, we
were not telling the linker that these globals had definitions. As far as it was
concerned, they were just declarations.
Attempt to resolve this by inserting module-level ASM functions and globals into
the '_symbol' set so that the linker will know that they have values.
This gets us further towards our goal of compiling LLVM, but it still has
problems when linking libLTO.dylib because of the `-dead_strip' flag that's
passed to the linker.
<rdar://problem/11124216>
llvm-svn: 153638
This is a code change to add support for changing instruction sequences of the form:
load
inc/dec of 8/16/32/64 bits
store
into the appropriate X86 inc/dec through memory instruction:
inc[qlwb] / dec[qlwb]
The checks that were in X86DAGToDAGISel::Select(SDNode *Node)>>ISD::STORE have been extracted to isLoadIncOrDecStore and reworked to use the better
named wrappers for getOperand(unsigned) (e.g. getOffset()) and replaced Chain.getNode() with LoadNode. The comments have also been expanded.
llvm-svn: 153635
This is a code change to add support for changing instruction sequences of the form:
load
inc/dec of 8/16/32/64 bits
store
into the appropriate X86 inc/dec through memory instruction:
inc[qlwb] / dec[qlwb]
The checks that were in X86DAGToDAGISel::Select(SDNode *Node)>>ISD::STORE have been extracted to isLoadIncOrDecStore and reworked to use the better
named wrappers for getOperand(unsigned) (e.g. getOffset()) and replaced Chain.getNode() with LoadNode. The comments have also been expanded.
llvm-svn: 153617
Some targets still mess up the liveness information, but that isn't
verified after MRI->invalidateLiveness().
The verifier can still check other useful things like register classes
and CFG, so it should be enabled after all passes.
llvm-svn: 153615
The late scheduler depends on accurate liveness information if it is
breaking anti-dependencies, so we should be able to verify it.
Relax the terminator checking in the machine code verifier so it can
handle the basic blocks created by if conversion.
llvm-svn: 153614
When an strd instruction doesn't get the registers it wants, it can be
expanded into two str instructions. Make sure the first str doesn't kill
the base register in the case where the base and data registers are
identical:
t2STRi12 %R0<kill>, %R0, 4, pred:14, pred:%noreg
t2STRi12 %R2<kill>, %R0, 8, pred:14, pred:%noreg
<rdar://problem/11101911>
llvm-svn: 153611
When a number of sub-register VLRDS instructions are combined into a
VLDM, preserve any super-register implicit defs. This is required to
keep the register scavenger and machine code verifier happy.
Enable machine code verification after ARMLoadStoreOptimizer.
ARM/2012-01-26-CopyPropKills.ll was failing because of this.
llvm-svn: 153610
The arm_neon intrinsics can create virtual registers from the DPair
register class which allows both even-odd and odd-even D-register pairs.
This fixes PR12389.
llvm-svn: 153603
Extract the liveness verification into its own method.
This makes it possible to run the machine code verifier after liveness
information is no longer required to be valid.
llvm-svn: 153596
Revert r153519: "ARMLoadStoreOptimizer invalidates register liveness."
These patches caused miscompilations in povray by turning off branch
folding's updating of live-in lists.
It turns out the the late scheduler depends on the live-in lists, even
if it doesn't need correct kill flags.
<rdar://problem/11139228>
llvm-svn: 153593
Original commit message for r153521 (aka r153423):
Use the new range metadata in computeMaskedBits and add a new optimization to
instruction simplify that lets us remove an and when loding a boolean value.
llvm-svn: 153587
blocks in the function cloner. This removes the last case of trivially
dead code that I've been seeing in the wild getting inlined, analyzed,
re-inlined, optimized, only to be deleted. Nukes a FIXME from the
cleanup tests.
llvm-svn: 153572
them as machine instructions. Directives ".set noat" and ".set at" are now
emitted only at the beginning and end of a function except in the case where
they are emitted to enclose .cpload with an immediate operand that doesn't fit
in 16-bit field or unaligned load/stores.
Also, make the following changes:
- Remove function isUnalignedLoadStore and use a switch-case statement to
determine whether an instruction is an unaligned load or store.
- Define helper function CreateMCInst which generates an instance of an MCInst
from an opcode and a list of operands.
llvm-svn: 153552
executable has been moved to another machine). If that's not available
(read-only or something), then exit gracefully.
<rdar://problem/11111686>
llvm-svn: 153538
undefined behavior, which Rafael was kind enough to fix.
Original commit message for r153423:
Use the new range metadata in computeMaskedBits and add a new optimization to
instruction simplify that lets us remove an and when loding a boolean value.
llvm-svn: 153521
This pass tries to update kill flags, but there are still many bugs.
Passes after the load/store optimizer don't need accurate liveness, so
don't even try.
<rdar://problem/11101911>
llvm-svn: 153519
Branch folding can use a register scavenger to update liveness
information when required. Don't do that if liveness information is
already invalid.
llvm-svn: 153517
Late optimization passes like branch folding and tail duplication can
transform the machine code in a way that makes it expensive to keep the
register liveness information up to date. There is a fuzzy line between
register allocation and late scheduling where the liveness information
degrades.
The MRI::tracksLiveness() flag makes the line clear: While true,
liveness information is accurate, and can be used for register
scavenging. Once the flag is false, liveness information is not
accurate, and can only be used as a hint.
Late passes generally don't need the liveness information, but they will
sometimes use the register scavenger to help update it. The scavenger
enforces strict correctness, and we have to spend a lot of code to
update register liveness that may never be used.
llvm-svn: 153511
size bloat. Unfortunately, I expect this to disable the majority of the
benefit from r152737. I'm hopeful at least that it will fix PR12345. To
explain this requires... quite a bit of backstory I'm afraid.
TL;DR: The change in r152737 actually did The Wrong Thing for
linkonce-odr functions. This change makes it do the right thing. The
benefits we saw were simple luck, not any actual strategy. Benchmark
numbers after a mini-blog-post so that I've written down my thoughts on
why all of this works and doesn't work...
To understand what's going on here, you have to understand how the
"bottom-up" inliner actually works. There are two fundamental modes to
the inliner:
1) Standard fixed-cost bottom-up inlining. This is the mode we usually
think about. It walks from the bottom of the CFG up to the top,
looking at callsites, taking information about the callsite and the
called function and computing th expected cost of inlining into that
callsite. If the cost is under a fixed threshold, it inlines. It's
a touch more complicated than that due to all the bonuses, weights,
etc. Inlining the last callsite to an internal function gets higher
weighth, etc. But essentially, this is the mode of operation.
2) Deferred bottom-up inlining (a term I just made up). This is the
interesting mode for this patch an r152737. Initially, this works
just like mode #1, but once we have the cost of inlining into the
callsite, we don't just compare it with a fixed threshold. First, we
check something else. Let's give some names to the entities at this
point, or we'll end up hopelessly confused. We're considering
inlining a function 'A' into its callsite within a function 'B'. We
want to check whether 'B' has any callers, and whether it might be
inlined into those callers. If so, we also check whether inlining 'A'
into 'B' would block any of the opportunities for inlining 'B' into
its callers. We take the sum of the costs of inlining 'B' into its
callers where that inlining would be blocked by inlining 'A' into
'B', and if that cost is less than the cost of inlining 'A' into 'B',
then we skip inlining 'A' into 'B'.
Now, in order for #2 to make sense, we have to have some confidence that
we will actually have the opportunity to inline 'B' into its callers
when cheaper, *and* that we'll be able to revisit the decision and
inline 'A' into 'B' if that ever becomes the correct tradeoff. This
often isn't true for external functions -- we can see very few of their
callers, and we won't be able to re-consider inlining 'A' into 'B' if
'B' is external when we finally see more callers of 'B'. There are two
cases where we believe this to be true for C/C++ code: functions local
to a translation unit, and functions with an inline definition in every
translation unit which uses them. These are represented as internal
linkage and linkonce-odr (resp.) in LLVM. I enabled this logic for
linkonce-odr in r152737.
Unfortunately, when I did that, I also introduced a subtle bug. There
was an implicit assumption that the last caller of the function within
the TU was the last caller of the function in the program. We want to
bonus the last caller of the function in the program by a huge amount
for inlining because inlining that callsite has very little cost.
Unfortunately, the last caller in the TU of a linkonce-odr function is
*not* the last caller in the program, and so we don't want to apply this
bonus. If we do, we can apply it to one callsite *per-TU*. Because of
the way deferred inlining works, when it sees this bonus applied to one
callsite in the TU for 'B', it decides that inlining 'B' is of the
*utmost* importance just so we can get that final bonus. It then
proceeds to essentially force deferred inlining regardless of the actual
cost tradeoff.
The result? PR12345: code bloat, code bloat, code bloat. Another result
is getting *damn* lucky on a few benchmarks, and the over-inlining
exposing critically important optimizations. I would very much like
a list of benchmarks that regress after this change goes in, with
bitcode before and after. This will help me greatly understand what
opportunities the current cost analysis is missing.
Initial benchmark numbers look very good. WebKit files that exhibited
the worst of PR12345 went from growing to shrinking compared to Clang
with r152737 reverted.
- Bootstrapped Clang is 3% smaller with this change.
- Bootstrapped Clang -O0 over a single-source-file of lib/Lex is 4%
faster with this change.
Please let me know about any other performance impact you see. Thanks to
Nico for reporting and urging me to actually fix, Richard Smith, Duncan
Sands, Manuel Klimek, and Benjamin Kramer for talking through the issues
today.
llvm-svn: 153506
MachinePointerInfo when getStore is called to create a node that stores an
argument passed in register to the stack. Without this change, the post RA
scheduler will fail to discover the dependencies between the stores
instructions and the instructions that load from a structure passed by value.
The link to the related discussion is here:
http://lists.cs.uiuc.edu/pipermail/llvmdev/2012-March/048055.html
llvm-svn: 153499
copies being considered for removal. Make sure to track all of the copies,
rather than just the most recent encountered, by holding a DenseSet instead of
an unsigned in SrcMap.
No test case - couldn't reduce something with a sane size.
llvm-svn: 153487
produces a 32-bit immediate which is consumed by the use. It tries to
fold the immediate by breaking it into two parts and fold them into the
immmediate fields of two uses. e.g
movw r2, #40885
movt r3, #46540
add r0, r0, r3
=>
add.w r0, r0, #3019898880
add.w r0, r0, #30146560
;
However, this transformation is incorrect if the user produces a flag. e.g.
movw r2, #40885
movt r3, #46540
adds r0, r0, r3
=>
add.w r0, r0, #3019898880
adds.w r0, r0, #30146560
Note the adds.w may not set the carry flag even if the original sequence
would.
rdar://11116189
llvm-svn: 153484
relocations. The algorithm is the same as
that for x86_64. Scattered relocations, a
feature present in i386 but not on x86_64,
are not yet supported.
llvm-svn: 153466
Original commit message:
Use the new range metadata in computeMaskedBits and add a new optimization to
instruction simplify that lets us remove an and when loading a boolean value.
llvm-svn: 153452
constant-offsets of a common base using the generic GEP-walking logic
I added for computing pointer differences in the same situation.
llvm-svn: 153419
inbounds GEPs. This isn't really necessary for simplifying pointer
differences, but I'm planning to re-use the same code to simplify
pointer comparisons where it is necessary. Since real code almost
exclusively uses inbounds GEPs, it doesn't seem worth it to support the
extra complexity of turning it on and off. If anyone would like that
back, feel free to shout. Note that instcombine will still catch any of
these patterns.
llvm-svn: 153418
aggressively. There are lots of dire warnings about this being expensive
that seem to predate switching to the TrackingVH-based value remapper
that is automatically updated on RAUW. This makes it easy to not just
prune single-entry PHIs, but to fully simplify PHIs, and to recursively
simplify the newly inlined code to propagate PHINode simplifications.
This introduces a bit of a thorny problem though. We may end up
simplifying a branch condition to a constant when we fold PHINodes, and
we would like to nuke any dead blocks resulting from this so that time
isn't wasted continually analyzing them, but this isn't easy. Deleting
basic blocks *after* they are fully cloned and mapped into the new
function currently requires manually updating the value map. The last
piece of the simplification-during-inlining puzzle will require either
switching to WeakVH mappings or some other piece of refactoring. I've
left a FIXME in the testcase about this.
llvm-svn: 153410
* Removed test/lib/llvm.exp - it is no longer needed
* Deleted the dg.exp reading code from test/lit.cfg. There are no dg.exp files
left in the test suite so this code is no longer required. test/lit.cfg is
now much shorter and clearer
* Removed a lot of duplicate code in lit.local.cfg files that need access to
the root configuration, by adding a "root" attribute to the TestingConfig
object. This attribute is dynamically computed to provide the same
information as was previously provided by the custom getRoot functions.
* Documented the config.root attribute in docs/CommandGuide/lit.pod
llvm-svn: 153408
to instead rely on much more generic and powerful instruction
simplification in the function cloner (and thus inliner).
This teaches the pruning function cloner to use instsimplify rather than
just the constant folder to fold values during cloning. This can
simplify a large number of things that constant folding alone cannot
begin to touch. For example, it will realize that 'or' and 'and'
instructions with certain constant operands actually become constants
regardless of what their other operand is. It also can thread back
through the caller to perform simplifications that are only possible by
looking up a few levels. In particular, GEPs and pointer testing tend to
fold much more heavily with this change.
This should (in some cases) have a positive impact on compile times with
optimizations on because the inliner itself will simply avoid cloning
a great deal of code. It already attempted to prune proven-dead code,
but now it will be use the stronger simplifications to prove more code
dead.
llvm-svn: 153403
fire if anything ever invalidates the assumption of a terminator
instruction being unchanged throughout the routine.
I've convinced myself that the current definition of simplification
precludes such a transformation, so I think getting some asserts
coverage that we don't violate this agreement is sufficient to make this
code safe for the foreseeable future.
Comments to the contrary or other suggestions are of course welcome. =]
The bots are now happy with this code though, so it appears the bug here
has indeed been fixed.
llvm-svn: 153401
list. This is a bad idea. ;] I'm hopeful this is the bug that's showing
up with the MSVC bots, but we'll see.
It is definitely unnecessary. InstSimplify won't do anything to
a terminator instruction, we don't need to even include it in the
iteration range. We can also skip the now dead terminator check,
although I've made it an assert to help document that this is an
important invariant.
I'm still a bit queasy about this because there is an implicit
assumption that the terminator instruction cannot be RAUW'ed by the
simplification code. While that appears to be true at the moment, I see
no guarantee that would ensure it remains true in the future. I'm
looking at the cleanest way to solve that...
llvm-svn: 153399
spotted by inspection, and I've crafted no test case that triggers it on
my machine, but some of the windows builders are hitting what looks like
memory corruption, so *something* is amiss here.
This patch takes a more generalized approach to eliminating
double-visits. Imagine code such as:
%x = ...
%y = add %x, 1
%z = add %x, %y
You can imagine that if we simplify %x, we would add %y and %z to the
list. If the use-chain order happens to cause us to add them in reverse
order, we could pull %y off first, and simplify it, adding %z to the
list. We now have %z on the list twice, and will reference it after it
is deleted.
Currently, all my test cases happen to not trigger this, likely due to
the use-chain ordering, but there seems no guarantee that such
a situation could not occur, so we should handle it correctly.
Again, if anyone knows how to craft a testcase that actually triggers
this, please let me know.
llvm-svn: 153397
worklist. This can happen in theory when an instruction uses itself,
such as a PHI node. This was spotted by inspection, and unfortunately
I've not been able to come up with a test case that would trigger it. If
anyone has ideas, let me know...
llvm-svn: 153396
regressed seriously here, we are no longer removing allocas during
inline cleanup. This appears to be because of lifetime markers "using"
them. =/ I'll look into this shortly.
llvm-svn: 153394
bit simpler by handling a common case explicitly.
Also, refactor the implementation to use a worklist based walk of the
recursive users, rather than trying to use value handles to detect and
recover from RAUWs during the recursive descent. This fixes a very
subtle bug in the previous implementation where degenerate control flow
structures could cause mutually recursive instructions (PHI nodes) to
collapse in just such a way that From became equal to To after some
amount of recursion. At that point, we hit the inf-loop that the assert
at the top attempted to guard against. This problem is defined away when
not using value handles in this manner. There are lots of comments
claiming that the WeakVH will protect against just this sort of error,
but they're not accurate about the actual implementation of WeakVHs,
which do still track RAUWs.
I don't have any test case for the bug this fixes because it requires
running the recursive simplification on unreachable phi nodes. I've no
way to either run this or easily write an input that triggers it. It was
found when using instruction simplification inside the inliner when
running over the nightly test-suite.
llvm-svn: 153393
The PPC64 SVR4 ABI requires integer stack arguments, and thus the var. args., that
are smaller than 64 bits be zero extended to 64 bits.
llvm-svn: 153373
Code such as:
%vreg100 = setcc %vreg10, -1, SETNE
brcond %vreg10, %tgt
was being incorrectly morphed into
%vreg100 = and %vreg10, 1
brcond %vreg10, %tgt
where the 'and' instruction could be eliminated since
such logic is on 1-bit types in the PTX back-end, leaving
us with just:
brcond %vreg10, %tgt
which essentially gives us inverted branch conditions.
llvm-svn: 153364
destination module, but one of them isn't used in the destination module. If
another module comes along and the uses the unused type, there could be type
conflicts when the modules are finally linked together. (This happened when
building LLVM.)
The test that was reduced is:
Module A:
%Z = type { %A }
%A = type { %B.1, [7 x x86_fp80] }
%B.1 = type { %C }
%C = type { i8* }
declare void @func_x(%C*, i64, i64)
declare void @func_z(%Z* nocapture)
Module B:
%B = type { %C.1 }
%C.1 = type { i8* }
%A.2 = type { %B.3, [5 x x86_fp80] }
%B.3 = type { %C.1 }
define void @func_z() {
%x = alloca %A.2, align 16
%y = getelementptr inbounds %A.2* %x, i64 0, i32 0, i32 0
call void @func_x(%C.1* %y, i64 37, i64 927) nounwind
ret void
}
declare void @func_x(%C.1*, i64, i64)
declare void @func_y(%B* nocapture)
(Unfortunately, this test doesn't fail under llvm-link, only during an LTO
linking.) The '%C' and '%C.1' clash. The destination module gets the '%C'
declaration. When merging Module B, it looks at the '%C.1' subtype of the '%B'
structure. It adds that in, because that's cool. And when '%B.3' is processed,
it uses the '%C.1'. But the '%B' has used '%C' and we prefer to use '%C'. So the
'@func_x' type is changed to 'void (%C*, i64, i64)', but the type of '%x' in
'@func_z' remains '%A.2'. The GEP resolves to a '%C.1', which conflicts with the
'@func_x' signature.
We can resolve this situation by making sure that the type is used in the
destination before saying that it should be used in the module being merged in.
With this fix, LLVM and Clang both compile under LTO.
<rdar://problem/10913281>
llvm-svn: 153351
same basic block, and it's not safe to insert code in the successor
blocks if the edges are critical edges. Splitting those edges is
possible, but undesirable, especially on the unwind side. Instead,
make the bottom-up code motion to consider invokes to be part of
their successor blocks, rather than part of their parent blocks, so
that it doesn't push code past them and onto the edges. This fixes
PR12307.
llvm-svn: 153343
This is necessary if the client wants to be able to mutate TargetOptions (for example, fast FP math mode) after the initial creation of the ExecutionEngine.
llvm-svn: 153342
dominated by Root, check that B is available throughout the scope. This
is obviously true (famous last words?) given the current logic, but the
check may be helpful if more complicated reasoning is added one day.
llvm-svn: 153323
(and hopefully on Windows). The bots have been down most of the day
because of this, and it's not clear to me what all will be required to
fix it.
The commits started with r153205, then r153207, r153208, and r153221.
The first commit seems to be the real culprit, but I couldn't revert
a smaller number of patches.
When resubmitting, r153207 and r153208 should be folded into r153205,
they were simple build fixes.
llvm-svn: 153241
execution-time regression for nsieve-bits on the ARMv7 -O0 -g nightly tester.
This may also improve compile-time on architectures that would otherwise
generate a libcall for urem (e.g., ARM) or fall back to the DAG selector.
rdar://10810716
llvm-svn: 153230
This is in braces so that it doesn't conflict with the existing %p.
It uses braces instead of parens because parens would have to be
regex-escaped.
llvm-svn: 153213
1. Declare a virtual function getPointerToNamedFunction() in JITMemoryManager
2. Move the implementation of getPointerToNamedFunction() form JIT/MCJIT to DefaultJITMemoryManager.
llvm-svn: 153205
Type legalization can zero-extend the elements of the build_vector node, so,
for example, we may have an <8 x i8> with i32 elements of value 255. That
should return 'true' for the vector being all ones.
llvm-svn: 153203
not attched to a basic block or function. There are conservatively
correct answers in these cases, and this makes the analysis more useful
in contexts where we have a partially formed bit of IR.
I don't have any way to test this directly... suggestions welcome here,
but I'm not seeing anything sadly. I only found this using a subsequent
patch to the inliner which runs instsimplify on partially inlined
instructions, and even then only on a quite large program. I never got
a reasonable testcase out of it, and anything I do get is likely to be
quite fragile due to requiring an interaction of two different passes,
and the only result being a segfault if it goes wrong.
llvm-svn: 153176
Adds /usr/lib/debug early to list, as some systems (debian) have unstripped libs in there
Adds /lib/i386-linux-gnu for systems that does multiarch (debian)
llvm-svn: 153174
We can simply confirm the handle released to open it with EXCLUSIVE. Attempting renaming was bad.
Disable win32file at ImportError. Thanks to Francois to let me know.
FIXME: Could we report warning or notification if win32file were not found?
llvm-svn: 153172
Remaining "uncategorized" functions have been organized into their
proper place in the hierarchy. Some functions were moved around so
groups are defined together.
No code changes were made.
llvm-svn: 153169
This gives a lot of love to the docs for the C API. Like Clang's
documentation, the C API is now organized into a Doxygen "module"
(LLVMC). Each C header file is a child of the main module. Some modules
(like Core) have a hierarchy of there own. The produced documentation is
thus better organized (before everything was in one monolithic list).
This patch also includes a lot of new documentation for APIs in Core.h.
It doesn't document them all, but is better than none. Function docs are
missing @param and @return annotation, but the documentation body now
commonly provides help details (like the expected llvm::Value sub-type
to expect).
llvm-svn: 153157
These changes allow us to compile big endian from the command line for 32 bit
Mips targets. This patch will result in code and data actually being produced
in the correct endianess.
llvm-svn: 153153
ImmutAVLTree uses random unsigned values as keys into a DenseMap,
which could possibly happen to be the same value as the Tombstone or
Entry keys in the DenseMap.
Test case is hard to come up with. We randomly get failures on the
internal static analyzer bot, which most likely hits this issue
(hard to be 100% sure without the full stack).
llvm-svn: 153148
relocations (i.e., pieces of data whose addresses
are referred to elsewhere in the binary image) and
update the references when the section containing
the relocations moves. The way this works is that
there is a map from section IDs to lists of
relocations.
Because the relocations are associated with the
section containing the data being referred to, they
are updated only when the target moves. However,
many data references are relative and also depend
on the location of the referrer.
To solve this problem, I introduced a new data
structure, Referrer, which simply contains the
section being referred to and the index of the
relocation in that section. These referrers are
associated with the source containing the
reference that needs to be updated, so now
regardless of which end of the relocation moves,
the relocation will now be updated correctly.
llvm-svn: 153147
Do not call SplitBlockPredecessors on a loop preheader when one of the
predecessors is an indirectbr. Otherwise, you will hit this assert:
!isa<IndirectBrInst>(Preds[i]->getTerminator()) && "Cannot split an edge from an IndirectBrInst"
llvm-svn: 153134
instead of skipping the current loop.
My prior fix was incomplete because of an overzealous compile-time optimization:
Better fix for: <rdar://problem/11049788> Segmentation fault: 11 in LoopStrengthReduce
llvm-svn: 153131
ARMBaseRegisterInfo::canRealignStack was checking for variable-sized objects
but not for stack adjustments around calls. Use hasReservedCallFrame() to
check for both. The hasBasePointer function was already correctly checking
both conditions, so the effect of this was that a base pointer would be used
without checking whether the base pointer register could be reserved. I don't
have a small testcase for this.
<rdar://problem/11075906>
llvm-svn: 153110
This results in things such as
vmovups 16(%rdi), %xmm0
vinsertf128 $1, %xmm0, %ymm0, %ymm0
to be combined to
vinsertf128 $1, 16(%rdi), %ymm0, %ymm0
rdar://11076953
llvm-svn: 153092
i128). In that case, we may not be able to print out the MCExpr as an
expression. For instance, we could have an MCExpr like this:
0xBEEF0000BEEF0000 | (0xBEEF0000BEEF0000 << 64)
The MCExpr printer handles sizes up to 64-bits, but this expression would
require 128-bits. In this situation, try to evaluate the constant expression and
emit that as the value into 64-bit chunks.
<rdar://problem/11070338>
llvm-svn: 153081
a variable. The previous code would break the debug info changing
code invariant. This will regress debug info for arguments where
we elide the alloca created.
Fixes rdar://11066468
llvm-svn: 153074
1) opt is not usually in the same path as the target program. Even for
the bugpoint as a standalone app, it should be more portable to search
in PATH, isn't it?
2) bugpoint driver accounts opt plugins, but does not list them in the
final output command.
Patch by Dmitry Mikushin!
llvm-svn: 153066
X86InstrCompiler.td.
It also adds –mcpu-generic to the legalize-shift-64.ll test so the test
will pass if run on an Intel Atom CPU, which would otherwise
produce an instruction schedule which differs from that which the test expects.
llvm-svn: 153033
violations I introduced. Also sort some of the instructions to get
a more consistent ordering.
Suggestions on still better / more consistent formatting would be
welcome. I'm actually tempted to use a macro to define all of the
delegate methods...
llvm-svn: 153030
fast-isel before emitting code. If the program bails after code was emitted,
then it could lead to the stack being adjusted more than once (two
CALLSEQ_BEGINs emitted) but being adjuste back only once after the call. This
leads to general badness and gnashing of teeth.
<rdar://problem/11050630>
llvm-svn: 152959
It's not a good style idea, as the registers will be laid down in memory in
numerical order, not the order they're in the list, but it's legal. vldm/vstm
are stricter.
rdar://11064740
llvm-svn: 152943
In previous case,
RUN: foo -o %t
RUN: FileCheck < %t
RUN: bar -o %t
2nd read handle might prevent manipulation of 3rd %t in bar, to remove and rename.
llvm-svn: 152916
alignment. If that's the case, then we want to make sure that we don't increase
the alignment of the store instruction. Because if we increase it to be "more
aligned" than the pointer, code-gen may use instructions which require a greater
alignment than the pointer guarantees.
<rdar://problem/11043589>
llvm-svn: 152907
It was added in 2007 as the first cut at supporting no-inline
attributes, but we didn't have function attributes of any form at the
time. However, it was added without any mention in the LangRef or other
documentation.
Later on, in 2008, Devang added function notes for 'inline=never' and
then turned them into proper function attributes. From that point
onward, as far as I can tell, the world moved on, and no one has touched
'llvm.noinline' in any meaningful way since.
It's time has now come. We have had better mechanisms for doing this for
a long time, all the frontends I'm aware of use them, and this is just
holding back progress. Given that it was never a documented feature of
the IR, I've provided no auto-upgrade support. If people know of real,
in-the-wild bitcode that relies on this, yell at me and I'll add it, but
I *seriously* doubt anyone cares.
llvm-svn: 152904
directly query the function information which this set was representing.
This simplifies the interface of the inline cost analysis, and makes the
always-inline pass significantly more efficient.
Previously, always-inline would first make a single set of every
function in the module *except* those marked with the always-inline
attribute. It would then query this set at every call site to see if the
function was a member of the set, and if so, refuse to inline it. This
is quite wasteful. Instead, simply check the function attribute directly
when looking at the callsite.
The normal inliner also had similar redundancy. It added every function
in the module with the noinline attribute to its set to ignore, even
though inside the cost analysis function we *already tested* the
noinline attribute and produced the same result.
The only tricky part of removing this is that we have to be able to
correctly remove only the functions inlined by the always-inline pass
when finalizing, which requires a bit of a hack. Still, much less of
a hack than the set of all non-always-inline functions was. While I was
touching this function, I switched a heavy-weight set to a vector with
sort+unique. The algorithm already had a two-phase insert and removal
pattern, we were just needlessly paying the uniquing cost on every
insert.
This probably speeds up some compiles by a small amount (-O0 compiles
with lots of always-inline, so potentially heavy libc++ users), but I've
not tried to measure it.
I believe there is no functional change here, but yell if you spot one.
None are intended.
Finally, the direction this is going in is to greatly simplify the
inline cost query interface so that we can replace its implementation
with a much more clever one. Along the way, all the APIs get simplified,
so it seems incrementally good.
llvm-svn: 152903
analysis implementation. The header was already separated. Also cleanup
all the comments in the header to follow a nice modern doxygen form.
There is still plenty of cruft here, but some of that will fall out in
subsequent refactorings and this was an easy step in the right
direction. No functionality changed here.
llvm-svn: 152898
These edges are not really necessary, but it is consistent with the
way we currently create physreg edges. Scheduler heuristics that
expect a DAG edge to the block terminator could benefit from this
change. Although in the future I hope we have a better mechanism for
modeling latency across scheduling regions.
llvm-svn: 152895
Only record IVUsers that are dominated by simplified loop
headers. Otherwise SCEVExpander will crash while looking for a
preheader.
I previously tried to work around this in LSR itself, but that was
insufficient. This way, LSR can continue to run if some uses are not
in simple loops, as long as we don't attempt to analyze those users.
Fixes <rdar://problem/11049788> Segmentation fault: 11 in LoopStrengthReduce
llvm-svn: 152892
on our internal nightly testers. So, basically revert r152486 again.
Abbreviated original commit message:
Implement a more intelligent way of spilling uses across an invoke boundary.
It looks as if Chander's inlining work, r152737, exposed an issue.
llvm-svn: 152887
It caused MSP430DAGToDAGISel::SelectIndexedBinOp() to be miscompiled.
When two ReplaceUses()'s are expanded as inline, vtable in base class is stored to latter (ISelUpdater)ISU.
llvm-svn: 152877
theoretical fix since it only matters for types with >= 2^63 bits (!) and also
only matters if pointers have more than 64 bits, which is not supported anyway.
llvm-svn: 152831
We cannot limit the concatenated instruction names to 64K. ARM is
already at 32K, and it is easy to imagine a target with more
instructions.
llvm-svn: 152817
This patch limited the concatenated register names to 64K which meant
that the total number of registers was many times less than 64K.
If any compilers actually enforce the 64K limit on string literals, and
it turns out to be a problem, we should fix that problem by not using
long string literals.
llvm-svn: 152816
This needs a test, but it will take some time to figure
out the best way to get an input that will produce > 2^16 relocs.
Patch by Graydon Hoare!
llvm-svn: 152787
out the DW_AT_name. Older gdbs unfortunately still use it to
disambiguate member functions in templated classes (gdb.cp/templates.exp).
rdar://11043421 (which is now deferred for a bit)
llvm-svn: 152782
This results in things such as
vmovaps -96(%rbx), %xmm1
vinsertf128 $1, %xmm1, %ymm0, %ymm0
to be combined to
vinsertf128 $1, -96(%rbx), %ymm0, %ymm0
rdar://10643481
llvm-svn: 152762
correlated pairs of pointer arguments at the callsite. This is designed
to recognize the common C++ idiom of begin/end pointer pairs when the
end pointer is a constant offset from the begin pointer. With the
C-based idiom of a pointer and size, the inline cost saw the constant
size calculation, and this provides the same level of information for
begin/end pairs.
In order to propagate this information we have to search for candidate
operations on a pair of pointer function arguments (or derived from
them) which would be simplified if the pointers had a known constant
offset. Then the callsite analysis looks for such pointer pairs in the
argument list, and applies the appropriate bonus.
This helps LLVM detect that half of bounds-checked STL algorithms
(such as hash_combine_range, and some hybrid sort implementations)
disappear when inlined with a constant size input. However, it's not
a complete fix due the inaccuracy of our cost metric for constants in
general. I'm looking into that next.
Benchmarks showed no significant code size change, and very minor
performance changes. However, specific code such as hashing is showing
significantly cleaner inlining decisions.
llvm-svn: 152752
Commit r152704 exposed a latent MSVC limitation (aka bug).
Both ilist and and iplist contains the same function:
template<class InIt> void insert(iterator where, InIt first, InIt last) {
for (; first != last; ++first) insert(where, *first);
}
Also ilist inherits from iplist and ilist contains a "using iplist<NodeTy>::insert".
MSVC doesn't know which one to pick and complain with an error.
I think it is safe to delete ilist::insert since it is redundant anyway.
llvm-svn: 152746
which are small enough to themselves be inlined. Delaying in this manner
can be harmful if the function is inelligible for inlining in some (or
many) contexts as it pessimizes the code of the function itself in the
event that inlining does not eventually happen.
Previously the check was written to only do this delaying of inlining
for static functions in the hope that they could be entirely deleted and
in the knowledge that all callers of static functions will have the
opportunity to inline if it is in fact profitable. However, with C++ we
get two other important sources of functions where the definition is
always available for inlining: inline functions and templated functions.
This patch generalizes the inliner to allow linkonce-ODR (the linkage
such C++ routines receive) to also qualify for this delay-based
inlining.
Benchmarking across a range of large real-world applications shows
roughly 2% size increase across the board, but an average speedup of
about 0.5%. Some benhcmarks improved over 2%, and the 'clang' binary
itself (when bootstrapped with this feature) shows a 1% -O0 performance
improvement when run over all Sema, Lex, and Parse source code smashed
into a single file. A clean re-build of Clang+LLVM with a bootstrapped
Clang shows approximately 2% improvement, but that measurement is often
noisy.
llvm-svn: 152737
There were cases where a value could be used and it's both crossing an invoke
and NOT crossing an invoke. This could happen in the landing pads. In that case,
we will demote the value to the stack like we did before.
<rdar://problem/10609139>
llvm-svn: 152705
New flags: -misched-topdown, -misched-bottomup. They can be used with
the default scheduler or with -misched=shuffle. Without either
topdown/bottomup flag -misched=shuffle now alternates scheduling
direction.
LiveIntervals update is unimplemented with bottom-up scheduling, so
only -misched-topdown currently works.
Capped the ScheduleDAG hierarchy with a concrete ScheduleDAGMI class.
ScheduleDAGMI is aware of the top and bottom of the unscheduled zone
within the current region. Scheduling policy can be plugged into
the ScheduleDAGMI driver by implementing MachineSchedStrategy.
ConvergingScheduler is now the default scheduling algorithm.
It exercises the new driver but still does no reordering.
llvm-svn: 152700
output (we're emitting a specification already and the information
isn't changing).
Saves 1% on the debug information for a build of llvm.
Fixes rdar://11043421
llvm-svn: 152697
(i16 load $addr+c*sizeof(i16)) and replace uses of (i32 vextract) with the
i16 load. It should issue an extload instead: (i32 extload $addr+c*sizeof(i16)).
rdar://11035895
llvm-svn: 152675
take a TargetLibraryInfo parameter. Internally, rather than passing TD, TLI
and DT parameters around all over the place, introduce a struct for holding
them.
llvm-svn: 152623
Also refactor the existing OProfile profiling code to reuse the same interfaces with the VTune profiling code.
In addition, unit tests for the profiling interfaces were added.
This patch was prepared by Andrew Kaylor and Daniel Malea, and reviewed in the llvm-commits list by Jim Grosbach
llvm-svn: 152620
offset accumulation to use a boring APInt instead of ConstantExprs.
I didn't go all the way to an 'int64_t' because I wanted APInt to handle
any magic required to properly wrap the arithmetic when the pointer
width is <64 bits. If there is a significant penalty from using APInt
here, first off WTF, and secondly let me know and I'll do the math by
hand.
I've left one layer still operating w/ ConstantExpr because it makes the
interface quite a bit simpler, and that one isn't iterative so has much
lower cost.
I suppose this may potentially speed up some strang compilation
situations, but I don't really expect much. It should have no functional
impact either way.
llvm-svn: 152590
candidate set for subsequent inlining, try to simplify the arguments to
the inner call site now that inlining has been performed.
The goal here is to propagate and fold constants through deeply nested
call chains. Without doing this, we loose the inliner bonus that should
be applied because the arguments don't match the exact pattern the cost
estimator uses.
Reviewed on IRC by Benjamin Kramer.
llvm-svn: 152556
Typically instcombine has handled this, but pointer differences show up
in several contexts where we would like to get constant folding, and
cannot afford to run instcombine. Specifically, I'm working on improving
the constant folding of arguments used in inline cost analysis with
instsimplify.
Doing this in instsimplify implies some algorithm changes. We have to
handle multiple layers of all-constant GEPs because instsimplify cannot
fold them into a single GEP the way instcombine can. Also, we're only
interested in all-constant GEPs. The result is that this doesn't really
replace the instcombine logic, it's just complimentary and focused on
constant folding.
Reviewed on IRC by Benjamin Kramer.
llvm-svn: 152555
Renamed methods caseBegin, caseEnd and caseDefault with case_begin, case_end, and case_default.
Added some notes relative to case iterators.
llvm-svn: 152532
This requires a C++ change to EDDisassembler's ctor to function properly
(the llvm::InitializeAll* functions aren't being called currently and
there is no way to call them from Python).
Code is partially tested and works well enough for initial commit. There
are probably many small bugs.
llvm-svn: 152506
Bill Wendling