Summary:
When I initially introduced -pass-remarks, I thought it would be a
neat idea to make it additive. So, if one used it as:
$ llc -pass-remarks=inliner --pass-remarks=loop.*
the compiler would build the regular expression '(inliner)|(loop.*)'.
The more I think about it, the more I regret it. This is not how
other flags work. The standard semantics are right-to-left overrides.
This is how clang interprets -Rpass. And I think the two should be
compatible in this respect.
Reviewers: qcolombet
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D3614
llvm-svn: 208122
Before this patch, the backend always emitted a store+load sequence to
bitconvert from f64 to i64 the input operand of a ISD::BITCAST dag node that
performed a bitconvert from type MVT::f64 to type MVT::v2i32. The resulting
i64 node was then used to build a v2i32 vector.
With this patch, the backend now produces a cheaper SCALAR_TO_VECTOR from
MVT::f64 to MVT::v2f64. That SCALAR_TO_VECTOR is then followed by a "free"
bitcast to type MVT::v4i32. The elements of the resulting
v4i32 are then extracted to build a v2i32 vector (which is illegal and
therefore promoted to MVT::v2i64).
This is in general cheaper than emitting a stack store+load sequence
to bitconvert the operand from type f64 to type i64.
llvm-svn: 208107
This patch implements the infrastructure to use named register constructs in
programs that need access to specific registers (bare metal, kernels, etc).
So far, only the stack pointer is supported as a technology preview, but as it
is, the intrinsic can already support all non-allocatable registers from any
architecture.
llvm-svn: 208104
An alias has the address of what it points to, so it also has the same
alignment.
This allows a few optimizations to see past aliases for free.
llvm-svn: 208103
fall back to the normal path without a cpu. While doing this fix
llc to just exit when we don't have a module to process instead of
asserting.
llvm-svn: 208102
Also, provide the ability to create temporary and non-temporary
declarations, as not all declarations may be replaced by definitions
later on.
This provides the necessary infrastructure for Clang to fix PR19598,
leaking temporary MDNodes in Clang's debug info generation.
llvm-svn: 208054
The Win64 docs are very clear that anything larger than 8 bytes is
passed by reference, and GCC MinGW64 honors that for __modti3 and
friends.
Patch by Jameson Nash!
llvm-svn: 208029
On x64, windows.h doesn't include intrin.h for intrinsics. It just
declares them in the global namespace and uses them, expecting the
compiler to lower it as a builtin. We basically need to do this in
clang, eventually.
llvm-svn: 208023
The number of tail call to loop conversions remains the same (1618 by my count).
The new algorithm does a local scan over the use-def chains to identify local "alloca-derived" values, as well as points where the alloca could escape. Then, a visit over the CFG marks blocks as being before or after the allocas have escaped, and annotates the calls accordingly.
llvm-svn: 208017
Summary:
Also ran clang-format on the function. The code added is the last else
if block.
Reviewers: nadav, craig.topper
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D3518
llvm-svn: 207992
Visibility is meaningless when the linkage is local. Change
`-internalize` to reset the visibility to `default`.
<rdar://problem/16141113>
llvm-svn: 207979
Tested that the right -target-cpu is set in the clang -cc1 command line
when running "clang -march=native -E -v - </dev/null" on both an FX-8150
and an FX-8350. Both are family 15h; the FX-8150 (Bulldozer processor)
reports a model number of 1, and the FX-8350 (Piledriver processor)
reports a model number of 2.
llvm-svn: 207973
Windows on ARM does not conform to AEABI. However, memset would be emitted
using the AEABI signature, resulting in inverted parameters. Handle this
special case appropriately.
llvm-svn: 207943
Add handling for FK_SecRel_4 (4-byte section relative relocations). These are
used by the generation of DWARF debug information (the abbrevations use section
relative relocations). This will also be used in generation of CodeView line
tables.
llvm-svn: 207941
Otherwise we use the same threshold as for complete unrolling, which is
way too high. This made us unroll any loop smaller than 150 instructions
by 8 times, but only if someone specified -march=core2 or better,
which happens to be the default on darwin.
llvm-svn: 207940
When can't assume a vectorized tree is rooted in an instruction. The IRBuilder
could have constant folded it. When we rebuild the build_vector (the series of
InsertElement instructions) use the last original InsertElement instruction. The
vectorized tree root is guaranteed to be before it.
Also, we can't assume that the n-th InsertElement inserts the n-th element into
a vector.
This reverts r207746 which reverted the revert of the revert of r205018 or so.
Fixes the test case in PR19621.
llvm-svn: 207939
operations on the call graph. This one forms a cycle, and while not as
complex as removing an internal edge from an SCC, it involves
a reasonable amount of work to find all of the nodes newly connected in
a cycle.
Also somewhat alarming is the worst case complexity here: it might have
to walk roughly the entire SCC inverse DAG to insert a single edge. This
is carefully documented in the API (I hope).
llvm-svn: 207935
Both MinGW and cygwin (i686) construct export directives without the global
leader prefix. This is mostly due to the fact that they use GNU ld which does
not correctly handle the export directive. This apparently has been been broken
for a while. However, this was recently reported as being broken by
mingwandroid and diorcety of the msys2 project.
Remove the global leader prefix if targeting MinGW or cygwin, otherwise, retain
the global leader prefix. Add an explicit test for cygwin's behaviour of export
directives.
llvm-svn: 207926
Create a helper function to generate the export directive. This was previously
duplicated inline to handle export directives for variables and functions. This
also enables the use of range-based iterators for the generation of the
directive rather than the traditional loops. NFC.
llvm-svn: 207925
This fix simply ensures that both metadata nodes are path-aware before
performing path-aware alias analysis.
This issue isn't normally triggered in LLVM, because we perform an autoupgrade
of the TBAA metadata to the new format when reading in LL or BC files. This
issue only appears when a client creates the IR manually and mixes old and new
TBAA metadata format.
This fixes <rdar://problem/16760860>.
llvm-svn: 207923
The fix itself is fairly simple: move getAccessVariant to MCValue so that we
replace the old weak expression evaluation with the far more general
EvaluateAsRelocatable.
This then requires that EvaluateAsRelocatable stop when it finds a non
trivial reference kind. And that in turn requires the ELF writer to look
harder for weak references.
Last but not least, this found a case where we were being bug by bug
compatible with gas and accepting an invalid input. I reported pr19647
to track it.
llvm-svn: 207920
There is no point in creating it if we're not going to vectorize
anything. Creating the map is expensive as it creates large values.
No functionality change.
llvm-svn: 207916
which are corresponding to the current target read from the ELF file.
This fix cannot be tested until obj2yaml does not support ELF format.
llvm-svn: 207905
Committed initially in r207724-r207726 and reverted due to compiler-rt
crashes in r207732.
Instead, fix this harder with unordered_map and store the LexicalScopes
by value in the map. This did necessitate moving the definition of
LexicalScope above the definition of LexicalScopes.
Let's see how the buildbots/compilers tolerate unordered_map::emplace +
std::piecewise_construct + std::forward_as_tuple...
llvm-svn: 207876
There are public functions that mutate various members as well as
another private member already, so make all the members private to
avoid the discontinuity and add accessors for the values. Should
be no functional change.
llvm-svn: 207868
Reading line tables in llvm-cov was pretty broken, but would happen to
work as long as no line in the table was 0. It's not clear to me
whether a line of zero *should* show up in these tables, but deciding
to read a string in the middle of the line table is certainly the
wrong thing to do if it does.
I've also added some comments, as trying to figure out what this block
of code was doing was fairly unpleasant.
llvm-svn: 207866
This moves most of GlobalOpt's constructor optimization
code out of GlobalOpt into Transforms/Utils/CDtorUtils.{h,cpp}. The
public interface is a single function OptimizeGlobalCtorsList() that
takes a predicate returning which constructors to remove.
GlobalOpt calls this with a function that statically evaluates all
constructors, just like it did before. This part of the change is
behavior-preserving.
Also add a call to this from GlobalDCE with a filter that removes global
constructors that contain a "ret" instruction and nothing else – this
fixes PR19590.
llvm-svn: 207856
address to AnalyzeLoadFromClobberingLoad. This fixes a bug in load-PRE where
PRE is applied to a load that is not partially redundant.
<rdar://problem/16638765>.
llvm-svn: 207853
.file records are supposed to have a section identifier of 65534
(IMAGE_SCN_DEBUG) rather than 0. This is spelt out clearly within the PE/COFF
specification. Fix this minor oversight with the implementation for support for
.file records.
llvm-svn: 207851
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Samuel Li <samuel.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
llvm-svn: 207846
The register spiller assumes that only one new instruction is created
when spilling and restoring registers, so we need to emit pseudo
instructions for vector register spills and lower them after
register allocation.
v2:
- Fix calculation of lane index
- Extend VGPR liveness to end of program.
v3:
- Use SIMM16 field of S_NOP to specify multiple NOPs.
https://bugs.freedesktop.org/show_bug.cgi?id=75005
llvm-svn: 207843
Previously, LLVM had no knowledge that these instructions actually
modified their address register: fine if they never end up in CodeGen,
but when I'd rather like to write some patterns for them it becomes a
disaster.
The change is mostly straightforward, I think the most significant
design decision was to *always* put the address write-back first. This
allows loads and stores to be accessed more uniformly, for example
permitting the continued sharing of the InstAlias definitions.
I also discovered that the custom Decode logic is no longer needed, so
I removed it.
No tests, because there should be no functionality change.
llvm-svn: 207839
While post-indexed LD1/ST1 instructions do exist for vector loads,
this patch makes use of the more flexible addressing-modes in LDR/STR
instructions.
llvm-svn: 207838
Given the following C code llvm currently generates suboptimal code for
x86-64:
__m128 bss4( const __m128 *ptr, size_t i, size_t j )
{
float f = ptr[i][j];
return (__m128) { f, f, f, f };
}
=================================================
define <4 x float> @_Z4bss4PKDv4_fmm(<4 x float>* nocapture readonly %ptr, i64 %i, i64 %j) #0 {
%a1 = getelementptr inbounds <4 x float>* %ptr, i64 %i
%a2 = load <4 x float>* %a1, align 16, !tbaa !1
%a3 = trunc i64 %j to i32
%a4 = extractelement <4 x float> %a2, i32 %a3
%a5 = insertelement <4 x float> undef, float %a4, i32 0
%a6 = insertelement <4 x float> %a5, float %a4, i32 1
%a7 = insertelement <4 x float> %a6, float %a4, i32 2
%a8 = insertelement <4 x float> %a7, float %a4, i32 3
ret <4 x float> %a8
}
=================================================
shlq $4, %rsi
addq %rdi, %rsi
movslq %edx, %rax
vbroadcastss (%rsi,%rax,4), %xmm0
retq
=================================================
The movslq is uneeded, but is present because of the trunc to i32 and then
sext back to i64 that the backend adds for vbroadcastss.
We can't remove it because it changes the meaning. The IR that clang
generates is already suboptimal. What clang really should emit is:
%a4 = extractelement <4 x float> %a2, i64 %j
This patch makes that legal. A separate patch will teach clang to do it.
Differential Revision: http://reviews.llvm.org/D3519
llvm-svn: 207801
This creates a lot of core infrastructure in which to add, with little
effort, quite a bit more to mips fast-isel
Test Plan: simplestore.ll
Reviewers: dsanders
Reviewed By: dsanders
Differential Revision: http://reviews.llvm.org/D3527
llvm-svn: 207790
This was initialized by llvm-mc (calling setDwarfVersion) but other
clients (such as clang, llc, etc) aren't necessarily initializing this
so we were getting garbage DWARF version values in the output.
Initialize it to a reasonable default (the same default used in llvm-mc,
though this is higher than it was (2) previously).
llvm-svn: 207788
This optimization merges the common part of a group of GEPs, so we can compute
each pointer address by adding a simple offset to the common part.
The optimization is currently only enabled for the NVPTX backend, where it has
a large payoff on some benchmarks.
Review: http://reviews.llvm.org/D3462
Patch by Jingyue Wu.
llvm-svn: 207783
This the LLVM portion that will allow Clang and other frontends to emit
typedefs of void by providing a null type for the typedef's underlying
type.
llvm-svn: 207777
We currently force symbols to be globals in .thumb_set. The intent
seems to be that given
.thumb_set foo, bar
we emit an undefined symbol to bar if it is never defined. The side
effect is that we mark bar as global, even if it is defined, which gas
does not.
Producing an undefined reference to bar is a general difference from MC and gas.
For example, given
a = b
gas will produce an undefined reference to b, MC will not. I would be surprised
if any code depends on this, but it it does, we should fix the general
difference, not special case .thumb_set.
llvm-svn: 207757
The canonical form of the BFM instruction is always one of the more explicit
extract or insert operations, which makes reading output much easier.
llvm-svn: 207752
just connects an SCC to one of its descendants directly. Not much of an
impact. The last one is the hard one -- connecting an SCC to one of its
ancestors, and thereby forming a cycle such that we have to merge all
the SCCs participating in the cycle.
llvm-svn: 207751
of SCCs in the SCC DAG. Exercise them in the big graph test case. These
will be especially useful for establishing invariants in insertion
logic.
llvm-svn: 207749
=[
Turns out that this was the root cause of PR19621. We found a crasher
only recently (likely due to improvements elsewhere in the SLP
vectorizer) but the reduced test case failed all the way back to here.
I've confirmed that reverting this patch both fixes the reduced test
case in PR19621 and the actual source file that led to it, so it seems
to really be rooted here. I've replied to the commit thread with
discussion of my (feeble) attempts to debug this. Didn't make it very
far, so reverting now that we have a good test case so that things can
get back to healthy while the debugging carries on.
llvm-svn: 207746
Summary:
There are two functional changes:
1) The directive is not expanded for the ASM->ASM code path.
2) If PIC is not set, there's no expansion for the ASM->OBJ code path (same behaviour as GAS).
Reviewers: dsanders
Reviewed By: dsanders
Differential Revision: http://reviews.llvm.org/D3482
llvm-svn: 207741
This fixes the memory leak introduced with the initial addition of support for
WoA stack probing. Now that the pseudo-instruction expansion can handle an
external symbol, use that to generate the load which simplifies the logic as
well as avoids the memory leak.
llvm-svn: 207737
This enhances the expansion of the mov32imm pseudo-instruction to support an
external symbol reference. This is motivated by a simplification of the stack
probe emission for Windows on ARM (and fixing a leak).
llvm-svn: 207736
This makes the coff writer compute the correct symbol value for the test in
pr19147. The section is still incorrect, that will be fixed in a followup patch.
llvm-svn: 207728
Summary:
Get rid of UserVariables set, and turn DbgValues into MapVector
to get a fixed ordering, as suggested in review for http://reviews.llvm.org/D3573.
Test Plan: llvm regression tests
Reviewers: dblaikie
Reviewed By: dblaikie
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D3579
llvm-svn: 207720
Breaks GDB buildbot
(http://lab.llvm.org:8011/builders/clang-x86_64-ubuntu-gdb-75/builds/14517)
GCC emits DW_AT_object_pointer /everywhere/ (declaration, abstract
definition, inlined subroutine), but it looks like GCC relies on it
being somewhere other than the declaration, at least. I'll experiment
further & can hopefully still remove it from the inlined_subroutine.
This reverts commit r207705.
llvm-svn: 207719
They just don't need to be there - they're inherited from the abstract
definition. In theory I would like them to be inherited from the
declaration, but the DWARF standard doesn't quite say that... we can
probably do it anyway but I'm less confident about that so I'll leave it
for a separate commit.
llvm-svn: 207717
Not all address taken blocks get inlined. The reason is
that a blocks new address is known only when it is cloned. But e.g.
a branch instruction in a different block could need that address earlier
while it gets cloned. The solution is to collect the set of all
blocks that can potentially get inlined and compute a new block address
up front. Then clone and cleanup.
rdar://16427209
llvm-svn: 207713
This effectively reverts r164326, but adds some comments and
justification and ensures we /don't/ emit the DW_AT_object_pointer on
the (abstract and concrete) definitions. (while still preserving it on
standalone definitions involving ObjC Blocks)
This does increase the size of member function declarations from 7 to 11
bytes, unfortunately, but still seems like the Right Thing to do so that
callers that see only the declaration still have the information about
the object pointer. That said, I don't know what, if any, DWARF
consumers don't have a heuristic to guess this in the case of normal
C++ member functions - perhaps we can remove it entirely.
llvm-svn: 207705
For pattern like ((x >> C1) & Mask) << C2, DAG combiner may convert it
into (x >> (C1-C2)) & (Mask << C2), which makes pattern matching of ubfx
more difficult.
For example:
Given
%shr = lshr i64 %x, 4
%and = and i64 %shr, 15
%arrayidx = getelementptr inbounds [8 x [64 x i64]]* @arr, i64 0, %i64 2, i64 %and
%0 = load i64* %arrayidx
With current shift folding, it takes 3 instrs to compute base address:
lsr x8, x0, #1
and x8, x8, #0x78
add x8, x9, x8
If using ubfx, it only needs 2 instrs:
ubfx x8, x0, #4, #4
add x8, x9, x8, lsl #3
This fixes bug 19589
llvm-svn: 207702
DwarfDebug.h has a SmallVector member containing a unique_ptr of an
incomplete type. MSVC doesn't have key functions, so the vtable and
dtor are emitted in AsmPrinter.cpp, where DwarfDebug's ctor is called.
AsmPrinter.cpp include DwarfUnit.h and doesn't get a complete definition
of DwarfTypeUnit. We could fix the problem by including DwarfUnit.h in
DwarfDebug.h, but that would increase header bloat. Instead, define
~DwarfDebug out of line.
llvm-svn: 207701
There is no need to check if we want to hoist the immediate value of an
shift instruction. Simply return TCC_Free right away.
This change is like r206101, but for X86.
rdar://problem/16190769
llvm-svn: 207692
Summary: negu $reg is equivalent to negu $reg, $reg.
Reviewers: dsanders
Reviewed By: dsanders
Differential Revision: http://reviews.llvm.org/D3510
llvm-svn: 207673
Summary:
The pattern sltu $r1, $r2, $imm is found in handwritten assembly which
is just a shorthand version of sltui $r1, $r2, $imm.
Reviewers: dsanders
Reviewed By: dsanders
Differential Revision: http://reviews.llvm.org/D3508
llvm-svn: 207671
We already do this for shstrtab, so might as well do it for strtab. This
extracts the string table building code into a separate class. The idea
is to use it for other object formats too.
I mostly wanted to do this for the general principle, but it does save a
little bit on object file size. I tried this on a clang bootstrap and
saved 0.54% on the sum of object file sizes (1.14 MB out of 212 MB for
a release build).
Differential Revision: http://reviews.llvm.org/D3533
llvm-svn: 207670
It's been decided that in the future, the floating-point immediate in
instructions like "fcmeq v0.2s, v1.2s, #0.0" will be canonically "0.0", which
has been implemented on AArch64 already but not ARM64.
This fixes that issue.
llvm-svn: 207666
Pretty straightforward, we weren't propagating whether or not an
AllocaInst had 'inalloca' marked on it when it came time to clone it.
The inliner exposed this bug. A reduced testcase is forthcoming.
llvm-svn: 207665
Summary:
The pattern dsll/dsrl $rd, $rt, $rs is found in handwritten assembly which
is just a shorthand version of dsllv/dsrlv $rd, $rt, $rs.
Reviewers: dsanders
Reviewed By: dsanders
Differential Revision: http://reviews.llvm.org/D3486
llvm-svn: 207664
Summary:
The pattern sll/srl $rd, $rt, $rs is found in handwritten assembly which
is just a shorthand version of sllv/srlv $rd, $rt, $rs.
Reviewers: dsanders
Reviewed By: dsanders
Differential Revision: http://reviews.llvm.org/D3483
llvm-svn: 207657
target cannot be determined accurately. This is the case for NaCl where the
sandboxing instructions are added in MC layer, after the MipsLongBranch pass.
It is also the case when the code has inline assembly. Instead of calculating
offset in the MipsLongBranch pass, use %hi(sym1 - sym2) and %lo(sym1 - sym2)
expressions that are resolved during the fixup.
This patch also deletes microMIPS test file test/CodeGen/Mips/micromips-long-branch.ll
and implements microMIPS CHECKs in a much simpler way in a file
test/CodeGen/Mips/longbranch.ll, together with MIPS32 and MIPS64.
llvm-svn: 207656
Only emit calls to compiler-rt asm routines on platforms where they are
present (currently limited to linux i386/x86_64).
Patch by Yuri Gorshenin.
llvm-svn: 207651
The canonical syntax for shifts by a variable amount does not end with 'v', but
that syntax should be supported as an alias (presumably for legacy reasons).
llvm-svn: 207649
AArch64 does not have a CPSR register in the same way that AArch32 does. Most
of its compiler-relevant roles have been taken over by the more specific NZCV
register (representing just the flags set by normal instructions).
Its system control functions still remain, but are now under the
pseudo-register referred to as "PSTATE". They're accessed via various MRS & MSR
instructions described in the reference manual.
llvm-svn: 207645
On instructions using the NZCV register, a couple of conditions have dual
representations: HS/CS and LO/CC (meaning unsigned-higher-or-same/carry-set and
unsigned-lower/carry-clear). The first of these is more descriptive in most
circumstances, so we should print it.
llvm-svn: 207644
Summary:
This isn't supported directly so we rotate the vector by the desired number of
elements, insert to element zero, then rotate back.
The i64 case generates rather poor code on MIPS32. There is an obvious
optimisation to be made in future (do both insert.w's inside a shared
rotate/unrotate sequence) but for now it's sufficient to select valid code
instead of aborting.
Depends on D3536
Reviewers: matheusalmeida
Reviewed By: matheusalmeida
Differential Revision: http://reviews.llvm.org/D3537
llvm-svn: 207640
Summary:
This directive is used for setting up $gp in the beginning of a function.
It expands to three instructions if PIC is enabled:
lui $gp, %hi(_gp_disp)
addui $gp, $gp, %lo(_gp_disp)
addu $gp, $gp, $reg
_gp_disp is a special symbol that the linker sets to the distance between
the lui instruction and the context pointer (_gp).
Reviewers: dsanders
Reviewed By: dsanders
Differential Revision: http://reviews.llvm.org/D3480
llvm-svn: 207637
Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to
piece together an immediate in 16-bit chunks, hex is probably the most
appropriate format.
llvm-svn: 207635
This is mostly aimed at the NEON logical operations and MOVI/MVNI (since they
accept weird shifts which are more naturally understandable in hex notation).
Also changes BRK/HINT etc, which is probably a neutral change, but easier than
the alternative.
llvm-svn: 207634
Since these instructions only accept a 12-bit immediate, possibly shifted left
by 12, the canonical syntax used by the architecture reference manual is "#N {,
lsl #12 }". We should accept an immediate that has already been shifted, (e.g.
Also, print a comment giving the full addend since it can be helpful.
llvm-svn: 207633
edge entirely within an existing SCC. Shockingly, making the connected
component more connected is ... a total snooze fest. =]
Anyways, its wired up, and I even added a test case to make sure it
pretty much sorta works. =D
llvm-svn: 207631
A bunch of switch cases were missing, not just for ARM64 but also for
AArch64_BE. I've fixed all those, but there's zero testing as
ExecutionEngine tests are disabled when crosscompiling and I don't
have a native platform available to test on.
llvm-svn: 207626
bits), and discover that it's totally broken. Yay tests. Boo bug. Fix
the basic edge removal so that it works by nulling out the removed edges
rather than actually removing them. This leaves the indices valid in the
map from callee to index, and preserves some of the locality for
iterating over edges. The iterator is made bidirectional to reflect that
it now has to skip over null entries, and the skipping logic is layered
onto it.
As future work, I would like to track essentially the "load factor" of
the edge list, and when it falls below a threshold do a compaction.
An alternative I considered (and continue to consider) is storing the
callees in a doubly linked list where each element of the list is in
a set (which is essentially the classical linked-hash-table
datastructure). The problem with that approach is that either you need
to heap allocate the linked list nodes and use pointers to them, or use
a bucket hash table (with even *more* linked list pointer overhead!),
etc. It's pretty easy to get 5x overhead for values that are just
pointers. So far, I think punching holes in the vector, and periodic
compaction is likely to be much more efficient overall in the space/time
tradeoff.
llvm-svn: 207619
This introduces the stack lowering emission of the stack probe function for
Windows on ARM. The stack on Windows on ARM is a dynamically paged stack where
any page allocation which crosses a page boundary of the following guard page
will cause a page fault. This page fault must be handled by the kernel to
ensure that the page is faulted in. If this does not occur and a write access
any memory beyond that, the page fault will go unserviced, resulting in an
abnormal program termination.
The watermark for the stack probe appears to be at 4080 bytes (for
accommodating the stack guard canaries and stack alignment) when SSP is
enabled. Otherwise, the stack probe is emitted on the page size boundary of
4096 bytes.
llvm-svn: 207615
Emit the COFF header when printing out the function. This is important as the
header contains two important pieces of information: the storage class for the
symbol and the symbol type information. This bit of information is required for
the linker to correctly identify the type of symbol that it is dealing with.
llvm-svn: 207613
When building with -Werror=covered-switch-default (as on the buildbots), the
build would fail since all cases are covered by the switch. Move the
llvm_unreachable to the end of the function as an annotation.
llvm-svn: 207609
IMAGE_REL_ARM_MOV32T relocations require that the movw/movt pair-wise
relocation is not split up and reordered. When expanding the mov32imm
pseudo-instruction, create a bundle if the machine operand is referencing an
address. This helps ensure that the relocatable address load is not reordered
by subsequent passes.
Unfortunately, this only partially handles the case as the Constant Island Pass
occurs after the instructions are unbundled and does not properly handle
bundles. That is a more fundamental issue with the pass itself and beyond the
scope of this change.
llvm-svn: 207608
Streamline parsing and dumping line tables:
Prefer composition to multiple inheritance in DWARFDebugLine::ParsingState.
Get rid of the weird concept of "DumpingState" structure.
was:
DWARFDebugLine::DumpingState state(OS);
DWARFDebugLine::parseStatementTable(..., state);
now:
DWARFDebugLine::LineTable LineTable;
LineTable.parse(...);
LineTable.dump(OS);
No functionality change.
llvm-svn: 207599
Currently, musttail codegen is relying on sibcall optimization, and
reporting a fatal error if fails. Sibcall optimization fails when stack
arguments need to be modified, which is insufficient for musttail.
The logic for moving arguments in memory safely is already implemented
for GuaranteedTailCallOpt. This change merely arranges for musttail
calls to use it.
No functional change for GuaranteedTailCallOpt.
Reviewers: espindola
Differential Revision: http://reviews.llvm.org/D3493
llvm-svn: 207598
MSVC 2013 provides std::make_unique, which it finds with ADL when one of
the parameters is std::unique_ptr, leading to an ambiguous overload.
llvm-svn: 207597
It's already set in AMDGPUISelLowering for all GPUs
Patch By: Jan Vesely
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 207592
SI_IF and SI_ELSE are terminators which also produce a value. For
these instructions ISel always inserts a COPY to move their value
to another basic block. This COPY ends up between SI_(IF|ELSE)
and the S_BRANCH* instruction at the end of the block.
This breaks MachineBasicBlock::getFirstTerminator() and also the
machine verifier which assumes that terminators are grouped together at
the end of blocks.
To solve this we coalesce the copy away right after ISel to make sure
there are no instructions in between terminators at the end of blocks.
llvm-svn: 207591
SALU instructions ignore control flow, so it is not always safe to use
them within branches. This is a partial solution to this problem
until we can come up with something better.
llvm-svn: 207590
This is a squash of several optimization commits:
- calculate DIV_Lo and DIV_Hi separately
- use BFE_U32 if we are operating on 32bit values
- use precomputed constants instead of shifting in UDVIREM
- skip the first 32 iterations of udivrem
v2: Check whether BFE is supported before using it
Patch by: Jan Vesely
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 207589
Initial implementation, rather slow
Patch by: Jan Vesely
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 207588
When legalizing ops, with UDIV/UREM set to expand, they automatically
expand to UDIVREM (if legal or custom).
We need to do this manually for legalize types.
v2:
SI should be set to Expand because the type is legal, and it is
automatically lowered to UDIVREM if UDIVREM is Legal/Custom
R600 should set to UDIV/UREM to Custom because it needs to lower them
during type legalization
Patch by: Jan Vesely
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 207587
Seems MSVC wants to be able to codegen inline-definitions of virtual
functions even in TUs that don't define the key function - and it's well
within its rights to do so.
llvm-svn: 207581
This starts in MCJIT::getSymbolAddress where the
unique_ptr<object::Binary> is release()d and (after a cast) passed to a
single caller, MCJIT::addObjectFile.
addObjectFile calls RuntimeDyld::loadObject.
RuntimeDld::loadObject calls RuntimeDyldELF::createObjectFromFile
And the pointer is never owned at this point. I say this point, because
the alternative codepath, RuntimeDyldMachO::createObjectFile certainly
does take ownership, so this seemed like a good hint that this was a/the
right place to take ownership.
llvm-svn: 207580
Move several function definitions into .cpp, unify constructors
and clear() methods (fixing a couple of latent bugs from copy-paste),
turn static function parsePrologue() into Prologue::parse().
More work needed here to untangle weird multiple inheritance
in table parsing and dumping.
No functionality change.
llvm-svn: 207579
The instcomine logic to handle vpermilvar's pd and 256 variants was incorrect.
The _256 variants have indexes into the individual 128 bit lanes and in all
cases it also has to mask out unused bits.
llvm-svn: 207577
This patch changes the vectorization remarks to also inform when
vectorization is possible but not beneficial.
Added tests to exercise some loop remarks.
llvm-svn: 207574
DIE doesn't need to store a pointer to its parent: we can traverse the DIE tree
only with functions getFirstChild() and getSibling(). Parents must be known
only when we construct the tree. Rewrite setDIERelations() procedure in a more
straightforward way, and get rid of lots of now unused DIEMinimal methods.
No functionality change.
llvm-svn: 207563
Summary:
The InstSE class already initializes Predicates to [HasStdEnc].
No functional change (confirmed by diffing tablegen-erated files before and
after)
Differential Revision: http://reviews.llvm.org/D3548
llvm-svn: 207558
Summary:
The InstSE class already initializes Predicates to [HasStdEnc].
No functional change (confirmed by diffing tablegen-erated files before and
after)
Differential Revision: http://reviews.llvm.org/D3547
llvm-svn: 207551
Change `BlockFrequency` to defer to `BranchProbability::scale()` and
`BranchProbability::scaleByInverse()`.
This removes `BlockFrequency::scale()` from its API (and drops the
ability to see the remainder), but the only user was the unit tests. If
some code in the future needs an API that exposes the remainder, we can
add something to `BranchProbability`, but I find that unlikely.
llvm-svn: 207550
Summary:
The MipsPat class already initializes Predicates to [HasStdEnc].
No functional change (confirmed by diffing tablegen-erated files before and
after)
Differential Revision: http://reviews.llvm.org/D3546
llvm-svn: 207548
These were called from distinct places and had significant distinct
behavior. No need to make that a dynamic check inside the function
rather than just having two functions (refactoring some common code into
a helper function to be called from the two separate functions).
llvm-svn: 207539
Summary:
This calls emitOptimizationRemark from the loop unroller and vectorizer
at the point where they make a positive transformation. For the
vectorizer, it reports vectorization and interleave factors. For the
loop unroller, it reports all the different supported types of
unrolling.
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D3456
llvm-svn: 207528
Summary:
This isn't supported directly so we splat the vector element and extract
the most convenient copy.
Reviewers: matheusalmeida
Reviewed By: matheusalmeida
Differential Revision: http://reviews.llvm.org/D3530
llvm-svn: 207524
This patch centralizes the handling of the thumb bit around
MCStreamer::isThumbFunc and makes isThumbFunc handle aliases.
This fixes a corner case, but the main advantage is having just one
way to check if a MCSymbol is thumb or not. This should still be
refactored to be ARM only, but at least now it is just one predicate
that has to be refactored instead of 3 (isThumbFunc,
ELF_Other_ThumbFunc, and SF_ThumbFunc).
llvm-svn: 207522
Seems at some point the intent was to emit fission ranges_base as unique
per CU but the code today emits ranges_base as the start of the ranges
section for all CUs being compiled and all the ranges_base relative
addresses are relative to that. So removing this dead code and leaving
the status quo until there's a reason to change it (perhaps something's
faster if it has distinct ranges for each CU).
llvm-svn: 207464
clang directly from the LLVM test suite! That doesn't work. I've
followed up on the review thread to try and get a viable solution sorted
out, but trying to get the tree clean here.
llvm-svn: 207462
FrameEntry doesn't need to hold a reference to the section it is
located in. Instead, pass DataExtractor as an argument of parsing
function.
No functionality change.
llvm-svn: 207461
There is no need to keep the whole contents of .debug_aranges section
in memory when we build address ranges table. Memory optimization that
used to be in this code (precalculate the size of vector of ranges before
filling it) is not really needed - later we will compact and resize this
vector anyway.
llvm-svn: 207457
(Clang doesn't warn here because it knows the string is benign - the
assert still checks what it's intended to - though putting the correct
parens does make clang-format format the code a little better)
llvm-svn: 207456
Since all 4 ctor calls in DwarfDebug just pass in a trivially
constructed DIE with the right tag type, sink the tag selection down
into the Dwarf*Unit ctors (removing the argument entirely from callers
in DwarfDebug) and initialize the DIE member in DwarfUnit.
llvm-svn: 207448
When evaluating an assembly expression for a relocation, we want to
stop at MCSymbols that are in the symbol table, even if they are variables.
This is needed since the semantics may require that the relocation use them.
That is not the case when computing the value of a symbol in the symbol table.
There are no relocations in this case and we have to keep going until we hit
a section or find out that the expression doesn't have an assembly time
value.
llvm-svn: 207445
Now that the subtle constructScopeDIE has been refactored into two
functions - one returning memory to take ownership of, one returning a
pointer to already owning memory - push unique_ptr through more APIs.
I think this completes most of the unique_ptr ownership of DIEs.
llvm-svn: 207442
While refactoring out constructScopeDIE into two functions I realized we
were emitting DW_AT_object_pointer in the inlined subroutine when we
didn't need to (GCC doesn't, and the abstract subprogram definition has
the information already).
So here's the refactoring and the bug fix. This is one step of
refactoring to remove some subtle memory ownership semantics. It turns
out the original constructScopeDIE returned ownership in its return
value in some cases and not in others. The split into two functions now
separates those two semantics - further cleanup (unique_ptr, etc) will
follow.
llvm-svn: 207441
This reverts commit r207287, reapplying r207286.
I'm hoping that declaring an explicit struct and instantiating
`addBlockEdges()` directly works around the GCC crash from r207286.
This is a lot more boilerplate, though.
llvm-svn: 207438
This commit provides the necessary C/C++ APIs and infastructure to enable fine-
grain progress report and safe suspension points after each pass in the pass
manager.
Clients can provide a callback function to the pass manager to call after each
pass. This can be used in a variety of ways (progress report, dumping of IR
between passes, safe suspension of threads, etc).
The run listener list is maintained in the LLVMContext, which allows a multi-
threaded client to be only informed for it's own thread. This of course assumes
that the client created a LLVMContext for each thread.
This fixes <rdar://problem/16728690>
llvm-svn: 207430
It's bad enough that I have to look up 5 different levels of TableGen class
definitions to work out what bits go where in a simple NEON instruction anyway,
without having to keep track of umpteen unused parameters.
llvm-svn: 207420
The symbol table itself has no relocations, so it is not possible to represent
things like
a = undefined + 1
With the patch we just omit these variables. That matches the behaviour of the
gnu assembler.
llvm-svn: 207419
X86_MAX_OPERANDS is changed to unsigned.
Also, add range-based for loops for affected loops. This in turn
needed an ArrayRef instead of a pointer-to-array in
InternalInstruction.
llvm-svn: 207413
Someone couldn't bear to have a completely orthogonal set of floating-point
registers, so we've got some instructions that only accept v0-v15 (coming in
ARMv9, V128_prime: you're allowed v2, v3, v5, v7, ...).
Anyway, we were permitting even the out of range registers during assembly
(CodeGen handled it correctly). This adds a diagnostic.
llvm-svn: 207412
contract (and be much more useful). It now provides exactly the
post-order traversal a caller might need to perform on newly formed
SCCs.
llvm-svn: 207410
mapping from a basic block to an incoming value, either for removal or
just lookup, is linear in the number of predecessors, and we were doing
this for every entry in the 'Preds' list which is in many cases almost
all of them!
Unfortunately, the fixes are quite ugly. PHI nodes just don't make this
operation easy. The efficient way to fix this is to have a clever
'remove_if' operation on PHI nodes that lets us do a single pass over
all the incoming values of the original PHI node, extracting the ones we
care about. Then we could quickly construct the new phi node from this
list. This would remove the remaining underlying quadratic movement of
unrelated incoming values and the need for silly backwards looping to
"minimize" how often we hit the quadratic case.
This is the last obvious fix for PR19499. It shaves another 20% off the
compile time for me, and while UpdatePHINodes remains in the profile,
most of the time is now stemming from the well known inefficiencies of
LVI and jump threading.
llvm-svn: 207409
by avoiding inlining massive switches merely because they have no
instructions in them. These switches still show up where we fail to form
lookup tables, and in those cases they are actually going to cause
a very significant code size hit anyways, so inlining them is not the
right call. The right way to fix any performance regressions stemming
from this is to enhance the switch-to-lookup-table logic to fire in more
places.
This makes PR19499 about 5x less bad. It uncovers a second compile time
problem in that test case that is unrelated (surprisingly!).
llvm-svn: 207403
each line. This is particularly nice for tracking which run of
a particular pass over a particular function was slow.
This also required making the TimeValue string much more useful. First,
there is a standard format for writing out a date and time. Let's use
that rather than strings that would have to be parsed. Second, actually
output the nanosecond resolution that timevalue claims to have.
This is proving useful working on PR19499, so I figured it would be
generally useful to commit.
llvm-svn: 207385
entry. This is in preparation for generic DW_OP_piece support.
No functional change so far.
http://reviews.llvm.org/D3373
rdar://problem/15928306
llvm-svn: 207368
Only the object streamers need to track if a symbol should be marked thumb or
not. This ports the ELF case. The COFF case is not ported since it is currently
not working for some other reason (I will report a bug).
llvm-svn: 207366
This restores the previous behaviour of just assuming that if you dont specify a
valid triple that you really meant the default triple with an ELF object file.
llvm-svn: 207349
Introduce support for WoA PE/COFF object file emission from LLVM. Add the new
target specific PE/COFF Streamer (ARMWinCOFFStreamer) that handles the ARM
specific behaviour of PE/COFF object emission. ARM exception information is not
yet emitted and is a TODO item.
The ARM specific object writer (ARMWinCOFFObjectWriter) handles the ARM specific
relocation handling in conjunction with the WinCOFFObjectWriter in the MC layer.
The MC layer needs to be updated to deal with the relocation adjustments.
Branch relocations are adjusted by 4 bytes (unlikely their ELF counterparts).
Minor tweaks to switch multiple conditional checks into equivalent switch
statements. The ObjectFileInfo is updated to relax the object file setup for
Windows COFF. Move the architecture checks into an assertion. Windows COFF is
currently only supported on x86, x86_64, and ARM (thumb). Rather than
defaulting to ELF, we will refuse to generate an object file. This is better
though as you do not get an (arbitrary) object file which is different from the
request.
llvm-svn: 207345
This introduces a target specific streamer, X86WinCOFFStreamer, which handles
the target specific behaviour (e.g. WinEH). This is mostly to ensure that
differences between ARM and X86 remain disjoint and do not accidentally cross
boundaries. This is the final staging change for enabling object emission for
Windows on ARM.
llvm-svn: 207344
This is in preparation for promoting WinCOFFStreamer to a base class which will
be shared by the X86 and ARM specific target COFF streamers. Also add a new
getOrCreateSymbolData interface (like MCELFStreamer) for the ARM COFF Streamer.
This makes the COFFStreamer more similar to the ELFStreamer.
llvm-svn: 207343
Stylistic changes to prepare for splitting up the COFFStreamer into target
specific streamers. Tweak some assertion messages. No functional change.
llvm-svn: 207342
Currently, the integrated assembler is the only choice for assembling Windows on
ARM binaries. IAS supports the .file <filename> directive which emits the file
symbol into the resulting object binary. Mark the GNU COFF information to
indicate support for this feature.
llvm-svn: 207341
API requirements much more obvious.
The key here is that there are two totally different use cases for
mutating the graph. Prior to doing any SCC formation, it is very easy to
mutate the graph. There may be users that want to do small tweaks here,
and then use the already-built graph for their SCC-based operations.
This method remains on the graph itself and is documented carefully as
being cheap but unavailable once SCCs are formed.
Once SCCs are formed, and there is some in-flight DFS building them, we
have to be much more careful in how we mutate the graph. These mutation
operations are sunk onto the SCCs themselves, which both simplifies
things (the code was already there!) and helps make it obvious that
these interfaces are only applicable within that context. The other
primary constraint is that the edge being mutated is actually related to
the SCC on which we call the method. This helps make it obvious that you
cannot arbitrarily mutate some other SCC.
I've tried to write much more complete documentation for the interesting
mutation API -- intra-SCC edge removal. Currently one aspect of this
documentation is a lie (the result list of SCCs) but we also don't even
have tests for that API. =[ I'm going to add tests and fix it to match
the documentation next.
llvm-svn: 207339
Since there's no way to ensure the type unit in the .dwo and the type
unit skeleton in the .o are correlated, this cannot work.
This implementation is a bit inefficient for a few reasons, called out
in comments.
llvm-svn: 207323
Sinking addition of the declaration attribute down to where the
signature is added. So that if the signature is not added neither is the
declaration attribute (this will come in handy when aborting type unit
construction to instead emit the type into the CU directly in some
cases)
Pull out type unit identifier hashing just to simplify the function a
little, it'll be getting longer.
llvm-svn: 207321
Otherwise the legalizer would just scalarize everything. Support for
mulhi in the targets isn't that great yet so on most targets we get
exactly the same scalarized output. Add a test for x86 vector udiv.
I had to disable the mulhi nodes on ARM because there aren't any patterns
for it. As far as I know ARM has instructions for getting the high part of
a multiply so this should be fixed.
llvm-svn: 207315
them, just skip over any DFS-numbered nodes when finding the next root
of a DFS. This allows the entry set to just be a vector as we populate
it from a uniqued source. It also removes the possibility for a linear
scan of the entry set to actually do the removal which can make things
go quadratic if we get unlucky.
llvm-svn: 207312
the DFS stack for leaves in the call graph. As mentioned in my previous
commit, this is particularly interesting for graphs which have high fan
out but low connectivity resulting in many leaves. For such graphs, this
can remove a large % of the DFS stack traffic even though it doesn't
make the stack much smaller.
It's a bit easier to formulate this for the full algorithm because that
one stops completely for each SCC. For example, I was able to directly
eliminate the "Recurse" boolean used to continue an outer loop from the
inner loop.
llvm-svn: 207311
makes working through the worklist much cleaner, and makes it possible
to avoid the 'bool-to-continue-the-outer-loop' hack. Not a huge
difference, but I think this is approaching as polished as I can make
it.
llvm-svn: 207310
more than 1 instruction. The caller need to be aware of this
and adjust instruction iterators accordingly.
rdar://16679376
Repaired r207302.
llvm-svn: 207309
processed in the DFS out of the stack completely. Keep it exclusively in
a variable. Re-shuffle some code structure to make this easier. This can
have a very dramatic effect in some cases because call graphs tend to
look like a high fan-out spanning tree. As a consequence, there are
a large number of leaf nodes in the graph, and this technique causes
leaf nodes to never even go into the stack. While this only reduces the
max depth by 1, it may cause the total number of round trips through the
stack to drop by a lot.
Now, most of this isn't really relevant for the incremental version. =]
But I wanted to prototype it first here as this variant is in ways more
complex. As long as I can get the code factored well here, I'll next
make the primary walk look the same. There are several refactorings this
exposes I think.
llvm-svn: 207306
graph in any way because we don't track edges in the SCC graph, just
nodes. This also lets us add a nice assert about the invariant that
we're working on at least a certain number of nodes within the SCC.
llvm-svn: 207305
The included test case would return the incorrect results, because the expansion
of an shift with a constant shift amount of 0 would generate undefined behavior.
This is because ExpandShiftByConstant assumes that all shifts by constants with
a value of 0 have already been optimized away. This doesn't happen for opaque
constants and usually this isn't a problem, because opaque constants won't take
this code path - they are not supposed to. In the case that the opaque constant
has to be expanded by the legalizer, the legalizer would drop the opaque flag.
In this case we hit the limitations of ExpandShiftByConstant and create incorrect
code.
This commit fixes the legalizer by not dropping the opaque flag when expanding
opaque constants and adding an assertion to ExpandShiftByConstant to catch this
not supported case in the future.
This fixes <rdar://problem/16718472>
llvm-svn: 207304
Reid Kleckner