Polly was accidently modifying a debug info metadata node when
attempting to generate a new unique metadata node for the loop id.
The problem was that we had dwarf metadata that referred to a
metadata node with a null value, like this:
!6 = ... some dwarf metadata referring to !7 ...
!7 = {null}
When we attempt to generate a new metadata node, we reserve the
first space for self-referential node by setting the first argument
to null and then mutating the node later to refer to itself.
However, because the nodes are uniqued based on pointer values, when
we get the new metadata node it actually referred to an existing
node (!7 in the example). When we went to modify the metadata to
point to itself, we were accidently mutating the dwarf metatdata. We
ended up in this situation:
!6 = ... some dwarf metadata referring to !7 ...
!7 = {!7}
and this causes an assert when generating the debug info. The fix is
simple, we just need to use a unique value when getting a new
metadata node. The MDNode::getTemporary() provides exactly the API
we need (and it is used in clang to generate the unique nodes).
Differential Revision: http://reviews.llvm.org/D6174
llvm-svn: 221550
This introduces the symbol rewriter. This is an IR->IR transformation that is
implemented as a CodeGenPrepare pass. This allows for the transparent
adjustment of the symbols during compilation.
It provides a clean, simple, elegant solution for symbol inter-positioning. This
technique is often used, such as in the various sanitizers and performance
analysis.
The control of this is via a custom YAML syntax map file that indicates source
to destination mapping, so as to avoid having the compiler to know the exact
details of the source to destination transformations.
llvm-svn: 221548
The darwin linker lets you rearrange functions and data for better locality
(less paging). You do this with the -order_file option which supplies a text
file containing one symbol per line.
Implementing this required a small change to LayoutPass to add a custom sorter
hook.
llvm-svn: 221545
When FileArchive loads a member, it instantiates a temporary MemoryBuffer
which points to the member range of the archive file. The problem is that the
object file parsers call getBufferIndentifer() on that temporary MemoryBuffer
and store that StringRef as the _path data member for that lld::File. When
FileArchive::instantiateMember() goes out of scope the MemoryBuffer is deleted
and the File::._path becomes a dangling reference.
The fix adds a vector<> to FileArchive to own the instantiated MemoryBuffers.
In addition it fixes member's path to be the standard format
(e.g. "/path/libfoo.a(foo.o)") instead of just the leaf name.
llvm-svn: 221544
Summary:
In the Power architecture, all branch instructions ignore the 2 least
significant bits of the target address. Consequently if you branch to an
invalid address, the address reported along with the SIGSEGV will have
been rounded down to a multiple of 4. Tweak this test accordingly.
This may fix the test for ARM too, in which case we could remove
the XFAIL, but I have no way of testing that.
Reviewers: kcc, willschm, glider, samsonov
Reviewed By: samsonov
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D6167
llvm-svn: 221542
Request `getPairRelocation()` function to get paired relocation type.
That allows us to look up another pairs like R_MICROMIPS_HI16/LO16
in the future.
llvm-svn: 221539
I.E., there is no value is having
void foo() override = 0;
If it is override it is already present in a base class. Since it is pure,
some other class will have to implement it.
llvm-svn: 221537
If clang was configured with a custom gcc toolchain (either by using GCC_INSTALL_PREFIX in cmake or the equivalent configure command), the path to the custom gcc toolchain path takes precedence to the one specified by -ccc-install-dir. This causes several regression tests to fail as they will be using an unexpected path. Adding the switch --gcc-toolchain="" in each test command is not enough as the hexagon toolchain implementation in the driver is not evaluating this argument. This commit modifies the hexagon toolchain to take the --gcc-toolchain="" argument into account when deciding the toolchain path, similarly to what is already done for other targets toolchains. Additionally, the faulty regression tests are modified in order to --gcc-toolchain="" be passed to the commands.
llvm-svn: 221535
Summary:
... and after all that refactoring, it's possible to distinguish softfloat
floating point values from integers so this patch no longer breaks softfloat to
do it.
Remove direct handling of i32's in the N32/N64 ABI by promoting them to
i64. This more closely reflects the ABI documentation and also fixes
problems with stack arguments on big-endian targets.
We now rely on signext/zeroext annotations (already generated by clang) and
the Assert[SZ]ext nodes to avoid the introduction of unnecessary sign/zero
extends.
It was not possible to convert three tests to use signext/zeroext. These tests
are bswap.ll, ctlz-v.ll, ctlz-v.ll. It's not possible to put signext on a
vector type so we just accept the sign extends here for now. These tests don't
pass the vectors the same way clang does (clang puts multiple elements in the
same argument, these map 1 element to 1 argument) so we don't need to worry too
much about it.
With this patch, all known N32/N64 bugs should be fixed and we now pass the
first 10,000 tests generated by ABITest.py.
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D6117
llvm-svn: 221534
A shared library (unlike a .a), has its dependencies recorded in the library and
we can pass PRIVATE to target_link_libraries.
This patch then removes some bogus dependencies when using
BUILD_SHARED_LIBS=ON. For example, we go from
build lib/CodeGen/CMakeFiles/LLVMCodeGen.dir/AggressiveAntiDepBreaker.cpp.o:
CXX_COMPILER /home/espindola/llvm/llvm/lib/CodeGen/AggressiveAntiDepBreaker.cpp
|| include/llvm/IR/intrinsics_gen lib/libLLVMSupport.so
lib/libLLVMCore.so lib/libLLVMBitReader.so
lib/libLLVMTransformUtils.so lib/libLLVMInstCombine.so
lib/libLLVMScalarOpts.so lib/libLLVMipa.so lib/libLLVMAnalysis.so
lib/libLLVMMCParser.so lib/libLLVMMC.so lib/libLLVMObject.so
lib/libLLVMTarget.so lib/libLLVMProfileData.so
to
build lib/CodeGen/CMakeFiles/LLVMCodeGen.dir/AggressiveAntiDepBreaker.cpp.o:
CXX_COMPILER /home/espindola/llvm/llvm/lib/CodeGen/AggressiveAntiDepBreaker.cpp
|| include/llvm/IR/intrinsics_gen lib/libLLVMSupport.so
lib/libLLVMCore.so lib/libLLVMTransformUtils.so
lib/libLLVMScalarOpts.so lib/libLLVMAnalysis.so lib/libLLVMMC.so
lib/libLLVMTarget.so
In fact, build.ninja goes from 5231028 bytes to 4896759 bytes.
With this, old verisons of bfd ld (2.24 is OK, 2.23 warns) will print a bogus
warning when building with BUILD_SHARED_LIBS.
llvm-svn: 221530
Summary:
One of the calls to AllocateStack (the one in LowerCall) doesn't look like
it should be there but it was there before and removing it breaks the
frame size calculation.
Reviewers: vmedic, theraven
Reviewed By: theraven
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D6116
llvm-svn: 221529
ELFLinkingContext had these two functions, which is really not needed since
the Writer uses a llvm::object template composed of Endianness, Alignment,
Is32bit/64bit. We could just use that and not duplicate functionality.
No Change In Functionality.
llvm-svn: 221523
Summary:
In addition to the usual f128 workaround, it was also necessary to provide
a means of accessing ArgListEntry::IsFixed.
Reviewers: theraven, vmedic
Reviewed By: vmedic
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D6111
llvm-svn: 221518
Summary:
In the long run, it should probably become a calling convention in its own
right but for now just move it out of
MipsISelLowering::analyzeCallOperands() so that we can drop this function
in favour of CCState::AnalyzeCallOperands().
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D6085
llvm-svn: 221517
Summary:
CCState objects already carry this information in their isVarArg() method.
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D6084
llvm-svn: 221516
Summary:
Teach llvm-symbolizer about PowerPC64 ELF function descriptors. Symbols in the .opd section point to function descriptors, the first word of which is a pointer to the real function. For the purposes of symbolizing we pretend that the symbol points directly to the function.
This is enough to get decent function names in stack traces for unoptimized binaries, which fixes the sanitizer print-stack-trace test on PowerPC64 Linux.
Reviewers: kcc, willschm, samsonov
Reviewed By: samsonov
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D6110
llvm-svn: 221514
We would attempt to fold away a call instruction which had been marked
overdefined. However, it's not valid to transition to constant from
overdefined.
This fixes PR21512.
llvm-svn: 221513
We would blindly assume that RTTI data should have the same linkage as
the vtable because we didn't think the RTTI data was external. This
oversight stemmed because we didn't take dllimport into account.
This fixes PR21512.
llvm-svn: 221511
Summary:
This makes PIC levels a Module flag attribute, which can be queried by the
backend. The flag is named `PIC Level`, and can have a value of:
0 - Backend-default
1 - Small-model (-fpic)
2 - Large-model (-fPIC)
These match the `-pic-level' command line argument for clang, and the value of the
preprocessor macro `__PIC__'.
Test Plan:
New flags tests specific for the 'PIC Level' module flag.
Tests to be added as part of a future commit for PowerPC, which will use this new API.
Reviewers: rafael, echristo
Reviewed By: rafael, echristo
Subscribers: rafael, llvm-commits
Differential Revision: http://reviews.llvm.org/D5882
llvm-svn: 221510
Mach-O normalized file reader assumes that the entire file is aligned
to a large boundary. If the in-memory file is not aligned properly, it will
abort with an assertion failure in read32/read64. This patch forces the
in-memory file for the unit test to be aligned at 64-byte boundary.
I found these tests are failing on Windows, but theoretically they could
fail on other platform.
llvm-svn: 221508
Reversing a CB* instruction used to drop the flags on the condition. On the
included testcase, this lead to a read from an undefined vreg.
Using addOperand keeps the flags, here <undef>.
Differential Revision: http://reviews.llvm.org/D6159
llvm-svn: 221507
A pointer's pointee might not be sized: the pointee could be a function.
Report this as IK_NoInduction when calculating isInductionVariable.
This fixes PR21508.
llvm-svn: 221501
Renato Golin