Include search paths can be relative paths. The loadSubdirectoryModuleMaps function
should account for that and respect the -working-directory parameter given to Clang.
rdar://46045849
Differential Revision: https://reviews.llvm.org/D54503
llvm-svn: 346822
The scan was incorrectly skipping the first instruction, so a register
could appear to be dead when it was actually live. This eventually leads
to a machine verifier failure and miscompile in arm-ldst-opt.
Differential Revision: https://reviews.llvm.org/D54491
llvm-svn: 346821
This is needed when cross-compiling for a different target since
CFLAGS may contain additional flags like -resource-dir which
change the location in which compiler-rt builtins are found.
Differential Revision: https://reviews.llvm.org/D54371
llvm-svn: 346820
- Make mergeSymbolRecords a method of PDBLinker to reduce the number of
parameters it needs.
- Remove a stale FIXME comment about error handling. We already drop
unknown symbol records, log them, and continue.
- Update a comment about why we're copying the symbol record. We do it
to realign the record. We can already mutate the symbol record memory,
it's memory allocated by relocateDebugChunk.
- Avoid the extra `CVSymbol NewSym` variable. We can mutate Sym in
place, which is best, since we're mutating the underlying record anyway.
llvm-svn: 346817
If we keep track of if the ContainsCalls bit is set in the MBB flags for each
candidate, then we have a better chance of not checking the candidate for calls
at all.
This saves quite a few checks in some CTMark tests (~200 in Bullet, for
example.)
llvm-svn: 346816
The shift amount of a funnel shift is modulo the scalar bitwidth:
http://llvm.org/docs/LangRef.html#llvm-fshl-intrinsic
...so we can use demanded bits analysis on that operand to simplify it
when we have a power-of-2 bitwidth.
This is another step towards canonicalizing {shift/shift/or} to the
intrinsics in IR.
Differential Revision: https://reviews.llvm.org/D54478
llvm-svn: 346814
LC_BUILD_VERSION records are of variable length. The original code
would use uninitialized memory when the size of a record was exactly 24.
rdar://problem/46032185
llvm-svn: 346812
This patch adds the cxx-benchmark-unittests target so we can start
getting test coverage on the benchmarks, including building with
sanitizers. Because we're only looking for test-coverage, the benchmarks
run for the shortest time possible, and in parallel.
The target is excluded from all by default. It only
builds and runs the libcxx configurations of the benchmarks, and not
any versions built against the systems native standard library.
llvm-svn: 346811
LoopUtils.cpp contains a utility that splits an loop exit block, so that the new block contains only edges coming from the loop. In the case of nested loops, the exit path for the inner loop might also be the back-edge of the outer loop. The new block which is inserted on this path, is now a latch for the outer loop, and it needs to hold the loop metadata for the outer loop. (The test case gives a more concrete view of the situation.)
Patch by Chang Lin (clin1)
Differential Revision: https://reviews.llvm.org/D53876
llvm-svn: 346810
We already determine a bunch of information about an MBB in
getMachineOutlinerMBBFlags. We can reuse that information to avoid calculating
things that must be false/true.
The first thing we can easily check is if an outlined sequence could ever
contain calls. There's no reason to walk over the outlined range, checking for
calls, if we already know that there are no calls in the block containing the
sequence.
llvm-svn: 346809
The cmp+branch variant of this pattern is shown in:
https://bugs.llvm.org/show_bug.cgi?id=34924
...and as discussed there, we probably can't transform
that without a rotate intrinsic. We do have that now
via funnel shift, but we're not quite ready to
canonicalize IR to that form yet. The case with 'select'
should already be transformed though, so that's this patch.
The sequence with negation followed by masking is what we
use in the backend and partly in clang (though that part
should be updated).
https://rise4fun.com/Alive/TplC
%cmp = icmp eq i32 %shamt, 0
%sub = sub i32 32, %shamt
%shr = lshr i32 %x, %shamt
%shl = shl i32 %x, %sub
%or = or i32 %shr, %shl
%r = select i1 %cmp, i32 %x, i32 %or
=>
%neg = sub i32 0, %shamt
%masked = and i32 %shamt, 31
%maskedneg = and i32 %neg, 31
%shl2 = lshr i32 %x, %masked
%shr2 = shl i32 %x, %maskedneg
%r = or i32 %shl2, %shr2
llvm-svn: 346807
This avoids spurious warnings, but could use
a lot of work. For example the number of vector
elements is not verified, and the passed
value type is not checked.
Fixes bug 39486
llvm-svn: 346806
After the change, the tests started failing, as skipped sections can be
equal in size to kMaxSegName.
Changing `<` to `<=` to address the off-by-one problem.
llvm-svn: 346804
Since we never outline anything with fewer than 2 occurrences, there's no
reason to compute cost model information if there's less than that.
llvm-svn: 346803
ray's gcc installation puts C++ headers in PREFIX/include/g++ without
indicating a gcc version at all. Typically this is because the version
is encoded somewhere in PREFIX.
Differential Revision: https://reviews.llvm.org/D53770
llvm-svn: 346802
An extractelement with non-constant index will be lowered either to
scratch or movrel loop in most cases. This patch converts such
instruction into a set of selects if vector size is not too big.
Differential Revision: https://reviews.llvm.org/D54351
llvm-svn: 346800
gdb-index-multiple-cu-2.s puts the symbol in question to another object file %t1.o, so that its CuIndex is affected by the number of CUs in %t.o
Also change `Kind:` in a comment to `Attributes:` as a follow-up of D54480 and D54481
llvm-svn: 346796
Legalizer used to request an ext load from i8 to i1 when promoting
vector element type to i8. Fixed.
Differential Revision: https://reviews.llvm.org/D54440
llvm-svn: 346795
Summary:
NameTypeEntry::Type is a bit-packed value of CU index+attributes (https://sourceware.org/gdb//onlinedocs/gdb/Index-Section-Format.html), which is named cu_index_and_attrs in a local variable in gdb/dwarf2read.c:dw2_symtab_iter_next
The new name CuIndexAndAttrs is more meaningful.
Reviewers: ruiu, dblaikie, espindola
Reviewed By: dblaikie
Subscribers: emaste, aprantl, arichardson, JDevlieghere, arphaman, llvm-commits
Differential Revision: https://reviews.llvm.org/D54481
llvm-svn: 346794
Reorder the bit-field classes and the members of the anonymous union
so that they both match the order in StmtNodes.td.
There is already a fair amount of them, and this is not going to
improve. Therefore lets try to keep some order here.
Strictly NFC.
llvm-svn: 346793
Calling ARRAY_SIZE on a char* will not actually compute it's size, but just the pointer size.
A new Clang warning enabled by default warns about this.
Replaced the call with internal_strnlen.
Differential Revision: https://reviews.llvm.org/D54484
llvm-svn: 346792
undname prints them, and the information is in the decorated name, so we probably shouldn't lose it when undecorating.
I spot-checked a few of the funnier-looking outputs, and undname has the same output.
Differential Revision: https://reviews.llvm.org/D54396
llvm-svn: 346791
Summary:
The comment refers to the field as "Kind:". However, in gdb,
https://sourceware.org/gdb//onlinedocs/gdb/Index-Section-Format.html names it "attributes",
gdb/dwarf2read.c:dw2_symtab_iter_next refers to the whole value as "cu_index_and_attrs"
Change it to `Attributes:` for consistency.
Reviewers: dblaikie
Reviewed By: dblaikie
Subscribers: aprantl, JDevlieghere, arphaman, llvm-commits
Differential Revision: https://reviews.llvm.org/D54480
llvm-svn: 346790
Summary:
This saves a lot of relocations in optimized object files (at the cost
of some cost/increase in linked executable bytes), but gold's 32 bit
gdb-index support has a bug (
https://sourceware.org/bugzilla/show_bug.cgi?id=21894 ) so we can't
switch to this unconditionally. (& even if it weren't for that bug, one
might argue that some users would want to optimize in one direction or
the other - prioritizing object size or linked executable size)
Differential Revision: https://reviews.llvm.org/D54243
llvm-svn: 346789
Summary:
Ranges base address specifiers can save a lot of object size in
relocation records especially in optimized builds.
For an optimized self-host build of Clang with split DWARF and debug
info compression in object files, but uncompressed debug info in the
executable, this change produces about 18% smaller object files and 6%
larger executable.
While it would've been nice to turn this on by default, gold's 32 bit
gdb-index support crashes on this input & I don't think there's any
perfect heuristic to implement solely in LLVM that would suffice - so
we'll need a flag one way or another (also possible people might want to
aggressively optimized for executable size that contains debug info
(even with compression this would still come at some cost to executable
size)) - so let's plumb it through.
Differential Revision: https://reviews.llvm.org/D54242
llvm-svn: 346788
clang-cl does not emit these, but MSVC does, so we need to be able to
handle them.
Because clang-cl does not generate them, it was a bit hard to write a
test. So what I had to do was get an PDB file with some S_CONSTANT
records in using cl and link, dump it using llvm-pdbutil dump -globals
-sym-data to get the bytes of the records, generate the same object file
using clang-cl but with -S to emit an assembly file, and replace all the
S_LDATA32 records with the bytes of the S_CONSTANT records. This way, we
can compile the file using llvm-mc and link it with lld-link.
Differential Revision: https://reviews.llvm.org/D54452
llvm-svn: 346787
In a previous patch, we pre-processed the TPI stream in order to build
the reverse mapping from nested type -> parent type so that we could
accurately reconstruct a DeclContext hierarchy.
However, there were some issues. An LF_NESTTYPE record is really just a
typedef, so although it happens to be used to indicate the name of the
nested type and referring to the global record which defines the type,
it is also used for every other kind of nested typedef. When we rebuild
the DeclContext hierarchy, we want it to be as accurate as possible,
which means that if we have something like:
struct A {
struct B {};
using C = B;
};
We don't want to create two CXXRecordDecls in the AST each with the
exact same definition. We just want to create one for B and then
define C as an alias to B. Previously, however, it would not be able
to distinguish between the two cases and it would treat A::B and
A::C as being two classes each with separate definitions. We address
the first half of improving the pre-processing logic so that only
actual definitions are treated this way.
Later, in a followup patch, we can handle the case of nested
typedefs since we're already going to be enumerating the field list
anyway and this patch introduces the general framework for
distinguishing between the two cases.
Differential Revision: https://reviews.llvm.org/D54357
llvm-svn: 346786
Previously, the extend_vector_inreg opcode required their input register to be the same total width as their output. But this doesn't match up with how the X86 instructions are defined. For X86 the input just needs to be a legal type with at least enough elements to cover the output.
This patch weakens the check on these nodes and allows them to be used as long as they have more input elements than output elements. I haven't changed type legalization behavior so it will still create them with matching input and output sizes.
X86 will custom legalize these nodes by shrinking the input to be a 128 bit vector and once we've done that we treat them as legal operations. We still have one case during type legalization where we must custom handle v64i8 on avx512f targets without avx512bw where v64i8 isn't a legal type. In this case we will custom type legalize to a *extend_vector_inreg with a v16i8 input. After that the input is a legal type so type legalization should ignore the node and doesn't need to know about the relaxed restriction. We are no longer allowed to use the default expansion for these nodes during vector op legalization since the default expansion uses a shuffle which required the widths to match. Custom legalization for all types will prevent us from reaching the default expansion code.
I believe DAG combine works correctly with the released restriction because it doesn't check the number of input elements.
The rest of the patch is changing X86 to use either the vector_inreg nodes or the regular zero_extend/sign_extend nodes. I had to add additional isel patterns to handle any_extend during isel since simplifydemandedbits can create them at any time so we can't legalize to zero_extend before isel. We don't yet create any_extend_vector_inreg in simplifydemandedbits.
Differential Revision: https://reviews.llvm.org/D54346
llvm-svn: 346784
Summary:
llvm-objcopy/strip support `--keep` (for sections) and `--keep-symbols` (for symbols). For consistency and clarity, rename `--keep` to `--keep-section`.
In fact, for GNU compatability, -K is --keep-symbol, so it's weird that the alias `-K` is not the same as the short-ish `--keep`.
Reviewers: jakehehrlich, jhenderson, alexshap, MaskRay, espindola
Reviewed By: jakehehrlich, MaskRay
Subscribers: emaste, arichardson, llvm-commits
Differential Revision: https://reviews.llvm.org/D54477
llvm-svn: 346782
In preparation for the patch which will move some data to the bit-fields
of Stmt. In particular, rename the private variable "Val" -> "Operand"
since the substatement is the operand of the unary operator.
Run clang-format on UnaryOperator. NFC otherwise.
llvm-svn: 346781
The usage of aligned_storage failed to pass the alignment it wanted,
which caused it to have a larger size and alignment that the
std::string's it was intended to store.
This patch manually specifies the alignment, as well as cleaning up
type alias bugs.
llvm-svn: 346779
In D54435 there was some discussion about the expand_tilde flag for
real_path that I wanted to expose through the VFS. The consensus is that
these two things should be separate functions. Since we already have the
code for this I went ahead and added a function expand_tilde that does
just that.
Differential revision: https://reviews.llvm.org/D54448
llvm-svn: 346776
The IEEE-754 Standard makes it clear that fneg(x) and
fsub(-0.0, x) are two different operations. The former is a bitwise
operation, while the latter is an arithmetic operation. This patch
creates a dedicated FNeg IR Instruction to model that behavior.
Differential Revision: https://reviews.llvm.org/D53877
llvm-svn: 346774
Usually MIPS hosts uses a legacy (non IEEE 754-2008) encoding for NaNs.
Tests like `nan_f32` failed in attempt to compare hard-coded IEEE
754-2008 NaN value and a legacy NaN value provided by a system.
llvm-svn: 346773
Alex Lorenz