| Age | Commit message (Collapse) | Author |
|---|
|
Emscripten uses the ARM ABI for pointers to member functions and doesn't
require that member functions are aligned.
|
|
Emscripten uses the Itanium C++ ABI for most things, except that it uses
ARM C++ ABI pointers to member functions, to avoid the overhead of aligning
functions.
|
|
Also, fix the way we override DefaultABIInfo methods, since
classifyArgumentType and classifyReturnType are not actually virtual.
Also, DefaultABIInfo does the right thing for floating-point types,
so we don't need to special-case them.
|
|
For us, staying in sync with upstream on issues related to the memory model
is more important. None of clang, gcc, nor icc appear to offer these features.
If such features are desirable, they should ideally be implemented in upstream
clang, in a target-independent way, since this issue would affect users of all
platforms which support threads, not just PNaCl or Emscripten.
|
|
Notable changes from le32-unknown-nacl so far include:
- Set i32 as the legal integer set, to help the optimizer avoid creating
needlessly inefficient code for asm.js.
- We can use llvm.pow.
- Don't predefine __ELF__ or __pnacl__ so that we don't need to undefine
them later.
- Do predefine asm.js and Emscripten macros, so that we don't need to
define them later.
- Don't provide __has_feature(pnacl).
|
|
libcalls.
For integer types of sizes 1, 2, 4 and 8, libcompiler-rt (and libgcc)
provide atomic functions that pass parameters by value and return
results directly.
libgcc and libcompiler-rt only provide optimized libcalls for
__atomic_fetch_*, as generic libcalls on non-integer types would make
little sense. This means that we can finally make __atomic_fetch_* work
on architectures for which we don't provide these operations as builtins
(e.g. ARM).
This should fix the dreaded "cannot compile this atomic library call
yet" error that would pop up once every while.
These code generation issues are encountered because PNaCl doesn't
inline some of the atomic instructions, whereas other targets do. This
patch is just after the 3.3 branch and applies cleanly, but it doesn't
fix all issues: there is still at least one with atomic operations on
pointers which isn't as clean to fix as applying one patch, so I'll
leave that one as-is for now.
R=dschuff@chromium.org
BUG= https://code.google.com/p/nativeclient/issues/detail?id=3623
TEST= ./pnacl/scripts/llvm-test.py --libcxx-tests
Review URL: https://codereview.chromium.org/59793007
|
|
This is a companion patch to:
https://codereview.chromium.org/22240002/
https://codereview.chromium.org/22474008/
and deals with the Clang-side of things.
The above patch will handle the fallouts of this Clang patch, including
some changes to un-duplicate work that RewriteAsmDirectives.cpp
does. The goal of this patch is to force some extra ordering on
non-atomics for le32 which LLVM doesn't necessarily provide.
R=eliben@chromium.org
TEST= ninja check-all
BUG= https://code.google.com/p/nativeclient/issues/detail?id=3475
BUG= https://code.google.com/p/nativeclient/issues/detail?id=3611
Review URL: https://codereview.chromium.org/22294002
|
|
signature.
__pnacl_profile_function_x takes a single argument - the name of the current function as a constant string, rather than a pair of function addresses like __cyg takes. This makes it work even in PNaCl and removes the need to track symbol information separately.
BUG=none
R=dschuff@chromium.org
Review URL: https://codereview.chromium.org/20000003
|
|
filtering. If omitted entirely, the original behavior is restored.
This also undos the string and __pnacl_profile stuff from the previous
CL. Finally, it fixes and updates the -finstrument-function tests.
BUG=none
R=bradnelson@google.com, dschuff@chromium.org
Review URL: https://codereview.chromium.org/19793007
|
|
Specifically:
r186564 - Fix volatile _Complex alignment test on platforms where 64-bit floating point isn't 64-bit aligned
r186490 - Propagate alignment for _Complex
These should fix GCC torture test failures, as well as the all-important uses of volatile _Complex numbers in C, and their alignment being incorrect.
BUG= PNaCl FYI bots red on torture tests
TEST= ./tools/toolchain_tester/torture_test.py pnacl x86-64 --concurrency=32 >& torture-x86-64.log ; ./tools/toolchain_tester/torture_test.py pnacl x86-32 --concurrency=32 >& torture-x86-32.log ; ./tools/toolchain_tester/torture_test.py pnacl arm --concurrency=32 >& torture-arm.log
R=stichnot@chromium.org
Review URL: https://codereview.chromium.org/19915003
|
|
|
|
Conflicts:
lib/CodeGen/ItaniumCXXABI.cpp
|
|
|
|
SVN log from upstream clang:
r185568 | eliben | 2013-07-03 12:19:12 -0700 (Wed, 03 Jul 2013)
Add target hook CodeGen queries when generating builtin pow*.
Without fmath-errno, Clang currently generates calls to @llvm.pow.* intrinsics
when it sees pow*(). This may not be suitable for all targets (for
example le32/PNaCl), so the attached patch adds a target hook that CodeGen
queries. The target can state its preference for having or not having the
intrinsic generated. Non-PNaCl behavior remains unchanged;
PNaCl-specific test added.
BUG= https://code.google.com/p/nativeclient/issues/detail?id=3513
R=dschuff@chromium.org
Review URL: https://codereview.chromium.org/18953003
|
|
Instead, it should generate calls to the pow* library functions, which
get found within the pexe.
This is a LOCALMOD for now, but I'm working on a more generic solution
that can be upstreamed to Clang.
BUG=https://code.google.com/p/nativeclient/issues/detail?id=3513
R=jvoung@chromium.org
Review URL: https://codereview.chromium.org/18135002
|
|
My previous change set IsARM=true for PNaCl in ItaniumCXXABI.cpp.
This gives us ARM-style representation of method pointers, which we
want, and ARM-style usage of guard variables, which we don't
necessarily want.
Switch the latter back so that the guard variable is tested via "load
i8 and compare with zero" rather than a "load i32 and test the bottom
bit". This should make the Clang-generated code match with how
libstdc++ is using the guard variable.
This makes the code match the patch I sent upstream (which hasn't been
committed yet).
BUG=https://code.google.com/p/nativeclient/issues/detail?id=3450
TEST=test/CodeGenCXX/static-init-pnacl.cpp
Review URL: https://codereview.chromium.org/17616003
|
|
This avoids baking into pexes an assumption that function pointers are
0 mod 2, which might not be the case in future sandboxing models.
BUG=https://code.google.com/p/nativeclient/issues/detail?id=3450
TEST=run_method_pointer_repr_test in NaCl + llvm-lit test
Review URL: https://codereview.chromium.org/17419005
|
|
Was "Implement __declspec(selectany) under -fms-extensions ..."
git-svn-id: https://llvm.org/svn/llvm-project/cfe/branches/release_33@182381 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
------------------------------------------------------------------------
r182266 | rnk | 2013-05-20 07:02:37 -0700 (Mon, 20 May 2013) | 13 lines
Implement __declspec(selectany) under -fms-extensions
selectany only applies to externally visible global variables. It has
the effect of making the data weak_odr.
The MSDN docs suggest that unused definitions can only be dropped at
linktime, so Clang uses weak instead of linkonce. MSVC optimizes away
references to constant selectany data, so it must assume that there is
only one definition, hence weak_odr.
Reviewers: espindola
Differential Revision: http://llvm-reviews.chandlerc.com/D814
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/cfe/branches/release_33@182337 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
------------------------------------------------------------------------
r181368 | rsmith | 2013-05-07 14:53:22 -0700 (Tue, 07 May 2013) | 3 lines
Don't crash in IRGen if a conditional with 'throw' in one of its branches is
used as a branch condition.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/cfe/branches/release_33@181401 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@181272 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
This patch then adds all the usual platform-specific pieces for SystemZ:
driver support, basic target info, register names and constraints,
ABI info and vararg support. It also adds new tests to verify pre-defined
macros and inline asm, and updates a test for the minimum alignment change.
This version of the patch incorporates feedback from reviews by
Eric Christopher and John McCall. Thanks to all reviewers!
Patch by Richard Sandiford.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@181211 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
This patch adds a new common code feature that allows platform code to
request minimum alignment of global symbols. The background for this is
that on SystemZ, the most efficient way to load addresses of global symbol
is the LOAD ADDRESS RELATIVE LONG (LARL) instruction. This instruction
provides PC-relative addressing, but only to *even* addresses. For this
reason, existing compilers will guarantee that global symbols are always
aligned to at least 2. [ Since symbols would otherwise already use a
default alignment based on their type, this will usually only affect global
objects of character type or character arrays. ] GCC also allows creating
symbols without that extra alignment by using explicit "aligned" attributes
(which then need to be used on both definition and each use of the symbol).
To enable support for this with Clang, this patch adds a
TargetInfo::MinGlobalAlign variable that provides a global minimum for the
alignment of every global object (unless overridden via explicit alignment
attribute), and adds code to respect this setting. Within this patch, no
platform actually sets the value to anything but the default 1, resulting
in no change in behaviour on any existing target.
This version of the patch incorporates feedback from reviews by
Eric Christopher and John McCall. Thanks to all reviewers!
Patch by Richard Sandiford.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@181210 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
I was not able to find a case (other than the fix in r181163) where this
makes a difference, but it is a more obviously correct API to have.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@181165 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
constructor from None
Patch by Robert Wilhelm.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@181139 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@181113 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
libgcc provides a __clear_cache intrinsic on AArch64, much like it
does on 32-bit ARM.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@181111 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Un-break the gdb buildbot.
- Use the debug location of the return expression for the cleanup code
if the return expression is trivially evaluatable, regardless of the
number of stop points in the function.
- Ensure that any EH code in the cleanup still gets the line number of
the closing } of the lexical scope.
- Added a testcase with EH in the cleanup.
rdar://problem/13442648
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@181056 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
a lambda.
Bug #1 is that CGF's CurFuncDecl was "stuck" at lambda invocation
functions. Fix that by generally improving getNonClosureContext
to look through lambdas and captured statements but only report
code contexts, which is generally what's wanted. Audit uses of
CurFuncDecl and getNonClosureAncestor for correctness.
Bug #2 is that lambdas weren't specially mapping 'self' when inside
an ObjC method. Fix that by removing the requirement for that
and using the normal EmitDeclRefLValue path in LoadObjCSelf.
rdar://13800041
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@181000 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
This reverts commit 180982.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180990 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Without any conversion, this is pretty straightforward. Most of the
fields can be zeros. The order is:
- field offset or pointer
- nonvirtual adjustment (for MI functions)
- vbptr offset (for unspecified)
- virtual adjustment offset (for virtual inheritance)
Differential Revision: http://llvm-reviews.chandlerc.com/D699
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180985 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
- Use the debug location of the return expression for the cleanup code
if the return expression is trivially evaluatable, regardless of the
number of stop points in the function.
- Ensure that any EH code in the cleanup still gets the line number of
the closing } of the lexical scope.
- Added a testcase with EH in the cleanup.
rdar://problem/13442648
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180982 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
the actual parser and support arbitrary id-expressions.
We're actually basically set up to do arbitrary expressions here
if we wanted to.
Assembly operands permit things like A::x to be written regardless
of language mode, which forces us to embellish the evaluation
context logic somewhat. The logic here under template instantiation
is incorrect; we need to preserve the fact that an expression was
unevaluated. Of course, template instantiation in general is fishy
here because we have no way of delaying semantic analysis in the
MC parser. It's all just fishy.
I've also fixed the serialization of MS asm statements.
This commit depends on an LLVM commit.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180976 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
After some discussion, it was decided to use the Itanium ABI for thread_local on
Darwin OS X platforms. This involved a couple of changes. First, we use
"_tlv_atexit" instead of "__cxa_thread_atexit". Secondly, the global variables
are marked with 'internal' linkage, because we want all access to be calls to
the Itanium-specific entry point, which has normal linkage.
<rdar://problem/13733006>
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180941 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
If there is cleanup code, the cleanup code gets the debug location of
the closing '}'. The subsequent ret IR-instruction does not get a
debug location. The return _expression_ will get the debug location
of the return statement.
If the function contains only a single, simple return statement,
the cleanup code may become the first breakpoint in the function.
In this case we set the debug location for the cleanup code
to the location of the return statement.
rdar://problem/13442648
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180932 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180931 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180823 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
correctly represented as breg+0 locations in the backend.
(Paired commit with LLVM: r180815)
rdar://problem/13658587
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180817 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
systems and did not do the right thing on Darwin.
Original commit message:
Emit the TLS intialization functions into a list.
Add the TLS initialization functions to a list of initialization functions. The
back-end takes this list and places the function pointers into the correct
section. This way they're called before `main().'
<rdar://problem/13733006>
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180809 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Summary:
Like Itanium, comparisons are basically bitwise comparisons of the two
values, with an exception for null member function pointers. If two
function pointers are null, only the function pointer field matters for
comparison purposes. The rest of the bits can be arbitrary. We take
advantage of this in isZeroInitializable(), and it may matter once we
start emitting conversions.
Reviewers: rjmccall
CC: cfe-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D695
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180800 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180795 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
are now two distinct canonical 'AutoType's: one is the undeduced 'auto'
placeholder type, and the other is a deduced-but-dependent type. All
deduced-to-a-non-dependent-type cases are still non-canonical.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180789 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Add the TLS initialization functions to a list of initialization functions. The
back-end takes this list and places the function pointers into the correct
section. This way they're called before `main().'
<rdar://problem/13733006>
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180739 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180656 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
We switch the order of offset and field type to make TBAAStructType node
(name, parent node, offset) similar to scalar TBAA node (name, parent node).
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180653 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180632 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180594 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
For ms structs, zero-length bitfields following non-bitfield members are
completely ignored, we should not increase the field index.
Before the fix, we will have an assertion failure.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180038 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
This is debugging code so functionality isn't a concern, but mingw32 warns
because it doesn't understand the %llx format specifier.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180024 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
thread_local keyword.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@180004 91177308-0d34-0410-b5e6-96231b3b80d8
|
