Frederic Riss [Fri, 12 Dec 2014 17:31:24 +0000 (17:31 +0000)]
Initial dsymutil tool commit.
The goal of this tool is to replicate Darwin's dsymutil functionality
based on LLVM. dsymutil is a DWARF linker. Darwin's linker (ld64) does
not link the debug information, it leaves it in the object files in
relocatable form, but embbeds a `debug map` into the executable that
describes where to find the debug information and how to relocate it.
When releasing/archiving a binary, dsymutil is called to link all the DWARF
information into a `dsym bundle` that can distributed/stored along with
the binary.
With this commit, the LLVM based dsymutil is just able to parse the STABS
debug maps embedded by ld64 in linked binaries (and not all of them, for
example archives aren't supported yet).
Note that the tool directory is called dsymutil, but the executable is
currently called llvm-dsymutil. This discrepancy will disappear once the
tool will be feature complete. At this point the executable will be renamed
to dsymutil, but until then you do not want it to override the system one.
Summary:
This commit enables the MIPS-III target and adds support for code
generation of SELECT nodes. We have to use pseudo-instructions with
custom inserters for these nodes as MIPS-III CPUs do not have
conditional-move instructions.
Andrea Di Biagio [Fri, 12 Dec 2014 15:09:58 +0000 (15:09 +0000)]
Reapply "[MachineScheduler] Fix for PR21807: minor code difference building with/without -g."
This reapplies r224118 with a fix for test 'misched-code-difference-with-debug.ll'.
That test was failing on some buildbots because it was x86 specific but it was
missing a target triple.
Added an explicit triple to test misched-code-difference-with-debug.ll.
[mips] Support SELECT nodes for targets that don't have conditional-move instructions.
Summary:
For Mips targets that do not have conditional-move instructions, ie. targets
before MIPS32 and MIPS-IV, we have to insert a diamond control-flow
pattern in order to support SELECT nodes. In order to do that, we add
pseudo-instructions with a custom inserter that emits the necessary
control-flow that selects the correct value.
With this patch we add complete support for code generation of Mips-II targets
based on the LLVM test-suite.
Andrea Di Biagio [Fri, 12 Dec 2014 12:41:22 +0000 (12:41 +0000)]
[MachineScheduler] Fix for PR21807: minor code difference building with/without -g.
This patch fixes the issue reported as PR21807. There was a minor difference
in the generated code depending on the -g flag.
The cause was that with -g the machine scheduler used a different
scheduling strategy. This decision was based on the number of instructions
in a schedule region and included debug instructions in that count.
This patch fixes the issue in MISched and provides a test.
Charlie Turner [Fri, 12 Dec 2014 11:59:18 +0000 (11:59 +0000)]
Emit Tag_ABI_FP_16bit_format build attribute.
The __fp16 type is unconditionally exposed. Since -mfp16-format is not yet
supported, there is not a user switch to change this behaviour. This build
attribute should capture the default behaviour of the compiler, which is to
expose the IEEE 754 version of __fp16.
When -mfp16-format is emitted, that will be the way to control the value of
this build attribute.
Yaron Keren [Fri, 12 Dec 2014 11:07:51 +0000 (11:07 +0000)]
Silence gcc 4.9.1 warning 'xyz' is used uninitialized in this function.
In release builds this is actually possible as without asserts there is
no testing of the actual read bytes and the variables could be partially
uninitialized.
DW_OP_const <const> doesn't describe a constant value, but a value at a constant address.
The proper way to describe a constant value is DW_OP_constu <const>, DW_OP_stack_value.
Added DW_OP_stack_value to the stack.
Marked incorrect-variable-debugloc1.ll to xfail for PowerPC64, while the the failure (PR21881)
is being investigated.
Steven Wu [Fri, 12 Dec 2014 04:34:07 +0000 (04:34 +0000)]
Fix another infinite loop in InstCombine
Summary:
InstCombine infinite-loops for the testcase added
It is because InstCombine is generating instructions that can be
optimized by itself. Fix by not optimizing frem if the optimized
type is the same as original type.
rdar://problem/19150820
Matt Arsenault [Fri, 12 Dec 2014 02:30:37 +0000 (02:30 +0000)]
R600: Fix min/max matching problems with unordered compares
The returned operand needs to be permuted for the unordered
compares. Also fix incorrectly producing fmin_legacy / fmax_legacy
for f64, which don't exist.
Matt Arsenault [Fri, 12 Dec 2014 02:30:29 +0000 (02:30 +0000)]
R600/SI: Don't promote f32 select to i32
This is nice for the instruction patterns, but it complicates
min / max matching. The select doesn't have the correct type and would
require looking through the bitcasts for the real float operands.
Matthias Braun [Fri, 12 Dec 2014 01:27:01 +0000 (01:27 +0000)]
Document that PassManager::add() may delete the pass right away.
Also remove redundant documentation:
- doxygen will copy documentation to overriden methods.
- Use \copydoc on PIMPL classes instead of replicating the text.
Philip Reames [Fri, 12 Dec 2014 00:49:03 +0000 (00:49 +0000)]
Comment and minor code cleanup for GCStrategy (NFC)
Updating comments to reflect the current state of the world after my recent changes to ownership structure and generally better describe what a GCStrategy is and how it works.
Matt Arsenault [Fri, 12 Dec 2014 00:00:24 +0000 (00:00 +0000)]
Add target hook for whether it is profitable to reduce load widths
Add an option to disable optimization to shrink truncated larger type
loads to smaller type loads. On SI this prevents using scalar load
instructions in some cases, since there are no scalar extloads.
Matt Arsenault [Thu, 11 Dec 2014 23:37:32 +0000 (23:37 +0000)]
R600/SI: Don't verify constant bus usage of flag ops
This was checking if pseudo-operands like the source
modifiers were using the constant bus, which happens to work
because the values these all can be happen to be valid inline
immediates.
This fixes a later commit which starts checking the register class
of the operands.
`MDString`s can have arbitrary characters in them. Prevent an assertion
that fired in `BitcodeWriter` because of sign extension by copying the
characters into the record as `unsigned char`s.
This reflects the typelessness of `Metadata` in the bitcode format,
removing types from all metadata operands.
`METADATA_VALUE` represents a `ValueAsMetadata`, and always has two
fields: the type and the value.
`METADATA_NODE` represents an `MDNode`, and unlike `METADATA_OLD_NODE`,
doesn't store types. It stores operands at their ID+1 so that `0` can
reference `nullptr` operands.
Hal Finkel [Thu, 11 Dec 2014 22:51:06 +0000 (22:51 +0000)]
[PowerPC] Better lowering for add/or of a FrameIndex
If we have an add (or an or that is really an add), where one operand is a
FrameIndex and the other operand is a small constant, we can combine the
lowering of the FrameIndex (which is lowered as an add of the FI and a zero
offset) with the constant operand.
Amusingly, this is an old potential improvement entry from
lib/Target/PowerPC/README.txt which had never been resolved. In short, we used
to lower:
Alexey Samsonov [Thu, 11 Dec 2014 21:53:03 +0000 (21:53 +0000)]
[ASan] Change fake stack and local variables handling.
This commit changes the way we get fake stack from ASan runtime
(to find use-after-return errors) and the way we represent local
variables:
- __asan_stack_malloc function now returns pointer to newly allocated
fake stack frame, or NULL if frame cannot be allocated. It doesn't
take pointer to real stack as an input argument, it is calculated
inside the runtime.
- __asan_stack_free function doesn't take pointer to real stack as
an input argument. Now this function is never called if fake stack
frame wasn't allocated.
- __asan_init version is bumped to reflect changes in the ABI.
- new flag "-asan-stack-dynamic-alloca" allows to store all the
function local variables in a dynamic alloca, instead of the static
one. It reduces the stack space usage in use-after-return mode
(dynamic alloca will not be called if the local variables are stored
in a fake stack), and improves the debug info quality for local
variables (they will not be described relatively to %rbp/%rsp, which
are assumed to be clobbered by function calls). This flag is turned
off by default for now, but I plan to turn it on after more
testing.
Since `MachineInstr` is required to have a trivial destructor, it cannot
remove itself from `LeakDetection`. Remove the calls.
As it happens, this requirement is because `MachineFunction` allocates
all `MachineInstr`s in a custom allocator; when the `MachineFunction` is
destroyed they're dropped of the edge. There's no benefit to detecting
leaks.
IR: Store MDNodes in a separate LeakDetector container
This gives us better leak detection messages, like `Value` has.
This also has the side effect of papering over a problem where
`MachineInstr`s are added as garbage to the leak detector and then
deleted without being removed. If `MDNode::getTemporary()` allocates an
`MDNodeFwdDecl` in the same spot, the leak detector asserts. By
separating `MDNode`s into their own container we lose that assertion.
Since `MachineInstr` is required to have a trivial destructor, its usage
of `LeakDetector` at all is pretty suspect. I'll be sending a patch
soon to strip that out.
Matthias Braun [Thu, 11 Dec 2014 21:26:47 +0000 (21:26 +0000)]
[CodeGen] Add print and verify pass after each MachineFunctionPass by default
Previously print+verify passes were added in a very unsystematic way, which is
annoying when debugging as you miss intermediate steps and allows bugs to stay
unnotice when no verification is performed.
To make this change practical I added the possibility to explicitely disable
verification. I used this option on all places where no verification was
performed previously (because alot of places actually don't pass the
MachineVerifier).
In the long term these problems should be fixed properly and verification
enabled after each pass. I'll enable some more verification in subsequent
commits.
This is the 2nd attempt at this after realizing that PassManager::add() may
actually delete the pass.
Rather than requiring overloads in the wrapper and the impl, just
overload the impl and use templates in the wrapper. This makes it less
error prone to add more overloads (`void *` defeats any chance the
compiler has at noticing bugs, so the easier the better).
At the same time, correct the comment that was lying about not changing
functionality for `Value`.
Andrea Di Biagio [Thu, 11 Dec 2014 20:44:59 +0000 (20:44 +0000)]
[InstCombine][X86] Improved folding of calls to Intrinsic::x86_sse4a_insertqi.
This patch teaches the instruction combiner how to fold a call to 'insertqi' if
the 'length field' (3rd operand) is set to zero, and if the sum between
field 'length' and 'bit index' (4th operand) is bigger than 64.
From the AMD64 Architecture Programmer's Manual:
1. If the sum of the bit index + length field is greater than 64, then the
results are undefined;
2. A value of zero in the field length is defined as a length of 64.
This patch improves the existing combining logic for intrinsic 'insertqi'
adding extra checks to address both point 1. and point 2.
Matthias Braun [Thu, 11 Dec 2014 19:42:05 +0000 (19:42 +0000)]
[CodeGen] Add print and verify pass after each MachineFunctionPass by default
Previously print+verify passes were added in a very unsystematic way, which is
annoying when debugging as you miss intermediate steps and allows bugs to stay
unnotice when no verification is performed.
To make this change practical I added the possibility to explicitely disable
verification. I used this option on all places where no verification was
performed previously (because alot of places actually don't pass the
MachineVerifier).
In the long term these problems should be fixed properly and verification
enabled after each pass. I'll enable some more verification in subsequent
commits.
Tim Northover [Thu, 11 Dec 2014 18:49:37 +0000 (18:49 +0000)]
ARM: convert isTargetIOS checks to isTargetDarwin.
The distinction is mostly useful in the front-end. By the time we get here,
there are very few situations where we actually want different behaviour for
Darwin and IOS (in fact Darwin mostly just exists in a few tests). So this
should reduce any surprising weirdness for anyone using it.
No functional change on anything anyone actually cares about.
Hal Finkel [Thu, 11 Dec 2014 18:37:52 +0000 (18:37 +0000)]
[PowerPC] Implement BuildSDIVPow2, lower i64 pow2 sdiv using sradi
PPCISelDAGToDAG contained existing code to lower i32 sdiv by a power-of-2 using
srawi/addze, but did not implement the i64 case. DAGCombine now contains a
callback specifically designed for this purpose (BuildSDIVPow2), and part of
the logic has been moved to an implementation of that callback. Doing this
lowering using BuildSDIVPow2 likely does not matter, compared to handling
everything in PPCISelDAGToDAG, for the positive divisor case, but the negative
divisor case, which generates an additional negation, can potentially benefit
from additional folding from DAGCombine. Now, both the i32 and the i64 cases
have been implemented.
Canonicalize formatting of metadata to make it easier to upgrade via
scripts -- in particular, one line per metadata definition makes it more
`sed`-able.
This is preparation for changing the assembly syntax for metadata [1].
Reverting commit 223981, because the test that I added (incorrect-variable-debugloc1.ll) failed for llvm-ppc64.
The test is failing for llvm-ppc64 because for this platform the location list is not being generated at all (most likely because of the bug in PPC code optimization or generation). I will file a bug agains PPC compiler, but meanwhile, until PPC bug is fixed, I will have to revert my change.
Philip Reames [Thu, 11 Dec 2014 01:47:23 +0000 (01:47 +0000)]
GCStrategy should not own GCFunctionInfo
This change moves the ownership and access of GCFunctionInfo (the object which describes the safepoints associated with a safepoint under GCRoot) to GCModuleInfo. Previously, this was owned by GCStrategy which was in turned owned by GCModuleInfo. This made GCStrategy module specific which is 'surprising' given it's name and other purposes.
There's a few more changes needed, but we're getting towards the point we can reuse GCStrategy for gc.statepoint as well.
p.s. The style of this code ends up being a mess. I was trying to move code around without otherwise changing much. Once I get the ownership structure rearranged, I will go through and fixup spacing, naming, comments etc.