Commit Graph

468 Commits

Author SHA1 Message Date
Andrew Trick 7bcb0100df Revert "Liveness Analysis Pass"
This reverts commit r197254.

This was an accidental merge of Juergen's patch. It will be checked in
shortly, but wasn't meant to go in quite yet.

Conflicts:
	include/llvm/CodeGen/StackMaps.h
	lib/CodeGen/StackMaps.cpp
	test/CodeGen/X86/stackmap-liveness.ll

llvm-svn: 197260
2013-12-13 18:57:20 +00:00
Andrew Trick 8d6a658430 Liveness Analysis Pass
llvm-svn: 197254
2013-12-13 18:37:03 +00:00
Pekka Jaaskelainen eb4a6e7c28 Guard the debug temp variable with NDEBUG to avoid warning/error with NDEBUG defined.
llvm-svn: 192709
2013-10-15 14:40:46 +00:00
Pekka Jaaskelainen eb08e2e0c8 Do not assert when trying to add a meta data operand with
MachineInstr::addOperand().

llvm-svn: 192707
2013-10-15 14:18:10 +00:00
Matthias Braun 1965bfa4c7 Rename parameter: defined regs are not incoming.
llvm-svn: 192391
2013-10-10 21:28:38 +00:00
Jakob Stoklund Olesen c8a4e3857a Remove dead function.
llvm-svn: 185731
2013-07-05 23:04:55 +00:00
Manman Ren 983a16c08a Debug Info: clean up usage of Verify.
No functionality change.
It should suffice to check the type of a debug info metadata, instead of
calling Verify. For cases where we know the type of a DI metadata, use
assert.

Also update testing cases to make them conform to the format of DI classes.

llvm-svn: 185135
2013-06-28 05:43:10 +00:00
Jakob Stoklund Olesen 4acf7dd866 Remove liveout lists from MachineRegisterInfo.
All targets are now adding return value registers as implicit uses on
return instructions, and there is no longer a need for the live out
lists.

llvm-svn: 174417
2013-02-05 18:21:56 +00:00
Andrew Trick b36388a1cb ScheduleDAG: colorize the DOT graph and improve formatting.
llvm-svn: 173431
2013-01-25 07:45:25 +00:00
Eric Christopher 0cb6fd930e For inline asm:
- recognize string "{memory}" in the MI generation
- mark as mayload/maystore when there's a memory clobber constraint.

PR14859.

Patch by Krzysztof Parzyszek

llvm-svn: 172228
2013-01-11 18:12:39 +00:00
Jakob Stoklund Olesen f0615c71fa Allow hasProperty() to be called on bundle-internal instructions.
When calling hasProperty() on an instruction inside a bundle, it should
always behave as if IgnoreBundle was passed, and just return properties
for the current instruction.

Only attempt to aggregate bundle properties whan asked about the bundle
header.

The assertion fires on existing ARM test cases without this fix.

llvm-svn: 172082
2013-01-10 18:42:44 +00:00
Jakob Stoklund Olesen 55a7be237e Support headerless bundles in MachineInstr::hasProperty().
This function can still work without a BUNDLE header instruction.

llvm-svn: 172029
2013-01-10 01:29:42 +00:00
Jakob Stoklund Olesen 6922e9ca7e Don't print bundle flags.
The bundle flags are used by MachineBasicBlock::print(), they don't need
to clutter up individual MachineInstrs.

llvm-svn: 171986
2013-01-09 18:35:09 +00:00
Jakob Stoklund Olesen 68d752bf6b Don't require BUNDLE headers in MachineInstr::getBundleSize().
It is possible to build MI bundles that don't begin with a BUNDLE
header. Add support for such bundles, counting all instructions inside
the bundle.

llvm-svn: 171985
2013-01-09 18:28:16 +00:00
Sergei Larin 3b46d7ea26 Fix a typo in MachineInstr::unbundleFromSucc() method.
llvm-svn: 171983
2013-01-09 17:54:33 +00:00
Jakob Stoklund Olesen a1b246df57 Pack MachineOperand bitfields better.
Previously, 4 bits were unused.

llvm-svn: 171814
2013-01-07 23:21:44 +00:00
Jakob Stoklund Olesen 5adc4a12fb Pack MachineInstr fields better.
This shrinks MachineInstr to 64 bytes (from 72).

llvm-svn: 171813
2013-01-07 23:21:41 +00:00
Jakob Stoklund Olesen dc5285f102 Don't call destructors on MachineInstr and MachineOperand.
The series of patches leading up to this one makes llc -O0 run 8% faster.

When deallocating a MachineFunction, there is no need to visit all
MachineInstr and MachineOperand objects to deallocate them. All their
memory come from a BumpPtrAllocator that is about to be purged, and they
have empty destructors anyway.

This only applies when deallocating the MachineFunction.
DeleteMachineInstr() should still be used to recycle MI memory during
the codegen passes.

Remove the LeakDetector support for MachineInstr. I've never seen it
used before, and now it definitely doesn't work. With this patch, leaked
MachineInstrs would be much less of a problem since all of their memory
will be reclaimed by ~MachineFunction().

llvm-svn: 171599
2013-01-05 05:05:51 +00:00
Jakob Stoklund Olesen 1bfeecb491 Use ArrayRecycler for MachineInstr operand lists.
Instead of an std::vector<MachineOperand>, use MachineOperand arrays
from an ArrayRecycler living in MachineFunction.

This has several advantages:

- MachineInstr now has a trivial destructor, making it possible to
  delete them in batches when destroying MachineFunction. This will be
  enabled in a later patch.

- Bypassing malloc() and free() can be faster, depending on the system
  library.

- MachineInstr objects and their operands are allocated from the same
  BumpPtrAllocator, so they will usually be next to each other in
  memory, providing better locality of reference.

- Reduce MachineInstr footprint. A std::vector is 24 bytes, the new
  operand array representation only uses 8+4+1 bytes in MachineInstr.

- Better control over operand array reallocations. In the old
  representation, the use-def chains would be reordered whenever a
  std::vector reached its capacity. The new implementation never changes
  the use-def chain order.

Note that some decisions in the code generator depend on the use-def
chain orders, so this patch may cause different assembly to be produced
in a few cases.

llvm-svn: 171598
2013-01-05 05:00:09 +00:00
Chandler Carruth 9fb823bbd4 Move all of the header files which are involved in modelling the LLVM IR
into their new header subdirectory: include/llvm/IR. This matches the
directory structure of lib, and begins to correct a long standing point
of file layout clutter in LLVM.

There are still more header files to move here, but I wanted to handle
them in separate commits to make tracking what files make sense at each
layer easier.

The only really questionable files here are the target intrinsic
tablegen files. But that's a battle I'd rather not fight today.

I've updated both CMake and Makefile build systems (I think, and my
tests think, but I may have missed something).

I've also re-sorted the includes throughout the project. I'll be
committing updates to Clang, DragonEgg, and Polly momentarily.

llvm-svn: 171366
2013-01-02 11:36:10 +00:00
Jakob Stoklund Olesen 7bca670a8b Remove a special case that doesn't seem necessary any longer.
Back when this exception was added, it was skipping a lot more code, but
now it just looks like a premature optimization.

llvm-svn: 170989
2012-12-22 17:33:22 +00:00
Jakob Stoklund Olesen b089483993 Use getNumOperands() instead of Operands.size().
The representation of the Operands array is going to change soon so it
can be allocated from a BumpPtrAllocator.

llvm-svn: 170988
2012-12-22 17:13:06 +00:00
Jakob Stoklund Olesen 2455b58551 Require the two-argument MI::addOperand(MF, MO) for dangling instructions.
Instructions that are inserted in a basic block can still be decorated
with addOperand(MO).

Make the two-argument addOperand() function contain the actual
implementation. This function will now always have a valid MF reference
that it can use for memory allocation.

llvm-svn: 170798
2012-12-20 22:54:05 +00:00
Jakob Stoklund Olesen 33f5d1492d Add an MF argument to MI::copyImplicitOps().
This function is often used to decorate dangling instructions, so a
context reference is required to allocate memory for the operands.

Also add a corresponding MachineInstrBuilder method.

llvm-svn: 170797
2012-12-20 22:54:02 +00:00
Jakob Stoklund Olesen ac4210eacb Use two-arg addOperand(MF, MO) internally in MachineInstr when possible.
llvm-svn: 170796
2012-12-20 22:53:58 +00:00