llvm/lib/Target/Sparc, branch stablehttp://llvm.orghttps://git.amat.us/llvm/atom/lib/Target/Sparc?h=stable2013-03-24T19:37:04ZClean up Sparc patterns.2013-03-24T19:37:04ZJakob Stoklund Olesenstoklund@2pi.dk2013-03-24T19:37:04Zurn:sha1:d28e30fcf44b22c897914343fce9061eb62d3b47
The types of register variables no longer need to be specified in output
patterns.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@177845 91177308-0d34-0410-b5e6-96231b3b80d8
Give Sparc instruction patterns direct types instead of register classes.2013-03-24T00:56:20ZJakob Stoklund Olesenstoklund@2pi.dk2013-03-24T00:56:20Zurn:sha1:15a3c18623a05a8b9f2f4ea0b0c15965fda4fe6f
Also update the documentation since Sparc is the nicest backend, and
used as an example in WritingAnLLVMBackend.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@177835 91177308-0d34-0410-b5e6-96231b3b80d8
Use direct types in Sparc def : Pat patterns.2013-03-23T20:35:05ZJakob Stoklund Olesenstoklund@2pi.dk2013-03-23T20:35:05Zurn:sha1:edf03820f1d10bbefcb2821e1e0326cad114ed2e
The SelectionDAG graph has MVT type labels, not register classes, so
this makes it clearer what is happening.
This notation is also robust against adding more types to the IntRegs
register class.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@177829 91177308-0d34-0410-b5e6-96231b3b80d8
Provide the register scavenger to processFunctionBeforeFrameFinalized2013-03-14T20:33:40ZHal Finkelhfinkel@anl.gov2013-03-14T20:33:40Zurn:sha1:3080d23fde4981835d8a7faf46c152441fadb11f
Add the current PEI register scavenger as a parameter to the
processFunctionBeforeFrameFinalized callback.
This change is necessary in order to allow the PowerPC target code to
set the register scavenger frame index after the save-area offset
adjustments performed by processFunctionBeforeFrameFinalized. Only
after these adjustments have been made is it possible to estimate
the size of the stack frame.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@177108 91177308-0d34-0410-b5e6-96231b3b80d8
ArrayRefize some code. No functionality change.2013-03-07T20:33:29ZBenjamin Kramerbenny.kra@googlemail.com2013-03-07T20:33:29Zurn:sha1:3853f74aba301ef08b699bac2fa8e53230714a58
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@176648 91177308-0d34-0410-b5e6-96231b3b80d8
Move the eliminateCallFramePseudoInstr method from TargetRegisterInfo2013-02-21T20:05:00ZEli Benderskyeliben@google.com2013-02-21T20:05:00Zurn:sha1:700ed80d3da5e98e05ceb90e9bfb66058581a6db
to TargetFrameLowering, where it belongs. Incidentally, this allows us
to delete some duplicated (and slightly different!) code in TRI.
There are potentially other layering problems that can be cleaned up
as a result, or in a similar manner.
The refactoring was OK'd by Anton Korobeynikov on llvmdev.
Note: this touches the target interfaces, so out-of-tree targets may
be affected.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@175788 91177308-0d34-0410-b5e6-96231b3b80d8
Move MRI liveouts to Sparc return instructions.2013-02-05T18:16:58ZJakob Stoklund Olesenstoklund@2pi.dk2013-02-05T18:16:58Zurn:sha1:067e5a2a1a3ed35bce7dc9e9b2eedee501db53c9
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@174413 91177308-0d34-0410-b5e6-96231b3b80d8
[PEI] Pass the frame index operand number to the eliminateFrameIndex function.2013-01-31T20:02:54ZChad Rosiermcrosier@apple.com2013-01-31T20:02:54Zurn:sha1:108fb3202af6f500073cdbb7be32c25d7a273a2e
Each target implementation was needlessly recomputing the index.
Part of rdar://13076458
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@174083 91177308-0d34-0410-b5e6-96231b3b80d8
Teach SDISel to combine fsin / fcos into a fsincos node if the following2013-01-29T02:32:37ZEvan Chengevan.cheng@apple.com2013-01-29T02:32:37Zurn:sha1:8688a58c53b46d2dda9bf50dafd5195790a7ed58
conditions are met:
1. They share the same operand and are in the same BB.
2. Both outputs are used.
3. The target has a native instruction that maps to ISD::FSINCOS node or
the target provides a sincos library call.
Implemented the generic optimization in sdisel and enabled it for
Mac OSX. Also added an additional optimization for x86_64 Mac OSX by
using an alternative entry point __sincos_stret which returns the two
results in xmm0 / xmm1.
rdar://13087969
PR13204
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@173755 91177308-0d34-0410-b5e6-96231b3b80d8
Clean up assignment of CalleeSaveStackSlotSize: get rid of the default and explicitly set this in every target that needs to change it from the default.2013-01-23T16:22:04ZEli Benderskyeliben@google.com2013-01-23T16:22:04Zurn:sha1:e752feee5228bfa33acee35ef9c606ce12f0f173
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@173270 91177308-0d34-0410-b5e6-96231b3b80d8