aboutsummaryrefslogtreecommitdiff
path: root/lib/Target/SparcV9
AgeCommit message (Collapse)Author
2003-07-02(1) Major bug fix: DecomposeArrayRef() replaces its argument instr. andVikram S. Adve
deletes it, but we were merrily trying to fix the operands of that instruction anyway! Instead, fix the replacement instruction. (2) An Improvement: Check for and extract global values in all operands, not just in known pointer operands. For example, they can occur in call arguments, and probably other unforeseeable places as well. This also eliminates the special-case handling of Load and Store. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7053 91177308-0d34-0410-b5e6-96231b3b80d8
2003-07-02A def. operand of a machine instruction may be an ordinary Value*,Vikram S. Adve
not just an Instruction*, at least in one unfortunate case: the first operand to the va_arg instruction. Modify ValueToDefVecMap to map from Value*, not Instruction*. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7052 91177308-0d34-0410-b5e6-96231b3b80d8
2003-07-02Bug/case fixes:Vikram S. Adve
(1) select: Ok to convert a pointer to a float or double. (2) regalloc: Some MachineInstr* for caller-saving code before a call were being inserted before and after the call! (3) Don't insert the caller-saving instructions in the MachineCodeForInstruction for the Call instruction. *All* instructions generated by register allocation need to be recorded in those maps, but it needs to be done uniformly. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7051 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-30Merged in autoconf branch. This provides configuration via the autoconfJohn Criswell
system. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@7014 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-23Add the padding needed for variable-size alloca's, which should work now.Vikram S. Adve
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6859 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-22Remove a ton of extraneous #includesChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6842 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-20Some preprocessors doen't support // comments and get confusedChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6821 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-20RDCCR defines arg. #1, not arg. #2.Vikram S. Adve
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6796 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-18lib/Target/Sparc/Sparc.cpp:Brian Gaeke
Move LowerAllocations, PrintFunction, and SymbolStripping passes, and the corresponding -disable-strip and -d options, over here to the SPARC target-specific bits of llc. Rename -d to -dump-asm. tools/llc/Makefile: Reindent. Add x86 library so that llc compiles again. tools/llc/llc.cpp: Remove support for running arbitrary optimization passes. Use opt instead. Remove LowerAllocations, PrintFunction, and SymbolStripping passes, as noted above. Allow user to select a backend (x86 or SPARC); default to guessing from the endianness/pointer size of the input bytecode file. Fix typos. Delete empty .s file and exit with error status if target does not support static compilation. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6776 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-18These instructions really take three operands. This fixes some assertionsChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6765 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-16Actually, change it to use explicit new/delete, which is more likely to beChris Lattner
optimized INTO an alloca git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6727 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-16Remove two using declsChris Lattner
Remove usage of alloca git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6725 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-16Isolate machine-dependent use of <alloca.h> in "Support/Alloca.h",Brian Gaeke
so that we can easily change its use to be conditional on the result of an autoconf test later. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6723 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-16Rename FInfo.cpp to FunctionInfo.cpp, eliminate FInfo.hChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6712 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-16move contents of include/llvm/Reoptimizer/Mapping/FInfo.h into here, it is ↵Chris Lattner
sparc internal git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6711 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-16Fix invalid number of arguments problemChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6692 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-11Included assert.h so that the code compiles under newer versions of GCC.John Criswell
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6682 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-11Updated for the new projects Makefile.John Criswell
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6678 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-10add some commentsGuochun Shi
add a function ModuloScheduling::dumpFinalSchedule() to print out final schedule git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6677 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-10add an brief instruction what this pass isGuochun Shi
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6676 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-10a simple introduction to this passGuochun Shi
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6675 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-10cleaned codeGuochun Shi
add some comments git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6674 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-08delete useless functionsGuochun Shi
add comment git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6673 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-08 change DEBUG to DEBUG_PRINTGuochun Shi
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6672 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-07Do not hastily change the Opcode from 'r' to 'i' type if we're not actuallyMisha Brukman
SETTING the operand to be an immediate or have verified that one of the operands is really a SignExtended or Unextended immediate value already, which warrants an 'i' opcode. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6662 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-06Print address out as hex.Misha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6657 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-06Added 'r' and 'i' versions to WRCCR.Misha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6656 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-06* Changed Bcc instructions to behave like BPcc instructionsMisha Brukman
* BPA and BPN do not take a %cc register as a parameter * SLL/SRL/SRA{r,i}5 are there for a reason - they are ONLY 32-bit instructions * Likewise, SLL/SRL/SRAX{r,i}6 are only 64-bit * Added WRCCR{r,i} opcodes git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6655 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-06* Removed PreSelection pass because that is now done in the JITMisha Brukman
* Removed instruction scheduling as it is too slow to run in a JIT environment * Removed other passes because they aren't necessary and can slow JIT down git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6652 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-06Fixed a bunch of test cases in test/Regression/Jello which could not get theMisha Brukman
address of a floating-point (allocated via ConstantPool) correctly. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6647 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-06* If a global is not a function, just ask the MachineCodeEmitter for the addrMisha Brukman
* Do not block a print statement with a DEBUG() guard if we're going to abort() git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6645 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-06The SUB*i instructions belong to a different class than their SUB*r brethren.Misha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6644 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-06Put all debug print statements under the DEBUG() guard to make output clean soMisha Brukman
that tests can automatically diff the output. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6642 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-05Fixed confusion between register classes and register types.Misha Brukman
Now %fcc registers are recognized correctly. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6640 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-05Added missing directive to store the instruction name.Misha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6639 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-05Moved predict and annul fields to the end of each individual instructionMisha Brukman
class, because they are currently unused. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6638 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-05Do not preset the cc register, the instructions actually use it.Misha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6637 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-05Minor tuning -- avoid a non-inlinable function call on every operand.Vikram S. Adve
Also, reorder a couple of functions for inlining. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6635 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-05Added lazy function resolution to the JIT.Misha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6633 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-05* The textual output of (non-)predicted FP branches is the same.Misha Brukman
* Stop mapping FBcc instructions to deprecated opcodes, map to FBPcc instead. * Fixed opf in FCMPxy instructions. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6632 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-05All store instructions really want 'rd' in the first field.Misha Brukman
Special cases: STFSRx and STXFSRx - they operate on predefined rd=0 or rd=1, and expect %fsr as the parameter in assembly. They are disabled (since not used) until an encoding, both for code generation and output, is chosen. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6619 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-05Added missing 'rs1' field to F3_rdrs1imm13, 'rd' to F3_rdrs1rs2.Misha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6618 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-04lib/CodeGen/Mapping/MappingInfo.cpp:Brian Gaeke
Update file comment to contain a bunch of the overview mapping-info documentation previously buried within the file. Remove some unnecessary include/using stmts. Rename pass to MappingInfoCollector. Rewrite a lot of it so it doesn't use global instance variables and so it outputs into MappingInfo objects and then dumps those out, instead of going straight to an assembly file. Change name of factory to getMappingInfoCollector. Fold prologue & epilogue writers into MappingInfo methods. lib/Target/Sparc/FInfo.cpp: Correct file comment to reflect above change lib/Target/Sparc/Sparc.cpp: Change name of factory to getMappingInfoCollector. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6617 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-04Add file comment. Include <vector> and <string>. Update include guardsBrian Gaeke
to reflect file's current location. Add definition of class MappingInfo. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6616 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-04Revert brians patch to get mapping info working againChris Lattner
sorry dude git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6615 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-04Had to comment out a line in outByte() to get it to compile because Out and ↵Tanya Lattner
tmp were undeclared. I was not sure what Brian wanted, so I will let him fix this. But now it compiles. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6614 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-04* Instead of re-inventing the MachineConstantPool emitter that's already givenMisha Brukman
in Emitter.cpp, just convert the Sparc version of the constant pool into what's already supported and inter-operate. * Implemented a first pass at lazy function resolution in the JITResolver. That required adding a SparcV9CodeEmitter pointer to simplify generating bit-patterns of the instructions. * SparcV9CodeEmitter now creates and destroys static TheJITResolver, which makes sense because the SparcV9CodeEmitter is the only user of TheJITResolver, and lives for the entire duration of the JIT (via PassManager which lives in VM). * Changed all return values in the JITResolver to uint64_t because of the 64-bit Sparc architecture. * Added a new version of getting the value of a GlobalValue in the SparcV9CodeEmitter, which now works for already-generated functions (JITted or library functions). * Removed little-used and unused functions, cleaning up the internal view of the SparcV9CodeEmitter. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6612 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-04Make writeNumber() void. Get ready to decouple it from .byte directive output.Brian Gaeke
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6609 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-04I have finally seen the light. The code to change the opcode must live higher inMisha Brukman
the loop, and in both cases. In the first case, it is a VReg that is a constant so it may be actually converted to a constant. In the second case, it is already a constant, but then if it doesn't change its type (e.g. to become a register and have the value loaded from memory if it is too large to live in its instruction field), we must change the opcode BEFORE the 'continue', otherwise we miss the opportunity. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6602 91177308-0d34-0410-b5e6-96231b3b80d8
2003-06-04Added the 4.7 instruction class and all the FMOVcc instructions in them.Misha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@6601 91177308-0d34-0410-b5e6-96231b3b80d8