Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | Fix a case where constantexprs could leak into the PPC isel. | Chris Lattner | 2004-08-11 | 1 | -1/+4 | |
| | | | | llvm-svn: 15661 | |||||
* | Remove a bunch of ad-hoc target-specific flags that were only used by the | Chris Lattner | 2004-08-11 | 3 | -57/+16 | |
| | | | | | | old asmprinter. llvm-svn: 15660 | |||||
* | Remove a dead method | Chris Lattner | 2004-08-11 | 1 | -32/+0 | |
| | | | | llvm-svn: 15659 | |||||
* | Finally, the entire instruction asmprinter is now generated from tblgen, woo! | Chris Lattner | 2004-08-11 | 1 | -296/+8 | |
| | | | | llvm-svn: 15658 | |||||
* | Add asmprintergen support for the last X86 instruction that needs it: ↵ | Chris Lattner | 2004-08-11 | 2 | -1/+12 | |
| | | | | | | pcrelative calls. llvm-svn: 15657 | |||||
* | This file is long dead | Chris Lattner | 2004-08-11 | 1 | -241/+0 | |
| | | | | llvm-svn: 15656 | |||||
* | Scrunch memoperands, add a few more for floating point memops | Chris Lattner | 2004-08-11 | 1 | -68/+61 | |
| | | | | | | Eliminate the FPI*m classes, converting them to use FPI instead. llvm-svn: 15655 | |||||
* | Move hacks up | Chris Lattner | 2004-08-11 | 1 | -32/+30 | |
| | | | | llvm-svn: 15654 | |||||
* | Make FPI take asm string and operand list | Chris Lattner | 2004-08-11 | 1 | -39/+33 | |
| | | | | llvm-svn: 15653 | |||||
* | Nuke the Im*i* patterns, by asmprintergenifying all users. | Chris Lattner | 2004-08-11 | 1 | -70/+73 | |
| | | | | llvm-svn: 15652 | |||||
* | X86 instructions that read-modify-write memory are not LLVM two-address ↵ | Chris Lattner | 2004-08-11 | 1 | -78/+60 | |
| | | | | | | instructions. llvm-svn: 15651 | |||||
* | Get rid of the Im8, Im16, Im32 classes, converting more instructions over to | Chris Lattner | 2004-08-11 | 1 | -224/+300 | |
| | | | | | | asmprintergeneration llvm-svn: 15650 | |||||
* | Fix 255.vortex by using getClassB instead of getClass | Nate Begeman | 2004-08-11 | 1 | -2/+1 | |
| | | | | llvm-svn: 15648 | |||||
* | Remove dead method | Chris Lattner | 2004-08-11 | 1 | -13/+0 | |
| | | | | llvm-svn: 15647 | |||||
* | Convert asmprinter to new style of instruction printer | Chris Lattner | 2004-08-11 | 2 | -9/+50 | |
| | | | | | | Start asmprintergen'ifying machine instrs with memory operands. llvm-svn: 15646 | |||||
* | Fill out immediate operand classes, add a new Operand class | Chris Lattner | 2004-08-11 | 1 | -4/+14 | |
| | | | | llvm-svn: 15642 | |||||
* | Fix InstCombine/2004-08-10-BoolSetCC.ll, a bug that is miscompiling | Chris Lattner | 2004-08-11 | 1 | -22/+21 | |
| | | | | | | | 176.gcc. Note that this is apparently not the only bug miscompiling gcc though. :( llvm-svn: 15639 | |||||
* | Breaking up the PowerPC target into 32- and 64-bit subparts, Part III: the rest. | Misha Brukman | 2004-08-11 | 8 | -4348/+26 | |
| | | | | llvm-svn: 15636 | |||||
* | Breaking up the PowerPC target into 32- and 64-bit subparts: Part II: 64-bit. | Misha Brukman | 2004-08-11 | 5 | -0/+269 | |
| | | | | llvm-svn: 15635 | |||||
* | Breaking up the PowerPC target into 32- and 64-bit subparts, Part I: 32-bit. | Misha Brukman | 2004-08-11 | 7 | -29/+277 | |
| | | | | llvm-svn: 15634 | |||||
* | Implement new constructor. | Misha Brukman | 2004-08-10 | 1 | -0/+7 | |
| | | | | llvm-svn: 15633 | |||||
* | Renamed PPC32 (namespace for regs, opcodes) to PPC to include 64-bit targets | Misha Brukman | 2004-08-10 | 11 | -856/+856 | |
| | | | | llvm-svn: 15631 | |||||
* | * Fix file header to use tablegen emacs mode instead of c++ | Misha Brukman | 2004-08-10 | 1 | -2/+3 | |
| | | | | | | * Wrap long line to 80 cols llvm-svn: 15630 | |||||
* | This is purely a formatting patch that gets us closer to the mecca of fitting | Chris Lattner | 2004-08-10 | 1 | -60/+103 | |
| | | | | | | X86InstrInfo.td into 80 columns llvm-svn: 15629 | |||||
* | Drop the first argument of FPI, and asmprinterify fxch | Chris Lattner | 2004-08-10 | 1 | -28/+32 | |
| | | | | llvm-svn: 15628 | |||||
* | Fix casts of float to unsigned long | Nate Begeman | 2004-08-10 | 4 | -69/+176 | |
| | | | | | | | | Replace STDX (store 64 bit int indexed) with STFDX (store double indexed) Fix latent bug in indexed load generation Generate indexed loads and stores in many more cases llvm-svn: 15626 | |||||
* | Fix file header comment: update filename, set tablegen emacs mode. | Misha Brukman | 2004-08-10 | 1 | -1/+1 | |
| | | | | llvm-svn: 15625 | |||||
* | This purely mechanical patch gives the "I" tblgen class operand list and asm | Chris Lattner | 2004-08-10 | 1 | -312/+239 | |
| | | | | | | string operands, and adjusts all users to pass them in instead of using II. llvm-svn: 15624 | |||||
* | Convert Ii32 instructions over to use the asmprinter generator | Chris Lattner | 2004-08-10 | 1 | -16/+16 | |
| | | | | llvm-svn: 15621 | |||||
* | DForm 5/6 extended mneumonics take 3 arguments. | Misha Brukman | 2004-08-10 | 1 | -0/+10 | |
| | | | | llvm-svn: 15620 | |||||
* | * Instruction definitions moved to SparcV9InstrInfo.td for consistency | Misha Brukman | 2004-08-10 | 2 | -753/+805 | |
| | | | | | | * Defined PHI instruction and SparcV9 subclass of Target llvm-svn: 15615 | |||||
* | Renamed SparcV9_Reg.td -> SparcV9RegisterInfo.td for consistency. | Misha Brukman | 2004-08-10 | 1 | -0/+0 | |
| | | | | llvm-svn: 15614 | |||||
* | Fix DForm_4: format is `op r, r, i' | Misha Brukman | 2004-08-10 | 1 | -2/+3 | |
| | | | | llvm-svn: 15613 | |||||
* | Fix comment header, specify type of file `tablegen'. | Misha Brukman | 2004-08-10 | 3 | -3/+3 | |
| | | | | llvm-svn: 15612 | |||||
* | Convert the Ii16 instructions over | Chris Lattner | 2004-08-10 | 1 | -23/+26 | |
| | | | | llvm-svn: 15606 | |||||
* | Convert all Ii8 instructions over to the autogenerated asmprinter. | Chris Lattner | 2004-08-10 | 1 | -36/+36 | |
| | | | | llvm-svn: 15605 | |||||
* | Define the target name so we can use it via TableGen. | Misha Brukman | 2004-08-10 | 1 | -0/+1 | |
| | | | | llvm-svn: 15604 | |||||
* | Paper over CBackend/2004-08-09-va-end-null.ll | Chris Lattner | 2004-08-10 | 1 | -3/+7 | |
| | | | | | | | | | Note that this indicates a serious problem with the way we are emitting varargs, but this should not be properly fixed until after 1.3. This patch SHOULD go into 1.3. llvm-svn: 15602 | |||||
* | Stub definition of the PowerPC CodeEmitter class; this isn't functional (yet). | Misha Brukman | 2004-08-09 | 2 | -8/+122 | |
| | | | | llvm-svn: 15600 | |||||
* | CodePrinter -> AsmPrinter | Misha Brukman | 2004-08-09 | 4 | -8/+8 | |
| | | | | llvm-svn: 15599 | |||||
* | Fix InstCombine/2004-08-09-RemInfLoop.llx | Chris Lattner | 2004-08-09 | 1 | -1/+1 | |
| | | | | | | This should go into the 1.3 branch llvm-svn: 15593 | |||||
* | Remove ClassPrefix variable as it's no longer used. | Misha Brukman | 2004-08-09 | 4 | -4/+0 | |
| | | | | llvm-svn: 15586 | |||||
* | The (future) SparcV8 JIT would do well to have a class prefix. | Misha Brukman | 2004-08-09 | 1 | -0/+1 | |
| | | | | llvm-svn: 15583 | |||||
* | Define the ClassPrefix for SparcV9. | Misha Brukman | 2004-08-09 | 1 | -0/+1 | |
| | | | | llvm-svn: 15581 | |||||
* | Define a ClassPrefix for PowerPC. | Misha Brukman | 2004-08-09 | 1 | -0/+1 | |
| | | | | llvm-svn: 15580 | |||||
* | Classes need to have a prefix name, so that they can be tacked on to the pieces | Misha Brukman | 2004-08-09 | 1 | -0/+1 | |
| | | | | | | that TableGen creates for them, e.g. CodeEmitter, AsmPrinter, etc. llvm-svn: 15579 | |||||
* | Generate a code emitter for PowerPC as well, this will be used in the JIT. | Misha Brukman | 2004-08-09 | 1 | -1/+5 | |
| | | | | llvm-svn: 15578 | |||||
* | Use instruction formats as defined in the PowerPC ISA manual | Misha Brukman | 2004-08-09 | 2 | -608/+472 | |
| | | | | llvm-svn: 15577 | |||||
* | Fix stack size processing now that the return address isn't an implied | Reid Spencer | 2004-08-09 | 1 | -2/+4 | |
| | | | | | | | push onto the top of the stack like x86, which uses the local area offset. This will allow the removal of PowerPCPEI.cpp soon. llvm-svn: 15573 | |||||
* | Fix another really nasty regression that Anshu pointed out. In cases where | Chris Lattner | 2004-08-08 | 1 | -24/+24 | |
| | | | | | | | | | | dangling constant users were removed from a function, causing it to be dead, we never removed the call graph edge from the external node to the function. In most cases, this didn't cause a problem (by luck). This should definitely go into 1.3 llvm-svn: 15570 |