| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Let us play simplify the td file (and fix a few missed sub and mul patterns). | Andrew Lenharth | 2006-10-31 | 2 | -83/+63 |
| | | | | | llvm-svn: 31322 | ||||
| * | handle "st" as "st(0)" | Chris Lattner | 2006-10-31 | 1 | -3/+12 |
| | | | | | llvm-svn: 31320 | ||||
| * | TargetLowering::isOperandValidForConstraint | Chris Lattner | 2006-10-31 | 1 | -2/+7 |
| | | | | | llvm-svn: 31319 | ||||
| * | Change the prototype for TargetLowering::isOperandValidForConstraint | Chris Lattner | 2006-10-31 | 3 | -17/+28 |
| | | | | | llvm-svn: 31318 | ||||
| * | generalize the fix for PR977 to also fix | Chris Lattner | 2006-10-31 | 1 | -28/+26 |
| | | | | | | | Transforms/LCSSA/2006-10-31-UnreachableBlock-2.ll llvm-svn: 31317 | ||||
| * | Fix PR977 and Transforms/LCSSA/2006-10-31-UnreachableBlock.ll | Chris Lattner | 2006-10-31 | 1 | -1/+8 |
| | | | | | llvm-svn: 31315 | ||||
| * | Add all that branch mangling niftiness | Andrew Lenharth | 2006-10-31 | 9 | -129/+414 |
| | | | | | llvm-svn: 31313 | ||||
| * | add support for calling functions when the caller has variable sized objects | Rafael Espindola | 2006-10-31 | 1 | -1/+19 |
| | | | | | llvm-svn: 31312 | ||||
| * | 1. Clean up code due to changes in SwitchTo*Section(2) | Anton Korobeynikov | 2006-10-31 | 9 | -52/+80 |
| | | | | | | | | | | | 2. Added partial debug support for mingw\cygwin targets (the same as Linux\ELF). Please note, that currently mingw\cygwin uses 'stabs' format for storing debug info by default, thus many (runtime) libraries has this information included. These formats shouldn't be mixed in one binary ('stabs' & 'DWARF'), otherwise binutils tools will be confused. llvm-svn: 31311 | ||||
| * | Turn an assert into an error message. This is commonly triggered when | Chris Lattner | 2006-10-31 | 1 | -1/+5 |
| | | | | | | | | we don't support a specific constraint yet. When this happens, print the unsupported constraint. llvm-svn: 31310 | ||||
| * | Unbreaking static ctors patch. | Anton Korobeynikov | 2006-10-31 | 1 | -1/+1 |
| | | | | | | | | Defaulting second arguments of SwitchTo*Section, this should make things somehow clearer. llvm-svn: 31306 | ||||
| * | Fix a typo which can break jumptables. | Evan Cheng | 2006-10-31 | 1 | -1/+1 |
| | | | | | llvm-svn: 31305 | ||||
| * | Make this compile again. | Reid Spencer | 2006-10-31 | 1 | -1/+1 |
| | | | | | llvm-svn: 31304 | ||||
| * | Apply Aton's LLVM patch for PR973: Linux ctors / dtors support. | Evan Cheng | 2006-10-31 | 1 | -4/+22 |
| | | | | | llvm-svn: 31303 | ||||
| * | Fix a problem introduced by a last-minute change (logic negation). | Reid Spencer | 2006-10-30 | 1 | -2/+2 |
| | | | | | llvm-svn: 31302 | ||||
| * | fix miscompilation of llvm.isunordered, where we branched on the opposite | Chris Lattner | 2006-10-30 | 1 | -2/+2 |
| | | | | | | | condition. This fixes miscompilation of Olden/bh and many others. llvm-svn: 31301 | ||||
| * | Add debug support for X86/ELF targets (Linux). This allows llvm-gcc4 | Reid Spencer | 2006-10-30 | 8 | -17/+52 |
| | | | | | | | | generated object modules to be debugged with gdb. Hopefully this helps pre-release debugging. llvm-svn: 31299 | ||||
| * | fix wonky indentation | Chris Lattner | 2006-10-30 | 1 | -6/+6 |
| | | | | | llvm-svn: 31298 | ||||
| * | Update comment. Valid option is -debug-pass=Structures. | Devang Patel | 2006-10-30 | 1 | -1/+1 |
| | | | | | llvm-svn: 31297 | ||||
| * | Switch abbreviations to use the folding set.a. | Jim Laskey | 2006-10-30 | 1 | -127/+111 |
| | | | | | llvm-svn: 31296 | ||||
| * | Simplify DwarfWriter header. | Jim Laskey | 2006-10-30 | 1 | -97/+508 |
| | | | | | llvm-svn: 31295 | ||||
| * | All targets expand BR_JT for now. | Evan Cheng | 2006-10-30 | 6 | -4/+11 |
| | | | | | llvm-svn: 31294 | ||||
| * | Lower jumptable to BR_JT. The legalizer can lower it to a BRIND or let the ↵ | Evan Cheng | 2006-10-30 | 2 | -28/+55 |
| | | | | | | | target custom lower it. llvm-svn: 31293 | ||||
| * | Added a new SDNode type: BR_JT for jumptable branch. | Evan Cheng | 2006-10-30 | 1 | -0/+1 |
| | | | | | llvm-svn: 31292 | ||||
| * | add a highly efficient hash table that is specialized for mapping C strings | Chris Lattner | 2006-10-29 | 1 | -0/+134 |
| | | | | | | | to some other type. llvm-svn: 31286 | ||||
| * | Add a new llvm::Allocator abstraction, which will be used by a container | Chris Lattner | 2006-10-29 | 1 | -0/+106 |
| | | | | | | | | I'm about to add. This is similar to, but necessarily different than, the STL allocator class. llvm-svn: 31285 | ||||
| * | Fix SimplifyCFG/2006-10-29-InvokeCrash.ll, a crash compiling QT. | Chris Lattner | 2006-10-29 | 1 | -1/+1 |
| | | | | | llvm-svn: 31284 | ||||
| * | Make CanFallThrough more intelligent (so it can handle blocks with (e.g.) no | Chris Lattner | 2006-10-29 | 1 | -62/+94 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | successors), and make island block movement more general. This compiles CodeGen/X86/2006-04-27-ISelFoldingBug.ll to: _loadAndRLEsource_no_exit_2E_1_label_2E_0: subl $8, %esp movl %esi, 4(%esp) movl %ebx, (%esp) movl 16(%esp), %eax movl 12(%esp), %ecx LBB1_3: #label.0 movl _last, %edx movl %edx, %esi incl %esi movl %esi, _last movl %ecx, %ebx # TRUNCATE movb %bl, %bl movl _block, %esi movb %bl, 1(%esi,%edx) cmpl %eax, _last jge LBB1_2 #codeRepl5.exitStub LBB1_4: #label.0 cmpl $257, %ecx je LBB1_2 #codeRepl5.exitStub LBB1_1: #label.0.no_exit.1_crit_edge.exitStub movl $1, %eax movl (%esp), %ebx movl 4(%esp), %esi addl $8, %esp ret LBB1_2: #codeRepl5.exitStub xorl %eax, %eax movl (%esp), %ebx movl 4(%esp), %esi addl $8, %esp ret instead of: _loadAndRLEsource_no_exit_2E_1_label_2E_0: subl $8, %esp movl %esi, 4(%esp) movl %ebx, (%esp) movl 16(%esp), %eax movl 12(%esp), %ecx jmp LBB1_3 #label.0 LBB1_1: #label.0.no_exit.1_crit_edge.exitStub movl $1, %eax movl (%esp), %ebx movl 4(%esp), %esi addl $8, %esp ret LBB1_2: #codeRepl5.exitStub xorl %eax, %eax movl (%esp), %ebx movl 4(%esp), %esi addl $8, %esp ret LBB1_3: #label.0 movl _last, %edx movl %edx, %esi incl %esi movl %esi, _last movl %ecx, %ebx # TRUNCATE movb %bl, %bl movl _block, %esi movb %bl, 1(%esi,%edx) cmpl %eax, _last jge LBB1_2 #codeRepl5.exitStub LBB1_4: #label.0 cmpl $257, %ecx jne LBB1_1 #label.0.no_exit.1_crit_edge.exitStub jmp LBB1_2 #codeRepl5.exitStub ... which is much better layout :) llvm-svn: 31282 | ||||
| * | fix Generic/2006-10-29-Crash.ll | Chris Lattner | 2006-10-29 | 1 | -1/+9 |
| | | | | | llvm-svn: 31281 | ||||
| * | Fix a load folding issue that Evan noticed: there is no need to export values | Chris Lattner | 2006-10-29 | 1 | -7/+17 |
| | | | | | | | used by comparisons in the main block. llvm-svn: 31279 | ||||
| * | Try again. | Jim Laskey | 2006-10-29 | 1 | -1/+4 |
| | | | | | llvm-svn: 31278 | ||||
| * | Not handling zero length strings. | Jim Laskey | 2006-10-29 | 1 | -2/+2 |
| | | | | | llvm-svn: 31277 | ||||
| * | VLOAD is not the LoadSDNode opcode. | Evan Cheng | 2006-10-29 | 1 | -1/+0 |
| | | | | | llvm-svn: 31276 | ||||
| * | Remove spurious case. EXTLOAD is not one of the node opcodes. | Nick Lewycky | 2006-10-29 | 1 | -1/+0 |
| | | | | | llvm-svn: 31275 | ||||
| * | split critical edges more carefully and intelligently. In particular, critical | Chris Lattner | 2006-10-28 | 1 | -4/+61 |
| | | | | | | | | | edges whose destinations are not phi nodes don't bother us. Also, share split edges, since the split edge can't have a phi. This significantly reduces the complexity of generated code in some cases. llvm-svn: 31274 | ||||
| * | Teach branch folding to fold identical jump tables together and to delete | Chris Lattner | 2006-10-28 | 1 | -5/+53 |
| | | | | | | | jump tables that are dead. llvm-svn: 31273 | ||||
| * | const'ify jump table stuff | Chris Lattner | 2006-10-28 | 1 | -1/+1 |
| | | | | | llvm-svn: 31269 | ||||
| * | add an assert | Chris Lattner | 2006-10-28 | 1 | -0/+1 |
| | | | | | llvm-svn: 31267 | ||||
| * | don't print dead jump tables | Chris Lattner | 2006-10-28 | 1 | -0/+3 |
| | | | | | llvm-svn: 31266 | ||||
| * | implement the BlockHasNoFallThrough hook | Chris Lattner | 2006-10-28 | 2 | -0/+12 |
| | | | | | llvm-svn: 31264 | ||||
| * | improve deletion of blocks that just contain branches by knowing that | Chris Lattner | 2006-10-28 | 1 | -33/+38 |
| | | | | | | | the pred block doesn't fall through into them if it's a jumptable. llvm-svn: 31263 | ||||
| * | add another target hook for branch folding. | Chris Lattner | 2006-10-28 | 2 | -0/+13 |
| | | | | | llvm-svn: 31262 | ||||
| * | Load and stores have not been uniqued properly. | Jim Laskey | 2006-10-28 | 1 | -13/+42 |
| | | | | | llvm-svn: 31261 | ||||
| * | Split *all* critical edges before isel. This resolves issues with spill code | Chris Lattner | 2006-10-28 | 1 | -44/+8 |
| | | | | | | | | | | being inserted on unsplit critical edges, which introduces (sometimes large amounts of) partially dead spill code. This also fixes PR925 + CodeGen/Generic/switch-crit-edge-constant.ll llvm-svn: 31260 | ||||
| * | add option to isCriticalEdge | Chris Lattner | 2006-10-28 | 1 | -3/+12 |
| | | | | | llvm-svn: 31258 | ||||
| * | break edges more intelligently | Chris Lattner | 2006-10-28 | 1 | -2/+3 |
| | | | | | llvm-svn: 31257 | ||||
| * | Expose a smarter way to break critical edges. | Chris Lattner | 2006-10-28 | 1 | -5/+24 |
| | | | | | llvm-svn: 31256 | ||||
| * | SplitCriticalEdge checks to see if an edge is critical, don't check twice | Chris Lattner | 2006-10-28 | 1 | -2/+1 |
| | | | | | llvm-svn: 31255 | ||||
| * | Fix a serious bug that caused any x86 vector stuff to infinite loop | Chris Lattner | 2006-10-28 | 1 | -1/+1 |
| | | | | | llvm-svn: 31254 | ||||
| * | Doh. Must check if GV is constant first before putting it in .cstring. | Evan Cheng | 2006-10-28 | 1 | -2/+2 |
| | | | | | llvm-svn: 31253 | ||||

