| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Add DAG combine for ladd / lsub. | Richard Osborne | 2010-03-09 | 3 | -0/+98 |
| | | | | | llvm-svn: 98057 | ||||
| * | Add an Objective-C test for token annotation | Douglas Gregor | 2010-03-09 | 1 | -0/+22 |
| | | | | | llvm-svn: 98056 | ||||
| * | Replace copy loops with memcpy. | Benjamin Kramer | 2010-03-09 | 1 | -4/+2 |
| | | | | | llvm-svn: 98055 | ||||
| * | Attempt to fix random build failures seen when doing highly | Duncan Sands | 2010-03-09 | 1 | -2/+1 |
| | | | | | | | | parallel builds: the gold plugin fails to link because the lto library is in the middle of being written out by the linker. llvm-svn: 98054 | ||||
| * | Allow more cross-rc coalescing. | Evan Cheng | 2010-03-09 | 1 | -10/+13 |
| | | | | | llvm-svn: 98048 | ||||
| * | reapply r98035: | Chris Lattner | 2010-03-09 | 5 | -28/+29 |
| | | | | | | | | Now that setStartLabel takes an MCSymbol, we can de-ID'ize beginScope and RecordSourceLine. llvm-svn: 98047 | ||||
| * | add some extra checks. I'm not sure why, but this does unbreak a | Chris Lattner | 2010-03-09 | 1 | -0/+3 |
| | | | | | | | failure remaining on mainline. llvm-svn: 98046 | ||||
| * | Revert accidental commit. | Evan Cheng | 2010-03-09 | 1 | -1/+1 |
| | | | | | llvm-svn: 98045 | ||||
| * | Don't do illegal cross-class coalescing. | Jakob Stoklund Olesen | 2010-03-09 | 1 | -1/+1 |
| | | | | | llvm-svn: 98044 | ||||
| * | - Make the machine cse dumb coalescer (as opposed to the more awesome simple | Evan Cheng | 2010-03-09 | 2 | -15/+59 |
| | | | | | | | | | | coalescer) handle sub-register classes. - Add heuristics to avoid non-profitable cse. Given the current lack of live range splitting, avoid cse when an expression has PHI use and the would be new use is in a BB where the expression wasn't already being used. llvm-svn: 98043 | ||||
| * | Don't try to fold V_SET0 and V_SETALLONES to loads in medium and | Dan Gohman | 2010-03-09 | 1 | -0/+5 |
| | | | | | | | large code models. llvm-svn: 98042 | ||||
| * | This is part of an LLC-beta test used to test <rdar://problem/6804645>. Please | Bill Wendling | 2010-03-09 | 1 | -1/+31 |
| | | | | | | | bear with the awful code. It won't last in its current state beyond tonight. llvm-svn: 98040 | ||||
| * | Speculatively revert r98035. It appears to have caused a set of buildbot | Eric Christopher | 2010-03-09 | 5 | -29/+28 |
| | | | | | | | failures. llvm-svn: 98039 | ||||
| * | Attempt to make this debug output meaningful, both in the case of | Dan Gohman | 2010-03-09 | 1 | -3/+6 |
| | | | | | | | multibyte opcodes and in the case of multiple scopes. llvm-svn: 98036 | ||||
| * | Now that setStartLabel takes an MCSymbol, we can de-ID'ize | Chris Lattner | 2010-03-09 | 5 | -28/+29 |
| | | | | | | | beginScope and RecordSourceLine. llvm-svn: 98035 | ||||
| * | change DbgScope to keep track of the start/end label as MCSymbol* | Chris Lattner | 2010-03-09 | 1 | -41/+36 |
| | | | | | | | now that the dependence on ID is removed from MMI. llvm-svn: 98034 | ||||
| * | Make isLCSSA ignore uses in blocks not reachable from the entry block, | Dan Gohman | 2010-03-09 | 2 | -3/+41 |
| | | | | | | | as LCSSA no longer transforms such uses. llvm-svn: 98033 | ||||
| * | remove a useless optimization: now that label replacement never | Chris Lattner | 2010-03-09 | 1 | -7/+0 |
| | | | | | | | happens, the start/end of a scope can never be the same. llvm-svn: 98032 | ||||
| * | strength reduce MMI::MappedLabel to MMI::isLabelDeleted, | Chris Lattner | 2010-03-09 | 5 | -33/+34 |
| | | | | | | | | and add a FIXME about how we are eventually going to zap this lookup table once mc world domination is complete. llvm-svn: 98031 | ||||
| * | inline RemapLabel into its only caller and simplify. | Chris Lattner | 2010-03-09 | 1 | -11/+3 |
| | | | | | llvm-svn: 98029 | ||||
| * | MC/Mach-O: Tweak .dump() formatting. | Daniel Dunbar | 2010-03-09 | 1 | -4/+4 |
| | | | | | llvm-svn: 98028 | ||||
| * | MC/Mach-O: Don't adjust section sizes when aligning zero fill sections, just ↵ | Daniel Dunbar | 2010-03-09 | 1 | -5/+1 |
| | | | | | | | pad the address. llvm-svn: 98027 | ||||
| * | Another place where debug info affected codegen. | Dale Johannesen | 2010-03-09 | 1 | -0/+2 |
| | | | | | llvm-svn: 98026 | ||||
| * | mcstreamerize AsmPrinter::printLabel. | Chris Lattner | 2010-03-09 | 2 | -6/+8 |
| | | | | | llvm-svn: 98025 | ||||
| * | Add inlining threshold to log output. | Jakob Stoklund Olesen | 2010-03-09 | 1 | -1/+4 |
| | | | | | llvm-svn: 98024 | ||||
| * | Disable physical register coalescing when the number of live ranges for the | Jakob Stoklund Olesen | 2010-03-09 | 1 | -0/+12 |
| | | | | | | | | | | | | | | physreg becomes ridiculously high. std::upper_bound may be log(N), but for sufficiently large live intervals, it becomes log(N)*cachemiss = a long long time. This patch improves coalescer time by 4500x for a function with 20000 function calls. The generated code is different, but not significantly worse - the allocator hints are almost as good as physreg coalescing anyway. llvm-svn: 98023 | ||||
| * | Start using DIFile. Corresponding llvm patch is r98020. | Devang Patel | 2010-03-09 | 2 | -150/+129 |
| | | | | | llvm-svn: 98021 | ||||
| * | Start using DIFile. See updated SourceLevelDebugging.html for more information. | Devang Patel | 2010-03-09 | 5 | -142/+201 |
| | | | | | | | | | | This patch updates LLVMDebugVersion to 8. Debug info descriptors encoded using LLVMDebugVersion 7 is supported. Corresponding llvmgcc and clang FE commits are required. llvm-svn: 98020 | ||||
| * | Print blank line and clear stubs vector. | Bill Wendling | 2010-03-09 | 1 | -2/+13 |
| | | | | | llvm-svn: 98019 | ||||
| * | MC-ize the stub printing in ARM. | Bill Wendling | 2010-03-09 | 2 | -5/+18 |
| | | | | | llvm-svn: 98018 | ||||
| * | add a EmitSymbolValue convenience method to MCStreamer. | Chris Lattner | 2010-03-09 | 5 | -9/+14 |
| | | | | | llvm-svn: 98017 | ||||
| * | make InlineInfoLabels hold MCSymbol*'s, avoiding | Chris Lattner | 2010-03-09 | 2 | -11/+11 |
| | | | | | | | recomputation of the labels. llvm-svn: 98016 | ||||
| * | mc'ize the last use of PrintLabelName and eliminate PrintLabelName. | Chris Lattner | 2010-03-09 | 3 | -20/+7 |
| | | | | | llvm-svn: 98015 | ||||
| * | eliminate an argument from PrintRelDirective, sinking | Chris Lattner | 2010-03-09 | 2 | -12/+17 |
| | | | | | | | | the one special case into EmitSectionOffset. MCize the non-special case in EmitSectionOffset. llvm-svn: 98014 | ||||
| * | Print the correct index in the "match failed at index" message. | Dan Gohman | 2010-03-09 | 1 | -1/+1 |
| | | | | | llvm-svn: 98013 | ||||
| * | remove the suffix form of PrintLabelName, which was only | Chris Lattner | 2010-03-09 | 2 | -10/+0 |
| | | | | | | | used for 'flavor'. llvm-svn: 98012 | ||||
| * | now that the debug and eh emitters use a common .set counter, | Chris Lattner | 2010-03-09 | 4 | -11/+5 |
| | | | | | | | we can eliminate "flavor". llvm-svn: 98011 | ||||
| * | move .set generation out of DwarfPrinter into AsmPrinter and | Chris Lattner | 2010-03-08 | 5 | -26/+41 |
| | | | | | | | MCize it. llvm-svn: 98010 | ||||
| * | Don't waste time trying to CSE labels, phis, inline asm. Definitely avoid ↵ | Evan Cheng | 2010-03-08 | 1 | -0/+4 |
| | | | | | | | cse implicit-def for obvious performance reason. llvm-svn: 98009 | ||||
| * | Add test cases for r98003 and r98006. | Ted Kremenek | 2010-03-08 | 2 | -0/+29 |
| | | | | | llvm-svn: 98008 | ||||
| * | Restrict machine cse to really trivial coalescing. Leave the heavy lifting ↵ | Evan Cheng | 2010-03-08 | 1 | -0/+3 |
| | | | | | | | to a real coalescer. llvm-svn: 98007 | ||||
| * | Add preprocessor guards to the definitions of size_t and wchar_t, and #undef ↵ | Ted Kremenek | 2010-03-08 | 1 | -0/+7 |
| | | | | | | | | | NULL before defining it. This addresses potential issues with system headers reported in <rdar://problem/7727159>. llvm-svn: 98006 | ||||
| * | simplify EmitSectionOffset to always use .set if it is | Chris Lattner | 2010-03-08 | 8 | -39/+17 |
| | | | | | | | | | | available, the only thing this affects is that we produce .set in one case we didn't before, which shouldn't harm anything. Make EmitSectionOffset call EmitDifference instead of duplicating it. llvm-svn: 98005 | ||||
| * | don't reset defaults. | Chris Lattner | 2010-03-08 | 2 | -3/+0 |
| | | | | | llvm-svn: 98004 | ||||
| * | Place the definition of 'va_list' within a preprocessor guard. This matches ↵ | Ted Kremenek | 2010-03-08 | 1 | -0/+3 |
| | | | | | | | the behavior of GCC, and avoids potential conflicts with system headers (e.g., stdio.h). Fixes <rdar://problem/7727145>. llvm-svn: 98003 | ||||
| * | Remove a version of EmitDifference. | Chris Lattner | 2010-03-08 | 4 | -51/+26 |
| | | | | | llvm-svn: 98002 | ||||
| * | Fix a crash compiling 254.gap for Thumb2. The Thumb2 add/sub with 12-bit | Bob Wilson | 2010-03-08 | 3 | -8/+292 |
| | | | | | | | | | immediate instructions cannot set the condition codes, so they do not have the extra cc_out operand. We hit an assertion during tail duplication because the instruction being duplicated had more operands that expected. llvm-svn: 98001 | ||||
| * | Re-commit 97860 with fix. getMallocAllocatedType may return null. | Evan Cheng | 2010-03-08 | 2 | -0/+26 |
| | | | | | llvm-svn: 98000 | ||||
| * | eliminate a form of PrintLabelName. | Chris Lattner | 2010-03-08 | 3 | -11/+5 |
| | | | | | llvm-svn: 97999 | ||||
| * | remove another form of EmitReference. | Chris Lattner | 2010-03-08 | 3 | -10/+3 |
| | | | | | llvm-svn: 97998 | ||||

