Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | Split up subtracts into add+negate if they have a reassociable use or operand | Chris Lattner | 2008-02-17 | 2 | -5/+21 | |
| | | | | | | that is also a subtract. This implements PR2047 and Transforms/Reassociate/subtest2.ll llvm-svn: 47241 | |||||
* | upgrade and simplify this test. | Chris Lattner | 2008-02-17 | 1 | -6/+7 | |
| | | | | llvm-svn: 47240 | |||||
* | make the logic for breaking up subtracts more explicit, no | Chris Lattner | 2008-02-17 | 1 | -9/+22 | |
| | | | | | | functionality change. llvm-svn: 47239 | |||||
* | Remove llvm-upgrade. | Tanya Lattner | 2008-02-17 | 50 | -906/+906 | |
| | | | | llvm-svn: 47238 | |||||
* | move PR2053 to here. | Chris Lattner | 2008-02-17 | 1 | -0/+12 | |
| | | | | llvm-svn: 47237 | |||||
* | fix this test. | Chris Lattner | 2008-02-17 | 1 | -1/+1 | |
| | | | | llvm-svn: 47232 | |||||
* | Remove llvm-upgrade | Tanya Lattner | 2008-02-17 | 1 | -10/+6 | |
| | | | | llvm-svn: 47231 | |||||
* | Remove llvm-upgrade. | Tanya Lattner | 2008-02-17 | 2 | -17/+21 | |
| | | | | llvm-svn: 47230 | |||||
* | this test isn't useful since we added @ notation for globals. | Chris Lattner | 2008-02-17 | 1 | -21/+0 | |
| | | | | llvm-svn: 47229 | |||||
* | this line was commented out. | Chris Lattner | 2008-02-17 | 1 | -1/+1 | |
| | | | | llvm-svn: 47228 | |||||
* | Remove llvm-upgrade and update tests. | Tanya Lattner | 2008-02-16 | 2 | -31/+8 | |
| | | | | llvm-svn: 47227 | |||||
* | Removing llvm upgrade, so remove tests specific to llvm-upgrade and update ↵ | Tanya Lattner | 2008-02-16 | 27 | -13270/+13054 | |
| | | | | | | the tests that used it. llvm-svn: 47225 | |||||
* | Remove any 'nest' parameter attributes if the function | Duncan Sands | 2008-02-16 | 2 | -10/+109 | |
| | | | | | | is not passed as an argument to a trampoline intrinsic. llvm-svn: 47220 | |||||
* | Some micro-optimizations. | Duncan Sands | 2008-02-16 | 1 | -4/+5 | |
| | | | | llvm-svn: 47219 | |||||
* | I cannot find a libgcc function for this builtin. Therefor expanding it to ↵ | Andrew Lenharth | 2008-02-16 | 9 | -9/+33 | |
| | | | | | | a noop (which is how it use to be treated). If someone who knows the x86 backend better than me could tell me how to get a lock prefix on an instruction, that would be nice to complete x86 support. llvm-svn: 47213 | |||||
* | Teach LegalizeTypes how to expand the operands of | Duncan Sands | 2008-02-16 | 2 | -4/+25 | |
| | | | | | | br_cc. This fixes 5 "make check" failures. llvm-svn: 47212 | |||||
* | Try to clarify which compilers can be used for the | Duncan Sands | 2008-02-16 | 1 | -6/+7 | |
| | | | | | | build. llvm-svn: 47211 | |||||
* | Refactor some code; check if commuteInstruction is able to commute the ↵ | Evan Cheng | 2008-02-16 | 2 | -17/+34 | |
| | | | | | | instruction. llvm-svn: 47208 | |||||
* | llvm.memory.barrier, and impl for x86 and alpha | Andrew Lenharth | 2008-02-16 | 17 | -1/+256 | |
| | | | | llvm-svn: 47204 | |||||
* | Fix typos. | Bill Wendling | 2008-02-16 | 1 | -2/+2 | |
| | | | | llvm-svn: 47200 | |||||
* | Rename CountMemOperands to ComputeMemOperandsEnd to reflect what | Dan Gohman | 2008-02-16 | 2 | -17/+15 | |
| | | | | | | | it actually does. Simplify CountOperands a little by reusing ComputeMemOperandsEnd. And reword some comments for both. llvm-svn: 47198 | |||||
* | Revert 47177, which was incorrect. | Dan Gohman | 2008-02-16 | 1 | -1/+1 | |
| | | | | llvm-svn: 47196 | |||||
* | upgrade this test, which wasn't testing the right thing since llvm-upgrade ↵ | Chris Lattner | 2008-02-15 | 1 | -7/+7 | |
| | | | | | | came around. llvm-svn: 47194 | |||||
* | rename llx -> ll | Chris Lattner | 2008-02-15 | 1 | -0/+0 | |
| | | | | llvm-svn: 47192 | |||||
* | This test is not interesting. | Evan Cheng | 2008-02-15 | 1 | -21/+0 | |
| | | | | llvm-svn: 47189 | |||||
* | Make tblgen a little smarter about constants smaller than i32. Currently, | Scott Michel | 2008-02-15 | 4 | -8/+29 | |
| | | | | | | | | | | | | | | | | | | | | | | | | tblgen will complain if a sign-extended constant does not fit into a data type smaller than i32, e.g., i16. This causes a problem when certain hex constants are used, such as 0xff for byte masks or immediate xor values. tblgen will try the sign-extended value first and, if the sign extended value would overflow, it tries to see if the unsigned value will fit. Consequently, a software developer can now safely incant: (XORHIr16 R16C:$rA, 0xffff) which is somewhat clearer and more informative than incanting: (XORHIr16 R16C:$rA, (i16 -1)) even if the two are bitwise equivalent. Tblgen also outputs the 64-bit unsigned constant in the generated ISel code when getTargetConstant() is invoked. llvm-svn: 47188 | |||||
* | simplify this. | Chris Lattner | 2008-02-15 | 1 | -3/+1 | |
| | | | | llvm-svn: 47187 | |||||
* | Modify test to expect improved code. | Dale Johannesen | 2008-02-15 | 1 | -4/+4 | |
| | | | | llvm-svn: 47182 | |||||
* | The copy instruction being coalesced will be removed, it is not a kill. | Evan Cheng | 2008-02-15 | 1 | -2/+2 | |
| | | | | llvm-svn: 47179 | |||||
* | Fixed bug in FoldingSetIteratorImpl where we did not correctly check if | Ted Kremenek | 2008-02-15 | 1 | -2/+4 | |
| | | | | | | | we had reached the "fake bucket" after the last bucket, allowing the iterator in some cases to run off the end of the hashtable. llvm-svn: 47178 | |||||
* | Skip over the defs and start at the uses when looking for operands | Dan Gohman | 2008-02-15 | 1 | -1/+1 | |
| | | | | | | with the TIED_TO attribute. llvm-svn: 47177 | |||||
* | Use the TargetInstrDescr to determine the number of operands | Dan Gohman | 2008-02-15 | 1 | -3/+3 | |
| | | | | | | | that should be checked for the TIED_TO attribute instead of using CountOperands. llvm-svn: 47176 | |||||
* | Fix this test on linux, which returns S242 | Duncan Sands | 2008-02-15 | 1 | -1/+2 | |
| | | | | | | using sret. llvm-svn: 47173 | |||||
* | Teach LegalizeTypes how to promote the flags | Duncan Sands | 2008-02-15 | 2 | -7/+34 | |
| | | | | | | | | | in a ret node. These are created as i32 constants but on some platforms i32 is not legal. This fixes 26 "make check" failures, for example Alpha/2005-07-12-TwoMallocCalls.ll. llvm-svn: 47172 | |||||
* | Handle \n's in value names for more targets. The asm printers | Chris Lattner | 2008-02-15 | 3 | -13/+40 | |
| | | | | | | really really really need refactoring :( llvm-svn: 47171 | |||||
* | If the llvm name contains an unprintable character, don't print it in | Chris Lattner | 2008-02-15 | 1 | -3/+15 | |
| | | | | | | | | | | | the global comment. This prevents printing things like: ... # foo bar when the name is "foo\nbar". llvm-svn: 47170 | |||||
* | targets that support quotes for mangled names still need to escape newlines | Chris Lattner | 2008-02-15 | 1 | -0/+2 | |
| | | | | | | when they occur in the name, just like " is escaped. llvm-svn: 47169 | |||||
* | Cosmetics. | Dale Johannesen | 2008-02-15 | 1 | -6/+3 | |
| | | | | llvm-svn: 47168 | |||||
* | - Removing the infamous r2rMap_ and rep() method. Now the coalescer will update | Evan Cheng | 2008-02-15 | 6 | -402/+327 | |
| | | | | | | | register defs and uses after each successful coalescing. - Also removed a number of hacks and fixed some subtle kill information bugs. llvm-svn: 47167 | |||||
* | Added CommuteChangesDestination(). This returns true if commuting the specified | Evan Cheng | 2008-02-15 | 2 | -1/+31 | |
| | | | | | | machine instr will change its definition register. llvm-svn: 47166 | |||||
* | Remove unnecessary #include. | Evan Cheng | 2008-02-15 | 1 | -1/+0 | |
| | | | | llvm-svn: 47164 | |||||
* | Use StoreSDNode::getValue instead of calling getOperand directly | Dan Gohman | 2008-02-15 | 1 | -4/+4 | |
| | | | | | | with a hard-coded operand number. llvm-svn: 47163 | |||||
* | Remove warning about 64-bit code on processor | Dale Johannesen | 2008-02-15 | 2 | -10/+0 | |
| | | | | | | that doesn't support it. Per Chris. llvm-svn: 47162 | |||||
* | It is not safe to call fork in PrintStackTrace. Sometimes it freezes the ↵ | Lauro Ramos Venancio | 2008-02-15 | 1 | -44/+2 | |
| | | | | | | program. llvm-svn: 47161 | |||||
* | Change MemOperand's size and offset to be 64-bit. | Dan Gohman | 2008-02-15 | 1 | -5/+6 | |
| | | | | llvm-svn: 47151 | |||||
* | Fix PR2028 | Devang Patel | 2008-02-15 | 2 | -2/+7 | |
| | | | | llvm-svn: 47150 | |||||
* | nocona, core2 and penryn support 64 bit. | Dale Johannesen | 2008-02-15 | 1 | -3/+3 | |
| | | | | llvm-svn: 47149 | |||||
* | Support vector constant zeros, thanks to Zack Rusin for the testcase. | Chris Lattner | 2008-02-15 | 1 | -3/+3 | |
| | | | | llvm-svn: 47148 | |||||
* | Rewrite tblgen handling of subtarget features so | Dale Johannesen | 2008-02-14 | 5 | -2/+10 | |
| | | | | | | | | | | it follows the order of the enum, not alphabetical. The motivation is to make -mattr=+ssse3,+sse41 select SSE41 as it ought to. Added "ignored" enum values of 0 to PPC and SPU to avoid compiler warnings. llvm-svn: 47143 | |||||
* | If loop header is also loop exiting block then OrigPN is incoming value for ↵ | Devang Patel | 2008-02-14 | 2 | -1/+471 | |
| | | | | | | | | B loop header. Fixes PR 2030. llvm-svn: 47141 |