| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | this final call to canLosslesslyBitCastTo is dead, because ValueRequiresCast | Chris Lattner | 2007-01-06 | 1 | -4/+0 | 
| | | | | | | | is only called on integers. llvm-svn: 32949 | ||||
| * | simplify some more code now that there are not multiple different integer | Chris Lattner | 2007-01-06 | 1 | -8/+4 | 
| | | | | | | | types of the same size llvm-svn: 32948 | ||||
| * | eliminate some uses of canLosslesslyBitCastTo, this actually makes the code ↵ | Chris Lattner | 2007-01-06 | 1 | -15/+14 | 
| | | | | | | | | | stronger, by nuking relational pointer comparisons with casts. llvm-svn: 32947 | ||||
| * | no need to worry about int vs uint any more. | Chris Lattner | 2007-01-06 | 1 | -2/+1 | 
| | | | | | llvm-svn: 32946 | ||||
| * | new note | Chris Lattner | 2007-01-06 | 1 | -0/+5 | 
| | | | | | llvm-svn: 32945 | ||||
| * | setSetCCIsExpensive is gone. | Evan Cheng | 2007-01-05 | 1 | -2/+0 | 
| | | | | | llvm-svn: 32941 | ||||
| * | Expand fcopysign to the bitwise sequence if select is marked as expensive. | Evan Cheng | 2007-01-05 | 1 | -4/+6 | 
| | | | | | llvm-svn: 32940 | ||||
| * | 1) Remove old AnalysisResolver. | Devang Patel | 2007-01-05 | 2 | -16/+9 | 
| | | | | | | | 2) Rename AnalysisResolver_New as AnalysisResolver llvm-svn: 32938 | ||||
| * | Regenerate. | Reid Spencer | 2007-01-05 | 3 | -319/+281 | 
| | | | | | llvm-svn: 32934 | ||||
| * | For PR1077: | Reid Spencer | 2007-01-05 | 1 | -26/+7 | 
| | | | | | | | | | Disallow merging of dupliate global variables. It is now illegal to declare or define two global variables of the same name and same type. llvm-gcc3 is dead in 2.0 and llvm-gcc4 doesn't have that problem nor need the hack. llvm-svn: 32933 | ||||
| * | - FCOPYSIGN custom lowering bug. Clear the sign bit of operand 0 first before | Evan Cheng | 2007-01-05 | 3 | -21/+39 | 
| | | | | | | | | | or'ing in the sign bit of operand 1. - Tweaking: rather than left shift the sign bit, fp_extend operand 1 first before taking its sign bit if its type is smaller than that of operand 0. llvm-svn: 32932 | ||||
| * | Bug in ExpandFCOPYSIGNToBitwiseOps(). Clear the old sign bit of operand 0 | Evan Cheng | 2007-01-05 | 1 | -4/+14 | 
| | | | | | | | before or'ing in the sign bit of operand 1. llvm-svn: 32930 | ||||
| * | CopyToReg source operand can be a register as well. e.g. Copy from ↵ | Evan Cheng | 2007-01-05 | 1 | -2/+6 | 
| | | | | | | | GlobalBaseReg. llvm-svn: 32929 | ||||
| * | Remove PassManagerT.h | Devang Patel | 2007-01-05 | 1 | -0/+0 | 
| | | | | | llvm-svn: 32928 | ||||
| * | Remove old pass manager. | Devang Patel | 2007-01-05 | 3 | -1099/+1 | 
| | | | | | llvm-svn: 32927 | ||||
| * | Regenerate. | Reid Spencer | 2007-01-05 | 3 | -1170/+1159 | 
| | | | | | llvm-svn: 32905 | ||||
| * | Change the syntax for parameter attributes: | Reid Spencer | 2007-01-05 | 3 | -72/+79 | 
| | | | | | | | | | | | | | | | | | | | | 1. The @ sign is no longer necessary. 2. We now support "function attributes" as parameter attribute 0. 3. Instead of locating the return type attributes after the type of a function result, they are now located after the function header's closing paranthesis and before any alignment or section options. 4. The way has been prepared for a new "noreturn" function attribute but there is no support for recognizing it in the lexer nor doing anything with it if it does get set. 5. The FunctionType::getParamAttrsText method now has support for returning multiple attributes. This required a change in its interface. I'm unhappy that this change leads to 6 new shift/reduce conflicts, but in each case bison's decision to choose the shift is correct so there shouldn't be any damage from these conflicts. llvm-svn: 32904 | ||||
| * | Typo | Evan Cheng | 2007-01-05 | 1 | -1/+1 | 
| | | | | | llvm-svn: 32902 | ||||
| * | With SSE2, expand FCOPYSIGN to a series of SSE bitwise operations. | Evan Cheng | 2007-01-05 | 3 | -8/+93 | 
| | | | | | llvm-svn: 32900 | ||||
| * | Implement InstCombine/vec_shuffle.ll:%test7, simplifying shuffles with | Chris Lattner | 2007-01-05 | 1 | -1/+23 | 
| | | | | | | | undef operands. llvm-svn: 32899 | ||||
| * | fold things like a^b != c^a -> b != c. This implements ↵ | Chris Lattner | 2007-01-05 | 1 | -12/+33 | 
| | | | | | | | InstCombine/xor.ll:test27 llvm-svn: 32893 | ||||
| * | Compile X + ~X to -1. This implements Instcombine/add.ll:test34 | Chris Lattner | 2007-01-05 | 1 | -1/+7 | 
| | | | | | llvm-svn: 32890 | ||||
| * | GEP subscript is interpreted as a signed value. | Evan Cheng | 2007-01-05 | 1 | -1/+1 | 
| | | | | | llvm-svn: 32888 | ||||
| * | fix PowerPC/2007-01-04-ArgExtension.ll, a bug handling K&R prototypes with | Chris Lattner | 2007-01-04 | 1 | -4/+6 | 
| | | | | | | | the recent signless changes. llvm-svn: 32884 | ||||
| * | Expand fcopysign to a series of bitwise of operations when it's profitable to | Evan Cheng | 2007-01-04 | 1 | -19/+46 | 
| | | | | | | | do so. llvm-svn: 32881 | ||||
| * | Expand SELECT (f32/f64) and FCOPYSIGN (f32/f64). | Lauro Ramos Venancio | 2007-01-04 | 1 | -0/+5 | 
| | | | | | llvm-svn: 32870 | ||||
| * | Death to useless bitcast instructions! | Reid Spencer | 2007-01-04 | 1 | -33/+3 | 
| | | | | | llvm-svn: 32866 | ||||
| * | Do not allow packed types for icmp and fcmp instructions. | Reid Spencer | 2007-01-04 | 1 | -5/+2 | 
| | | | | | llvm-svn: 32865 | ||||
| * | Regenerate. | Reid Spencer | 2007-01-04 | 2 | -27/+35 | 
| | | | | | llvm-svn: 32862 | ||||
| * | Disallow packed types in icmp/fcmp instructions. The code generator is | Reid Spencer | 2007-01-04 | 1 | -0/+4 | 
| | | | | | | | not prepared to handle them yet. llvm-svn: 32861 | ||||
| * | Now that setcondinst has been eliminated, we can mark Value::SubclassID | Chris Lattner | 2007-01-04 | 1 | -4/+0 | 
| | | | | | | | const and remove the ugly mutator methods. llvm-svn: 32860 | ||||
| * | fix typo | Chris Lattner | 2007-01-04 | 1 | -1/+1 | 
| | | | | | llvm-svn: 32859 | ||||
| * | Enable a couple xforms for packed vectors (undef | v) -> -1 for packed. | Chris Lattner | 2007-01-04 | 1 | -8/+9 | 
| | | | | | llvm-svn: 32858 | ||||
| * | fix some bugs handling vectors, avoid host-specific handling of undefined ↵ | Chris Lattner | 2007-01-04 | 1 | -5/+15 | 
| | | | | | | | shift results. llvm-svn: 32857 | ||||
| * | Add a new ConstantPacked::getAllOnesValue method | Chris Lattner | 2007-01-04 | 1 | -0/+12 | 
| | | | | | llvm-svn: 32856 | ||||
| * | Regenerate. | Reid Spencer | 2007-01-04 | 2 | -155/+149 | 
| | | | | | llvm-svn: 32853 | ||||
| * | Permit icmp and fcmp to have packed operands. | Reid Spencer | 2007-01-04 | 1 | -5/+2 | 
| | | | | | | | Make an error message a little more useful. llvm-svn: 32852 | ||||
| * | The previous implementation of LLVM Streams wasn't removing symbols. This | Bill Wendling | 2007-01-03 | 2 | -2/+2 | 
| | | | | | | | one should. llvm-svn: 32845 | ||||
| * | fix testcase. It's not safe to strictly evaluate a load that should be lazy. | Chris Lattner | 2007-01-03 | 1 | -1/+2 | 
| | | | | | llvm-svn: 32842 | ||||
| * | Private labels start with .L on linux, not just . | Chris Lattner | 2007-01-03 | 1 | -1/+1 | 
| | | | | | llvm-svn: 32841 | ||||
| * | Fix a comment that referred to the now defunct ubyte type. | Reid Spencer | 2007-01-03 | 1 | -1/+1 | 
| | | | | | llvm-svn: 32840 | ||||
| * | Remove two useless bit casts. | Reid Spencer | 2007-01-03 | 1 | -6/+0 | 
| | | | | | llvm-svn: 32839 | ||||
| * | Legalizer doesn't do an ANY_EXTEND if we don't ask for one so make sure | Reid Spencer | 2007-01-03 | 1 | -1/+1 | 
| | | | | | | | | that we default to an ANY_EXTEND if no parameter attribute is set on the result value of a function. llvm-svn: 32836 | ||||
| * | NULL names should pass validation. | Jim Laskey | 2007-01-03 | 1 | -1/+2 | 
| | | | | | llvm-svn: 32835 | ||||
| * | Silence warning. | Jim Laskey | 2007-01-03 | 1 | -1/+1 | 
| | | | | | llvm-svn: 32834 | ||||
| * | Really big cleanup. | Anton Korobeynikov | 2007-01-03 | 10 | -24/+33 | 
| | | | | | | | | | | | | - New target type "mingw" was introduced - Same things for both mingw & cygwin are marked as "cygming" (as in gcc) - .lcomm is supported here, so allow LLVM to use it - Correctly use underscored versions of setjmp & _longjmp for both mingw & cygwin llvm-svn: 32833 | ||||
| * | Restore previous behavior of defaulting to ZEXT. This works around two | Reid Spencer | 2007-01-03 | 1 | -7/+2 | 
| | | | | | | | | | things: (1) preventing PR1071 and (2) working around missing parameter attributes for bool type. (2) will be fixed shortly. When PR1071 is fixed, this patch should be undone. llvm-svn: 32831 | ||||
| * | Two changes: | Reid Spencer | 2007-01-03 | 1 | -5/+12 | 
| | | | | | | | | | | | | 1. Switch expression and cases are compared signed and are sign extended. 2. For function results needing extended, do SIGN_EXTEND if the SExtAttribute is set and ZERO_EXTEND if the ZExtAttribute is set, otherwise just let the Legalizer do ANY_EXTEND. This fixes the recent regression in kimwitu++ and probably the llvm-gcc bootstrap issue we had today. llvm-svn: 32830 | ||||
| * | Clean up from recent changes. Comment the new parameter to ExpandLibCall. | Reid Spencer | 2007-01-03 | 1 | -24/+20 | 
| | | | | | | | Consolidate some lines of code and remove duplication. llvm-svn: 32829 | ||||
| * | Vectors are not supported by ConstantInt::getAllOnesValue. | Jim Laskey | 2007-01-03 | 1 | -5/+10 | 
| | | | | | llvm-svn: 32827 | ||||

