Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | Implement CodeGen/CBackend/2005-02-14-VolatileOperations.ll | Chris Lattner | 2005-02-14 | 1 | -0/+15 | |
| | | | | | | Volatile loads and stores need to emit volatile pointer operations in C. llvm-svn: 20177 | |||||
* | Fix hyphenation in output comment | Misha Brukman | 2005-01-31 | 1 | -1/+1 | |
| | | | | llvm-svn: 19954 | |||||
* | Fix alloca support for Cygwin. On cygwin its __alloca not __builtin_alloca | Reid Spencer | 2005-01-23 | 1 | -1/+4 | |
| | | | | llvm-svn: 19776 | |||||
* | Fix CBE code so that it compiles with VC++. | Jeff Cohen | 2005-01-06 | 1 | -1/+1 | |
| | | | | llvm-svn: 19303 | |||||
* | Fix PR490 | Chris Lattner | 2004-12-29 | 1 | -0/+6 | |
| | | | | | | Fix testcase CodeGen/CBackend/2004-12-28-LogicalConstantExprs.ll llvm-svn: 19176 | |||||
* | Fix PR485, instead of emitting zero sized arrays, emit arrays of size 1. | Chris Lattner | 2004-12-15 | 1 | -0/+1 | |
| | | | | llvm-svn: 18974 | |||||
* | When generating code for X86 targets, make sure the fp control word is set | Chris Lattner | 2004-12-13 | 1 | -0/+13 | |
| | | | | | | to 64-bit precision, not 80 bits. llvm-svn: 18915 | |||||
* | Emit correct prototype for __builtin_alloca on V8. | Brian Gaeke | 2004-12-10 | 1 | -1/+8 | |
| | | | | llvm-svn: 18745 | |||||
* | Move lower intrinsics before FP constant emission, in case | Chris Lattner | 2004-12-05 | 1 | -2/+7 | |
| | | | | | | | | | intrinsic lowering ever introduces constants. Rename local symbols before printing function bodies, fixing 255.vortex with the CBE!!! llvm-svn: 18534 | |||||
* | Fix test/Regression/CodeGen/CBackend/2004-12-03-ExternStatics.ll and | Chris Lattner | 2004-12-03 | 1 | -1/+4 | |
| | | | | | | PR472 llvm-svn: 18459 | |||||
* | Reverting revision 1.209. | John Criswell | 2004-12-02 | 1 | -1/+1 | |
| | | | | | | | Including alloca.h on Solaris brings in the prototype of strftime(), which breaks compilation of CBE generated code. llvm-svn: 18435 | |||||
* | Do not let GCC emit a warning for INT64_MIN | Chris Lattner | 2004-11-30 | 1 | -2/+5 | |
| | | | | llvm-svn: 18398 | |||||
* | Sparcs behave better if we use <alloca.h> and avoid messing with ↵ | Brian Gaeke | 2004-11-30 | 1 | -1/+1 | |
| | | | | | | __builtin_alloca. llvm-svn: 18397 | |||||
* | Hack around stupidity in GCC, fixing Burg with the CBE and | Chris Lattner | 2004-11-13 | 1 | -9/+56 | |
| | | | | | | CBackend/2004-11-13-FunctionPointerCast.llx llvm-svn: 17710 | |||||
* | Change Library Names Not To Conflict With Others When Installed | Reid Spencer | 2004-10-27 | 1 | -1/+1 | |
| | | | | llvm-svn: 17286 | |||||
* | Removed dead method, printPHICopiesForSuccessors(). | John Criswell | 2004-10-25 | 1 | -21/+0 | |
| | | | | llvm-svn: 17216 | |||||
* | Modified switch generation so that only the phi values associated with the | John Criswell | 2004-10-25 | 1 | -1/+2 | |
| | | | | | | destination basic block are copied. llvm-svn: 17212 | |||||
* | We won't use automake | Reid Spencer | 2004-10-22 | 2 | -673/+0 | |
| | | | | llvm-svn: 17155 | |||||
* | Small performance improvement in generated C code: | John Criswell | 2004-10-20 | 1 | -1/+23 | |
| | | | | | | | | | | | | Instead of unconditionally copying all phi node values into temporaries for all successor blocks, generate code that will determine what successor block will be called and then copy only those phi node values needed by the successor block. This seems to cut down namd execution time from being 8% higher than GCC to 4% higher than GCC. llvm-svn: 17144 | |||||
* | Initial automake generated Makefile template | Reid Spencer | 2004-10-18 | 1 | -0/+657 | |
| | | | | llvm-svn: 17136 | |||||
* | Print a semicolon for the unreacahble instruction. This fixes problems | Chris Lattner | 2004-10-17 | 1 | -1/+1 | |
| | | | | | | where C requires semicolons in some cases to indicate null statements. llvm-svn: 17107 | |||||
* | The first hunk corrects a bug when printing undef null values. We would print | Chris Lattner | 2004-10-17 | 1 | -6/+11 | |
| | | | | | | | | 0->field, which is illegal. Now we print ((foo*)0)->field. The second hunk is an optimization to not print undefined phi values. llvm-svn: 17094 | |||||
* | Add support for unreachable and undef | Chris Lattner | 2004-10-16 | 1 | -2/+10 | |
| | | | | llvm-svn: 17048 | |||||
* | Update to reflect changes in Makefile rules. | Reid Spencer | 2004-10-13 | 1 | -4/+5 | |
| | | | | llvm-svn: 16950 | |||||
* | Fix a warning that is emitted on the suns | Chris Lattner | 2004-10-11 | 1 | -1/+1 | |
| | | | | llvm-svn: 16917 | |||||
* | Initial version of automake Makefile.am file. | Reid Spencer | 2004-10-10 | 1 | -0/+15 | |
| | | | | llvm-svn: 16893 | |||||
* | Really fix FreeBSD, which apparently doesn't tolerate the extern. | Chris Lattner | 2004-10-06 | 1 | -2/+3 | |
| | | | | | | Thanks to Jeff Cohen for pointing out my goof. llvm-svn: 16762 | |||||
* | FreeBSD uses GCC. Patch contributed by Jeff Cohen! | Chris Lattner | 2004-10-06 | 1 | -3/+2 | |
| | | | | llvm-svn: 16756 | |||||
* | Pull assignment out of for loop conditional in order for this to | Alkis Evlogimenos | 2004-09-28 | 1 | -2/+2 | |
| | | | | | | compile under windows. Patch contributed by Paolo Invernizzi! llvm-svn: 16534 | |||||
* | 'Pass' should now not be derived from by clients. Instead, they should derive | Chris Lattner | 2004-09-20 | 1 | -3/+3 | |
| | | | | | | | from ModulePass. Instead of implementing Pass::run, then should implement ModulePass::runOnModule. llvm-svn: 16436 | |||||
* | Changes For Bug 352 | Reid Spencer | 2004-09-01 | 1 | -3/+3 | |
| | | | | | | | | Move include/Config and include/Support into include/llvm/Config, include/llvm/ADT and include/llvm/Support. From here on out, all LLVM public header files must be under include/llvm/. llvm-svn: 16137 | |||||
* | Previous checkin broke printf(%a) support for fp constants-- re-fix it. | Brian Gaeke | 2004-08-25 | 1 | -1/+9 | |
| | | | | llvm-svn: 16051 | |||||
* | New version of Bill Wendling's PR33 patch. | Brian Gaeke | 2004-08-25 | 1 | -24/+38 | |
| | | | | llvm-svn: 16050 | |||||
* | 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 | |||||
* | Temporarily disable this code, as it is emitting LLVM_NAN("nan") which ↵ | Chris Lattner | 2004-07-25 | 1 | -2/+2 | |
| | | | | | | | | | | | results in a call to the glibc 'nan' function because the initializer is not a string. This breaks when used in a global initializer. Try compiling this testcase for example: %X = global float <some nan value> llvm-svn: 15223 | |||||
* | Emit NaNs and INFs bit-identically to the bytecode file, if the system has | Brian Gaeke | 2004-07-21 | 1 | -2/+64 | |
| | | | | | | | printf("%a") support. Patch contributed by Bill Wendling. llvm-svn: 15056 | |||||
* | bug 122: | Reid Spencer | 2004-07-18 | 1 | -6/+4 | |
| | | | | | | - Replace ConstantPointerRef usage with GlobalValue usage llvm-svn: 14953 | |||||
* | Patches towards fixing PR341 | Chris Lattner | 2004-07-15 | 1 | -3/+3 | |
| | | | | llvm-svn: 14841 | |||||
* | Apple's MacOS X is another OS which does not provide alloca() via <alloca.h> | Misha Brukman | 2004-07-12 | 1 | -1/+1 | |
| | | | | llvm-svn: 14781 | |||||
* | Delete the allocate*TargetMachine function, which is now dead . | Chris Lattner | 2004-07-11 | 1 | -6/+0 | |
| | | | | llvm-svn: 14755 | |||||
* | Make these format a bit nicer | Chris Lattner | 2004-07-11 | 1 | -1/+1 | |
| | | | | llvm-svn: 14747 | |||||
* | Auto-registrate target | Chris Lattner | 2004-07-11 | 2 | -0/+7 | |
| | | | | llvm-svn: 14745 | |||||
* | Add #include <iostream> since Value.h does not #include it any more. | Reid Spencer | 2004-07-04 | 1 | -1/+2 | |
| | | | | llvm-svn: 14622 | |||||
* | Move the IntrinsicLowering header into the CodeGen directory, as per PR346 | Chris Lattner | 2004-06-20 | 1 | -1/+1 | |
| | | | | llvm-svn: 14266 | |||||
* | Rename Type::PrimitiveID to TypeId and ::getPrimitiveID() to ::getTypeID() | Chris Lattner | 2004-06-17 | 1 | -3/+3 | |
| | | | | llvm-svn: 14201 | |||||
* | <alloca.h> on cygwin pulls in a ton of stuff (macros and function protos) | Chris Lattner | 2004-06-02 | 1 | -1/+1 | |
| | | | | | | | | | that we REALLY don't want in the CBE code. With this fix, the CBE passes all of the MultiSource tests on cygwin that it does on linux. Yaay! llvm-svn: 13975 | |||||
* | Stubs are no longer needed | Chris Lattner | 2004-06-02 | 1 | -5/+0 | |
| | | | | llvm-svn: 13951 | |||||
* | Fix the big regression that has been killing the nightly tester these last | Chris Lattner | 2004-05-28 | 1 | -1/+3 | |
| | | | | | | | | few days. Apparently the old symbol table used to auto rename collisions in the type symbol table and the new one does not. It doesn't really make sense for the new one to do so, so we just make the client do it. llvm-svn: 13877 | |||||
* | Minor changes. Switch to a SymbolTable remove that does not take linear time | Chris Lattner | 2004-05-28 | 1 | -2/+2 | |
| | | | | llvm-svn: 13874 | |||||
* | Fix PR344: the incorrect remove was being used. | Chris Lattner | 2004-05-26 | 1 | -1/+1 | |
| | | | | llvm-svn: 13790 |