Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Add #ifdef hack for MAP_ANONYMOUS being spelled MAP_ANON on some platforms. | Brian Gaeke | 2003-06-17 | 1 | -0/+3 |
| | | | | | | | (We're already talking about autoconf'ing this, so I'm assuming this hack will be short-lived...I just don't want it to get lost in my working files.) llvm-svn: 6761 | ||||
* | Use $(PLATFORMLIBDL) to selectively bring in -ldl only on those platforms where | Brian Gaeke | 2003-06-17 | 1 | -1/+1 |
| | | | | | | it is needed. llvm-svn: 6753 | ||||
* | Use std::isnan instead of isnan. Brought back to you from the | Brian Gaeke | 2003-06-17 | 1 | -1/+1 |
| | | | | | | future, by the reconciliation of the C++ and C99 standards. Someday. llvm-svn: 6751 | ||||
* | Life is too short. Link in too much stuff on Linux to make building on sun ↵ | Chris Lattner | 2003-06-17 | 1 | -4/+3 |
| | | | | | | easier llvm-svn: 6748 | ||||
* | The never-ending odyssey trying to get sparc to link | Chris Lattner | 2003-06-17 | 1 | -1/+1 |
| | | | | llvm-svn: 6747 | ||||
* | Make sure to get the value of ARCH before we use it | Chris Lattner | 2003-06-17 | 1 | -0/+5 |
| | | | | llvm-svn: 6746 | ||||
* | Do not link in the Sparc JIT when building on X86. Eventually the sparc ↵ | Chris Lattner | 2003-06-17 | 1 | -6/+6 |
| | | | | | | | | | | will not link in the X86 JIT either, but this makes testing easier. For some reason, the sparc JIT was breaking the X86 JIT when it was linked in. :( llvm-svn: 6745 | ||||
* | Apparently "sparc" is a macro on sparcs. Ugh. :) | Chris Lattner | 2003-06-17 | 1 | -4/+4 |
| | | | | llvm-svn: 6744 | ||||
* | Whoops, didn't mean to check that in :( | Chris Lattner | 2003-06-17 | 1 | -3/+19 |
| | | | | llvm-svn: 6743 | ||||
* | Use more structured command line option processing | Chris Lattner | 2003-06-17 | 2 | -41/+31 |
| | | | | llvm-svn: 6742 | ||||
* | #ifdef out code that only applies when the HOSTARCH = sparc | Chris Lattner | 2003-06-17 | 1 | -1/+4 |
| | | | | llvm-svn: 6741 | ||||
* | Add #include for older GCC's | Chris Lattner | 2003-06-08 | 1 | -0/+1 |
| | | | | llvm-svn: 6670 | ||||
* | ::: HACK ALERT ::: HACK ALERT ::: HACK ALERT ::: HACK ALERT ::: HACK ALERT ::: | Misha Brukman | 2003-06-06 | 1 | -1/+14 |
| | | | | | | | | | | | | | | | | | The JIT is designed to code-generate a function at-a-time. That means that any pass can only make local changes to its function. Period. Because the Sparc PreSelection pass claims to be a BasicBlock pass while adding globals to the Module, it cannot be run with the other passes, because by this time, the globals have been output already by the JIT, and the addresses of any globals appearing AFTER this point are not recognized. However, the PreSelection pass is a requirement for correctness in the Sparc codegen path, so it MUST be run. ::: HACK ALERT ::: HACK ALERT ::: HACK ALERT ::: HACK ALERT ::: HACK ALERT ::: llvm-svn: 6650 | ||||
* | Output function address as hex. | Misha Brukman | 2003-06-06 | 1 | -3/+2 |
| | | | | llvm-svn: 6649 | ||||
* | Removed debug print statement. | Misha Brukman | 2003-06-06 | 1 | -1/+0 |
| | | | | llvm-svn: 6641 | ||||
* | * Institute a hack for the Sparc call to mmap() to get our generated code to be | Misha Brukman | 2003-06-04 | 1 | -10/+16 |
| | | | | | | | | | | | laid out closer to the VM so that calls to library functions (e.g. puts()) and callback (e.g. JITResolver::CompilationCallback) fit into 30 bits of the call instruction. * Abort if architecture is not yet supported (not X86 or Sparc) because it likely requires a different set of parameters to mmap() . * Stop using hard-coded values for page size; use sysconf(_SC_PAGESIZE) instead. llvm-svn: 6610 | ||||
* | Sparc's dlsym() requires the special operand RTLD_SELF to find a symbol in the | Misha Brukman | 2003-06-04 | 1 | -0/+4 |
| | | | | | | currently-running process. llvm-svn: 6592 | ||||
* | * Removed SparcEmitter.cpp; rolled into lib/Target/Sparc/SparcV9CodeEmitter.cpp | Misha Brukman | 2003-06-02 | 4 | -260/+23 |
| | | | | | | | | | * No more createX86Emitter() vs. createSparcEmitter() -- there can be only one * As a result, the memory management semantics must be handled according to platform -- the parameters to mmap() are particularly sensitive to the host architecture. llvm-svn: 6527 | ||||
* | The flag modifications weren't picking up the old values of the | Brian Gaeke | 2003-06-02 | 1 | -5/+10 |
| | | | | | | | flags before. Save them in a temporary variable, then restore them from the temporary after creating the new constant. llvm-svn: 6520 | ||||
* | Remove obsolete code | Chris Lattner | 2003-06-02 | 1 | -51/+0 |
| | | | | llvm-svn: 6518 | ||||
* | Move target specific code to target files. The new MachineCodeEmitter | Chris Lattner | 2003-06-01 | 5 | -171/+76 |
| | | | | | | class is actually target independent! llvm-svn: 6517 | ||||
* | Fix induction variable name clash in for loops, in finishFunction(). | Brian Gaeke | 2003-06-01 | 1 | -3/+8 |
| | | | | | | | Modify new MachineOperand so that its flags match the old MachineOperand's flags, for the flags that matter. llvm-svn: 6513 | ||||
* | Fixed rewriting of branches -- they now work forward and backward. | Misha Brukman | 2003-05-31 | 1 | -4/+9 |
| | | | | llvm-svn: 6463 | ||||
* | Since malloc is no longer used, no need to free() memory. | Misha Brukman | 2003-05-30 | 1 | -15/+11 |
| | | | | | | Fixed BasicBlock patching by supplying correct type for the displacement. llvm-svn: 6453 | ||||
* | Fix call to mmap, so that it can be used on sparc. | Brian Gaeke | 2003-05-30 | 1 | -10/+5 |
| | | | | llvm-svn: 6424 | ||||
* | mmap() seems to be failing on Sparc, so just use malloc()/free() . | Misha Brukman | 2003-05-28 | 1 | -4/+29 |
| | | | | llvm-svn: 6387 | ||||
* | Link in Sparc libs for the JIT, even on X86 to be able to support debugging | Misha Brukman | 2003-05-27 | 1 | -3/+19 |
| | | | | | | of Sparc JIT (printing out instrs) on X86. Con: this increases linking time. llvm-svn: 6361 | ||||
* | Allow for specification of which JIT to run on the commandline. | Misha Brukman | 2003-05-27 | 4 | -13/+269 |
| | | | | | | | | | | | | | | | | `lli -march=x86' or `lli -march=sparc' will forcefully select the JIT even on a different platform. Running lli without the -march option will select the JIT for the platform that it's currently running on. Pro: can test Sparc JIT (debug printing mode) on X86 -- faster to compile/link LLVM source base to test changes. Con: Linking lli on x86 now pulls in all the Sparc libs -> longer link time (but X86 can bear it, right?) In the future, perhaps this should be a ./configure option to enable/disable target JITting... llvm-svn: 6360 | ||||
* | Fix typeos | Chris Lattner | 2003-05-14 | 1 | -4/+4 |
| | | | | llvm-svn: 6204 | ||||
* | Add support for more constant expressions | Chris Lattner | 2003-05-14 | 1 | -6/+35 |
| | | | | llvm-svn: 6203 | ||||
* | Add support for atexit function, remove support for __main function | Chris Lattner | 2003-05-14 | 5 | -14/+37 |
| | | | | llvm-svn: 6194 | ||||
* | Add support for atexit handlers to the JIT, fixing 2003-05-14-AtExit.c | Chris Lattner | 2003-05-14 | 3 | -4/+28 |
| | | | | llvm-svn: 6193 | ||||
* | Fix compilation problems with previous checking *blush* | Chris Lattner | 2003-05-14 | 1 | -3/+4 |
| | | | | llvm-svn: 6191 | ||||
* | Add a framework for intercepting system calls | Chris Lattner | 2003-05-14 | 2 | -21/+54 |
| | | | | llvm-svn: 6190 | ||||
* | Clean up cast | Chris Lattner | 2003-05-13 | 1 | -1/+1 |
| | | | | llvm-svn: 6174 | ||||
* | Make sure that globals are emitted AFTER the passmanager is set up for the JIT, | Chris Lattner | 2003-05-12 | 2 | -0/+2 |
| | | | | | | because the globals may refer to functions that need to be compiled! llvm-svn: 6105 | ||||
* | switch main LLI core execution to use an InstVisitor instead of a switch ↵ | Chris Lattner | 2003-05-10 | 2 | -67/+61 |
| | | | | | | statement llvm-svn: 6081 | ||||
* | Fix testcase: SingleSource/UnitTests/2003-05-02-DependantPHI.c | Chris Lattner | 2003-05-10 | 2 | -34/+55 |
| | | | | llvm-svn: 6074 | ||||
* | Add support for function stubs, which allow calling functions which need to | Chris Lattner | 2003-05-09 | 4 | -8/+73 |
| | | | | | | have an address available, but have not yet been code generated. llvm-svn: 6059 | ||||
* | Minor speedup by avoiding callbacks to functions already generated | Chris Lattner | 2003-05-08 | 2 | -6/+9 |
| | | | | llvm-svn: 6052 | ||||
* | Improve efficiency of JIT by having it use direct function calls instead of | Chris Lattner | 2003-05-08 | 3 | -36/+29 |
| | | | | | | signals to regain control from the executing code llvm-svn: 6051 | ||||
* | assert early instead of late for unimplemented feature | Chris Lattner | 2003-05-08 | 1 | -3/+10 |
| | | | | llvm-svn: 6050 | ||||
* | Implement varargs support for LLI! | Chris Lattner | 2003-05-08 | 2 | -73/+56 |
| | | | | llvm-svn: 6043 | ||||
* | Add a pointersize/endianness safe load routine to match the store routine | Chris Lattner | 2003-05-08 | 1 | -0/+76 |
| | | | | llvm-svn: 6042 | ||||
* | A large number of simple changes: | Chris Lattner | 2003-05-08 | 7 | -178/+95 |
| | | | | | | | * s/Method/Function * Kill some obsolete (external) functions that used to be to support tracing llvm-svn: 6041 | ||||
* | Add support for recording arguments passed through the ... of a varargs function | Chris Lattner | 2003-05-08 | 2 | -10/+17 |
| | | | | llvm-svn: 6040 | ||||
* | Fix bug: Jello/2003-05-07-ArgumentTest.llx | Chris Lattner | 2003-05-07 | 1 | -17/+33 |
| | | | | | | Also fixes yacr2/ks benchmarks llvm-svn: 6017 | ||||
* | Remove two fields from TargetData which are target specific. | Chris Lattner | 2003-04-26 | 1 | -1/+0 |
| | | | | llvm-svn: 5963 | ||||
* | Only do the %ld -> %lld promotion when running a 64 bit bytecode on a 32 bit ↵ | Chris Lattner | 2003-04-25 | 1 | -1/+2 |
| | | | | | | host llvm-svn: 5942 | ||||
* | Add __strdup | Chris Lattner | 2003-04-25 | 1 | -0/+7 |
| | | | | llvm-svn: 5941 |