| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | 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 | ||||
| * | Fix the JIT when being used from llvm-db | Chris Lattner | 2004-11-30 | 1 | -1/+1 |
| | | | | | llvm-svn: 18391 | ||||
| * | This pass is completely broken. | Chris Lattner | 2004-11-30 | 1 | -11/+4 |
| | | | | | llvm-svn: 18387 | ||||
| * | RevisionNum is read by error(), initialize it early. | Chris Lattner | 2004-11-30 | 1 | -0/+1 |
| | | | | | llvm-svn: 18386 | ||||
| * | Update list of failing benchmarks. | Brian Gaeke | 2004-11-30 | 1 | -1/+1 |
| | | | | | llvm-svn: 18384 | ||||
| * | If we're about to emit something like: | Brian Gaeke | 2004-11-30 | 1 | -13/+25 |
| | | | | | | | | | | | %f0 = fmovs %f0 %f1 = fmovs %f1 then just delete the FpMOVD pseudo-instruction instead. Also, add statistics and debug printouts. llvm-svn: 18383 | ||||
| * | Squelch warning | Chris Lattner | 2004-11-30 | 1 | -1/+1 |
| | | | | | llvm-svn: 18381 | ||||
| * | Fix several bugs in 'op x, imm' handling. Foremost is that we now emit | Chris Lattner | 2004-11-30 | 1 | -14/+10 |
| | | | | | | | | | | | | | | | addi r3, r3, -1 instead of addi r3, r3, 1 for 'sub int X, 1'. Secondarily, this fixes several cases where we could crash given an unsigned constant. And fixes a couple of minor missed optimization cases, such as xor X, ~0U -> not X llvm-svn: 18379 | ||||
| * | Up the compression threshold to 64K so we avoid it for all but the largest | Reid Spencer | 2004-11-30 | 1 | -1/+1 |
| | | | | | | | bytecode files. This should help linking substantially. llvm-svn: 18378 | ||||
| * | Fix test/Regression/Transforms/LICM/2004-09-14-AliasAnalysisInvalidate.llx | Chris Lattner | 2004-11-30 | 1 | -0/+2 |
| | | | | | | | This only fails on darwin or on X86 under valgrind. llvm-svn: 18377 | ||||
| * | Fix CodeGen/PowerPC/2004-11-30-shr-var-crash.ll | Chris Lattner | 2004-11-30 | 1 | -1/+1 |
| | | | | | llvm-svn: 18376 | ||||
| * | Fix test/Regression/CodeGen/PowerPC/2004-11-29-ShrCrash.ll | Chris Lattner | 2004-11-30 | 1 | -1/+3 |
| | | | | | llvm-svn: 18374 | ||||
| * | Fix test/Regression/CodeGen/PowerPC/2004-11-30-shift-crash.ll | Chris Lattner | 2004-11-30 | 1 | -1/+6 |
| | | | | | llvm-svn: 18371 | ||||
| * | Alkis noticed that this variable is dead. Thanks! | Chris Lattner | 2004-11-30 | 1 | -2/+0 |
| | | | | | llvm-svn: 18369 | ||||
| * | Add method | Chris Lattner | 2004-11-30 | 1 | -2/+27 |
| | | | | | llvm-svn: 18368 | ||||
| * | If we have something like this: | Chris Lattner | 2004-11-30 | 1 | -0/+105 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | if (x) { code ... } else { code ... } Turn it into: code if (x) { ... } else { ... } This reduces code size and in some common cases allows us to completely eliminate the conditional. This turns several if/then/else blocks in loops into straightline code in 179.art, turning the loops into single basic blocks (good for modsched even!). Maybe now brg will leave me alone ;-) llvm-svn: 18366 | ||||
| * | Remove extraneous namespacification. In particular, don't define ↵ | Chris Lattner | 2004-11-30 | 1 | -11/+9 |
| | | | | | | | llvm::llvm::createInternalGlobalMapperPass llvm-svn: 18365 | ||||
| * | Allow hoisting loads of globals and alloca's in conditionals. | Chris Lattner | 2004-11-29 | 1 | -0/+6 |
| | | | | | llvm-svn: 18363 | ||||
| * | Functionality moved to portable lib/System/DynamicLibrary.cpp | Reid Spencer | 2004-11-29 | 1 | -96/+0 |
| | | | | | | | implementation llvm-svn: 18358 | ||||
| * | Use System/DynamicLibrary instead of Support/DynamicLinker | Reid Spencer | 2004-11-29 | 4 | -12/+20 |
| | | | | | llvm-svn: 18357 | ||||
| * | Use System/DynamicLibrary instead of Support/DynamicLinker to implement. | Reid Spencer | 2004-11-29 | 1 | -2/+12 |
| | | | | | llvm-svn: 18356 | ||||
| * | Implement two new functions: LoadLibraryPermanently and | Reid Spencer | 2004-11-29 | 1 | -28/+62 |
| | | | | | | | SearchForAddressOfSymbol. llvm-svn: 18355 | ||||
| * | Shared library extension is now in LTDL_SHLIB_EXT | Reid Spencer | 2004-11-29 | 2 | -3/+3 |
| | | | | | llvm-svn: 18353 | ||||
| * | We just use ltdl's implementation for this abstraction now. Its portable to | Reid Spencer | 2004-11-29 | 1 | -13/+3 |
| | | | | | | | more platforms than LLVM supports. llvm-svn: 18352 | ||||
| * | Mods for compilation with llvm. | Reid Spencer | 2004-11-29 | 1 | -7/+10 |
| | | | | | llvm-svn: 18346 | ||||
| * | Original version of ltdl.h from libtool 1.5.10 | Reid Spencer | 2004-11-29 | 1 | -0/+366 |
| | | | | | llvm-svn: 18345 | ||||
| * | Original version of ltdl.c from libtool 1.5.10 | Reid Spencer | 2004-11-29 | 1 | -0/+4495 |
| | | | | | llvm-svn: 18344 | ||||
| * | Implement the default constructor which causes the current program to be | Reid Spencer | 2004-11-29 | 3 | -5/+29 |
| | | | | | | | opened as if it was a dynamic library so its symbols can be searched too. llvm-svn: 18341 | ||||
| * | Revamp long/ulong comparisons to use a much more efficient sequence (thanks | Chris Lattner | 2004-11-29 | 1 | -36/+15 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | to Brian and the Sun compiler for pointing out that the obvious works :) This also enables folding all long comparisons into setcc and branch instructions: before we could only do == and != For example, for: void test(unsigned long long A, unsigned long long B) { if (A < B) foo(); } We now generate: test: subl $4, %esp movl %esi, (%esp) movl 8(%esp), %eax movl 12(%esp), %ecx movl 16(%esp), %edx movl 20(%esp), %esi subl %edx, %eax sbbl %esi, %ecx jae .LBBtest_2 # UnifiedReturnBlock .LBBtest_1: # then call foo movl (%esp), %esi addl $4, %esp ret .LBBtest_2: # UnifiedReturnBlock movl (%esp), %esi addl $4, %esp ret Instead of: test: subl $12, %esp movl %esi, 8(%esp) movl %ebx, 4(%esp) movl 16(%esp), %eax movl 20(%esp), %ecx movl 24(%esp), %edx movl 28(%esp), %esi cmpl %edx, %eax setb %al cmpl %esi, %ecx setb %bl cmove %ax, %bx testb %bl, %bl je .LBBtest_2 # UnifiedReturnBlock .LBBtest_1: # then call foo movl 4(%esp), %ebx movl 8(%esp), %esi addl $12, %esp ret .LBBtest_2: # UnifiedReturnBlock movl 4(%esp), %ebx movl 8(%esp), %esi addl $12, %esp ret llvm-svn: 18330 | ||||
| * | Reworked branching so we don't handle BAs specially. It just updates the ↵ | Tanya Lattner | 2004-11-29 | 2 | -139/+97 |
| | | | | | | | branchTO regardless of what type of branch it is. llvm-svn: 18322 | ||||
| * | Fixed bug where instructions in the kernel were not ordered right to ↵ | Tanya Lattner | 2004-11-28 | 4 | -7/+32 |
| | | | | | | | preserve dependencies in a cycle. llvm-svn: 18314 | ||||
| * | Fix for PR454: | Reid Spencer | 2004-11-28 | 1 | -41/+80 |
| | | | | | | | | * Make sure we handle signed to unsigned conversion correctly * Move this visitSetCondInst case to its own method. llvm-svn: 18312 | ||||
| * | The LLVM bool type shall have 1 byte alignment on PPC. | Chris Lattner | 2004-11-28 | 1 | -2/+2 |
| | | | | | llvm-svn: 18311 | ||||
| * | Make DSE potentially more aggressive by being more specific about alloca sizes. | Chris Lattner | 2004-11-28 | 1 | -4/+10 |
| | | | | | llvm-svn: 18309 | ||||
| * | Fix DeadStoreElimination/2004-11-28-LiveStoreDeleted.ll | Chris Lattner | 2004-11-28 | 1 | -3/+3 |
| | | | | | llvm-svn: 18308 | ||||
| * | Fix SingleSource/UnitTests/2004-11-28-GlobalBoolLayout.c, and hopefully | Chris Lattner | 2004-11-28 | 1 | -1/+1 |
| | | | | | | | PR449 llvm-svn: 18306 | ||||
| * | Fix PR463 | Chris Lattner | 2004-11-28 | 1 | -1/+4 |
| | | | | | llvm-svn: 18303 | ||||
| * | Compute the firstFileOffset correctly after reading the LLVM symbol table. | Reid Spencer | 2004-11-28 | 1 | -2/+3 |
| | | | | | llvm-svn: 18300 | ||||
| * | When merging to alias sets, if they are both must alias, the result is not | Chris Lattner | 2004-11-27 | 1 | -8/+22 |
| | | | | | | | | a must alias set unless all of the pointers in the resultant set are must aliased together. llvm-svn: 18275 | ||||
| * | Implement Regression/Transforms/InstCombine/getelementptr_cast.ll, which | Chris Lattner | 2004-11-27 | 1 | -0/+15 |
| | | | | | | | occurs many times in crafty llvm-svn: 18273 | ||||
| * | Remove the ISel->AsmPrinter link via the TargetMachine that was put in | Nate Begeman | 2004-11-27 | 3 | -24/+2 |
| | | | | | | | | | place to help bring up the PowerPC back end on Darwin. This code is no longer serves any purpose now that the AsmPrinter does the right thing all the time printing GlobalValues. --Cruft. llvm-svn: 18267 | ||||
| * | Add a new interface | Chris Lattner | 2004-11-26 | 1 | -0/+12 |
| | | | | | llvm-svn: 18266 | ||||
| * | Provide size information when checking to see if we can LICM a load, this | Chris Lattner | 2004-11-26 | 1 | -3/+6 |
| | | | | | | | allows us to hoist more loads in some cases. llvm-svn: 18265 | ||||
| * | When evaluating an AA, pass in size info | Chris Lattner | 2004-11-26 | 1 | -28/+38 |
| | | | | | llvm-svn: 18264 | ||||
| * | There is no reason to store <x,x>, just store <x>. | Chris Lattner | 2004-11-26 | 1 | -5/+3 |
| | | | | | llvm-svn: 18263 | ||||
| * | The trick with globals actually works with allocas and malloc too | Chris Lattner | 2004-11-26 | 1 | -12/+12 |
| | | | | | llvm-svn: 18262 | ||||
| * | A store or load cannot alias a global if the accessed amount is larger then | Chris Lattner | 2004-11-26 | 1 | -20/+49 |
| | | | | | | | | | the global. This implements Regression/Analysis/BasicAA/global-size.ll llvm-svn: 18261 | ||||
| * | Add bzip2 subdirectory | Reid Spencer | 2004-11-25 | 1 | -0/+1 |
| | | | | | llvm-svn: 18251 | ||||
| * | Remove zlib support in favor of our own bzip2 library | Reid Spencer | 2004-11-25 | 1 | -177/+77 |
| | | | | | llvm-svn: 18250 | ||||

