| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Fix PR577 and testcase InstCombine/2005-06-15-ShiftSetCCCrash.ll. | Chris Lattner | 2005-06-15 | 1 | -2/+16 |
| | | | | | | | Do not perform undefined out of range shifts. llvm-svn: 22217 | ||||
| * | Put the hack back in that removes features, causes regressions to fail, but | Reid Spencer | 2005-06-15 | 1 | -0/+2 |
| | | | | | | | allows test programs to succeed. Actual fix for this is forthcoming. llvm-svn: 22213 | ||||
| * | Commit fix for generating conditional branch pseudo instructions that | Nate Begeman | 2005-06-15 | 1 | -3/+8 |
| | | | | | | | | avoids dereferencing the end() iterator when selecting the fallthrough block. This requires an ilist change. llvm-svn: 22212 | ||||
| * | Commit a small improvement that is already in the x86 and ia64 backends to | Nate Begeman | 2005-06-14 | 1 | -0/+5 |
| | | | | | | | | not generate unnecessary register copies. This improves compile time by 2-5% depending on the test. llvm-svn: 22210 | ||||
| * | Unbreak several InstCombine regression checks introduced by a hack to | Reid Spencer | 2005-06-13 | 1 | -2/+0 |
| | | | | | | | fix the bzip2 test. A better hack is needed. llvm-svn: 22209 | ||||
| * | When compiled with GCC 4.0, a latent bug was exposed where both SparcV9 | Nate Begeman | 2005-06-12 | 16 | -106/+108 |
| | | | | | | | | | | and the target independant register allocator were both using a class named 'LiveRange'. This lead to the target independant code calling code in the SparcV9 backend, which crashed. Fixed by renaming SparcV9's LiveRange to V9LiveRange. llvm-svn: 22208 | ||||
| * | Fix a memory smasher caught by Mac OS X's debug malloc library. We were | Nate Begeman | 2005-06-12 | 1 | -8/+19 |
| | | | | | | | incorrectly using an iterator after it was invalid. llvm-svn: 22207 | ||||
| * | Fix a 64-bit problem, passing (int)0 through ... instead of (void*)0 | Chris Lattner | 2005-06-09 | 1 | -4/+4 |
| | | | | | llvm-svn: 22206 | ||||
| * | Fix a problem on 64-bit targets where we passed (int)0 through ... instead of | Chris Lattner | 2005-06-09 | 1 | -7/+7 |
| | | | | | | | (void*)0. llvm-svn: 22205 | ||||
| * | fix BranchCC with a setCC with an arg of 0 | Andrew Lenharth | 2005-06-08 | 1 | -3/+3 |
| | | | | | llvm-svn: 22203 | ||||
| * | For PR572: | Reid Spencer | 2005-06-08 | 1 | -0/+6 |
| | | | | | | | | Undefine the PPC symbol which is defined by Linux/PPC (erroneously) so it doesn't pollute the user namespace and clash with our namespace declarations. llvm-svn: 22202 | ||||
| * | Handle some more real world cases of rlwimi. These don't come up that | Nate Begeman | 2005-06-08 | 1 | -10/+42 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | regularly in "normal" code, but for things like software graphics, they make a big difference. For the following code: unsigned short Trans16Bit(unsigned srcA,unsigned srcB,unsigned alpha) { unsigned tmpA,tmpB,mixed; tmpA = ((srcA & 0x03E0) << 15) | (srcA & 0x7C1F); tmpB = ((srcB & 0x03E0) << 15) | (srcB & 0x7C1F); mixed = (tmpA * alpha) + (tmpB * (32 - alpha)); return ((mixed >> 5) & 0x7C1F) | ((mixed >> 20) & 0x03E0); } We now generate: _Trans16Bit: .LBB_Trans16Bit_0: ; entry andi. r2, r4, 31775 rlwimi r2, r4, 15, 7, 11 subfic r4, r5, 32 mullw r2, r2, r4 andi. r4, r3, 31775 rlwimi r4, r3, 15, 7, 11 mullw r3, r4, r5 add r2, r2, r3 srwi r3, r2, 5 andi. r3, r3, 31775 rlwimi r3, r2, 12, 22, 26 blr Instead of: _Trans16Bit: .LBB_Trans16Bit_0: ; entry slwi r2, r4, 15 rlwinm r2, r2, 0, 7, 11 andi. r4, r4, 31775 or r2, r2, r4 subfic r4, r5, 32 mullw r2, r2, r4 slwi r4, r3, 15 rlwinm r4, r4, 0, 7, 11 andi. r3, r3, 31775 or r3, r4, r3 mullw r3, r3, r5 add r2, r2, r3 srwi r3, r2, 5 andi. r3, r3, 31775 srwi r2, r2, 20 rlwimi r3, r2, 0, 22, 26 blr llvm-svn: 22201 | ||||
| * | Fix lli linking on Mac OS X 10.4.1 for Intel. | Nate Begeman | 2005-06-08 | 1 | -2/+2 |
| | | | | | llvm-svn: 22200 | ||||
| * | Remove code for conversion from old style va_args. Preparing the way for | Andrew Lenharth | 2005-06-08 | 1 | -76/+3 |
| | | | | | | | returning to the old style :) llvm-svn: 22199 | ||||
| * | Fix spelling of `correlate' | Misha Brukman | 2005-06-06 | 1 | -1/+1 |
| | | | | | llvm-svn: 22196 | ||||
| * | allow marking of loads and stores in the instruction stream with enough ↵ | Andrew Lenharth | 2005-06-06 | 2 | -0/+31 |
| | | | | | | | information to reconstruct the Value* if it existed llvm-svn: 22195 | ||||
| * | hide basic block labels. The utility of these for debuging is long since passed | Andrew Lenharth | 2005-06-06 | 1 | -2/+2 |
| | | | | | llvm-svn: 22194 | ||||
| * | * Replace block of commented-out lines with #if 0 | Misha Brukman | 2005-06-06 | 1 | -26/+29 |
| | | | | | | | * Remove warning "control reaches end of non-void function" llvm-svn: 22193 | ||||
| * | hack to fix bzip2 (bug 571) | Andrew Lenharth | 2005-06-04 | 1 | -0/+2 |
| | | | | | llvm-svn: 22192 | ||||
| * | Make sure that Cygwin assembly includes _ as part of function names. | Reid Spencer | 2005-06-02 | 1 | -0/+9 |
| | | | | | llvm-svn: 22190 | ||||
| * | Put in a hack for Cygwin that prevents mkdtemp from being used since | Reid Spencer | 2005-06-02 | 1 | -0/+5 |
| | | | | | | | configure seems to find it on Cygwin but linking against it fails. llvm-svn: 22189 | ||||
| * | try custom expanders, doesn't seem to expand yet, so disabled | Andrew Lenharth | 2005-05-31 | 1 | -0/+61 |
| | | | | | llvm-svn: 22188 | ||||
| * | switch to the new live in thing. Really, this time it works | Andrew Lenharth | 2005-05-31 | 1 | -1/+1 |
| | | | | | llvm-svn: 22187 | ||||
| * | switch to the new live in thing | Andrew Lenharth | 2005-05-31 | 1 | -2/+2 |
| | | | | | llvm-svn: 22186 | ||||
| * | switch to the new live in thing | Andrew Lenharth | 2005-05-31 | 1 | -65/+60 |
| | | | | | llvm-svn: 22185 | ||||
| * | match gcc, makes diff easier | Andrew Lenharth | 2005-05-27 | 1 | -2/+2 |
| | | | | | llvm-svn: 22179 | ||||
| * | Fix 2005-05-12-Int64ToFP | Andrew Lenharth | 2005-05-26 | 1 | -0/+26 |
| | | | | | | | | | | The issue is there is no unsigned -> double conversion, only signed. So I need to test the sign and do a different thing depending on it. Ideally this should be in a different BB, but in the mean time, I use a branch free method. llvm-svn: 22177 | ||||
| * | Make the registration hash_map static. No other module needs it. Also, | Reid Spencer | 2005-05-21 | 1 | -1/+5 |
| | | | | | | | document what its for a little better. llvm-svn: 22164 | ||||
| * | Adjust the file comment to read a little easier. | Reid Spencer | 2005-05-21 | 1 | -6/+7 |
| | | | | | llvm-svn: 22163 | ||||
| * | Make sure ... arguments are casted to sbyte* where needed. | Reid Spencer | 2005-05-21 | 1 | -13/+12 |
| | | | | | llvm-svn: 22162 | ||||
| * | Add a "brief" comment for CastToCStr | Reid Spencer | 2005-05-21 | 1 | -0/+1 |
| | | | | | llvm-svn: 22161 | ||||
| * | Fix mismatched type problem that crashed on cases like this: | Chris Lattner | 2005-05-20 | 1 | -8/+17 |
| | | | | | | | | | sprintf(P, "%s", X); Where X is not an sbyte*. This fixes the bug JohnMC reported on llvm-bugs. llvm-svn: 22159 | ||||
| * | Fix Transforms/SimplifyCFG/switch-simplify-crash.ll | Chris Lattner | 2005-05-20 | 1 | -0/+7 |
| | | | | | llvm-svn: 22158 | ||||
| * | C'mon everybody, let's modify X86JITInfo.cpp. This time, we add <iostream> | Nate Begeman | 2005-05-20 | 1 | -1/+2 |
| | | | | | | | so that the shiny new use of std::cerr is defined. llvm-svn: 22156 | ||||
| * | Since everyone else has "fixed" this file, might as well join in the fun. | Misha Brukman | 2005-05-20 | 1 | -4/+3 |
| | | | | | | | | * Change assert() to std::cerr printout, as it will not appear in opt builds * Add comments to clarify what #ifdef/#else/#endif match what condition(s) llvm-svn: 22154 | ||||
| * | Fix this a 3rd time :) | Chris Lattner | 2005-05-20 | 1 | -1/+2 |
| | | | | | llvm-svn: 22151 | ||||
| * | fix compilation error due to no abort being defined. There is probably a ↵ | Andrew Lenharth | 2005-05-20 | 1 | -0/+1 |
| | | | | | | | better way to do this llvm-svn: 22150 | ||||
| * | re-enable direct calls, this should just be a performance boost | Duraid Madina | 2005-05-20 | 1 | -6/+5 |
| | | | | | llvm-svn: 22148 | ||||
| * | this seems dead (and broke the ia64 build, so..) | Duraid Madina | 2005-05-20 | 1 | -1/+0 |
| | | | | | llvm-svn: 22147 | ||||
| * | Give the asmparser the ability to parse strings. Patch contributed by | Chris Lattner | 2005-05-20 | 4 | -13/+49 |
| | | | | | | | Alexander Friedman llvm-svn: 22146 | ||||
| * | Fix tail call support in VC++ builds | Jeff Cohen | 2005-05-20 | 1 | -38/+45 |
| | | | | | llvm-svn: 22143 | ||||
| * | Fastcc passes arguments in EAX and EDX, make sure the JIT doesn't clobber them | Chris Lattner | 2005-05-19 | 1 | -6/+39 |
| | | | | | llvm-svn: 22137 | ||||
| * | Tailcalls require stubs to be emitted. Otherwise, the compilation callback | Chris Lattner | 2005-05-19 | 2 | -9/+16 |
| | | | | | | | doesn't know who 'called' it. llvm-svn: 22136 | ||||
| * | Wrap long lines | Misha Brukman | 2005-05-18 | 1 | -5/+14 |
| | | | | | llvm-svn: 22125 | ||||
| * | teach the inliner about coldcc and noreturn functions | Chris Lattner | 2005-05-18 | 1 | -0/+15 |
| | | | | | llvm-svn: 22113 | ||||
| * | comment the hpux bit | Duraid Madina | 2005-05-16 | 1 | -0/+1 |
| | | | | | llvm-svn: 22081 | ||||
| * | HP-UX system headers make a mess of isinf(), so much so that gcc fixincludes | Duraid Madina | 2005-05-16 | 1 | -0/+4 |
| | | | | | | | can't patch it up for c++ (works in C though) llvm-svn: 22079 | ||||
| * | Don't look for __builtin_ffs, we'll never see it from llvm-gcc and there's | Reid Spencer | 2005-05-15 | 1 | -14/+0 |
| | | | | | | | not reason to include it for other front ends. llvm-svn: 22070 | ||||
| * | Provide this optimization as well: | Reid Spencer | 2005-05-15 | 1 | -5/+45 |
| | | | | | | | ffs(x) -> (x == 0 ? 0 : 1+llvm.cttz(x)) llvm-svn: 22068 | ||||
| * | PPC "branch and link" instructions are branches in the PPC sense, but not | Chris Lattner | 2005-05-15 | 1 | -1/+1 |
| | | | | | | | | in the LLVM code generator sense (they are calls). Don't mark them as such, which fixes the regressions on the ppc tester last night llvm-svn: 22065 | ||||

