Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | For PR950: | Reid Spencer | 2006-12-08 | 1 | -0/+16 | |
| | | | | | | Add a test for a regression in combining of ICmp instructions. llvm-svn: 32358 | |||||
* | Manually convert this to new assembly. | Reid Spencer | 2006-12-07 | 1 | -3/+3 | |
| | | | | llvm-svn: 32301 | |||||
* | Instcombine shouldn't fold these casts. | Chris Lattner | 2006-12-05 | 1 | -0/+13 | |
| | | | | llvm-svn: 32243 | |||||
* | new testcase | Chris Lattner | 2006-12-05 | 1 | -0/+7 | |
| | | | | llvm-svn: 32210 | |||||
* | SCCP does not handle Packed Type properly. Disable Packed Type handling | Devang Patel | 2006-12-04 | 1 | -0/+149 | |
| | | | | | | for now. llvm-svn: 32208 | |||||
* | Purpose of test is lost in upgrade. | Jim Laskey | 2006-12-04 | 1 | -23/+0 | |
| | | | | llvm-svn: 32182 | |||||
* | Remove debug info. | Jim Laskey | 2006-12-04 | 1 | -5/+0 | |
| | | | | llvm-svn: 32181 | |||||
* | Upgrade to version 7 bytecode format. | Reid Spencer | 2006-12-03 | 1 | -0/+0 | |
| | | | | llvm-svn: 32143 | |||||
* | Upgrade intrinsic function calls manually. | Reid Spencer | 2006-12-02 | 4 | -18/+18 | |
| | | | | llvm-svn: 32137 | |||||
* | Prevent a GEP from getting upgraded to avoid resulting test failure. | Reid Spencer | 2006-12-02 | 1 | -1/+1 | |
| | | | | llvm-svn: 32123 | |||||
* | Update this test to use llvm-update. | Reid Spencer | 2006-12-02 | 1 | -2/+2 | |
| | | | | llvm-svn: 32117 | |||||
* | Use the llvm-upgrade program to upgrade llvm assembly. | Reid Spencer | 2006-12-02 | 636 | -713/+714 | |
| | | | | llvm-svn: 32115 | |||||
* | new testcase | Chris Lattner | 2006-12-02 | 1 | -0/+9 | |
| | | | | llvm-svn: 32111 | |||||
* | fix incorrectly translated testcase | Chris Lattner | 2006-12-01 | 1 | -1/+1 | |
| | | | | llvm-svn: 32093 | |||||
* | test1 is miscompiled | Chris Lattner | 2006-12-01 | 1 | -0/+13 | |
| | | | | llvm-svn: 32086 | |||||
* | new 'cast' miscompilation, reduced from perlbmk by Evan | Chris Lattner | 2006-12-01 | 1 | -0/+5 | |
| | | | | llvm-svn: 32079 | |||||
* | new testcase | Chris Lattner | 2006-12-01 | 1 | -0/+7 | |
| | | | | llvm-svn: 32061 | |||||
* | The InstCombine hack has been removed, no need to XFAIL this any more. | Reid Spencer | 2006-12-01 | 1 | -4/+0 | |
| | | | | llvm-svn: 32058 | |||||
* | new bswap idiom | Chris Lattner | 2006-11-29 | 1 | -1/+20 | |
| | | | | llvm-svn: 32012 | |||||
* | new testcase | Chris Lattner | 2006-11-29 | 1 | -0/+7 | |
| | | | | llvm-svn: 32010 | |||||
* | new testcase | Chris Lattner | 2006-11-29 | 1 | -0/+8 | |
| | | | | llvm-svn: 32008 | |||||
* | new testcase, bswaps should be eliminated. | Chris Lattner | 2006-11-29 | 1 | -0/+26 | |
| | | | | llvm-svn: 32005 | |||||
* | Fix a typo. | Reid Spencer | 2006-11-28 | 1 | -1/+1 | |
| | | | | llvm-svn: 31959 | |||||
* | this doesn't pass with CVS HEAD. | Chris Lattner | 2006-11-27 | 1 | -0/+5 | |
| | | | | llvm-svn: 31942 | |||||
* | testcase for PR1014 | Chris Lattner | 2006-11-27 | 1 | -0/+11 | |
| | | | | llvm-svn: 31939 | |||||
* | Add a test to ensure a bitcast/and/trunc combination eliminates the bitcast. | Reid Spencer | 2006-11-27 | 1 | -0/+14 | |
| | | | | llvm-svn: 31934 | |||||
* | For PR950: | Reid Spencer | 2006-11-27 | 25 | -29/+134 | |
| | | | | | | | | | | The long awaited CAST patch. This introduces 12 new instructions into LLVM to replace the cast instruction. Corresponding changes throughout LLVM are provided. This passes llvm-test, llvm/test, and SPEC CPUINT2000 with the exception of 175.vpr which fails only on a slight floating point output difference. llvm-svn: 31931 | |||||
* | Promote GEP ubyte indices to uint. Backwards compatibility for 1.2 and | Reid Spencer | 2006-11-23 | 31 | -59/+59 | |
| | | | | | | | older features will be dropped soon and these test cases must not rely on the upgrade capability. llvm-svn: 31896 | |||||
* | Update to new predicate simplifier VRP design. Fixes PR966 and PR967. | Nick Lewycky | 2006-11-22 | 5 | -0/+148 | |
| | | | | | | | | Remove predicate simplifier from default gcc3 pipeline. New design is too slow to enable by default. Add new testcases for problems encountered in development. llvm-svn: 31895 | |||||
* | new testcase | Chris Lattner | 2006-11-20 | 1 | -0/+9 | |
| | | | | llvm-svn: 31885 | |||||
* | converting massive blocks of phis into selects like this is silly. | Chris Lattner | 2006-11-18 | 1 | -28/+0 | |
| | | | | llvm-svn: 31852 | |||||
* | New testcase. | Chris Lattner | 2006-11-17 | 1 | -0/+48 | |
| | | | | llvm-svn: 31808 | |||||
* | new testcase | Chris Lattner | 2006-11-14 | 1 | -0/+22 | |
| | | | | llvm-svn: 31728 | |||||
* | new testcase | Chris Lattner | 2006-11-14 | 1 | -0/+10 | |
| | | | | llvm-svn: 31726 | |||||
* | For PR950: | Reid Spencer | 2006-11-13 | 6 | -6/+6 | |
| | | | | | | Changes for new cast instructions that are backwards compatible. llvm-svn: 31707 | |||||
* | Don't hard code test names into the test. | Reid Spencer | 2006-11-13 | 1 | -1/+1 | |
| | | | | llvm-svn: 31706 | |||||
* | Fix 80 cols violations. | Reid Spencer | 2006-11-13 | 1 | -5/+5 | |
| | | | | llvm-svn: 31705 | |||||
* | For PR950: | Reid Spencer | 2006-11-13 | 1 | -1/+1 | |
| | | | | | | Use the notcast script in preparation for 12 casts. llvm-svn: 31704 | |||||
* | For PR950: | Reid Spencer | 2006-11-13 | 1 | -1/+1 | |
| | | | | | | Use the notcast script in preparation for 12 cast instructions. llvm-svn: 31703 | |||||
* | This testcase is miscompiled by instcombine, reduced from 176.gcc | Chris Lattner | 2006-11-10 | 1 | -0/+7 | |
| | | | | llvm-svn: 31652 | |||||
* | new testcases for PR992/993 | Chris Lattner | 2006-11-09 | 2 | -0/+590 | |
| | | | | llvm-svn: 31609 | |||||
* | new testcases | Chris Lattner | 2006-11-09 | 2 | -2/+24 | |
| | | | | llvm-svn: 31585 | |||||
* | For PR950: | Reid Spencer | 2006-11-08 | 1 | -1/+1 | |
| | | | | | | | | This patch converts the old SHR instruction into two instructions, AShr (Arithmetic) and LShr (Logical). The Shr instructions now are not dependent on the sign of their operands. llvm-svn: 31542 | |||||
* | scalarrepl should not split the two elements of the vsiidx array: | Chris Lattner | 2006-11-07 | 1 | -0/+19 | |
| | | | | | | | | | | | | | int func(vFloat v0, vFloat v1) { int ii; vSInt32 vsiidx[2]; vsiidx[0] = _mm_cvttps_epi32(v0); vsiidx[1] = _mm_cvttps_epi32(v1); ii = ((int *) vsiidx)[4]; return ii; } llvm-svn: 31523 | |||||
* | Got a little ahead of myself here | Andrew Lenharth | 2006-11-07 | 1 | -0/+0 | |
| | | | | llvm-svn: 31502 | |||||
* | add a regression for memmove | Andrew Lenharth | 2006-11-03 | 1 | -0/+18 | |
| | | | | llvm-svn: 31430 | |||||
* | Testcase from http://llvm.org/bugs/show_bug.cgi?id=979 | Devang Patel | 2006-11-01 | 1 | -0/+1087 | |
| | | | | llvm-svn: 31357 | |||||
* | extra pass is required now | Chris Lattner | 2006-11-01 | 1 | -1/+1 | |
| | | | | llvm-svn: 31353 | |||||
* | new testcase that crashes global opt, reduced from chomp | Chris Lattner | 2006-11-01 | 1 | -0/+33 | |
| | | | | llvm-svn: 31351 | |||||
* | remove a testcase for a marginal feature | Chris Lattner | 2006-11-01 | 1 | -14/+0 | |
| | | | | llvm-svn: 31345 |