| Commit message (Collapse) | Author | Age | Files | Lines | ||
|---|---|---|---|---|---|---|
| ... | ||||||
| * | add a new DSGraph::spliceFrom method, which violently takes the content of | Chris Lattner | 2005-03-24 | 1 | -0/+41 | |
| | | | | | | | one graph and plops it into another, without breaking a sweat. llvm-svn: 20824 | |||||
| * | add new spliceFrom methods. | Chris Lattner | 2005-03-24 | 1 | -0/+10 | |
| | | | | | llvm-svn: 20823 | |||||
| * | Implement next round of Pattern ISel fixes | Nate Begeman | 2005-03-24 | 1 | -38/+162 | |
| | | | | | | | | | 1. void returns 2. multiplies 3. calls llvm-svn: 20822 | |||||
| * | Fix an incorrect argument being passed to BuildMI for indirect calls. | Nate Begeman | 2005-03-24 | 1 | -1/+1 | |
| | | | | | llvm-svn: 20821 | |||||
| * | This replaces the correct but slow code with a more aggressive scc-finder | Chris Lattner | 2005-03-24 | 1 | -54/+97 | |
| | | | | | | | | | based approach to find globals and call sites that need to be copied. This speeds up the BU pass on 176.gcc from 22s back up to 2.3s. Not as good as 1.5s, but at least it's correct :) llvm-svn: 20820 | |||||
| * | only look at successors of globals. This gets us down to "only" 22s in the | Chris Lattner | 2005-03-24 | 1 | -4/+9 | |
| | | | | | | | bu pass for 176.gcc llvm-svn: 20818 | |||||
| * | Unfortunately, a previous patch was not safe. Revert it, reimplement | Chris Lattner | 2005-03-24 | 1 | -40/+93 | |
| | | | | | | | | something correct. Unfortunately this takes 176.gcc's BU phase back up to 29s from 1.5. This fixes DSGraph/2005-03-24-Global-Arg-Alias.ll llvm-svn: 20817 | |||||
| * | Commit Gabor Greif's patch to use iterators in lowering intrinsics. | Nate Begeman | 2005-03-24 | 1 | -7/+8 | |
| | | | | | llvm-svn: 20816 | |||||
| * | don't bother |'ing in 0's | Chris Lattner | 2005-03-24 | 3 | -4/+2 | |
| | | | | | llvm-svn: 20815 | |||||
| * | be more aggressive about incompleteness marking | Chris Lattner | 2005-03-24 | 1 | -3/+2 | |
| | | | | | llvm-svn: 20814 | |||||
| * | new testcase that DS-AA is causing to be miscompiled. | Chris Lattner | 2005-03-24 | 1 | -0/+26 | |
| | | | | | llvm-svn: 20813 | |||||
| * | eliminate dead variables, patch contributed by Gabor Greif! | Chris Lattner | 2005-03-24 | 3 | -6/+0 | |
| | | | | | llvm-svn: 20812 | |||||
| * | Implement more of the PPC32 Pattern ISel: | Nate Begeman | 2005-03-24 | 3 | -35/+124 | |
| | | | | | | | | | | | | | | 1) dynamic stack alloc 2) loads 3) shifts 4) subtract 5) immediate form of add, and, or, xor 6) change flag from -pattern-isel to -enable-ppc-pattern-isel Remove dead arguments from getGlobalBaseReg in the simple ISel llvm-svn: 20810 | |||||
| * | Fix silly "no newline at end of file" warning | Chris Lattner | 2005-03-24 | 1 | -1/+2 | |
| | | | | | llvm-svn: 20809 | |||||
| * | Fix compilation errors, patch contributed by the fabulous Bill Wendling! | Chris Lattner | 2005-03-24 | 1 | -0/+1 | |
| | | | | | llvm-svn: 20808 | |||||
| * | fix a compilation error, patch contributed by Bill Wendling! | Chris Lattner | 2005-03-24 | 1 | -0/+1 | |
| | | | | | llvm-svn: 20807 | |||||
| * | Addition of the PPC32 Pattern ISel. While it is far from complete, it will | Nate Begeman | 2005-03-24 | 3 | -0/+761 | |
| | | | | | | | | be brought up to parity with the current simple ISel in the coming days. Currently, -pattern-isel is required to trigger it. llvm-svn: 20805 | |||||
| * | Remove comments that are now meaningless from the pattern ISels, at Chris's | Nate Begeman | 2005-03-24 | 3 | -6/+3 | |
| | | | | | | | request. llvm-svn: 20804 | |||||
| * | Fix a crash while promoting a value out of a loop from a global variable | Chris Lattner | 2005-03-24 | 2 | -2/+20 | |
| | | | | | | | when using ds-aa llvm-svn: 20802 | |||||
| * | teach ds-aa about mod/ref for external function calls. | Chris Lattner | 2005-03-24 | 1 | -1/+25 | |
| | | | | | llvm-svn: 20801 | |||||
| * | Simplify dead code into a fixme :) | Chris Lattner | 2005-03-24 | 1 | -13/+4 | |
| | | | | | llvm-svn: 20800 | |||||
| * | fix a compiler crash in runtime/libprofile | Chris Lattner | 2005-03-24 | 1 | -1/+1 | |
| | | | | | llvm-svn: 20799 | |||||
| * | wrap a long line | Chris Lattner | 2005-03-23 | 1 | -1/+2 | |
| | | | | | llvm-svn: 20797 | |||||
| * | If we are calling an external function, chain to another AA to potentially | Chris Lattner | 2005-03-23 | 1 | -8/+20 | |
| | | | | | | | | | decide, don't just immediately give up. This implements GlobalsModRef/chaining-analysis.ll llvm-svn: 20796 | |||||
| * | GlobalMR should be able to analyze this function. | Chris Lattner | 2005-03-23 | 1 | -0/+20 | |
| | | | | | llvm-svn: 20795 | |||||
| * | if a function doesn't access memory at all, it definitely doesn't read it. | Chris Lattner | 2005-03-23 | 1 | -1/+2 | |
| | | | | | llvm-svn: 20794 | |||||
| * | Make this more efficient by only making one virtual method call. | Chris Lattner | 2005-03-23 | 1 | -4/+6 | |
| | | | | | llvm-svn: 20793 | |||||
| * | Make this a bit more aggressive | Chris Lattner | 2005-03-23 | 1 | -1/+1 | |
| | | | | | llvm-svn: 20792 | |||||
| * | a hack to allow count-aa to work with ds-aa :( | Chris Lattner | 2005-03-23 | 2 | -6/+5 | |
| | | | | | llvm-svn: 20791 | |||||
| * | Add two options to allow -count-aa to print queries either (1) all queries, | Chris Lattner | 2005-03-23 | 1 | -25/+59 | |
| | | | | | | | or (2) only queries that are not successful (e.g. return may alias) llvm-svn: 20790 | |||||
| * | Fix grammar | Misha Brukman | 2005-03-23 | 1 | -1/+1 | |
| | | | | | llvm-svn: 20789 | |||||
| * | enable -debug-only=licm | Chris Lattner | 2005-03-23 | 1 | -0/+1 | |
| | | | | | llvm-svn: 20788 | |||||
| * | turn a dead conditional into an assert. | Chris Lattner | 2005-03-23 | 1 | -9/+5 | |
| | | | | | llvm-svn: 20787 | |||||
| * | Totally gut mergeInGraph. There is absolutely no reason to be merging | Chris Lattner | 2005-03-23 | 1 | -73/+47 | |
| | | | | | | | | | | | | global roots in from callees to callers. The BU graphs do not have accurate globals information and all of the clients know it. Instead, just make sure the GG is up-to-date, and they will be perfectly satiated. This speeds up the BU pass on 176.gcc from 5.5s to 1.5s, and Loc+BU+TD from 7s to 2.7s. llvm-svn: 20786 | |||||
| * | wrap a long line | Chris Lattner | 2005-03-23 | 1 | -1/+2 | |
| | | | | | llvm-svn: 20785 | |||||
| * | don't lie to the register allocator | Andrew Lenharth | 2005-03-23 | 1 | -2/+2 | |
| | | | | | llvm-svn: 20784 | |||||
| * | Make -steens-aa more conservative (aka correct) by making sure to obey | Chris Lattner | 2005-03-23 | 1 | -8/+18 | |
| | | | | | | | | | | incompleteness flags. Make it more aggressive by taking field sensitive information into account. llvm-svn: 20781 | |||||
| * | Added alias analysis. | Tanya Lattner | 2005-03-23 | 6 | -193/+724 | |
| | | | | | | | | Fixed many many bugs. This now works on almost all Singlesource , and most of MultiSource. llvm-svn: 20780 | |||||
| * | implement Analysis/DSGraph/field-sensitive.ll | Chris Lattner | 2005-03-23 | 1 | -3/+2 | |
| | | | | | llvm-svn: 20779 | |||||
| * | new testcase to verify that we have field sensitive alias analysis info. This | Chris Lattner | 2005-03-23 | 1 | -0/+21 | |
| | | | | | | | test cannot be satisfied without interprocedural information. llvm-svn: 20778 | |||||
| * | make sure this test tests the intended target. | Chris Lattner | 2005-03-23 | 1 | -1/+1 | |
| | | | | | llvm-svn: 20777 | |||||
| * | don't crash in some bad cases. | Chris Lattner | 2005-03-23 | 1 | -0/+3 | |
| | | | | | llvm-svn: 20776 | |||||
| * | Several changes here: | Chris Lattner | 2005-03-22 | 1 | -19/+42 | |
| | | | | | | | | | | | | | 1. Increase max node size from 64->256 to avoid collapsing an important structure in 181.mcf 2. If we have multiple calls to an indirect call node with an indirect callee, fold these call nodes together, to avoid DSA turning apoc into a flaming fireball of death when analyzing 176.gcc. With this change, 176.gcc now takes ~7s to analyze for loc+bu+td, with 5.7s of that in the BU pass. llvm-svn: 20775 | |||||
| * | We might as well check the TD graphs as well, even though they are fine. | Chris Lattner | 2005-03-22 | 1 | -2/+2 | |
| | | | | | llvm-svn: 20774 | |||||
| * | Mark external globals incomplete in the BU Globals graph, fixing | Chris Lattner | 2005-03-22 | 1 | -0/+3 | |
| | | | | | | | Analysis/DSGraph/2005-03-22-IncompleteGlobal.ll llvm-svn: 20773 | |||||
| * | New testcase, the BU pass is marking the global complete in the globals graph. | Chris Lattner | 2005-03-22 | 1 | -0/+12 | |
| | | | | | llvm-svn: 20772 | |||||
| * | Updating my entry. | Tanya Lattner | 2005-03-22 | 1 | -5/+6 | |
| | | | | | llvm-svn: 20771 | |||||
| * | Fix a serious bug where we didn't insert globals into the globalset when | Chris Lattner | 2005-03-22 | 1 | -1/+6 | |
| | | | | | | | cloning a graph. llvm-svn: 20770 | |||||
| * | hum, it is good to use real instructions | Andrew Lenharth | 2005-03-22 | 1 | -2/+2 | |
| | | | | | llvm-svn: 20769 | |||||
| * | Directly count the number of memory instructions. | Chris Lattner | 2005-03-22 | 1 | -4/+9 | |
| | | | | | llvm-svn: 20766 | |||||

