June 2009 Archives by date

• Messages sorted by: [ thread ] [ subject ] [ author ]
• More info on this list...

Starting: Mon Jun 1 12:50:41 IST 2009
Ending: Tue Jun 30 23:25:30 IST 2009
Messages: 338

• [phc-internals] [phc commit] r2719 - The "reduce" tools have become severly bit-rotten, and they were never very well written ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2720 - Allow reduce to just take in the contents of a file, and to run on either XML or PHP code...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2721 - Fix check_against_interpreted, remove the use of remove_all, and switch the default pass ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2722 - Fix an off-by-one error, which stopped the reduce algorithm when the program got quite sm...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2723 - Change the svn:ignore flags for manual to ignore the .build directory.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2724 - Rename core.php to Reduce.php, and document it.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2725 - Add a header to Reduce.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2726 - Add initial support for the --reduce flag for the test suite.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2727 - Remove unused file.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2728 - Pretend that Internal methods have no formal parameters. It may be better to get a count ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2729 - Handle reading from $b[$i], where $i is UNKNOWN (we add an index_node ABSVAL(b)->UNKNOWN,...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2730 - Add more information about debugging phc. There is quite few nuances to it, which we've b...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2731 - Make |phc| appear in bold.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2732 - Fix get_owner(). Sometimes its called on a "fake" index node, one which doesnt represent ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2733 - Optimize return statements with a known value.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2734 - The function update_phi_node() now compiles. I've made it compatible with new Alias_name...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2735 - Deleting wiki page CompilingPHP.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2736 - Fix monotonicity in backward_bind in WPA_lattice (CCP and type-inference).   codesite-noreply at google.com
• [phc-internals] code-reviews   Paul Biggar
• [phc-internals] [phc commit] r2737 - The command line argument --dump-uppered is long gone, but it was not fixed in raised_com...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2738 - Fix monotonicity in Aliasing and WPA_lattice (hence Type-inference and CCP). It was typic...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2739 - Add a "BasicOptimizeTest" to track the optimizer's progress. It just checks if the optimi...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2740 - Revised my previous editing of Def_use_web::update_phi_node() as per Paul's recommendatio...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2741 - Fixed previously commented out section in HSSA::convert_to_hssa_form(), now adds phi node...   codesite-noreply at google.com
• [phc-internals] Comment on revision r2740 in phc   codesite-noreply at google.com
• [phc-internals] Comment on revision r2741 in phc   codesite-noreply at google.com
• [phc-internals] [phc commit] r2742 - Fix build.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2743 - Add a "raised" optimize test.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2744 - I accidentally added an extra BasicOptimizeTest instead of a BasicRaisedOptimizeTest. Fixed.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2745 - Minor style modifications, spaces etc.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2746 - Slightly reorganize the Optimized tests, and add dependencies.   codesite-noreply at google.com
• [phc-internals] Comment on revision r2741 in phc   codesite-noreply at google.com
• [phc-internals] Comment on revision r2740 in phc   codesite-noreply at google.com
• [phc-internals] [phc commit] r2747 - Slight rework on the Monotonicity, after talking it through with David. We can clear the ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2748 - reduce.php is no longer used, its been replaced with old_driver.php, and the functionalit...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2749 - Its not necessary to include references in pull_possible_null.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2750 - Removing unconditional print which should ahve been predicated on --debug.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2751 - New, very simple, test case.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2752 - Edited HSSA::rename_vars() so that it now also renames phi nodes   codesite-noreply at google.com
• [phc-internals] [phc commit] r2753 - In DCE::sweep_pass(), I commented out the phc_TODO() and uncommented the some other code ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2754 - Remove an incorrect TODO, and fix an off-by-one error.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2755 - Add stream-of-consciousness comments to Def-use, which outlines most of how it needs to w...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2756 - Add a fix for the link-visited color (there is none). It will work in the next version of...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2757 - Make the def-use analysis a lot clearer. the comments are clearer, the code is much clear...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2758 - Fix summary propagation.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2759 - Minor syntax fix.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2760 - Fix two problems with uses not propagation:   codesite-noreply at google.com
• [phc-internals] [phc commit] r2761 - Edge's had an executable flag to the SCCP algorithm. However, in the presence of (possibl...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2762 - Add test for references.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2763 - I'm trying to fix analysing (potentially mutually) recursive functions. This adds a check...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2764 - Minor renaming of variables to make things clearer.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2766 - Add comment about how to fix reverse dominance information for infinite loops.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2767 - Tidy up and greatly improve dead-code elimination. The def-use information should be good...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2765 - Don't replace expressions which are Methods or News in Optimization_transformer.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2768 - trunk and dataflow were slightly out-of-sync, this fixes problems, which were all minor.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2769 - Go most of the way to modelling constants. Constants need to be modelled as an Abstract_v...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2770 - Merge all non-optimization differences from dataflow/ to trunk/.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2771 - Make constant analysis work.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2772 - This combines two large changes. The latter was required for the former (I should have ch...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2773 - Fix logic when handling constants.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2774 - After an Assign_var, I should have reset saved_is_ref.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2775 - If there are multiple receivers, check them all before putting literals into arguments.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2776 - outssa isnt a pass anymore. Use codegen instead.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2777 - Fix reading from $a[$i] when $a isnt created yet.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2778 - Unsetting a variable should kill its value as well as its reference. When doing so, we sh...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2779 - Move -Wno-unused-parameter from CXXFLAGS to CPPFLAGS.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2780 - This implements using the reduce algorithm automatically on test cases. In some cases, it...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2781 - Handle $$a and $a["f"]. These both re-use whats already there, so this is just a small ch...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2782 - Fix   codesite-noreply at google.com
• [phc-internals] [phc commit] r2783 - The concept of ABSVAL is different from an Abstract_value, even though they started meani...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2784 - Handle more operators.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2785 - Following some odd seg-faults, this makes all classes properly inherit from the GC.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2786 - Properly fix "Bailed out without an address" errors.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2787 - Phi nodes now show up at the start of basic blocks in CFGs.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2788 - Implement compact ().   codesite-noreply at google.com
• [phc-internals] [phc commit] r2789 - Handle reads from arrays indexed by a literal thats not a string or integer.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2790 - Implement range() for bench_sieve.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2791 - Don't put literals into the RHS of assign_array if the assign_array is by-ref.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2792 - Handle array_indices in unset.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2793 - Fix the logic for DEFINITE vs POSSIBLE (should have been > 1, not >= 1). Also take into a...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2794 - There was contention over whether to remove a field edge. In the case of killing referenc...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2795 - Model var_export.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2796 - Mention the presence of .sgml files in the README.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2797 - Re-order modelled functions alphabetically.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2798 - Fixed section in HSSA::convert_to_hssa_form() where I'd previously confused 'bb' and 'fro...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2799 - Handle variable-variables.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2800 - Optimize branch blocks with known values.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2801 - Make unknowns more difficult.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2802 - Changed var stacks to use a string rather than Alias_name for a key (Alias_name.name).   codesite-noreply at google.com
• [phc-internals] Comment on revision r2802 in phc   codesite-noreply at google.com
• [phc-internals] [phc commit] r2803 - Add "Getting Started" information. A couple of people have been looking at getting involv...   codesite-noreply at google.com
• [phc-internals] Newbies   Paul Biggar
• [phc-internals] [phc commit] r2804 - Make storage nodes abstract when they are created more than once.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2805 - The Whole_program::assign_* methods do too much. This starts splitting them up.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2806 - Move worklist to separate files.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2807 - Reorganize the Whole_program.h header to group assignment related functions together.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2808 - Main change is in Dead_code_elimination.cpp - now marks RDF of a block immediately after ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2808 - Main change is in Dead_code_elimination.cpp - now marks RDF of a block immediately after ...   Paul Biggar
• [phc-internals] [phc commit] r2809 - Revert accidental commit.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2810 - Rename assign_storage to assign_path_value. Use create_empty_storage instead of what was ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2811 - Cleaned up messy code.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2812 - Re-enabled assertions, removed some unnecessary code.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2813 - Add test for identity assignments.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2814 - Fixed segfault - stopped pop_var_stack () from popping stack if empty.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2815 - I had been killing definitions before constructing the values to be assigned to them. Thi...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2816 - Refactor to avoid duplication when assigning abstract values.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2817 - Rename most of assign_* to assign_path_* to indicate they work on paths. The signature sa...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2818 - Remove start of comment. This produces a warning in gcc, which drags my editor back to it...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2819 - Instead of copying contexts everywhere, use context pointers. I expected this to reduce t...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2820 - With the changes to use Context*s, I hadnt fixed the debugging, which started to print po...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2821 - We didn't copy_value from nodes which didnt exist. Copy them from unknown in that case.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2822 - Handle assignments to variable-variables, and add tests for edges cases (which we fail).   codesite-noreply at google.com
• [phc-internals] [phc commit] r2823 - Handle visit_new. However, invoke_method doesn't yet deal with targets, so that's the nex...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2824 - Support calling constructors.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2825 - Use the type name for storage nodes - this makes it more precise, but more importantly ma...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2826 - Get receivers for method calls. This makes method calls work, but doesn't take into accou...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2827 - Add standard_lhs and standard_rhs, which refactors the standard assignment functionality ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2828 - Make paths work for Fields.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2829 - Add a 'use_object' field to Path, so that we track when object handlers need to be called.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2830 - The use_object idea was ill thought out. Revert.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2831 - Add 'this' parameter for method calls. The object still needs to be copied to it.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2832 - Add the 'this' parameter to methods before the call (before, it had only been added to co...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2833 - Support visit_new in Optimization_transformer. To make this easier, get_possible_receiver...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2834 - In callgraph, use Method_infos rather than strings for the method names. This requires a ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2835 - Programs with method calls and field assignments should now complete.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2836 - Makr the value of the RHS of an assignment-by-reference as used.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2838 - Fixed Graphviz labels in Basic Block.cpp   codesite-noreply at google.com
• [phc-internals] [phc commit] r2837 - Fixed var_stacks indexing - unique string for each variable can now be gotten from its Al...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2838 - Fixed Graphviz labels in Basic Block.cpp   Paul Biggar
• [phc-internals] [phc commit] r2837 - Fixed var_stacks indexing - unique string for each variable can now be gotten from its Al...   Paul Biggar
• [phc-internals] [phc commit] r2839 - Clear out old code and comments.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2840 - Fix missing --extension flag.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2841 - After comparing the textual output of the html files generated by sphinx and the ones fro...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2842 - Remove old docbook manual, now that I'm convinved the new manual is up-to-scratch.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2843 - Fix var-var analysis.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2844 - Added new visitor class: Stat_collector, for gathering statistics.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2844 - Added new visitor class: Stat_collector, for gathering statistics.   Paul Biggar
• [phc-internals] [phc commit] r2845 - Make test work.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2848 - Added a new map for stats involving sets of strings and funcitons to manipulate it.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2846 - In the labels file, it was easy to say no-sometestname, but not possible to say 'dont use...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2847 - Expand casts test.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2849 - Add bugs reported by Stig Vestøl.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2850 - Add test: simple_may_def.php   codesite-noreply at google.com
• [phc-internals] [phc commit] r2853 - Don't fail if there is an attribute, and start to handle it instead. (handling is just a ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2854 - Handle error_reporting().   codesite-noreply at google.com
• [phc-internals] [phc commit] r2855 - Handle attributes (so long as they are literals).   codesite-noreply at google.com
• [phc-internals] [phc commit] r2857 - Stat dumps no longer unconditional, --stats flag will dump all stats.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2851 - Added May defs into CFG.x   codesite-noreply at google.com
• [phc-internals] [phc commit] r2856 - Handle return-by-reference and returning literals.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2852 - Updated Stat_collector and stats.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2857 - Stat dumps no longer unconditional, --stats flag will dump all stats.   Paul Biggar
• [phc-internals] [phc commit] r2858 - Handle var_export properly.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2860 - Model die(). This doesn't model the fact that code after die() is unreachable, which woul...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2859 - Handle a meothd invocation on a NULL object.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2862 - Handle 'empty' and 'defined'. This looks like all of the easy functions from the test suite.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2861 - Model more functions.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2857 - Stat dumps no longer unconditional, --stats flag will dump all stats.   Paul Biggar
• [phc-internals] [phc commit] r2866 - I hadn't added Result_state stuff to Constant_state, meaning I was getting 0 in places.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2865 - Add usage information, and make the results easier to read.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2864 - Add a test to compare logs. This will find tests which used to pass and now fail, and vic...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2863 - Adds a result_state (either IN, OUT, WORKING or POSTBIND), representing the   codesite-noreply at google.com
• [phc-internals] [phc commit] r2868 - Handle bin_op types where the types are not just scalars.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2867 - After contexts have been merged, get_possible_receivers would fail. This was because it r...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2869 - Add a count of new time-outs.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2870 - I broke test with loops in them before, but they became timeouts, so I hadnt noticed they...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2871 - Model get_parent_class (conservatively - don't look for the parent class).   codesite-noreply at google.com
• [phc-internals] [phc commit] r2872 - I had left out a few statement and expr types from the CFG_visitor (my own fault for not ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2873 - Collect stats for more statements and expressions and make a few other minor changes.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2874 - Since the CFG_visitor was missing methods, any class derived from it was also missign met...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2877 - Handle array_push() by pretending its the same as visit_assign_next.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2875 - The abstract count wasn't monotonic. Make it an abstract state instead, and let it satura...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2879 - --call-string-length should not be mandatory.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2878 - Add command line options to allow flow-insensitivity, object-insensitivity, and call-stri...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2876 - Wrap Points-to in a copy-on-write shell. This avoids lots of cloning, and generally resul...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2880 - Add a simple implementation of flow-sensitivity, which is that definitions can never be k...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2881 - Change call-string-length to take an integer.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2882 - Refactor out BasicTest from Basic{Parse, Optimize}Test. This will allow me to add extra si...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2883 - Fix the build when configured with --without-php.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2885 - Merge max_execution_time fix from dataflow branch.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2884 - Add some troubleshooting tips to the manual.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2886 - When we marked the RDF of a block we didn't keep going and mark the RDF's RDF (etc). Fixed.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2888 - Replace the optimize test with two, one that is "precise", and one that is "fast".   codesite-noreply at google.com
• [phc-internals] [phc commit] r2889 - Remove checks for 'diff' and 'patch', which we haven't used in a long time.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2893 - Add the start of support for the __toString handler.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2891 - Handle call-string-length argument.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2895 - Add more tests to the list of 'don't cares'.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2887 - Resolve CFG edges if we know their direction (before we only cleared up loops that were c...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2890 - This adds a warning at the end of configure for those using the svn repository. It tells ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2892 - Fix % and \ operators (they were reversed).   codesite-noreply at google.com
• [phc-internals] [phc commit] r2894 - Add in quite a few other coerce_to_string calls.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2896 - Change comments to echos to make the test more readable.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2897 - Make the string_handler test more readable.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2898 - Add an extra coerce_to_string calls, and some comments.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2899 - There were a few places where we created an array, then set a type for its values. Refact...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2916 - Unless we have full context-sensitivity, don't close the scope of a symtable.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2915 - In many cases we were re-using the fake nodes, and getting conservative results as a resu...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2904 - Add support for literal default values.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2910 - Add LCTS for 'compile-time-stats with line-number'   codesite-noreply at google.com
• [phc-internals] [phc commit] r2911 - Remove unneeded headers. Stat_collector was building unnecessarily when I was changing Po...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2905 - Give reasons why the no-Optimiz label is set.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2902 - Collect stats for BBs processed, BBs eliminated by DCE and inlinable methods.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2914 - Add the new options to phc_bash_completion.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2901 - Handle casting of arrays and objects. However, this still leaves casting to scalars, whic...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2906 - Allow assign_field to pass Optimization_transformer. This is predicated on nothing weird ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2907 - Ignore assignments to "FAKE" variables.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2900 - Ignore clone for optimization.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2919 - The scope should only be closed with an infinite call-string, which is 'call_string_lengt...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2920 - Support all types for division (the result will be integer, bool, or real).   codesite-noreply at google.com
• [phc-internals] [phc commit] r2908 - Add a comment about handlers.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2913 - Collect stats for uninitialised variables (in SSA form).   codesite-noreply at google.com
• [phc-internals] [phc commit] r2917 - unique_count does need to be reset, just not every time. Instead, we reset it as part of ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2918 - Set the call-string-length to 0 (aka infinite) by default.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2903 - Script which runs specified tests with --stats and records the results in an SQLite datab...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2912 - Fix copy-on-write Points_to graphs. I had been sharing the Points_to between multiple pla...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2921 - Rather than resetting to TOP upon a call to WPA_lattice::kill_value, remove the results f...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2909 - Add more exceptions to the labels file.   codesite-noreply at google.com
• [phc-internals] Re: PHC Dataflow   Paul Biggar
• [phc-internals] [phc commit] r2923 - Add a comment for the last commit.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2924 - Add a test for what happens when an aray index is NULL (its treated as "").   codesite-noreply at google.com
• [phc-internals] [phc commit] r2925 - Dont allow self references. They make no sense.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2902 - Collect stats for BBs processed, BBs eliminated by DCE and inlinable methods.   Paul Biggar
• [phc-internals] [phc commit] r2873 - Collect stats for more statements and expressions and make a few other minor changes.   Paul Biggar
• [phc-internals] [phc commit] r2913 - Collect stats for uninitialised variables (in SSA form).   Paul Biggar
• [phc-internals] [phc commit] r2927 - Simplify collecting stats regarding number of inlinable methods.   codesite-noreply at google.com
• [phc-internals] Comment on revision r2926 in phc   codesite-noreply at google.com
• [phc-internals] Comment on revision r2927 in phc   codesite-noreply at google.com
• [phc-internals] [phc commit] r2929 - Combine contexts which have values.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2928 - Count the number of tests missing from the previous test logs.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2930 - Start handling foreach. This does foreach_reset, foreach_has_key and foreach_get_key.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2934 - Eliminate dynamic_cast at start of visit_method_invocation () by using more appropriate v...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2936 - Minor tidy-up.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2935 - Clone abstract values without segfaulting.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2922 - Fix a number of problems involving stacks.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2932 - Minor fixes to small tests:   codesite-noreply at google.com
• [phc-internals] [phc commit] r2926 - Remove index number and print just Basic Block IDs on BBs in CFG   codesite-noreply at google.com
• [phc-internals] [phc commit] r2937 - Collect stats for total number of BBs and total number of contexts   codesite-noreply at google.com
• [phc-internals] [phc commit] r2933 - Get rid of useless check in collect_uninit_var_stats ()   codesite-noreply at google.com
• [phc-internals] [phc commit] r2938 - Collect stats for number of BBs changed in Optimization Transformer.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2931 - Handle casts to scalars (except to string, which needs to handle __toString).   codesite-noreply at google.com
• [phc-internals] [phc commit] r2939 - Add spaces after commas etc. in stats.php   codesite-noreply at google.com
• [phc-internals] [phc commit] r2940 - Consider resource to be a scalar type.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2941 - Handle reading from variable-variables. In this case, we need to be able to copy from cyc...   codesite-noreply at google.com
• [phc-internals] Re: [PHP-DEV] Re: PHC Dataflow   Paul Biggar
• [phc-internals] [phc commit] r2942 - Handle the rest of the foreach statements. The most important is foreach_get_val, which i...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2943 - There's nothing to optimize in foreach loops.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2945 - Make the output for DCE more readable.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2944 - Model is_array and is_object better, to give useful results.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2946 - This fixes If_simplification, however, it turns out that a lot of what it relies on is br...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2947 - Add the current php-benchmarks to the repository. This will allow us modify them slightly...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2948 - Add a script to run all the benchmarks through the optimizer.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2949 - Tidy up the test_benches script, and print stats.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2950 - Remove useless tests.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2951 - Model srand, max, min and in_array.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2952 - Split assign_path_by_ref to use a fake index_node.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2956 - Add more unsupported tests.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2955 - Support isset in the optimizer.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2954 - Fix reading an array index by reference, where the array variable is NIL. We need to find...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2957 - When modelling var_export, we were storing the address of a stack variable in the heap, w...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2953 - Support casting to strings if the value to be casted is known.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2958 - Its perfectly reasonable to look for the values pointed to by a variable which doesnt exist.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2959 - Mark more unsupported (optimization) test cases, mostly due to eval.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2960 - Optimizing constants is handled, so don't fail on it.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2961 - Allow debugging of the Optimization_transformer with --debug=transformer.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2968 - Foreach_get_val can be by-ref, Foreach_get_key cannot.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2962 - When merging the context in Def_use, we didn't delete the old results. This led to Optimi...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2965 - The assign_path_scalar function was too simplistic. It needs to go through copy_value ins...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2967 - Mark field accesses of static fields as TODO.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2970 - Fix test to include properly.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2969 - Handle method_invocations by reference. They required checking saved_is_ref() during back...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2963 - Don't fail in optimization of var-vars. We'll do them later, if needs be.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2971 - If we don't have a result for the constant used, get it from the Zend Engine.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2966 - Add --include to the Optimization tests.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2973 - Ignore test with __destruct.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2974 - Don't fail for assign_var_var.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2964 - Don't transform the LHS of an assign_var to a literal if the assign_var is by-ref.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2972 - If we read try to fold a string index, handle potential failure (this happens when readin...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2975 - Add more unsupported files.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2976 - Handle class statics.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2977 - Handle inheritence.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2978 - Start making test_whirl compile. It needed the directories played with, and --include to ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2980 - Minor changes to Whirl to make phc like it:   codesite-noreply at google.com
• [phc-internals] [phc commit] r2979 - Handle static_arrays as attributes.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2981 - Pass $this to parent::__construct calls.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2982 - Handle lower-case lookups of method and class names. toLower in its current form was nast...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2983 - Model a few more functions.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2984 - Don't pass the $this parameter for static methods.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2985 - During check_owner_type, we were killing the owner's value if it had unset as a possible ...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2987 - Instead of converting   codesite-noreply at google.com
• [phc-internals] [phc commit] r2986 - Don't dispatch on any scalar type, or on arrays.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2988 - I had changed the variable names before the last commit, but somehow forgot to check that...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2989 - Mark another test as no-Optimiz.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2990 - Don't consider the block after a die() or an exit() to be executable (unless its the only...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2994 - Model is_null() and pow().   codesite-noreply at google.com
• [phc-internals] Comment on revision r2995 in phc   codesite-noreply at google.com
• [phc-internals] [phc commit] r2993 - Remove unused variable.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2997 - Fix If-Simplification.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2996 - Ugh, didn't mean to make that last commit as it breaks the build, this fixes it. The com...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2995 - Up until now, we'd been creating the Def Use Web from the Def Use and using the DUW to co...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2991 - Merging the points-to graphs wasnt working (I think it hasnt been for some time), because...   codesite-noreply at google.com
• [phc-internals] [phc commit] r2992 - When a meothd is being invoked on an index_node, we can delete all non-objects from the i...   codesite-noreply at google.com
• [phc-internals] [phc commit] r3001 - Handle pass-by-ref for Internal_method_infos (the conditions were reversed).   codesite-noreply at google.com
• [phc-internals] [phc commit] r3002 - Fix minor bug in richards (probably one of a few, but it was stopping the analysis working).   codesite-noreply at google.com
• [phc-internals] [phc commit] r2999 - Fix arrows between defs and uses on CFGs.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3003 - Model array_pop.   codesite-noreply at google.com
• [phc-internals] [phc commit] r2998 - Some minor changes trying to make richards.php work.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3000 - Comment out body of If_simplification, since it's broken.   codesite-noreply at google.com
• [phc-internals] Re: PHC Dataflow   Paul Biggar
• [phc-internals] [phc commit] r3005 - Add test for array_pop.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3004 - Add a small comment.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3010 - Add a copy of the zend benchmark without recursive tests.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3008 - Fix for richards from upstream (it didnt work before, it looks like it does now).   codesite-noreply at google.com
• [phc-internals] [phc commit] r3009 - Make test_benches loop, and send a mail after every test run.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3007 - Fix includes to make raytrace easier to run.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3006 - Model method_exists.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3011 - Add other tests we should be checking.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3012 - Add a dirname to make --include work.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3013 - Fix raytracer includes (properly).   codesite-noreply at google.com
• [phc-internals] [phc commit] r3016 - Model sizeof, strval and substr.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3014 - Model sqrt. Use the word "real" instead of "float", and fix a place I had used the word "...   codesite-noreply at google.com
• [phc-internals] [phc commit] r3015 - Remove exceptions from whirl.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3017 - Model feof().   codesite-noreply at google.com
• [phc-internals] [phc commit] r3019 - Model floor() and ord().   codesite-noreply at google.com
• [phc-internals] [phc commit] r3018 - Model ob_start.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3020 - If an object's type does not support a method, that result in an error. After that point,...   codesite-noreply at google.com
• [phc-internals] [phc commit] r3022 - We dont model function static variables (though they would not be hard). Changed to a glo...   codesite-noreply at google.com
• [phc-internals] [phc commit] r3021 - Model fgets and split.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3024 - This converts Map to use a hashtable as its underlying store. I've been meaning to do thi...   codesite-noreply at google.com
• [phc-internals] [phc commit] r3023 - Model gettype.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3025 - Model rtrim.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3026 - Allow code generation for "return 5; "   codesite-noreply at google.com
• [phc-internals] [phc commit] r3027 - Add a test that rtrim gets optimized out.   codesite-noreply at google.com
• [phc-internals] [phc commit] r3028 - This disables the SSA def-use web, making DCE rely on the unfactored def-use web (which w...   codesite-noreply at google.com
• [phc-internals] [phc commit] r3029 - A long time ago, it was useful to consider the possability of a set with every possible v...   codesite-noreply at google.com
• [phc-internals] [phc commit] r3030 - If a method is pure and all of its parameters are literals, we'll just execute it instead...   codesite-noreply at google.com
• [phc-internals] [phc commit] r3031 - Convert Sets to use unordered_set. This results in about an 8% speed-up.   codesite-noreply at google.com
• [phc-internals] Re: PHC Dataflow   Paul Biggar
• [phc-internals] Alias_names   Paul Biggar
• [phc-internals] [phc commit] r3032 - Remove the operator< from Alias_name, and implement operator== in terms of str(). This sh...   codesite-noreply at google.com

Last message date: Tue Jun 30 23:25:30 IST 2009
Archived on: Tue Jun 30 23:25:38 IST 2009

• Messages sorted by: [ thread ] [ subject ] [ author ]
• More info on this list...