aboutsummaryrefslogtreecommitdiff
path: root/flow.c
Commit message (Expand)AuthorAgeFilesLines
* Unhardcode byte size being 8 bits.David Given2008-12-171-4/+6
* Fix typos in commentsJosh Triplett2007-03-091-5/+4
* Update usage chain for dead branch instruction.Christopher Li2007-01-161-1/+1
* Update usage chain for dead instructionsChristopher Li2007-01-161-0/+1
* Add instruction to pseudo user tracking.Christopher Li2007-01-161-25/+26
* Change the symbol access list to a pseudo listChristopher Li2007-01-161-4/+4
* Use the new per-instruction position information for betterLinus Torvalds2005-04-071-1/+1
* Make "check_access()" take the size of the access properlyLinus Torvalds2005-04-071-2/+2
* Add warning for accessing outside of a symbolLinus Torvalds2005-04-071-1/+19
* Split OP_SETVAL into OP_SETVAL (fp expressions and labels) and OP_SYMADDRLinus Torvalds2005-04-071-10/+5
* Remove phi source merging.Linus Torvalds2005-04-071-10/+0
* Remove unreachable bb's from the list of bb's when we kill them.Linus Torvalds2005-04-071-0/+3
* When we simplify branches-to-branches, assert that we actually foundLinus Torvalds2005-04-071-5/+8
* Mark asms as having side effects, so that we don'tLinus Torvalds2005-04-071-0/+8
* Much cleaner entrypoint dominance handling: instead of makingLinus Torvalds2005-04-071-1/+1
* Fix conditional branch to same target simplification.Linus Torvalds2005-04-071-8/+0
* Make the "entrypoint" be a special OP_ENTRY instruction instead ofLinus Torvalds2005-04-071-7/+2
* Fix thinko. If we follow a conditional branch, we should _not_Linus Torvalds2005-04-071-1/+1
* Allow conditional branch following to trigger more often.Linus Torvalds2005-04-071-17/+15
* Implement conditional branch following when simplifying branches.Linus Torvalds2005-04-071-5/+50
* Dang. Bitten by that non-canonical phi-list thing again.Linus Torvalds2005-04-071-2/+6
* Split up branch flow simplification a bit in preparation for addingLinus Torvalds2005-04-071-15/+19
* A sparse "context" instruction has side effects. Don't allowLinus Torvalds2005-04-071-0/+1
* Do real flow simplification only after liveness analysis.Linus Torvalds2005-04-071-26/+68
* Move constant switch simplification to the instructionLinus Torvalds2005-04-071-37/+0
* Associate pseudos with the symbol name whose value they got.Linus Torvalds2005-04-071-0/+2
* Start using instruction sizes properly.Linus Torvalds2005-04-071-8/+6
* Don't get confused by infinite loops. They happen, and weLinus Torvalds2005-04-071-1/+6
* Pair up removals from parents and children properly.Linus Torvalds2005-04-071-0/+1
* Remember to kill the use of the conditional if bothLinus Torvalds2005-04-071-0/+1
* Fix entrypoint branch rewriting..Linus Torvalds2005-04-071-1/+3
* When turning a conditional branch into an unconditional one,Linus Torvalds2005-04-071-1/+5
* Clean up bb merging a bit. Still chasing "the bug".Linus Torvalds2005-04-071-9/+5
* Check switch and computed goto target lists too when verifying.Linus Torvalds2005-04-071-1/+7
* Add a flow verification thing.Linus Torvalds2005-04-071-0/+64
* Be more careful when rewriting parent branches.Linus Torvalds2005-04-071-7/+9
* Allow CSE to run after bb packing. Linus Torvalds2005-04-071-0/+18
* Make list-ptr remove/replace take a count.Linus Torvalds2005-04-071-4/+4
* Leave symbol pseudo usage intact when doing phi-node conversion.Linus Torvalds2005-04-071-1/+3
* We had a big comment about destroying the use chains. So fix it already.Linus Torvalds2005-04-071-10/+1
* Expose "dominates()" function for memop domination checking.Linus Torvalds2005-04-071-8/+7
* Export the load instruction conversion functions.Linus Torvalds2005-04-071-6/+6
* Fix check for "local" symbols.Linus Torvalds2005-04-071-2/+2
* Nonlocal memops are non-determinedly dominated by the entry.Linus Torvalds2005-04-071-0/+3
* Be a lot more careful when re-writing branches.Linus Torvalds2005-04-071-11/+3
* Fix total braino in empty block packing.Linus Torvalds2005-04-071-3/+1
* Use "has_use_list()" to determine whether to clear it on killingLinus Torvalds2005-04-071-6/+4
* Be more careful about insn->bb pointers.Linus Torvalds2005-04-071-5/+1
* When packing basic blocks, update the instructions bb pointers.Linus Torvalds2005-04-071-2/+8
* Make the CSE "repeat" logic be more fine-grained than justLinus Torvalds2005-04-071-0/+5