aboutsummaryrefslogtreecommitdiff
path: root/lib/CodeGen
AgeCommit message (Collapse)Author
2008-11-21Rename "ADDO" to "SADDO" and "UADDO". The "UADDO" isn't equivalent to "ADDC"Bill Wendling
because the boolean it returns to indicate an overflow may not be treated like as a flag. It could be stored to memory, for instance. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59780 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-21Implement the sadd_with_overflow intrinsic. This is converted intoBill Wendling
"ISD::ADDO". ISD::ADDO is lowered into a target-independent form that does the addition and then checks if the result is less than one of the operands. (If it is, then there was an overflow.) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59779 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-21Use ComputeLatency in the MachineInstr scheduler.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59777 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-21Remove the CycleBound computation code from the ScheduleDAGRRListDan Gohman
schedulers. This doesn't have much immediate impact because targets that use these schedulers by default don't yet provide pipeline information. This code also didn't have the benefit of register pressure information. Also, removing it will avoid problems with list-burr suddenly starting to do latency-oriented scheduling on x86 when we start providing pipeline data, which would increase spilling. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59775 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-21Implement ComputeLatency for MachineInstr ScheduleDAGs. FactorDan Gohman
some of the latency computation logic out of the SDNode ScheduleDAG code into a TargetInstrItineraries helper method to help with this. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59761 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-21Add UADDO and SADDO nodes. These will be used for determining an overflowBill Wendling
condition in an addition operation. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59760 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-21Change these schedulers to not emit no-ops. It turns out thatDan Gohman
the RR scheduler actually does look at latency values, but it doesn't use a hazard recognizer so it has no way to know when a no-op is needed, as opposed to just stalling and incrementing the cycle count. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59759 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-20Treat mid-block labels the same as terminators when building theDan Gohman
MachineInstr scheduling DAG, meaning they implicitly depend on all preceding defs. This fixes Benchmarks/Shootout-C++/except and Regression/C++/EH/simple_rethrow in -relocation-model=pic -disable-post-RA-scheduler=false mode. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59747 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-20Add another machine-code printing pass when post-pass scheduling is run.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59746 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-20Add some documentation.Duncan Sands
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59727 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-2080-column violation.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59718 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-20Remove a remnant of list-burr's fast mode.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59702 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-20Factor out the SethiUllman numbering logic from the list-burr andDan Gohman
list-tdrr schedulers into a common base class. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59701 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-20Remove the "fast" form of the list-burr scheduler, and use theDan Gohman
dedicated "fast" scheduler in -fast mode instead, which is faster. This speeds up llc -fast by a few percent on some testcases -- the speedup only happens for code not handled by fast-isel. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59700 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-20Facter AddPseudoTwoAddrDeps and associated infrasructure out ofDan Gohman
the list-burr scheduler so that it can be used by the list-tdrr scheduler too. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59698 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-20- Register scavenger should use MachineRegisterInfo and internal map to find ↵Evan Cheng
the first use of a register after a given machine instruction. - When scavenging a register, in addition to the spill, insert a restore before the first use. - Abort if client is looking to scavenge a register even when a previously scavenged register is still live. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59697 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-20Add #include <climits> to get the definition of INT_MAX.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59692 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-20Factor out the code for verifying the work of the scheduler,Dan Gohman
extend it a bit, and make use of it in all schedulers, to ensure consistent checking. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59689 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-19Simplify this code a little. In the fast scheduler, CreateNewSUnitDan Gohman
and CreateClone don't add any extra value. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59679 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-19Experimental post-pass scheduling support. Post-pass schedulingDan Gohman
is currently off by default, and can be enabled with -disable-post-RA-scheduler=false. This doesn't have a significant impact on most code yet because it doesn't yet do anything to address anti-dependencies and it doesn't attempt to disambiguate memory references. Also, several popular targets don't have pipeline descriptions yet. The majority of the changes here are splitting the SelectionDAG-specific code out of ScheduleDAG, so that ScheduleDAG can be moved to libLLVMCodeGen.a. The interface between ScheduleDAG-using code and the rest of the scheduling code is somewhat rough and will evolve. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59676 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-19Move the code for printing a graph node label for an SUnit intoDan Gohman
a virtual method of SelectionDAG. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59667 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-19Convert SUnit's dump method into a print method and implementDan Gohman
dump in terms of it. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59665 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-19Fix compilation error on MSVC.Argyrios Kyrtzidis
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59629 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-19Add support for rematerialization in pre-alloc-splitting.Owen Anderson
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59587 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-19Rearrange code to reduce the nesting level. No functionality change.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59580 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-19Fix debug printing of flagged SDNodes in SUnits so that theyDan Gohman
print in the correct order. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59567 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Make the same change to RegScavenger::backward.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59566 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18We also need to keep the operand index for two address check.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59562 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Register scavenger should process early clobber defs first. A dead early ↵Evan Cheng
clobber def should not interfere with a normal def which happens one slot later. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59559 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Tidy up ScheduleNodeBottomUp methods, and make them moreDan Gohman
consistent with ScheduleNodeTopDown methods. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59550 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Update a comment to reflect the current code.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59549 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Remove integer promotion support for FP_EXTENDDuncan Sands
and FP_ROUND. Not sure what these were doing here - probably they were sometimes (wrongly) created with integer operands somewhere that has since been fixed. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59548 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Simplify code using helper routines. There is notDuncan Sands
supposed to be any functionality change. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59545 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Fix a bug introduced by my previous patch. With this change, SPEC is now ↵Owen Anderson
clean with prealloc splitting enabled. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59544 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Add more const qualifiers. This fixes build breakage from r59540.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59542 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Make some methods const.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59540 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Whitespace cleanups.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59532 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18LegalizeTypes support for splitting and scalarizingDuncan Sands
SCALAR_TO_VECTOR. I didn't add the testcase, because once llc gets past scalar-to-vector it hits a SPU target lowering bug and explodes. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59530 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Rename stackprotector_create intrinsic to stackprotector.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59519 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Reapply r59464, this time using the correct typeDuncan Sands
when softening FNEG. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59513 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Remove the stackprotector_check intrinsic. Use a volatile load instead.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59504 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18- Use "moveAfter" instead of "remove/insert" of a basic block.Bill Wendling
- Use less indentation in coding. - Shorten description. - Update comments. - Move code around git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59496 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Fix a typo in a comment.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59489 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Change SUnit's dump method to take a ScheduleDAG* instead ofDan Gohman
a SelectionDAG*. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59488 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Revert r59464. It was causing this failure:Bill Wendling
Running /Volumes/Sandbox/Buildbot/llvm/full-llvm/build/llvm.src/test/CodeGen/XCore/dg.exp ... FAIL: /Volumes/Sandbox/Buildbot/llvm/full-llvm/build/llvm.src/test/CodeGen/XCore/fneg.ll Failed with signal(SIGABRT) at line 1 while running: llvm-as < /Volumes/Sandbox/Buildbot/llvm/full-llvm/build/llvm.src/test/CodeGen/XCore/fneg.ll | llc -march=xcore > fneg.ll.tmp1.s Assertion failed: (VT.isFloatingPoint() && "Cannot create integer FP constant!"), function getConstantFP, file /Volumes/Sandbox/Buildbot/llvm/full-llvm/build/llvm.src/lib/CodeGen/SelectionDAG/SelectionDAG.cpp, line 913. 0 llc 0x0092115c _ZN4llvm3sys18RemoveFileOnSignalERKNS0_4PathEPSs + 844 1 libSystem.B.dylib 0x9217809b _sigtramp + 43 2 ??? 0xffffffff 0x0 + 4294967295 3 libSystem.B.dylib 0x921f0ec2 raise + 26 4 libSystem.B.dylib 0x9220047f abort + 73 5 libSystem.B.dylib 0x921f2063 __assert_rtn + 101 6 llc 0x005a5b0a _ZN4llvm12SelectionDAG13getConmake[1]: *** [check-local] Error 1 make: *** [check] Error 2 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59487 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18Avoid using a loop in ReleasePred and ReleaseSucc methods to compute theDan Gohman
new CycleBound value. Instead, just update CycleBound on each call. Also, make ReleasePred and ReleaseSucc methods more consistent accross the various schedulers. This also happens to make ScheduleDAGRRList's CycleBound computation somewhat more interesting, though it still doesn't have any noticeable effect, because no current targets that use the register-pressure reduction scheduler provide pipeline models. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59475 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-17Eliminate some trivial differences between the ScheduleNodeTopDownDan Gohman
functions in these two schedulers. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59465 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-17Add soft float support for a bunch more operations. OriginalDuncan Sands
patch by Richard Osborne, tweaked and extended by your humble servant. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59464 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-17Don't bother doing latency calculations in the "fast" scheduler.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59461 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-17Use SUnit's CycleBound field instead of duplicating it inDan Gohman
a side-car datastructure git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59458 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-18-g5258 (git 2.32.0) at 2025-07-19 14:38:42 +0000