aboutsummaryrefslogtreecommitdiff
path: root/lib/CodeGen/SelectionDAG
AgeCommit message (Collapse)Author
2007-09-27Silence a compiler warning.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42389 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-27Boogs.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42388 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-27Be smarter about which node to force schedule. Reduce # of duplications + ↵Evan Cheng
copies; Added statistics. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42387 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-27Backtracking only when it won't create a cycle.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42384 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-26- Move getPhysicalRegisterRegClass() from ScheduleDAG to MRegisterInfo.Evan Cheng
- Added ability to emit cross class register copies to the BBRU scheduler. - More aggressive backtracking. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42375 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-26Enable codegen for long double abs, sin, cosDale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42368 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-26Fix f80 UNDEF.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42359 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-26Allow copyRegToReg to emit cross register classes copies.Evan Cheng
Tested with "make check"! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42346 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-25Move the setOperationAction(ISD::DEBUG_LOC, MVT::Other, Expand) andDan Gohman
the check to see if the assembler supports .loc from X86TargetLowering into the superclass TargetLowering. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42297 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-25Added major new capabilities to scheduler (only BURR for now) to support ↵Evan Cheng
physical register dependency. The BURR scheduler can now backtrace and duplicate instructions in order to avoid "expensive / impossible to copy" values (e.g. status flag EFLAGS for x86) from being clobbered. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42284 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-24Use the correct result value type instead of using getValueType(0)Dan Gohman
in ExpandEXTRACT_VECTOR_ELT and SplitVectorOp. This fixes an abort in the included testcase. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42264 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-22initialize isstore/isload fields in ctor, fixing PR1695Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42222 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-21Change APFloat::convertFromInteger to take the incomingDale Johannesen
bit width instead of number of words allocated, which makes it actually work for int->APF conversions. Adjust callers. Add const to one of the APInt constructors to prevent surprising match when called with const argument. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42210 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-21initialize SetCCResultContents, fixing PR1693Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42193 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-19More long double fixes. x86_64 should build now.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42155 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-19Fix longdouble -> uint conversion.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42143 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-19Use struct SDep instead of std::pair for SUnit pred and succ lists. First stepEvan Cheng
in tracking physical register output dependencies. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42125 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-18Fix a bogus splat xform:Evan Cheng
shuffle <undef, undef, x, undef>, <undef, undef, undef, undef>, <2, 2, 2, 2> != <undef, undef, x, undef> git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42111 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-18Prevent crash on long double.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42103 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-17Do not hide APInt::dump() inside #ifndef NDEBUG.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42068 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-17This is not ideal but unbreaks build failure.Devang Patel
APInt::dump() is inside #ifndef NDEBUG, however SelectionDAG dump() routines are not. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42047 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-16Adjust per revew comments.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42002 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-14Remove the assumption that FP's are either float orDale Johannesen
double from some of the many places in the optimizers it appears, and do something reasonable with x86 long double. Make APInt::dump() public, remove newline, use it to dump ConstantSDNode's. Allow APFloats in FoldingSet. Expand X86 backend handling of long doubles (conversions to/from int, mostly). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41967 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-13Fix build problems on Cygwin (PR1652), patch by Patrick Walton.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41923 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-13Bug fixes.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41900 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-12Remove dead code.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41899 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-12Yet another getTargetNode variant.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41898 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-12Revise previous patch per review comments.Dale Johannesen
Next round of x87 long double stuff. Getting close now, basically works. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41875 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-11Add APInt interfaces to APFloat (allows directlyDale Johannesen
access to bits). Use them in place of float and double interfaces where appropriate. First bits of x86 long double constants handling (untested, probably does not work). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41858 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-11Fold the adjust_trampoline intrinsic intoDuncan Sands
init_trampoline. There is now only one trampoline intrinsic. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41841 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-10Emit:Chris Lattner
cmpl %eax, %ecx setae %al movzbl %al, %eax instead of: cmpl %eax, %ecx setb %al xorb $1, %al movzbl %al, %eax when using logical not of a C comparison. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41807 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-101. Don't call Value::getName(), which is slow.Chris Lattner
2. Lower calls to fabs and friends to FABS nodes etc unless the function has internal linkage. Before we wouldn't lower if it had a definition, which is incorrect. This allows us to compile: define double @fabs(double %f) { %tmp2 = tail call double @fabs( double %f ) ret double %tmp2 } into: _fabs: fabs f1, f1 blr git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41805 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-08Implement misaligned FP loads and stores.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41786 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-07Add support for having different alignment for objects on call frames.Rafael Espindola
The x86-64 ABI states that objects passed on the stack have 8 byte alignment. Implement that. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41768 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-07Split eh.select / eh.typeid.for intrinsics into i32/i64 versions. This is ↵Anton Korobeynikov
needed, because they just "mark" register liveins and we let frontend solve type issue, not lowering code :) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41763 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-07Add lengthof and endof templates that hide a lot of sizeof computations.Owen Anderson
Patch by Sterling Stein! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41758 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-06Next round of APFloat changes.Dale Johannesen
Use APFloat in UpgradeParser and AsmParser. Change all references to ConstantFP to use the APFloat interface rather than double. Remove the ConstantFP double interfaces. Use APFloat functions for constant folding arithmetic and comparisons. (There are still way too many places APFloat is just a wrapper around host float/double, but we're getting there.) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41747 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-05Fix PR1628. When exception handling is turned on,Duncan Sands
labels are generated bracketing each call (not just invokes). This is used to generate entries in the exception table required by the C++ personality. However it gets in the way of tail-merging. This patch solves the problem by no longer placing labels around ordinary calls. Instead we generate entries in the exception table that cover every instruction in the function that wasn't covered by an invoke range (the range given by the labels around the invoke). As an optimization, such entries are only generated for parts of the function that contain a call, since for the moment those are the only instructions that can throw an exception [1]. As a happy consequence, we now get a smaller exception table, since the same region can cover many calls. While there, I also implemented folding of invoke ranges - successive ranges are merged when safe to do so. Finally, if a selector contains only a cleanup, there's a special shorthand for it - place a 0 in the call-site entry. I implemented this while there. As a result, the exception table output (excluding filters) is now optimal - it cannot be made smaller [2]. The problem with throw filters is that folding them optimally is hard, and the benefit of folding them is minimal. [1] I tested that having trapping instructions (eg divide by zero) in such a region doesn't cause trouble. [2] It could be made smaller with the help of higher layers, eg by having branch folding reorder basic blocks ending in invokes with the same landing pad so they follow each other. I don't know if this is worth doing. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41718 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-04Fix for PR1632. EHSELECTION always produces a i32 value.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41712 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-31Add mod, copysign, abs operations to APFloat.Dale Johannesen
Implement some constant folding in SelectionDAG and DAGCombiner using APFloat. Remove double versions of constructor and getValue from ConstantFPSDNode. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41664 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-31Revise per review of previous patch.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41645 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-31Enhance APFloat to retain bits of NaNs (fixes oggenc).Dale Johannesen
Use APFloat interfaces for more references, mostly of ConstantFPSDNode. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41632 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-30Change LegalFPImmediates to use APFloat.Dale Johannesen
Add APFloat interfaces to ConstantFP, SelectionDAG. Fix integer bit in double->APFloat conversion. Convert LegalizeDAG to use APFloat interface in ConstantFPSDNode uses. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41587 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-29Fix use of declaration inside case blockAnton Korobeynikov
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41584 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-29Lower FRAME_TO_ADDR_OFFSET to zero by default (if not custom lowered)Anton Korobeynikov
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41578 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-28Add an option, -view-sunit-dags, for viewing the actual SUnit DAGs used byDan Gohman
scheduling. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41556 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-27Make DAGCombiner's global alias analysis query more precise in the caseDan Gohman
where both pointers have non-zero offsets. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41491 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-27If the source and destination pointers in an llvm.memmove are knownDan Gohman
to not alias each other, it can be translated as an llvm.memcpy. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41489 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-27There is an impedance matching problem between LLVM andDuncan Sands
gcc exception handling: if an exception unwinds through an invoke, then execution must branch to the invoke's unwind target. We previously tried to enforce this by appending a cleanup action to every selector, however this does not always work correctly due to an optimization in the C++ unwinding runtime: if only cleanups would be run while unwinding an exception, then the program just terminates without actually executing the cleanups, as invoke semantics would require. I was hoping this wouldn't be a problem, but in fact it turns out to be the cause of all the remaining failures in the LLVM testsuite (these also fail with -enable-correct-eh-support, so turning on -enable-eh didn't make things worse!). Instead we need to append a full-blown catch-all to the end of each selector. The correct way of doing this depends on the personality function, i.e. it is language dependent, so can only be done by gcc. Thus this patch which generalizes the eh.selector intrinsic so that it can handle all possible kinds of action table entries (before it didn't accomodate cleanups): now 0 indicates a cleanup, and filters have to be specified using the number of type infos plus one rather than the number of type infos. Related gcc patches will cause Ada to pass a cleanup (0) to force the selector to always fire, while C++ will use a C++ catch-all (null). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41484 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-26Revise per review comments.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41409 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2026-03-26 09:58:07 +0000