aboutsummaryrefslogtreecommitdiff
path: root/lib
AgeCommit message (Collapse)Author
2008-08-26Put a heuristic in place to prevent GVN from falling into bad cases with ↵Owen Anderson
massively complicated CFGs. This speeds up a particular testcase from 12+ hours to 5 seconds with little perceptible loss of quality. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55391 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26Optimize SelectionDAG's topological sort to use one pass insteadDan Gohman
of two, and to not need a scratch std::vector. Also, use the SelectionDAG's topological sort in LegalizeDAG instead of having a separate implementation. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55389 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26Refactor the bitcast code into its own function.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55387 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26Make FastISel use the correct argument type when casting GEP indices.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55384 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26Don't select binary instructions with illegal types.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55383 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26Add support for fast isel of sitofp, and remove some unnecessary and ↵Owen Anderson
imprecise legality checks. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55381 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26Use a combination of copyRegToReg and ISD::BIT_CONVERT when doing fast isel ↵Owen Anderson
of bitcasts, allowing it to support the full range of conversions people might ask for in a correct manner. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55378 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26These assertions should be return false's instead, allowing the client to ↵Owen Anderson
detect the failure. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55377 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26Make TargetInstrInfo::copyRegToReg return a bool indicating whether the copy ↵Owen Anderson
requested was inserted or not. This allows bitcast in fast isel to properly handle the case where an appropriate reg-to-reg copy is not available. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55375 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26If IV is used in a int-to-float cast inside the loop then try to eliminate ↵Devang Patel
the cast operation. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55374 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26Add support for fast isel of non-constant fptosi instructions.Owen Anderson
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55373 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26If an xmm register is referenced explicitly in an inline asm, make sure to Chris Lattner
assign it to a version of the xmm register with the regclass that matches its type. This fixes PR2715, a bug handling some crazy xpcom case in mozilla. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55358 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26typo fix.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55355 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26Actually recycle SDNode allocations. SelectionDAG is usingDan Gohman
RecyclingAllocator, but this change is needed for the nodes to actually be recycled. This cuts SelectionDAG's memory usage high-water-mark in half in some cases. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55351 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26Avoid a warning about isTargetNullPtr being unused in release builds.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55350 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26This is done.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55348 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25Add a RetVT parameter to emitted FastISel methods, so that we will be able ↵Owen Anderson
to pass the desired return type down. This is not currently used. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55345 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25Implement 32 & 64 bit versions of PPC atomicDale Johannesen
binary primitives. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55343 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25Unbreak build.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55342 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-2580 col. violations.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55341 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25Expand bitcast support in fast isel to support bitcasts of non-constant ↵Owen Anderson
values by emitting reg-reg copies. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55340 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25Try approach to moving call address load inside of callseq_start. Now it's ↵Evan Cheng
done during the preprocess of x86 isel. callseq_start's chain is changed to load's chain node; while load's chain is the last of callseq_start or the loads or copytoreg nodes inserted to move arguments to the right spot. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55338 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25Remove PPC-specific lowering for atomics; theDale Johannesen
generic stuff works fine. Mark rewritten cmp-and-swap as not using CR1. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55336 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25Add support for fast isel of (integer) immediate materialization pattens, ↵Owen Anderson
and use them to support bitcast of constants in fast isel. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55325 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25It's important for the cmp-and-swap to balanceDale Johannesen
loads and stores but it's even more important for it to store the right value.:( git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55319 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25Nevermind. This broke the bootstrap (?!).Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55318 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25MOVQ2DQ and MOVQ2DQ use SSE2. We should conditionalize the use of theseBill Wendling
instructions on having SSE2. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55317 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25Fix pasto (values -> types) noticed by Josh ConnerChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55315 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25flush stream after dumping.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55301 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-25Fix asm printing of MOVSDto64mr and MOV64toSDrm.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55300 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-24Temporarily reverting r55292. It's causing a bootstraping failure:Bill Wendling
/Volumes/Sandbox/Buildbot/llvm/full-llvm/build/llvm-gcc.obj/./gcc/xgcc ... src/libiberty/make-temp-file.c -o make-temp-file.o Assertion failed: (Node2Index[SU->NodeNum] > Node2Index[I->Dep->NodeNum] && "Wrong topological sorting"), function InitDAGTopologicalSorting, file /Volumes/Sandbox/Buildbot/llvm/full-llvm/build/llvm.src/lib/CodeGen/SelectionDAG/ScheduleDAGRRList.cpp, line 508. ../../../../llvm-gcc.src/libiberty/hashtab.c:955: internal compiler error: Abort trap Please submit a full bug report, with preprocessed source if appropriate. See <URL:http://developer.apple.com/bugreporter> for instructions. make[4]: *** [hashtab.o] Error 1 make[4]: *** Waiting for unfinished jobs.... make[3]: *** [multi-do] Error 1 make[2]: *** [all] Error 2 make[1]: *** [all-target-libiberty] Error 2 make: *** [all] Error 2 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55295 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-24remove the type checking logic already done by tblgen, just keep theChris Lattner
parts tblgen doesn't do (checking for constant, alloca, enclosing function has gc marker). This passes dj! :) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55294 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-24simplify PseudoSourceValue printing a bit. Unnest all of ↵Chris Lattner
PseudoSourceValue.cpp from the llvm namespace. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55293 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-24Move callseq_start above the call address load to allow load to be folded ↵Evan Cheng
into the call node. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55292 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-24Print PseudoSourceValue.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55291 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-24add a hack to temporarily allow pseudo source values.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55289 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-24make sure to flush the stream after dumping, to make sure it goes out ↵Chris Lattner
immediately. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55288 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-24Use additionnal include directory instead of ../ in #include.Cedric Venet
Suggested by aKor. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55282 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-24Updating VC++ project.Cedric Venet
Adding one include file and correct one declaration from class to struct in order to make llvm compile on VC2005. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55279 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23improve encapsulation of the BBExecutable set.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55271 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23Switch an assortment of maps, sets and vectors to more efficient versions,Chris Lattner
patch contributed by m-s! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55270 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23get MachineConstantPool off std::ostream, onto raw_ostream. It would beChris Lattner
really nice if someone converted MachineFunction::print to raw_ostream. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55268 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23minor cleanups.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55267 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23Add raw_stream adaptors that write into an std::string and ↵Chris Lattner
SmallVector/SmallString. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55265 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23Switch the asmprinter (.ll) and all the stuff it requires over toChris Lattner
use raw_ostream instead of std::ostream. Among other goodness, this speeds up llvm-dis of kc++ with a release build from 0.85s to 0.49s (88% faster). Other interesting changes: 1) This makes Value::print be non-virtual. 2) AP[S]Int and ConstantRange can no longer print to ostream directly, use raw_ostream instead. 3) This fixes a bug in raw_os_ostream where it didn't flush itself when destroyed. 4) This adds a new SDNode::print method, instead of only allowing "dump". A lot of APIs have both std::ostream and raw_ostream versions, it would be useful to go through and systematically anihilate the std::ostream versions. This passes dejagnu, but there may be minor fallout, plz let me know if so and I'll fix it. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55263 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23Add #includeChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55258 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23add #includeChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55257 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23add #includeChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55252 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23add a simple mechanism for formatted output. This gives raw_ostream's Chris Lattner
all the power and risk of fprintf format strings. Use them like this: OS << format("%10.4f", 42.0) << "\n" << format("%x", 42) << '\n'; git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55246 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23Provide a 64 bit variant of mmx.maskmovq intrinsic lowering.Anton Korobeynikov
Is there way to avoid explicit target check? git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55238 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-09-22 03:02:47 +0000