aboutsummaryrefslogtreecommitdiff
path: root/lib/Transforms/Scalar/LoopStrengthReduce.cpp
AgeCommit message (Collapse)Author
2007-01-15Eliminate calls to isInteger, generalizing code and tightening checks as needed.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@33218 91177308-0d34-0410-b5e6-96231b3b80d8
2007-01-08For PR1097:Reid Spencer
Enable complex addressing modes on 64-bit platforms involving two induction variables by keeping a size and scale in 64-bits not 32. Patch by Dan Gohman. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@33011 91177308-0d34-0410-b5e6-96231b3b80d8
2007-01-06no need to worry about int vs uint any more.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32946 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-31For PR950:Reid Spencer
This patch replaces signed integer types with signless ones: 1. [US]Byte -> Int8 2. [U]Short -> Int16 3. [U]Int -> Int32 4. [U]Long -> Int64. 5. Removal of isSigned, isUnsigned, getSignedVersion, getUnsignedVersion and other methods related to signedness. In a few places this warranted identifying the signedness information from other sources. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32785 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-23For PR950:Reid Spencer
This patch removes the SetCC instructions and replaces them with the ICmp and FCmp instructions. The SetCondInst instruction has been removed and been replaced with ICmpInst and FCmpInst. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32751 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-19Switch over Transforms/Scalar to use the STATISTIC macro. For each statisticChris Lattner
converted, we lose a static initializer. This also allows GCC to emit warnings about unused statistics. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32690 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-13Change the interface to SCEVExpander::InsertCastOfTo to take a cast opcodeReid Spencer
so the decision of which opcode to use is pushed upward to the caller. Adjust the callers to pass the expected opcode. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32535 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-12Change inferred getCast into specific getCast. Passes all tests.Reid Spencer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32469 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-07Changed llvm_ostream et all to OStream. llvm_cerr, llvm_cout, llvm_null, areBill Wendling
now cerr, cout, and NullStream resp. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32298 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-06Detemplatize the Statistic class. The only type it is instantiated withChris Lattner
is 'unsigned'. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32279 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-27For PR950:Reid Spencer
The long awaited CAST patch. This introduces 12 new instructions into LLVM to replace the cast instruction. Corresponding changes throughout LLVM are provided. This passes llvm-test, llvm/test, and SPEC CPUINT2000 with the exception of 175.vpr which fails only on a slight floating point output difference. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31931 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-26Removed #include <iostream> and replaced with llvm_* streams.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31923 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-17If an indvar with a variable stride is used by the exit condition, go aheadChris Lattner
and handle it like constant stride vars. This fixes some bad codegen in variable stride cases. For example, it compiles this: void foo(int k, int i) { for (k=i+i; k <= 8192; k+=i) flags2[k] = 0; } to: LBB1_1: #bb.preheader movl %eax, %ecx addl %ecx, %ecx movl L_flags2$non_lazy_ptr, %edx LBB1_2: #bb movb $0, (%edx,%ecx) addl %eax, %ecx cmpl $8192, %ecx jle LBB1_2 #bb LBB1_5: #return ret or (if the array is local and we are in dynamic-nonpic or static mode): LBB3_2: #bb movb $0, _flags2(%ecx) addl %eax, %ecx cmpl $8192, %ecx jle LBB3_2 #bb and: lis r2, ha16(L_flags2$non_lazy_ptr) lwz r2, lo16(L_flags2$non_lazy_ptr)(r2) slwi r3, r4, 1 LBB1_2: ;bb li r5, 0 add r6, r4, r3 stbx r5, r2, r3 cmpwi cr0, r6, 8192 bgt cr0, LBB1_5 ;return instead of: leal (%eax,%eax,2), %ecx movl %eax, %edx addl %edx, %edx addl L_flags2$non_lazy_ptr, %edx xorl %esi, %esi LBB1_2: #bb movb $0, (%edx,%esi) movl %eax, %edi addl %esi, %edi addl %ecx, %esi cmpl $8192, %esi jg LBB1_5 #return and: lis r2, ha16(L_flags2$non_lazy_ptr) lwz r2, lo16(L_flags2$non_lazy_ptr)(r2) mulli r3, r4, 3 slwi r5, r4, 1 li r6, 0 add r2, r2, r5 LBB1_2: ;bb li r5, 0 add r7, r3, r6 stbx r5, r2, r6 add r6, r4, r6 cmpwi cr0, r7, 8192 ble cr0, LBB1_2 ;bb This speeds up Benchmarks/Shootout/sieve from 8.533s to 6.464s and implements LoopStrengthReduce/var_stride_used_by_compare.ll git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31809 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-02For PR786:Reid Spencer
Turn on -Wunused and -Wno-unused-parameter. Clean up most of the resulting fall out by removing unused variables. Remaining warnings have to do with unused functions (I didn't want to delete code without review) and unused variables in generated code. Maintainers should clean up the remaining issues when they see them. All changes pass DejaGnu tests and Olden. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31380 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-28break edges more intelligentlyChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31257 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-28prepare for a change I'm about to makeChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31248 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-20For PR950:Reid Spencer
This patch implements the first increment for the Signless Types feature. All changes pertain to removing the ConstantSInt and ConstantUInt classes in favor of just using ConstantInt. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31063 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-27eliminate RegisterOpt. It does the same thing as RegisterPass.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29925 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-27s|llvm/Support/Visibility.h|llvm/Support/Compiler.h|Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29911 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-03Changes:Chris Lattner
1. Update an obsolete comment. 2. Make the sorting by base an explicit (though still N^2) step, so that the code is more clear on what it is doing. 3. Partition uses so that uses inside the loop are handled before uses outside the loop. Note that none of these changes currently changes the code inserted by LSR, but they are a stepping stone to getting there. This code is the result of some crazy pair programming with Nate. :) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29493 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-18Only reuse a previous IV if it would not require a type conversion.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29186 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-28Use hidden visibility to make symbols in an anonymous namespace getChris Lattner
dropped. This shrinks libllvmgcc.dylib another 67K git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28975 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-09RewriteExpr, either the new PHI node of induction variable or theEvan Cheng
post-increment value, should be first cast to the appropriated type (to the type of the common expr). Otherwise, the rewrite of a use based on (common + iv) may end up with an incorrect type. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28735 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-12Get rid of a signed/unsigned compare warning.Reid Spencer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27625 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-24Fix spelloChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27052 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-22silence a bogus gcc warningChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26953 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-18- Fixed a bogus if condition.Evan Cheng
- Added more debugging info. - Allow reuse of IV of negative stride. e.g. -4 stride == 2 * iv of -2 stride. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26841 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-18Sort StrideOrder so we can process the smallest strides first. This allowsEvan Cheng
for more IV reuses. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26837 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-17Allow users of iv / stride to be rewritten with expression that is a multiplyEvan Cheng
of a smaller stride even if they have a common loop invariant expression part. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26828 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-16For each loop, keep track of all the IV expressions inserted indexed byEvan Cheng
stride. For a set of uses of the IV of a stride which is a multiple of another stride, do not insert a new IV expression. Rather, reuse the previous IV and rewrite the uses as uses of IV expression multiplied by the factor. e.g. x = 0 ...; x ++ y = 0 ...; y += 4 then use of y can be rewritten as use of 4*x for x86. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26803 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-13Added target lowering hooks which LSR consults to make more intelligentEvan Cheng
transformation decisions. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26738 91177308-0d34-0410-b5e6-96231b3b80d8
2006-02-04Use SCEVExpander::InsertCastOfTo instead of our own code. This reducesChris Lattner
#LLVM LOC, and auto-cse's cast instructions. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@25974 91177308-0d34-0410-b5e6-96231b3b80d8
2006-02-04Fix two significant bugs in LSR:Chris Lattner
1. When rewriting code in outer loops, sometimes we would insert code into inner loops that is invariant in that loop. 2. Notice that 4*(2+x) is 8+4*x and use that to simplify expressions. This is a performance neutral change. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@25964 91177308-0d34-0410-b5e6-96231b3b80d8
2006-01-22Make iostream #inclusion explicitChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@25514 91177308-0d34-0410-b5e6-96231b3b80d8
2006-01-11Switch these to using ETForest instead of DominatorSet to compute itself.Chris Lattner
Patch written by Daniel Berlin! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@25202 91177308-0d34-0410-b5e6-96231b3b80d8
2005-12-05getRawValue zero extens for unsigned values, use getsextvalue so that weChris Lattner
know that small negative values fit into the immediate field of addressing modes. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24608 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-21My previous patch was too conservative. Reject FP and void types, but doChris Lattner
allow pointer types. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23859 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-20Do NOT touch FP ops with LSR. This fixes a testcase Nate sent me from anChris Lattner
inner loop like this: LBB_RateConvertMono8AltiVec_2: ; no_exit lis r2, ha16(.CPI_RateConvertMono8AltiVec_0) lfs f3, lo16(.CPI_RateConvertMono8AltiVec_0)(r2) fmr f3, f3 fadd f0, f2, f0 fadd f3, f0, f3 fcmpu cr0, f3, f1 bge cr0, LBB_RateConvertMono8AltiVec_2 ; no_exit to an inner loop like this: LBB_RateConvertMono8AltiVec_1: ; no_exit fsub f2, f2, f1 fcmpu cr0, f2, f1 fmr f0, f2 bge cr0, LBB_RateConvertMono8AltiVec_1 ; no_exit Doh! good catch! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23838 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-11Fix (hopefully the last) issue where LSR is nondeterminstic. When pullingChris Lattner
out CSE's of base expressions it could build a result whose order was nondet. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23698 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-11Fix another problem where LSR was being nondeterminstic. Also remove elementsChris Lattner
from the end of a vector instead of the beginning git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23697 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-11Fix another lsr-is-nondeterministic caseChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23695 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-09Hrm, you didn't see this.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23673 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-09Fix a source of non-determinism in the backend: the order of processingChris Lattner
IV strides dependend on the pointer order of the strides in memory. Non-determinism is bad. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23672 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-03Make IVUseShouldUsePostIncValue more aggressive when the use is a PHI. InChris Lattner
particular, it should realize that phi's use their values in the pred block not the phi block itself. This change turns our em3d loop from this: _test: cmpwi cr0, r4, 0 bgt cr0, LBB_test_2 ; entry.no_exit_crit_edge LBB_test_1: ; entry.loopexit_crit_edge li r2, 0 b LBB_test_6 ; loopexit LBB_test_2: ; entry.no_exit_crit_edge li r6, 0 LBB_test_3: ; no_exit or r2, r6, r6 lwz r6, 0(r3) cmpw cr0, r6, r5 beq cr0, LBB_test_6 ; loopexit LBB_test_4: ; endif addi r3, r3, 4 addi r6, r2, 1 cmpw cr0, r6, r4 blt cr0, LBB_test_3 ; no_exit LBB_test_5: ; endif.loopexit.loopexit_crit_edge addi r3, r2, 1 blr LBB_test_6: ; loopexit or r3, r2, r2 blr into: _test: cmpwi cr0, r4, 0 bgt cr0, LBB_test_2 ; entry.no_exit_crit_edge LBB_test_1: ; entry.loopexit_crit_edge li r2, 0 b LBB_test_5 ; loopexit LBB_test_2: ; entry.no_exit_crit_edge li r6, 0 LBB_test_3: ; no_exit lwz r2, 0(r3) cmpw cr0, r2, r5 or r2, r6, r6 beq cr0, LBB_test_5 ; loopexit LBB_test_4: ; endif addi r3, r3, 4 addi r6, r6, 1 cmpw cr0, r6, r4 or r2, r6, r6 blt cr0, LBB_test_3 ; no_exit LBB_test_5: ; loopexit or r3, r2, r2 blr Unfortunately, this is actually worse code, because the register coallescer is getting confused somehow. If it were doing its job right, it could turn the code into this: _test: cmpwi cr0, r4, 0 bgt cr0, LBB_test_2 ; entry.no_exit_crit_edge LBB_test_1: ; entry.loopexit_crit_edge li r6, 0 b LBB_test_5 ; loopexit LBB_test_2: ; entry.no_exit_crit_edge li r6, 0 LBB_test_3: ; no_exit lwz r2, 0(r3) cmpw cr0, r2, r5 beq cr0, LBB_test_5 ; loopexit LBB_test_4: ; endif addi r3, r3, 4 addi r6, r6, 1 cmpw cr0, r6, r4 blt cr0, LBB_test_3 ; no_exit LBB_test_5: ; loopexit or r3, r6, r6 blr ... which I'll work on next. :) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23604 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-03Refactor some code into a functionChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23603 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-03This break is bogus and I have no idea why it was there. Basically it preventsChris Lattner
memoizing code when IV's are used by phinodes outside of loops. In a simple example, we were getting this code before (note that r6 and r7 are isomorphic IV's): li r6, 0 or r7, r6, r6 LBB_test_3: ; no_exit lwz r2, 0(r3) cmpw cr0, r2, r5 or r2, r7, r7 beq cr0, LBB_test_5 ; loopexit LBB_test_4: ; endif addi r2, r7, 1 addi r7, r7, 1 addi r3, r3, 4 addi r6, r6, 1 cmpw cr0, r6, r4 blt cr0, LBB_test_3 ; no_exit Now we get: li r6, 0 LBB_test_3: ; no_exit or r2, r6, r6 lwz r6, 0(r3) cmpw cr0, r6, r5 beq cr0, LBB_test_6 ; loopexit LBB_test_4: ; endif addi r3, r3, 4 addi r6, r2, 1 cmpw cr0, r6, r4 blt cr0, LBB_test_3 ; no_exit this was noticed in em3d. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23602 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-03when checking if we should move a split edge block outside of a loop,Chris Lattner
check the presplit pred, not the post-split pred. This was causing us to make the wrong decision in some cases, leaving the critical edge block in the loop. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23601 91177308-0d34-0410-b5e6-96231b3b80d8
2005-09-27Make the pass name simplerChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23476 91177308-0d34-0410-b5e6-96231b3b80d8
2005-09-13Fix an issue where LSR would miss rewriting a use of an IV expression by a ↵Chris Lattner
PHI node that is not the original PHI. This fixes up a dot-product loop in galgel, speeding it up from 18.47s to 16.13s. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23327 91177308-0d34-0410-b5e6-96231b3b80d8
2005-09-12Fix a regression from last night, which caused this pass to create invalidChris Lattner
code for IV uses outside of loops that are not dominated by the latch block. We should only convert these uses to use the post-inc value if they ARE dominated by the latch block. Also use a new LoopInfo method to simplify some code. This fixes Transforms/LoopStrengthReduce/2005-09-12-UsesOutOutsideOfLoop.ll git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23318 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-18-g5258 (git 2.32.0) at 2025-06-22 18:20:44 +0000