maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
57,989 Commits 23 Branches 0 Tags 2.2 GiB
Commit Graph

5159 Commits

Author SHA1 Message Date
Dan Gohman 8b6ebb1112 Minor code simplifications. Don't attempt LSR on theoretical 
targets with pointers larger than 64 bits, due to the code not
yet being APInt clean.

llvm-svn: 69296
 2009-04-16 16:49:48 +00:00
Dan Gohman e2ead2c328 LSR is no longer a GEP optimizer. It is now an IV expression 
optimizer, which just happen to frequently involve optimizing GEPs.

llvm-svn: 69295
 2009-04-16 16:46:01 +00:00
Dan Gohman a8be04b2db Use ConstantExpr::getIntToPtr instead of SCEVExpander::InsertCastOfTo, 
since the operand is always a constant.

llvm-svn: 69291
 2009-04-16 15:48:38 +00:00
Dan Gohman 71bccd3e0e Use a SCEV expression cast instead of immediately inserting a 
new instruction with SCEVExpander::InsertCastOfTo.

llvm-svn: 69290
 2009-04-16 15:47:35 +00:00
Dan Gohman 0a40ad93a9 Expand GEPs in ScalarEvolution expressions. SCEV expressions can now 
have pointer types, though in contrast to C pointer types, SCEV
addition is never implicitly scaled. This not only eliminates the
need for special code like IndVars' EliminatePointerRecurrence
and LSR's own GEP expansion code, it also does a better job because
it lets the normal optimizations handle pointer expressions just
like integer expressions.

Also, since LLVM IR GEPs can't directly index into multi-dimensional
VLAs, moving the GEP analysis out of client code and into the SCEV
framework makes it easier for clients to handle multi-dimensional
VLAs the same way as other arrays.

Some existing regression tests show improved optimization.
test/CodeGen/ARM/2007-03-13-InstrSched.ll in particular improved to
the point where if-conversion started kicking in; I turned it off
for this test to preserve the intent of the test.

llvm-svn: 69258
 2009-04-16 03:18:22 +00:00
Dale Johannesen a71daa83c6 Eliminate zext over (iv | const) or (signed iv), 
and sext over (iv | const), if a longer iv is
available.  Allow expressions to have more than
one zext/sext parent.  All from OpenSSL.

llvm-svn: 69241
 2009-04-15 23:31:51 +00:00
Dale Johannesen 82230b5b17 Eliminate zext over (iv & const) or ((iv+const)&const) 
if a longer iv is available.  These subscript forms are
not common; they're a bottleneck in OpenSSL.

llvm-svn: 69215
 2009-04-15 20:41:02 +00:00
Dale Johannesen 7ffb7d5728 Enhance induction variable code to remove the 
sext around sext(shorter IV + constant), using a
longer IV instead, when it can figure out the
add can't overflow.  This comes up a lot in
subscripting; mainly affects 64 bit.

llvm-svn: 69123
 2009-04-15 01:10:12 +00:00
Evan Cheng ffb83a155e Avoid making the transformation enabled by my last patch if the new destinations have phi nodes. 
llvm-svn: 69121
 2009-04-15 00:43:54 +00:00
Devang Patel 046bf624b9 While inlining, clone llvm.dbg.func.start intrinsic and adjust 
llvm.dbg.region.end instrinsic. This nested llvm.dbg.func.start/llvm.dbg.region.end pair now enables DW_TAG_inlined_subroutine support in code generator.

llvm-svn: 69118
 2009-04-15 00:17:06 +00:00
Evan Cheng 5ebf2acd84 Optimize conditional branch on i1 phis with non-constant inputs. 
This turns:

eq:
        %3 = icmp eq i32 %1, %2
        br label %join

ne:
        %4 = icmp ne i32 %1, %2
        br label %join

join:
        %5 = phi i1 [%3, %eq], [%4, %ne]
        br i1 %5, label %yes, label %no

=>

eq:
        %3 = icmp eq i32 %1, %2
        br i1 %3, label %yes, label %no

ne:
        %4 = icmp ne i32 %1, %2
        br i1 %4, label %yes, label %no

llvm-svn: 69102
 2009-04-14 23:40:03 +00:00
Owen Anderson a1902318e3 LoopIndexSplit needs to inform the loop pass manager of the instructions it is 
deleting, not just the basic block.

llvm-svn: 69011
 2009-04-14 01:04:19 +00:00
Chris Lattner 836e77d161 eliminate unneeded parens. 
llvm-svn: 68939
 2009-04-13 05:38:23 +00:00
Chris Lattner 6cd82fb430 "There was a typo in my previous patch which leads to miscompilation of 
strncat :(

strncat(foo, "bar", 99)
would be optimized to
memcpy(foo+strlen(foo), "bar", 100, 1)
instead of
memcpy(foo+strlen(foo), "bar", 4, 1)"

Patch by Benjamin Kramer!

llvm-svn: 68905
 2009-04-12 18:22:33 +00:00
Chris Lattner 91b6af24ac add some optimizations for strncpy/strncat and factor some 
code.  Patch by Benjamin Kramer!

llvm-svn: 68885
 2009-04-12 05:06:39 +00:00
Chris Lattner eb510d6b3d Instcombine should not promote whole computation trees to "strange" 
integer types, unless they are already strange.  This prevents it from
turning the code produced by SROA into crazy libcalls and stuff that 
the code generator can't handle.  In the attached example, the result
was an i96 multiply that caused the x86 backend to assert.

Note that if TargetData had an idea of what the legal types are for
a target that this could be used to stop instcombine from introducing
i64 muls, as Scott wanted.

llvm-svn: 68598
 2009-04-08 05:41:03 +00:00
Chris Lattner 321741af5f fix rdar://6762290, a crash compiling cxx filt with clang. 
llvm-svn: 68500
 2009-04-07 05:03:34 +00:00
Chris Lattner 47d6e7b93e remove empty section 
llvm-svn: 68485
 2009-04-07 02:55:53 +00:00
Ed Schouten 01aa6ec97a Let the strcat optimizer return the pointer to the start of the buffer, 
instead of the place where it started to perform the string copy.

- PR3661
- Patch by Benjamin Kramer!

llvm-svn: 68443
 2009-04-06 13:06:48 +00:00
Owen Anderson 98f912bf13 Reapply r68211, with the miscompilations it caused fixed. 
llvm-svn: 68262
 2009-04-01 23:53:49 +00:00
Dan Gohman c4971721ea Revert r68172. It caused regressions in 
Applications/Burg/burg
  Applications/ClamAV/clamscan
and many other tests.

llvm-svn: 68211
 2009-04-01 16:37:47 +00:00
Owen Anderson ff5961b46c Enhance GVN to propagate simple conditionals. This fixes PR3921. 
llvm-svn: 68172
 2009-04-01 01:20:45 +00:00
Chris Lattner f72ce6ea8b Make the key of ValueRankMap an AssertingVH, so that we die violently 
if it dangles.

llvm-svn: 68150
 2009-03-31 22:13:29 +00:00
Evan Cheng 826b6f0f7c Throttle back "fold select into operand" transformation. InstCombine should not generate selects of two constants unless they are selects of 0 and 1. 
e.g.
define i32 @t1(i32 %c, i32 %x) nounwind {
       %t1 = icmp eq i32 %c, 0
       %t2 = lshr i32 %x, 18
       %t3 = select i1 %t1, i32 %t2, i32 %x
       ret i32 %t3
}

was turned into

define i32 @t2(i32 %c, i32 %x) nounwind {
       %t1 = icmp eq i32 %c, 0
       %t2 = select i1 %t1, i32 18, i32 0
       %t3 = lshr i32 %x, %t2
       ret i32 %t3
}

For most targets, that means materializing two constants and then a select. e.g. On x86-64

movl    %esi, %eax
shrl    $18, %eax
testl   %edi, %edi
cmovne  %esi, %eax
ret

=>

xorl    %eax, %eax
testl   %edi, %edi
movl    $18, %ecx
cmovne  %eax, %ecx
movl    %esi, %eax
shrl    %cl, %eax
ret

Also, the optimizer and codegen can reason about shl / and / add, etc. by a constant. This optimization will hinder optimizations using ComputeMaskedBits.

llvm-svn: 68142
 2009-03-31 20:42:45 +00:00
Devang Patel 4ce6e69022 Update call graph after inlining invoke. 
Patch by Jay Foad.

llvm-svn: 68120
 2009-03-31 17:36:12 +00:00
Devang Patel 6e68bd007a Loop Index Split can eliminate a loop if it can determin if loop body is executed only once. There was a bug in determining IV based value of the iteration for which the loop body is executed. Fix it. 
llvm-svn: 68071
 2009-03-30 22:24:10 +00:00
Duncan Sands 3241b74f69 Revert r67798: it breaks llvm-gcc bootstrap on x86-64-linux, presumably due to 
a miscompilation.

make[4]: Entering directory `gcc-4.2.llvm-objects/x86_64-unknown-linux-gnu/libstdc++-v3/include'
if [ ! -d "./x86_64-unknown-linux-gnu/bits/stdtr1c++.h.gch" ]; then \
          mkdir -p ./x86_64-unknown-linux-gnu/bits/stdtr1c++.h.gch; \
        fi; \
        gcc-4.2.llvm-objects/./gcc/xgcc -shared-libgcc -Bgcc-4.2.llvm-objects/./gcc -nostdinc++ 
-Lgcc-4.2.llvm-objects/x86_64-unknown-linux-gnu/libstdc++-v3/src -Lgcc-4.2.llvm-objects/x86_64-unknown-linux-gnu/libstdc++-v3/src/.libs 
-B/usr/local/gnat-llvm/x86_64-unknown-linux-gnu/bin/ -B/usr/local/gnat-llvm/x86_64-unknown-linux-gnu/lib/ -isystem 
/usr/local/gnat-llvm/x86_64-unknown-linux-gnu/include -isystem /usr/local/gnat-llvm/x86_64-unknown-linux-gnu/sys-include -Winvalid-pch -Wno-deprecated -x 
c++-header -g -O2  -D_GNU_SOURCE -Igcc-4.2.llvm-objects/x86_64-unknown-linux-gnu/libstdc++-v3/include/x86_64-unknown-linux-gnu 
-Igcc-4.2.llvm-objects/x86_64-unknown-linux-gnu/libstdc++-v3/include -Igcc-4.2.llvm/libstdc++-v3/libsupc++ -O2 -g 
gcc-4.2.llvm/libstdc++-v3/include/precompiled/stdtr1c++.h -o x86_64-unknown-linux-gnu/bits/stdtr1c++.h.gch/O2g.gch
In file included from gcc-4.2.llvm-objects/x86_64-unknown-linux-gnu/libstdc++-v3/include/tr1/repeat.h:247,
                 from gcc-4.2.llvm-objects/x86_64-unknown-linux-gnu/libstdc++-v3/include/tr1/functional:1098,
                 from gcc-4.2.llvm/libstdc++-v3/include/precompiled/stdtr1c++.h:53:
gcc-4.2.llvm-objects/x86_64-unknown-linux-gnu/libstdc++-v3/include/tr1/functional_iterate.h:417: internal compiler error: in ggc_recalculate_in_use_p, at 
ggc-page.c:1602
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://llvm.org/bugs/> for instructions.
make[4]: *** [x86_64-unknown-linux-gnu/bits/stdtr1c++.h.gch/O2g.gch] Error 1

llvm-svn: 67839
 2009-03-27 14:56:47 +00:00
Dale Johannesen 4026b041ce One more place to skip debug info. 
llvm-svn: 67811
 2009-03-27 01:13:37 +00:00
Devang Patel fe7c0492a0 While hoisting an instruction, update alias info set tracker. 
llvm-svn: 67798
 2009-03-26 23:48:52 +00:00
Dale Johannesen db90560c1c Skip debug info one more place. (This one gets 
called from llc, not opt, but it's an IR level
optimization nevertheless.)

llvm-svn: 67724
 2009-03-26 01:15:07 +00:00
Devang Patel 4555618854 Before deleting a basic block, give other loop passes a chance cleanup analysis values, related to the instructions in the basic block. 
llvm-svn: 67719
 2009-03-25 23:57:48 +00:00
Chris Lattner c3b2111d97 Fix PR3874 by restoring a condition I removed, but making it more 
precise than it used to be.

llvm-svn: 67662
 2009-03-25 00:28:58 +00:00
Chris Lattner 9e94538005 oops, I intended to remove this, not comment it out. Thanks Duncan! 
llvm-svn: 67657
 2009-03-24 23:48:25 +00:00
Chris Lattner 306813cbbb canonicalize inttoptr and ptrtoint instructions which cast pointers 
to/from integer types that are not intptr_t to convert to intptr_t
then do an integer conversion to the dest type.  This exposes the
cast to the optimizer.

llvm-svn: 67638
 2009-03-24 18:35:40 +00:00
Chris Lattner d9eb41177a two changes: 
1. Make instcombine always canonicalize trunc x to i1 into an icmp(x&1).  This 
   exposes the AND to other instcombine xforms and is more of what the code
   generator expects.
2. Rewrite the remaining trunc pattern match to use 'match', which 
   simplifies it a lot.
   

llvm-svn: 67635
 2009-03-24 18:15:30 +00:00
Dale Johannesen 32dfb35281 Use a SmallPtrSet instead of std::set. 
llvm-svn: 67578
 2009-03-23 23:39:20 +00:00
Dan Gohman 4f2fea1a21 Now that errs() is properly non-buffered, there's no need to 
explicitly flush it.

llvm-svn: 67526
 2009-03-23 15:57:19 +00:00
Duncan Sands 1f15ca7c7a Factorize out a concept - no functionality change. 
llvm-svn: 67454
 2009-03-21 21:27:31 +00:00
Chris Lattner 0a981d1d36 Fix instcombine to not introduce undefined shifts when merging two 
shifts together.  This fixes PR3851.

llvm-svn: 67411
 2009-03-20 22:41:15 +00:00
Duncan Sands a09e0afe74 Don't load values out of global constants with weak 
linkage: the value may be replaced with something
different at link time.  (Frontends that want to
allow values to be loaded out of weak constants can
give their constants weak_odr linkage).

llvm-svn: 67407
 2009-03-20 21:53:29 +00:00
Dale Johannesen 2050968df9 Clear the cached cost when removing a function in 
the inliner; prevents nondeterministic behavior
when the same address is reallocated.
Don't build call graph nodes for debug intrinsic calls;
they're useless, and there were typically a lot of them.

llvm-svn: 67311
 2009-03-19 18:03:56 +00:00
Dale Johannesen e4f361212b Fix comment typo. 
llvm-svn: 67307
 2009-03-19 17:23:29 +00:00
Dale Johannesen 52bc2aac8a This pass keeps a map of Instructions to Rank numbers, 
and was deleting Instructions without clearing the
corresponding map entry.  This led to nondeterministic
behavior if the same address got allocated to another
Instruction within a short time.

llvm-svn: 67306
 2009-03-19 17:22:53 +00:00
Nick Lewycky bfd4ad67c7 Remove strange extra semicolons. 
llvm-svn: 67287
 2009-03-19 05:51:39 +00:00
Chris Lattner 514fc5b143 aha, DAE does have to think about PHI nodes. Many thanks to "Dr Evil" (aka Duncan) 
for pointing this out :)

llvm-svn: 67212
 2009-03-18 16:48:45 +00:00
Chris Lattner 595923ff75 Fix PR3826 - InstComb assert with vector shift, by not calling ComputeNumSignBits on a vector. 
llvm-svn: 67211
 2009-03-18 16:32:19 +00:00
Chris Lattner ab8022055a add an assertion to make it clear that PHI nodes are not allowed. 
llvm-svn: 67210
 2009-03-18 16:23:56 +00:00
Zhou Sheng 4e2af3cb55 Explicitly check for StoreInst, do not lose the chance to delete 
unused loads or bitcasts.

llvm-svn: 67202
 2009-03-18 12:48:48 +00:00
Zhou Sheng 05bea906c1 Revert my previous change on Local.cpp, instead, fix the bug on scalarrepl. 
If the instruction has no users, it is also not only used by debug info 
and should not be deleted.

llvm-svn: 67194
 2009-03-18 10:13:08 +00:00
Zhou Sheng 64a6a092b1 Fix a bug. 
If I->use_empty(), this method should return false.

llvm-svn: 67180
 2009-03-18 07:56:13 +00:00