7989 Commits

Author SHA1 Message Date
Daniel Dunbar
e19b42b927 tests: Add a %abs_tmp substitution which is guaranteed to be a full path.
llvm-svn: 81662
2009-09-13 01:39:50 +00:00
Daniel Dunbar
4f5ca6bbf4 Sink llvm-gcc dependent tests into distinct subdirs.
llvm-svn: 81661
2009-09-13 01:39:08 +00:00
Daniel Dunbar
03e575c936 Rewrite tests to not use Tcl substitution.
llvm-svn: 81660
2009-09-13 01:37:07 +00:00
Daniel Dunbar
19d95a536c Simplify LLVMC tests.
llvm-svn: 81659
2009-09-13 01:36:19 +00:00
Anton Korobeynikov
fa4c7562d5 Define proper subreg sets for arm - this should fix bunch of subtle problems
with subreg - superreg mapping and also fix PR4965.

llvm-svn: 81657
2009-09-13 00:59:43 +00:00
Dan Gohman
2ef4d82155 Add -mattr=+sse2 to the -march=x86 version of this test. Without
sse, this code falls back to SelectionDAG isel which uses an x87
instruction, which is fine, but not what this test is testing for.

llvm-svn: 81656
2009-09-12 23:45:47 +00:00
Dan Gohman
8e55f0f55b Remove an unnecessary -f.
llvm-svn: 81546
2009-09-11 18:41:06 +00:00
Dan Gohman
f2c290dfa6 Convert more tests to avoid llvm-as.
llvm-svn: 81545
2009-09-11 18:36:27 +00:00
Dan Gohman
c9307490ce Eliminate more redundant llvm-as calls.
llvm-svn: 81540
2009-09-11 18:17:12 +00:00
Dan Gohman
ded1a79805 Fix this test to test what it was originally intended to test.
llvm-svn: 81539
2009-09-11 18:16:43 +00:00
Dan Gohman
205b641954 Change tests from "opt %s" to "opt < %s" so that opt doesn't see the
input filename so that opt doesn't print the input filename in the
output so that grep lines in the tests don't unintentionally match
strings in the input filename.

llvm-svn: 81537
2009-09-11 18:01:28 +00:00
Chris Lattner
ebfcd8d588 turn on -experimental-asm-printer for x86 / AT&T by default.
llvm-svn: 81532
2009-09-11 17:07:27 +00:00
Chris Lattner
abca9e3a0d another random update
llvm-svn: 81531
2009-09-11 17:07:01 +00:00
Chris Lattner
e8f2153059 fix a bunch of spurious failures for people whose home directory
is sabre.

llvm-svn: 81528
2009-09-11 17:02:12 +00:00
Chris Lattner
6e3ee2e578 this test is using invalid "intrinsics".
llvm-svn: 81527
2009-09-11 16:49:18 +00:00
Chris Lattner
817a8fc5b0 fix test to not get a moduleid that matches 'br'
llvm-svn: 81526
2009-09-11 16:47:41 +00:00
Evan Cheng
c3ef5b0802 Follow up to 81494. When the folded reload is narrowed to a 32-bit load then change the destination register to a 32-bit one or add a sub-register index.
llvm-svn: 81496
2009-09-11 01:01:31 +00:00
Evan Cheng
93831c07de It's not legal to fold a load from a narrower stack slot into a wider instruction. If done, the instruction does a 64-bit load and that's not
safe. This can happen we a subreg_to_reg 0 has been coalesced. One
exception is when the instruction that folds the load is a move, then we
can simply turn it into a 32-bit load from the stack slot.                                                                                                                    

rdar://7170444

llvm-svn: 81494
2009-09-11 00:39:26 +00:00
Dan Gohman
964c0b8333 Reapply r81171 with a fix: don't try to use i64 when it
isn't legal.

llvm-svn: 81492
2009-09-11 00:34:46 +00:00
Dan Gohman
aa66e3d968 Teach lib/VMCore/ConstantFold.cpp how to set the inbounds keyword and
how to fold notionally-out-of-bounds array getelementptr indices instead
of just doing these in lib/Analysis/ConstantFolding.cpp, because it can
be done in a fairly general way without TargetData, and because not all
constants are visited by lib/Analysis/ConstantFolding.cpp. This enables
more constant folding.

Also, set the "inbounds" flag when the getelementptr indices are
one-past-the-end.

llvm-svn: 81483
2009-09-11 00:04:14 +00:00
Dan Gohman
58a0550024 Factor out the code for checking that all indices in a getelementptr are
within the notional bounds of the static type of the getelementptr (which
is not the same as "inbounds") from GlobalOpt into a utility routine,
and use it in ConstantFold.cpp to check whether there are any mis-behaved
indices.

llvm-svn: 81478
2009-09-10 23:37:55 +00:00
Bob Wilson
7b39f31422 Don't swap the operands of a subtraction when trying to create a
post-decrement load/store.

llvm-svn: 81464
2009-09-10 22:09:31 +00:00
Evan Cheng
d3f98b1d15 Add a test case for r81431.
llvm-svn: 81432
2009-09-10 05:08:51 +00:00
Daniel Dunbar
b15a757410 Remove prcontext.
llvm-svn: 81427
2009-09-10 04:56:59 +00:00
Bob Wilson
76cffeddd1 Revert r81171 which was causing pr4927.
llvm-svn: 81415
2009-09-10 00:49:22 +00:00
Dale Johannesen
181999c0a6 Testcase for (llvm-gcc) 81399. Adjust an older
test case to allow for different, functionally
identical output.

llvm-svn: 81400
2009-09-09 23:37:46 +00:00
Bob Wilson
877a857b4b Fix pr4939: Change FPCCToARMCC to translate SETOLE to ARMCC::LS.
See the bug report for details.

llvm-svn: 81397
2009-09-09 23:14:54 +00:00
Dan Gohman
0357e0117e Fix SplitCriticalEdge to properly update LCSSA form when splitting a
loop exit edge -- new PHIs may be needed not only for the additional
splits that are made to preserve LoopSimplify form, but also for the
original split. Factor out the code that inserts new PHIs so that it
can be used for both. Remove LoopRotation.cpp's code for manually
updating LCSSA form, as it is now redundant. This fixes PR4934.

llvm-svn: 81363
2009-09-09 18:18:18 +00:00
Dan Gohman
beccd0690e When widening a vector load, use the correct chain. This fixes PR4891.
llvm-svn: 81343
2009-09-09 14:22:57 +00:00
Torok Edwin
556197eec9 Add testcase for r81322 (PR4933).
llvm-svn: 81327
2009-09-09 09:34:43 +00:00
Chris Lattner
83d3519ec4 add a testacse for the objc problem that required required r81305
to be temporarily disabled.

llvm-svn: 81320
2009-09-09 06:19:34 +00:00
Chris Lattner
2dfd5f5963 disable the new asmprinter by default. Both the Mangler and MCSymbol
printing stuff are quoting symbols now, breaking objc testcases.

llvm-svn: 81319
2009-09-09 06:11:14 +00:00
Evan Cheng
91e719a6a1 Cast MO.getImm() to unsigned before comparing with an unsigned limit.
llvm-svn: 81318
2009-09-09 06:05:16 +00:00
Daniel Dunbar
61043be3ae Update test.
llvm-svn: 81314
2009-09-09 02:41:50 +00:00
Chris Lattner
c74e9c2c37 turn the mcinst asmprinter on by default for x86, tweaking two tests to
expect the slight syntax differences in the generated code.

llvm-svn: 81305
2009-09-09 00:41:36 +00:00
Bill Wendling
2809c62757 Remove failing test...
llvm-svn: 81303
2009-09-09 00:30:25 +00:00
Dan Gohman
0df0f8323c Use "opt < %s" instead of "opt %s" to keep the testname away from the grep.
llvm-svn: 81299
2009-09-09 00:22:49 +00:00
Chris Lattner
e2e20d69a3 this got merged into lea.ll
llvm-svn: 81298
2009-09-09 00:22:31 +00:00
Chris Lattner
d5bec5079a filecheckize
llvm-svn: 81297
2009-09-09 00:19:46 +00:00
Bill Wendling
95b3d14496 Fix PR4865. This syncs up the JIT's DWARF emitter with what's in the
'DwarfException.cpp' file, which changed how CIEs were emitted, the sizes of
some fields, etc.

llvm-svn: 81295
2009-09-09 00:11:02 +00:00
Dan Gohman
142428ce64 Eliminate more uses of llvm-as and llvm-dis.
llvm-svn: 81293
2009-09-09 00:09:15 +00:00
Dan Gohman
df2896d609 Eliminate more uses of llvm-as and llvm-dis.
llvm-svn: 81290
2009-09-08 23:54:48 +00:00
Chris Lattner
ded5762e28 update various tests for signedness changes in .s file.
llvm-svn: 81289
2009-09-08 23:51:06 +00:00
Chris Lattner
837420e12f adjust for signedness change. I'd appreciate it if an ARM flavored person
could look at this: the top undefined bits of an immediate shouldn't affect
isel (cmp vs cmp.w)

llvm-svn: 81288
2009-09-08 23:44:53 +00:00
Daniel Dunbar
648d754b7b Merge Archive/extract* tests into one; this avoids a race when tests are run in
parallel (the test should really use temps for the output, though).

llvm-svn: 81287
2009-09-08 23:44:24 +00:00
Chris Lattner
09a6252bb8 merge thumb2-bic2.ll into thumb2-bic.ll and update for signedness changes.
llvm-svn: 81285
2009-09-08 23:41:06 +00:00
Chris Lattner
a97bedf017 tweak this to pass on linux.
llvm-svn: 81273
2009-09-08 23:32:40 +00:00
Chris Lattner
840fbf6897 convert to filecheck syntax
llvm-svn: 81267
2009-09-08 23:16:26 +00:00
Chris Lattner
0b34068b2b change selectiondag to add the sign extended versions of immediate operands
to instructions instead of zero extended ones.  This makes the asmprinter
print signed values more consistently.  This apparently only really affects
the X86 backend.

llvm-svn: 81265
2009-09-08 23:05:44 +00:00
Dan Gohman
cd0fb89725 Use "opt < %s" instead of "opt %s" so that opt doesn't print the test
filename in the output, which interferes with the tests' grep lines.

llvm-svn: 81263
2009-09-08 22:57:49 +00:00