llvm/lib
Chandler Carruth e6329cf303 [x86] Fix a crash and wrong-code bug in the new vector lowering all
found by a single test reduced out of a failure on llvm-stress.

The start of the problem (and the crash) came when we tried to use
a find of a non-used slot in the move-to half of the move-mask as the
target for two bad-half inputs. While if lucky this will be the first of
a pair of slots which we can place the bad-half inputs into, it isn't
actually guaranteed. This really isn't surprising, not sure what I was
thinking. The correct way to find the two unused slots is to look for
one of the *used* slots. We know it isn't that pair, and we can use some
modular arithmetic to find the other pair by masking off the odd bit and
adding 2 modulo 4. With this, we reliably found a viable pair of slots
for the bad-half inputs.

Sadly, that wasn't enough. We also had a wrong code bug that surfaced
when I reduced the test case for this where we would use the same slot
twice for the two bad inputs. This is because both of the bad inputs
could be in odd slots originally and thus the mod-2 mapping would
actually be the same. The whole point of the weird indexing into the
pair of empty slots was to try to leverage when the end result needed
the two bad-half inputs to be paired in a dword and pre-pair them in the
correct orrientation. This is less important with the powerful combining
we're now doing, and also easier and more reliable to achieve be noting
that we add the bad-half inputs in order. Thus, if they are in a dword
pair, the low part of that will be the first input in the sequence.
Always putting that in the low element will just do the right thing in
addition to computing the correct result.

Test case added. =]

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@214849 91177308-0d34-0410-b5e6-96231b3b80d8
2014-08-05 08:19:21 +00:00
..
Analysis Fix ScalarEvolutionExpander when creating a PHI in a block with duplicate predecessors 2014-07-31 19:13:38 +00:00
AsmParser Use "weak alias" instead of "alias weak" 2014-07-30 22:51:54 +00:00
Bitcode UseListOrder: Fix blockaddress use-list order 2014-08-01 22:27:19 +00:00
CodeGen Have MachineFunction cache a pointer to the subtarget to make lookups 2014-08-05 02:39:49 +00:00
DebugInfo DWOHolder takes ownership of the argument constructor, use std::unique_ptr. 2014-07-31 20:26:42 +00:00
ExecutionEngine Remove the TargetMachine forwards for TargetSubtargetInfo based 2014-08-04 21:25:23 +00:00
IR Provide convenient access to the zext/sext attributes of function arguments. NFC. 2014-08-05 05:43:41 +00:00
IRReader Update the MemoryBuffer API to use ErrorOr. 2014-07-06 17:43:13 +00:00
LineEditor
Linker Include <tuple> to make buildbots happy 2014-06-27 18:38:12 +00:00
LTO Remove the TargetMachine forwards for TargetSubtargetInfo based 2014-08-04 21:25:23 +00:00
MC Enable Darwin vararg parameters support in assembler macros. 2014-08-04 23:14:37 +00:00
Object A std::unique_ptr case I missed in the previous patch. 2014-07-31 03:36:00 +00:00
Option Generic: add range-adapter for option parsing. 2014-07-09 13:03:37 +00:00
ProfileData Coverage: add HasCodeBefore flag to a mapping region. 2014-08-04 18:00:51 +00:00
Support Path: Stop claiming path::const_iterator is bidirectional 2014-08-04 17:36:41 +00:00
TableGen Don't fail tablegen immediately after failing to set a value. 2014-07-31 01:44:00 +00:00
Target [x86] Fix a crash and wrong-code bug in the new vector lowering all 2014-08-05 08:19:21 +00:00
Transforms [SimplifyCFG] fix accessing deleted PHINodes in switch-to-table conversion. 2014-08-02 23:41:54 +00:00
CMakeLists.txt
LLVMBuild.txt
Makefile