OSDN Git Service

ART: x86_64 RoundDouble/Float intrinsics should initialize out value.
authorPavel Vyssotski <pavel.n.vyssotski@intel.com>
Fri, 31 Jul 2015 07:03:17 +0000 (13:03 +0600)
committerBrian Carlstrom <bdc@google.com>
Thu, 6 Aug 2015 02:16:47 +0000 (19:16 -0700)
commitce4b1329ca903d6b98734a27a46b54bb9cfd6d5b
tree460b5aa557b9a7b5d01ec59e4ec9c4c4e51d8ca3
parent2ada67b3145d38b352297bdb9d0546f58882e208
ART: x86_64 RoundDouble/Float intrinsics should initialize out value.

x86_64 RoundDouble intrinsic should initialize output register for the case of
"inPlusPointFive >= maxLong" as expected. The same for the RoundFloat intrinsic.
Fixed also the out register type in CreateSSE41FPToIntLocations provoking
a DCHECK failure.

Signed-off-by: Pavel Vyssotski <pavel.n.vyssotski@intel.com>
(cherry picked from commit 9ca257196b46fd7629bce0b338580e571e4113a8)

Bug: 22973442
Change-Id: If974e79d33311587d0b541a01ca8a4c9c11b9468
compiler/optimizing/intrinsics_x86_64.cc