http://llvm.org/bugs/show_bug.cgi?id=10903
Summary: [VECTOR-SELECT] needed dagcombine optimization for
loading illegal types from memory
Product: new-bugs
Version: 2.8
Platform: PC
OS/Version: Windows NT
Status: NEW
Severity: normal
Priority: P
Component: new bugs
AssignedTo: [email protected]
ReportedBy: [email protected]
CC: [email protected]
When enabling the promote-element type legalization we often load illegal types
such as <4 x i8> and legalize them into legal types such as <4 x i32>.
Currently we perform scalarization of the load operation and insert the
promoted elements one by one. This is inefficient and can be done by a single
load operation followed by a shuffle instruction. I had implemented something
similar for the trunc-store operation.
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
_______________________________________________
LLVMbugs mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs
