[llvm-bugs] [Bug 25602] New: Inliner generates 'bitcast' instead of 'addrspacecast'

via llvm-bugs llvm-bugs at lists.llvm.org
Sun Nov 22 11:32:51 PST 2015


https://llvm.org/bugs/show_bug.cgi?id=25602

            Bug ID: 25602
           Summary: Inliner generates 'bitcast' instead of 'addrspacecast'
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P
         Component: Transformation Utilities
          Assignee: unassignedbugs at nondot.org
          Reporter: zvi.rackover at intel.com
                CC: llvm-bugs at lists.llvm.org
    Classification: Unclassified

Created attachment 15330
  --> https://llvm.org/bugs/attachment.cgi?id=15330&action=edit
Reproducer

To reproduce:
opt -inline -S < repro.ll

Output with debug flag on:
Args: opt -inline -S -debug
Inliner visiting SCC: bar: 0 call sites.
Inliner visiting SCC: foo: 1 call sites.
    Inlining: cost=always, Call:   %call = call <32 x i16> @bar(%S
addrspace(4)* byval %d)
    -> Deleting dead function: bar
CGSCCPASSMGR: Refreshing SCC with 1 nodes:
Call graph node for function: 'foo'<<0x60141d0>>  #uses=1

CGSCCPASSMGR: SCC Refresh didn't change call graph.
Invalid bitcast
  %lo.i = bitcast %S* %d1 to <32 x i16> addrspace(4)*
LLVM ERROR: Broken function found, compilation aborted!
------------------------------------------------------------

Seems like the Inliner is creating a 'bitcast' instead of an 'addrspacecast'.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20151122/109db450/attachment.html>


More information about the llvm-bugs mailing list