[LLVMbugs] [Bug 5327] New: Systemz backend - verify-machineinstrs error

[bugzilla-daemon at cs.uiuc.edu]

http://llvm.org/bugs/show_bug.cgi?id=5327

           Summary: Systemz backend - verify-machineinstrs error
           Product: new-bugs
           Version: trunk
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: quickslyver at free.fr
                CC: llvmbugs at cs.uiuc.edu


Created an attachment (id=3716)
 --> (http://llvm.org/bugs/attachment.cgi?id=3716)
C source

$clang-cc __udivhi3.c -O3 -o - -S -triple=s390x-unknown-unknown -debug
-verify-machineinstrs

(...)

while.body11: 0xa9bbed4, LLVM BB @0xa98ea30, ID#12:
    Predecessors according to CFG: 0xa9bb780 (#7) 0xa9bbe88 (#11)
        %reg1053<def> = PHI %reg1031, mbb<while.body11,0xa9bbe88>, %reg1052,
mbb<while.body11,0xa9bb780>
        JL mbb<while.body11,0xa9bbfc4>, %PSW<imp-use>
    Successors according to CFG: 0xa9bbf78 (#13) 0xa9bbfc4 (#14)

(...)

*** Bad machine code: Using an undefined physical register ***
- function:    __udivhi3
- basic block: while.body11 0xa9bbed4 (#12)
- instruction: JL mbb<while.body11,0xa9bbfc4>, %PSW<imp-use>
- operand 1:   %PSW<imp-use>
fatal error: error in backend: Found 1 machine code errors.

and the final code is:
(...)
        jl      .LBB1_8
        lr      %r3, %r5
.LBB1_8:
        jl      .LBB1_10
        lr      %r2, %r11
.LBB1_10:
(...)

there is a big problem somewhere... but looking at debug output it's hard for
me to see how basic bloc #12 is created.


-- 
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.