[LLVMdev] "Cannot select" error in 2.9

[Gregory Junker]

Hi all

I didn't find anything relevant in the archives about this so I wanted to
ask here.

I am generating a simple function for JIT using the API. The function
structure and code work fine with 2.8 but give me "Cannot select" with 2.9.


Assembly dump of the function:

define double @0(double %f0, double %f1, double %f2, double %f3, double %f4,
double %f5, double %f6, double %f7, double %f8, double %f9) {
  %1 = add double %f0, %f1
  %2 = add double %1, %f2
  %3 = add double %2, %f3
  %4 = add double %3, %f4
  %5 = add double %4, %f5
  %6 = add double %5, %f6
  %7 = add double %6, %f7
  %8 = add double %7, %f8
  %9 = add double %8, %f9
  ret double %9
}


LLVM error I get:

LLVM ERROR: Cannot select: 0xd1b720: f64 = add 0xd1b620, 0xd1ae20 [ORD=9]
[ID=32]
  0xd1b620: f64 = add 0xd1b520, 0xd1ac20 [ORD=8] [ID=31]
    0xd1b520: f64 = add 0xd1b420, 0xd1a920 [ORD=7] [ID=30]
      0xd1b420: f64 = add 0xd1b320, 0xd1a610 [ORD=6] [ID=29]
        0xd1b320: f64 = add 0xd1b220, 0xd1a410 [ORD=5] [ID=28]
          0xd1b220: f64 = add 0xd1b120, 0xd1a210 [ORD=4] [ID=27]
            0xd1b120: f64 = add 0xd1b020, 0xd1a010 [ORD=3] [ID=26]
              0xd1b020: f64 = add 0xd1af20, 0xd19e10 [ORD=2] [ID=25]
                0xd1af20: f64 = add 0xd19a10, 0xd19c10 [ORD=1] [ID=24]
                  0xd19a10: f64,ch = CopyFromReg 0xcee028, 0xd19910 [ORD=1]
[ID=14]

                  0xd19c10: f64,ch = CopyFromReg 0xcee028, 0xd19b10 [ORD=1]
[ID=15]

                0xd19e10: f64,ch = CopyFromReg 0xcee028, 0xd19d10 [ORD=2]
[ID=16]
                  0xd19d10: f64 = Register %vreg2 [ORD=2] [ID=3]
              0xd1a010: f64,ch = CopyFromReg 0xcee028, 0xd19f10 [ORD=3]
[ID=17]
                0xd19f10: f64 = Register %vreg3 [ORD=3] [ID=4]
            0xd1a210: f64,ch = CopyFromReg 0xcee028, 0xd1a110 [ORD=4]
[ID=18]
              0xd1a110: f64 = Register %vreg4 [ORD=4] [ID=5]
          0xd1a410: f64,ch = CopyFromReg 0xcee028, 0xd1a310 [ORD=5] [ID=19]
            0xd1a310: f64 = Register %vreg5 [ORD=5] [ID=6]
        0xd1a610: f64,ch = CopyFromReg 0xcee028, 0xd1a510 [ORD=6] [ID=20]
          0xd1a510: f64 = Register %vreg6 [ORD=6] [ID=7]
      0xd1a920: f64,ch = CopyFromReg 0xcee028, 0xd1a710 [ORD=7] [ID=21]
        0xd1a710: f64 = Register %vreg7 [ORD=7] [ID=8]
    0xd1ac20: f64,ch = load 0xcee028, 0xd1aa20,
0xd1ab20<LD8[FixedStack-1](align=16)> [ORD=8] [ID=22]
      0xd1aa20: i64 = FrameIndex<-1> [ORD=8] [ID=9]
      0xd1ab20: i64 = undef [ORD=8] [ID=10]
  0xd1ae20: f64,ch = load 0xcee028, 0xd1ad20, 0xd1ab20<LD8[FixedStack-2]>
[ORD=9] [ID=23]
    0xd1ad20: i64 = FrameIndex<-2> [ORD=9] [ID=11]
    0xd1ab20: i64 = undef [ORD=8] [ID=10]



Is there something different I need to do in my function definition or
structure to get by this? Or is this a bug?

Thanks
Greg