This is the last r300g piglit regression from 7.11 to 8.0. I'm not sure if this is a valid bug, there is lots of errors:

r300: ERROR: FS input generic 20 unassigned, not enough hardware slots (it's not a bug, do not report it).
r300: ERROR: FS input generic 21 unassigned, not enough hardware slots (it's not a bug, do not report it).

However since this is a regression I'm reporting it nevertheless. May also be related to bug 34201 


Output:
Vertical axis: Increasing numbers of varyings.
Horizontal axis: Which of the varyings contains the color.
GL_MAX_VARYING_FLOATS = 40
Probe at (8,98)
  Expected: 0.000000 1.000000 0.000000
  Observed: 0.000000 0.000000 0.000000
  Failure with 9 vec4 varyings used in varying index 0
Probe at (8,110)
  Expected: 0.000000 1.000000 0.000000
  Observed: 0.000000 0.000000 0.000000
  Failure with 10 vec4 varyings used in varying index 0


1ded658ce074a85bc08c989ff17840b840ff3051 is the first bad commit
commit 1ded658ce074a85bc08c989ff17840b840ff3051
Author: Marek Olšák <maraeo@gmail.com>
Date:   Tue Nov 22 15:05:29 2011 +0100

    st/mesa: add color varyings to MaxVarying
    
    The linker now adds color varyings to the number of used varyings and checks
    against that limit.
    
    NOTE: This is a candidate for the 7.11 branch.


GPU: RV530
Mesa: df1cd55ebf362948788c04d2fa7da55c80991605
Kernel: 3.2.3
Libdrm: 2.4.31