Multi-cpu configs don't properly update ASTE for prds on non-boot CPUs

[Gary Dixon]

On GHM, I was researching length of prds (per-processor) segments. The config deck shows 2 cpus running.

pcd cpu

cpu  a  7  on  dps8  70.  8.
cpu  b  6  on  dps8  70.  8.
cpu  c  5  off  dps8  70.  8.
cpu  d  4  off  dps8  70.  8.
cpu  e  3  off  dps8  70.  8.
cpu  f  2  off  dps8  70.  8.

r 09:53 0.083 3

But lengths reported for these prds files show a positive length only for the bootload CPU A.

ls >sl1>*.prds

Segments = 6, Lengths = 9.

       9  cpu_A.prds
       0  cpu_B.prds
       0  cpu_C.prds
       0  cpu_D.prds
       0  cpu_E.prds
       0  cpu_F.prds

r 09:51 0.152 4

Does this mean that when 2nd thru nth CPU is added, the dynamic CPU adding code does not properly update the ASTE for these prds file to reflect length of these segments (whose pages must be wired in memory)?

I have not investigated this bug, but only entered it to record this anomaly.

I was investigate bug #170 when I discovered this anomaly.