Some (bad) news about SS20 [was Re: SS20 SMP success!]

[bertrand <joel.bertrand%systella.fr@localhost>]

Hello,

After my last message about HS SMP bug, I have tested some different 
configurations. I can reproduce a bug when my SS20 is idle : cpu1 cannot 
ping cpu0.



NetBSD 5.99.56 (GENERIC.MP) #0: Wed Nov 23 18:31:32 UTC 2011
        
builds%b6.netbsd.org@localhost:/home/builds/ab/HEAD/sparc/201111230730Z-obj/...
...
mainbus0 (root): SUNW,SPARCstation-20: hostid 72786e83
cpu0 at mainbus0: mid 8: RT620/625 @ 200 MHz, on-chip FPU
cpu0: 512K byte write-back, 32 bytes/line, sw flush: cache enabled
cpu1 at mainbus0: mid 10: RT620/625 @ 200 MHz, on-chip FPU
cpu1: 512K byte write-back, 32 bytes/line, sw flush: cache enabled
...

db{0}> bt
cpu_Debugger(0xf4221638, 0x0, 0x41, 0x59a9, 0xffff, 0xf03f11b0) at 
netbsd:zsc_in
tr_hard+0xf4
zsc_intr_hard(0xf42215f8, 0x0, 0x0, 0x0, 0xffff, 0xf4211000) at 
netbsd:zshard+0x
8
zshard(0xf42215f8, 0xf033502c, 0xf00, 0x1e8000e6, 0x0, 0xf4211000) at 
netbsd:spa
rc_interrupt44c+0x150
sparc_interrupt44c(0xf0502000, 0xf408eb00, 0x0, 0x0, 0x1e4000e7, 
0xf03fbaf0) atn
etbsd:sched_curcpu_runnable_p+0x24
sched_curcpu_runnable_p(0x0, 0x0, 0x1, 0xf0401140, 0xf03f4c00, 
0xf0002800) at ne
tbsd:idle_loop+0xf8
idle_loop(0xf408eb00, 0xf408eb00, 0x1, 0x0, 0xf0502000, 0xf03d35d8) at 
netbsd:lw
p_setfunc_trampoline

After bt command, debugger hangs :-(

Regards,

JKB