./qemu-system-sparc64 -M niagara -L /git/md/bin/tlb_fix/2048 -nographic -m 2048 -drive if=pflash,readonly=on,file=/zroot/iso/sol-11-1111-text-sparc.iso
cpu Probing I/O buses


Sun Fire T2000, No Keyboard
Copyright 2005 Sun Microsystems, Inc.  All rights reserved.
OpenBoot 4.20.0, 2048 MB memory available, Serial #1122867.
[mo23723 obp4.20.0 #0]
Ethernet address 0:80:3:de:ad:3, Host ID: 80112233.



ok boot -kdv
Boot device: /virtual-devices/disk@0  File and args: -kdv
hsfs-file-system
Loading: /platform/sun4v/boot_archive
ramdisk-root ufs-file-system
Loading: /platform/sun4v/kernel/sparcv9/unix
module /platform/sun4v/kernel/sparcv9/unix: text at [0x1000000, 0x10e2677] data at 0x1800000
module /platform/sun4v/kernel/sparcv9/genunix: text at [0x10e2678, 0x1306ae7] data at 0x19373c0
module /platform/sun4v/kernel/misc/sparcv9/platmod: text at [0x1306ae8, 0x1306aef] data at 0x198d698
module /platform/sun4v/kernel/cpu/sparcv9/SUNW,UltraSPARC-T1: text at [0x1306b00, 0x13096f7] data at 0x198dd00
Loading kmdb...
module /platform/sun4v/kernel/misc/sparcv9/kmdbmod: text at [0x1309700, 0x138a397] data at 0x19a5e50
module /kernel/misc/sparcv9/ctf: text at [0x138a398, 0x1391f9f] data at 0x19c33b0
/
Welcome to kmdb
kmdb: unable to determine terminal type: assuming `vt100'
Loaded modules: [ unix krtld genunix ]
[0]> :c
SunOS Release 5.11 Version 11.0 64-bit
Copyright (c) 1983, 2011, Oracle and/or its affiliates. All rights reserved.
os-io Ethernet address = 0:80:3:de:ad:3
Using default device instance data
mem = 2097152K (0x80000000)
avail mem = 1748688896
root nexus = Sun Fire T2000
pseudo0 at root
pseudo0 is /pseudo
scsi_vhci0 at root
scsi_vhci0 is /scsi_vhci
NOTICE: Disabling watchdog as watchdog services are not available

(hard hang -- no further console output at all after this point. Kernel
init itself completes cleanly (root nexus, pseudo bus, scsi_vhci,
watchdog-disable check all print normally); the hang happens AFTER
init, at the point where it should move on to device attach
(ramdisk0/dld0) and never does. Re-tested against the corrected build
(post hv-config memsize fix, see src/tlb_fix/2048/1up.hdesc history) --
this is a DIFFERENT outcome than before the fix, which got past device
attach (through "ramdisk0 at root" and "pseudo-device: dld0") before
panicking with "bad unexpected error from hypervisor call at TL 1".
With the corrected 2048M, it now hangs one step earlier in the device
attach sequence instead. Another genuine behavior change caused by the
memory correction, same as the sol-10-u8 and s10u11 cells.)

--- follow-up attempt: tried to break in via serial break (Ctrl-A b) ---
[0]> :c
kmdb: single-step stop on miscellaneous trap
kmdb: target stopped at:
kmdb_enter+4:   retl
[0]> :c
SunOS Release 5.11 Version 11.0 64-bit
Copyright (c) 1983, 2011, Oracle and/or its affiliates. All rights reserved.
os-io Ethernet address = 0:80:3:de:ad:3
Using default device instance data

Type  'go' to resume
ok go
mem = 2097152K (0x80000000)
avail mem = 1748688896

(hard hang again -- and this time further serial breaks stopped being
acknowledged at all ("kmdb stopped respond to break"). Two things worth
noting: (1) the first break landed at kmdb_enter+4 itself, an
uninformative trampoline location, not the actual stuck point; (2) the
second break dropped to the OBP "ok" prompt requesting 'go' to resume,
a different intercept path than kmdb catching it directly. This run
appears to hang around/right after "avail mem =", earlier than the
previous run's hang point (after the watchdog-disable notice) -- though
that could be genuine non-determinism in where this stalls, or just
where this particular attempt was stopped/observed. Unlike the
alloc_page_freelists hang (tlb_fix-2048__snv_134-ai-iso.txt etc.), this
one did not yield a clean, reproducible break-in point.)