[Qemu-devel] More playing with Win2k

[Derrik Pates <dpates@dsdk12.net>]

After trying to run it again with cores enabled, and tracing through the 
new core, it seems to be SEGVing in a different place, but with a 
strangely similar behavior. This time it died as follows:

#0  0x1000d484 in vga_update_display () at /misc/tmp/qemu/vga.c:1553
1553        if (s->ds->depth == 0) {
(gdb) bt
#0  0x1000d484 in vga_update_display () at /misc/tmp/qemu/vga.c:1553
#1  0x10011fe0 in sdl_refresh (ds=0x100794cc) at /misc/tmp/qemu/sdl.c:212
#2  0x10006b60 in main_loop (opaque=0x102090f8) at /misc/tmp/qemu/vl.c:3097
#3  0x100075b4 in main (argc=268763136, argv=0x7ffff2a0)
     at /misc/tmp/qemu/vl.c:3619
(gdb) print s
$1 = (VGAState *) 0x102090f8
(gdb) print s->ds
$2 = (DisplayState *) 0xff00aa55

Strangely, the last time I ran it, all the members of the ide_table[] 
array in ide.c got scribbled over with that same valie (0xff00aa55). It 
dies in different places, but with a common outcome. Any ideas how I 
could nail this problem down?

--
Derrik Pates
dpates@dsdk12.net