Re: [Qemu-devel] qemu-system-arm -M akita/terrier - which roms work? (more patches... :)

[Juergen Lock]

On Wed, Aug 01, 2007 at 01:32:00AM +0200, Juergen Lock wrote:
> On Tue, Jul 31, 2007 at 11:31:58PM +0200, andrzej zaborowski wrote:

> > can you try this change:
> > --- a/target-arm/helper.c
> > +++ b/target-arm/helper.c
> > @@ -840,7 +840,7 @@ void helper_set_cp15(CPUState *env, uint32_t insn, 
> > uint32_t
> > val)
> >              if (op2 == 0 && crm == 1) {
> >                  /* Changes cp0 to cp13 behavior, so needs a TB flush.  */
> >                  tb_flush(env);
> > -                env->cp15.c15_cpar = (val & 0x3fff) | 2;
> > +                env->cp15.c15_cpar = val & 0x3fff;
> >                  break;
> >              }
> >              goto bad_reg;
> > 
>  Hmm that fixed the undefined instruction crahes, at least ps, more
> and dc now work.  The gui still doesn't start tho.

Okay I now updated qemu again and added an strace to the gui (qpe)
invocation, and it ends like this:
...
[pid   299] [4086c154] open("/etc/group", O_RDONLY) = 0
[pid   299] [4086c37c] fcntl64(0, F_GETFD) = 0
[pid   299] [4086c37c] fcntl64(0, F_SETFD, FD_CLOEXEC) = 0
[pid   299] [4086b350] fstat64(0, {st_mode=S_IFREG|0664, st_size=254, ...}) = 0
[pid   299] [40875fcc] old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, 
MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40019000
[pid   299] [40878fa0] _llseek(0, 0, [0], SEEK_CUR) = 0
[pid   299] [4086c1a4] read(0, "root:x:0:\nwheel:x:10:\nbin:x:1:bi"..., 4096) = 
254
[pid   299] [4086c194] close(0)         = 0
[pid   299] [40875fe4] munmap(0x40019000, 4096) = 0
[pid   299] [40851cc4] geteuid32()      = 0
[pid   299] [40872f38] setreuid32(0x1f4, 0) = 0
[pid   299] [40873034] setregid32(0x1f4, 0x1f4) = 0
pid 299 stray syscall exit
) = 500
[pid   299] upeek: ptrace(PTRACE_PEEKUSER,299,60,0): No such process
[????????] +++ killed by SIGKILL +++
Process 300 detached

 Is the `stray syscall exit' caused by the SIGKILL?  And anyone have
an idea where a SIGKILL could come from?  I don't think it is running
out of memory, `free' shows about half of the memory free shortly
before it exits, and also it works on the real hardware...

 Oh and something else: even with a valid -sd image I can't seem
to mount /dev/mmcda1 in the guest, and I see the following in dmesg:

pxa_sd_wait_response: card removed (cmd=00)
pxa_sd_wait_response: card removed (cmd=00)
pxa_sd_wait_response: card removed (cmd=00)
pxa_sd_wait_id_response: card removed (cmd=00)
sharp_mmcsd 0.30 13 Oct 2004
pxa_sd_wait_response: card removed (cmd=00)
pxa_sd_wait_response: card removed (cmd=00)
pxa_sd_wait_response: card removed (cmd=00)
pxa_sd_wait_id_response: card removed (cmd=00)

 (It's not that big an issue tho, since at least with the terrier
emulation I do have -hda which works.  Of course working network
would be even nicer...  I guess emulating the usb netowrk thingy
would take some work, but could the ne2k optionally be hooked up to
pcmcia socket 0?)

 Cheers,
        Juergen