qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] qemu-system-ppc TCG assert with git master


From: Mark Cave-Ayland
Subject: Re: [Qemu-devel] qemu-system-ppc TCG assert with git master
Date: Thu, 12 Mar 2015 23:38:03 +0000
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Icedove/31.5.0

On 12/03/15 16:51, Bastian Koppelmann wrote:

Hi Bastian,

> On 03/12/2015 03:41 PM, Richard Henderson wrote:
>> On 03/12/2015 01:41 AM, Mark Cave-Ayland wrote:
>>> Whilst testing git master in preparation for some OpenBIOS updates, I'm
>>> seeing the following TCG assert in one of my older test images:
>>>
>>>
>>> $ gdb --args ./qemu-system-ppc -cdrom
>>> /home/build/src/qemu/image/ppc/ubuntu-5.10-live-powerpc.iso -boot d -g
>>> 800x600x8
>> Older, eh?  I can only find 5.04.  ;-)
>>
>> And, fwiw, that doesn't assert.  I get all the way to
>>
>> Kernel panic - not syncing: VFS: unable to mount root fs on
>> unknown-block(1,0)
>>
>> So, I guess I need something different to
>>
>> http://cdimage.ubuntu.com/ubuntu-mate/daily-live/current/vivid-desktop-powerpc.iso
>>
>>
>> to reproduce this.
>>
>>
>>
>> r~
>>
> 
> Hi Richard,
> 
> I can reproduce it fine here with
> http://old-releases.ubuntu.com/releases/5.10/.
> As far as I see it, the Instruction at 0xc9105830 generates a mov 0 to
> r30, which is optimized by tcg_opt_gen_mov and r30 is set to
> TCG_TEMP_CONST. When 0xc9105838 or r0, r30, r30 should be optimized r30
> is already in the state TCG_TEMP_CONST and the assert fails.

Ah great, you're able to reproduce the same issue locally. Hopefully
this will give you both enough information to figure out what is
happening in the optimiser...


Many thanks,

Mark.




reply via email to

[Prev in Thread] Current Thread [Next in Thread]