Isaku Yamahata wrote:
> On Mon, Oct 06, 2008 at 11:16:02AM +0800, Zhang, Jingke wrote:
>> Isaku Yamahata wrote:
>>> On Mon, Oct 06, 2008 at 09:54:13AM +0800, Zhang, Jingke wrote:
>>>> Hi all,
>>>> There is no new regression found in Cset#18580. Three old
>>>> issues still existed.
>>>> Old issue (3):
>>>> 1. Qemu will disappear and VTI hangs when we do 'xm restore'
>>>> and 'xm migrate'
>>> Hi. Does this still exist?
>>> The change set 18578:ba543f51c6f1 should fix this.
>>> At least works for me with xm save/restore. I haven't tried xm
>>> migrate, though.
>>> Does qemu-dm exit with the following message?
>>> from xen_vga_vram_map() in hw/vga.c
>>> fprintf(stderr, "Failed to map vram nr_pfn=0x%lx
>>> vram_addr=%llx: %s\n", nr_pfn, (unsigned long
>>> long)vram_addr, strerror(errno)); exit(1);
>> Yes, my qemu-dm log shows: Failed to map vram nr_pfn=0x200
>> vram_addr=c0000000: Invalid argument
>> I used GFW#129, image type is QCOW. Since it works for you, what is
>> the difference between us? Thanks. :)
> Ah, I'm using patched GFW. I sent out two patches as
> But I guess Tristan is too busy to commit those.
> The patch in msg00067.html would fix the issue. Could you try that
Thanks, I see.
GFW building needs too much to set up the compile environment. So, I am afraid
it may need some time for me to check it.
Anyway, I will mark the old issue1 as a GFW issue first. Hope Tristan can check
it in soon. :)
>>>> 2. With 4096M, VTI can not boot up emulated rtl8139
>>>> 3. With 4095M, qemu of VTI guest will disappear when booting
Xen-ia64-devel mailing list