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

Re: [Xen-devel] [Qemu-devel] Hvmloader: Modify ACPI to only supply _EJ0 methods for PCIslots that support hotplug by runtime patching



> -----Original Message-----
> From: Fabio Fantoni [mailto:fabio.fantoni@xxxxxxx]
> Sent: Thursday, October 24, 2013 8:58 PM
> To: Gonglei (Arei); Jan Beulich
> Cc: anthony.perard@xxxxxxxxxx; Ian Campbell; Stefano Stabellini; Gaowei (UVP);
> Hanweidong (Randy); Huangweidong (Hardware); Luonengjun; Yanqiangjun;
> xen-devel@xxxxxxxxxxxxx; qemu-devel@xxxxxxxxxx; Markus Armbruster
> Subject: Re: [Qemu-devel] [Xen-devel] Hvmloader: Modify ACPI to only supply
> _EJ0 methods for PCIslots that support hotplug by runtime patching
> 
> Il 24/10/2013 14:17, Gonglei (Arei) ha scritto:
> >> -----Original Message-----
> >> From: Jan Beulich [mailto:JBeulich@xxxxxxxx]
> >> Sent: Tuesday, October 22, 2013 4:06 PM
> >> To: Gonglei (Arei)
> >> Cc: anthony.perard@xxxxxxxxxx; Ian Campbell; Stefano Stabellini; Gaowei
> (UVP);
> >> Hanweidong (Randy); Huangweidong (Hardware); Luonengjun; Yanqiangjun;
> >> xen-devel@xxxxxxxxxxxxx; Fabio Fantoni; qemu-devel@xxxxxxxxxx; Markus
> >> Armbruster
> >> Subject: RE: [Qemu-devel] [Xen-devel] Hvmloader: Modify ACPI to only
> supply
> >> _EJ0 methods for PCIslots that support hotplug by runtime patching
> >>
> >>>>> On 22.10.13 at 06:08, "Gonglei (Arei)" <arei.gonglei@xxxxxxxxxx>
> wrote:
> >>> Hi, guys. The new patch has been modified based on the principles you
> >>> suggested, thank you so much.
> >>> Last time I test the patch based on the codes of 4.3.0.
> >>> This time, I found that the system based on the codes of trunk causes the
> VM
> >>> reboot again and again, which I have not found out the reason.
> >>> So i can not test the patch based on the codes of trunk (details in
> >>> EJ0_ACPI_PCI_Hotplug.patch)..
> >> I'm afraid we will need you to figure out that problem first, and
> >> then do the verification on -unstable. Even if the code shouldn't
> >> be that different from 4.3, we still don't want to apply completely
> >> untested stuff.
> > Hi, Jan. We found that the reason that we used a wrong seabios PATH, and
> the hvmloader can't load the bios.bin.
> > So the VM restart again and again after we start it. That's our fault.
> >
> > Now I test the patch based on the codes of trunk, which works well.
> > The patch has been modified after your suggestion.
> > The patch works well with upstream qemu and doesn't affect the system with
> traditional qemu.
> >
> >>
> >>> --- a/tools/firmware/hvmloader/ovmf.c
> >>> +++ b/tools/firmware/hvmloader/ovmf.c
> >>> @@ -79,7 +79,11 @@ static void ovmf_acpi_build_tables(void)
> >>>           .dsdt_anycpu = dsdt_anycpu,
> >>>           .dsdt_anycpu_len = dsdt_anycpu_len,
> >>>           .dsdt_15cpu = NULL,
> >>> -        .dsdt_15cpu_len = 0
> >>> +        .dsdt_15cpu_len = 0,
> >>> +        .aml_ej0_name = NULL,
> >>> +        .aml_adr_dword = NULL,
> >>> +        .aml_ej0_name_len = 0,
> >>> +        .aml_adr_dword_len = 0,
> >> I don't see why you're adding these.
> >>
> > Insurance purposes is that just initialize the struct.
> >
> > Signed-off-by: Gaowei <gao.gaowei@xxxxxxxxxx>
> > Signed-off-by: gonglei <arei.gonglei@xxxxxxxxxx>
> 
> Tested-by: Fabio Fantoni <fabio.fantoni@xxxxxxx>
> 
> Tested on xen unstable with qemu 1.6.1, no problem found for now.
> Only one question: this patch remove hotplug only from essentials pci
> device, right?
> On windows 7 hotplug continues to show: virtio-serial driver, xen pci
> device driver and hd audio.
> 
It depends on the property of hotplug of pci devices' class emulated by 
upstream qemu.
If you set k->no_hotplug = 1 in class_init function for those pci devices, 
which will not be shown in the Windows guest any more.

> Thanks for any reply.
> 

Best regards,
-Gonglei
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.