Re: [Xen-devel] PCI passthrough to QEMU traditional stubdom not working when option ROM present

2016-10-24 Thread Jan Beulich
>>> On 24.10.16 at 15:56, wrote: > On Mon, Oct 24, 2016 at 9:12 AM, Jan Beulich wrote: > On 24.10.16 at 15:01, wrote: >>> On Mon, Oct 24, 2016 at 8:47 AM, Jan Beulich wrote: >>> On 24.10.16 at 14:30,

Re: [Xen-devel] PCI passthrough to QEMU traditional stubdom not working when option ROM present

2016-10-24 Thread Eric Shelton
On Mon, Oct 24, 2016 at 9:12 AM, Jan Beulich wrote: On 24.10.16 at 15:01, wrote: >> On Mon, Oct 24, 2016 at 8:47 AM, Jan Beulich wrote: >> On 24.10.16 at 14:30, wrote: As best as I can tell, the current

Re: [Xen-devel] PCI passthrough to QEMU traditional stubdom not working when option ROM present

2016-10-24 Thread Samuel Thibault
Jan Beulich, on Mon 24 Oct 2016 07:12:27 -0600, wrote: > And btw - is it really qemu that does this? It's pciaccess which does it. > I didn't think it would be a user space app doing any such sizing; I > thought this would get done elsewhere in stubdom/mini-os. mini-os just lets libpciaccess

Re: [Xen-devel] PCI passthrough to QEMU traditional stubdom not working when option ROM present

2016-10-24 Thread Jan Beulich
>>> On 24.10.16 at 15:01, wrote: > On Mon, Oct 24, 2016 at 8:47 AM, Jan Beulich wrote: > On 24.10.16 at 14:30, wrote: >>> As best as I can tell, the current code already writes out the proper >>> value for sizing ("pci_write_long(d,

Re: [Xen-devel] PCI passthrough to QEMU traditional stubdom not working when option ROM present

2016-10-24 Thread Eric Shelton
On Mon, Oct 24, 2016 at 8:47 AM, Jan Beulich wrote: On 24.10.16 at 14:30, wrote: >> As best as I can tell, the current code already writes out the proper >> value for sizing ("pci_write_long(d, reg, ~0)" in the above code >> snippet). > > No - for

Re: [Xen-devel] PCI passthrough to QEMU traditional stubdom not working when option ROM present

2016-10-24 Thread Jan Beulich
>>> On 24.10.16 at 14:30, wrote: > As best as I can tell, the current code already writes out the proper > value for sizing ("pci_write_long(d, reg, ~0)" in the above code > snippet). No - for sizing a ROM BAR, only the address portion is supposed to be written with all 1s as

Re: [Xen-devel] PCI passthrough to QEMU traditional stubdom not working when option ROM present

2016-10-24 Thread Eric Shelton
On Mon, Oct 24, 2016 at 6:19 AM, Jan Beulich wrote: On 21.10.16 at 15:23, wrote: >> Eric Shelton, on Fri 21 Oct 2016 09:01:43 -0400, wrote: >>> ERROR: PCI region size must be pow2 type=0x8, size=0xdf08 >> >>> u32 u =

Re: [Xen-devel] PCI passthrough to QEMU traditional stubdom not working when option ROM present

2016-10-24 Thread Jan Beulich
>>> On 21.10.16 at 15:23, wrote: > Eric Shelton, on Fri 21 Oct 2016 09:01:43 -0400, wrote: >> ERROR: PCI region size must be pow2 type=0x8, size=0xdf08 > >> u32 u = pci_read_long(d, reg); >> if (u != 0x) >> -

Re: [Xen-devel] PCI passthrough to QEMU traditional stubdom not working when option ROM present

2016-10-21 Thread Samuel Thibault
Hello, Eric Shelton, on Fri 21 Oct 2016 09:01:43 -0400, wrote: > ERROR: PCI region size must be pow2 type=0x8, size=0xdf08 > u32 u = pci_read_long(d, reg); > if (u != 0x) > - d->rom_base_addr = u; > +{ > + d->rom_base_addr = u; >