On Tue, 22 Aug 2017 16:27:38 +0200 Halil Pasic <pa...@linux.vnet.ibm.com> wrote:
> On 08/22/2017 04:06 PM, Cornelia Huck wrote: > > On Tue, 22 Aug 2017 15:24:34 +0200 > > Cornelia Huck <coh...@redhat.com> wrote: > > > >> On Tue, 22 Aug 2017 14:58:37 +0200 > >> Halil Pasic <pa...@linux.vnet.ibm.com> wrote: > >> > >>> On 08/22/2017 11:39 AM, Cornelia Huck wrote: > >>>> On Tue, 22 Aug 2017 11:20:51 +0200 > >>>> Halil Pasic <pa...@linux.vnet.ibm.com> wrote: > > > >>>>> Well, the QEMU name of the command is misleading misleading. In the AR > >>>>> it's called 'Configure I/O Adapter'. The PCI comes into the picture via > >>>>> byte 8 of the SCCB, the so called adapter type. Valid values for the > >>>>> adapter type are: 00-01 reserved; 02 PCI function; 03-FF reserved. So > >>>>> at this point we only have PCI. > > > > OK, I need one more piece of information. > > > > We obviously need to check whether the sccb we got is long enough > > before we try to access the command-specific field. How long is the > > sccb supposed to be for configure I/O adapter? For pci, 16 bytes; in > > general, I would guess that it needs to include at least atype and some > > placeholder for the payload. What does the AR say? > > > > The first 2 bytes of the SCCB designate it's length. For this particular > command it's at least 16 bytes (regardless of pci). The length is > marked as may be changed by the SCLP. Thanks for the info, this makes implementing it correctly much easier! > > > > Looking at the pci code, I also noted that it cheerfully uses the aid > > field of the sccb before checking whether it is actually long enough... > > >
