7 Mayıs 2019 Salı 08:52:46 UTC+3 tarihinde Jan Kiszka yazdı:
> On 06.05.19 20:38, Hakkı Kurumahmut wrote:
> > 6 Mayıs 2019 Pazartesi 18:19:07 UTC+3 tarihinde [email protected] yazdı:
> >> For sure ! You can find the requested file attached.
> >>
> >> best regards,
> >>
> >>
> >> Le lundi 6 mai 2019 16:11:05 UTC+2, Jan Kiszka a écrit :
> >>> On 06.05.19 15:47, [email protected] wrote:
> >>>> Hello,
> >>>>
> >>>> I'm trying to run jailhouse on a HP ProLiant ML350 Gen10.
> >>>> ACPI DMAR is available, but my hardware-check fails with this
> >>>>
> >>>> error :
> >>>>  "Unsupported DMAR Device Scope Structure"
> >>>>
> >>>> I also tried on a HP ProLiant DL360p Gen8 but I ended up with the same 
> >>>> error...
> >>>>
> >>>>
> >>>> Any idea of this trouble ?
> >>>>
> >>>
> >>> Can you share an archive generated by "jailhouse config collect"? Then we 
> >>> can
> >>> analyze locally.
> >>>
> >>> Jan
> >>>
> >>> -- 
> >>> Siemens AG, Corporate Technology, CT RDA IOT SES-DE
> >>> Corporate Competence Center Embedded Linux
> > 
> > DMAR region parser is detected Device Scope size more than 8 byte.
> > 
> > 
> > VT-d Spec at 8.3.1
> > 
> > If the ‘Path’ field length is more than 2 bytes (N > 1), the Device Scope
> > Entry identifies a device behind one or more system software visible PCI-
> > PCI bridges. Bus rebalancing actions by system software modifying bus
> > assignments of the device’s parent bridge impacts the bus number portion
> > of device’s requester-id.
> > 
> > This devices
> > 
> > 1C.7.0.0
> > 1C.7.0.2
> > 
> > 
> > Current Script does not support this type device... Scripts must be read 
> > Dev Scope length and than it is read the DMAR file to collect correct data.
> > 
> > 
> > Jan can provide technical information on this issue. I have no idea why 
> > it's not supported.
> > 
> 
> I think I was "just" too lazy to implement the needed logic behind that.
> 
> The tricky thing here is surely the identification of the final bus number 
> the 
> BIOS or Linux as chosen for the device described that way. It will be needed 
> in 
> order to identify the device entry in our configuration. Likely, we have to 
> walk 
> the topology, first identify the bridge and then query that one for its bus 
> number in order to describe the device.
> 
> Jan
> 
> -- 
> Siemens AG, Corporate Technology, CT RDA IOT SES-DE
> Corporate Competence Center Embedded Linux

Hi,

Temporary solution but is not solution is below. Change relative function with 
them. Hardware check will be pass but new problems may be seen.


def parse_dmar_devscope(f):
    (scope_type, scope_len) = struct.unpack('<BB', f.read(2))

    N = (int)((scope_len - 6) / 2) - 1

    (id, starting_bus, starting_dev, starting_fn) = \
            struct.unpack('<xxBBBB', f.read(6))

    while N != 0:
        N-=1
        (secondary_dev, secondary_fn) = \
            struct.unpack('<BB', f.read(2))

    return (scope_type, scope_len, id, starting_bus, starting_dev, starting_fn)


Feature                         Availability
------------------------------  ------------------
Number of CPUs > 1              ok
Long mode                       ok
x2APIC                          ok

VT-x (VMX)                      ok
  VMX outside SMX               ok
  VMX inside SMX                ok
  IA32_TRUE_*_CLTS              ok
  NMI exiting                   ok
  Preemption timer              ok
  I/O bitmap                    ok
  MSR bitmap                    ok
  Secondary controls            ok
  Optional CR3 interception     ok
  Virtualize APIC access        ok
  RDTSCP                        ok
  Unrestricted guest            ok
  EPT                           ok
    4-level page walk           ok
    EPTP write-back             ok
    2M pages                    ok
    1G pages                    missing (optional)
    INVEPT                      ok
      Single or all-context     ok
  VM-exit save IA32_PAT         ok
  VM-exit load IA32_PAT         ok
  VM-exit save IA32_EFER        ok
  VM-exit load IA32_EFER        ok
  VM-entry load IA32_PAT        ok
  VM-entry load IA32_EFER       ok
  Activity state HLT            ok

VT-d (IOMMU #0)                 ok
  39-bit AGAW                   ok
  48-bit AGAW                   ok
  2M pages                      ok
  1G pages                      ok
  Queued invalidation           ok
  Interrupt remapping           ok
  Extended interrupt mode       ok

VT-d (IOMMU #1)                 ok
  39-bit AGAW                   ok
  48-bit AGAW                   ok
  2M pages                      ok
  1G pages                      ok
  Queued invalidation           ok
  Interrupt remapping           ok
  Extended interrupt mode       ok

Check passed!

-- 
You received this message because you are subscribed to the Google Groups 
"Jailhouse" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jailhouse-dev/36a1c92b-732e-4f4a-81e4-159b89919f84%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to