[haiku-bugs] Re: [Haiku] #14557: KDL when booting on Intel "Hades Canyon" NUC in XHCI::XHCI(pci_info info, Stack stack)

From: "Haiku" <trac@xxxxxxxxxxxx>
To: undisclosed-recipients: ;
Date: Mon, 15 Apr 2019 19:05:20 -0000

#14557: KDL when booting on Intel "Hades Canyon" NUC in XHCI::XHCI(pci_info
*info,
Stack *stack)
--------------------------------+-------------------------
   Reporter:  robg2             |      Owner:  nobody
       Type:  bug               |     Status:  reopened
   Priority:  normal            |  Milestone:  Unscheduled
  Component:  Drivers/USB/XHCI  |    Version:  R1/beta1
Resolution:                    |   Keywords:
Blocked By:  12885             |   Blocking:
Has a Patch:  0                 |   Platform:  All
--------------------------------+-------------------------

Comment (by waddlesplash):

Also, it seems there is no reference to any unexpected data being read

so I guess your check for 0xffffffff is actually hiding a different bug.

0xffffffff is most definitely invalid data; it's traditional for all reads
to PCI I/O space of dead devices to just return that. So it's clear the
controller is not booting anyway.

If I understand this correctly, it actually finds 4 controllers.

Nope, it sees 4 busses on 2 controllers. Notice:
{{{
[ 0.840234] xhci_hcd 0000:00:14.0: xHCI Host Controller
[ 0.843654] xhci_hcd 0000:02:00.0: xHCI Host Controller
}}}
Each of these lines appears twice, with a different bus ID on the
following line. So Haiku is at least seeing the correct information here.

--
Ticket URL: <https://dev.haiku-os.org/ticket/14557#comment:9>
Haiku <https://dev.haiku-os.org>
The Haiku operating system.

[haiku-bugs] Re: [Haiku] #14557: KDL when booting on Intel "Hades Canyon" NUC in XHCI::XHCI(pci_info *info, Stack *stack)

Other related posts:

[haiku-bugs] Re: [Haiku] #14557: KDL when booting on Intel "Hades Canyon" NUC in XHCI::XHCI(pci_info info, Stack stack)