Red Hat Bugzilla – Bug 161810
modules.pcimap handling broken
Last modified: 2014-03-16 22:54:42 EDT
+++ This bug was initially created as a clone of Bug #146213 +++
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.5)
Gecko/20041228 Firefox/1.0 Fedora/1.0-8
Description of problem:
I have discovered the reason why the nvidia driver doesn't autoload
properly at bootup, like any other PCI device should. Kudzu's
handling of modules.pcimap doesn't match the kernel's algorithm.
Specifically, Kudzu doesn't handle the case where a vendor is defined
but device is PCI_ANY_ID (0xffffffff). It also totally ignores the
class and class_mask. I added another device entry for my specific
card, and then kmodule properly reported "VIDEO nvidia".
# pci module vendor device subvendor subdevice class
nvidia 0x000010de 0xffffffff 0xffffffff 0xffffffff
0x00030000 0xffffffff 0x0
nvidia 0x000010de 0x00000140 0xffffffff 0xffffffff
0x00000000 0x00000000 0x0
This problem isn't specific to the nvidia driver. Any PCI device can
legitamately (by the kernel definition) use such a construct.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
Use a kernel module which defines a PCI vendor and class, but no
Actual Results: kmodule did not discover the device.
Expected Results: kmodule returns the proper driver module for the
Fixed in 220.127.116.11-1.
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on the solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.