From Bugzilla Helper: User-Agent: Mozilla/4.0 (compatible; MSIE 5.0; Windows NT; DigExt) Description of problem: Upgrade from 2.4.2-2 kernel to 2.4.3-12 kernel. Running on a IBM netfinity 7000 quad processor host with 1024 MB of memory. I have 2 Qlogic 2200 and an Intel ethernet pro100 nic card in the pci slots. After upgrading successfully rebooting the host modules for those cards would not load, saying that there are no devices available. I have an adaptec scsi host adapter and that loads fine with the aic7xxx module. How reproducible: Always Steps to Reproduce: 1.Load kernel 2.4.3-12 2.Create new ramdisk and edit lilo.conf to load new kernel 3.Reboot the host Actual Results: All pci devices are no longer accessible. Expected Results: They should still be accessible. Additional info: Output from lspci: 00:00.0 ISA bridge: Intel Corporation 82371AB PIIX4 ISA (rev 02) 00:00.1 IDE interface: Intel Corporation 82371AB PIIX4 IDE (rev 01) 00:00.2 USB Controller: Intel Corporation 82371AB PIIX4 USB (rev 01) 00:00.3 Bridge: Intel Corporation 82371AB PIIX4 ACPI (rev 02) 00:01.0 PCI bridge: IBM IBM27-82351 (rev 07) 00:0a.0 VGA compatible controller: S3 Inc. Trio 64V2/DX or /GX (rev 16) 00:10.0 Host bridge: Intel Corporation 450NX - 82451NX Memory & I/O Controller (rev 03) 00:12.0 Host bridge: Intel Corporation 450NX - 82454NX/84460GX PCI Expander Bridge (rev 02) 00:13.0 Host bridge: Intel Corporation 450NX - 82454NX/84460GX PCI Expander Bridge (rev 02) 00:14.0 Host bridge: Intel Corporation 450NX - 82454NX/84460GX PCI Expander Bridge (rev 02) 01:02.0 Bridge: IBM: Unknown device 00dc (rev 02) 01:04.0 SCSI storage controller: Adaptec AHA-2940U/UW / AHA-39xx / AIC- 7895 (rev 04) 01:04.1 SCSI storage controller: Adaptec AHA-2940U/UW / AHA-39xx / AIC- 7895 (rev 04)
I assume you reran lilo (no kidding, some people forget that)
yup ! I created a new ramdisk and ran lilo after making all the necessary changes in lilo.conf.
Ok I've to admit this puzzles me. I don't think the PCI code changed THAT much between 2.4.2-2 and 2.4.3-12....
I have other smp host that was upgraded to the 2.4.3-12smp kernel with the same set of pci devices and they are running fine. The host are Dell PowerEdge 2400.
Also see Bug #50856
We believe this will be fixed in the next rawhide kernel, 2.4.7-2.8 or later. If so, the bug was caused by a fix for another machine with the 450NX chipset...