Hide Forgot
After system boot up, cxgb4 interfaces do not show up. In dmesg, we saw the following errors: Chelsio T4 Network Driver - version 1.3.0-ko cxgb4 0001:01:00.4: Using 32-bit DMA via iommu cxgb4 0001:01:00.4: Using 32-bit DMA via iommu cxgb4 0001:01:00.4: corrupted VPD EEPROM, actual csum 101 cxgb4: probe of 0001:01:00.4 failed with error -22 ---uname output--- 2.6.32-118.el6.ppc64 ---Steps to Reproduce--- After system boot up, you can see above error messages in /var/log/messages. ---Kernel - Network Drivers Component Data--- Stack trace output: no Oops output: no System Dump Info: The system is not configured to capture a system dump. Here is the rhel6.1 feature bug to enable this new hardware in RHEL6.1. https://bugzilla.redhat.com/show_bug.cgi?id=647006 Please provide the following - 1.Server architecture(s) (please list all effected) (x86/POWER6/Z/etc.): Power 7 2.Server type (9117-MMA/HS20/s390/etc.): Jupiter 3.General component (desktop/kernel/base OS/dev tools/etc.): kernel 4.Other components involved (ixgbe/java/emulex/etc.): no 5.Does the server have the latest GA firmware? yes 6.Has the problem been shown to occur on more than one system? yes 7.Is a tested patch available? no If yes to the above, has it been approved upstream? n/a 8.What is the latest official Red Hat build on which this bug has been seen? RHEL 6.1 Alpha
As it stands today, https://bugzilla.redhat.com/attachment.cgi?id=467853 is the only cxgb4 patch pulled in RHEL6.1. But it is not the problem here. The problem is that the cxgb4 driver included in 2.6.32-118.el6 does not match the driver submitted originally by Chelsio. Chelsio's cxgb4 original update included a file named cxgb4_compat.h. This file was implementing backports of kernel routines used upstream, but not present in RHEL6.0. For example, it was backporting the kernel routines implementing VPD management - pci_vpd_lrdt_size(), pci_vpd_info_field_size(), pci_vpd_find_info_keyword(). Now cxgb4_compat.h was moved out of the the cxgb4 code in 2.6.32-118.el6. And the upstream commits using these backported kernel routines were also filtered out of the cxgb4 code. For example, in this case, this upstream commit has been dropped: commit 226ec5fd6746c0ef2e9efb583b44d01264ea0bb5 Author: Dimitris Michailidis <dm> Date: Tue Apr 27 12:24:15 2010 +0000 cxgb4: parse the VPD instead of relying on a static VPD layout This is the root cause of this particular bug. In the presetn case, it looks like these VPD management routines have been backported from upstream to RHEL6.1-alpha1 core code and are in use by 2 drivers, tg3 and bnx2. So cxgb4 commit 226ec5fd6746c0ef2e9efb583b44d01264ea0bb5 can be pulled. We are in the process of identifying the missing upstream patches out of the original submission.
------- Comment From tpnoonan.com 2011-03-15 16:38 EDT------- please consider as exception for rhel6.1
https://brewweb.devel.redhat.com/taskinfo?taskID=3205274 That build should have the vpd fix for you steve, thanks!
changes are coming in for this with bz 647006 *** This bug has been marked as a duplicate of bug 647006 ***