Public bug reported: The e1000e driver in 2.6.27 (not tested with 2.6.26, but it was absolutely fine with whatever driver was used by 2.6.24) seems to support writing to the LAN chip's firmware, and if it follows the behaviour of the OpenBSD driver[0] then the firmware is mapped as writable the entire time the driver is loaded. As has happened to my Thinkpad X300 and several other people[1][2], it seems that other kernel bugs which trash memory can end up spewing nonsense into the LAN firmware and thus rewriting it with something that isn't the code required to drive a LAN port. There is no simple recovery strategy from this, and Intel's downloadable tools (IABUTIL.EXE) will not repair it, and worse, they will destroy the firmware enough that the device will no longer enumerate on the PCI bus. Intel subsequently say that these tools are only for use with Desktop parts and the fact that they run at all on Laptop parts is a bug[3].
I am returning my laptop to Lenovo to have the motherboard replaced, and without a strategy to be able to repair the firmware myself, I am very much reluctant to test this any further when I get the machine back, as if I hit the bug again I will be once again left with broken hardware. [0] http://www.blahonga.org/~art/rant.html (search for "em0") [1] http://www.mail-archive.com/[EMAIL PROTECTED]/msg00360.html [2] http://bugzilla.kernel.org/show_bug.cgi?id=11382 [3]http://www.mail-archive.com/[EMAIL PROTECTED]/msg00398.html ** Affects: linux (Ubuntu) Importance: Critical Status: New ** Changed in: linux (Ubuntu) Importance: Undecided => Critical -- [intrepid] 2.6.27 e1000e kernel places Intel gigE chipsets at risk https://bugs.launchpad.net/bugs/263555 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
