RE: [Bug 263555] Re: [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 gigE chipsets at risk
Thanks, I still don't understand how it is that e1000e_recover didn't work for you, but I admit that I have been using it as a tool, and don't understand its inner workings. Let's try another approach to get the invalid NVM content listed, this time by the driver when it reads the data. I attach a patch "e1000e-1.0.15.shownvm.patch" which can be applied our latest e1000e sourceforge release. To install the driver, and collect that result , please proceeed as follows 1) Copy this patch to a local directory 2) Download e1000e-1.0.15.tar.gz from http://sourceforge.net/projects/e1000/files/ 3) Untar the tarball to a local directory, tar xvzf e1000e-1.0.15.tar.gz 4) cd e1000e-1.0.15/src 5) Apply the patch patch -p2 <../../e1000e-1.0.15.shownvm.patch 6) Remove the old driver, build & install the new one rmmod e1000e make insmod e1000e.ko 7) The system message log should have the NVM content that was read. The driver should also load even in the presence of the errored NVM. Please let me know whether it does load and work, and .send me the dmesg log that includes the NVM dump, and I will see if I can fix it up and return it to you with instructions on how to apply the fixed-up version, Thanks Dave From: boun...@canonical.com [boun...@canonical.com] On Behalf Of bonsiware [bonsiw...@gmail.com] Sent: Wednesday, October 07, 2009 10:28 AM To: Graham, David Subject: [Bug 263555] Re: [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 gigE chipsets at risk lspci -tv: +-19.0 Intel Corporation 82566DC Gigabit Network Connection lspci -xxx: 00:19.0 Ethernet controller: Intel Corporation 82566DC Gigabit Network Connection (rev 03) 00: 86 80 4b 10 03 01 10 00 03 00 00 02 00 00 00 00 10: 00 00 40 fe 00 40 42 fe 21 18 00 00 00 00 00 00 20: 00 00 00 00 00 00 00 00 00 00 00 00 86 80 00 00 30: 00 00 00 00 c8 00 00 00 00 00 00 00 0b 01 00 00 dmesg: [1.652371] e1000e: Intel(R) PRO/1000 Network Driver - 1.0.2-k2 [1.652374] e1000e: Copyright (c) 1999-2008 Intel Corporation. [1.652440] e1000e :00:19.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16 [1.652447] e1000e :00:19.0: pci_enable_pcie_error_reporting failed 0xfffb [1.652456] e1000e :00:19.0: setting latency timer to 64 [1.652628] alloc irq_desc for 28 on node -1 [1.652630] alloc kstat_irqs on node -1 [1.652647] e1000e :00:19.0: irq 28 for MSI/MSI-X [1.739443] ohci1394 :1c:03.4: PCI INT A -> GSI 16 (level, low) -> IRQ 16 [1.757784] :00:19.0: :00:19.0: The NVM Checksum Is Not Valid [1.787486] e1000e :00:19.0: PCI INT A disabled [1.787495] e1000e: probe of :00:19.0 failed with error -5 -- [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 gigE chipsets at risk https://bugs.launchpad.net/bugs/263555 You received this bug notification because you are a direct subscriber of the bug. Status in The Linux Kernel: Fix Released Status in “linux” package in Ubuntu: Fix Released Status in “linux-lpia” package in Ubuntu: Fix Released Status in linux in Ubuntu Intrepid: Fix Released Status in linux-lpia in Ubuntu Intrepid: Fix Released Status in “linux” package in Fedora: Fix Released Status in “linux” package in Gentoo Linux: Fix Released Status in “linux” package in Mandriva: Fix Released Status in “linux” package in Suse: Fix Released Bug description: In some circumstances it appears possible for the 2.6.27-rc kernels to corrupt the NVRAM used by some Intel network parts to store data such as MAC addresses. This is limited to the new e1000e driver, and reports have only appeared from users of "82566 and 82567 based LAN parts (ich8 and ich9)" (to quote Intel). The reports seem to be isolated to laptops, but it is not clear if this is because desktop/server parts are not vulnerable, or if use cases simply increase the chances of laptop users being hit. Once this corruption has occurred, recovery may be possible via a BIOS update, but may well require replacement of the hardware. Use of Intel's IABUTIL.EXE is strongly discouraged, as it will worsen the problem to the point where the network part will no longer appear on the PCI bus. (this is a new description, the original one was based on too much guesswork. Below are the URLs originally referenced) (the driver i blacklisted in Ubuntu for 2.6.27-rc in the latest releases, so if your network is not working, it doesn't have to be damaged, but just disabled in order to prevent any accidents until this bug is solved, don't wary!) http://www.blahonga.org/~art/rant.html (search for "em0") http://www.mail-archive.com/e1000-de...@lists.sourceforge.net/msg00360.html http://www.mail-archive.com/e1000-de...@lists.sourceforge.net/msg00398.html ** Attachment added: "e1000e-1.0.15.shownvm.patch" http://launchpadlibrarian.net/33264068/e1000e-1.0.15.shownvm.patch -- [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 g
RE: [Bug 263555] Re: [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 gigE chipsets at risk
This is unusual, as the e1000e_nvm -r –d eth0 -o ethtool.dmp command normally dumps out the 1Gb portion of the system flash even if it _does_ have a bad checksum, and then I've been fixing the checksum & content. Are you sure that there isn't an ethtool.dmp file created in the local directory from which you ran e1000e_nvm ? If there really is no ethtool.dmp, please send me 1) lspci -tv 2) lspci -xxx 3) dmesg (that includes the failure of the e1000e driver to load) and I'll send you an instrumented driver that will dump out the 1Gb flash content, and we may be able to fix it from there. At least I hope so. As I say, I've fixed a lot of these corruptions, but do not recall seeing this particular failure mode before. Dave -Original Message- From: boun...@canonical.com [mailto:boun...@canonical.com] On Behalf Of bonsiware Sent: Wednesday, October 07, 2009 3:26 AM To: Graham, David Subject: [Bug 263555] Re: [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 gigE chipsets at risk Burned the iso, followed your instructions, but: eth0 EEprom len 4096 checksum ed0e wrong should be 830e So I can't send you my ethtool.dmp My notebook is a Fujitsu Siemens Lifebook E8410 Thank you Dave!!! -- [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 gigE chipsets at risk https://bugs.launchpad.net/bugs/263555 You received this bug notification because you are a direct subscriber of the bug. Status in The Linux Kernel: Fix Released Status in “linux” package in Ubuntu: Fix Released Status in “linux-lpia” package in Ubuntu: Fix Released Status in linux in Ubuntu Intrepid: Fix Released Status in linux-lpia in Ubuntu Intrepid: Fix Released Status in “linux” package in Fedora: Fix Released Status in “linux” package in Gentoo Linux: Fix Released Status in “linux” package in Mandriva: Fix Released Status in “linux” package in Suse: Fix Released Bug description: In some circumstances it appears possible for the 2.6.27-rc kernels to corrupt the NVRAM used by some Intel network parts to store data such as MAC addresses. This is limited to the new e1000e driver, and reports have only appeared from users of "82566 and 82567 based LAN parts (ich8 and ich9)" (to quote Intel). The reports seem to be isolated to laptops, but it is not clear if this is because desktop/server parts are not vulnerable, or if use cases simply increase the chances of laptop users being hit. Once this corruption has occurred, recovery may be possible via a BIOS update, but may well require replacement of the hardware. Use of Intel's IABUTIL.EXE is strongly discouraged, as it will worsen the problem to the point where the network part will no longer appear on the PCI bus. (this is a new description, the original one was based on too much guesswork. Below are the URLs originally referenced) (the driver i blacklisted in Ubuntu for 2.6.27-rc in the latest releases, so if your network is not working, it doesn't have to be damaged, but just disabled in order to prevent any accidents until this bug is solved, don't wary!) http://www.blahonga.org/~art/rant.html (search for "em0") http://www.mail-archive.com/e1000-de...@lists.sourceforge.net/msg00360.html http://www.mail-archive.com/e1000-de...@lists.sourceforge.net/msg00398.html -- [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 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
[Bug 263555] Re: [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 gigE chipsets at risk
I am still contacted about once per month for instructions on how to recover ethernet functionality on systems that have had their 1Gb flash content corrupted, possibly by this defect. If you believe that you are affected by this isssue, you can safely perform steps 1 through 5 from the bullet list below, then contact me with the result, from which I will prepare a fully repaired image, and post it back to you. You can then continue with steps 6,7 & 8. 1) Download a CD image of the recovery program (originally created by Karsten Keil formerly of SuSE) from http://e1000.sourceforge.net/e1000e_recover.iso. Please type the address in your browser window and choose "save to fle"- you cannot search for this file. 2) Burn the iso to CD, & boot the CD. When prompted, select “Rescue System” Linux will load, you’ll see an openSUSE splashscreen, and eventually a login prompt. 3) Log on as root. There's no password, so just hit return. 4) Read the current eeprom and save it to file. Be patient ! e1000e_nvm -r –d eth0 -o ethtool.dmp 5) mount a USB disk to save the file, and send the file to me david_dot_graham_at_intel_dot_com I will then fix up the image, and mail it back to you as ethtoola.dmp. When you receive the updated file: 6) Write the new eeprom content back to your system NVM e1000e_nvm –d eth0 -P 108C8086 ethtoola.dmp 7) You will see some warnings, select YES when prompted. 8) You should then be able to remove the recovery CD, and successfully boot back to a working ethernet using Linux, Windows, OpenSolaris, or anything else. -- [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 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
[Bug 323256] Re: 2.6.27-11 Intel Ethernet e100e Remains active after shutdown
Stefan, Beige. SOrry about the delay. I still have no repro here, though I have taken this up with our customer issue repro team to keep trying. In the meantime, if you do decide that you want to restore the EEPROM (to re-enable he WOL capability), you can do this: ethtool -E eth0 magic 0x109a8086 offset 0x49 value 0x4D -- 2.6.27-11 Intel Ethernet e100e Remains active after shutdown https://bugs.launchpad.net/bugs/323256 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
[Bug 323256] Re: 2.6.27-11 Intel Ethernet e100e Remains active after shutdown
I agree that the temproary fix shohuld be safe. It was a good idea. I've been trying to debug the issue myself, and though I have not been sucessful, can still share with you what I have done - it may be of some use. I don't have a Toshiba Satellite Pro but have tried to repro on a Lenovo T60P, which has the same INTEL 82573L. On the T60P, I don't see the problem you report, so will continue hunting for a Toshiba. Here is what I see, using the same eeprom settings as you have, on my T60P. 1) shutdown -h now 2) As long as there is any active WOL setting (there is in your eeprom image, and that's the setting that you disabled using the ethtool command), the link "bounces" down and then up as the driver unloads. At the connected equipment, I can see that the link is back to 1Gb for a few seconds, and the system continues its shutdown. If there is no WOL setting, the link does not come back, and this is becuase the driver requests that the NIC phy powers down. 3) If there is still power to the NIC (ie there was some type of WOL active as indicated in ethtool eth0), AND the BIOS is configured for wol enabled, AND the system is either docked or powered by power cord, the link is then reestablished at 100Mbps while the system is otherwise powered down. If the NIC was already powered down, OR the system is on battery power, OR the BIOS was configured with WOL disabled, there is no link. I looked at the e1000e differences between the 2.6.27.9 and 2.6.27.11 builds, and there's very little changed: nothing that I thin could account for your problem. So its likely something in the kernel changes, interracting with the Toshiba BIOS. Once I get a Toshiba, I should be able to go a llittle further. Could you provide the BIOS version number on your Toshiba, it might help me align my testing with yours when I get hold of one. Its interesting that you have shown that you can temporarily resolve the issue using "ethtool -s eth0 wol d", and it makes sense too, as this allows the driver to request that the NIC powers down before the main kernel shuts-down. Because the initial WOL settings are loaded from the EEPROM, we could also disable them in the EEPROM instead, so you wouldn't need to use a script to run "ethtool -s eth0 wol d", and it should be a workable solution until we figure out the kernel/BIOS interaction. If you do this, of course there will be no Wake On Lan capability unless you later restore the EEPROM settings, and we don't want to keep changing the EEPROM settings to dynamically control the feature. If you do want to do this, you can disable WOL in the eeprom by: ethtool -E eth0 magic 0x109a8086 offset 0x49 value 0x49 -- 2.6.27-11 Intel Ethernet e100e Remains active after shutdown https://bugs.launchpad.net/bugs/323256 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
[Bug 263555] Re: [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 gigE chipsets at risk
While we should expect no further reports of flash corruptions due to this bug, I would like to know of any systems which did fall foul of the bug, and have not yet had their flash restored. Pleae let me know if you have system that had proper (e1000e) LAN functionality proor to installing a 2.6.27-rc kernel, and lost it while running the rc kernel. So as not to confuse this bug report, please contact me offline and I will try to help you restore your LAN. david.graham_at_intel_dot_com -- [intrepid] 2.6.27 e1000e driver places Intel ICH8 and ICH9 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