Bug#543398: linux-image-2.6.30-1-686: e100 is missing a dependency on linux-firmware
Package: linux-image-2.6.30-1-686 Version: 2.6.30-5 Severity: important In the 2.6.30 kernel, but not in 2.6.26, e100 depends on d101m_ucode.bin, from firmware-linux. The kernel package is not marked as depending on this. Consequently, if you reboot into your newly installed 2.6.30 kernel, your networking, if using an E10/100 card for instance, is irretrievably broken. -- Package-specific info: ** Version: Linux version 2.6.30-1-686 (Debian 2.6.30-5) (m...@debian.org) (gcc version 4.3.3 (Debian 4.3.3-15) ) #1 SMP Mon Aug 3 16:18:30 UTC 2009 ** Command line: root=/dev/hda1 ro ** Not tainted ** Kernel log: [3.439776] hdb: unknown partition table [3.463618] hdc: max request size: 128KiB [3.486058] hdc: 150136560 sectors (76869 MB) w/1916KiB Cache, CHS=65535/16/63 [3.486190] hdc: cache flushes not supported [3.486318] hdc: hdc1 [3.503498] ide-cd: hdd: ATAPI 52X CD-ROM drive, 128kB Cache [3.503686] Uniform CD-ROM driver Revision: 3.20 [ 16.480023] scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 7.0 [ 16.480026] Adaptec 2902/04/10/15/20C/30C SCSI adapter [ 16.480028] aic7850: Single Channel A, SCSI Id=7, 3/253 SCBs [ 16.480030] [ 16.481500] ohci_hcd :02:01.0: PCI INT A - GSI 19 (level, low) - IRQ 19 [ 16.481587] ohci_hcd :02:01.0: OHCI Host Controller [ 16.481709] ohci_hcd :02:01.0: new USB bus registered, assigned bus number 1 [ 16.481800] ohci_hcd :02:01.0: irq 19, io mem 0xfeaf5000 [ 16.536048] usb usb1: New USB device found, idVendor=1d6b, idProduct=0001 [ 16.536106] usb usb1: New USB device strings: Mfr=3, Product=2, SerialNumber=1 [ 16.536165] usb usb1: Product: OHCI Host Controller [ 16.536213] usb usb1: Manufacturer: Linux 2.6.30-1-686 ohci_hcd [ 16.536262] usb usb1: SerialNumber: :02:01.0 [ 16.536425] usb usb1: configuration #1 chosen from 1 choice [ 16.536518] hub 1-0:1.0: USB hub found [ 16.536577] hub 1-0:1.0: 2 ports detected [ 16.536800] sata_via :02:0c.0: version 2.4 [ 16.536826] sata_via :02:0c.0: PCI INT A - GSI 17 (level, low) - IRQ 17 [ 16.536944] sata_via :02:0c.0: routed to hard irq line 10 [ 16.538599] scsi1 : sata_via [ 16.538781] scsi2 : sata_via [ 16.539155] scsi3 : sata_via [ 16.539266] ata1: SATA max UDMA/133 port i...@0xdfa0 bmdma 0xdf40 irq 17 [ 16.539319] ata2: SATA max UDMA/133 port i...@0xdf90 bmdma 0xdf48 irq 17 [ 16.539369] ata3: PATA max UDMA/133 port i...@0xdf80 bmdma 0xdf50 irq 17 [ 16.539489] ohci_hcd :02:01.1: PCI INT B - GSI 18 (level, low) - IRQ 18 [ 16.539575] ohci_hcd :02:01.1: OHCI Host Controller [ 16.539666] ohci_hcd :02:01.1: new USB bus registered, assigned bus number 2 [ 16.539757] ohci_hcd :02:01.1: irq 18, io mem 0xfeaf6000 [ 16.592045] usb usb2: New USB device found, idVendor=1d6b, idProduct=0001 [ 16.592102] usb usb2: New USB device strings: Mfr=3, Product=2, SerialNumber=1 [ 16.592160] usb usb2: Product: OHCI Host Controller [ 16.592208] usb usb2: Manufacturer: Linux 2.6.30-1-686 ohci_hcd [ 16.592258] usb usb2: SerialNumber: :02:01.1 [ 16.592424] usb usb2: configuration #1 chosen from 1 choice [ 16.592518] hub 2-0:1.0: USB hub found [ 16.592575] hub 2-0:1.0: 2 ports detected [ 16.870893] ata1: SATA link down (SStatus 0 SControl 310) [ 17.19] ata2: SATA link down (SStatus 0 SControl 310) [ 17.870387] PM: Starting manual resume from disk [ 17.983514] kjournald starting. Commit interval 5 seconds [ 17.983578] EXT3-fs: mounted filesystem with ordered data mode. [ 19.881515] udev: starting version 141 [ 20.118188] input: Power Button as /devices/LNXSYSTM:00/LNXPWRBN:00/input/input1 [ 20.118258] ACPI: Power Button [PWRF] [ 20.118398] input: Power Button as /devices/LNXSYSTM:00/device:00/PNP0C0C:00/input/input2 [ 20.118463] ACPI: Power Button [PBTN] [ 20.124898] processor ACPI_CPU:00: registered as cooling_device0 [ 20.279539] input: PC Speaker as /devices/platform/pcspkr/input/input3 [ 20.303906] parport_pc 00:0a: reported by Plug and Play ACPI [ 20.304052] parport0: PC-style at 0x378 (0x778), irq 7 [PCSPP,TRISTATE] [ 20.357868] pci_hotplug: PCI Hot Plug PCI Core version: 0.5 [ 20.386242] shpchp: Standard Hot Plug PCI Controller Driver version: 0.4 [ 20.388427] intel_rng: Firmware space is locked read-only. If you can't or [ 20.388430] intel_rng: don't want to disable this in firmware setup, and if [ 20.388432] intel_rng: you are certain that your system has a functional [ 20.388434] intel_rng: RNG, try using the 'no_fwh_detect' option. [ 20.394626] i801_smbus :00:1f.3: PCI INT B - GSI 17 (level, low) - IRQ 17 [ 20.682646] input: ImExPS/2 Generic Explorer Mouse as /devices/platform/i8042/serio1/input/input4 [ 22.887083] Adding 2650684k swap on /dev/hda5. Priority:-1 extents:1 across:2650684k [ 23.344244] EXT3 FS on hda1, internal journal [ 23.942484] loop: module loaded [ 24.423183] kjournald starting.
Bug#543398: linux-image-2.6.30-1-686: e100 is missing a dependency on linux-firmware
On Mon, Aug 24, 2009 at 07:24:28PM +0100, Jon Thackray wrote: Package: linux-image-2.6.30-1-686 Version: 2.6.30-5 Severity: important In the 2.6.30 kernel, but not in 2.6.26, e100 depends on d101m_ucode.bin, from firmware-linux. The kernel package is not marked as depending on this. Packages in main cannot depend upon packages in non-free. Further, this isn't a strict dependency - only certain devices require firmware files. I won't argue that the current mechanism is ideal - there is a warning on upgrade (from initramfs-tools), but it doesn't tell you precisely what to do about it. Consequently, if you reboot into your newly installed 2.6.30 kernel, your networking, if using an E10/100 card for instance, is irretrievably broken. Irretrievably? Can't you reboot your old kernel and install firmware-linux? -- dann frazier -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#543398: linux-image-2.6.30-1-686: e100 is missing a dependency on linux-firmware
Well, it seems a shame to have gone from a situation where the system worked to one where it doesn't. If you must do this, you should at least force the user to interact with the upgrade process (ie it won't proceed until the user makes a decision (cf the pam package for example), either to ignore the problem because it isn't going to affect him, or to install the missing package, or indeed to decide he can't cope with the moral dilemma of having a 5 year old network card that has suddenly been found to depend on non-free software, in which case he can take the card out and replace it with something else). Can one at least hope that an unencumbered driver will be produced in the future? Sorry if this all sounds overly pragmatic and a bit irritated, but for the naive user at least, this sort of thing can only serve to push him back towards the evil empire where such problems don't exist. Is that what the open source movement wants? As for irretrievably, I merely meant that your system, as is, was broken in a way that couldn' be repaired (unlike for example, a missing sound card driver where you could go and download it). Of course one can reboot into an earlier kernel, that's how I fixed the problem (after several rounds of diagnosis). - Original Message - From: dann frazier da...@debian.org To: Jon Thackray j...@pobox.com; 543...@bugs.debian.org Sent: Monday, August 24, 2009 8:27 PM Subject: Re: Bug#543398: linux-image-2.6.30-1-686: e100 is missing a dependency on linux-firmware On Mon, Aug 24, 2009 at 07:24:28PM +0100, Jon Thackray wrote: Package: linux-image-2.6.30-1-686 Version: 2.6.30-5 Severity: important In the 2.6.30 kernel, but not in 2.6.26, e100 depends on d101m_ucode.bin, from firmware-linux. The kernel package is not marked as depending on this. Packages in main cannot depend upon packages in non-free. Further, this isn't a strict dependency - only certain devices require firmware files. I won't argue that the current mechanism is ideal - there is a warning on upgrade (from initramfs-tools), but it doesn't tell you precisely what to do about it. Consequently, if you reboot into your newly installed 2.6.30 kernel, your networking, if using an E10/100 card for instance, is irretrievably broken. Irretrievably? Can't you reboot your old kernel and install firmware-linux? -- dann frazier -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#543398: linux-image-2.6.30-1-686: e100 is missing a dependency on linux-firmware
On Mon, 2009-08-24 at 21:02 +0100, Jon Thackray wrote: Well, it seems a shame to have gone from a situation where the system worked to one where it doesn't. If you must do this, you should at least force the user to interact with the upgrade process (ie it won't proceed until the user makes a decision (cf the pam package for example), either to ignore the problem because it isn't going to affect him, or to install the missing package, or indeed to decide he can't cope with the moral dilemma of having a 5 year old network card that has suddenly been found to depend on non-free software, in which case he can take the card out and replace it with something else). Starting from 2.6.31-1, there will be a high-priority notice for missing firmware whenever you upgrade to a newer upstream version. Unless you explicitly request a non-interactive upgrade, this *will* appear. The notice appears after the package's files have been installed, at which point installation cannot be cancelled. However, since each new upstream kernel version has new package names, it does not replace the older version, which will still be installed at this point. Can one at least hope that an unencumbered driver will be produced in the future? [...] This seems unlikely. Ben. -- Ben Hutchings If at first you don't succeed, you're doing about average. signature.asc Description: This is a digitally signed message part
