Ah, danke, sehr schöner Tip. In den syslog finde ich nun endlich, was zwischen den Boot-Vorgängen unterschiedlich war:
[ 0.000000] Fast TSC calibration failed
[ 0.000000] TSC: PIT calibration matches PMTIMER. 1 loops
findet sich in dem, wo udev gar nicht geladen wird. Im anderen Fall hießt das:
[ 0.000000] Fast TSC calibration using PIT
Keine Ahnung, was das ist.
Hm, dann werden einige Dinge manchmal in umgekehrter Reihenfolge aufgerufen, aber es scheint eine Weile alles da zu sein. Nur dann wird eben in dem einen Fall udev nicht aufgerufen, was dann so aussieht:
[ 2.073270] sr0: scsi3-mmc drive: 24x/24x writer dvd-ram cd/rw xa/form2 cdda tray
[ 2.073273] Uniform CD-ROM driver Revision: 3.20
[ 2.073358] sr 1:0:0:0: Attached scsi CD-ROM sr0
[ 2.084273] sda1 sda2 sda3 < sda5 sda6 sda7 sda8 sda9 sda10 > sda4
[ 2.171779] sd 0:0:0:0: [sda] Attached SCSI disk
[ 2.176967] sd 0:0:0:0: Attached scsi generic sg0 type 0
[ 2.177010] sr 1:0:0:0: Attached scsi generic sg1 type 5
[ 2.875790] PM: Starting manual resume from disk
[ 2.938751] kjournald starting. Commit interval 5 seconds
[ 2.938763] EXT3-fs: mounted filesystem with ordered data mode.
[ 5.445771] loop: AES key scrubbing enabled
[ 5.446129] loop: loaded (max 8 devices)
[ 5.668229] Adding 3413768k swap on /dev/loop1. Priority:-1 extents:1 across:3413768k
[ 5.904859] EXT3 FS on sda10, internal journal
[ 6.276119] device-mapper: uevent: version 1.0.3
[ 6.276205] device-mapper: ioctl: 4.14.0-ioctl (2008-04-23) initialised:
dm-devel@redhat.com
[ 7.659835] fuse init (API version 7.11)
[ 7.741404] kjournald starting. Commit interval 5 seconds
[ 7.746697] EXT3 FS on sda2, internal journal
[ 7.746702] EXT3-fs: mounted filesystem with ordered data mode.
Wenn aber udev mit aufgerufen wird, kommt folgende Ausgabe bei syslog (ich glaube die oben schon beschriebene Fehlermeldung fehlt hier):
[ 3.343767] kjournald starting. Commit interval 5 seconds
[ 3.343780] EXT3-fs: recovery complete.
[ 3.344343] EXT3-fs: mounted filesystem with ordered data mode.
[ 5.054129] udev: starting version 146
[ 5.164258] input: Power Button as /devices/LNXSYSTM:00/LNXPWRBN:00/input/input2
[ 5.164266] ACPI: Power Button [PWRF]
[ 5.164338] input: Lid Switch as /devices/LNXSYSTM:00/device:00/PNP0C0D:00/input/input3
[ 5.164423] ACPI: Lid Switch [LID]
[ 5.164481] input: Power Button as /devices/LNXSYSTM:00/device:00/PNP0C0C:00/input/input4
[ 5.164485] ACPI: Power Button [PWRB]
[ 5.333453] ACPI: SSDT 7fe7e9a0 001A8 (v01 PmRef Cpu0Ist 00003000 INTL 20050624)
[ 5.333770] ACPI: SSDT 7fe7e73f 001DC (v01 PmRef Cpu0Cst 00003001 INTL 20050624)
[ 5.334110] Marking TSC unstable due to TSC halts in idle
[ 5.334158] ACPI: CPU0 (power states: C1[C1] C2[C2])
[ 5.334186] processor ACPI_CPU:00: registered as cooling_device1
[ 5.334191] ACPI: Processor [CPU0] (supports 8 throttling states)
[ 5.334532] ACPI: SSDT 7fe7eb48 00089 (v01 PmRef Cpu1Ist 00003000 INTL 20050624)
[ 5.334819] ACPI: SSDT 7fe7e91b 00085 (v01 PmRef Cpu1Cst 00003000 INTL 20050624)
[ 5.335267] ACPI: CPU1 (power states: C1[C1] C2[C2])
[ 5.335291] processor ACPI_CPU:01: registered as cooling_device2
[ 5.335296] ACPI: Processor [CPU1] (supports 8 throttling states)
[ 5.336587] ACPI: EC: non-query interrupt received, switching to interrupt mode
[ 5.353377] ACPI: AC Adapter [ADP0] (on-line)
[ 5.362796] ACPI: Battery Slot [BAT0] (battery present)
[ 5.648096] cfg80211: Using static regulatory domain info
[ 5.648101] cfg80211: Regulatory domain: US
[ 5.648104] ^I(start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
[ 5.648110] ^I(2402000 KHz - 2472000 KHz @ 40000 KHz), (600 mBi, 2700 mBm)
[ 5.648116] ^I(5170000 KHz - 5190000 KHz @ 40000 KHz), (600 mBi, 2300 mBm)
[ 5.648121] ^I(5190000 KHz - 5210000 KHz @ 40000 KHz), (600 mBi, 2300 mBm)
[ 5.648126] ^I(5210000 KHz - 5230000 KHz @ 40000 KHz), (600 mBi, 2300 mBm)
[ 5.648130] ^I(5230000 KHz - 5330000 KHz @ 40000 KHz), (600 mBi, 2300 mBm)
[ 5.648135] ^I(5735000 KHz - 5835000 KHz @ 40000 KHz), (600 mBi, 3000 mBm)
[ 5.648147] cfg80211: Calling CRDA for country: US
[ 5.712762] input: PC Speaker as /devices/platform/pcspkr/input/input5
[ 5.945256] iwl3945: Intel(R) PRO/Wireless 3945ABG/BG Network Connection driver for Linux, 1.2.26ks
[ 5.945259] iwl3945: Copyright(c) 2003-2009 Intel Corporation
[ 5.945370] iwl3945 0000:05:00.0: PCI INT A -> GSI 18 (level, low) -> IRQ 18
[ 5.945384] iwl3945 0000:05:00.0: setting latency timer to 64
[ 5.999276] iwl3945 0000:05:00.0: Tunable channels: 13 802.11bg, 23 802.11a channels
[ 5.999279] iwl3945 0000:05:00.0: Detected Intel Wireless WiFi Link 3945ABG
[ 5.999451] iwl3945 0000:05:00.0: irq 28 for MSI/MSI-X
[ 6.112791] yenta_cardbus 0000:07:06.0: CardBus bridge found [1179:ff10]
[ 6.112813] yenta_cardbus 0000:07:06.0: Enabling burst memory read transactions
[ 6.112819] yenta_cardbus 0000:07:06.0: Using CSCINT to route CSC interrupts to PCI
[ 6.112821] yenta_cardbus 0000:07:06.0: Routing CardBus interrupts to PCI
[ 6.112829] yenta_cardbus 0000:07:06.0: TI: mfunc 0x01a01b22, devctl 0x66
[ 6.123825] i801_smbus 0000:00:1f.3: PCI INT B -> GSI 19 (level, low) -> IRQ 19
[ 6.125175] intel_rng: FWH not detected
[ 6.157069] Synaptics Touchpad, model: 1, fw: 6.2, id: 0x25a0b1, caps: 0xa04713/0x0
[ 6.157079] synaptics: Toshiba Satellite A100 detected, limiting rate to 40pps.
[ 6.192097] input: SynPS/2 Synaptics TouchPad as /devices/platform/i8042/serio4/input/input6
[ 6.345114] yenta_cardbus 0000:07:06.0: ISA IRQ mask 0x0cf8, PCI irq 18
[ 6.345119] yenta_cardbus 0000:07:06.0: Socket status: 30000006
[ 6.345124] pci_bus 0000:07: Raising subordinate bus# of parent bus (#07) from #07 to #0b
[ 6.345133] yenta_cardbus 0000:07:06.0: pcmcia: parent PCI bridge I/O window: 0x5000 - 0x5fff
[ 6.345136] pcmcia_socket pcmcia_socket0: cs: IO port probe 0x5000-0x5fff: clean.
[ 6.345380] yenta_cardbus 0000:07:06.0: pcmcia: parent PCI bridge Memory window: 0xde000000 - 0xde0fffff
[ 6.345383] yenta_cardbus 0000:07:06.0: pcmcia: parent PCI bridge Memory window: 0x88000000 - 0x8bffffff
[ 6.428507] tifm_7xx1 0000:07:06.2: PCI INT A -> GSI 18 (level, low) -> IRQ 18
[ 6.961916] pcmcia_socket pcmcia_socket0: cs: IO port probe 0x100-0x3af: clean.
[ 6.963916] pcmcia_socket pcmcia_socket0: cs: IO port probe 0x3e0-0x4ff: excluding 0x4d0-0x4d7
[ 6.964802] pcmcia_socket pcmcia_socket0: cs: IO port probe 0x820-0x8ff: clean.
[ 6.967264] pcmcia_socket pcmcia_socket0: cs: IO port probe 0xc00-0xcf7: clean.
[ 6.968130] pcmcia_socket pcmcia_socket0: cs: IO port probe 0xa00-0xaff: clean.
[ 7.076618] HDA Intel 0000:00:1b.0: PCI INT A -> GSI 22 (level, low) -> IRQ 22
[ 7.076657] HDA Intel 0000:00:1b.0: setting latency timer to 64
[ 7.159640] phy0: Selected rate control algorithm 'iwl-3945-rs'
[ 7.186142] udev: renamed network interface wlan0 to eth3
[ 7.187152] input: HDA Digital PCBeep as /devices/pci0000:00/0000:00:1b.0/input/input7
[ 9.161072] iwl3945 0000:05:00.0: Radio Frequency Kill Switch is On:
[ 9.161075] Kill switch must be turned off for wireless networking to work.
[ 14.239476] EXT3 FS on sda10, internal journal
[ 14.455050] loop: AES key scrubbing enabled
[ 14.455408] loop: loaded (max 8 devices)
Die ersten und letzten Zeilen sind identisch. Ich hab sie drin gelassen, um deutlich zu machen, was wohl zu udev gehört. Die Ausgabe von udev, die sich auch beim Einzelaufruf wiederfindet, hab ich dazu noch fett gemacht.
Ja, und nun? Daraus kann ich zumindest noch nicht erkennen, warum udev beim einen Mal startet, beim anderen aber nicht. Und es leuchtet mir auch noch nicht ein, warum ich anschließend udev über 20 Mal im System laufen habe, wobei der Prozess dann wirklich 25% Leistung beansprucht.
Zum System: Die source-list sieht so aus:
Lebe so, dass es noch immer o.k. wäre, wenn alle Menschen so leben würden