Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep

[Mauro D'Aloisio]

When I turn on my computer I can read this on my screen

"modprobe: Fatal: Could not load /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"

After 4-5 seconds, plymouth starts and the system works fine.

salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi rimane con lo schermo nero..ed è impossibile lavorare...! grazie

hi, I have the same problem, I asked the psw keychain and then left with a black screen .. and it is impossible to work ...! thanks

[Omer Akram (om26er)]

this started with linux-image-2.6.35-21-generic and persists with 2.6.35-22

[Bernhard (b.a.koenig)]

I think it's a simple mix-up because /lib/modules/2.6.35-22-generic/ exists, but /lib/modules/2.6.35-22/ doesn't exist.

[DickMc (rhm47)]

I have the same problem and it is not a mix up as my message reads "modprobe: Fatal: Could not load /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory". Note the message includes "generic" which is not included in the original bug report (I suspect due to an oversight versus the message being different.)

[Mauro D'Aloisio (maurodaloisio)]

I apologize; I have the same message of DickMc. I update the bug description

[Bernhard (b.a.koenig)]

As for the mix-up: I get the following in the terminal

% update-initramfs -c -k 2.6.35 ~
update-initramfs: Generating /boot/initrd.img-2.6.35
touch: cannot touch `/boot/initrd.img-2.6.35.new': Permission denied
grep: /boot/config-2.6.35: No such file or directory
WARNING: missing /lib/modules/2.6.35
Device driver support needs thus be built-in linux image!
WARNING: Couldn't open directory /lib/modules/2.6.35: No such file or directory
FATAL: Could not open /lib/modules/2.6.35/modules.dep.temp for writing: No such file or directory
FATAL: Could not load /lib/modules/2.6.35/modules.dep: No such file or directory

[Bernhard (b.a.koenig)]

and maybe I should add

% update-initramfs -c -k 2.6.35-generic
update-initramfs: Generating /boot/initrd.img-2.6.35-generic
touch: cannot touch `/boot/initrd.img-2.6.35-generic.new': Permission denied
grep: /boot/config-2.6.35-generic: No such file or directory
WARNING: missing /lib/modules/2.6.35-generic
Device driver support needs thus be built-in linux image!
WARNING: Couldn't open directory /lib/modules/2.6.35-generic: No such file or directory
FATAL: Could not open /lib/modules/2.6.35-generic/modules.dep.temp for writing: No such file or directory
FATAL: Could not load /lib/modules/2.6.35-generic/modules.dep: No such file or directory

[ramadasan7 (ramadasan7-gmail)]

I have the same problem. But the OS works.

[Andrew Betts (andrew-betts)]

I felt this was a timing issue, so I set my cpu to a slower speed in the BIOS and this error message disappeared. (I still have other problems though, like bug #595321. I don't know if they are related.)

[0N3_xyz (tophitter225)]

I have the same problem as above although its not effecting me in anyway if anything just slowing down the boot time

[eric (obrowny06)]

Same thing : modprobe: FATAL: Could not load /lib/modules/2.6.35-22-generic/modules.dep No such file or directory.
This happened after the update from lucid to maverick.
Everything seems to work except a very slow boot up and the panel disappears sometimes while loading a program with applications or system... No idea if it is related to this bug.

[Louis (louisgag)]

Same issue here.

[Andrew Betts (andrew-betts)]

This is fixed for me in kernel mainline v2.6.36-rc6-maverick

(https://wiki.ubuntu.com/Kernel/MainlineBuilds?action=show)

[Gufo (thrasherxxx)]

Same exact problem here.
i think we should wait the 2.6.36?

[durlo (rsanver)]

I have upgraded to 10.10-RC and I have the same problem. Running dmesg, I have noticed that there are several lines like this:

[ 17.588275] EXT4-fs (sda1): re-mounted. Opts: errors=remount-ro,commit=0

Can it be possible that modprobe is trying to read modules.dep while the FS is being re-mounted?

[durlo (rsanver)]

Sorry, my previous message was wrong; the error messages appear before the filesystem is mounted, so this lines aren't related to this bug.

[larochea103 (larochea103)]

j'ai le meme probleme

[Virgil Brummond (uraharakisuke153)]

If I boot with nosmp, the problem seems to not happen. Back next normal boot.

[durlo (rsanver)]

For me, booting with smp doesn't solve the problem.

[max63 (maxime-chassagneux)]

Same probleme on my netbook (KUBUNTU) with the 2.6.35-generic-pae kernel

[Julien Aubin (gojulgarbmail)]

Same problem for me with kernel 2.6.35-22-generic.

Hardware :
Mobo : Asus M2N-SLI Deluxe
CPU : AMD Phenom 9850
4 GB RAM
Graphic card : NVidia

[Blaster (holst-niels)]

Same problem running 2.6.35-22-generic. Started with 2.6.35-20 I think.

[manzur (sl-solaris)]

same problem in here

[Ron Chapman (ronchap)]

Just installed 10.10 with a bunch of updates after install. The updates did not fix this. Same problem here.
Hoping we have a clean boot on 10.10.10 release.

[Omer Akram (om26er)]

although the system boots just fine. but this bug ruins the boot experience for many. can anyone from the kernel team please have a say on this.

[Panos (pgalatis)]

Same here folks. Updated from Maverick beta to RC (2.6.35-22-generic kernel). Running Ubuntu Maverick on a Lenovo X200s laptop.

[Lucazade (lucazade)]

Please stop with "Same here"
Doesn't add anything to the bug report, only send a lot of email notifications!

[Bernhard (b.a.koenig)]

Yes, just click the "affects me" button at the top of the page.

[Yann (ylevot)]

Same issue for me with Maverick RC on ASUS A6J laptop.

[manzur (sl-solaris)]

I hate when ubuntu goes 'Final Release' with this kind of errors

[durlo (rsanver)]

I have found a workaround that fixes the issue for me. I've edited the file /etc/initramfs-tools/initramfs.conf and I've changed the line MODULES=most to MODULES=dep. Then, I've reinstalled initramfs-tools and the error messages have disappeared.

[Blaster (holst-niels)]

@ #31: Thanks durlo. Worked for me too.

[Yann (ylevot)]

Thanks. Actually error messages have disappeared for me as well with your workaround, but boot time is still quite long and I've lost my mouse pointer with gdm :(
I've got this trace in dmesg:

[ 51.864127] Modules linked in: rfcomm binfmt_misc sco bnep l2cap parport_pc ppdev joydev snd_hda_codec_si3054 snd_hda_codec_realtek arc4 snd_hda_intel radeon snd_hda_codec snd_hwdep snd_pcm iwl3945 snd_seq_midi snd_rawmidi ttm snd_seq_midi_event iwlcore snd_seq pcmcia drm_kms_helper mac80211 snd_timer gspca_m5602 btusb snd_seq_device gspca_main firewire_ohci drm bluetooth videodev v4l1_compat r8169 sdhci_pci yenta_socket intel_agp firewire_core sdhci mii snd psmouse pcmcia_rsrc crc_itu_t serio_raw cfg80211 pcmcia_core i2c_algo_bit soundcore video snd_page_alloc agpgart output asus_laptop sparse_keymap led_class lp parport
[ 51.864238] Pid: 1028, comm: Xorg Not tainted 2.6.35-22-generic #33-Ubuntu
[ 51.864243] Call Trace:
[ 51.864258] [<c014ac52>] warn_slowpath_common+0x72/0xa0
[ 51.864292] [<f8b8c2de>] ? radeon_fence_wait+0x23e/0x290 [radeon]
[ 51.864326] [<f8b8c2de>] ? radeon_fence_wait+0x23e/0x290 [radeon]
[ 51.864335] [<c014ad23>] warn_slowpath_fmt+0x33/0x40
[ 51.864369] [<f8b8c2de>] radeon_fence_wait+0x23e/0x290 [radeon]
[ 51.864380] [<c0165e10>] ? autoremove_wake_function+0x0/0x50
[ 51.864414] [<f8b8ca01>] radeon_sync_obj_wait+0x11/0x20 [radeon]
[ 51.864430] [<f8a39e67>] ttm_bo_wait+0xe7/0x180 [ttm]
[ 51.864469] [<f8ba2049>] radeon_gem_wait_idle_ioctl+0x89/0x110 [radeon]
[ 51.864505] [<f86ac98d>] drm_ioctl+0x1ad/0x430 [drm]
[ 51.864530] [<f8ba1fc0>] ? radeon_gem_wait_idle_ioctl+0x0/0x110 [radeon]
[ 51.864538] [<c010afb2>] ? restore_i387_xstate+0xe2/0x210
[ 51.864544] [<c0226622>] vfs_ioctl+0x32/0xb0
[ 51.864557] [<f86ac7e0>] ? drm_ioctl+0x0/0x430 [drm]
[ 51.864562] [<c0226eb9>] do_vfs_ioctl+0x79/0x2d0
[ 51.864567] [<c01021d6>] ? restore_sigcontext+0xc6/0xe0
[ 51.864571] [<c0227177>] sys_ioctl+0x67/0x80
[ 51.864577] [<c05c90a4>] syscall_call+0x7/0xb
[ 51.864580] ---[ end trace cb6f809ca6822860 ]---
[ 52.021556] Failed to wait GUI idle while programming pipes. Bad things might happen.

[Petras I. Florin (petras-florin)]

Same here, running Ubuntu 10.10 Final.

[Hadrien Titeux (hadware)]

Same, running the final.
It did it on an updated-to-10.10 Thinkpad X200

[Martitza (martitzam)]

I also experience this exact bug on every boot since I updated from lucid to maverick.

Intel Core2 Duo E6850 @ 3GHz
2GiB RAM
booting from /dev/sdb1 per grub
x64 smp kernel 2.6.35-22-generic
nvidia graphics

Also maybe related is I get a really awful looking boot screen...even worse than it looked in lucid

[Blaster (holst-niels)]

Try to edit /etc/initramfs-tools/initramfs.conf (sudo gedit /etc/initramfs-tools/initramfs.conf in the console) and change the line MODULES=most to MODULES=dep.

Then use Synaptic to reinstall initramfs-tools like "durlo" suggested.

It's not perfect, but it works.

[Marc-solutions2u (marc-solutions2u)]

Same problem here..
Went from 10.04 to BETa.... no problems...then when i upgraded on 10/10 to Maverick, the same error above.

Thanks Blaster, your fix is working but boot up experience is long..*-(

Please work to fix it - thansk

[Roberto Zedda (zedrob)]

Fixed on latest 2.6.35-22. Thanks!

[Alexander (porsche911)]

The same problem.

I have tried to reinstall kernels using synaptic, but this message is still appears

[russ (russthered)]

I upgraded from 10.04 to 10.10 using network upgrade as recommended - no issues or alerts during procedure.

I too get 'modprobe:FATAL: could not load /lib/modules/2.6.35-22-generic/modules.dep No such file or directory' twice on the screen, boot up then hangs for at least 20seconds before booting into OS.

I'm running an ACER Aspire 5738z dual booting with Windows 7

Not keen on using durlo's 'workaround' - not confident enough :) but thanks for it anway.

What's the chance of a proper fix?

[FergusB (fergus-bremner)]

Upgraded to 10.10, now also getting modprobe:FATAL etc. (exactly as above).
System unusable.

[FergusB (fergus-bremner)]

Errors gone with durlo's edit. Able to log in and do stuff. (compiz broken)

[Omer Akram (om26er)]

the workaround should/does not break anything. I think

On Mon, Oct 11, 2010 at 10:34 PM, FergusB <email address hidden> wrote:

> Errors gone with durlo's edit. Able to log in and do stuff. (compiz
> broken)
>
> --
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
> https://bugs.launchpad.net/bugs/642421
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in “linux” package in Ubuntu: Confirmed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi rimane
> con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left with a
> black screen .. and it is impossible to work ...! thanks
>
>
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/642421/+subscribe
>

[Aydos (jesse-cresswell)]

I did a clean install of the release version of Ubuntu 10.10 64 bit and was also have this bug. Is there going to be any kind of official fix for it?

[Michael (michaeljt)]

Just a shot in the dark, but might this be a glitch while switching from the initramfs to the real root filesystem? I took a quick look at the source of modprobe, and I can't immediately see any way that the message might occur other than the file genuinely not being present.

[Jiri Kosinka (jiri-kosinka)]

same problem here after a clean install of 10.10 Final

[Vanillalite (vanillasky26)]

Ditto same problem with a 10.10 64 bit clean install of the Final release on launch day...

running ext4 as my file system if that matters to anyone!

[Michael (michaeljt)]

I investigated a bit further and my idea above is clearly wrong. If I put a printf right at the beginning of the initramfs init script that triggers after the "FATAL" message, so this is clearly a modprobe triggered by kernel code before init starts running. My next suspicion is that the kernel calls modprobe before it has finished unpacking the initramfs, so that /sbin/modprobe is already available but modules.dep is not. And when I repacked the initramfs with modules.dep as the very first file (well, after ".", "./lib", "./lib/modules" etc) the error went away.

There are a few places in the kernel where modprobe can be called from. Sticking printk statements in there should show which it is and what command line is being called. That will have to wait until I get round to it though.

[John Stewart (cane-cubo)]

I'm curious about something: how can Ubuntu ship with such a massive error right at boot? I'd like to understand the release process better. Surely this sort of thing could benefit form a post-release analysis? It's absolutely a priority-1 blocker!

[Andrew Somerville (andy16666)]

Yes, I agree. This release actually got worse and worse right up to release day. When I rolled over from RC to final the number of bugs show up significantly. Now my machine won't even standby. Even the betas were far more stable.

[Alessandro Isaia (alex69)]

same problem for me upgrading to Maverick from 10.04, after that to solve this issue I've tried to make a clean installation but the "modprobe: Fatal..." error remain anyway every time I boot.
Another new release bug is in the Gnome-panel, often disappearing.

HOW COULD CANONICAL DELIVER SUCH A BUGGY OFFICIAL RELEASE?

Alessandro

[Michael (michaeljt)]

I can't comment on the other bugs, but this error message at boot is hardly a critical bug. It is a minor visual blemish that affects a small handful (79 registered that they are affected by this bug) of the many users Ubuntu has, and unless I am mistaken it doesn't prevent anyone from getting their work done using Ubuntu. I would guess that the people at Canonical who might fix this can't reproduce it on the hardware they have available and have got more urgent things to look after. Delivering software that works perfectly for everyone on random hardware is next to impossible - even a company with the resources of Microsoft doesn't manage. It is also not new and not specific to Ubuntu, as e.g. http://kerneltrap.org/node/13918 shows. (I have downloaded the mainline kernels corresponding to the current "generic" one from the kernel ppa to see if they show this too.)

[russ (russthered)]

"but this error message at boot is hardly a critical bug."

Very true Michael, but to the less savvy amongst us anything that comes up with the words FATAL, in capital letters, is at the very least, disconcerting. :)

One of the 'bragging' points of vers. 10.04 was its short boot up time, with the upgrade to 10.10 this has increased by at least 25secs on my system.

Linux for humans don't forget :)

[John Stewart (cane-cubo)]

Michael,

I hear you, but I read your reply as reproductive of the head-in-the-sand mentality that won't examine cases such as these as opportunities for improvement. I've been a regular user of Ubuntu for many years, and I've always felt the releases are rushed, whenever a "Final" comes out, several weeks, sometimes months, go by before it reaches the stability of the previous one. That my be fun for geeks, but it's not 'Linux for Humans'. So if this keeps happening over and over again, there's a real problem to be solved.

Saying that this affects some small subset, that it's happened before, see there's a Debian thread from two years ago ... is so not the point. Here we have a case of an error that occurs across, it seems, many hardware configurations -- I'm using a stock netbook that (after a few months of updates) works more-or-less fine under 10.04, I'm not on some exotic motherboard etc. I suspect that many, many users are seeing this. And an error that hits you immediately upon the first boot with FATAL really doesn't inspire a lot of confidence, whatever the impact later. And a link to an old discussion that ends with another link to a thread on how to recompile the kernel is hardly an effective argument for why this shouldn't be a release blocker.

The broader point here is: what needs to happen to harden these releases? Maybe the 6-month thing should be rethought, or at least the non-LTS releases should be tagged as geek-oriented, unstable experiments (I wish the LTS releases would be more stable on the release date). Instead Canonical makes a big deal that this is the latest and greatest OS for the masses. There is such a gap between the rhetoric and the reality that people naturally just walk off to buy a Mac.

[Bernhard (b.a.koenig)]

John Stewart is right. I guess I could also find a number of bugs like this one that just creeped in shortly before release and then weren't fixed later. I guess there's two choices here:

1) either fix those bugs more quickly, I understand that this requires a lot of manpower, or
2) be a bit more conservative with new features and new packages. If this is a kernel problem, maybe hold back the latest kernel for a while?

[Alessandro Isaia (alex69)]

Imagine if this error was committed at Microsoft corp., there would happen a platetarian scandal in the ICT environment and all the technical guru would speak about that for years, if it does not happen to Canonical, is just because is still a common opinion that Ubuntu is not an operating system suitable for production environments, and therefore no one is really surprised.

[Michael (michaeljt)]

Just to be clear, I don't work for Canonical - I am a professional open source developer on an unrelated project. Like Ubuntu though, our software is used by many people, and as such many bugs are uncovered. We only have the manpower to fix a subset of those, so we concentrate on those we judge the most serious or the most problematic for our users. Obviously those users affected by a particular bug that we don't have time to look at often disagree with our judgement, and we often see very harsh comments on our bugtracker - from users who do not pay to use our software - criticising us and demanding our immediate attention, and sometimes even that we change the focus of our work in a direction that they consider better. Of course, we have to take care of our paying customers before we can help the non-paying ones, even though we go to great pains to help the non-paying ones too, and we also have to set our development priorities to suit the paying customers (and attract more).

And just to make the point that other companies are not immune to bugs in released software either - http://news.bbc.co.uk/2/hi/8304229.stm

[John Stewart (cane-cubo)]

Michael,

Again, I'm totally sympathetic with what you're saying, but I don't believe this case is comparable to the experience you have in your organization. Here we don't have a minor bug that some bunch of free-loaders are whinging about. In fact, as far as my own usage is concerned, I could care less whether this gets fixed -- it's not yet clear whether it affects the operation of the system, and in any case I can just stick to LTS. The point is: how can it be that such monster bugs make it into a final release? The real bug is the process, not this particular exemplar. And yet one sees little effort on the part of Canonical to alter the process to prevent this sort of thing. And thus potential users turn away. That is the frustration.

[Martitza (martitzam)]

John is on the right path here. I every reason to believe that
everyone at Canonical is doing their job with passion, sincerity and
competence. The issue is the process. A process exhibiting a pattern
of failures (unstable major releases in this case) is a strong signal
of an opportunity. And I hasten to say that I believe the majority of
people on the Ubuntu team are already focused on that because they
know that user experience matters (this is a very disconcerting bug
even if harmless) and they know that finite manpower and triage are
facts of life which do not absolve leaders of the consequences of
decisions to release.

Defects describe us.
How we react to defects defines us.

~M

P.S. I don't believe this defect is rare or obscure. I have it on
three quite varied x64 systems ranging in age from two months to two
years.

On 10/12/10, John Stewart <email address hidden> wrote:
> Michael,
>
> Again, I'm totally sympathetic with what you're saying, but I don't
> believe this case is comparable to the experience you have in your
> organization. Here we don't have a minor bug that some bunch of free-
> loaders are whinging about. In fact, as far as my own usage is
> concerned, I could care less whether this gets fixed -- it's not yet
> clear whether it affects the operation of the system, and in any case I
> can just stick to LTS. The point is: how can it be that such monster
> bugs make it into a final release? The real bug is the process, not
> this particular exemplar. And yet one sees little effort on the part of
> Canonical to alter the process to prevent this sort of thing. And thus
> potential users turn away. That is the frustration.
>
> --
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
> https://bugs.launchpad.net/bugs/642421
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in “linux” package in Ubuntu: Confirmed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load /lib/modules/2.6.35-22-generic/modules.dep:
> no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi rimane
> con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left with a
> black screen .. and it is impossible to work ...! thanks
>
>
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/642421/+subscribe
>

[René Moser (resmo)]

Same problem here.

Hardware (new):
Netbook Asus EeePC 1015PE,
Atom N450, 10",
Product number 1015PE-BLK067S

[Michael (michaeljt)]

I just tried booting linux-image-2.6.35-02063504-generic and linux-image-2.6.35-02063507-generic from the mainline kernel ppa, and they do not show these error messages (although they are no faster booting, and just sit there with a blinking cursor at the point where the Ubuntu kernel shows the errors). Since Debian systems also seem to be affected by this, I think that there is a good chance that this is the result of some Debian (not Ubuntu)-specific kernel change.

[Mark Harper (markfh)]

On booting after the normal BIOS screen then selecting "Ubuntu, with Linux 2.6.35-22-generic" in GNU GRUB I also get the message :-

"modprobe: Fatal: Could not load /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"

twice on screen and then after a delay of about 20 seconds I get a Ubuntu screen with three dots followed shortly by the login screen.

My hardware is an Acer Revo R3610 with 2 gig of RAM
I did a clean install of 32 bit Ubuntu from a USB on 11 Oct 2010 and have had the error on every boot since then. I have the "NVIDIA accelerated graphics driver (current version)" installed (this was as it was recommended and also it resolved an issue I was having getting sound to work via HDMI output)

Also doing a search for "modprobe: FATAl:" in all open ubuntu forums shows at least 4 active threads discussing this issue, so it may well be more widespread than the 90 odd reports in this bug report suggests.

One question - there seems to be an assumption in this thread that because the computer then proceeds to the normal login that everything else is working correctly - what is the evidence to support this assumption ?

[Michael (michaeljt)]

Off-topic: John, I may have sounded rather unkind towards some of our users. There are certainly a number who do experience bugs which "degrade their user experience" (nice industry jargon!) or even affect the way they work with our software, but which we have had to put aside as we were not able to reproduce them with a reasonable effort, and because they only affect a small subset (and particularly no paying customers). In the end we have to tell them that the source is available so that other people can look at problems we don't have time to - we know that this is rather a high barrier for people who aren't software developers, but we don't really have anything better to offer.

[Michael (michaeljt)]

Mark, the assumption that everything else is working correctly is based on the fact that no one has yet linked this to any issues with their day-to-day work using Ubuntu. That is only correct until someone does of course, but I would have thought that anything non-trivial would have been noticed by now.

[Michael (michaeljt)]

A quick random glance through the Ubuntu diff against the kernel source code has thrown up the following:

-static int __init populate_rootfs(void)
+LIST_HEAD(populate_rootfs_domain);
+
+static void __init async_populate_rootfs(void)

If my theory that this bug is caused by a modprobe parallel to the root filesystem population process is correct then that looks suspicious to say the least.

[ZioNemo (zionemo)]

FYI: I have this bug and a strange USB-mouse malfunction (mouse detected, but only buttons work, cursor does not move), which *might* be related.
I will try now the "workaround"

[Michael (michaeljt)]

I think that the culprits are

https://lists.ubuntu.com/archives/kernel-team/2010-February/008658.html ([PATCH 1/2] Make populate_rootfs asynchronous)
https://lists.ubuntu.com/archives/kernel-team/2010-February/008659.html ([PATCH 2/2] async_populate_rootfs: move rootfs init earlier)

Actually these do seem to be Ubuntu changes and not Debian ones - they make initramfs unpacking asynchronous to gain a few tenths of a second in boot time. I am building a local kernel package from the Ubuntu kernel sources minus these two patches and I will see whether that fixes the issue next time I boot (tonight or tomorrow).

[Marceau (marceau-dewilde)]

1) @Michael: You said "I can't comment on the other bugs, but this error message at boot is hardly a critical bug. It is a minor visual blemish that affects a small handful (79 registered that they are affected by this bug) of the many users Ubuntu has, and unless I am mistaken it doesn't prevent anyone from getting their work done using Ubuntu."

I think that is very flawed reasoning. I'm experiencing this bug and hadn't even thought of filing a bug report when I found this page by chance. Even so, I believe a launchpad account is needed to declare oneself a sufferer of the bug. So even if the number of people affected is now at 98, there are probably many many more people who are not bothering with this bug business and/or wouldn't know how if they had to.

2) I've not seen any ill consequences of this bug, except for one person who notes that, like in my case, compiz is broken. Are compiz/compositing issues a known effect of this issue?

My computer: Asus EeePC 1000HE, running the 10.10 official release, installed on a formatted partition.

[ZioNemo (zionemo)]

Workaround made error disappear, but:
- Delay (about 20s) is still there.
- mouse does not work (might be unrelated).

[Michael (michaeljt)]

I think that the delay is also a separate issue, as I saw that even with the mainline kernel which doesn't seem to have this issue.

Assuming that reverting the patches I mentioned above fixes the issue, the next thing would be to find who is calling modprobe at the wrong moment. If anyone knows which kernel function can be called to insert a current stack trace into the kernel log they would save me from searching for it. Presumably inserting a synchronisation point at the right place (which the stack trace should show) would delay the modprobe until modules.dep is available.

[Andy Whitcroft (apw)]

@Michael -- nominally the patches you are suspicious of would make it more likely the files are present than without them as the unpacking starts sooner. I could see how the error might be triggered if the probe occurs in parallel with the unpack, however before the patches the modprobe would have failed as modprobe itself would also not be present before the patches; though it might be quieter. We would then expect udev to sort out the situation when it does its coldplug phase. But I would be interested in the results of your testing, please report back your testing with those patches reverted. If they do affect the result I can help to add the syncronisation to the probe routine as the next step.

Overall I am pretty cirtain that this error is purely cosmetic in nature and would encourage those who have other issues (after this error) should also file separate bugs for those issues, its unlikely this error is the cause of them.

While it is undesirable that this error is emitted so prominantly during boot on any system, even had this error been on the release teams radar, it is unlikely it would have been a release blocker as it does not seem to affect functionality, nor does it affect a large percentage of the install base. All of the release team and the development teams run the development release for the entire time it is being created, and QA test the release on the hardware they have available, and personally I have boot tested the final release code on all the H/W I have and none of it triggers this error. Every release has some bugs that is unavoidable, and why we also produce stable release updates to address them.

[John Stewart (cane-cubo)]

"While it is undesirable that this error is emitted so prominantly during boot on any system, even had this error been on the release teams radar, it is unlikely it would have been a release blocker as it does not seem to affect functionality, nor does it affect a large percentage of the install base."

And right here you have, in its tightest expression, the core reason why Ubuntu will never reach levels of popularity beyond the tiniest fraction of computer users world-wide.

[Andrew Linux (aglinux)]

You're asserting that error messages that can't even be tentatively linked
to any problem with functionality is the reason that Ubuntu won't be popular
with more than a tiny fraction of world-wide computer users? I really don't
think that's right. Windows operating systems are known for "the blue
screen of death" which display cryptic error messages and are typically
followed by system crashes or applications failing and shutting down, yet
Windows OSes have been and remain popular. I don't think an error message
that pops up for 3 seconds during boot is going to prevent Ubuntu from being
popular with more than "the tiniest fraction of computer users wordwide."

I usually don't even watch my computer boot and I had booted the Ubuntu
10.10 a number of times before even noticing the error message.

I think the reason Linux isn't as popular as Windows is because Linux can be
more complicated to use. For example, the command line interface takes
quite a while to learn even though the instructions for it (the man pages)
are quite detailed. Also, computer games are very, very popular and most
games are probably designed with Windows compatibility in mind and no
attention is paid to Linux compatibility by software companies. The same
can be said for the newest pieces of hardware such as video cards, sound
cards, printers, etc. They all often come with CDs for Windows
installation, while Linux can require that the computer's interface for each
new piece of hardware has to be developed and tested individually.

Just my two cents

On Wed, Oct 13, 2010 at 2:29 PM, John Stewart <email address hidden>wrote:

> "While it is undesirable that this error is emitted so prominantly
> during boot on any system, even had this error been on the release teams
> radar, it is unlikely it would have been a release blocker as it does
> not seem to affect functionality, nor does it affect a large percentage
> of the install base."
>
> And right here you have, in its tightest expression, the core reason why
> Ubuntu will never reach levels of popularity beyond the tiniest fraction
> of computer users world-wide.
>
>

[Mustafa Kaya (mustail)]

I want to add that similar to several other users having problems also with their mice, I started to have my mouse pointer (the touchpad or trackpoint, I'm not using a USB mouse) lagging for several seconds out of nothing, after a fresh install of 10.10. I don't know if the modprobe error warning and mouse pointer problem might have anything in common, but I just wanted to note. There is a bug reported for that (#658538) on launchpad.

[John Stewart (cane-cubo)]

Andrew,

Thanks for your reply, and for keeping the (totally misplaced) discussion going :) I too for a long time held the position you have stated.

My opinion has changed though. First observe that Macs also have a command-line interface (though one can get away with using it rarely) and that game support for Macs is extremely limited compared to Windows. But mainly, in my professional life I have observed many software developers -- leading-edge users who otherwise know Unix up and down -- simply wonder away from Linux (to Macs) precisely because of issues such as this: that Linux installations feel brittle, untested, held together by hacks and tape. That's why the optics *are* important. The attitude that "oh well it's just a message that flashes by, and hey it doesn't actually affect the system" in my view is so mistaken. Things that make the system seem broken and unstable send the wrong message to end users.

I was very excited when Ubuntu kicked off because I thought, at last here's an attempt to bring rigour and polish to Linux packaging. Instead, release after release I see that the old attitudes are still prevalent. The new netbook interface in 10.10 is *just terrible*, how anyone thought to release it without saying, "this is really rough so play with it and file bugs" is beyond me. Compare the language here http://www.ubuntu.com/netbook to the reality here http://ubuntuforums.org/showpost.php?p=9953240&postcount=1

[Michael (michaeljt)]

Andy, reverting those patches indeed made the messages go away. Perhaps I have misunderstood something (due to lack of time I didn't study them in detail) but I though that the first of the two made the initramfs unpacking asynchronous, so that kernel activity that would previously have waited for the unpacking to finish was now allowed to proceed in parallel. And it seems to me that that kernel activity includes calls to modprobe which are triggered from the kernel code, not from the initramfs (and/or udev).

I am currently rebuilding with the patches reapplied and the following local change:

--- ../kmod.c 2010-08-02 00:11:14.000000000 +0200
+++ kernel/kmod.c 2010-10-14 01:21:29.275212001 +0200
@@ -120,6 +120,10 @@
    wait ? UMH_WAIT_PROC : UMH_WAIT_EXEC,
    NULL, NULL, NULL);

+ printk(KERN_INFO "Request to load module %s, wait=%d, ret=%d\n",
+ module_name, (int) wait, ret);
+ dump_stack();
+
  atomic_dec(&kmod_concurrent);
  return ret;
 }

I hope that does what I want (it is a five minute patch based on quick googling) and may provide more information about why modprobe is getting called. By the way, is there any way to do a quick rebuild of the kernel package when only a couple of files were changed?

To all others, the delay at boot time really doesn't seem to be related to this error message. I'm sure though that anyone interested can find out more about what is causing it using bootchart or something similar.

[Michael (michaeljt)]

It looks to me like the two modprobe requests that cause the error messages are these ones shown below. If so, perhaps cryptomgr should be compiled into the kernel if it is loaded unconditionally anyway? I will also attach my full bootlog in case it is of interest to anyone.

[ 0.432970] Request to load module cryptomgr, wait=1, ret=256
[ 0.432973] Pid: 1, comm: swapper Not tainted 2.6.35-22-generic #34
[ 0.432975] Call Trace:
[ 0.432983] [<ffffffff81079e2c>] __request_module+0x1dc/0x240
[ 0.432987] [<ffffffff81088f89>] ? sysctl_check_table+0x239/0x320
[ 0.432991] [<ffffffff81580072>] ? detect_ht+0xfe/0x1b2
[ 0.432994] [<ffffffff81084f35>] ? __blocking_notifier_call_chain+0x65/0x80
[ 0.432999] [<ffffffff812932e3>] crypto_probing_notify+0x43/0x60
[ 0.433001] [<ffffffff81295242>] crypto_wait_for_test+0x22/0x80
[ 0.433004] [<ffffffff812953ec>] crypto_register_alg+0x6c/0x80
[ 0.433008] [<ffffffff81b15c8c>] ? md5_mod_init+0x0/0x12
[ 0.433011] [<ffffffff8129a4ee>] crypto_register_shash+0x6e/0xc0
[ 0.433013] [<ffffffff81b15c9c>] md5_mod_init+0x10/0x12
[ 0.433017] [<ffffffff8100204c>] do_one_initcall+0x3c/0x1a0
[ 0.433021] [<ffffffff81aed70a>] do_basic_setup+0x54/0x66
[ 0.433023] [<ffffffff81aed84f>] kernel_init+0x133/0x1bd
[ 0.433026] [<ffffffff8100aee4>] kernel_thread_helper+0x4/0x10
[ 0.433029] [<ffffffff81aed71c>] ? kernel_init+0x0/0x1bd
[ 0.433031] [<ffffffff8100aee0>] ? kernel_thread_helper+0x0/0x10
[ 0.434245] Request to load module cryptomgr, wait=1, ret=256
[ 0.434248] Pid: 1, comm: swapper Not tainted 2.6.35-22-generic #34
[ 0.434250] Call Trace:
[ 0.434254] [<ffffffff81079e2c>] __request_module+0x1dc/0x240
[ 0.434258] [<ffffffff81589255>] ? _raw_spin_lock_irq+0x15/0x20
[ 0.434260] [<ffffffff81580072>] ? detect_ht+0xfe/0x1b2
[ 0.434263] [<ffffffff81084f35>] ? __blocking_notifier_call_chain+0x65/0x80
[ 0.434267] [<ffffffff812932e3>] crypto_probing_notify+0x43/0x60
[ 0.434269] [<ffffffff81295242>] crypto_wait_for_test+0x22/0x80
[ 0.434272] [<ffffffff812953ec>] crypto_register_alg+0x6c/0x80
[ 0.434275] [<ffffffff81b15cc2>] ? krng_mod_init+0x0/0x12
[ 0.434277] [<ffffffff81b15cd2>] krng_mod_init+0x10/0x12
[ 0.434280] [<ffffffff8100204c>] do_one_initcall+0x3c/0x1a0
[ 0.434283] [<ffffffff81aed70a>] do_basic_setup+0x54/0x66
[ 0.434285] [<ffffffff81aed84f>] kernel_init+0x133/0x1bd
[ 0.434287] [<ffffffff8100aee4>] kernel_thread_helper+0x4/0x10
[ 0.434290] [<ffffffff81aed71c>] ? kernel_init+0x0/0x1bd
[ 0.434292] [<ffffffff8100aee0>] ? kernel_thread_helper+0x0/0x10

[Michael (michaeljt)]

Putting a synchronisation point against the end of the root fs unpacking process in __request_module might also be a solution of course.

[Michael (michaeljt)]

I am currently rebuilding with the patch below applied to see if that will fix the issue. It feels very much like a workaround though, and I fear that even if it does the job other problems might pop up elsewhere for other people. Perhaps you might consider posting the original patches to LKML for some more eyeballing?

--- ../api.c 2010-08-02 00:11:14.000000000 +0200
+++ crypto/api.c 2010-10-15 10:19:55.306433001 +0200
@@ -15,8 +15,10 @@
  *
  */

+#include <linux/async.h>
 #include <linux/err.h>
 #include <linux/errno.h>
+#include <linux/init.h>
 #include <linux/kernel.h>
 #include <linux/kmod.h>
 #include <linux/module.h>
@@ -215,6 +217,7 @@
  mask &= ~(CRYPTO_ALG_LARVAL | CRYPTO_ALG_DEAD);
  type &= mask;

+ async_synchronize_full_domain(&populate_rootfs_domain);
  alg = crypto_alg_lookup(name, type, mask);
  if (!alg) {
   request_module("%s", name);
@@ -239,6 +242,7 @@

  ok = blocking_notifier_call_chain(&crypto_chain, val, v);
  if (ok == NOTIFY_DONE) {
+ async_synchronize_full_domain(&populate_rootfs_domain);
   request_module("cryptomgr");
   ok = blocking_notifier_call_chain(&crypto_chain, val, v);
  }

[Michael (michaeljt)]

I hope the patch above is actually correct - I don't do much kernel programming, so I just copied and pasted that async_synchronize_full_domain() from your patch...

[Michael (michaeljt)]

Update: the patch does make the messages go away, although as expected the long boot time (long is relative - compared to my other, older system running Maverick it is long) has not gone away. I am still somewhat dubious though since the original assumption that the kernel won't try to access the root filesystem before init is started does not seem to be quite right.

[joshua christmas (eaerth)]

i was just curious - this is also occurring on my system and i noticed that Mark Harper installed maverick via usb, "I did a clean install of 32 bit Ubuntu from a USB on 11 Oct 2010 and have had the error on every boot since then." i also installed the 32-bit version but my system is amd 64 - i'm not sure whether his is 32 or 64 or if that even has anything to do with it but, might this bug have something to do with installing through a usb and the files are then unable to be located?

best of luck getting this sorted out.

[Davide Capodaglio (davidecapod)]

I am affected by this bug, I am using maverick amd64 (upgrade from lucid). I don't think it's related to usb install.
However, I don't have any particular slowdowns in the boot sequence (maybe just a few seconds more than lucid)

[Serdar Soytetir (tullio)]

I installed Kubuntu 10.10 Final on same laptop twice, using same CD media. Differences between two installation:

First installation: NO seperate HOME partition selected, used only / . Installation started clicking "Install Kubuntu" desktop icon. This installation is clear and NO "could not load /lib/modules/2.6.35-22-generic/modules.dep" error message.

Second installation: seperate HOME part. selected. Installation started clicking "Install Kubuntu" button whlile boot process. The error message appears.
 This bug may be releated things above or it may be random error.

[Nabor Erices (nabor)]

I get this error on a Ubuntu 10.10 fresh install with HOME on a separate partition, after a couple of seconds it shows plymouth splash.
I reintalled evererything on a single partition and the error disappeared , but it doesnt show the plymouth splash.

[Bernhard (b.a.koenig)]

I have two partitions, one lucid and one maverick and I get this. So maybe it's a timing problem of finding and mounting the boot partition.

[djwoodard (woodard-dave)]

@joshua #84:

It's not specific to USB installations. I'm experiencing the bug after an upgrade from 10.04 to 10.10 beta with "upgrade-manager -d".

[Trevor (ths7)]

Is there a problem with Grub that is causing some of this?

I have had problems since trying 10.10 64-bit on a spare partition using a clean install from USB created using 10.04 64-bit

When I updated to kernel 2.6.35-25 using the update manager it didn't update grub, just left the 2.6.35-22 lines in there, so there is an issue there

So I re-installed from usb. My grub.cfg has keywords like msdos & refers to HD.. rather than SD.. (from memory i'm afraid). It certainly doesn't look like the Grub from 10.04

[vexorian (vexorian)]

Once I upgraded to 10.10 I had this issue and also a slow boot issue.

I think both issues are unrelated. Because the boot issue has apparently stopped being a problem after I made sure to run fsck correctly on my huge partitions. For some reason the slow down happened right after fsck ran. I also remember that pressing some keys could make it faster.

But of course the FATAL error remains. Today I was trying to fix it because it occured to me that was the reason I had nvidia driver issues. But apparently many people have the issue and they could not link to it. I have been wondering if the people that had this issue were having issues with restricted drivers like me.

Also, I think that completely-OT posts specially anecdotes about developers supposedly moving to mac are counter productive to helping solve this issue for the few users that have it.

[Michael (michaeljt)]

As mentioned above, this seems to me to be a mainly cosmetic issue due to two Ubuntu-specific patches in the Ubuntu kernel, and those whom it disturbs can work around it by installing the 2.6.35 kernel from the Mainline ppa, which corresponds to the "original" 3.6.35 kernel without any Ubuntu-specific patches. I'm not sure if it is possible to have the kernel from the Mainline ppa automatically update to match new upstream 2.6.35 micro-versions though.

[Matthias Niess (mniess)]

I got this error on Samsung N130 netbook (Atom N270), so it shouldn't have to do with nvidia drivers. Also i DON'T have this on my desktop (AMD + nvidia). I did not notice any slowdowns, on the contrary, the netbook boots faster than before. So any slowdowns others experience might be unrelated. As michaeljt said, this seems to be a cosmetic issue.

[mityi (mityi)]

I think I got this error after reinstalling nvidia drivers.

[Davide Capodaglio (davidecapod)]

I reinstalled nvidia drivers too... because after the upgrade from lucid I wanted to try nouveau drivers with 3d accel.
After that I reinstalled nvidia drivers.
Just a coincidence?

[Robert Zelnik (rzelnik)]

I got this error on HP 510 notebook that has Intel graphic chipset.

[cojerman (marquezuk)]

Same problem, i have windows 7 too, on other partition!!
AMD 64 x2
2 gb ram
80 sata
geforce 8400gs

Help!!! Please!!!

[vexorian (vexorian)]

I can say that after I upgraded to the mainline 2.6.36-rc7 kernel this was fixed.

[phobus (romualdas-striela)]

Thanks durlo. Worked for me too.

[cojerman (marquezuk)]

solution is reinstall el initramfs-tools with synaptic!

C-you!

[Andrew Linux (aglinux)]

I'm not sure that you mean by "el initramfs-tools" but I reinstaled
initramfs-tools with synaptic package manager and it had no effect. I still
twice see the "Fatal: Could not load
/lib/moduls/2.6.35-22-generic/modules.dep: no such file or directory"

On Tue, Oct 19, 2010 at 7:40 PM, cojerman <email address hidden> wrote:

> solution is reinstall el initramfs-tools with synaptic!
>
> C-you!
>
> --
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
> https://bugs.launchpad.net/bugs/642421
>

[russ (russthered)]

Latest kernel upgrade this morning seems to have removed the message. :)

[Raul Morera (rmorera)]

unfortunately not in my case.
I'll try to do a dpkg-reconfigure initramfs-tools with the new kernel to test if this solves the problem.

russ wrote:
>Latest kernel upgrade this morning seems to have removed the message. :)

[GallegoT (gallegot-hotmail)]

No one of the two last "solutions" worked for me neither. Not reinstalling initramfs-tools or updating kernel modules this morning.
Waiting for new kernel modules...
Regards.

[GallegoT (gallegot-hotmail)]

I also have another clean installation in a different hardware and there is no message. Well one is Kubuntu64 another is Ubuntu32. Kubuntu64 is a desktop computer no branded, but with good hardware components and NVIDIA restricted drivers, Ubuntu32 is ASUS U5A laptop.
Has no sense for me.

[Blaster (holst-niels)]

@ #104
You need to edit /etc/initramfs-tools/initramfs.conf (sudo gedit /etc/initramfs-tools/initramfs.conf in the console) and change the line "MODULES=most" to "MODULES=dep" - Then reinstall initramfs using synaptic.

[Michael (michaeljt)]

I can happily provide a test build of the Ubuntu kernel with either my workaround patch applied or the problematic Ubuntu change reverted if someone can tell me what the best way to go about it is. I have a locally build 64bit .deb, but I'm a bit reluctant to just upload that somewhere as I don't want to encourage people to download .debs from sources they don't know...

[Michael (michaeljt)]

By the way, I think the explanation for the "MODULES=most" vs "MODULES=dep" thing is that the modules come before modules.dep in the initramfs archive, so by reducing the number of modules in the archive the time before modules.dep is unpacked into the initial root fs is also reduced and the race condition is less likely to trigger.

[GallegoT (gallegot-hotmail)]

@ #106
"Modules=most" to "Modules=dep" at initramfs.conf worked for me. Thanks.
Same booting time, so.... it wasn't so important, just a "not nice to see" message.

[Blaster (holst-niels)]

@ #108

You're definitely on the right track, Michael.

[Alejandro Arcos (aarcos)]

@ #106
"Modules=most" to "Modules=dep" at initramfs.conf worked for me. Thanks.
Same booting time, so.... it wasn't so important, just a "not nice to see" message.

works for me!!!!

[Andrew Linux (aglinux)]

Hello,

I tried modifying /etc/initramfs-tools/initramfs.conf as suggest:
"MODULES=most" was replaced by "MODULES=dep" and I still receive the error
message, twice.

On Wed, Oct 20, 2010 at 12:21 PM, satara <email address hidden> wrote:

> @ #106
> "Modules=most" to "Modules=dep" at initramfs.conf worked for me. Thanks.
> Same booting time, so.... it wasn't so important, just a "not nice to see"
> message.
>
> works for me!!!!
>
> --
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
> https://bugs.launchpad.net/bugs/642421
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in The Linux Kernel: Confirmed
> Status in “linux” package in Ubuntu: Confirmed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi rimane
> con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left with a
> black screen .. and it is impossible to work ...! thanks
>
>
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

[fralk (fmonte)]

After editing /etc/initramfs-tools/initramfs.conf (sudo gedit /etc/initramfs-tools/initramfs.conf in the console) and changing the line "MODULES=most" to "MODULES=dep", I think it is enough to run "sudo update-initramfs -u" instead of reinstalling initramfs using synaptic. At least it worked for me :)

[Robert Zelnik (rzelnik)]

After editing /etc/initramfs-tools/initramfs.conf - "MODULES=most" replaced to "MODULES=dep", neither "sudo
update-initramfs -u" nor synaptic reinstalling initramfs did not help. The error messages are still present.

[Andrew Linux (aglinux)]

As I mentioned, I tried editing /etc/initramfs-tools/initramfs.conf so that
"MODULES=most" was replaced by "MODULES=dep" and rebooted. I still saw the
error message, however, after trying "sudo update-initramfs -u" and
rebooting, the error message was no longer present. I rebooted a second
time just to make sure.

I also noticed that when I shut down, Ubuntu 10.10 gives me information on
what processes are being shut down (I think this is what it is: a lot of the
lines start of with "init: ").

So, it worked for me.

Anyone think they can explain why this works? I see the two commented out
lines in initramfs.conf

"# most - Add most filesystem and all harddrive drivers.
#
# dep - Try and guess which modules to load."

How does it "guess" which modules to load and why would this get rid of the
error?

On Wed, Oct 20, 2010 at 11:35 PM, Robert Zelnik
<email address hidden>wrote:

> After editing /etc/initramfs-tools/initramfs.conf - "MODULES=most" replaced
> to "MODULES=dep", neither "sudo
> update-initramfs -u" nor synaptic reinstalling initramfs did not help. The
> error messages are still present.
>
> --
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
> https://bugs.launchpad.net/bugs/642421
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in The Linux Kernel: Confirmed
> Status in “linux” package in Ubuntu: Confirmed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi rimane
> con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left with a
> black screen .. and it is impossible to work ...! thanks
>
>
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

[Michael (michaeljt)]

Interestingly my system would not even boot with the new kernel - after the two modules.dep messages I got one about udev trigger not being usable when udev is not configured and was dropped to a busy box shell. I haven't yet build a patched (or unpatched) version of the new kernel, but e.g. using my rebuilt initramfs (from the last kernel) with modules.dep at the beginning and a couple of debug traces did let me boot. Of course this may also be a new, unrelated timing issue.

[Michael (michaeljt)]

Indeed, after reverting https://lists.ubuntu.com/archives/kernel-team/2010-February/008658.html and
https://lists.ubuntu.com/archives/kernel-team/2010-February/008659.html I can boot again. Of course, this could still just be a timing co-incidence, but I don't really have time to investigate properly just now.

[equiman (equiman)]

I'm try to do the Durlo solution... but i have the same messsage.

[equiman (equiman)]

Sorry i'm don add specifications: Ubuntu 10.10 x64 (Fresh Install). I think that this message appear when update to a new Kernel. I update before install and message appear.

Trying with this and i have the same message:

editing /etc/initramfs-tools/initramfs.conf so that "MODULES=most" was replaced by "MODULES=dep".

sudo update-initramfs -u
update-initramfs: Generating /boot/initrd.img-2.6.35.22-generic
grep: /boot/config-2.6.35.22-generic: No existe el archivo o directorio
WARNING: missing /lib/modules/2.6.35.22-generic
Device driver support needs thus be built-in linux image!
WARNING: Couldn't open directory /lib/modules/2.6.35.22-generic: No such file or directory
FATAL: Could not open /lib/modules/2.6.35.22-generic/modules.dep.temp for writing: No such file or directory
FATAL: Could not load /lib/modules/2.6.35.22-generic/modules.dep: No such file or directory

My problem is that the file: /lib/modules/2.6.35.22-generic/modules.dep don't exist but exist this file /lib/modules/2.6.35.22-generic/modules.dep. De diference is between this 2.6.35.22-generic (with ".") and this 2.6.35.22-generic (with "-")

DP: Sorry my english isn't very good.

[equiman (equiman)]

Sorry... a correction. My problem is that the file: /lib/modules/2.6.35.22.generic/modules.dep don't exist but exist this file /lib/modules/2.6.35-22-generic/modules.dep. The diference is between this 2.6.35.22-generic (with ".") and this 2.6.35.22-generic (with "-")

I'm trying to with:
You can try:

$ sudo update && sudo upgrade
$ sudo update-initramfs -u -k all

Or only for that kernel:
$ sudo update-initramfs -u -k 2.6.35.22-generic

But i have the same result.

[russ (russthered)]

Hmmn! See entry #102

:) Just logged on and message is back!

Have done nothing, no updates.

Why should it return?

[lunix (bob-l-lewis)]

What bothers me is that the error message is not accurate.

modprobe: FATAL: Could not load /lib/modules/2.6.35-22-generic/modules.dep:
No such file or directory

modprobe: FATAL: Could not load /lib/modules/2.6.35-22-generic/modules.dep:
No such file or directory

When I go to this directory/folder here is what I see.
-rw-r--r-- 1 root root 278561 2010-10-19 20:00 /lib/modules/2.6.35-22-generic/modules.dep

[Andrew Linux (aglinux)]

I think there's no doubt that the file exists. There's been some discussion
on this in the thread previously and speculation is that, basically, some
process is trying to read modules.dep before the entire filesystem has been
mounted. I may be wrong about this description as I'm not that familiar
with the Linux boot process and terminology, but that's the gist of my
limited understanding.

On Fri, Oct 22, 2010 at 8:07 PM, lunix <email address hidden> wrote:

> What bothers me is that the error message is not accurate.
>
> modprobe: FATAL: Could not load /lib/modules/2.6.35-22-generic/modules.dep:
> No such file or directory
>
> modprobe: FATAL: Could not load /lib/modules/2.6.35-22-generic/modules.dep:
> No such file or directory
>
> When I go to this directory/folder here is what I see.
> -rw-r--r-- 1 root root 278561 2010-10-19 20:00
> /lib/modules/2.6.35-22-generic/modules.dep
>
> --
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
> https://bugs.launchpad.net/bugs/642421
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in The Linux Kernel: Confirmed
> Status in “linux” package in Ubuntu: Confirmed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi rimane
> con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left with a
> black screen .. and it is impossible to work ...! thanks
>
>
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

[Mhpeake (mhpeake)]

I think it is a timing issue, i had installed ubuntu 10.10 on my dell inspiron 9400, it has a intel C2D 2.0 ghz processor, but i had the amd Ubuntu 10.10 on, i just installed the i386 10.10 one and its no longer having that problem, make sure to have the right ubuntu for the right processor, and you shouldnt have any problems

[markofealing (mark-ferns16)]

I agree about this being a possible timing issue, yesterday upgraded to 10.10 on three PCs:

Dell Latitude 640C (P4M 2.4GHz) running Kubuntu - No problem
Athlon 64 Socket 939 x2 4200, with DDR RAM running Kubuntu 64-bit - No problem
Athlon 64 Socket AM2 x2 5600, with DDR2 RAM running Ubuntu 32-bit - Problem

Both Athlon 64 systems use Adaptec SCSI controllers and 72Gb disks for the OS so the only real difference is CPU and RAM performance. I don't think 64-bit makes a difference.

[lunix (bob-l-lewis)]

I could easily be mis-remembering but after doing an upgrade from 10.04 -> 10.10 everything worked fine. This system is an I5 64-bit installation and it has Ugh, windows 7 as a bootable partition. I never boot into Windows 7 but a relative needed something done from Windows 7 so I did allowed it to get the latest updates. After that I could no longer boot into Ubuntu or Windows 7 as Windows 7 had rewritten the MBR.

To fix this I followed the following article.
https://help.ubuntu.com/community/RecoveringUbuntuAfterInstallingWindows
Which told me to do:
sudo grub-install --root-directory=/media/0d104aff-ec8c-44c8-b811-92b993823444 /dev/sda
Where I replaced the hex numbers with mine. This allowed me to see grub2 menus and select Ubuntu again. All went well, except that is when I noticed the issue we are discussing here.

[wendelmax (jacksonwendel)]

Here was giving error. solved with this command: "sudo update-initramfs -c -k 2.6.35-22-generic"

[lunix (bob-l-lewis)]

"sudo update-initramfs -c -k 2.6.35-22-generic"
worked without an error message but did not solve the problem for me being discussed here.

[Martin Plaček (placek-martin)]

"sudo update-initramfs -c -k 2.6.35-22-generic" had no effect for me. "MODULES=dep" caused that the messages are away BUT it slowed down my system.

[Michael (michaeljt)]

For those who are interested, installing bootchart shows that ureadahead runs for nearly 13 seconds here during boot. During part of that plymouthd is also accessing the disk, and for over eight seconds ureadahead is the only active process on the system.

[Michael (michaeljt)]

Bug 552165 may be relevant.

[Vanillalite (vanillasky26)]

Figured I'd chime in again with my system specs...

Mobo: Abit Ab9
CPU: c2s 6600
RAM: 2 gigs
GPU: Ati 4830
HDD: 7200 RPM 250 gig SATA

Also I'm dual booting with Windows 7 using grub on 11.04 64bit regular disc install (not wubi).

[Vanillalite (vanillasky26)]

EDIT: For last post I meant to say 10.10 not 11.04 and also my cpu I meant c2d 6600. The s was a typo!

[srinivas (docbsreenu)]

thank you it worked for me, i no more see the terrifying message.

[ThomasLee82 (kamiyasha)]

Still having this issue after updating to 2.6.35-23-generic.

[Martin Plaček (placek-martin)]

Same situation after upgrade to 2.6.35-23-generic.

[rquint (richard-quint)]

One more user with the same bug after upgrades.

I'm running 2.6.35-23-generic on an Intel(R) Pentium(R) 4 CPU 2.60GHz

[halil kömürcü (halilkomurcu)]

ı installed ubuntu 10.10 with wubi,
and ı got same problem.
when ı do MODULES=most to MODULES=dep, ı couldn't update initramfs
ı get below message
E: initramfs-tools: installed post-installation script alt işlemi çıkış durumunda hata döndürdü

W: Ignoring file 'yorba-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'opera-beta.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'opera.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'ubuntu-mozilla-security-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'emesene-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'pidgin-developers-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'pdfmod-team-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'elegant-gnome-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'ubuntu-wine-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'ubuntu-tweak-stable.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'bisigi-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'yorba-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'opera-beta.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'opera.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'ubuntu-mozilla-security-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'emesene-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'pidgin-developers-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'pdfmod-team-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'elegant-gnome-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'ubuntu-wine-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'ubuntu-tweak-stable.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension
W: Ignoring file 'bisigi-ppa.list.save' in directory '/etc/apt/sources.list.d/' as it has an invalid filename extension

and ı couldn't clear the sources.list.d

[halil kömürcü (halilkomurcu)]

when ı do MODULES=most to MODULES=dep, it couldn't solve my problem.
also couldn't install any packet
today ı changed MODULES=most to MODULES=list
and ı reinstalled initramfs-tool,
when ı restarted ubuntu ı didn't get modprobe: FATAL... error.
but opening time is same as before.

[valeriy (val-varnava)]

Yes 'halil' MODULES=most -> MODULES=list (but not MODULES=dep)

[Ubucolors (ubucolors)]

Ok, it solved that problem with my boot too,

What i did: sudo nautilus, then change in the file initramfs.conf:
MODULES=most to MODULES=list, save the file, then
sudo update-initramfs -u and then a restart that shows me that i don't see
the message again, finally !

2010/11/28 valeriy <email address hidden>

> Yes 'halil' MODULES=most -> MODULES=list (but not MODULES=dep)
>
> --
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
> https://bugs.launchpad.net/bugs/642421
> You received this bug notification because you are a direct subscriber
> of the bug.
>

[Martin Plaček (placek-martin)]

MODULES=list makes my system dead :-(( I had to restore my system from live cd through chroot and repair initramfs.conf. I don't recommend this tip, it's too unreliable.

[Runar Ingebrigtsen (ringe)]

modules=dep solved this issue for me on an Acer Aspire One A250 netbook with dual boot Maverick and Windows 7.

[GallegoT (gallegot-hotmail)]

modules=list killed my system too. It was necesary to boot from previous kernel, change to modules=dep and so...
I don't recomend modules=list.

[Artem Yakimenko (temik)]

People! Read the manpages and do not change initramfs settings if you don't
know what you're doing!

       MODULES
             Specifies the modules for the initramfs image. The default
set-
             ting is most.

             *most* adds all the framebuffer, acpi, file system, ide,
sata,
             scsi and usb drivers.

             *dep* tries to guess which modules are necessary for the
 running
             box.

             *netboot* adds the base modules, network modules, but skips
block
             devices.

             *list* includes only modules from the additional modules list.

Explaining the dep function "guess" thing. This setting loads the essential
kernel modules plus the dependencies.
The error does not appear in some cases because dep setting loads less
modules and takes less time and resourses, so as far as I know it's a timing
issue.
It was explained in much detail in the start of this discussion.

Regards,
Artem Yakimenko

On Mon, Nov 29, 2010 at 10:58 PM, GallegoT <email address hidden>
wrote:
> modules=list killed my system too. It was necesary to boot from previous
kernel, change to modules=dep and so...
> I don't recomend modules=list.
>
> --
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
> https://bugs.launchpad.net/bugs/642421
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in The Linux Kernel: Confirmed
> Status in “linux” package in Ubuntu: Confirmed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
/lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left with a
black screen .. and it is impossible to work ...! thanks
>
>
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

[Danny (koffeehaus)]

This happens when you tick "Install updates" during installation of the system. The installer first places the bootloader for current kernel at BIOS; but if the "Install updates" was ticked, the system installs the newer version of kernel and, apparently, does not the old one.

For example, the module on my Live CD was "generic-module-2.6.35". During installation I ticked for automatic updates, which resulted in the installer downloading a newer version. The bootloader, however was not updated. Hence its showing the "No such file or directory" message.

The error message is not fatal, but inconvenient as it cripples the aesthetics. To fix this I had to go an extra mile and reinstall the system, but this time with automatic updating switched off.

[lunix (bob-l-lewis)]

Perhaps someone can explain to me why Ubuntu hasn't released a patch to fix this issue? Given the number of people having the issue and the dialogue that has taken place one would think something substantive would have happened by now.

[markofealing (mark-ferns16)]

Danny, thank you for the explaination, this makes a lot of sense but does not really explain how this was not picked up in testing before release. After all, doing an upgrade and installing updates is in my view "best practice" when doing an upgrade.

May be Launchpad should have a way for the community members to prioritise the importance of fixing the bugs rather than being dependent on the developers who may have their own separate agendas?

[Davide Capodaglio (davidecapod)]

@Danny: you are wrong, the message has nothing to do with GRUB bootloader, it is printed by Ubuntu boot sequence.
(modbprobe is the command to load a linux kernel module).
I have this nasty message and I upgraded from 10.04, so the reason is not the one you explain.
As others have said, it is a timing issue in filesystem mounting during the boot sequence.

[Narcis Garcia (narcisgarcia)]

I think that some of the modules listed in /lib/modules/2.6.35-xx-generic/modules.dep does not probe properly.

The MODULES=dep in /etc/initramfs-tools/initramfs.conf I believe that only makes update-initramfs to select the modules before including them in the kernel load.

My questions:
- Could this mean that with MODULES=dep will be necessary to update-initramfs on hardware changes? (to re-select modules to load)
- Can come the problem from 1 of the modules listed in /lib/modules/2.6.35-xx-generic/modules.dep ?

[Mustafa Kaya (mustail)]

I had not seen the error message for some time now (except the wait was there), until today when I did an update. The command uname -r gives the result as 2.6.35-24-generic
My system is Ubuntu 10.10 x86.

[Grewmj (grewmj)]

Some additional information... maybe...

I have just installed Ubuntu 10.10 on 2 systems (both dual boot, both fresh install, both 64 bit).

System #1 is a home built media pc with Asus m4a77d with an athlon x2 BE-2350 and Radeon HD-4670 PCIe, with windows xp as the other operating system. This system never got the message and boots quite fast for the slow hard drive.

System #2 is a HP G60 Laptop with an intel core 2 duo T6600 @2.20 GHz with windows 7 64 bit home premium as the other opersating system. This system got the message, I did the "Modules=dep" fix, and the messages went away, however like others the boot time remains the same... approx 15-20 seconds of black screen... then it runs like normal and is lightning fast.

hope that helps... If I get time I am planning on performing another install on my desktop which is another asus motherboard and AMD, however this system has a quad core, onboard graphics, and has windows 7 ultimate 64 bit.

-Mike

[Andy Whitcroft (apw)]

On Sat, Dec 04, 2010 at 07:32:56AM -0000, markofealing wrote:
> Danny, thank you for the explaination, this makes a lot of sense but
> does not really explain how this was not picked up in testing before
> release. After all, doing an upgrade and installing updates is in my
> view "best practice" when doing an upgrade.

It was not located during testing as it did not affect any of the
systems on which the tests were performed. As Michaels very cogent
analysis demonstrates this is a race between unpacking the initramfs
and modules attempting to load. Now the way the boot works they may
fail then but will then load during the udev cold boot later in
userspace. This makes it highly unlikely that the messages are anything
other than cosmetic.

> May be Launchpad should have a way for the community members to
> prioritise the importance of fixing the bugs rather than being dependent
> on the developers who may have their own separate agendas?

This bug is known about, and due to Michaels detailed work likely is
fixable once we have done some performance analysis. But it is almost
cirtain from that analysis that the message (though prominant and bad
sounding) is benign in actuality. Though it may seem important to have
prettyness during boot, and understand we have a stated goal of a pretty
and non-techie friendly boot, achieving that is not a priority over
other types of bugs such as data corruption or crashes.

Michael once more thank you for your detailed and thoughtful analysis of
this issue. I will try and figure out the best way to use that to
remove the error without penalising performance.

-apw

[Andy Whitcroft (apw)]

Ok using Michaels analysis as a basis I have spun a more comprehensive patch to ensure the root disk is unpacked in its entirety before any userspace helper is triggered. As I have no hardware which exhibits this behaviour I have been unable to confirm that the patch is complete or indeed fixes anything at all. So if those of you who see this issue could test the Natty kernels below (they should work on Maverick userspace too) and let me know. Please report back here:

    http://people.canonical.com/~apw/lp642421-natty/

[Tim (tim-eilers)]

That kernel made the messages go away at my system. However i can not succesful boot into graphical user interface and virtualbox (vboxdrv) causes a kernel oops.
Perhaps i could have solved that, but didn't want to completely reconfigure my system.
Is it possible to backport that patch to the current Maverick kernel?

[Yann (ylevot)]

The Natty kernel provided for test seems to fix the issue for my ASUS A6J laptop. I don't see any regression so far.
Do we get a chance to see the patch backported to a Maverick update?

[Triwanto Simanjuntak (lu176)]

@Andy, the kernel for Natty solves the problem for my 32 bit ThinkPad X60 too.

[Andy Whitcroft (apw)]

Ok I've attempted a backport to Maverick, could those who have this issue please test the kernels at the URL below and report back here:

    http://people.canonical.com/~apw/lp642421-maverick/

[Triwanto Simanjuntak (lu176)]

@Andy, I just tested the backported kernel, it also solves the problem on my 32 bit Thinkpad X60. Many thanks.

[Michael (michaeljt)]

Andy, sorry for the slow reaction. The updated Maverick kernel seems to do the trick here too. Thanks.

[Matthias Niess (mniess)]

@Andy: here is another "me too". Works on my maverick netbook.

On a sidenote: you can play this down as being a cosmetic issue. I for myself don't care about boot aesthetics. But a casual or non-savvy user will ALWAYS think "Ubuntu *seems* to work nicely, but something is broken" when they get a *fatal* error on every boot, every day.

[lunix (bob-l-lewis)]

Its not clear to me that I would call it a cosmetic issue since quite a delay is introduced on each boot.
Does the fix eliminate that delay? Also, what is considered cosmetic to a programmer is not cosmetic
to grandpa or an ordinary user of the OS in my opinion.

Will this patch be distributed in a normal update sometime soon?

[Andrew Somerville (andy16666)]

Agreed. I haven't seen it mentioned here, but sometimes the system will actually hang with that error, and won't continue to boot until a key is pressed. On my system is usually doesn't hang like this, but occasionally it does. I have a Phenom X4 with 8GB of RAM and a 64GB SSD as a root drive.

[Davide Capodaglio (davidecapod)]

Fixed for me too!

[Yann (ylevot)]

The Maverick patch works for me too on my Asus laptop. Thanks !

[Andrew Betts (andrew-betts)]

Patch works for me on amd 64.

[Davide Capodaglio (davidecapod)]

I forgot to say, I am using maverick amd64, upgraded from lucid.

[abrianb (abrianb2003)]

I did not have this problem the first time I installed Maverick on my thinkpad t-60. Last week I decided to upgrade the hard drive from a 5400 rpm seagate to a 7200 rpm hitachi and do a new install of maverick. After the new install I got the modprobe error on boot. My boot time went from 60 seconds with the seagate to 75 seconds with the hitachi.
 I installed the patched linux headers and image suggested by Andy Whitcroft and I no longer get the error on boot. Boot time is now 30 seconds. Thank you Andy!

[Gerald Williams (off-wfam)]

Affects my Toshiba A215-S7422 laptop. In addition to boot warnings, menus don't appear properly transparent and video artifacts appear.

Going back to latest stable kernel from 10.4 resolves all, recommended workarounds with initramfs and rebuilding modules.dep did not help. Haven't tried the backported kernel yet (have an AMD Turion64, was running 32-bit, didn't want to grab the wrong version).

[John Stewart (cane-cubo)]

Andy's patched kernel worked like a charm on my EEE PC 1001px, on which I
originally noticed the boot-time error message. On the netbook 10.10 was
installed in its own partition, separate from the bootable partition of the
system (which has 10.04 on it).

Thanks Andy, terrific work. I was also one of the early greasy-well
complainers about this bug, arguing that it's much more than merely
cosmetic. I still think that's true!

jds

On Thu, Jan 13, 2011 at 8:26 AM, Gerald Williams
<email address hidden>wrote:

> Affects my Toshiba A215-S7422 laptop. In addition to boot warnings,
> menus don't appear properly transparent and video artifacts appear.
>
> Going back to latest stable kernel from 10.4 resolves all, recommended
> workarounds with initramfs and rebuilding modules.dep did not help.
> Haven't tried the backported kernel yet (have an AMD Turion64, was
> running 32-bit, didn't want to grab the wrong version).
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/642421
>
> Title:
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>
> Status in The Linux Kernel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Confirmed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left
> with a black screen .. and it is impossible to work ...! thanks
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

[Savvas Radevic (medigeek)]

Thank you for your effort Andy! When can we expect the fix in the repos? :)

[Joe Niski (joeniski)]

The patch works for 64-bit Maverick me on my CLEVO:M570RU-1.00.04SL (a.k.a. Sager 5791, also packaged & reold by AlienWare) with a 2x Intel(R) Core(TM)2 Duo CPU T7700 @ 2.40GHz.

Boot time isn't noticeably faster, but the FATAL error is gone. Maverick has always been much slower to boot than Lucid on this machine.

Once i figured out to install headers-all before headers, it went smoothly - thanks so much for the patch!

[lunix (bob-l-lewis)]

>
> Once i figured out to install headers-all before headers, it went
> smoothly - thanks so much for the patch!
>
> --
>

I have been resisting putting the patch on as I share
this machine with others and I didn't want to upset anything or screw up in
the order. I suppose I could just boot from one of the other kernels if
something went wrong.

Perhaps someone who has been successful, apparently all of you might post an
extra procedure for those that are used to just going through the normal
update options.
Will this patch be released as an update?

I have a 64-bit machine. It looks like I need to install:
linux-image-2.6.35-25-generic_2.6.35-25.43lp642421v201101091714_amd64.deb

linux-headers-2.6.35-25_2.6.35-25.43lp642421v201101091714_all.deb

linux-headers-2.6.35-25-generic_2.6.35-25.43lp642421v201101091714_amd64.deb

Then do I reboot and come up on this and then apply the patch or apply the
patch straight away while running the older kernel?

0001-UBUNTU-SAUCE-ensure-root-is-ready-before-running-use.patch

Do I just type as root: patch
0001-UBUNTU-SAUCE-ensure-root-is-ready-before-running-use.patch

Does it matter what directory the above patch file is in when I run the
patch command?

[John Stewart (cane-cubo)]

I didn't use the patch command at all. I just installed the two header
files and the kernel using 'dpkg -i <file>', then rebooted. Grub will show
you the new kernel on the boot screen.

jds

On Sat, Jan 15, 2011 at 1:59 PM, lunix <email address hidden> wrote:

> >
> > Once i figured out to install headers-all before headers, it went
> > smoothly - thanks so much for the patch!
> >
> > --
> >
>
> I have been resisting putting the patch on as I share
> this machine with others and I didn't want to upset anything or screw up in
> the order. I suppose I could just boot from one of the other kernels if
> something went wrong.
>
> Perhaps someone who has been successful, apparently all of you might post
> an
> extra procedure for those that are used to just going through the normal
> update options.
> Will this patch be released as an update?
>
> I have a 64-bit machine. It looks like I need to install:
> linux-image-2.6.35-25-generic_2.6.35-25.43lp642421v201101091714_amd64.deb
>
> linux-headers-2.6.35-25_2.6.35-25.43lp642421v201101091714_all.deb
>
> linux-
> headers-2.6.35-25-generic_2.6.35-25.43lp642421v201101091714_amd64.deb
>
> Then do I reboot and come up on this and then apply the patch or apply the
> patch straight away while running the older kernel?
>
> 0001-UBUNTU-SAUCE-ensure-root-is-ready-before-running-use.patch
>
> Do I just type as root: patch
> 0001-UBUNTU-SAUCE-ensure-root-is-ready-before-running-use.patch
>
> Does it matter what directory the above patch file is in when I run the
> patch command?
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/642421
>
> Title:
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>
> Status in The Linux Kernel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Confirmed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left
> with a black screen .. and it is impossible to work ...! thanks
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

[Denis Turrina (denis86pm-deactivatedaccount)]

@Andy:

First, sorry for late answer.
Patch worked on AMD Turion64 X2 TL-50, Maverick upgraded from Lucid on
Compaq Presario V6131EU.
Didn't slow boot time.

Thanks a lot!

DT

2011/1/15, John Stewart <email address hidden>:
> I didn't use the patch command at all. I just installed the two header
> files and the kernel using 'dpkg -i <file>', then rebooted. Grub will show
> you the new kernel on the boot screen.
>
> jds
>
> On Sat, Jan 15, 2011 at 1:59 PM, lunix <email address hidden>
> wrote:
>
>> >
>> > Once i figured out to install headers-all before headers, it went
>> > smoothly - thanks so much for the patch!
>> >
>> > --
>> >
>>
>> I have been resisting putting the patch on as I share
>> this machine with others and I didn't want to upset anything or screw up
>> in
>> the order. I suppose I could just boot from one of the other kernels if
>> something went wrong.
>>
>> Perhaps someone who has been successful, apparently all of you might post
>> an
>> extra procedure for those that are used to just going through the normal
>> update options.
>> Will this patch be released as an update?
>>
>> I have a 64-bit machine. It looks like I need to install:
>> linux-image-2.6.35-25-generic_2.6.35-25.43lp642421v201101091714_amd64.deb
>>
>> linux-headers-2.6.35-25_2.6.35-25.43lp642421v201101091714_all.deb
>>
>> linux-
>> headers-2.6.35-25-generic_2.6.35-25.43lp642421v201101091714_amd64.deb
>>
>> Then do I reboot and come up on this and then apply the patch or apply the
>> patch straight away while running the older kernel?
>>
>> 0001-UBUNTU-SAUCE-ensure-root-is-ready-before-running-use.patch
>>
>> Do I just type as root: patch
>> 0001-UBUNTU-SAUCE-ensure-root-is-ready-before-running-use.patch
>>
>> Does it matter what directory the above patch file is in when I run the
>> patch command?
>>
>> --
>> You received this bug notification because you are a direct subscriber
>> of the bug.
>> https://bugs.launchpad.net/bugs/642421
>>
>> Title:
>> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>>
>> Status in The Linux Kernel:
>> Confirmed
>> Status in “linux” package in Ubuntu:
>> Confirmed
>>
>> Bug description:
>> When I turn on my computer I can read this on my screen
>>
>> "modprobe: Fatal: Could not load
>> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>>
>> After 4-5 seconds, plymouth starts and the system works fine.
>>
>> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
>> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>>
>> hi, I have the same problem, I asked the psw keychain and then left
>> with a black screen .. and it is impossible to work ...! thanks
>>
>> To unsubscribe from this bug, go to:
>> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>>
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/642421
>
> Title:
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>
> Status in The Linux Kernel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Confirmed
>
> Bug description:
> When I turn on my computer I can read this o...

[lunix (bob-l-lewis)]

I finally got up the nerve to add the kernel to my Intel based 64-bit system.
The error message is indeed gone. I got out a stop watch and the boot time
for me is 25-seconds prior and after the patch so no increase in speed booting
but I don't have to look at the error message any longer which is nice.
25-seconds booting is pretty phenomenal in my view.

Thanks Andy and sorry for the delay.

Bob

[lecomte (le-comte)]

I' have the same problem and read all your comments: it's all ok but can you detail how to install the patch?
Besides, which one I've to install considering my pc is an Acer TimeLineX Intel i3?
Thanks
Carlo

[Martin Plaček (placek-martin)]

I have updated kernel to 2.6.35-25 today. It has had no effect for me. Strange...

[Nikita Malyavin (nikitamalyavin)]

It seems that 2.6.37 rc2 from mainline also fixes this bug, BTW it also boots up faster

[Denis Turrina (denis86pm-deactivatedaccount)]

Same as Martin: No more benefits with official kernel release update.
As I told before, patch worked on my laptop.

2011/1/28, Nikita Malyavin <email address hidden>:
> It seems that 2.6.37 rc2 from mainline also fixes this bug, BTW it also
> boots up faster
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/642421
>
> Title:
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>
> Status in The Linux Kernel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Fix Released
> Status in “linux” source package in Maverick:
> Fix Committed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left
> with a black screen .. and it is impossible to work ...! thanks
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

--
Denis Turrina
Dottore in Sicurezza dei Sistemi e delle Reti Informatiche

"Il computer non è una macchina intelligente che aiuta le persone
stupide, anzi è una macchina stupida che funziona solo nelle mani
delle persone intelligenti" U. Eco

[Willeo Soeurs (ding-yeen)]

its weird that kernel 2.6.35.25 does not update my kernel at all

still stuck at 2.6.35.24

[Oscar (mail4oscar)]

Hi Guys,
I'm struggling with this issue for quite some time now. I think I tried every suggestion I found on the web (except reinstall.... that's not an option :-()
I'm using 10.10 on an Asus F70 laptop. Unlike most responders I get the "Fatal" error during boot, but the boot process does *not* continue at all!
After changing MODULES=most to MODULES=dep, the message did disappear, but I still cannot boot (booted late at night and next morning the system still wasn't there.... just to make sure nobady claims I'm impatient :-S). All I can do to get my system running, is boot the 2.6.32-26-generic; I'm having this issue with both 2.6.35 kernels. I don't dare to update to 2.6.35-25 since I can't afford to loose this system (or better; to loose the time I'ld need to setup my development environment again).

So in my opinion, this is a critical issue.....

[Andy Whitcroft (apw)]

On Mon, Feb 07, 2011 at 10:35:09AM -0000, Oscar wrote:
> Hi Guys,
> I'm struggling with this issue for quite some time now. I think I tried every suggestion I found on the web (except reinstall.... that's not an option :-()
> I'm using 10.10 on an Asus F70 laptop. Unlike most responders I get the "Fatal" error during boot, but the boot process does *not* continue at all!
> After changing MODULES=most to MODULES=dep, the message did disappear, but I still cannot boot (booted late at night and next morning the system still wasn't there.... just to make sure nobady claims I'm impatient :-S). All I can do to get my system running, is boot the 2.6.32-26-generic; I'm having this issue with both 2.6.35 kernels. I don't dare to update to 2.6.35-25 since I can't afford to loose this system (or better; to loose the time I'ld need to setup my development environment again).
>
> So in my opinion, this is a critical issue.....

If you managed to switch to MODULES=dep and the the messages were gone
but you still could not boot then that indicates you have a separate
issue and this kernel is not going to fix your issues. I would strongly
recommend you start your own bug with additional symtoms (Fatal message
etc) prominantly reported.

-apw

[Oscar (mail4oscar)]

Thnx for your answer Andy, but I'm afraid I need an extra hand filing a new but; the bug would contain something like "system won't boot"; all msgs I ever got is the 'modprobe: Fatal (...)'
There's no bootlog (since I can't boot the failing kernel, so any suggestions how I can gather more info? (sorry for being off-topic then :-S)

[matt_callaghan (matt-callaghan)]

Andy, I wasn't able to install the headers/image deb files.

During the headers installation:
{{{
 linux-headers-2.6.35-25-generic depends on linux-headers-2.6.35-25; however:
  Package linux-headers-2.6.35-25 is not installed.
dpkg: error processing linux-headers-2.6.35-25-generic (--install):
 dependency problems - leaving unconfigured
}}}

During the image installation:
{{{
 linux-headers-2.6.35-25-generic depends on linux-headers-2.6.35-25; however:
  Package linux-headers-2.6.35-25 is not installed.
dpkg: error processing linux-headers-2.6.35-25-generic (--install):
 dependency problems - leaving unconfigured
}}}

Full text: http://tinypaste.com/40b51d

[matt_callaghan (matt-callaghan)]

correction:
During the image installation:
{{{
-2.6.35-25-generic
 * dkms: running auto installation service for kernel 2.6.35-25-generic
 * virtualbox-ose (3.2.8)... [fail]
 * nvidia-current (260.19.06)... [fail]
}}}

[matt_callaghan (matt-callaghan)]

sigh, sorry, i had some bad copy/paste output in the previous tinypaste, proper: http://tinypaste.com/9064c

[Omer Akram (om26er)]

matt, you also have to install this
http://people.canonical.com/~apw/lp642421-maverick/linux-headers-2.6.35-25_2.6.35-25.43lp642421v201101091714_all.debat
the same time.

On Sun, Feb 20, 2011 at 11:55 AM, matt_callaghan
<email address hidden>wrote:

> sigh, sorry, i had some bad copy/paste output in the previous tinypaste,
> proper: http://tinypaste.com/9064c
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/642421
>
> Title:
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>
> Status in The Linux Kernel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Fix Released
> Status in “linux” source package in Maverick:
> Fix Committed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left
> with a black screen .. and it is impossible to work ...! thanks
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

[matt_callaghan (matt-callaghan)]

Thanks Omer, good call! On i386, installing these three worked:
{{{
linux-headers-2.6.35-25_2.6.35-25.43lp642421v201101091714_all.deb
linux-headers-2.6.35-25-generic_2.6.35-25.43lp642421v201101091714_i386.deb
linux-image-2.6.35-25-generic_2.6.35-25.43lp642421v201101091714_i386.deb
}}}

[matt_callaghan (matt-callaghan)]

Curious; now that we've manually added these linux-headers/image, will ubuntu auto-update upgrade us automatically once a later kernel is released upstream?

[matt_callaghan (matt-callaghan)]

bah; comment spamming (sorry).

I did:
 1) installed the patched kernel/images/headers, seemed flawless.
 2) rebooted, yay! was fixed. the "shiny" ubuntu logo appears as expected, and there were no FATAL messages.
 3) ubuntu prompted me for a whack of updates (it looked like several applications needed to be updated w/ the newer kernel)
 4) rebooted, darn. got the same error as before. ... FATAL message present again.

[John Stewart (cane-cubo)]

Sorry if this is an obvious question, but are you sure you rebooted from the
patched kernel and not some updated ubuntu kernel?

jds

On Sun, Feb 20, 2011 at 6:02 PM, matt_callaghan
<email address hidden>wrote:

> bah; comment spamming (sorry).
>
> I did:
> 1) installed the patched kernel/images/headers, seemed flawless.
> 2) rebooted, yay! was fixed. the "shiny" ubuntu logo appears as expected,
> and there were no FATAL messages.
> 3) ubuntu prompted me for a whack of updates (it looked like several
> applications needed to be updated w/ the newer kernel)
> 4) rebooted, darn. got the same error as before. ... FATAL message
> present again.
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/642421
>
> Title:
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>
> Status in The Linux Kernel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Fix Released
> Status in “linux” source package in Maverick:
> Fix Committed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left
> with a black screen .. and it is impossible to work ...! thanks
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

[matt_callaghan (matt-callaghan)]

Yeah it auto updates grub, and it was the new default.
Plus, after the reboot, when ubuntu wanted to update packages, the
kernel version was the new one.

-------- Original Message --------
From: John Stewart <email address hidden>
Sent: Sun 20 Feb 2011 07:48:57 PM EST
To: <email address hidden>
Cc:
Subject: Re: [Bug 642421] Re: Maverick could not load
/lib/modules/2.6.35-22-generic/modules.dep

Sorry if this is an obvious question, but are you sure you rebooted from the
patched kernel and not some updated ubuntu kernel?

jds

On Sun, Feb 20, 2011 at 6:02 PM, matt_callaghan
<email address hidden>wrote:

bah; comment spamming (sorry).

I did:
   1) installed the patched kernel/images/headers, seemed flawless.
   2) rebooted, yay! was fixed. the "shiny" ubuntu logo appears as
expected,
and there were no FATAL messages.
   3) ubuntu prompted me for a whack of updates (it looked like several
applications needed to be updated w/ the newer kernel)
   4) rebooted, darn. got the same error as before. ... FATAL message
present again.

--
You received this bug notification because you are a direct subscriber
of the bug.
https://bugs.launchpad.net/bugs/642421

Title:
   Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep

Status in The Linux Kernel:
   Confirmed
Status in “linux” package in Ubuntu:
   Fix Released
Status in “linux” source package in Maverick:
   Fix Committed

Bug description:
   When I turn on my computer I can read this on my screen

   "modprobe: Fatal: Could not load
   /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"

   After 4-5 seconds, plymouth starts and the system works fine.

   salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
   rimane con lo schermo nero..ed è impossibile lavorare...! grazie

   hi, I have the same problem, I asked the psw keychain and then left
   with a black screen .. and it is impossible to work ...! thanks

To unsubscribe from this bug, go to:
https://bugs.launchpad.net/linux/+bug/642421/+subscribe

[Joe Niski (joeniski)]

I've had the same issue - installed the patch, it worked fine until a recent
set of updates, now seeing the FATAL message on startup again.
On Feb 20, 2011 5:20 PM, "matt_callaghan" <email address hidden> wrote:
> Yeah it auto updates grub, and it was the new default.
> Plus, after the reboot, when ubuntu wanted to update packages, the
> kernel version was the new one.
>
> -------- Original Message --------
> From: John Stewart <email address hidden>
> Sent: Sun 20 Feb 2011 07:48:57 PM EST
> To: <email address hidden>
> Cc:
> Subject: Re: [Bug 642421] Re: Maverick could not load
> /lib/modules/2.6.35-22-generic/modules.dep
>
>
> Sorry if this is an obvious question, but are you sure you rebooted from
the
> patched kernel and not some updated ubuntu kernel?
>
> jds
>
> On Sun, Feb 20, 2011 at 6:02 PM, matt_callaghan
> <email address hidden>wrote:
>
> bah; comment spamming (sorry).
>
> I did:
> 1) installed the patched kernel/images/headers, seemed flawless.
> 2) rebooted, yay! was fixed. the "shiny" ubuntu logo appears as
> expected,
> and there were no FATAL messages.
> 3) ubuntu prompted me for a whack of updates (it looked like several
> applications needed to be updated w/ the newer kernel)
> 4) rebooted, darn. got the same error as before. ... FATAL message
> present again.
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/642421
>
> Title:
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>
> Status in The Linux Kernel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Fix Released
> Status in “linux” source package in Maverick:
> Fix Committed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left
> with a black screen .. and it is impossible to work ...! thanks
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/642421
>
> Title:
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>
> Status in The Linux Kernel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Fix Released
> Status in “linux” source package in Maverick:
> Fix Committed
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left
> with a black screen .. and...

[Wayne Stark (wastark-gmail)]

Same problem for me using kernel 2.6.35.25.
I found that installing kernel 2.6.35.11 from http://kernel.ubuntu.com/~kernel-ppa/mainline/ solved the problem.
Strange because I thought the 2 kernels corresponded to one another, but either way it worked.

[Jack (jacknoen)]

Problem was solved for me with installation of kernel 2.6.35.28.

Note: kernel 2.6.35.28 was patched as explained on http://ubuntuforums.org/showthread.php?t=1594981&page=9 (related to bug 636045) but don't think this is related anyhow.

[Andy Whitcroft (apw)]

Closing Maverick tasks Fix Released as the issue which triggers this message is now patched for Maverick.

[Sree15081947 (sree15081947)]

Hi friends, once me also disturbed because of the same problem. this can be avoid by reinstalling the following files http://people.canonical.com/~apw/lp642421-maverick/
just check...
thanks.

[Stuart Ledwich (stuart-ledwich)]

Downloaded regular kernel update today which is 2.6.35-27 and this has fixed my problem, error mess has now gone when booting - thanks

[Larry Whitehead (llw8008)]

I upgraded both of my computers with the regular kernel update last night
and things are looking much better now with 2.6.35-27. Thanks and best
regards.

On Sat, Mar 5, 2011 at 8:49 AM, Stuart Ledwich <email address hidden>wrote:

> Downloaded regular kernel update today which is 2.6.35-27 and this has
> fixed my problem, error mess has now gone when booting - thanks
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/642421
>
> Title:
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>
> Status in The Linux Kernel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Fix Released
> Status in “linux” source package in Maverick:
> Fix Released
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left
> with a black screen .. and it is impossible to work ...! thanks
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

[Denis Turrina (denis86pm-deactivatedaccount)]

Same here, regular kernel update fixes the bug.
Kind regards.

2011/3/5, Larry Whitehead <email address hidden>:
> I upgraded both of my computers with the regular kernel update last night
> and things are looking much better now with 2.6.35-27. Thanks and best
> regards.
>
>
> On Sat, Mar 5, 2011 at 8:49 AM, Stuart Ledwich
> <email address hidden>wrote:
>
>> Downloaded regular kernel update today which is 2.6.35-27 and this has
>> fixed my problem, error mess has now gone when booting - thanks
>>
>> --
>> You received this bug notification because you are a direct subscriber
>> of the bug.
>> https://bugs.launchpad.net/bugs/642421
>>
>> Title:
>> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>>
>> Status in The Linux Kernel:
>> Confirmed
>> Status in “linux” package in Ubuntu:
>> Fix Released
>> Status in “linux” source package in Maverick:
>> Fix Released
>>
>> Bug description:
>> When I turn on my computer I can read this on my screen
>>
>> "modprobe: Fatal: Could not load
>> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>>
>> After 4-5 seconds, plymouth starts and the system works fine.
>>
>> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
>> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>>
>> hi, I have the same problem, I asked the psw keychain and then left
>> with a black screen .. and it is impossible to work ...! thanks
>>
>> To unsubscribe from this bug, go to:
>> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>>
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/642421
>
> Title:
> Maverick could not load /lib/modules/2.6.35-22-generic/modules.dep
>
> Status in The Linux Kernel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Fix Released
> Status in “linux” source package in Maverick:
> Fix Released
>
> Bug description:
> When I turn on my computer I can read this on my screen
>
> "modprobe: Fatal: Could not load
> /lib/modules/2.6.35-22-generic/modules.dep: no such file or directory"
>
> After 4-5 seconds, plymouth starts and the system works fine.
>
> salve, ho lo stesso problema, mi chiede la psw del portachiavi e poi
> rimane con lo schermo nero..ed è impossibile lavorare...! grazie
>
> hi, I have the same problem, I asked the psw keychain and then left
> with a black screen .. and it is impossible to work ...! thanks
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/linux/+bug/642421/+subscribe
>

--
Denis Turrina
Dottore in Sicurezza dei Sistemi e delle Reti Informatiche
Computer systems and networks security B.Sc.

"Il computer non è una macchina intelligente che aiuta le persone
stupide, anzi è una macchina stupida che funziona solo nelle mani
delle persone intelligenti" U. Eco

"Computer is not an intelligent machine that helps stupid people, on
the contrary it is a stupid machine that works only in intelligent
people's hands" U. Eco

[matt_callaghan (matt-callaghan)]

Verified with 2.6.35.28.

NOTE: Due to the workarounds supplied above, I was no longer getting automatic kernel updates. Apparently, when the above workaround failed, and I reverted the manual installation of that kernel, I accidentally removed the wrapper packages to get the latest kernel (linux-image-generic linux-headers-generic)

{{{
callaghan@family:~$ dpkg -l 'linux-image*'

||/ Name Version Description
+++-========================-========================-================================================================
un linux-image <none> (no description available)
un linux-image-2.6 <none> (no description available)
ii linux-image-2.6.35-22-ge 2.6.35-22.35 Linux kernel image for version 2.6.35 on x86/x86_64
rc linux-image-2.6.35-25-ge 2.6.35-25.44 Linux kernel image for version 2.6.35 on x86/x86_64
un linux-image-generic 2.6.35.28.36 Generic Linux kernel image
}}}

{{{
callaghan@family:~$ dpkg --list linux-headers-*
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad)
||/ Name Version Description
+++-=================================-=================================-==================================================================================
un linux-headers-2.6 <none> (no description available)
un linux-headers-2.6-686 <none> (no description available)
un linux-headers-2.6-amd64 <none> (no description available)
ii linux-headers-2.6.35-22 2.6.35-22.35 Header files related to Linux kernel version 2.6.35
ii linux-headers-2.6.35-22-generic 2.6.35-22.35 Linux kernel headers for version 2.6.35 on x86/x86_64
un linux-headers-2.6.35-25 <none> (no description available)
un linux-headers-2.6.35-25-generic <none> (no description available)
un linux-headers-generic 2.6.35.28.36 Generic Linux kernel headers
}}}

In order to get automatic kernel updates again, I had to do:
{{{
sudo apt-get install linux-image-generic linux-headers-generic
}}}

Credit: trism on IRC (freenode, #ubuntu)

[Andrew Somerville (andy16666)]

I just wanted to point out that whatever fixed this bug could possibly be causing a random freeze-up I've been having during boot ever since I updated from 2.6.35-25. I've reported this issue as bug #771418, but I'm posting here as it may be related.

[Del Socorro Françoise (waterreedshimmer)]

modprobe -d *
FATAL: Could not load Bureau/lib/modules/3.11.0-19-generic/modules.dep: No such file or directory

uname -r
3.11.0-19-generic

sudo update-initramfs -c -k 3.11.0-19-generic
update-initramfs: Generating /boot/initrd.img-3.11.0-19-generic
Found in: http://www.linuxquestions.org/questions/debian-26/modprobe-fatal-could-not-load-lib-modules-modules-dep-335214/

Don't know how to put it on.

modinfo i915 | grep filename
filename: /lib/modules/3.11.0-19-generic/kernel/drivers/gpu/drm/i915/i915.ko

[Del Socorro Françoise (waterreedshimmer)]

Oh it has changed ?!

locate modules.dep
/lib/modules/3.11.0-15-generic/modules.dep
/lib/modules/3.11.0-15-generic/modules.dep.bin
/usr/share/man/man5/modules.dep.5.gz
/usr/share/man/man5/modules.dep.bin.5.gz

fairy@tales:/$ modprobe -d *
FATAL: Could not load bin/lib/modules/3.11.0-19-generic/modules.dep: No such file or directory

[Del Socorro Françoise (waterreedshimmer)]

I'm on Precise.

cd /lib/modules/3.11.0-19-generic/build/lib/modules/3.11.0-19-generic/
bash: cd: /lib/modules/3.11.0-19-generic/build/lib/modules/3.11.0-19-generic/: Aucun fichier ou dossier de ce type

cd /lib/modules/3.11.0-19-generic/build/lib/modules/
bash: cd: /lib/modules/3.11.0-19-generic/build/lib/modules/: Aucun fichier ou dossier de ce type

cd /lib/modules/
ls
cd ./3.11.0-19-generic/
cd ./build
cd ./lib
ls
fonts Kconfig Kconfig.debug Kconfig.kgdb Kconfig.kmemcheck lz4 lzo Makefile mpi raid6 reed_solomon xz zlib_deflate zlib_inflate

cd ./modules
bash: cd: ./modules: Aucun fichier ou dossier de ce type

Je ne comprends plus rien. Aucune idée de ce que je fais là:

fairy@tales:/lib/modules/3.11.0-19-generic/build/lib$ make
make: *** Pas de règle pour fabriquer la cible « /gen_crc32table », nécessaire pour « /crc32table.h ». Arrêt.

cd /lib/modules/3.11.0-15-generic/build/lib
ls
fonts Kconfig Kconfig.debug Kconfig.kgdb Kconfig.kmemcheck lz4 lzo Makefile mpi raid6 reed_solomon xz zlib_deflate zlib_inflate
fairy@tales:/lib/modules/3.11.0-15-generic/build/lib$ make
make: *** Pas de règle pour fabriquer la cible « /gen_crc32table », nécessaire pour « /crc32table.h ». Arrêt.

modprobe -d *
FATAL: Could not load fonts/lib/modules/3.11.0-19-generic/modules.dep: No such file or directory