no buffer space available

[Hussam Al-Tayeb]

Binary package hint: udev

after upgrading from dapper to edgy, the computer hangs at bootup at loading hardware drivers.
It spits out this error: "error receiving uevent message : no buffer space available"
it hangs at this stage for a minute then it continues booting.

[Scott James Remnant (Canonical) (canonical-scott)]

Please provide a lot more information, what type of computer is this? What architecture is it running?

Copies of lspci, lsusb, dmidecode, etc. output would be useful
Also a copy of /var/log/udev and /var/log/dmesg

[Matthew Wardrop (mister.wardrop)]

I too receive this error - although not during boot up. It occurs for me when I attempt to use the internet, and after about a second or so, it cuts out. Ping manages to get through, but anything more demanding fails. Once it fails - NOTHING gets through. This has stopped my upgrading to 2.6.17 kernel in Ubuntu, although I am using the latest in everything else. I might also comment the nVidia drivers also fail to run before AND after this occurs.

The actual error message on the command line is (or something similar, as I attempt to read my scrawled writing):
[ 437.549280] sr 0:0:3:0: rejected i/o to offline device

I will restart my computer now, in order that I might send you the /var/log/dmesg and /var/log/udev files.

Kind Regards,
Matthew

[Simon Hepburn (sth)]

Following up from #60389

[Scott James Remnant (Canonical) (canonical-scott)]

I assume you still have this problem with the current edgy kernel?

[Hussam Al-Tayeb (hussam)]

I don't see it anymore but I don't know what caused it. It does however seem to have been fixed so you might want to close it.

[Scott James Remnant (Canonical) (canonical-scott)]

It was most likely a kernel problem. I've not had any recent reports of this, so will close it unless it appears again.

[Philipp Schlesinger (philipp-sadleder)]

I still see this problem like described above on an Acer TravelMate 4233.
If I press the "Radio Frequency Kill Switch" directly after grub, the problem doesn't occur, otherwise the boot process hangs for a minute and I see something like:

Dec 19 15:43:42 rcS: * Activating swapfile swap... [ ok ]
Dec 19 15:44:47 rcS: * Configuring network interfaces... [ ok ]

[Philipp Schlesinger (philipp-sadleder)]

Sorry, I forgot to mention that this is a ipw3945 device.
lspci gives:

05:00.0 Network controller: Intel Corporation PRO/Wireless 3945ABG Network Connection (rev 02)

[Philipp Schlesinger (philipp-sadleder)]

Is there any further input I can give to help fix this bug?

[Philipp Schlesinger (philipp-sadleder)]

I just want to state that this bug is NOT fixed for me, could you please reopen it?

[Bryce Harrington (bryce)]

Ran into this one while testing Tribe1 installation CD's.
On the same system I installed using 6 different media, and spotted this error message once.

ok - x86 Server
ok - x86 Alternate
ok - x86 Desktop
ok - AMD64 Server
ok - AMD64 Alternate
fail - AMD64 Desktop

Attached is a photo of what it printed (sorry so blurry). The system failed to boot into the LiveCD environment on the AMD64 desktop (I tried it twice, same results.) All the others installed without trouble.

[mobiusNZ (mobiusnz)]

I also have this problem, quite randomly it seems. I get up to "Loading hardware drivers..." and sometimes it will hang, with the message "Error receiving uevent message: no buffer space available". At this point I have to hold down the power button on my laptop and try again.

Other times it boots up fine... very odd. I also have to include kernel switches "noapic irqpoll" for anything to happen at all.
Laptop is a Compaq presario V6302AU (V6000), AMD64 architecture, and it has a broadcom wireless card.

Running Kubuntu with Ubuntu kernel 2.6.20-16-generic.

Anyone have any ideas? I *think* it may be related to the kernel switches, and the kernel not playing nicely with my bios, or something :S

[Alex F (frase)]

I get this message during boot too, but it doesn't cause booting to hang and doesn't seem to happen every time. I'm not even sure what the message means or if its anything to worry about; there are still a few quirks with my hardware setup that I'm (slowly) working on, but I don't know if this is related to any of them.

Feisty 7.04, 2.6.20-16-generic (pci=noacpi noacpi acpi=off)
AMD X2 6000+
ASUS M2N32-SLI Deluxe WiFi

[Hampton (hcatlin)]

I believe this to be two bugs. The first one from edgy. But, this is a new bug.

I can report that with Feisty 64-bit Desktop, this bug happens.

When I am using the live-cd or booting after the alternate install, it will hang with the "error receiving uevent message : no buffer space available" message.

I am re-opening this bug. If the maintainer wants to open a second ticket for it, feel free.

-hampton.

[Hampton (hcatlin)]

This is appearing again.

[ccinope (csinuous)]

in 7.10 x86-64 tribe 4 i see this at every boot.

[darryn.smith (darryn-smith)]

7.04 -- Kernel 2.6.20-16-generic
HP Laptop - DV9417ca - 2GB Ram - AMD X2 Turion 64 TL-56 - Nvidia 6150 video

Added pci=conf1 things started working correctly (video etc) several reboots later though laptop hangs with "no buffer space available" message.

It keeps displaying this error. THEN strangely after half a dozen or so reboots it boots up perfectly.

[Tim Post (tinkertim)]

I'm getting the same thing in a custom 2.6.20.3 kernel that I built from kernel.org , dmesg isn't showing anything (at all) out of the ordinary.

Same thing with older / newer, from 2.6.18 to latest 2.6.22, 32/32P and x64. These are all servers with no peripherals, so I see this only once and boot continues normally.

This appears to be a kernel problem, udev is talking and the kernel went deaf. I can't be _sure_ about that, but its a pretty good hunch. I'm going to take out all kernel debug, fs debug (anything noisy) in the 2.6.20.3 kernel and see if I can make this go away, that should say one way or another.

[Tim Post (tinkertim)]

Bah, I get the same issue with Xen kernels (2.6.18) from xen-unstable. Kernel debug is off, there is plenty of space yet the Kernel says 'no' when udev talks. This is a Kernel issue, pls close it.

[Tim Post (tinkertim)]

After taking out (a ton) of debug stuff, I was able to see the errors, it was secure_unlink() complianing that certain files do not exist.

This is a kernel bug folks, please close it.

[Scott James Remnant (Canonical) (canonical-scott)]

Closed at request of current reporter