Hide Forgot
Created attachment 495769 [details] output of dmesg Description of problem: I have Fedora 15 Beta installed. With clean install (Desktop live image), the boot went smoothly. There were (and still are) some errors like bug 699113 and bug 698674, but it all went quite well. After updating, I started experiencing one or sometimes two hangups during boot. At some point, the boot process suddenly ceases. The computer sits completely still. Unitill i press a key. Then it resumes the boot and continues normally. Sometimes this repeats, i hit the button and all is well again, but usually it happens only once. I was advised on IRC to delete quiet and rghb parameters form kernel line and try booting with that. In this run, computer hung up only once, just after spewing out line [ 14.191168] Bluetooth: SCO socket layer initialized. It can be found in attached output of dmesg (line 1209). Hitting the up arrow on keyboard resumed the boot and it went on uneventfully. I have also cell phone video of the boot, in case you have nothing else to watch: http://youtu.be/tZ1-yENubEk Hope it is the right file because I still do not have MPEG4 AVC codecs or Adobe Flash installed :D Sometimes hitting key is not enough and the boot is resumed only after hitting the power off button on my box, but that is rare. At IRC, one person pointed out it is clearly somehow related to cgroups. My /cgroup directory is completely empty. ls -alF /sys/fs/cgroup gives: total 0 drwxr-xr-x. 12 root root 240 Apr 29 2011 ./ drwxr-xr-x. 5 root root 0 Apr 29 2011 ../ drwxr-xr-x. 2 root root 0 Apr 29 2011 blkio/ drwxr-xr-x. 3 root root 0 Apr 29 2011 cpu/ drwxr-xr-x. 2 root root 0 Apr 29 2011 cpuacct/ drwxr-xr-x. 2 root root 0 Apr 29 2011 cpuset/ drwxr-xr-x. 2 root root 0 Apr 29 2011 devices/ drwxr-xr-x. 2 root root 0 Apr 29 2011 freezer/ drwxr-xr-x. 2 root root 0 Apr 29 2011 memory/ drwxr-xr-x. 2 root root 0 Apr 29 2011 net_cls/ drwxr-xr-x. 2 root root 0 Apr 29 2011 ns/ drwxr-xr-x. 4 root root 0 Apr 29 2011 systemd/ it was also suggested to me that bug 668849 can have to do something with this. My computer is desktop, but has only USB wi-fi dongle. Network cable is not connected.
Created attachment 495770 [details] /var/log/boot.log
One more thing. These hangups also appeared when I tried to update my Fedora 15 Beta instalation USB. I did not report it, because I read afterwards, that updating live USB is not supported and is an easy way to hell. So it has to be unrelated to the kernel version, because kernel is not updated when runnig live USB
Yet another thing. The bug I wrote could be related is a mistake. I was pointed not to that bug, but to page that is the search result for bugs related to cgroup, sorry
Sounds like https://bugzilla.kernel.org/show_bug.cgi?id=15289 You can try disabling AMD C1E support in then BIOS as a workaround.
Yes, that is it. I tried your fix and some other mentioned in the kernel bugzilla Disabling AMD C1E in BIOS will fix it for me. My standard kernel line is "kernel /boot/vmlinuz-2.6.38.3-18.fc15.x86_64 ro root=UUID=4b152f7e-0493-40ae-b1e9-2dc41e73b6e7 rd_NO_LUKS rd_NO_LVM rd_NO_MD rd_NO_DM LANG=en_US.UTF-8 SYSFONT=latarcyrheb-sun16 KEYTABLE=cz-lat2" + " rhgb quiet" booting with AMD C1E on and with various suggested parameters kernel line + " debug" will hang up several times, but not in kernel, like hpet and irqpool below kernel line + " debug nomsi" will hang up twice, but it will recover itself without pressing any keys. But this also hung up during shutdown and that need the helping key in the end kernel + " debug maxcpus=1" will work without error kernel + " acpi_skip_timer_override" will work without error kernel + " hpet=disable" & kernel + " irqpool" will hang up in kernel, after initializing USBs, in which moment key is not enough and power button has to be pushed. Then hangs up few more times which gets corrected by key kernel + " idle=mwait" works fine, no hang ups kernel + " ioacpi" hangs up My HW profile is at smolt.org as f8c35d19-a61d-4f3c-83ba-ca00bed65130, I have CPU0: AMD Phenom(tm) II X4 905e Processor stepping 02 and Gigabyte 790XTA-UD4 at kernel bugzilla, there is link to bug 648837 which is filled against Fedora 14 Beta. So should that bug be updated to include Fedora 15 and this bug be marked as a duplicate?
This is not due to having C1E enabled if it worked before. Did it start with the 2.6.38.4 kernel? There was a bug introduced in that one that has since been fixed in the pending 2.6.38.6 update. Look at the "flags" line in /proc/cpuinfo and see if it contains the string "arat". If it is present, boot an older kernel and see if that goes away.
kernel-2.6.38.6-26.rc1.fc15 has been submitted as an update for Fedora 15. https://admin.fedoraproject.org/updates/kernel-2.6.38.6-26.rc1.fc15
Please test the update and report whether it fixes this bug.
kernel-2.6.38.6-26.rc1.fc15 has been pushed to the Fedora 15 stable repository. If problems still persist, please make note of it in this bug report.
I have updated to [jirka@private ~]$ uname -a Linux private 2.6.38.6-27.fc15.x86_64 #1 SMP Sun May 15 17:23:28 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux and the problem still persits. Strangely, when I rebooted from 2.6.38.5 after installing the update, the first boot went well. But when I tried three more times, it got stuck at least once each time.
(In reply to comment #10) > I have updated to > > [jirka@private ~]$ uname -a > Linux private 2.6.38.6-27.fc15.x86_64 #1 SMP Sun May 15 17:23:28 UTC 2011 > x86_64 x86_64 x86_64 GNU/Linux > > and the problem still persits. > > Strangely, when I rebooted from 2.6.38.5 after installing the update, the first > boot went well. But when I tried three more times, it got stuck at least once > each time. Please attach the contents of /proc/cpuinfo after booting the latest update. (You can keep it from stalling by moving the mouse pointer around and/or hitting the shift key repeatedly.)
Created attachment 500171 [details] /proc/cpuinfo on 2.6.38.6-27.fc15.x86_64
Should be fixed in : https://admin.fedoraproject.org/updates/kernel-2.6.38.8-32.fc15
I just installed Fedora 15 on my wife's Sony Viao and now the boot process hangs at right after "Started SYSV: Emable monthly update of Smolt." Switching to a terminal window and performing "yum -y update smolt" does not fix the problem. During install, the GUI was trashed so after switching to a terminal I installed RPM Fusion repos and the Kmod nvidia drivers. The next boot and each subsequent boot stops as above. Is there hope to rescue this installation or must I re-install Fedora 14?
(In reply to comment #14) > I just installed Fedora 15 on my wife's Sony Viao and now the boot process > hangs at right after "Started SYSV: Emable monthly update of Smolt." Switching > to a terminal window and performing "yum -y update smolt" does not fix the > problem. > And does it continue if you keep hitting the "shift' key? Does that machine have an AMD processor in it?