Description of problem: When booting, from time to time (30%), after choosing the system to boot (grub), the kernel freezes, the screen is void. Typing ctrl-alt unfreezes the kernel, the tree cauldrons are displayed (Athlon II X3) and the system boots normally. Version-Release number of selected component (if applicable): kernel-2.6.38.3-1.mga1 (but it also occurred on previous 2.6.38 kernels) How reproducible: Occurs 30% of boot Steps to Reproduce: 1. just boot or reboot 2. the kernel freezes 3. type ctrl alt, and the boot continues normally
This one should have been fixed in kernel-2.6.38.5-0.rc1.1.mga1 Can you confirm that ?
CC: (none) => tmbAssignee: bugsquad => tmb
I rebooted several times with kernel 2.6.38.5-server-0.rc1.1.mga without reproducing the freeze. It effectively seems to be fixed. Thanks.
Status: NEW => RESOLVEDResolution: (none) => FIXED
It occurred again with kernel 2.6.38.6-server-1.mga
Status: RESOLVED => REOPENEDResolution: FIXED => (none)
It occurred again with kernel 2.6.38.6-desktop-2.mga. This only happen for i586 kernel, but it working fine with x86_64 kernel.
CC: (none) => pham182b
It occurred also with kernel 2.6.38.7-server-1.mga
so the only working kernel for you is kernel 2.6.38.5-server-0.rc1.1.mga ?
Yes
I can boot with the others, but from time to time, I have to press <ctrl> <alt> for the kernel to start (I found no other working key combination).
Can you test kernel-2.6.38.8-1.1.mga1 from: http://tmb.mine.nu/Mageia/1/bugs/901/ http://tmb2.mine.nu/Mageia/1/bugs/901/
I tested it. I rebooted 10 times without seeing any freeze. Do you know what can be done to increase the probability to be in the conditions where the bug may occur?
It was amd that screwed up both an errata check based on hw revisions, and also with srat enabling based on hw revisions. Theese are now fixed in upstream 2.6.38.8 (wich you tested) and any more changes to the errata and srat checks will not be backported to -stable tree, so you should be safe for Mageia 1. They might try to fix the errata/srat checks in upstream kernels with more intrusive code changes, so there is a slight possibility the issue might re-appear in newer kernels if they dont get it right. But as it seems now, you should be safe
Unfortunately not, it occurred again (AMD Athlon II X3 425 on Gigabyte MA770T-UD3P).
It also occurs with kernel-2.6.38.8-server-2.mga2
I tried 3.0.0-server-1.mga2 in Cauldron and I never got this trouble. I also installed it on Mageia 1 and I never got this trouble.
(In reply to comment #14) > I tried 3.0.0-server-1.mga2 in Cauldron and I never got this trouble. > > I also installed it on Mageia 1 and I never got this trouble. Can this bug be closed, then?
CC: (none) => shlomif
Yes, of course
Status: REOPENED => RESOLVEDResolution: (none) => FIXED