FS#33721 - [refind-efi] Kernel Hangs When Loaded with Refind Defaults
            Attached to Project:
            Arch Linux
            
Opened by Mark E. Lee (bluerider) - Tuesday, 05 February 2013, 01:02 GMT
Last edited by Tobias Powalowski (tpowa) - Wednesday, 13 February 2013, 15:57 GMT
          Opened by Mark E. Lee (bluerider) - Tuesday, 05 February 2013, 01:02 GMT
Last edited by Tobias Powalowski (tpowa) - Wednesday, 13 February 2013, 15:57 GMT
| 
 | Details
                    Description: When loading the arch linux kernel via defaults, refind hangs at starting vmlinuz. Strangely, when I load the kernels from the options menu (pressing +), the kernel loads normally. Additional info: linux 3.7.5-1 refind-efi 0.6.7-1 ### Refind_linux.conf ### "Boot with Defaults" "root=UUID=698a56e0-6683-49ea-94ab-0facd8153c9f rootflags=,relatime,data=ordered rootfstype=ext4 ro initrd=\EFI\arch\initramfs-linux.img add_efi_memmap" "Boot to Terminal" "root=UUID=698a56e0-6683-49ea-94ab-0facd8153c9f rootflags=,relatime,data=ordered rootfstype=ext4 ro initrd=\EFI\arch\initramfs-linux.img add_efi_memmap 3" Steps to reproduce: Set up refind to automatically load a kernel and have it automatically boot the kernel. | 
              This task depends upon
              
              
            
            
          
            Closed by  Tobias Powalowski (tpowa)
Wednesday, 13 February 2013, 15:57 GMT
Reason for closing: Fixed
          
          
        Wednesday, 13 February 2013, 15:57 GMT
Reason for closing: Fixed
 Comment by
            Mark E. Lee (bluerider) -
            Wednesday, 06 February 2013, 15:55 GMT
            
            Comment by
            Mark E. Lee (bluerider) -
            Wednesday, 06 February 2013, 15:55 GMT
          
           
          
              Downgrading to refind-efi 0.6.6-1 fixes the issue. There is
              something wrong in refind-efi 0.6.7-1.
            
           Comment by
            Mark E. Lee (bluerider) - Saturday,
            09 February 2013, 17:08 GMT
            
            Comment by
            Mark E. Lee (bluerider) - Saturday,
            09 February 2013, 17:08 GMT
          
           
          
              Upgrading to linux 3.7.6-1 and using refind-efi 0.6.7-1 also fixed
              the issue.
            
           Comment by
            Mark E. Lee (bluerider) - Saturday,
            09 February 2013, 17:09 GMT
            
            Comment by
            Mark E. Lee (bluerider) - Saturday,
            09 February 2013, 17:09 GMT
          
           
          
              I am requesting closure of the bug, since a kernel upgrade has
              fixed the problem.
            
           Comment by
            Mark E. Lee (bluerider) - Saturday,
            09 February 2013, 18:35 GMT
            
            Comment by
            Mark E. Lee (bluerider) - Saturday,
            09 February 2013, 18:35 GMT
          
           
          Oh wait, the problem has returned...
           Comment by
            Mark E. Lee (bluerider) - Sunday,
            10 February 2013, 15:40 GMT
            
            Comment by
            Mark E. Lee (bluerider) - Sunday,
            10 February 2013, 15:40 GMT
          
           
          
              I recompiled linux (using the ABS) and the problem has gone away
              (confirmed with several reboots over two days).
            
           Comment by
            Mark E. Lee (bluerider) - Tuesday,
            12 February 2013, 19:04 GMT
            
            Comment by
            Mark E. Lee (bluerider) - Tuesday,
            12 February 2013, 19:04 GMT
          
           
          
              A kernel upgrade to Linux 3.7.7-1 has introduced a new error.
              Systemd enters into recovery mode due to being unable to mount
              /boot/efi. Note, my linux-bfs installation does not have this
              problem. I would recompile the kernel for my system using ABS (but
              the ABS build scripts haven't updated to the new kernel yet). Is
              there any work being done on this particular error?
            
          