Description of problem: The install process is started from memory stick. The target drive is attached as primary master on a PCI attached "RAID" adapter operated in plain vanilla mode. There are no other drives of any kind attached to the machine. Installation appears to proceed normally until after the language is selected and the licence accepted. It then stops with the error message "INTERNAL ERROR: unknown device hda". I preserved a number of log files on an external USB drive (identified as sda (Seagate) at the end of the syslog). The logs suggest that the sd_mod module is required, but not found. When the Rescue Mode option is selected instead of Install Mageia 3 then the existing partitions on hda are correctly mounted at /mnt. When Hardware Detection is selected from the install medium menu the machine performs an instant reset. It is not impossible that this problem may also be related to Bug 9446 Version-Release number of selected component (if applicable): Mageia 3 beta 3 DVD iso for x86_64 How reproducible: every time Steps to Reproduce: 1. see above 2. 3. Reproducible: Steps to Reproduce:
Created attachment 3635 [details] Grub stage1 log from the installation session
Created attachment 3636 [details] syslog from the install session
Created attachment 3637 [details] installation debug log including final error message
Attachment 3635 description: log from the installation session => Grub stage1 log from the installation session
Hi Richard, sorry that this bug report was overlooked. Can you still reproduce with Mageia 5 RC ?
Keywords: (none) => NEEDINFO
Sorry for the delay. I have been trying to refresh my memory about this one. I have a couple of PCI ATA cards and I sort of recall there was an issue connected with the availability of a driver for one of them (or both perhaps) to enable scsi emulation to be used with it. Plain ol' parallel ATA drivers were perhaps not determined to be needed by, or available to the install process so it went toes up. I will get the hardware together again and try it out with 5RC but for now I reckon we should put this bug report out of our misery. "OLD" would seem to be the most appropriate closure label.
Thanks, closing then.
Status: NEW => RESOLVEDResolution: (none) => OLD