From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.9) Gecko/20020311 Description of problem: I have a disk with some bad sectors (probably many, from the sound the drive makes) on one partition. I had thought that perhaps that the formatter would mark those sectors as bad, and move on. Instead, anaconda crashed. I'm not too worried if anaconda can't use the disk (I've already replaced it) but the crash seems noteworthy. Version-Release number of selected component (if applicable): How reproducible: Didn't try Steps to Reproduce: 1. Get a disk with at least as many bad sectors as the one I've got (I'll attach a log to show the errors that had been echoed by anaconda up till the crash). 2. Set up an ext3 partition that covers those bad sectors. 3. Try to format/check for errors that partition. Actual Results: anaconda crashes, and pops up a dialog to save the crash trace (will be attached...) Expected Results: format either marks the sectors bad and anaconda continues if the adjusted partition size has room for the install, or anaconda gracefully reports a problem with the disk, and fails to install. Additional info:
Created attachment 53344 [details] anaconda dump from crash
We cannot correct errors on the fly. However we do need to allow the 'Check for bad sectors' test to create a map of the bad sectors which is used when creating the filesystem. This is a future enhancement.
I anticipate we're removing bad block check.