I'm reporting this under the kernel for reasons explained below. Attempts to burn an ISO with k3b fail with a message saying the first sectors can't be written. I've tried this now with a dozen or more blanks and 3 different drives on 2 different cauldron machines at the same level. Brasero fails as well, but doesn't give much in the way of detail. The resulting disks are not recognized by Plasma or dd as present media. Figuring that these probably all use wodim/cdrecord and that something was wrong with that, I brought up a Windows 7 VBox VM, attached a drive to that, copied the ISO to the VM, and tried burning from Windows, which I assume would have nothing between Windows burning code and the drive other than the kernel. That failed too. XBurn *thinks* it's working, and its progress bar goes all the way to 100%, but produces the same kind of coaster. On one of these tries (XBurn) I tried to verify the burn using "cmp -l" against the ISO file. The result I got was "EOF on /dev/sr2 after 2048 bytes". Is it possible that a kernel bug is writing every sector to sector 0 of the disk ? Can anyone else reproduce this ?
(In reply to Frank Griffin from comment #0) > > Is it possible that a kernel bug is writing every sector to sector 0 of the > disk ? > > Can anyone else reproduce this ? Don't have time to try, but will already assign this bug to the kernel & drivers maintainers. Did you try with a different kernel (version, or -linus, or...)
Assignee: bugsquad => kernelCC: (none) => marja11
I've since gotten some good burns with a USB drive on the second machine I tried, so I'm going to chalk the bad ones there up to bad blanks. The same drive gets immediate errors writing sector 0 on the original machine under both the current cauldron kernel and an old 4.2.2, so this has to be hardware-related, although how a single hardware error could accout for errors on both an internal SATA burner and an external USB burner mystifies me.
Status: NEW => RESOLVEDResolution: (none) => INVALID