Hi list,

I've commited the attached patch, which (should) get ride of the infamous
" Assertion (metadata_length > 0) at disk_dos.c in function
add_logical_part_metadata() failed " bug (at least for partitions schemes
generated by Parted...).

While I was writing it I discovered that the previous code was really
unsafe and could produce logical partitions overlapping their partition
table, which can lead to data corruption. This wasn't happening often
before 1.6.23 (but this could definitly happen) because the partitions
were aligned most of the time, in which case the bug doesn't occur.
But now, when using MB units to write the start / end of a partition on
a modern HD with cylinder size == 16065 sectors, partitions are very
often unaligned.

Cheers,
Guillaume Knispel


_______________________________________________
Bug-parted mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/bug-parted

Reply via email to