http://bugzilla.novell.com/show_bug.cgi?id=617945
http://bugzilla.novell.com/show_bug.cgi?id=617945#c3
Petr Uzel
Unfortunately parted-2.3 didn't make it into openSUSE-11.3 ....
parted-2.3 does not differ much from parted-2.2 wrt to partition alignment. IOW, I'm almost sure it suffers from this issue as well. This is what happens in this particular case: - the extended partition starts at sector 12032000 - user wants logical partition to start at cylinder 748, which is sector 12016620 - parted's algorithm to create the range of sectors, where the partition can start, is this: - centered on the sector provided by user - twice as big as the unit used by the user - this gives [12000556,12032684] - as a next constraint, the logical partition can not start before the extended; this gives this range of sectors: [12032001, 12032684] - the device does not provide information about optimum alignment size, so parted defaults to 1 MiB - none of the sectors from the above range is on MiB boundary => BANG Parted's algorithm clearly isn't optimal, but changing it so late in 11.3 RC phase isn't a good idea IMHO. Arvin, would you agree to leave it as is for 11.3 and address this issue in upstream? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.