Hi. I recently made some space on my old iBook G4 (1.2 GHz, late 2004 model) to put a few extra partitions, for a Linux, an OpenBSD 4.9, and a shared data partition using FAT32.
During the installation process, I briefly struggled with the disklabel tool before realizing that my problems came from the fact that it gave me an incorrect automatic layout to begin with: the root partition was starting one sector *ahead* of the OpenBSD partition itself. I fixed the layout manually, but the "boundstart" displayed by disklabel is still incorrect on my running system, as you can see from those outputs of pdisk and disklabel respectively: # pdisk wd0 Edit wd0 - Command (? for help): p Partition map (with 512 byte blocks) on 'wd0' #: type name length base ( size ) 1: Apple_partition_map Apple 63 @ 1 2: Apple_Bootstrap untitled 1954 @ 39935392 3: Apple_HFS Untitled 39673184 @ 262208 ( 18.9G) 4: Apple_UNIX_SVR2 LinuxRoot 10742188 @ 39937346 ( 5.1G) 5: Apple_UNIX_SVR2 swap 585938 @ 50679534 (286.1M) 6: Apple_UNIX_SVR2 Shared 2441407 @ 51265472 ( 1.2G) 7: Apple_Free Extra 262144 @ 64 (128.0M) 8: OpenBSD OpenBSD 4898241 @ 53706879 ( 2.3G) Device block size=512, Number of Blocks=58605120 (27.9G) DeviceType=0x0, DeviceId=0x0 Command (? for help): q # disklabel wd0 # /dev/rwd0c: type: ESDI disk: ESDI/IDE disk label: TOSHIBA MK3025GA duid: c2997bcd646b6636 flags: bytes/sector: 512 sectors/track: 63 tracks/cylinder: 16 sectors/cylinder: 1008 cylinders: 58140 total sectors: 58605120 boundstart: 53706878 boundend: 58605119 drivedata: 0 16 partitions: # size offset fstype [fsize bsize cpg] a: 4499968 53706880 4.2BSD 2048 16384 1 # / b: 398241 58206879 swap c: 58605120 0 unused i: 39673184 262208 HFS j: 2441407 51265472 MSDOS I read some documentation and source files to try to figure out where this problem comes from, and I stepped across a very suspicious line of code, in "src/sys/arch/macppc/macppc/disksubr.c" <http://www.openbsd.org/cgi-bin/cvsweb/src/sys/arch/macppc/macppc/disksubr.c?annotate=1.69>. Line 144: "hfspartoff = part->pmPyPartStart - LABELSECTOR;" And, if I'm not mistaken, LABELSECTOR is equal to 1 on MacPPC, and hfspartoff is used to initialize the boundstart 6 lines later... So coincidence or real bug ? :) -- Mathieu OLIVIER <moliv...@users.sourceforge.net>
