Rink Springer wrote: > On Mon, Mar 10, 2008 at 03:46:56PM +0200, Ian FREISLICH wrote: > > Rink Springer wrote: > > > Hi, > > > > > > On Mon, Mar 10, 2008 at 02:44:26PM +0200, Ian FREISLICH wrote: > > > > Yet, sysinstall has the size as 1255914653 sectors. > > > > > > Could you paste the output of 'sysctl -b kern.geom.conftxt' ? > > > > 0 DISK aacd0 639830589440 512 hd 255 sc 63 > > This is very odd.. the big number in the above line is the size in > bytes (len), and the 512 is the number of bytes per sector (s). So, the > size in sectors is simply len / s = 639830589440 / 512 = 1249669120, > what the dmesg claims. This is the same as the calculation made by > sysinstall - look at lib/libdisk/open_disk.c:Int_Open_Disk(). > > I don't see any obvious overflow errors there... everything seems to use > daddr_t types, which are 64 bit (at least on i386), so that should not > be a problem. I forgot to state that this was amd64. The thing is, that the installer actually gets the right number - 1249664157: Disk: aacd0 Partition name: aacd0s1 Free: 1249664157 blocks (595GB) This is how the disk is configured after my manual change using 'disklabel -e' to /var: Part Mount Size Newfs Part Mount Size Newfs ---- ----- ---- ----- ---- ----- ---- ----- aacd0s1a <none> 512MB * aacd0s1b swap 4056MB SWAP aacd0s1e <none> 512MB * aacd0s1d <none> 10240MB * aacd0s1f <none> 577GB * If I delete these partitions and manually enter the sizes in the installer, an extra (nonexistent) 3GB appears: Part Mount Size Newfs Part Mount Size Newfs ---- ----- ---- ----- ---- ----- ---- ----- aacd0s1a / 512MB UFS2 Y aacd0s1b swap 4056MB SWAP aacd0s1d /tmp 512MB UFS2+S Y aacd0s1e /usr 10240MB UFS2+S Y aacd0s1f /var 580GB UFS2+S Y > Could you perhaps add debugging printf()'s to Int_Open_Disk() to try to > trace what is going on? I'll see what I can do. Ian -- Ian FreislichReceived on Tue Mar 11 2008 - 07:10:12 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:28 UTC