Sorry, lost the previous mail so can't continue the thread. I took a look at the sources as cgf suggested and have the following question:
in fhandler_proc.cc revision 1.36, in function format_proc_partitions() : after getting the drive geometry with 'IOCTL_DISK_GET_DRIVE_GEOMETRY' the size of the disk is calculated in bytes and this number is right-shifted 6 bits, so that's a division by 64 and that maps with the scaling difference of 16 between df and /proc/partitions. likewise after getting the drive's partition layout with 'IOCTL_DISK_GET_DRIVE_LAYOUT' the length of each parititions (which is in bytes) is again right-shifted 6 bits. So why the right-shift 6 instead of 10 which would map with a blocksize of 1K? If this is an obvious or trivial thing or if this has been addressed on the mailing list before or on some website: I apologise. I'm neither a windows nor a cygwin developer and I have little time but this just struck me as a bit weird. Erik. PS: if the layout of this mail looks weird I blame it on Lotus Notes which I am forced to use here at work. :( -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
