Re: [PATCH] fs_ext2 sets min_size constraint in blocks, not sectors

[Andrew Clausen]

Matt Wilson wrote:
> 
> I'm not sure if there is code around that is depending on this broken
> behavior, but the fs_ext2 code was setting the min_size in blocks, not
> sectors.  Patch attached.

Thanks :-)

The code is used.  When you do ped_disk_set_partition_geom(),
you should pass the constraint from
ped_file_system_get_resize_constraint().
This will be particularly important, when I implement
ext2-resize-the-start, because ext2 file systems' blocks must remain
aligned the same way (i.e. block_start_sector % block_size == constant).

Anyway, when you do ped_disk_set_partition_geom(), it will bump the
partition size up, to satisfy the constraint (if necessary).  The
relevant function is ped_constraint_solve_nearest().  I should document
this...

Since it's under-estimating, there are no disasters ;-)

Andrew Clausen