Re: [patch] ext2_fill_super breakage

[Nikita Danilov]

Alexander Viro writes:
 > 
 > 
 > On Thu, 28 Mar 2002, Nikita Danilov wrote:
 > 
 > > Explicit initialization always leaves room for some "pad" field inserted
 > > by compiler for alignment to be left with garbage. This is more than
 > > just annoyance when structure is something that will be written to the
 > > disk. Reiserfs had such problems.
 > 
 > If your structure will be written on disk you'd better have full control
 > over alignment - otherwise you are risking incompatibilities between
 > platforms and compiler versions.

Yes, but such experience frequently is only gained after format is
already carved in stone^Wdisk.

 > 
 > 

Nikita.