On Jun 21, 2016, at 12:32 AM, David Holland wrote:

> 
> I am becoming fairly strongly inclined to enter the 64-bit inode
> number into the lfs filehandle structure as two 32-bit values instead
> of one 64-bit value, because this way the size will be 28 (it was 24
> before -- I don't know where the 28 originally came from) and none of
> the above rubbish is needed.
> 
> This would be kind of a hack, but it seems like the best approach for
> the time being... anyone want to comment?

David,

It looks like you could do the split-field just in struct lfid; that way
the rubbishy parts are confined to only lfs_fhtovp() and lfs_fhtovp().

It seems it would be better if struct lfid were not in a header file at all.

- Michael

Reply via email to