2009/387 [Pathname Reparse Points]

[Afshin Salek]

Garrett D'Amore wrote:
> Robert Thurlow wrote:
>> Garrett D'Amore wrote:
>>
>>> I'm confused.  Brian says that archivers Just Work with the current 
>>> form, because the attributes are retained.  Yet, you're saying that 
>>> the attributes are not necessarily retained.  Which is it?  Right 
>>> now, either way, you have an attribute... which I *think* means that 
>>> the you need support (which may or may not be present) in the archivers.
>>
>> I know that some archivers in use today will not create a different
>> bitstream for a reparse point than for a regular symlink, but I
>> believe our intent to set the bit on restore or create means that
>> we don't have to teach all archivers about the new bit.  If any
>> Sun-maintained archivers are aware, that's fine, but this should
>> work with GNU tar, too.
> 
> Okay, that's an important detail that I missed somewhere.  (Was it in 
> the original spec?)
> 

Yes, it is:

      fop_symlink() needs to be modified to recognize the reparse
      @{REPARSE} tag and pass the appropriate attribute (i.e. reparse
      system attribute) to VOP_SYMLINK to be set on the symlink.

Afshin

>>
>> Logically, that bit is only a way for the NFS/CIFS server
>> code to avoid having to read the symlink data to understand
>> that it's touched a reparse point.  We want this because the
>> referral is a two-part deal - we return an "it's not here"
>> error to the client first, and then respond to a followup
>> query later.  The sysattr makes an NFSv4 READDIR op able to
>> avoid N accesses to symlink data.
> 
> OK.
> 
>    -- Garrett
>>
>> Rob T
>