rsync - Jun 2004 - [Bug 1467] Hash table generation seems to be flawed

https://bugzilla.samba.org/show_bug.cgi?id=1467





------- Additional Comments From ripper@internode.on.net  2004-06-19 02:18
-------
The way the code seems to work is as follows:

A sorted list of hashes and their block numbers are generated (targets). 
They're sorted so that identical hashes are numerically next to each other.

A reverse lookup table is then generated, that (i presume) is supposed to 
return the index of the first of these targets for any given hash. 
Unfortunately, this doesn't look like it will always be the case and so
certain
blocks from the 'generator' might be ignored and their contents
retransmitted.

-- 
Configure bugmail: https://bugzilla.samba.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug, or are watching the QA contact.

https://bugzilla.samba.org/show_bug.cgi?id=1467


wayned@samba.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID




------- Additional Comments From wayned@samba.org  2004-06-19 10:48 -------
You'll note that the loop is running in reverse, with "i" going
from high to
low. This means that "i < tag_table[targets[i].t]" will always be
true for tags
that are not NULL_TAG (since all the values came from prior iterations of the
loop).  So, I believe that the if you added can never be false.

If you disagree with this analysis, please show me the mistake in my reasoning.

-- 
Configure bugmail: https://bugzilla.samba.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug, or are watching the QA contact.