: Revised older backlog
cover : Revised older backlog
This is a revision of the earliest patches in the backlog,the more recent ones will be sent separately.
Patch 1: Implements (signed) modular 48-bit arithmetic.
Now the macros follow the upper-case convention.
Patch 2: Make dccp_delta_seqno return signed numbers - good for code reuse.
Patch 3: Make `before' relation unambiguous - this now mirrors the TCP solution.
Patch 4: Code reuse 1: implement `follows48' on top of dccp_delta_seqno
Patch 5: Implements the INC48() macro I was asking about -- for 48 bit fields where
dccp_inc_seqno can not be used.
Quoted 3 lines · Gerrit Renker
This is a revision of the earliest patches in the backlog,the more recent ones will be sent separately.
Can you tell us what was revised?
Signature · 3 lines
Web: http://wand.net.nz/~iam4 Blog: http://imcdnzl.blogspot.com WAND Network Research Group
Quoting Ian McDonald:
| On 1/4/07, Gerrit Renker <[email protected]> wrote: | > This is a revision of the earliest patches in the backlog,the | > more recent ones will be sent separately. | > | Can you tell us what was revised?
The breakdown of individual changes is as follows: (you can also use interdiff against the previously submitted patches, I state the links below):
Patch 1: "Macros for 48-bit sequence number arithmetic"
Previous: http://www.mail-archive.com/[email protected]/msg01158.html
CHANGE: use the convention that macros are upper-case (was not previously)
Patch 2: "Signed variant of dccp_delta_seqno"
Previous: http://www.mail-archive.com/[email protected]/msg01159.html
CHANGE: 1) Tossed out the unnecessary _add48/_sub48 macros (covered already)
2) Clarified the commit message (was not very clear before)
Patch 3: "Make `before' relation unambiguous"
Previous: http://www.mail-archive.com/[email protected]/msg01161.html
CHANGE: Removed the hunk which affected `max48' (since not really necessary)
=> Please note: The TCP variant of this new definition has been withdrawn,
on the grounds that old code depends on it. I still maintain that the new definition is useful and a better choice for DCCP.
Patch 4: "Fix for follows48"
Previous: http://www.mail-archive.com/[email protected]/msg01160.html
CHANGE: No code change, merely the commit message has been reformatted.
Patch 5: "Introduce macro for incrementing modulo-48"
=> This is a brand-new patch :), it has not been posted before.
I decided to lump it in with the others since it belongs to the sequence number arithmetic (and relates to some email questions I sent).
Gerrit
Quoted 10 lines · Gerrit Renker
Quoting Ian McDonald:
| On 1/4/07, Gerrit Renker <[email protected]> wrote: | > This is a revision of the earliest patches in the backlog,the | > more recent ones will be sent separately. | > | Can you tell us what was revised?The breakdown of individual changes is as follows: (you can also use interdiff against the previously submitted patches, I state the links below):
I'm OK with these then. Thanks for the clarification.
Signature · 3 lines
Web: http://wand.net.nz/~iam4 Blog: http://imcdnzl.blogspot.com WAND Network Research Group