lwip-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[lwip-devel] Re: [task #7040] Work on tcp_enqueue


From: Jakob Stoklund Olesen
Subject: [lwip-devel] Re: [task #7040] Work on tcp_enqueue
Date: Sun, 01 Feb 2009 16:32:20 +0100
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/23.0.60 (gnu/linux)

"address@hidden" <address@hidden> writes:

> Jakob Stoklund Olesen wrote:
>> Sure, but UDP performance is rarely critical, and it seems wasteful to
>> extend struct pbuf just for that.
>
> I wouldn't think so. If you want high throughput and can cope with
> lost segments, UDP is the better alternative to TCP and would still
> require high throughput!

You are right, I didn't think of the multimedia protocols.

> However, the goal of that patch (and its
> discussion) was to NOT include another member in struct pbuf but to
> solve it in a different way (which is why it is still open and not
> finished).

Right. Do you think my proposals for TCP are a detour from a unified
solution? I think the requirements for TCP and UDP/UDPLite are different
enough that little can be shared.

I would probably make these functions:

inet_chksum_pseudo_header(): inet_chksum_pseudo without payload.
inet_chksum_memcpy(): Checksumming memcpy.

They would be useful for UDP as well.

For TCP, a checksum field in the pbuf is not really that useful. A
data_csum field in tcp_seg might be, but it is not strictly necessary.





reply via email to

[Prev in Thread] Current Thread [Next in Thread]