Hi Ying,

Yes, that makes sense.
I have updated the patches with your suggestion and am going to send them now.

As for patch #2, I guess you're referring to creating a preliminary node.
Let's consider it as separated from the tipc crypto as well. In fact, we had a 
boolean flag to control it, also in addition to the new kernel option, if the 
tipc crypto is disabled, it will be no meaningful at all but just fallback to 
the original one, so that the node must work exactly the same as usual (i.e. 
preliminary = false)... I think we can keep it without the "ifdef else" 
instructions, otherwise we might need to duplicate the code for two cases. Do 
you agree?

BR/Tuong

-----Original Message-----
From: Ying Xue <ying....@windriver.com> 
Sent: Tuesday, November 5, 2019 3:02 PM
To: Jon Maloy <jon.ma...@ericsson.com>; Tuong Tong Lien 
<tuong.t.l...@dektech.com.au>; tipc-discussion@lists.sourceforge.net; 
ma...@donjonn.com
Subject: Re: [PATCH RFC 0/5] TIPC encryption

On 11/5/19 5:33 AM, Jon Maloy wrote:
> Tuong, Ying
> I am ok with a kernel option, as long as it is enabled by default. I can 
> imagine smaller embedded systems where the deployer want a small module, and 
> encryption anyway is managed differently, or not at all.
> 
> ///jon

When I gave the suggestion to add a new kernel option, I also ever
figured out how to reach this goal based on this series of patches.

In my opinion, we don't need to modify the patches of making some
preparation things. For example, we don't need to change patch #1 at all.

Patch #3, #4 and #5 are almost completely separate with current code.
The only thing is that we need to consider how to modify patch #2. In
sum, it looks like it's not very difficult to introduce the new kernel
configuration option.

The kernel option not only can help us keep TIPC module size smaller,
but also can help us maintain it easily particularly with TIPC becoming
more and more complex. Lastly, probably it can help to make it possible
that the series of patches can be easily accepted by upstream if user
can disable the feature with a kernel option.




_______________________________________________
tipc-discussion mailing list
tipc-discussion@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/tipc-discussion

Reply via email to