Jan, I am using gmail from Windows. Please use the attached diffs if gmail mangles stuff. BTW: here gmail *is* the company mail server. I have re-attached a slight change.
The "stray" comments help when quickly grepping for things. The stats *do not* map onto anything RTNET provides. They are just stats that instrument low rtskb conditions and help in debugging when one scratches his head and asks "WTF is going on?". I think the reason for using a rtskb pool is that the underlying driver expects the packets to be transmitted to come from such a pool. I may be wrong. I do not intend to refactor the code further as of now. The only bit of refactoring has to do with my resource allocation/freeing paranoia. I will keep in mind your comments re: skb pool replacement. A good practice is that if you don't have to do anything in function then exit from it as soon as possible, e.g. when the xmit function gets a NULL rtskb then it has nothing to do, *regardless of a NULL skb being a bug*. C99 exists for a reason. Keeping the variable close to where you use them [and not at the top of the function] is good practice. I am not open to comments re: code formatting/style. I am getting enough grief from a Hungarian-notation aficionado at my other place of work. You may have a more successful project/community if casual contributors don't have to learn your project-specific coding rules. For the time being this is the patch and the time I have allocated for rtnetproxy. Splitting into 5 patches is just *make-work*. My patch fixes things, *works* and corrects a problem that has been around for ages. We have done our GPL bit and did enough due diligence to have it accepted. Regards, Vlad
rtnetproxy.c.diff
Description: Binary data
rtnetproxy.c
Description: Binary data
------------------------------------------------------------------------------
_______________________________________________ RTnet-users mailing list RTnet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/rtnet-users
