On Sat, Dec 24, 2016 at 11:17:18AM -0500, Geoff Lansberry wrote: > Mark - I'm sorry, but I did not write this code, and therefore was not > able to accurately describe it. It is fixing a different issue, not > the neard segfault that we are still chasing. Last week Jaret Cantu > sent a separate email explaining the purpose of the code, which had > you copied, did you see that? Hm, no, I didn't. I received an email from Justin Bronder but not from Jaret Cantu. Justin's email did help but is still pretty high-level. We need a clear understanding as to what is happening in the digital layer and the driver to know how execution is getting into a block of error handling code that should never be executed. Once we understand that we can start thinking about what the best fix is. > Does it explain why it was done to > your satisfaction? I've asked him to join in on the effort to push > the change upstream, however he will not be available until the new > year. I expect that it would help if he joins. After the holidays is fine - I think many people are taking it easy for the next week or so, anyway. > I know you did suggest that we split off that change from the others, > and if now is the time to do that, let me know. If you don't have > the email from Jaret, also please let me know and I will forward it to > you. I think it would help you if you split it off because the first two patches have a good chance of being accepted but this one doesn't (yet). If you separate the them, it will make it easier for Samuel to take the first two (or he may take the first two anyway but its always good to make it as easy maintainers as you can). Mark --
