On Tue, 12 Aug 2014 21:53:57 -0400 TR Reardon thomas_reardon@hotmail.com wrote:
One other additional thought regarding partial read/write: the current logic all will fail to erase&write all-zero newcontent. Instead, if the reader bails (read_all_first == 0), the copy_old_content() logic should still read old included regions. In other words, it should do the opposite of what it currently does.
Thanks for your reports. It definitely sounds like something has gone wrong while rebasing... I'll look into the issue(s) in the next days.