Hello, On Wed, Sep 30, 2020 at 07:34:35PM +0300, Jouni Roivas wrote: > > So, I'm not necessarily against the change, mostly in the spirit of "why > > not?". > > There's actual user space application failing because of this. Of course > can to fix the app, but think it's better to fix kernel as well. At > least prevents possible similar failures in future. Just for the record, none of these pseudo interface files are expected to behave like real files. The supported usage pattern is - open, read with sufficentily large buffer till EOF and then close, or open, write the content in one go and close. There are some exceptions and we add convenience features when the cost is low enough but beyond the core usage pattern the boundary between what's supported and not is pretty mushy. Thanks. -- tejun
