On Fri, 22 Mar 2019 15:39:11 +0000 Slavomir Kaslev <kaslevs@xxxxxxxxxx> wrote: > You're completely right. I still think that it makes sense to have > that patch go in, maybe with rewritten commit message. > In tracecmd_msg_read_data()[1] we do pointer arithmetic on void > pointer which is undefined behavior (though gcc allows it unless > -pedantic-errors is passed) and this patch will take care of it. > > Should I resend or drop it? I think the patch cleans up the code (removes a lot of the type casting). Resend it with a new change log. -- Steve
|