On 21.01.2026 14:14, Daniel P. Smith wrote: > Apologies, I've been on travel for the last two weeks and I wasn't > comfortable acking this with just a read of the diff. The thing that bothers > me that I want to understand better is why only after the else does it worry > about null terminated. Additionally, stepping back, a casual reader of the > code is going to wonder why only after some reads into the buffer does it > need a null while others do not.
I'm curious to know of an example or two which you refer to here, as ... > I think most people would find that as a red flag that an underlying issue is > getting papers papered over. I will be back from travel this weekend and I > will sit down and review with more context. > > V/r, > DPS > > On January 19, 2026 8:50:02 AM CST, Jan Beulich <[email protected]> wrote: >> Daniel, >> >> On 08.01.2026 10:18, Jan Beulich wrote: >>> Gcc's "threading" of conditionals can lead to undue warnings, as reported >>> in e.g. https://gcc.gnu.org/bugzilla/show_bug.cgi?id=116519 (no matter >>> that the overall situation is different there). While my gcc15 complains >>> ("buf[2] may be used uninitialized in this function") about only two of >>> the three instances (not about the one in type_read()), adjust all three >>> to be on the safe side. ... I've already extended the change to cover all three similar patterns, no matter that only two triggered a warning. Jan
