On Sun, 6 Dec 2020 00:18:44 +0100 Arnd Bergmann wrote: > > This is for evrey clang build or just W=1+? Would be annoying if clang > > produced this on every build with 5.10 (we need to decide fix vs -next). > > The -Wsometimes-uninitialized is enabled unconditionally for clang, > but this only happens for IPv4-only configurations with IPv6 disabled, > so most real configurations should not observe it, but the fix should still > go into v5.10.
Great, I guessed correctly :)
