http://bugzilla.opensuse.org/show_bug.cgi?id=1178241 Bug ID: 1178241 Summary: Building cilium on x86_64 fails after LLVM 11 update Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Kubic Assignee: kubic-bugs@opensuse.org Reporter: aaronpuchert@alice-dsl.net QA Contact: qa-bugs@suse.de Found By: --- Blocker: --- There is a new default warning, and cilium builds with -Werror: /home/abuild/rpmbuild/BUILD/go/src/github.com/cilium/cilium/bpf/lib/dbg.h:168:13: error: cast to smaller integer type '__u32' (aka 'unsigned int') from '__u32 *' (aka 'unsigned int *') [-Werror,-Wpointer-to-int-cast] .source = EVENT_SOURCE, ^~~~~~~~~~~~ bpf_lxc.c:21:22: note: expanded from macro 'EVENT_SOURCE' #define EVENT_SOURCE LXC_ID ^~~~~~ /home/abuild/rpmbuild/BUILD/go/src/github.com/cilium/cilium/bpf/lxc_config.h:28:16: note: expanded from macro 'LXC_ID' #define LXC_ID fetch_u32(LXC_ID) ^~~~~~~~~~~~~~~~~ /home/abuild/rpmbuild/BUILD/go/src/github.com/cilium/cilium/bpf/lib/utils.h:130:22: note: expanded from macro 'fetch_u32' #define fetch_u32(x) __fetch(x) ^~~~~~~~~~ /home/abuild/rpmbuild/BUILD/go/src/github.com/cilium/cilium/bpf/lib/utils.h:122:21: note: expanded from macro '__fetch' # define __fetch(x) (__u32)(&(x)) ^~~~~~~~~~~~~ There are more occurrences. The warning seems valid to me: pointers on x86_64 are 64-bit wide, so the conversion to a 32-bit integer is lossy. If this is intended, convert to uintptr_t first, then to __u32: # define __fetch(x) (__u32)(uintptr_t)(&(x)) That might require #include <stdint.h>. Generally only conversions to uintptr_t are guaranteed to have no information loss, i.e. you can convert them back to a pointer and get the original pointer back. LLVM 11 is currently staged in K, so you can see the full log here: https://build.opensuse.org/package/live_build_log/openSUSE:Factory:Staging:K... -- You are receiving this mail because: You are on the CC list for the bug.