Details
-
Bug
-
Status: Closed
-
Minor
-
Resolution: Fixed
-
3.6.2
Description
When compiled with -O3 and gcc-10 (which is the default for Ubuntu on ppc64el), compilation fails like this:
/bin/bash ./libtool -tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I./include -I./tests -I./generated -Wdate-time -D_FORTIFY_SOURCE=2 -Wall -Werror -g -O3 -fdebug-prefix-map=/<<PKGBUILDDIR>>=. -fstack-protector-strong -Wformat -Werror=format-security -MT zookeeper.lo -MD -MP -MF .deps/zookeeper.Tpo -c -o zookeeper.lo `test -f 'src/zookeeper.c' || echo './'`src/zookeeper.c libtool: compile: gcc -DHAVE_CONFIG_H -I. -I./include -I./tests -I./generated -Wdate-time -D_FORTIFY_SOURCE=2 -Wall -Werror -g -O3 -fdebug-prefix-map=/<<PKGBUILDDIR>>=. -fstack-protector-strong -Wformat -Werror=format-security -MT zookeeper.lo -MD -MP -MF .deps/zookeeper.Tpo -c src/zookeeper.c -fPIC -DPIC -o .libs/zookeeper.o src/zookeeper.c: In function 'free_completions': src/zookeeper.c:284:9: error: 'a_list.next' may be used uninitialized in this function [-Werror=maybe-uninitialized] 284 | tmp = a_list>next; | ~~~^~~~~~~~~~~~~ cc1: all warnings being treated as errors
What's happening here is that free_auth_completions is being inlined into free_completions, and this lets gcc see that members of a_list are being accessed without initialization. I don't know anything like enough about this code to see if this is a bug in code paths that are actually taken but at a glance it's certainly not obviously impossible: if the two if conditions at the top level of free_completions evaluate false, the function effectively looks like this:
void free_completions(zhandle_t *zh,int callCompletion,int reason) { auth_completion_list_t a_list; free_auth_completion(&a_list); }
so it's pretty clear that a_list is backed by uninitialized stack memory. Explicitly initializing the variable with "a_list = {NULL, NULL, NULL}" makes the warning go away.
