[Bug server/26839] Systemtap build failures with clang

["tbaeder at redhat dot com" <sourceware-bugzilla@sourceware.org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=26839

--- Comment #6 from Timm Bäder <tbaeder at redhat dot com> ---
Adding the GCC diagnostic pragmas to the printf-like functions was wrong btw
and just adding the appropriate __attribute__((format(printf, x, y)))
attributes is better, fixes the problems with clang and gives us better error
reporting.

It also shows a few problems with the current un-annotated functions like this
one in staprun/relay.c:

--- a/staprun/relay.c
+++ b/staprun/relay.c
@@ -272,7 +272,7 @@ static void switchfile_handler(int sig)
                pthread_mutex_lock(&mutex[avail_cpus[i]]);
                if (reader[avail_cpus[i]] && switch_file[avail_cpus[i]]) {
                        pthread_mutex_unlock(&mutex[avail_cpus[i]]);
-                       dbug(2, "file switching is progressing, signal
ignored.\n", sig);
+                       dbug(2, "file switching is progressing, signal %d
ignored.\n", sig);
                        return;
                }
                pthread_mutex_unlock(&mutex[avail_cpus[i]]);
diff --git a/staprun/staprun.h b/staprun/staprun.h


Frank, do you have an opinion about the printsig() issue? Or should I just post
the patches I already have?


Thanks,
Timm

-- 
You are receiving this mail because:
You are the assignee for the bug.