Re: [PATCH v8] scripts: ftrace - move the sort-processing in ftrace_init

[Steven Rostedt]

On Fri, 21 Jan 2022 11:42:28 +0100
Heiko Carstens <hca@xxxxxxxxxxxxx> wrote:

> This really should be addressed before rc1 is out, otherwise s390 is
> broken if somebody enables ftrace.
> Where "broken" translates to random crashes as soon as ftrace is
> enabled, which again is nowadays quite common.

Instead of reverting, should we just add this patch?

-- Steve

diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig
index f468767bc287..752ed89a293b 100644
--- a/kernel/trace/Kconfig
+++ b/kernel/trace/Kconfig
@@ -70,6 +70,13 @@ config HAVE_C_RECORDMCOUNT
 	help
 	  C version of recordmcount available?
 
+config BUILDTIME_MCOUNT_SORT
+       bool
+       default y
+       depends on BUILDTIME_TABLE_SORT && !S390
+       help
+         Sort the mcount_loc section at build time.
+
 config TRACER_MAX_TRACE
 	bool
 
@@ -918,7 +925,7 @@ config EVENT_TRACE_TEST_SYSCALLS
 config FTRACE_SORT_STARTUP_TEST
        bool "Verify compile time sorting of ftrace functions"
        depends on DYNAMIC_FTRACE
-       depends on BUILDTIME_TABLE_SORT
+       depends on BUILDTIME_MCOUNT_SORT
        help
 	 Sorting of the mcount_loc sections that is used to find the
 	 where the ftrace knows where to patch functions for tracing
diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
index 403e485bf091..b01e1fa62193 100644
--- a/kernel/trace/ftrace.c
+++ b/kernel/trace/ftrace.c
@@ -6429,10 +6429,10 @@ static int ftrace_process_locs(struct module *mod,
 
 	/*
 	 * Sorting mcount in vmlinux at build time depend on
-	 * CONFIG_BUILDTIME_TABLE_SORT, while mcount loc in
+	 * CONFIG_BUILDTIME_MCOUNT_SORT, while mcount loc in
 	 * modules can not be sorted at build time.
 	 */
-	if (!IS_ENABLED(CONFIG_BUILDTIME_TABLE_SORT) || mod) {
+	if (!IS_ENABLED(CONFIG_BUILDTIME_MCOUNT_SORT) || mod) {
 		sort(start, count, sizeof(*start),
 		     ftrace_cmp_ips, NULL);
 	} else {