Commit e5a11016 authored Mar 03, 2010 by Masami Hiramatsu Committed by Ingo Molnar Mar 04, 2010

x86: Issue at least one memory barrier in stop_machine_text_poke()



Fix stop_machine_text_poke() to issue smp_mb() before exiting
waiting loop, and use cpu_relax() for waiting.

Changes in v2:
 - Don't use ACCESS_ONCE().

Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com>
Acked-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: systemtap <systemtap@sources.redhat.com>
Cc: DLE <dle-develop@lists.sourceforge.net>
Cc: Jason Baron <jbaron@redhat.com>
LKML-Reference: <20100304033850.3819.74590.stgit@localhost6.localdomain6>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

parent 32cb0dd5

arch/x86/kernel/alternative.c

+2 −2

Original line number	Original line	Diff line number	Diff line
	@@ -595,8 +595,8 @@ static int __kprobes stop_machine_text_poke(void *data)
	wrote_text = 1;		wrote_text = 1;
	} else {		} else {
	while (!wrote_text)		while (!wrote_text)
	smp_rmb();		cpu_relax();
	sync_core();		smp_mb(); /* Load wrote_text before following execution */
	}		}

	flush_icache_range((unsigned long)tpp->addr,		flush_icache_range((unsigned long)tpp->addr,