This is the mail archive of the systemtap@sourceware.org mailing list for the systemtap project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFC PATCH 2/3] Djprobe improvement patches (Re: Dynamic djprobe)


Hi, Anil
Thank you for your review.

Keshavamurthy Anil S wrote:
On Wed, Aug 03, 2005 at 02:07:38AM +0900, Masami Hiramatsu wrote:

diff -Narup linux-2.6.12-djprobe.2/arch/i386/kernel/kprobes.c linux-2.6.12-djprobe.3/arch/i386/kernel/kprobes.c
--- linux-2.6.12-djprobe.2/arch/i386/kernel/kprobes.c	2005-08-01 22:00:17.000000000 +0900
+++ linux-2.6.12-djprobe.3/arch/i386/kernel/kprobes.c	2005-08-02 20:30:10.000000000 +0900
@@ -497,7 +497,7 @@ int djprobe_bypass_handler(struct kprobe
	kprobe_opcode_t *stub = djpi->stub.insn;
	int cpu = smp_processor_id();

-	if (!DJPI_CHECKED(djpi)) {
+	if (!DJPI_CHECKED(djpi) && check_safety_djprobe_instance(djpi)) {
		cpu_set(cpu, djpi->checked_cpus); /* check this cpu */
		
		if (DJPI_CHECKED(djpi)) { /* all cpus are checked */

Still you are assuming that all CPU's to enter djprobe_bypass_handler() which is incorrct. Please fix this logic.

OK, I think it is easy that I develop additional safety check point in other points. I am developing it based on the idea suggested by Mathieu.



Best Regards,


--
Masami HIRAMATSU
2nd Research Dept.
Hitachi, Ltd., Systems Development Laboratory
E-mail: hiramatu@sdl.hitachi.co.jp


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]