[LKML watch] Issues that systemtap team may be interest

[Masami Hiramatsu]

Hi,

Here are issues posted on LKML in the last week.
Some issues were solved, but other issues were not solved.

http://www.ussg.iu.edu/hypermail/linux/kernel/0512.2/1841.html
Linux-Kernel Archive: 4k stacks
# Measuring amount of consumed stack size by using
# a hand-made patch. SystemTap can do it more easily.

http://www.ussg.iu.edu/hypermail/linux/kernel/0512.3/0220.html
Linux-Kernel Archive: How to obtain process ID that created a packet
# They tried to catch PID that generated a packet by using netfilter.
# But the attempt failed. SystemTap can easily do that.

For example:
probe kernel.function("sys_send") {
        printf("packet generated by %d\n", pid());
}

http://www.ussg.iu.edu/hypermail/linux/kernel/0512.3/0455.html
Linux-Kernel Archive: [SCHED] Totally WRONG prority calculation
with specific test-case(since 2.6.10-bk12)
# Monitoring process priority by using top command.
# SystemTap can provide more precise log of priority.

http://www.ussg.iu.edu/hypermail/linux/kernel/0512.3/0342.html
Linux-Kernel Archive: How to get blocking information?
# He wants to know when the processes sleep inside system-calls.
# SystemTap may obtain these information easily.
# This issue is not solved.

Best regards,

--
Masami HIRAMATSU
2nd Research Dept.
Hitachi, Ltd., Systems Development Laboratory
E-mail: [hidden email]