March 2011 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

Create a one-to-many tunnel
by Vimal March 16, 2011

March 16, 2011

Hi, I wish to create an IP in IP tunnel interface that works as follows: At the sender: * The interface checks the dst IP address on the IP packet that it receives. * It encapsulates the IP packet inside another IP packet with: Field 1: source address = the interface's IP address Field 2: dst address = the packet's dst IP address * The interface now transmits the packet This is different from normal IPinIP tunnels because it is not point-to-point. The operation in Field 2 is similar to NAT. Also, this is different from SNAT, because I want the receiver to know what the original source IP. At the receiver, assume that there is a stack that understands this special IPinIP packet and has a way to handle it. If it's too specific, then I do not mind implementing it. Thanks, -- Vimal

2 2

project ideas
by mohit verma March 15, 2011

March 15, 2011

Hi all , I am seeking for any project on Networking related stuff ( in kernel land). I have visited the kernelnewbies.org janitors and kernel projects. Among them i found two projects of my interest but both of them are completed. So can anyone suggest me any ideas? I would appreciate any suggestions and ideas. -- ........................ *MOHIT VERMA*

7 7

UART and servo controller
by Carlo Caione March 15, 2011

March 15, 2011

Hi, as an exercise I was trying to write a very simple module to drive a micro serial servo controller. This servo controller should be interfaced through a RS232-UART converter to a free UART port on the OMAP processor present on a beagleboard. This servo controller is very simple: the serial commands are given through serial port following a strict protocol. The idea was to create an ad-hoc character driver and interact with it using ioctl() (or procfs) to give commands to the motors (which motor to move and the final position). To learn about serial and UART drivers I was looking at the serial drivers for the omap (drivers/serial/omap-serial.c) but I suspect that this is not the right choice. The problem is that I don't want a serial device (/dev/ttyXX) the user can interact with (i.e. reading or writing to). The user should be able just to use ioctl() on an ad-hoc character device (i.e. /dev/motors) whereas the communication part through the UART and the communication protocol (configuration of the packet to send) should be up to the driver, in a transparent way. Now my question is: * How can I manage to hide the serial communication part into the driver (kernel), exposing to the user just an ad-hoc character device? It is enough for me a pointer to a driver similar to mine to take a look at the code. I suspect that I should use the UART driver because I need to set the communication parameters such as baudrate, parity and so on. But also when I use uart_register_driver() to register the driver I need (need?) e pointer to the console structure and this is not I want. Thank you and sorry for the noob question. -- Carlo Caione

2 2

scheduling - more doubts
by sudheer.divakaran＠wipro.com March 15, 2011

March 15, 2011

Hi, I had asked a similar question before, but I've some doubts regarding the correctness of the following code. Please see the following code snippet from the file linux-2.6.37/arch/x86/kernel/apm_32.c. This is the core function of apm thread. Question: Suppose the apm thread has just completed execution of the line 1442 and the scheduling happens for some reason, AFAIK, since the apm thread's state has been set to TASK_INTERRUPTIBLE, it will be removed from the run-queue. So there is a chance for the apm thread to remain suspended indefinitely as nobody is there to wakeup the apm thread, correct? I'm asking this because the same function has been given as an example in the following link and would like to confirm the accuracy of the example. http://www.linuxjournal.com/node/8144/print 1437 static void apm_mainloop(void) 1438 { 1439 DECLARE_WAITQUEUE(wait, current); 1440 1441 add_wait_queue(&apm_waitqueue, &wait); 1442 set_current_state(TASK_INTERRUPTIBLE); 1443 for (;;) { 1444 schedule_timeout(APM_CHECK_TIMEOUT); 1445 if (kthread_should_stop()) 1446 break; 1447 /* 1448 * Ok, check all events, check for idle (and mark us sleeping 1449 * so as not to count towards the load average).. 1450 */ 1451 set_current_state(TASK_INTERRUPTIBLE); 1452 apm_event_handler(); 1453 } 1454 remove_wait_queue(&apm_waitqueue, &wait); 1455 } -- Thanks, Sudheer Please do not print this email unless it is absolutely necessary. The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments. WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email. www.wipro.com

4 4

SIGPROF received ( profile timer expired )
by अनिकेत आगाशे March 15, 2011

March 15, 2011

Hi all I am facing problem in running an application over character driver written by me. I wanted to profile that application (time required for each system call) so I compiled it using gcc -pg option. But some time i am receiving SIGPROF signal during ioctl() system call. Is there any way we can control profile timer's value ? or some other way to get rid of this problem. Thanks in advance

2 5

I have a problem with schedule() and sys_perf_event_open.
by Minwoo Lee March 14, 2011

March 14, 2011

Hello. I am hacking the kernel source for studying. I just modified schedule() in kernel/sched.c for profiling processes. I want to count the number of cache miss, so added the code using sys_perf_event_open. However, it seems to returns always error fd. I hope to know what should I do for getting the number of cache miss. Here is the diff code based at kernel version 2.6.37 Thanks for reading this e-mail. diff --git a/kernel/sched.c b/kernel/sched.c index 297d1a0..afe68bc 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -72,6 +72,7 @@ #include <linux/ctype.h> #include <linux/ftrace.h> #include <linux/slab.h> +#include <linux/syscalls.h> #include <asm/tlb.h> #include <asm/irq_regs.h> @@ -4061,6 +4062,30 @@ pick_next_task(struct rq *rq) BUG(); /* the idle class will always have a runnable task */ } +static void profile(struct task_struct *task) +{ + int fd; + size_t res; + u64 count; + struct perf_event_attr attr = + { + .type = PERF_TYPE_HARDWARE, + .config = PERF_COUNT_HW_CACHE_MISSES, + .size = sizeof(struct perf_event_attr) + }; + int res_close; + + BUG_ON(task == 0); + + fd = sys_perf_event_open(&attr, 0, -1, -1, 0); + BUG_ON(fd < 0); + res = sys_read(fd, (char *)&count, sizeof(u64)); + BUG_ON(res != sizeof(u64)); + res_close = sys_close(fd); + BUG_ON(res_close == -1); + pr_info("PERF: %d %u %llu %d", fd, res, count, res_close); +} + /* * schedule() is the main scheduler function. */ @@ -4080,6 +4105,7 @@ need_resched: release_kernel_lock(prev); need_resched_nonpreemptible: + profile(prev); schedule_debug(prev);

2 2

Gsoc participation
by Maksym Planeta March 14, 2011

March 14, 2011

Hello, I'm a student and I want to participate in Google Summer of Code (Gsoc) in project related to Linux kernel development. I've read that kernelnewbies is one of Gsoc project groups [1]. So I want to know what criteria should I fit to be good enough for taking part in Gsoc in kernel development project? 1. https://www.linuxfoundation.org/collaborate/workgroups/gsoc/google-summer-c…

5 7

release used memory
by Andreas Leppert March 13, 2011

March 13, 2011

Hello, for study reasons, I evaluate a kernel module by running a parallel version of bzip. Input data are some senseless data files (500 x 10MB for example). While the bzip program runs, I took a look at the output of top and noticed something which I do not understand. In the following, I'm refering to that two lines of top which look something like that: Mem: 8185716k total, 5603224k used, 2582492k free, 9104k buffers Swap: 8388604k total, 0k used, 8388604k free, 5374400k cached While the bzip program runs, it uses memory and thus, the value before "free" in the Mem: line is getting smaller. As you can see above, there is also some "cached stuff" (5374400k). What is meant by this value? Someone explained me that it has to do something with filesystem buffers which were read or written. Could you elaborate on this? If I have enough input data, the memory is not enough (the "free" value gets smaller and smaller, and the "cached" value gets bigger and bigger). Then kswapd appears in the list of running processes. I would expect that Linux gets rid of the "cached" stuff so that the program gets more memory. Instead, nothing happens and kswapd is activated. What's the sense of keeping some filesystem buffers when they are not used anymore? The main problem is that I do not understand what by "buffers" and "cached" is meant. It would be nice if you can help me on this. Thanks in advance Andreas

4 4

SPI over GPIO
by Alexandru Caramida March 13, 2011

March 13, 2011

Hello everyone, I am trying to emulate an SPI interface using four GPIO pins. This is a desktop computer with linux 2.6.28 kernel. The problem is I don't know if the pins are working. I read that if I have the sysfs support for the kernel (which I do :D) all I have to do is enable the respectiv pin using echo N > /sys/class/gpio/export And I would get a gpioN directory But I always get: "write error : invalid argument." Is there some other option i need to enable first? The people I am working with assured me that these pins are not used by any other device and they haven't found anything to suggest that they were turned off manualy. Please help,

6 5

where is __memory_barrier in kernel ?
by loody March 13, 2011

March 13, 2011

hi all: I grep kernel source and found cpu_relax is defined as __memory_barrier(), which seems not defined in kernel source. At beginning I think it may be the gcc build-in functions, but I cannot find in the gcc document. Where and what is that used for? -- Regards, miloody

5 7