September 2012 - kernelnewbies@kernelnewbies.org

what will happen when the jiffies overflow ?
by Fan Yang Sept. 20, 2012

Sept. 20, 2012

hi all: As far as I known, the var jiffies in the kernel will overflow after 497days as the Hz defined 100. So what waill happen when the jiffies overflow? thanks

3 2

Questions about process state manipulation
by Amit Gupta Sept. 20, 2012

Sept. 20, 2012

Hi All, I had a question about manipulating process state information in the kernel and am curious about the right way to go about it. (I'm just messing around with some stuff to try and understand the kernel better). I had 2 questions: 1. I'm trying to create a process in the TASK_INTERRUPTIBLE state in the kernel. Since do_fork() creates a process/task that is ready to be scheduled, I was thinking what might be a good way to pull it off the runqueue/stop its execution after do_fork() returns? 2.I'm curious if someone can explain the correct way to use the "set_task_state()" macro. Its implementation just seems to boil down to an assignment to the appropriate member of the task_struct. The question I had in mind was, what if the process is already running? Is it safe to do it? Don't we need to acquire some sort of lock before we manipulate a task_struct ? 3. I would also be interested to read the correct way to use task_lock() that locks on alloc_lock field of the task_struct. What exactly is this lock for and does acquiring it/or any other lock in the task_struct cause the process to be remoted from the runqueue? Would appreciate any pointers to appropriate code/reading or any explanation. Thanks, -- Amit Gupta

3 6

What is the difference between evdev_event and input_inject_event?
by Tom Xue Sept. 20, 2012

Sept. 20, 2012

Hi, I did a trial on my Beagle Board which installed with Android GingerBread2.3.4. That is to sendevent from console and trigger the "mouse click" action. Firstly I check what I will receive if I make some *real* mouse click action. If I click the right-bottom corner of the screen: # getevent /dev/input/event2: 0004 0004 00090001 /dev/input/event2: 0001 0110 00000001 /dev/input/event2: 0000 0000 00000000 /dev/input/event2: 0004 0004 00090001 /dev/input/event2: 0001 0110 00000000 /dev/input/event2: 0000 0000 00000000 That means if I use "sendevent" with the same data and sequence, the "mouse click" action shoule be triggered. But the trial failed. Nothing happened. Then someone gave a solution. Check this link: http://e2e.ti.com/support/embedded/android/f/509/p/207113/737993.aspx#737993 And I modified the kernel drivers/input/evdev.c file as below. --- evdev.c 2012-09-19 21:33:08.239665514 +0800 +++ evdev_modified.c 2012-09-19 21:32:56.307334229 +0800 @@ -352,6 +352,8 @@ while (retval < count) { if (input_event_from_user(buffer + retval, &event)) { retval = -EFAULT; goto out; } input_inject_event(&evdev->handle, event.type, event.code, event.value); retval += input_event_size(); } + evdev_event(&evdev->handle,event.type, event.code, event.value); + out: mutex_unlock(&evdev->mutex); return retval; And then recompile the kernel image and put it on the SD card, booting... and did the trial again. This time I succeeded. But why? evdev_event() has the totally same args as input_inject_event(), and I found that input_inject_event() was implemented by adding printk(). That shows when applying sendevent, the input_inject_event() will pass the event to handlers as well (my printk gives INPUT_PASS_TO_HANDLERS for 6 times, due to 6 data were sent). Could anyone please help to clarify what the difference is between them? And why only evdev_event() helps in this case? BTW, I have a HTC Sensation XE phone which installed with Android 4.0.3. And I checked its kernel source code, it is almost same to above code (without evdev_event()). But this phone can work when I sendevent to it, no need to add evdev_event(). Thanks! BR, Tom

1 0

stack trace implementation
by devendra.aaru Sept. 20, 2012

Sept. 20, 2012

Hi all, where can i look into the stack tracing implementation in kernel? is it likely same as the gcc's builtin functions? thanks,

2 1

Maximum size of data segment used by c program
by devendra.aaru Sept. 20, 2012

Sept. 20, 2012

Hi all, Is there any limit on the maximum data segment size used by c programs? Thanks,

3 5

R: Problem using and creating virtual interfaces
by Pietro Paolini Sept. 18, 2012

Sept. 18, 2012

Hello, thanks for your anser, I solved my problem using some virtual machine (kmv) but what I really would like was to do what virtual machine do: 1) create some virtual interfaces and assign them an IP addresses, each of them must have a different MAC address 2) make a bridge with my real one I am able to catch the second point but not first one, can you provide me an example of use of the ip command for a virtual interface creation ? Many thanks, Pietro Paolini. ________________________________________ Inizio: Jeff Haran [Jeff.Haran(a)citrix.com] Inviato: martedì 18 settembre 2012 19.36 Fine: Pietro Paolini Oggetto: RE: Problem using and creating virtual interfaces Not sure what you mean by “IP interface”, but perhaps network namespaces will provide the functionality you are looking for. Running ifconfig like you did below simply adds another IP address to an existing interface. “ethX:1” isn’t really a separate network interface in the kernel (there’s no separate struct netdevice associated with it), it’s just an alias for “ethX” that addresses can be assigned to using the old BSD derived IOCTLs that ifconfig uses. ifconfig is more or less a deprecated command in linuxland. You will probably want to make yourself familiar with the iproute2 package ip command, which is its more modern equivalent and provides an interface that allows you to avoid using network interface alias names like “ethX:1”. Jeff Haran From: kernelnewbies-bounces(a)kernelnewbies.org [mailto:kernelnewbies-bounces@kernelnewbies.org] On Behalf Of Pietro Paolini Sent: Tuesday, September 18, 2012 2:24 AM To: kernelnewbies(a)kernelnewbies.org Subject: Problem using and creating virtual interfaces Hi all, I am not really sure this is the correct newsletter for my question, if not please apologize me and suggest me a right newsletter, thanks. My goal is to simulate more than one IP interface using just a physical interface, for do that I tried using this alias Ifconfing ethX:1 <IP> But when I try to send a packet over that interface I see the IP source the same as the original, and real, Ethernet interface. Then I tried to use TAP interface but I was unable to “link” them with my physical one, I notice them can communicate each others but can not ping (ping –I tap0 <ip real IF>) the original one. It is still valid if I create a bridge with all them. I would like avoid to install many Virtual Box as I need but what I want is exactly what it do: build a new virtual interface and bridge it with a real one in order to have more IP interface attached at one physical. Any suggestion ? Thanks in advance, Pietro.

1 0

Problem using and creating virtual interfaces
by Pietro Paolini Sept. 18, 2012

Sept. 18, 2012

Hi all, I am not really sure this is the correct newsletter for my question, if not please apologize me and suggest me a right newsletter, thanks. My goal is to simulate more than one IP interface using just a physical interface, for do that I tried using this alias Ifconfing ethX:1 <IP> But when I try to send a packet over that interface I see the IP source the same as the original, and real, Ethernet interface. Then I tried to use TAP interface but I was unable to "link" them with my physical one, I notice them can communicate each others but can not ping (ping -I tap0 <ip real IF>) the original one. It is still valid if I create a bridge with all them. I would like avoid to install many Virtual Box as I need but what I want is exactly what it do: build a new virtual interface and bridge it with a real one in order to have more IP interface attached at one physical. Any suggestion ? Thanks in advance, Pietro.

1 0

About support USB3.0 LTM(Latency Tolerance Messaging)
by zhaoqiao Sept. 17, 2012

Sept. 17, 2012

Hi all, I have a problem. In commit 024f117... said "USB: Add a sysfs file to show LTM capabilities." speak about "+What: /sys/bus/usb/devices/.../ltm_capable", but i not found ltm_capable. Related information not found in the source code, maybe i'm missing something. :( Can you give me some help? Additional message: http://tomoyo.sourceforge.jp/cgi-bin/lxr/source/Documentation/ABI/testing/s… ----- Thanks, Qiao

1 0

some problem about make menuconfig
by g.wangweihong＠gmail.com Sept. 17, 2012

Sept. 17, 2012

hello all: I had a problem. I want to build a kernel-2.6.22.19 in my ubuntu which's kernel is 3.2 for embedded platform. But i failed in the step of /make menuconfig/. It tips as this: > kiongf@ubuntu:~/sigma/tarball.386/smp86xx_kernel_source_R2.6.22-35/linux$ > make menuconfig > HOSTCC scripts/kconfig/lxdialog/checklist.o > HOSTCC scripts/kconfig/lxdialog/inputbox.o > HOSTCC scripts/kconfig/lxdialog/menubox.o > HOSTCC scripts/kconfig/lxdialog/textbox.o > scripts/kconfig/lxdialog/textbox.c: 在函数‘print_line’中: > scripts/kconfig/lxdialog/textbox.c:323:9: 警告：变量‘x’被设定但未被使 > 用 [-Wunused-but-set-variable] > scripts/kconfig/lxdialog/textbox.c:323:6: 警告：变量‘y’被设定但未被使 > 用 [-Wunused-but-set-variable] > HOSTCC scripts/kconfig/lxdialog/util.o > HOSTCC scripts/kconfig/lxdialog/yesno.o > HOSTCC scripts/kconfig/mconf.o > HOSTLD -static scripts/kconfig/mconf > scripts/kconfig/lxdialog/checklist.o: In function `print_arrows': > checklist.c:(.text+0x5a): undefined reference to `acs_map' > checklist.c:(.text+0xbc): undefined reference to `acs_map' > checklist.c:(.text+0x105): undefined reference to `acs_map' > checklist.c:(.text+0x116): undefined reference to `acs_map' > checklist.c:(.text+0x127): undefined reference to `acs_map' > scripts/kconfig/lxdialog/checklist.o:checklist.c:(.text+0x138): more > undefined references to `acs_map' follow > scripts/kconfig/lxdialog/checklist.o: In function `dialog_checklist': > checklist.c:(.text+0x46d): undefined reference to `stdscr' > checklist.c:(.text+0x4c3): undefined reference to `COLS' > checklist.c:(.text+0x4d7): undefined reference to `LINES' > /usr/lib/gcc/i686-linux-gnu/4.6/../../../i386-linux-gnu/libncursesw.a(hashmap.o):(.text+0x9c2): > more undefined references to `SP' follow > /usr/lib/gcc/i686-linux-gnu/4.6/../../../i386-linux-gnu/libncursesw.a(lib_beep.o): > In function `beep': > (.text+0x5): undefined reference to `cur_term' > /usr/lib/gcc/i686-linux-gnu/4.6/../../../i386-linux-gnu/libncursesw.a(lib_beep.o): > In function `beep': > (.text+0x1b): undefined reference to `putp' > /usr/lib/gcc/i686-linux-gnu/4.6/../../../i386-linux-gnu/libncursesw.a(lib_beep.o): > In function `beep': > (.text+0x22): undefined reference to `_nc_flush' > /usr/lib/gcc/i686-linux-gnu/4.6/../../../i386-linux-gnu/libncursesw.a(lib_hline.o): > In function `whline': I google for this problem, it says i didn't install the /ncurses. / > i libncurses5 - shared libraries for terminal handling > i libncurses5-dbg - debugging/profiling libraries for ncurses > i libncurses5-dev - developer's libraries for ncurses But after I install the lib , it doesn't change. The script /linux/scripts/kconfig/lxdialog///check-lxdialog.sh/ contain these messages: > ldflags() > { > for ext in so a dylib ; do > for lib in ncursesw ncurses curses ; do > $cc -print-file-name=lib${lib}.${ext} | grep -q / > if [ $? -eq 0 ]; then > echo "-l${lib}" > exit > fi > done > done > exit 1 > } > > # Where is ncurses.h? > ccflags() > { > if [ -f /usr/include/ncurses/ncurses.h ]; then > echo '-I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>"' > elif [ -f /usr/include/ncurses/curses.h ]; then > echo '-I/usr/include/ncurses -DCURSES_LOC="<ncurses/curses.h>"' > elif [ -f /usr/include/ncurses.h ]; then > echo '-DCURSES_LOC="<ncurses.h>"' > else > echo '-DCURSES_LOC="<curses.h>"' > fi > } libncurses appear in /usr/lib ,and the head file appear in /usr/include. > kiongf@ubuntu:/usr/lib$ ls | grep curse > libncurses_g.a > libncurses++_g.a > libncurses.so.5.7 > libncurses++w_g.a > libncursesw_g.a > kiongf@ubuntu:/usr/include$ ls | grep ncurse > ncurses_dll.h > ncurses.h > ncursesw I have trapped in this problem for several days, please help me if you have any good ideas. Thanks.

2 1

kernel stack memory
by shubham sharma Sept. 16, 2012

Sept. 16, 2012

Hi, As far as i know, the size of stack allocated in the kernel space is 8Kb for each process. But in case i use more than 8Kb of memory from the stack then what will happen? I think that in that case the system would crash because i am accessing an illegal memory area. I wrote kernel module in which i defined an integer array whose size was 8000. But still it did not crash my system. Why? The module i wrote was as follows: #include <linux/kernel.h> #include <linux/module.h> int __init init_my_module(void) { int arr[8000]; printk("%s:%d\tmodule initilized\n", __func__, __LINE__); arr[1] = 1; arr[4000] = 1; arr[7999] = 1; printk("%s:%d\tarr[1]:%d, arr[4000]:%d, arr[7999]:%d\n", __func__, __LINE__, arr[1], arr[4000], arr[7999]); return 0; } void __exit cleanup_my_module(void) { printk("exiting\n"); return; } module_init(init_my_module); module_exit(cleanup_my_module); MODULE_LICENSE("GPL");

9 11