Linux Forums - The Linux Kernel

Basic question about device drivers

balteo — Thu, 19 Nov 2009 16:01:50 GMT

Hello,

I am tempted to write a device driver and I would like to better understand how these work before embarking on this adventure.

I recently purchased a usb tv tuner card i.e. a nova hauppage and followed the instructions given on linuxtv website and finally got a working tv on my Ubuntu box.

What I don't understand is:

-How many layers/drivers are there between the kernel and my hauppage tuner card? Just one or several?
-How do modules relate to each other?
-How would you define what the firmware is?

I also have a logitec webcam that was picked up automatically on installation by Ubuntu.
-Where is the driver for that specific webcam located on the hard drive?
-Can a given driver server several devices?

Any reply, albeit partial is welcome.

Thanks in advance,

Julien.

Using linux performance counters from within application

Wallin — Thu, 19 Nov 2009 02:25:46 GMT

I'm interested in using the new performance counter functionality from kernel 2.6.32. I've built the perf tool, and it seems like exactly what I want, but I'm wondering if there is a good way of embedding this functionality in an application.

Specifically, my goal is to profile only a portion of code, and trigger the beginning and end of the measurement from calls within the code. It seems that the perf tool is really just a wrapper for calling the sys_perf_event_open system call, so I assume that I probably want to call this directly from my code, but I haven't been able to find any decent documentation on that particular system call.

Does anyone know where I can find some documentation? Or is there perhaps a better way to go about this? The particular event counters I want (to start with at least) are the instruction count, cache miss counts, tlb misses, branch mispredictions, and malloc calls.

FTDI usb-serial converter causes near-hang

madengineer — Wed, 18 Nov 2009 19:50:49 GMT

I'm having a very strange problem on my HP Pavilion dv4t laptop (NVidia GeForce G105M GPU) with a clean install of Kubuntu Karmic (64 bit) and kernel 2.6.31-14.48. When I attempt to boot the system when it's connected to a board that holds several FTDI FT232RL USB-to-serial converter chips, the system begins to encounter serious problems as soon as X tries to start. However, this board works flawlessly with the same machine running Windows 7, and on a different laptop running Debian with 2.6.31.

From a user perspective, the system appears to hang immediately after the Kubuntu bootup screen ("kubuntu" and a progress bar) with the screen switching on and off a few times but showing only black. The machine does not respond in a timely manner to a reboot command using the magic SysRq keys (which normally works on this box), but it appears that some stuff is still happening, albeit very very VERY slowly (like something's in an infinite loop and is eating CPU). The only way to snap the machine out of it is a power-off by holding the power button down.

Examining /var/log/syslog (cut down to only the boot involved and attached, compressed with bz2) showed me that the kernel is up to some very strange things; first come buffer overflow errors involving the FTDI chips ("response buffer filled" and "nnn input overrun(s)"), then SATA communication faults. Meanwhile, gdm repeatedly fails to start and Xorg crashes. At the very end, I unplug the USB cable, and after this fails to improve matters (even though the kernel does recognize the disconnect) I do a manual power-down.

This problem occurred first on this machine with a kernel I built myself (identical to the stock 2.6.31-14.48 except with the patch to fix the ACPI quirk on the dv4 that prevents battery status from showing up after suspend/resume), and also occurred with 2.6.31-15.50. I did a clean install hoping to fix the problem, but there was no apparent change. I also tried removing the proprietary NVidia driver in favor of the open-source one, but this didn't help either. Also, if I plug the device in after booting, there is no crash, but it doesn't work properly (select() calls that one of my programs makes on the serial ports involved always return EINTR, when the same calls succeed on the Debian machine). In addition, plugging in a USB-serial converter containing a single FTDI chip (as opposed to the seven on the board mentioned above) does not cause any problem during booting, even though the same driver (ftdi_sio) is involved.

I'm currently out of ideas on what to try; kernel hacking is a bit out of my league. Does anyone have any suggestions?

EDIT: I tried removing the modemmanager and kppp packages on a wild hunch, and this appears to stop the problem from occurring. Tomorrow I'm going to reinstall the packages and confirm that the problem comes back. Still doesn't explain why probing the USB-serial converters to determine if modems are attached throws the kernel into chaos though...

Attached Files

syslog-cut.bz2 (20.4 KB)

kmemtrace problem

AnbuManikandan — Wed, 18 Nov 2009 13:30:21 GMT

Hi,
While building the linux kernel image, I enabled the config option CONFIG_KMEMTRACE. After loaded the image into target, It doesn't have the folder "kmemtrace" inside the /sys/kernel/debug folder.

If anyone have the idea, please help me.

Thanks & Regards,
Manikandan

Unpacking kernel source produces duplicate files

mucygni — Wed, 18 Nov 2009 07:52:38 GMT

Hi. Please excuse me I'm new at this, but the newbie section just didn't seem appropriate. I can't seem to unpack/extract the kernel source without difficulty. I've tried every 2.6 version. They all turn up several duplicate .h and .c files. I'm not sure whether I should agree to write over the previous or not. Maybe I need both copies. I've considered extracting twice, choosing both options, but then the files couldn't be configured properly. I should include that I'm using XP and I'm extracting with 7-zip. I can't see why either of those should be a problem. I'm obviously not going to develop on XP, I just want to study the source. I have good info for that, and I just want to be sure I have everything I need and it's in order. That's all. Any opinions out there?

Kernel 2.6.27.7-9 for Suse 11.1 help required with CPU Clock freq/speed/ticks

subhan222 — Tue, 17 Nov 2009 15:31:18 GMT

Problem Statement: High CPU clock speed/freq
Requirement: Obtain a stable CPU clock ticks/speed/freq on Suse 11.1 with kernel 2.6.27.7-9
Questions: Please refer to the bottom of this page

Artifacts in Use:
-----------------------------------
OS: Suse 11.1
GCC Version: 4.3.2 (Suse Linux) [gcc_4_3-brach revision 141291]
Kernel: 2.6.27.7-9
Kernel Mods:
* Un-check the Tickless option
* Timer Frequency - 1000Hz
* Preempt kernel

Others 1
-----------
OS: Suse 10.2
GCC Version: 4.1.2 20061115 (prerelease)(SUSE Linux)
Kernel: 2.6.18.2-34-default
Kernel Mods:
* Preemptible Kernel (Low-Latency Desktop) (NEW)
* Preempt The Big Kernel Lock (NEW)
* Timer frequency (NEW) - 1000 Hz (NEW)

Others 2
-----------
OS: Suse 9.3
GCC Version: 3.3.5 20050117 (prerelease)(SUSE Linux)
Kernel 2.6.11.4-20a-default
Kernel Mods: None

I executed the following c++ code on the above stated builds

code is in attached file as well

Code:

#include 

#include 

#include 

#include 





using namespace std;



long long int counter = 0;

struct timeval last_time;

struct timeval curr_time;

double dif = 0.0;

double freq_sum = 0.0;

double reps = 0.0;

double freq_min = 99999.0;

double freq_max = 0.0;

long long int num_too_low = 0;



void time_check(int data)

{

    // work out the frequecy of the SIGALRMs

    ++counter;

    if (counter >= 1000)

    {



        // need to work out time since last 1000 ticks

        gettimeofday(&curr_time, NULL);

        dif = ((curr_time.tv_usec + curr_time.tv_sec * 1000000)

                - (last_time.tv_usec + last_time.tv_sec * 1000000));

        double freq = (counter / dif)  * 1000000;



        freq_sum = ((freq_sum * reps) + freq) / (reps + 1) ;



        ++reps;

        cout << "Frequency = " << freq << " reps= " << dif / 1000000 << endl;

        if (freq_min > freq) freq_min = freq;

        if (freq_max < freq) freq_max = freq;

        if (freq < 950) 

            ++num_too_low;

        last_time = curr_time;

        counter = 0;

    }

}



int main (int argc, char * argv[])

{

    cout << "go" /* << "\a \a \a" */ << endl;

        

    int res;

    gettimeofday(&last_time, NULL);

    struct sigaction action;



    struct itimerval timer_value;



    timer_value.it_value.tv_sec = 0;

    timer_value.it_value.tv_usec = 1;

    timer_value.it_interval.tv_sec = 0;

    timer_value.it_interval.tv_usec = 1;



    action.sa_handler = time_check;

    action.sa_flags = SA_NOMASK;

    sigemptyset(&action.sa_mask);



    sigaction( SIGALRM, &action, NULL);

    res = setitimer( ITIMER_REAL, &timer_value, NULL );

    if (res == -1)

        cout << "error in setitimer" << endl;



    cout << "Starting CPU Test program" << endl;

    // loop for a long time



    while (reps < 3000.0)

    {

        sleep(1);

    }

    cout << "average CPU frequency = " << freq_sum << endl;

    cout << "reps = " << reps << endl;

    cout << "Min = " << freq_min << "  Max = " << freq_max << endl;

    cout << "number of low = " << num_too_low << endl;

    return 0;

}

Compiled the above code with "g++ test.cpp -o test"
Run the above code with "./test"

Code:

Observations:

SUSE Version    AVG. Frequency                        AVG. REPS

Suse 9.3                  999.xxx to 990.xxx                          reps > 1 (on average)

Suse 10.2                 999.xxx to 998.xxx                          reps (1 - 1.001)

Suse 10.3             1000.xxx to 912.xxx                          reps (1 - 1.090)

Suse 11.1       0 to  324263.xxx                                  reps (-2e-0.6 upto 0.005112)



Suse 11.1 - Options:



- Turn off ACPI in options when booting with clock=tsc | clock=pit



Observations for suse 11.1 with options acpi=off and clock=tsc | clock=pit



SUSE Version    AVG. Frequency                        AVG. REPS

Suse 11.1                 710.xxx to 999.xxx                      reps (1.37xx to 1.0002)

Hardware used for Suse 11.1 - HP Pavilion Slimline s3821uk Desktop PC (NQ808AA) specifications - HP Home & Home Office products

While testing no other application/s were running.

Code:

Question:



* Is my c++ code correct?

* Am I using the correct way to get the clock ticks or is there more reliable tchnique avaliable to do this?

* Is that a problem with new Kernel if so is it a known issue ?

* Do I need to modify the Kernel to obtain regular and stable clock tick at 1000 Hz ?

* Do I have to upgrade the Kernel ?

* Would be great, if pointed to the right direction for solution (e.g. another forum) if no solution at hand.



Thanks

Attached Files

test.txt (2.1 KB)

MMC Card suddenly changes to readonly file system

maheshgupta024 — Mon, 16 Nov 2009 13:46:51 GMT

Hi all.. I have a very strange problem..
My system has a sd memory card which is mounted with vfat filesystem with "rw" mode. Now when I transfer a number(say 150 ) files, my sd card suddenly changes to ro filesystem and the dmesg output gave the following :
FAT: Filesystem panic (dev sdmmc0p1)
fat_free_clusters: deleting FAT entry beyond EOF
File system has been set read-only
what is reason for the SD card getting suddenly changing to read only filesystem. Can anyone tell what is reason.

Thanks in advance.

Kernel 2.6.31.6 will boot all the way to loading services, then stop

Zenith88 — Sun, 15 Nov 2009 01:39:49 GMT

Hi,

I've been a Slackware/Redhat user for many years until v.6, then withdrawn for almost 10 years.
Now I am back and building 2 machines for my kids to use for school projects.

Installed Fedora 11, d/l latest stable kernel from .org, configured, compiled, installed and stalled at the point where kernel will boot all the way up to the point where it will start services. Instead of proceeding, it just sits there and I can type anything and it will appear on the screen. If at that point I would press CAD, it will say shutting down MD volumes and reboot.

What information should I provide for you to be able to help me troubleshoot this particular kernel?

I can boot into the kernel 2.6.30.9 that came with the latest update no problem and am writing it from that Fedora 11 machine.

Here's my Grub.conf with the kernel in question coming first:

Code:

# grub.conf generated by anaconda

#

# Note that you do not have to rerun grub after making changes to this file

# NOTICE:  You have a /boot partition.  This means that

#          all kernel and initrd paths are relative to /boot/, eg.

#          root (hd1,0)

#          kernel /vmlinuz-version ro root=/dev/mapper/vg_lion-lv_root

#          initrd /initrd-version.img

#boot=/dev/sdb

default=2

timeout=10

splashimage=(hd1,0)/grub/splash.xpm.gz

#hiddenmenu

title Fedora (2.6.31.6.-LionP4.fc11.i686)

        root (hd1,0)

        kernel /vmlinuz-2.6.31.6.LionP4.fc11.i686 ro root=/dev/mapper/vg_lion-lv_root rhgb vga=0x346 #quiet

        initrd /initrd-2.6.31.6-LionP4.fc11.i686.img

title Fedora (2.6.30.9-96.fc11.i586)

        root (hd1,0)

        kernel /vmlinuz-2.6.30.9-96.fc11.i586 ro root=/dev/mapper/vg_lion-lv_root rhgb quiet vga=0x346

        initrd /initrd-2.6.30.9-96.fc11.i586.img

title Fedora (2.6.30.9-96.fc11.i686.PAE)

        root (hd1,0)

        kernel /vmlinuz-2.6.30.9-96.fc11.i686.PAE ro root=/dev/mapper/vg_lion-lv_root rhgb quiet vga=0x346

        initrd /initrd-2.6.30.9-96.fc11.i686.PAE.img

title Fedora (2.6.29.4-167.fc11.i686.PAE)

        root (hd1,0)

        kernel /vmlinuz-2.6.29.4-167.fc11.i686.PAE ro root=/dev/mapper/vg_lion-lv_root rhgb quiet vga=0x346

        initrd /initrd-2.6.29.4-167.fc11.i686.PAE.img

title WinXP

        rootnoverify (hd0,0)

        chainloader +1

[SOLVED] No Sound after Kernel Update

RobinVossen — Fri, 13 Nov 2009 21:44:06 GMT

Today I start my system after rebooting and I dont have any sound.
when I press my little icon on the bottom I get this:

"The volume control did not find any elements and/or devices to control. This means either that you don't have the right GStreamer plugins installed, or that you don't have a sound card configured.

You can remove the volume control from the panel by right-clicking the speaker icon on the panel and selecting "Remove From Panel" from the menu"

I tried to downgrade my kernel and my gsteamer no luck.
Upgraded back blah it doesn't work it kinda pisses me off...

So well here are some logs:

Quote:

[2009-11-01 11:11] synchronizing package lists
[2009-11-01 11:11] starting full system upgrade
[2009-11-01 11:12] starting full system upgrade
[2009-11-01 11:23] removed gnome-mount (0.8-1)
[2009-11-01 11:23] upgraded libgsf (1.14.15-1 -> 1.14.16-1)
[2009-11-01 11:23] upgraded libgsf-gnome (1.14.15-1 -> 1.14.16-1)
[2009-11-01 11:23] upgraded abiword (2.6.8-2 -> 2.8.1-1)
[2009-11-01 11:23] upgraded kernel-headers (2.6.30.5-1 -> 2.6.31.4-2)
[2009-11-01 11:23] Generating locales...
[2009-11-01 11:23] en_US.UTF-8... done
[2009-11-01 11:23] en_US.ISO-8859-1... done
[2009-11-01 11:23] Generation complete.
[2009-11-01 11:23] upgraded glibc (2.10.1-4 -> 2.10.1-5)
[2009-11-01 11:23] upgraded attr (2.4.43-1 -> 2.4.44-1)
[2009-11-01 11:23] upgraded acl (2.2.47-2 -> 2.2.48-1)
[2009-11-01 11:23] upgraded binutils (2.19.1-5 -> 2.20-1)
[2009-11-01 11:23] upgraded libx11 (1.2.2-1 -> 1.3.2-1)
[2009-11-01 11:23] upgraded xextproto (7.0.5-1 -> 7.1.1-1)
[2009-11-01 11:23] upgraded libxext (1.0.5-2 -> 1.1.1-1)
[2009-11-01 11:23] upgraded libxt (1.0.6-1 -> 1.0.7-1)
[2009-11-01 11:23] upgraded libxmu (1.0.4-2 -> 1.0.5-1)
[2009-11-01 11:23] upgraded libxpm (3.5.7-2 -> 3.5.8-1)
[2009-11-01 11:23] upgraded libxaw (1.0.6-1 -> 1.0.7-1)
[2009-11-01 11:23] upgraded inputproto (1.5.1-2 -> 2.0-1)
[2009-11-01 11:23] upgraded libxi (1.2.1-1 -> 1.3-1)
[2009-11-01 11:23] upgraded libxtst (1.0.3-2 -> 1.1.0-1)
[2009-11-01 11:23] upgraded python (2.6.3-2 -> 2.6.4-1)
[2009-11-01 11:23] upgraded brltty (4.0-3 -> 4.1-1)
[2009-11-01 11:23] upgraded dbus-glib (0.82-1 -> 0.82-2)
[2009-11-01 11:23] upgraded gstreamer0.10 (0.10.24-1 -> 0.10.25-1)
[2009-11-01 11:23] upgraded libxv (1.0.4-1 -> 1.0.5-1)
[2009-11-01 11:23] upgraded gstreamer0.10-base (0.10.24-1 -> 0.10.25-1)
[2009-11-01 11:23] upgraded libtheora (1.0-1 -> 1.1.0-1)
[2009-11-01 11:23] upgraded gstreamer0.10-base-plugins (0.10.24-1 -> 0.10.25-1)
[2009-11-01 11:23] upgraded gstreamer0.10-python (0.10.16-1 -> 0.10.17-1)
[2009-11-01 11:23] upgraded exaile (0.3.0.1-2 -> 0.3.0.1-3)
[2009-11-01 11:23] upgraded gcc-libs (4.4.1-1 -> 4.4.2-2)
[2009-11-01 11:23] upgraded xulrunner (1.9.1.3-2 -> 1.9.1.4-1)
[2009-11-01 11:23] upgraded firefox (3.5.3-1 -> 3.5.4-1)
[2009-11-01 11:23] upgraded fixesproto (4.0-3 -> 4.1.1-1)
[2009-11-01 11:23] upgraded gcc (4.4.1-1 -> 4.4.2-2)
[2009-11-01 11:23] upgraded gnome-disk-utility (2.28.0-1 -> 2.28.0-2)
[2009-11-01 11:23] upgraded gnome-games (2.28.1-1 -> 2.28.1-2)
[2009-11-01 11:23] upgraded gnutls (2.8.3-1 -> 2.8.4-1)
[2009-11-01 11:23] upgraded gstreamer0.10-ffmpeg (0.10.8-1 -> 0.10.9-1)
[2009-11-01 11:23] upgraded module-init-tools (3.10-2 -> 3.11.1-1)
[2009-11-01 11:23] >>> Updating module dependencies. Please wait ...
[2009-11-01 11:24] >>> MKINITCPIO SETUP
[2009-11-01 11:24] >>> ----------------
[2009-11-01 11:24] >>> If you use LVM2, Encrypted root or software RAID,
[2009-11-01 11:24] >>> Ensure you enable support in /etc/mkinitcpio.conf .
[2009-11-01 11:24] >>> More information about mkinitcpio setup can be found here:
[2009-11-01 11:24] >>> Configuring mkinitcpio - ArchWiki
[2009-11-01 11:24]
[2009-11-01 11:24] >>> Generating initial ramdisk, using mkinitcpio. Please wait...
[2009-11-01 11:24] ==> Building image "default"
[2009-11-01 11:24] ==> Running command: /sbin/mkinitcpio -k 2.6.31-ARCH -c /etc/mkinitcpio.conf -g /boot/kernel26.img
[2009-11-01 11:24] :: Begin build
[2009-11-01 11:24] :: Parsing hook [base]
[2009-11-01 11:24] :: Parsing hook [udev]
[2009-11-01 11:24] :: Parsing hook [autodetect]
[2009-11-01 11:24] :: Parsing hook [pata]
[2009-11-01 11:24] :: Parsing hook [scsi]
[2009-11-01 11:24] :: Parsing hook [sata]
[2009-11-01 11:24] :: Parsing hook [filesystems]
[2009-11-01 11:24] :: Generating module dependencies
[2009-11-01 11:24] :: Generating image '/boot/kernel26.img'...SUCCESS
[2009-11-01 11:24] ==> SUCCESS
[2009-11-01 11:24] ==> Building image "fallback"
[2009-11-01 11:24] ==> Running command: /sbin/mkinitcpio -k 2.6.31-ARCH -c /etc/mkinitcpio.conf -g /boot/kernel26-fallback.img -S autodetect
[2009-11-01 11:24] :: Begin build
[2009-11-01 11:24] :: Parsing hook [base]
[2009-11-01 11:24] :: Parsing hook [udev]
[2009-11-01 11:24] :: Parsing hook [pata]
[2009-11-01 11:24] :: Parsing hook [scsi]
[2009-11-01 11:24] :: Parsing hook [sata]
[2009-11-01 11:24] :: Parsing hook [filesystems]
[2009-11-01 11:25] :: Generating module dependencies
[2009-11-01 11:25] :: Generating image '/boot/kernel26-fallback.img'...SUCCESS
[2009-11-01 11:25] ==> SUCCESS
[2009-11-01 11:25] upgraded kernel26 (2.6.31.4-1 -> 2.6.31.5-1)
[2009-11-01 11:25] upgraded libatasmart (0.16-1 -> 0.17-1)
[2009-11-01 11:25] upgraded libdmx (1.0.99.1-1 -> 1.1.0-1)
[2009-11-01 11:25] upgraded libdrm (2.4.13-1 -> 2.4.15-1)
[2009-11-01 11:25] upgraded libfetch (2.25-1 -> 2.26-1)
[2009-11-01 11:25] upgraded libgksu (2.0.8-1 -> 2.0.12-1)
[2009-11-01 11:25] upgraded libgphoto2 (2.4.6-3 -> 2.4.7-1)
[2009-11-01 11:25] upgraded libmpcdec (1.2.6-1 -> 1.2.6-2)
[2009-11-01 11:25] upgraded libsexy (0.1.11-1 -> 0.1.11-2)
[2009-11-01 11:25] upgraded libxfixes (4.0.3-2 -> 4.0.4-1)
[2009-11-01 11:25] upgraded libxcomposite (0.4.0-2 -> 0.4.1-1)
[2009-11-01 11:25] upgraded libxdamage (1.1.1-2 -> 1.1.2-1)
[2009-11-01 11:25] upgraded libxdmcp (1.0.2-2 -> 1.0.3-1)
[2009-11-01 11:25] upgraded libxfont (1.4.0-1 -> 1.4.1-1)
[2009-11-01 11:25] upgraded libxfontcache (1.0.4-2 -> 1.0.5-1)
[2009-11-01 11:25] upgraded renderproto (0.9.3-2 -> 0.11-1)
[2009-11-01 11:25] upgraded libxrender (0.9.4-2 -> 0.9.5-1)
[2009-11-01 11:25] upgraded libxft (2.1.13-1 -> 2.1.14-1)
[2009-11-01 11:25] upgraded libxinerama (1.0.99.1-1 -> 1.1-1)
[2009-11-01 11:25] upgraded libxkbfile (1.0.5-1 -> 1.0.6-1)
[2009-11-01 11:25] upgraded libxres (1.0.3-2 -> 1.0.4-1)
[2009-11-01 11:25] upgraded libxvmc (1.0.4-2 -> 1.0.5-1)
[2009-11-01 11:25] upgraded libxxf86dga (1.1.1-0.1 -> 1.1.1-1)
[2009-11-01 11:25] upgraded libxxf86misc (1.0.1-2 -> 1.0.2-1)
[2009-11-01 11:25] upgraded libxxf86vm (1.0.99.1-1 -> 1.1.0-1)
[2009-11-01 11:25] upgraded live-media (2009.09.28-1 -> 2009.09.28-2)
[2009-11-01 11:25] upgraded xf86-input-evdev (2.2.5-1 -> 2.3.0-1)
[2009-11-01 11:25] installed printproto (1.0.4-1)
[2009-11-01 11:25] installed libxp (1.0.0-3)
[2009-11-01 11:25] upgraded xorg-server-utils (7.4-7 -> 7.5-2)
[2009-11-01 11:25] upgraded xorg-server (1.6.3.901-1 -> 1.7.1-1)
[2009-11-01 11:25] upgraded nvidia-utils (185.18.36-1 -> 190.42-1)
[2009-11-01 11:25] upgraded mesa (7.5.1-2 -> 7.6-2)
[2009-11-01 11:25] upgraded mlocate (0.22.2-1 -> 0.22.2-2)
[2009-11-01 11:25] upgraded nautilus (2.28.0-1 -> 2.28.1-1)
[2009-11-01 11:25] upgraded nautilus-cd-burner (2.24.0-2 -> 2.25.3-1)
[2009-11-01 11:25] upgraded ntfs-3g (2009.4.4-1 -> 2009.4.4-2)
[2009-11-01 11:25] In order to use the new nvidia module, exit Xserver and unload it manually.
[2009-11-01 11:25] upgraded nvidia (185.18.36-2 -> 190.42-1)
[2009-11-01 11:25] upgraded popt (1.14-1 -> 1.15-1)
[2009-11-01 11:25] upgraded python-gdata (1.3.1-1 -> 2.0.4-1)
[2009-11-01 11:25] upgraded pywebkitgtk (1.1.6-2 -> 1.1.7-1)
[2009-11-01 11:25] upgraded shared-mime-info (0.70-1 -> 0.70-2)
[2009-11-01 11:25] upgraded totem (2.28.1-1 -> 2.28.2-1)
[2009-11-01 11:25] upgraded transmission-gtk (1.75-1 -> 1.76-1)
[2009-11-01 11:25] upgraded vte (0.22.2-1 -> 0.22.3-1)
[2009-11-01 11:25] upgraded xorg-apps (7.5-1.1 -> 7.5-2)
[2009-11-01 11:25] upgraded xorg-font-utils (7.4-3 -> 7.5-2)
[2009-11-01 11:25] upgraded xorg-fonts-alias (1.0.1-2 -> 1.0.2-1)
[2009-11-01 11:25] Regenerating font encodings... done.
[2009-11-01 11:25] upgraded xorg-fonts-encodings (1.0.2-3 -> 1.0.3-1)
[2009-11-01 11:25] upgraded xorg-utils (7.4-4 -> 7.5-1)
[2009-11-01 11:25] upgraded xorg-xauth (1.0.3-1 -> 1.0.4-1)
[2009-11-01 11:25] upgraded xorg-xkb-utils (7.4-2 -> 7.5-1)
[2009-11-01 11:25] upgraded xterm (246-1 -> 250-1)
[2009-11-01 13:32] removed libgnomeprintui (2.18.4-1)
[2009-11-01 13:32] removed libgnomeprint (2.18.6-1)
[2009-11-05 17:12] synchronizing package lists
[2009-11-05 17:13] starting full system upgrade
[2009-11-05 17:18] starting full system upgrade
[2009-11-05 17:22] upgraded bash (4.0.033-1 -> 4.0.035-1)
[2009-11-05 17:22] upgraded bluez (4.56-1 -> 4.57-1)
[2009-11-05 17:22] upgraded curl (7.19.6-2 -> 7.19.7-1)
[2009-11-05 17:22] upgraded libwebkit (1.1.15.3-1 -> 1.1.15.3-2)
[2009-11-05 17:22] upgraded epiphany (2.28.1-1 -> 2.28.1-2)
[2009-11-05 17:22] upgraded gnome-commander (1.2.8.2-1 -> 1.2.8.3-1)
[2009-11-05 17:22] upgraded libldap (2.4.18-1 -> 2.4.19-1)
[2009-11-05 17:22] upgraded libnice (0.0.9-1 -> 0.0.10-1)
[2009-11-05 17:22] upgraded libtheora (1.1.0-1 -> 1.1.1-1)
[2009-11-05 17:22] upgraded libtiff (3.9.1-1 -> 3.9.2-1)
[2009-11-05 17:22] upgraded libxi (1.3-1 -> 1.3-2)
[2009-11-05 17:22] installed tk (8.5.7-1)
[2009-11-05 17:22] upgraded mercurial (1.3.1-1 -> 1.3.1-3)
[2009-11-05 17:22] upgraded nspr (4.8-1 -> 4.8.2-1)
[2009-11-05 17:23] when you use a non-reparenting window manager
[2009-11-05 17:23] set _JAVA_AWT_WM_NONREPARENTING=1 in
[2009-11-05 17:23] /etc/profile.d/openjdk6.sh
[2009-11-05 17:23] upgraded openjdk6 (1.6.1-1 -> 1.6.1-2)
[2009-11-05 17:23] upgraded pam (1.0.4-2 -> 1.1.0-1)
[2009-11-05 17:23] upgraded openssh (5.3p1-1 -> 5.3p1-2)
[2009-11-05 17:23] upgraded pcre (7.9-1 -> 8.00-1)
[2009-11-05 17:23] installed libpurple (2.6.3-2)
[2009-11-05 17:23] upgraded pidgin (2.6.3-1 -> 2.6.3-2)
[2009-11-05 17:23] upgraded sg3_utils (1.27-2 -> 1.28-1)
[2009-11-05 17:23] upgraded taglib (1.6-1 -> 1.6.1-1)
[2009-11-05 17:23] upgraded tzdata (2009o-1 -> 2009q-1)
[2009-11-05 17:23] installed telepathy-glib (0.9.1-1)
[2009-11-05 17:23] upgraded vinagre (2.28.1-1 -> 2.28.1-2)
[2009-11-05 17:23] upgraded vino (2.28.1-1 -> 2.28.1-2)
[2009-11-05 17:23] upgraded wavpack (4.50.1-1 -> 4.60.0-1)
[2009-11-05 17:23] upgraded xfsprogs (3.0.3-1 -> 3.0.5-1)
[2009-11-05 17:23] upgraded xorg-server-utils (7.5-2 -> 7.5-3)
[2009-11-05 17:50] upgraded ardentryst (1.7-1 -> 1.71-1)
[2009-11-06 13:41] installed finch (2.6.3-2)
[2009-11-07 08:55] installed tabbed-hg (130-1)
[2009-11-07 09:03] installed dmenu (4.0-1)
[2009-11-12 17:11] synchronizing package lists
[2009-11-12 17:12] starting full system upgrade
[2009-11-12 17:12] starting full system upgrade
[2009-11-12 17:17] installed gstreamer0.10-ugly (0.10.13-1)
[2009-11-12 17:17] installed libmpeg2 (0.5.1-1)
[2009-11-12 17:17] installed a52dec (0.7.4-4)
[2009-11-12 17:17] installed libsidplay (1.36.59-3)
[2009-11-12 17:17] installed gstreamer0.10-ugly-plugins (0.10.13-1)
[2009-11-12 17:19] synchronizing package lists
[2009-11-12 17:20] starting full system upgrade
[2009-11-12 17:21] starting full system upgrade
[2009-11-12 17:52]
[2009-11-12 17:52] ==> awesome installation notes:
[2009-11-12 17:52] ----------------------------------------
[2009-11-12 17:52] During some updates of awesome, the config file syntax changes heavily.
[2009-11-12 17:52] This means that your configuration file (~/.config/awesome/rc.lua) for
[2009-11-12 17:52] any other than this particular version of awesome may or may not work.
[2009-11-12 17:52]
[2009-11-12 17:52] upgraded awesome (3.4-1 -> 3.4.1-1)
[2009-11-12 17:52] upgraded kernel-headers (2.6.31.4-2 -> 2.6.31.5-1)
[2009-11-12 17:52] warning: /etc/locale.gen installed as /etc/locale.gen.pacnew
[2009-11-12 17:52] Generating locales...
[2009-11-12 17:52] en_US.UTF-8... done
[2009-11-12 17:52] en_US.ISO-8859-1... done
[2009-11-12 17:52] Generation complete.
[2009-11-12 17:52] upgraded glibc (2.10.1-5 -> 2.11-1)
[2009-11-12 17:52] upgraded binutils (2.20-1 -> 2.20-2)
[2009-11-12 17:52] upgraded brasero (2.28.2-1 -> 2.28.2-2)
[2009-11-12 17:52] upgraded bzip2 (1.0.5-4 -> 1.0.5-5)
[2009-11-12 17:52] upgraded devicekit-disks (008-1 -> 009-1)
[2009-11-12 17:52] upgraded dhcpcd (5.1.1-1 -> 5.1.3-1)
[2009-11-12 17:52] upgraded libxml2 (2.7.3-3 -> 2.7.6-1)
[2009-11-12 17:52] upgraded docbook-xml (4.5-2 -> 4.5-3)
[2009-11-12 17:52] upgraded gcc-libs (4.4.2-2 -> 4.4.2-3)
[2009-11-12 17:52] upgraded exempi (2.1.0-1 -> 2.1.1-1)
[2009-11-12 17:52] The package is built using sysv ipc. If this causes segfaults
[2009-11-12 17:52] see included README file and rebuild the pkg --with-ipc=tcp
[2009-11-12 17:52] upgraded fakeroot (1.13.1-1 -> 1.14.3-1)
[2009-11-12 17:52] upgraded filezilla (3.2.8.1-1 -> 3.3.0-1)
[2009-11-12 17:52] upgraded gnutls (2.8.4-1 -> 2.8.5-1)
[2009-11-12 17:52] upgraded libpng (1.2.39-1 -> 1.2.40-1)
[2009-11-12 17:52] upgraded libcups (1.4.1-1 -> 1.4.2-1)
[2009-11-12 17:52] upgraded gtk2 (2.18.3-1 -> 2.18.3-2)
[2009-11-12 17:52] upgraded xulrunner (1.9.1.4-1 -> 1.9.1.5-1)
[2009-11-12 17:52] upgraded firefox (3.5.4-1 -> 3.5.5-1)
[2009-11-12 17:52] upgraded freetype2 (2.3.9-2 -> 2.3.11-1)
[2009-11-12 17:52] upgraded fuse (2.8.0-1 -> 2.8.1-1)
[2009-11-12 17:52] upgraded gcc (4.4.2-2 -> 4.4.2-3)
[2009-11-12 17:52] upgraded gnome-disk-utility (2.28.0-2 -> 2.28.1-1)
[2009-11-12 17:52] upgraded gnome-panel (2.28.0-1 -> 2.28.0-2)
[2009-11-12 17:52] upgraded gnome-python-extras (2.25.3-2 -> 2.25.3-3)
[2009-11-12 17:52] upgraded gstreamer0.10-good (0.10.16-1 -> 0.10.16-2)
[2009-11-12 17:52] upgraded hal (0.5.13-2 -> 0.5.13-3)
[2009-11-12 17:52] upgraded libdv (1.0.0-2 -> 1.0.0-3)
[2009-11-12 17:52] upgraded libv4l (0.6.0-1 -> 0.6.3-1)
[2009-11-12 17:52] upgraded gstreamer0.10-good-plugins (0.10.16-1 -> 0.10.16-2)
[2009-11-12 17:52] upgraded gtkspell (2.0.15-1.1 -> 2.0.16-1)
[2009-11-12 17:52] upgraded imagemagick (6.5.6.10-1 -> 6.5.7.5-1)

Edit:
I noticed, as root I do have sound.. :S

A question with respect to IRQ distribution

sunnyhay — Fri, 13 Nov 2009 17:26:19 GMT

Hi all,

I am a Linux newbie in kernel development. In class I was asked a question about IRQ distribution as followed.
IRQs can be distributed among the available CPUs (in a multiple-CPU system) in two ways:
(1) Static distribution
The IRQ signal is delivered to the local APICs listed in the corresponding Redirection Table entry. The interrupt is delivered to one specific CPU, to a subset of CPUs, or to all CPUs at once (broadcast mode).
(2) Dynamic distribution
The IRQ signal is delivered to the local APIC of the processor that is executing the process with the lowest priority. Task Priority Register (TPR) is used to compute the priority of the currently running process.
So I wonder which one is used, or both (how if so) in Linux kernel 2.6 or higher. I know this is a silly question but I did not find out any clear indication of the IRQ load-balance problem.
Thank you very much!

Help with compiling new kernel..

abhiram7 — Fri, 13 Nov 2009 05:59:52 GMT

Hi

I am trying to write a simple system call in the 2.6 linux kernel. While running the make mrproper command to recompile the kernel i am getting the following error

scripts/Makefile.clean:17: /usr/src/linux-2.6.31.6/block/mycall/Makefile: No such file or directory
make[1]: *** No rule to make target `/usr/src/linux-2.6.31.6/block/mycall/Makefile'. Stop.
make: *** [_clean_block/mycall] Error 2

What does block directory contain in linux? Do i need to add my syscall file( mycall) to this directory too???

Can someone help..

thanks

socket or not from struct file

persavon — Tue, 10 Nov 2009 13:55:01 GMT

Hi,

I want to be able to differentiate between disk I/O and network I/O from vfs_read(). This function takes a pointer to a struct file as a parameter. My question is how to figure whether that file is a socket or a physical file on disk. Thanks.

Rx/TX irqs are disabled during access to ATA

kuzba — Tue, 10 Nov 2009 09:39:40 GMT

Hello all,

I have powerpc system with compactflash card connected over ata. On this system is running process using raw socket to send and receive cyclic data each 4ms. When I start another iptraffic - sshftp - to write to compact flash, after certain time the IRQ-135 of ata driver locks task irqs from network card for cca 300ms. And my cyclic data transfer is broken. Any idea how to solve this problem?

Thanks a lot for hint. Jan

Output from sched_switch trace here:

pniodhld-2516 [000] 1305.162746: 2516: 33:T ==> [000] 6: 33:R sirq-net-rx/0
sirq-net-rx/0-6 [000] 1305.162755: 6: 33:R + [000] 2516: 33:T pniodhld
sirq-net-rx/0-6 [000] 1305.162766: 6: 49:R ==> [000] 2516: 33:S pniodhld
pniodhld-2516 [000] 1305.162835: 2516: 33:D ==> [000] 6: 49:R sirq-net-rx/0
sirq-net-rx/0-6 [000] 1305.162923: 6: 49:D ==> [000] 99:115:R ata/0
ata/0-99 [000] 1305.163050: 99:115:D ==> [000] 2409:120:R sftp-server
sftp-server-2409 [000] 1305.163100: 2409:120:R + [000] 710: 49:D IRQ-135
sftp-server-2409 [000] 1305.163119: 2409:120:R ==> [000] 710: 49:R IRQ-135
IRQ-135-710 [000] 1305.163227: 710: 49:D + [000] 710: 49:D IRQ-135
IRQ-135-710 [000] 1305.163324: 710: 49:D ==> [000] 2409:120:R sftp-server
sftp-server-2409 [000] 1305.163344: 2409:120:R + [000] 710: 49:D IRQ-135
sftp-server-2409 [000] 1305.163355: 2409:120:R ==> [000] 710: 49:R IRQ-135
IRQ-135-710 [000] 1305.163448: 710: 49:D + [000] 710: 49:D IRQ-135
IRQ-135-710 [000] 1305.163543: 710: 49:D ==> [000] 2409:120:R sftp-server
sftp-server-2409 [000] 1305.163562: 2409:120:R + [000] 710: 49:D IRQ-135
sftp-server-2409 [000] 1305.163574: 2409:120:R ==> [000] 710: 49:R IRQ-135
IRQ-135-710 [000] 1305.163634: 710: 49:R + [000] 4: 49:D sirq-timer/0
IRQ-135-710 [000] 1305.163768: 710: 49:D + [000] 710: 49:D IRQ-135
IRQ-135-710 [000] 1305.163870: 710: 49:D ==> [000] 4: 49:R sirq-timer/0
sirq-timer/0-4 [000] 1305.163894: 4: 49:R + [000] 710: 49:D IRQ-135
sirq-timer/0-4 [000] 1305.163918: 4: 49:D ==> [000] 710: 49:R IRQ-135

help with elf format

julvr2 — Mon, 09 Nov 2009 19:34:28 GMT

I have a problem that I'm out of my depth with, and was wondering if anyone could offer some suggestions on how to go foraward.

I have recently upgraded to the 2.6.30 kernel, and after that, I decided to update the toolchain (to GCC 4.4.2). The problem is that with the new toolchain, busybox (and any other program I've tried) exits immediately with a code 11 (SIGSEGV). (which causes a kernel panic, as it's the init task)

Everything works fine with the old toolchain, but as soon as I switch toolchains, I get my crash.

I did a dump of the elf file headers, and the only differences I found in the header were:

old toolchain:
e_shoff=0x00403ab0
e_shentsize=8

new toolchain:
e_shoff=004038e0
e_shentsize=9 (I'm not sure if this is a problem)

Also, and objdump shows a stack section in the new toolchain which did not show up in the old one:

STACK off 0x00000000 vaddr 0x00000000 paddr 0x00000000 align 2**2
filesz 0x00000000 memsz 0x00000000 flags rwx

My platform is a mips board. I've tried everything I can think of, but still have no clue why this does not work.

If anyone has any clues, I would be very greatful.

John

How does LD find /lib/tls/i686/cmov?

jdaviestx — Sun, 08 Nov 2009 22:15:13 GMT

I'm running Debian Lenny/Sid, BTW. More and more, I've been (forced to) download binary-only (no source code available) software for my linux system (for example, Flash and the Android SDK). The problem is, it's all failing lately with:

/lib/tls/i686/cmov/libc.so.6: version `GLIBC_2.4' not found

I look at /lib/tls/i686/cmov/libc.so.6 with objdump (it doesn't work with nm - (?)), and I see:

debian:tools$ objdump -p /lib/tls/i686/cmov/libc.so.6
...
13 0x00 0x0d696913 GLIBC_2.3
GLIBC_2.2.6
14 0x00 0x09691972 GLIBC_2.3.2
GLIBC_2.3
15 0x00 0x09691973 GLIBC_2.3.3
GLIBC_2.3.2
16 0x00 0x09691974 GLIBC_2.3.4
GLIBC_2.3.3
...

And no GLIBC_2.4.

So... I know I'm going to have to upgrade libc (although I've seen some pretty dire warnings against doing so). However, I'm still just trying to figure out where this is coming from!

LD_LIBRARY_PATH contains:

debian:tools$ echo $LD_LIBRARY_PATH
/usr/local/Trolltech/Qt-4.3.4/lib:/usr/local/lib

not /lib/tls/i686/cmov - and /etc/ld.so.conf.d contains one file whose contents also don't include this mysterious /lib/tls/i686/cmov:

debian:ld.so.conf.d$ ls
i486-linux-gnu.conf
debian:ld.so.conf.d$ cat i486-linux-gnu.conf
# Multiarch support
/lib/i486-linux-gnu
/usr/lib/i486-linux-gnu

Where the heck is "/lib/tls/i686/cmov" coming from? Is this just some "magic" inside ld.so? I don't see it documented anywhere - the man page says it should be looking in the following order:

Using the environment variable LD_LIBRARY_PATH
(LD_AOUT_LIBRARY_PATH for a.out programs). Except if the exe‐
cutable is a setuid/setgid binary, in which case it is ignored.

o From the cache file /etc/ld.so.cache which contains a compiled
list of candidate libraries previously found in the augmented
library path.

o In the default path /usr/lib, and then /lib.

I see a /lib/libc.so.6, but it's not a symlink to /lib/tls/i686/cmov...