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

Re: inittab is not executing after my Linux box boots up
by Prajosh Premdas June 21, 2011

June 21, 2011

Hi I have copied the generated rsa key and placed it in my nfs root filesystem. I find the problem still persists Please see the last leg of the logs below *eth0: link up (100/Full)* *VFS: Mounted root (nfs filesystem) on device 0:12.* *Freeing init memory: 128K* *Starting network...* *ip: RTNETLINK answers: File exists* *Starting dropbear sshd: OK* * * I feel the already generated key has been detected. And still the initab doesn't execute. Please advice On Tue, Jun 21, 2011 at 11:15 AM, Vladimir Murzin <murzin.v(a)gmail.com>wrote: > On Mon, Jun 20, 2011 at 10:24 AM, Prajosh Premdas > <premdas.prajosh(a)gmail.com> wrote: > > Hi Dave and Mulyadi > > Can you tell me where to place the previously generated RSA key. I even > > suspected this but this line > > Starting dropbear sshd: generating rsa key... generating dsa key... OK > > > > and the OK pulled me off the track. One more thing can you tell me what > > does the OK mean there... > > My mask is mask=255.0.0.0 so it is able to get the NFS server location > > I typed a wrong command after changing the commands i find no difference > > On Sun, Jun 19, 2011 at 11:22 PM, Dave Hylands <dhylands(a)gmail.com> > wrote: > >> > >> Hi Projish, > >> > >> On Sun, Jun 19, 2011 at 8:24 AM, Prajosh Premdas > >> <premdas.prajosh(a)gmail.com> wrote: > >> > Hi > >> > Please find the entire boot sequence here > >> > RomBOOT > >> ...snip... > >> > TCP cubic registered > >> > NET: Registered protocol family 17 > >> > rtc-at91sam9 at91_rtt.0: hctosys: unable to read the hardware clock > >> > IP-Config: Guessing netmask 255.0.0.0 > >> > IP-Config: Complete: > >> > device=eth0, addr=10.220.4.200, mask=255.0.0.0, gw=10.220.4.35, > >> > host=10.220.4.200, domain=, nis-domain=(none), > >> > bootserver=255.255.255.255, rootserver=10.0.0.1, rootpath= > >> > eth0: link up (100/Full) > >> > VFS: Mounted root (nfs filesystem) on device 0:12. > >> > >> So it looks like your root file system was mounted successfully. > >> > >> > Freeing init memory: 128K > >> > >> All message from this point onwards come from init. > >> > >> > Starting network... > >> > ip: RTNETLINK answers: File exists > >> > Starting dropbear sshd: generating rsa key... generating dsa key... OK > >> > >> And since sshd is starting, it's actually processing stuff from your > >> root file system. > >> > >> You may want to try disabling ssh. Generating the rsa key can take > >> several minutes, depending on how fast your target is. So it may look > >> like the system is hung, but it will actually continue to boot. > >> > >> You may want save away the generated RSA key and include it in your > >> generated image so that this step is faster. For a given filesystem, > >> it only happens the first time, and once the generated key is stored, > >> ssh will just use the previously generated key. But if you're > >> constantly rebuilding your filesystem without any generated keys then > >> the first time you boot the new filesystem you'lll gget a long delay. > >> > >> -- > >> Dave Hylands > >> Shuswap, BC, Canada > >> http://www.davehylands.com > > > > > > > > -- > > Regards, > > > > Prajosh Premdas > > > > _______________________________________________ > > Kernelnewbies mailing list > > Kernelnewbies(a)kernelnewbies.org > > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > > > > > > Hi Prajosh, > > I believe, that generating keys by sshd involve blocked operation on > /dev/random. > It's strict requirement to use /dev/random instead of /dev/urandom in such > case. > > Actually, it's well-known problem in embedded world to fill up entropy > pool, because of > lack random events into some solutions. > > I've heard it's possible to save and restore entropy pool, but it's > not safe, I think. > In my opinion, Dave's suggestion about saving keys into image is quite > suitable, as > soon as you aren't intend to distribute you solution on the market. > > P.S. Sorry for my English. > -- Regards, Prajosh Premdas -- Regards, Prajosh Premdas

2 1

request_mem_region
by Daniel Redway June 21, 2011

June 21, 2011

Hello As I understand ioremap() maps and returns virtual address for a given physical(bus) address associated with my PCI card. That is cool and clear. What is the purpose of request_mem_region()? They both take physical address as a parameter, but do not return any virt addresses. 1. What do they allocate/reserve and where? 2. Should I call reques_mem_region() before doing ioremap() or not? thanks

2 2

(no subject)
by Venkateswarlu P June 19, 2011

June 19, 2011

how to understand the kernerl source files in a simple way what header files i have to understand first for example to understand do_fork() function for process creation which is defined in kernel/fork.c -- ** P VENKATESWARLU M.Tech (CSE) II year NIT Calicut ,Kerala* ** *+91-9995115752

6 5

tools to identify linux kernel panics and system hangs
by Amit Nagal June 19, 2011

June 19, 2011

Hi , What are the best tools to trace linux kernel panics and system hangs for linux kernel . please recommend for x86 as well as arm platforms . regards Amit Nagal

5 8

mmap to fixed adr
by Volker Poplawski June 19, 2011

June 19, 2011

Hi, is there a VA range in userspace which is always safe for mmap() with fixed adr? (x86_64 mostly) I.e. a VA range where mmap() or ld will never map memory. I know I can try particular fixed adr with mmap() and check if the mapping succeeds. However this is not what I want. I want mmap() to stay clear of certain ranges and map memory there myself. Regards .....Volker

2 1

The problem with the installation of the latest kernel
by david June 19, 2011

June 19, 2011

Hi, guys: I am totally a beginner to Linux kernel and feel sorry to bother you with such an easy question. I encountered a problem which is failed to making the latest kernel 2.6.39.1 downloaded from kernel.org bootable under the environment of Ubuntu 10.10. Although I have tried to solve this problem for almost 3 days, I don't have any clue on the cause of not able to boot the newly-installed kernel. The procedure that I adopted to configure, build and install the kernel is shown bellow 1. I downloaded the tarball "linux-2.6.39.1.tar.bz2" and stored it in the directory "/home/lenovo/linux", then uncompressed it, a new folder "linux-2.6.39.1" was created. 2. I created a folder called "linux-2.6.39.1-install" in the same directory to store the compiled files from folder "linux-2.6.39.1". 3. I opened the terminal and typed in: cd ~/linux/linux-2.6.39.1 make O=../linux-2.6.39.1-install defconfig make -j4 O=../linux-2.6.39.1-install The result of these operations is: GZIP arch/x86/boot/compressed/vmlinux.bin.gz MKPIGGY arch/x86/boot/compressed/piggy.S AS arch/x86/boot/compressed/piggy.o LD arch/x86/boot/compressed/vmlinux ZOFFSET arch/x86/boot/zoffset.h OBJCOPY arch/x86/boot/vmlinux.bin AS arch/x86/boot/header.o LD arch/x86/boot/setup.elf OBJCOPY arch/x86/boot/setup.bin BUILD arch/x86/boot/bzImage Root device is (7, 0) Setup is 13148 bytes (padded to 13312 bytes) System is 4091 kB CRC 887907a3 Kernel: arch/x86/boot/bzImage is ready (#1) 4. I typed in the command: make O=../linux-2.6.39.1-install modules_install The result is: INSTALL arch/x86/kernel/test_nx.ko INSTALL drivers/scsi/scsi_wait_scan.ko INSTALL drivers/watchdog/hpwdt.ko INSTALL net/netfilter/xt_mark.ko DEPMOD 2.6.39.1 5. sudo make O=../linux-2.6.39.1-install install The result is: sh /home/lenovo/linux/linux-2.6.39.1/arch/x86/boot/install.sh 2.6.39.1 arch/x86/boot/bzImage \ System.map "/boot" 6. sudo update-initramfs -c -k 2.6.39.1 The result is: update-initramfs: Generating /boot/initrd.img-2.6.39.1 Warning: No support for locale: en_US.utf8 7. sudo update-grub The result is: Found linux image: /boot/vmlinuz-2.6.39.1 Found initrd image: /boot/initrd.img-2.6.39.1 Finally, after I rebooted the system and selected "Linux-2.6.39.1" in the grup, a problem arose along with a message showed up saying: mount: mounting none on /dev failed: No such device. w: devtmpfs not available, falling back to tmpfs for /dev fuse: device not found, try 'modprobe' fuse first. ...... Could not mount the partition /dev/sda8. ...... (file system = ntfs, error code = 21) best regards david.G

4 3

3.0.0-rc? kernel suddenly loses sound on my ubuntu 11.10 system
by Robert P. J. Day June 18, 2011

June 18, 2011

i've been asking about this on the ubuntu list recently, but i figured i'd throw this out here as well to see if anyone else has a solution to this. recently, i did a release upgrade of my ubuntu 10.10 system, on which i was running a hand-rolled 2.6.39-rc7-dirty kernel, for which sound worked just fine (intel hda). i also identically configured and built a newer 3.0.0-rcsomething kernel, but when i boot that newer kernel, i have no sound. in particular, when i go under System->Preferences->Sound, the list of sound output devices is empty. if i reboot to the slightly older kernel, sound is back. so the difference is clearly something in the kernel, and i'm wondering if anyone has run into this and has any suggestions or pointers to ALSA-related changes that might be causing this. thanks. rday -- ======================================================================== Robert P. J. Day Ottawa, Ontario, CANADA http://crashcourse.ca Twitter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================

2 2

Re: 回复： The problem with the installation of the latest kernel
by Prashant Shah June 18, 2011

June 18, 2011

Hi, On Fri, Jun 17, 2011 at 8:10 PM, david <gwj35(a)yahoo.cn> wrote: > HI: > I have tried this command, but it asked me a lot of other new config > symbols which I don't know whether to include, exclude or make them modules > to the kernel. Just hit enter - it will take the default value.

1 0

How to traverse the linked list of kernel modules
by amit mehta June 18, 2011

June 18, 2011

kernel modules are represented by the data structure(struct module) as defined in ~include/linux/module.h, and they seem to be connected by a doubly circular link list. Now suppose I want to access all the elements in this linked list of kernel modules(modules that show up in /proc/modules), then how should i go about doing this ? Regards, amit

2 1

some defines I couldnt understand '_IOC_NONE 1U'
by Daniel Redway June 18, 2011

June 18, 2011

Hello I dont understand what means 'U' character. I found this defines in asm/ioctl.h file. Thanks #define _IOC_NONE 1U #define _IOC_READ 2U #define _IOC_WRITE 4U

2 1