• State
  • Locked Locked
  • Replies 9 replies
  • Subscribers 101 subscribers
  • Views 1150 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

montavista linux start within 2 second

king king
Intellectual 650 points

 montavista  linux 2.6.18 on dm365.  how to start it within 2 second.

 

U-Boot 1.2.0 (Jan  6 2010 - 16:40:36)

 

I2C:   ready

DRAM:  256 MB

MY AMD Flash: 16 MB

In:    serial

Out:   serial

Err:   serial

ARM Clock :- 495MHz

DDR Clock :- 330MHz

Hit any key to stop autoboot:  0 

## Booting image at 42040000 ...

   Image Name:   Linux-2.6.10_mvl401-davinci_evm-

   Image Type:   ARM Linux Kernel Image (uncompressed)

   Data Size:    736252 Bytes = 719 kB

   Load Address: 80008000

   Entry Point:  80008000

   Verifying Checksum ... OK

OK

 

Starting kernel ...

 

Uncompressing Linux.................................................... done, booting the kernel.

Linux version 2.6.10_mvl401-davinci_evm-PSP_01_30_00_070 (root@localhost.localdomain) (gcc version 3.4.3 (MontaVista 3.4.3-25.0.104.0600975 2006-07-06)) #12 Sun Oct 24 10:28:34 CST 2010

CPU: ARM926EJ-Sid(wb) [41069265] revision 5 (ARMv5TEJ)

CPU0: D VIVT write-back cache

CPU0: I cache: 16384 bytes, associativity 4, 32 byte lines, 128 sets

CPU0: D cache: 8192 bytes, associativity 4, 32 byte lines, 64 sets

Machine: DaVinci DM6467 EVM

Memory policy: ECC disabled, Data cache writeback

DM6467

Built 1 zonelists

Kernel command line: mem=120M console=ttyS0,115200n8 root=/dev/mtdblock3 rootfstype=jffs2 noinitrd rw video=dm64xxfb:output=pal

PID hash table entries: 512 (order: 9, 8192 bytes)

Console: colour dummy device 80x30

Dentry cache hash table entries: 16384 (order: 4, 65536 bytes)

Inode-cache hash table entries: 8192 (order: 3, 32768 bytes)

Memory: 120MB = 120MB total

Memory: 119936KB available (1205K code, 283K data, 128K init)

Mount-cache hash table entries: 512 (order: 0, 4096 bytes)

CPU: Testing write buffer coherency: ok

spawn_desched_task(00000000)

desched cpu_callback 3/00000000

ksoftirqd started up.

desched cpu_callback 2/00000000

desched thread 0 started up.

Linux NoNET1.0 for Linux 2.6

DaVinci: 48 gpio irqs

Registering platform device 'serial8250.0'. Parent at platform

Registering platform device 'musb_hdrc'. Parent at platform

Registering platform device 'nor_davinci.0'. Parent at platform

DaVinci I2C DEBUG: 10:28:05 Oct 24 2010

Registering platform device 'i2c'. Parent at platform

SCSI subsystem initialized

JFFS2 version 2.2. (C) 2001-2003 Red Hat, Inc.

Registering platform device 'DavinciHD_vdce.2'. Parent at platform

Registering platform device 'relay_ctrl_dvs'. Parent at platform

Registering platform device 'tlp121'. Parent at platform

Major:253  Minor:0

chdrv: Driver initialized.

Serial: 8250/16550 driver $Revision: 1.90 $ 2 ports, IRQ sharing disabled

Registering platform device 'serial8250'. Parent at platform

ttyS0 at MMIO 0x1c20000 (irq = 40) is a ST16654

ttyS1 at MMIO 0x1c20400 (irq = 41) is a ST16650V2

io scheduler noop registered

RAMDISK driver initialized: 1 RAM disks of 8192K size 1024 blocksize

i2c /dev entries driver

elevator: using noop as default io scheduler

DaVinci flash: probing 16-bit flash bus

DaVinci flash: Found 1 x16 devices at 0x0 in 16-bit bank

 Amd/Fujitsu Extended Query Table at 0x0040

DaVinci flash: CFI does not contain boot bank location. Assuming top.

number of CFI chips: 1

cfi_cmdset_0002: Disabling erase-suspend-program due to code brokenness.

Creating 4 MTD partitions on "DaVinci flash":

0x00000000-0x00020000 : "bootloader"

0x00020000-0x00040000 : "params"

0x00040000-0x00240000 : "kernel"

0x00240000-0x01000000 : "filesystem"

mice: PS/2 mouse device common for all mice

 

before this the speed is quick enouch.

but after this the speed is slow. so the total time is long about 10second.

 

VFS: Mounted root (jffs2 filesystem).

Freeing init memory: 128K

Warning: unable to open an initial console.

 

MontaVista(R) Linux(R) Professional Edition 4.0.1 (0600980)

 

(none) login: 

I have cut the driver of network and other a lot of drivers.
help me thanks!!

  • 0
    Prodigy 50 points

    use the cramfs filesystme(read only) instead of jffs2 filesystem! 

  • 0 in reply to DAN LIU44148
    Intellectual 650 points

    do you have other way to achieve it ?

     

  • 0
    TI__Genius 9490 points

    King King,

    One observation that I wanted to point out from your log from the Hardware perspective was that it mentions the ARM/DDR speeds to be ARM Clock :- 495MHz & DDR Clock :- 330MHz, and this is a violation of the DM365 device specification of maximum of 300MHZ ARM, 270MHZ DDR.  Why are you running the device at these speeds?

    regards,

    miguel

  • 0 in reply to Miguel A. Alanis
    Expert 2315 points
    Hi Miguel, I am trying to sort out issues with my dm365evm booting a modified smaller filesystem that I made up from the demo one that cam with the evm where I added ssh and got it working. I recently went through a huge exercise to change from ubuntu 9.04 to 10.04 and from my backups the small filesystem I am using to boot from NAND is giving me some problems with ssh logging in. My investigations is also looking on the kernel log and the following output led me to this thread : "Warning: unable to open an initial console." Do you perhaps know why it get printed by the kernel? I am struggling to get the ssh consitant and would appreciate help on what I should do if I just want it to be able to open a proper telnet session for me - xinet only? Thanks, Jinh T.
  • 0 in reply to Jinh
    TI__Expert 4590 points

    I would suggest that you go through this wiki page as well: http://processors.wiki.ti.com/index.php/Optimize_Linux_Boot_Time

    Though it uses AM37x for illustration, most of the concepts are generic Linux and should apply even to your kernel as well.

  • 0 in reply to Jinh
    TI__Expert 4590 points

    Jinh said:
    My investigations is also looking on the kernel log and the following output led me to this thread : "Warning: unable to open an initial console." Do you perhaps know why it get printed by the kernel?

    This message indicates that "/dev/console" doesn't exist. The bare filesystem should have /dev/null and /dev/console. (This may not be true going forward - but it is safe to assume for this discussion). The device nodes are "special" files and cannot be created/copied by normal user privileges. You may want to check for either of these:

    • The file /dev/console really exists.
    • If the filesystem was extracted - from packaged tarball (or equiv) - as a normal user. Though file does exist, it is not a "device" (using term device quite liberally).
    • Is the target medium containing the filesystem 'writeable'. 'console" should be opened in write mode. Some filesystems create all devices on a tmpfs as part of the boot process to overcome the "read-only" limitation.

    $ ls -l /dev | grep console
    crw-------.  1 root root      5,   1 Jun 28  2011 console

  • 0 in reply to Sanjeev Premi
    Expert 2315 points
    (sorry about formatting but I don't know what is going on) Thanks Sanjeev, I got the kernel happy with 'dev and my telnetd working properly but picked up problems with ssh that was not there on the previous setup of my customised filesystem that came from the ubuntu 9.04 machine. I tar'ed it and use it again on my new ubuntu 10.04 machine and initially had problems with keys but now my dm365 boots up and when my app thats starts from flash is finished init and starts to run and tries to sync with my incoming video fram the kernel in the background is struggling with rsa keys and interferes with my app trying to sync. Eventually after the login: appears and ssh is silent my app syncs and runs fine. The following is the text that the kernel spits out during its workings and I don't know what changed apart from having now a new host pc and old filesystem coming from the old host pc. Ths is the print that appears on the console between the text my app spits out to indicate its status (I have removed it as it just shows timing data meassured between video frames - which are out of spec as I am doing realtime sync on video frames received and with ssh stealing processor bandwidth, struggles until ssh goes to sleep) , ###################Please see attachment of output I get on terminal.################################## , Please help me to get the dm365 (my host?) and the pc (client want to ssh to dm365?) happy with each other's key again. Thanks, Jinh T.
  • 0 in reply to Jinh
    TI__Expert 4590 points

    Don't worry about formatting.

    To understand you better, the SSH related messages continue to appear while the video application is running. Right?

    Few questions:
    1. Does your application use any SSH based auth at run time?
    2. Do the keys exchanged remain same OR do you notice any change in the vaues.
    3. What are the permissions to the files where SSH keys are being saved.

    Next time you run the application, can you run either/both of these commands to see who could be using "ssh"
    ps aux | grep ssh
    ps ef | grep ssh

  • 0 in reply to Sanjeev Premi
    Expert 2315 points
    Hi Sanjeev, thanks for the response. This looks like one of those situations where lots of things goes wrong and change but now it is working like before my upgrade. I was hoping it should or assuming it should as soon as all the hosts and clients have generated their new key - as far as I understand it. To answer the questions you had in short, my application does not use any ssh, I just use it as a back door to be able to look at the board when it boots from NAND and maybe gives problems or I want to restart the app manually without having to do a complete reboot which takes too much time. Secondly I assume now the keys that have been exchanged have ended where they should be because when I boot from NAND again (i was for a while booting the ramdisk from the host but now it is in flash and the board boots standalone) the login: prompt comes up before my application starts to run and it synchronises with the first bunch of video frames that it gets without missing frames where the kernel is busy sorting out keys. I am still confused where the keys are then because the ramdisk did noet get updated recently and those keys it would only have written to the filesystme on the host when I was still in NFS mode and after that only made the ramdisk which is used now. Thisrdly when II run the ps cmd's I get "641 root 3952 SW /usr/sbin/sshd". I am continuing on my other box again now but will keep an eye on this issue and post again if there are problems with my filesystem booting up. Thanks, Jinh T.