From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from SG2EHSOBE001.bigfish.com (outbound-sin.frontbridge.com [207.46.51.80]) by ozlabs.org (Postfix) with ESMTP id 78DB1DDF09 for ; Thu, 14 Aug 2008 02:43:34 +1000 (EST) MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C8FD63.B1E49866" Subject: RE: ml507 initrd problem Date: Wed, 13 Aug 2008 10:43:20 -0600 In-Reply-To: <8BB04FAC749BD24BBA250E9D09F2F18B220AD6@MAIL-SJC.quantum3d.com> References: <8BB04FAC749BD24BBA250E9D09F2F18B220ABB@MAIL-SJC.quantum3d.com> <20080813141107.92D4B948071@mail62-wa4.bigfish.com> <8BB04FAC749BD24BBA250E9D09F2F18B220AD6@MAIL-SJC.quantum3d.com> From: John Linn To: "Stu Bershtein" , Message-ID: <20080813164323.3F51F23004D@mail93-sin.bigfish.com> List-Id: Linux on Embedded PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , ------_=_NextPart_001_01C8FD63.B1E49866 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: quoted-printable Stu, = I'm assuming you're using arch/powerpc. The command line we use in the device tree file, arch/powerpc/boot/dts/ml507.dts, is bootargs =3D "console=3DttyS0 ip=3Don root=3D/dev/ram"; = It sounds like you're getting console output til it switches from kernel to user in the ram disk, but your message is a little confusing on that. You should expect the shell output to be on the same serial port at the same baud rate as the boot output of the kernel. = Yes the ML507 uses the 9 pin serial connector at 9600 baud. = We have removed the linux_support*.gz file when I put the bit streams and ram disk on the wiki site. If you have these files you have everything. = You should be able to following the directions on the wiki site to get a baseline working, although I didn't specifically say on the wiki to put the ram disk image in the boot directory before you build or the make command line. = The ram disk on our site should work correctly if you are using our kernel with the default configuration, ml507_defconfig. I always advice people to get a good baseline with our stuff then make changes after you see if baselined. = -- John = ________________________________ From: Stu Bershtein [mailto:sbershtein@Quantum3D.com] = Sent: Wednesday, August 13, 2008 9:34 AM To: John Linn; linuxppc-embedded@ozlabs.org Subject: RE: ml507 initrd problem = Morning John, = Thanks for the tip. Running said ramdisk things certainly look lively, I'll add a list of opens that I see at boot up at the bottom of this email. Looks like it's opening a shell and everything and yet I see no serial communication from the shell, just quiet. Am I correct to expect the console to be the serial port (9pin DB) on the ml507 board at 9600 baud? Also, if you could spare a desperate person a moment, could you send me the boot command line that you use for testing with the ramdisk? Finally, another member of this form suggested that there is a fully build BSP (rootfs and all) available as 'linux_support.tar.gz' on the Xilinx site. Are you familiar with this? I tried locating it a few days ago to no avail. = Thanks very much for your help! Stu = Freeing unused kernel memory: 144k init open /dev/console code 10000000 - 1009c134, data 1009a128 - 1009c134, stack bfdb9f00 open /lib/libcrypt.so.0 open /lib/libm.so.0 open /lib/libc.so.0 open /dev/console open /dev/console open /etc/TZ open /etc/inittab open /dev/console open /etc/TZ code 10000000 - 1009c134, data 1009a128 - 1009c134, stack bfd65eb0 open /lib/libcrypt.so.0 open /lib/libm.so.0 open /lib/libc.so.0 open /etc/rc.sh code 10000000 - 1009c134, data 1009a128 - 1009c134, stack bfef2ea0 open /lib/libcrypt.so.0 open /lib/libm.so.0 open /lib/libc.so.0 open /etc/mtab code 10000000 - 1009c134, data 1009a128 - 1009c134, stack bf9a8e90 open /lib/libcrypt.so.0 open /lib/libm.so.0 open /lib/libc.so.0 code 10000000 - 10039640, data 10039000 - 10039640, stack bfa47e70 open /lib/libm.so.0 open /lib/libcrypt.so.0 open /lib/libc.so.0 open /dev/null open /tmp/xinetd.pid open /etc/xinetd.conf open /etc/xinetd.d open /dev/console open /etc/TZ open /etc/xinetd.d/ftpd open /dev/console code 10000000 - 1009c134, data 1009a128 - 1009c134, stack bfda3ea0 open /etc/TZ open /etc/TZ code 10000000 - 1009c134, data 1009a128 - 1009c134, stack bf9c5ec0 open /etc/passwd open /etc/xinetd.d/telnet open /lib/libcrypt.so.0 open /etc/TZ open /lib/libm.so.0 open /lib/libcrypt.so.0 open /etc/passwd open /lib/libc.so.0 open /lib/libm.so.0 open /etc/protocols open /bin/application open /lib/libc.so.0 open /etc/services code 10000000 - 1009c134, data 1009a128 - 1009c134, stack bffa9ea0 open .profile open /etc/services open /etc/profile open /etc/protocols open /etc/passwd open /etc/services open /lib/libcrypt.so.0 open /etc/services open /lib/libm.so.0 open /etc/TZ open /lib/libc.so.0 open /etc/TZ code 10000000 - 1009c134, data 1009a128 - 1009c134, stack bfaa5eb0 open /lib/libcrypt.so.0 open /lib/libm.so.0 open /lib/libc.so.0 = = ________________________________ From: John Linn [mailto:John.Linn@xilinx.com] = Sent: Wednesday, August 13, 2008 7:11 AM To: Stu Bershtein; linuxppc-embedded@ozlabs.org Subject: RE: ml507 initrd problem = Hi Stu, = I realize you are trying to get your own ram disk to work, but if you want to get a baseline you can use one we have on our wiki site, at http://xilinx.wikidot.com/open-source-linux, under Files To Download. This is the same ram disk that is included in the ELDK. = This is the ram disk that I use to test the ML507 design in our Git tree. I also use NFS a lot for testing. = Looks like you're close with yours, but I'm certainly no expert in that area as I typically have something that works and don't change it. = Good luck, John = ________________________________ From: linuxppc-embedded-bounces+john.linn=3Dxilinx.com@ozlabs.org [mailto:linuxppc-embedded-bounces+john.linn=3Dxilinx.com@ozlabs.org] On Behalf Of Stu Bershtein Sent: Tuesday, August 12, 2008 5:28 PM To: linuxppc-embedded@ozlabs.org Subject: ml507 initrd problem = I have been trying to boot linux on an ml507 board for a few days now. I am using xilinx's linux-2.6-xlnx.git kernel. The rootfs is home rolled. I have had experience with NFS mounted root filesystems (not an option here), cramfs, and romfs but have never tried running root out of a ramdisk. Because of that I am quite sure I have dorked this up. Any help will be appreciated! I have added printk's in the kernel at do_sys_open to get an idea what is going on. Here is a snip of the boot output: = . Kernel command line: console=3DttyS0,9600 ip=3Don init=3D/linuxrc root=3D/dev/ram rw ram disk_size=3D9000 . Serial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing disabled 83e00000.serial: ttyS0 at MMIO 0x83e00003 (irq =3D 16) is a 16550A . eth0: XLlTemac: Send Threshold =3D 0, Receive Threshold =3D 0 eth0: XLlTemac: Send Wait bound =3D 0, Receive Wait bound =3D 0 IP-Config: Incomplete network configuration information. open /dev/ram fd 0 open /initrd.image fd 1 RAMDISK: Compressed image found at block 0 VFS: Mounted root (ext2 filesystem). Freeing unused kernel memory: 144k init open /dev/console fd 0 code 10000000 - 1019854c, data 10195000 - 1019854c, stack bf85df00 open /dev/null fd 3 open /etc/inittab fd 3 = And it gets very quiet at that point. The kernel is still running in ppc44x_idle. Does this suggest anything to anyone? I'll be more than happy to send or describe any config files to any one who might take pity. = Thanks, Stu = This email and any attachments are intended for the sole use of the named recipient(s) and contain(s) confidential information that may be proprietary, privileged or copyrighted under applicable law. If you are not the intended recipient, do not read, copy, or forward this email message or any attachments. Delete this email message and any attachments immediately. = This email and any attachments are intended for the sole use of the named r= ecipient(s) and contain(s) confidential information that may be proprietary= , privileged or copyrighted under applicable law. If you are not the intend= ed recipient, do not read, copy, or forward this email message or any attac= hments. Delete this email message and any attachments immediately. ------_=_NextPart_001_01C8FD63.B1E49866 Content-Type: text/html; charset="US-ASCII" Content-Transfer-Encoding: quoted-printable

Stu,

 

I’m assuming you’re using arch/powerpc. The command line we use in the device tree file, arch/powerpc/boot/dts/ml507.dts,  is bootargs =3D "console=3DttyS= 0 ip=3Don root=3D/dev/ram";

 

It sounds like you’re getting console output til it switches from kernel to user in the ram disk, but you= r message is a little confusing on that.  You should expect the shell ou= tput to be on the same serial port at the same baud rate as the boot output of the kernel.

 

Yes the ML507 uses the 9 pin serial connector at 9600 baud.

 

We have removed the linux_support*.gz = file when I put the bit streams and ram disk on the wiki site.  If you have= these files you have everything.

 

You should be able to following the directions on the wiki site to get a baseline working, although I didn̵= 7;t specifically say on the wiki to put the ram disk image in the boot director= y before you build or the make command line.

 

The ram disk on our site should work correctly if you are using our kernel with the default configuration, ml507_defconfig.  I always advice people to get a good baseline with o= ur stuff then make changes after you see if baselined.

 

-- John

 

From: Stu Bers= htein [mailto:sbershtein@Quantum3D.com]
Sent: Wednesday, August 13, = 2008 9:34 AM
To: John Linn; linuxppc-embedded@ozlabs.org
Subject: RE: ml507 initrd pr= oblem

 

Mo= rning John,

 

Th= anks for the tip.  Running said ramdisk things certainly look lively, I’ll add a list of opens that I see at boot up at the bottom of this email.  Looks like it’s opening a shell and everything and yet I= see no serial communication from the shell, just quiet.  Am I correct to expect the console to be the serial port (9pin DB) on the ml507 board at 96= 00 baud?  Also, if you could spare a desperate person a moment, could you= send me the boot command line that you use for testing with the ramdisk?  Finally, another member of this form suggested that there is a fully build BSP (rootfs and all) available as ‘linux_support.tar.gz’ = on the Xilinx site.  Are you familiar with this?  I tried locating i= t a few days ago to no avail.

 

Th= anks very much for your help!

St= u

 

Fr= eeing unused kernel memory: 144k init

op= en /dev/console

co= de 10000000 - 1009c134, data 1009a128 - 1009c134, stack bfdb9f00

op= en /lib/libcrypt.so.0

op= en /lib/libm.so.0

op= en /lib/libc.so.0

op= en /dev/console

op= en /dev/console

op= en /etc/TZ

op= en /etc/inittab

op= en /dev/console

op= en /etc/TZ

co= de 10000000 - 1009c134, data 1009a128 - 1009c134, stack bfd65eb0

op= en /lib/libcrypt.so.0

op= en /lib/libm.so.0

op= en /lib/libc.so.0

op= en /etc/rc.sh

co= de 10000000 - 1009c134, data 1009a128 - 1009c134, stack bfef2ea0

op= en /lib/libcrypt.so.0

op= en /lib/libm.so.0

op= en /lib/libc.so.0

op= en /etc/mtab

co= de 10000000 - 1009c134, data 1009a128 - 1009c134, stack bf9a8e90

op= en /lib/libcrypt.so.0

op= en /lib/libm.so.0

op= en /lib/libc.so.0

co= de 10000000 - 10039640, data 10039000 - 10039640, stack bfa47e70

op= en /lib/libm.so.0

op= en /lib/libcrypt.so.0

op= en /lib/libc.so.0

op= en /dev/null

op= en /tmp/xinetd.pid

op= en /etc/xinetd.conf

op= en /etc/xinetd.d

op= en /dev/console

op= en /etc/TZ

op= en /etc/xinetd.d/ftpd

op= en /dev/console

co= de 10000000 - 1009c134, data 1009a128 - 1009c134, stack bfda3ea0

op= en /etc/TZ

op= en /etc/TZ

co= de 10000000 - 1009c134, data 1009a128 - 1009c134, stack bf9c5ec0

op= en /etc/passwd

op= en /etc/xinetd.d/telnet

op= en /lib/libcrypt.so.0

op= en /etc/TZ

op= en /lib/libm.so.0

op= en /lib/libcrypt.so.0

op= en /etc/passwd

op= en /lib/libc.so.0

op= en /lib/libm.so.0

op= en /etc/protocols

op= en /bin/application

op= en /lib/libc.so.0

op= en /etc/services

co= de 10000000 - 1009c134, data 1009a128 - 1009c134, stack bffa9ea0

op= en .profile

op= en /etc/services

op= en /etc/profile

op= en /etc/protocols

op= en /etc/passwd

op= en /etc/services

op= en /lib/libcrypt.so.0

op= en /etc/services

op= en /lib/libm.so.0

op= en /etc/TZ

op= en /lib/libc.so.0

op= en /etc/TZ

co= de 10000000 - 1009c134, data 1009a128 - 1009c134, stack bfaa5eb0

op= en /lib/libcrypt.so.0

op= en /lib/libm.so.0

op= en /lib/libc.so.0

 

 

 

From: John Lin= n [mailto:John.Linn@xilinx.com]
Sent: Wednesday, August 13, = 2008 7:11 AM
To: Stu Bershtein; linuxppc-embedded@ozlabs.org
Subject: RE: ml507 initrd pr= oblem

 

Hi Stu,

 

I realize you are trying to get your o= wn ram disk to work, but if you want to get a baseline you can use one we have= on our wiki site, at h= ttp://xilinx.wikidot.com/open-source-linux, under Files To Download.  This is the same ram disk that is included i= n the ELDK.

 

This is the ram disk that I use to tes= t the ML507 design in our Git tree.  I also use NFS a lot for testing.

 

Looks like you’re close with you= rs, but I’m certainly no expert in that area as I typically have somethin= g that works and don’t change it.

 

Good luck,

John

 

From: linuxppc-embedded-bounces+john.linn=3Dxilinx.com@ozlabs.org [mailto:linuxpp= c-embedded-bounces+john.linn=3Dxilinx.com@ozlabs.org] On Behalf Of Stu Bershtein Sent: Tuesday, August 12, 20= 08 5:28 PM
To: linuxppc-embedded@ozlabs= .org
Subject: ml507 initrd proble= m

 

I = have been trying to boot linux on an ml507 board for a few days now.  I am using xilinx’s linux-2.6-xlnx.git kernel.  The rootfs is home rolled.  I have had experience with NFS mounted root filesystems (not = an option here), cramfs, and romfs but have never tried running root out of a ramdisk.  Because of that I am quite sure I have dorked this up. = Any help will be appreciated!  I have added printk’s in the kernel a= t do_sys_open to get an idea what is going on.  Here is a snip of the bo= ot output:

 

.<= o:p>

Ke= rnel command line: console=3DttyS0,9600 ip=3Don init=3D/linuxrc root=3D/dev/ram = rw ram disk_size=3D9000

.<= o:p>

Se= rial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing disabled 83e00000.serial: ttyS0 at MMIO 0x83e00003 (irq =3D 16) is a 16550A

.<= o:p>

et= h0: XLlTemac: Send Threshold =3D 0, Receive Threshold =3D 0

et= h0: XLlTemac: Send Wait bound =3D 0, Receive Wait bound =3D 0=

IP= -Config: Incomplete network configuration information.<= /p>

op= en /dev/ram

fd= 0

op= en /initrd.image

fd= 1

RA= MDISK: Compressed image found at block 0

VF= S: Mounted root (ext2 filesystem).

Fr= eeing unused kernel memory: 144k init

op= en /dev/console

fd= 0

co= de 10000000 - 1019854c, data 10195000 - 1019854c, stack bf85df00

op= en /dev/null

fd= 3

op= en /etc/inittab

fd= 3

 

An= d it gets very quiet at that point.  The kernel is still running in ppc44x_idle.  Does this suggest anything to anyone?  I’ll b= e more than happy to send or describe any config files to any one who might t= ake pity.

 

Th= anks,

St= u

 


This email and any attachments are intended for the sole use of the named recipient(s) and contain(s) confidential information that may be proprietar= y, privileged or copyrighted under applicable law. If you are not the intended= recipient, do not read, copy, or forward this email message or any attachme= nts. Delete this email message and any attachments immediately.


This email and any attachments are intended for the sole u= se of the named recipient(s) and contain(s) confidential information that m= ay be proprietary, privileged or copyrighted under applicable law. If you a= re not the intended recipient, do not read, copy, or forward this email mes= sage or any attachments. Delete this email message and any attachments imme= diately. ------_=_NextPart_001_01C8FD63.B1E49866--