[okl4-developer] OKL4 on openmoko simulator
Geoffrey Lee
glee at ok-labs.com
Wed Mar 19 23:16:50 EST 2008
On Wed, Mar 19, 2008 at 09:33:33AM +0100, Lukas HANEL wrote:
Hi
Please use bootelf 0x30100000 instead.
-gl
> <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
> <html>
> <head>
> <meta content="text/html;charset=UTF-8" http-equiv="Content-Type">
> </head>
> <body bgcolor="#ffffff" text="#000000">
> Hi again,<br>
> <br>
> I was using a wrong -kernel image. The outcome is now "better":<br>
> listing:<br>
> ...<br>
> GTA01Bv4 # boot<br>
> ## Booting image at 30100000 ...<br>
> Bad Magic Number<br>
> GTA01Bv4 # md 0x30100000<br>
> 30100000: 464c457f 61010101 00000000 00000000 .ELF...a........<br>
> 30100010: 00280002 00000001 30000000 00000034 ..(........04...<br>
> 30100020: 000de370 00000202 00200034 00280017 p.......4. ...(.<br>
> 30100030: 00420043 00000001 00008000 30000000 C.B............0<br>
> <br>
> <br>
> Lukas Hänel wrote:
> <blockquote cite="mid:47E0CD92.6010805 at st.com" type="cite">
> <meta content="text/html;charset=us-ascii" http-equiv="Content-Type">
> Hi,<br>
> <br>
> now it becomes more clear. Using md 0x3010000 I see now that the
> bootloader has no image in place, there are only zeros. Obviously
> bootelf does complain. 0x3100000, as expected for my version, is empty
> too.<br>
> <br>
> listing:<br>
> ~/qemu-neo1973$ arm-softmmu/qemu-system-arm -M gta01 -m 130 -mtdblock
> openmoko/openmoko-flash.image -kernel
> /home/lukas/okl4_release_1.5.2/build/images/image.boot -usb
> -show-cursor -nographic<br>
> neo_gsm_switch: GSM disabled.<br>
> <br>
> <br>
> U-Boot 1.3.01.3.0+git20071201+svnr3484 (Dec 1 2007 - 14:22:42)<br>
> <br>
> DRAM: 128 MB<br>
> NAND: 64 MiB<br>
> Found Environment offset in OOB..<br>
> Video: 640x480x8 31kHz 59Hz<br>
> <br>
> NAND read: device 0 offset 0x25c000, size 0x5000<br>
> <br>
> Reading data from 0x260e00 -- 100% complete.<br>
> 20480 bytes read: OK<br>
> USB: S3C2410 USB Deviced<br>
> In: serial<br>
> Out: serial<br>
> Err: serial<br>
> pcf_write: charging in Qualification Mode.<br>
> pcf_write: charge voltage 4.20V.<br>
> neo_lcd_rst_switch: LCD reset.<br>
> jbt6k74_command: Display on.<br>
> neo_vib_switch: Buzz, buzz.<br>
> neo_vib_switch: Vibrator stopped.<br>
> neo_bl_switch: LCD Backlight now on.<br>
> GTA01Bv4 # bootelf 0x3010000<br>
> ## No elf image at address 0x03010000<br>
> GTA01Bv4 # md 0x3010000<br>
> 03010000: 00000000 00000000 00000000 00000000 ................<br>
> 03010010: 00000000 00000000 00000000 00000000 ................<br>
> 03010020: 00000000 00000000 00000000 00000000 ................<br>
> 03010030: 00000000 00000000 00000000 00000000 ................<br>
> ...<br>
> <br>
> David Mirabito wrote:
> <blockquote
> cite="mid:F597253D-B722-41CC-B6D4-F6ADF23A37FB at ok-labs.com" type="cite">
> <pre wrap="">Hi Lukas,
>
> It sounds like you're running Qemu in graphical mode? If so, the
> naming example does not display anything to the framebuffer, so you
> wouldn't see it anyway :)
>
> Try adding the --nographic option to get a u-boot prompt and start
> with 'bootelf 0x31000000'
> (note that --nographic will forward ctrl-c to the simulation, to quit
> it's ctrl-a, x)
>
> Alternatively, you can pass '-serial stdio' and get both the graphical
> display and the serial console, and ctrl-c to the console will exit
> the simulation.
>
> Please let me know how this goes, or if you need assistance ...
> - David Mirabito
>
>
> On 19/03/2008, at 12:54 AM, Lukas HANEL wrote:
>
> </pre>
> <blockquote type="cite">
> <pre wrap="">Hi
>
> I have no boot loader shell. I can only use space and return.
>
> However, using qemu I switched to the qemu monitor and used memsave to
> dump the portion to memory. Don't know, however, if its the "same"
> memory. E.g memsave seems to operate on virtual memory:
> memsave addr size file -- save to disk virtual memory dump starting at
> 'addr' of size 'size'
> The outcome is, however the same as the image.boot file:
> 0000000 457f 464c 0101 6101 0000 0000 0000 0000
> 0000010 0002 0028 0001 0000 0000 3000 0034 0000
> 0000020 e370 000d 0202 0000 0034 0020 0017 0028
> 0000030 0043 0042 0001 0000 8000 0000 0000 3000
> ...
>
> Using following binaries:
> OpenMoko-openmoko-devel-image-glibc-ipk-P1-Snapshot-20080314-fic-
> gta01.rootfs.jffs2
> u-boot-gta01bv4-1.3.0+git20071201+svnr3484-r0.bin
> uImage-2.6.24+svnr4184-r4165-r3-neo1973.bin
>
>
> Geoffrey Lee wrote:
> </pre>
> <blockquote type="cite">
> <pre wrap="">On Tue, Mar 18, 2008 at 09:54:02AM +0100, Lukas HANEL wrote:
>
> </pre>
> <blockquote type="cite">
> <pre wrap="">Hi,
>
> I have a problem with the simulation of OKL4 gta01 platform.
> I build the qemu-neo173 simulator following the steps in the
> openmoko
> wiki. So my version is founded on the latest versions of the
> tools, as
> downloaded by the script.
>
> When I start the simulation using the following command, qemu
> starts but
> there is no output.
> ./tools/build.py machine=gta01 project=iguana example=naming
> simulate
>
> Using the qemu command like for openmoko, displays the boot options.
> After selecting Boot, the bootloader complains about missing magic
> numbers in the image.
>
> Listing:
> qemu-system-arm -M gta01 -m 130 -mtdblock openmoko/openmoko-
> flash.image
> -kernel /home/lukas/okl4_release_1.5.2/build/images/image.boot -usb
> -show-cursor
>
> ## Booting Image at 30100000 ...
> Bad Magic Number
>
> </pre>
> </blockquote>
> <pre wrap="">Hi
>
> Can you dump the memory in u-boot at 0x30100000 and see if there's
> anything there? e.g.
>
> GTA01Bv4 # md 0x31000000
> 31000000: 464c457f 61010101 00000000 00000000 .ELF...a........
> 31000010: 00280002 00000001 30000000 00000034 ..(........04...
> 31000020: 007e4510 00000202 00200034 00280013 .E~.....4. ...(.
> 31000030: 00530054 00000001 00008000 30000000 T.S............0
>
> (note that my local copy is a bit older so things are at 0x31000000.)
>
> -gl
>
>
> </pre>
> <blockquote type="cite">
> <pre wrap="">Could you give me some hints what version of qemu-neo to use and
> whether
> additional adoptions are necessary to run it with the 1.5.2 OKL4
> snapshot.
>
> Lukas
>
> PS: the skyeye simulation works.
>
> _______________________________________________
> Developer mailing list
> <a moz-do-not-send="true" class="moz-txt-link-abbreviated"
> href="mailto:Developer at okl4.org">Developer at okl4.org</a>
> <a moz-do-not-send="true" class="moz-txt-link-freetext"
> href="https://lists.okl4.org/mailman/listinfo/developer">https://lists.okl4.org/mailman/listinfo/developer</a>
>
> </pre>
> </blockquote>
> </blockquote>
> <pre wrap="">_______________________________________________
> Developer mailing list
> <a moz-do-not-send="true" class="moz-txt-link-abbreviated"
> href="mailto:Developer at okl4.org">Developer at okl4.org</a>
> <a moz-do-not-send="true" class="moz-txt-link-freetext"
> href="https://lists.okl4.org/mailman/listinfo/developer">https://lists.okl4.org/mailman/listinfo/developer</a>
> </pre>
> </blockquote>
> <pre wrap=""><!---->
>
> _______________________________________________
> Developer mailing list
> <a moz-do-not-send="true" class="moz-txt-link-abbreviated"
> href="mailto:Developer at okl4.org">Developer at okl4.org</a>
> <a moz-do-not-send="true" class="moz-txt-link-freetext"
> href="https://lists.okl4.org/mailman/listinfo/developer">https://lists.okl4.org/mailman/listinfo/developer</a>
>
> </pre>
> </blockquote>
> <br>
> </blockquote>
> <br>
> </body>
> </html>
>
> _______________________________________________
> Developer mailing list
> Developer at okl4.org
> https://lists.okl4.org/mailman/listinfo/developer
--
More information about the Developer
mailing list