The problem sees to be more complicated. I just recompile the bootldr
and kernel with exactly the same configuration on the same host machine.
And now, I get a new type of error: it hangs when copying compressed
ramdisk to DRAM:
copying compressed ramdisk from 00100004 to C0800000...
I have tried a few changes, say using different host kernel version to
compile the bootldr and kernel, remove or add a few options in the
kernel, none of them is working. It hangs at the same location.
I started to miss the "invalid compressed format" error now :)
Regards,
Leonard Ye
longfengcai wrote:
> Hello,
>
> I was always encountering the same problems a few days ago.
> (CRC error,length error,etc) I have solved these problems for
> myself. But I don't know if it can help you.
>
> I built the arm linux cross toolchain myself on an i686
> (Pentium III 667) and Red Hat 6.2 (Note:with kernel 2.2.14-5)
> host. Then I found this 2.2.14-5 kernel doesn't surport cramfs,
> so I built a new 2.4.0 kernel with cramfs surported for my
> host,but I didn't rebuilt the arm linux cross toolchain when
> running 2.4.0 kernel. So I always encounter these problems
> when I build the kernel for assabet while running 2.4.0 kernel
> on my host. And I solved these problems by booting 2.2.14-5
> kernel on my host.
>
> Perhaps you don't use Red Hat 6.2 host.
> Perhaps you don't have two different kernel running
> on your host.
> Perhaps my thought is absurd.
> Any way, it solve my problem.
>
> Hope this can help you!
>
> Regards
> Fengcai
>
>
> ----- Original Message -----
> From:John G Dorsey <[EMAIL PROTECTED]>
> To:Leonard Ye <[EMAIL PROTECTED]>
> Subject:Re: Novice Questions regarding Bootldr
> Date:Sun, 11 Mar 2001 22:36:24 +0800
> >on 3/11/01 7:30 AM, Leonard Ye at [EMAIL PROTECTED] wrote:
> >
> >> Uncompressing Linux.....................................
> >>
> >> invalid compressed format (err=1)
> >>
> >> -- System halted
> >
> >Well, at least it's not just me. =)
> >
> >I've been seeing this with 2.4.1 on both Assabet/Neponset and Spot for
> >several weeks now, and thought that I just had bad luck or something. I've
> >discovered that if I change the bits that get linked/compressed/downloaded
> >in some slight way (add some code, change the configuration slightly), it
> >goes away. Note that I've also seen errors regarding a bad CRC, and an
> >invalid length at this stage, and this seems amenable to the same solution.
> >
> >Since the behavior is generally reproducible across successive downloads of
> >the same image, I'm tempted to think that if this is a software bug, it may
> >be in the link or kernel image compression stages. If it were just CRC
> >corruption, I'd first blame the bootloader, but this seems like general
> >compression header damage. Perhaps if anyone else has observed this
> >recently, they might comment so that we can narrow the search.
> >
> >-jd
> >
> >
> >_______________________________________________
> >http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm
> >Please visit the above address for information on this list.
> >
> >
>
> ______________________________________
>
> ===================================================================
> ÐÂÀËÃâ·Ñµç×ÓÓÊÏä (http://mail.sina.com.cn)
> ÍøÀïÑ°Ëýǧ°Ù¶È!ûÓÐ"ÁÄÓÑËÙÅä",ÔõÄÜ"³ÉË«³É¶Ô"? (http://newchat.sina.com.cn)
_______________________________________________
http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm
Please visit the above address for information on this list.
