Results 1 to 8 of 8
Hello to all Linux Experts,
I am currently trying to boot arm linux on an ARM core processor board. But the kernel does not load and stops by displaying the ...
Enjoy an ad free experience by logging in. Not a member yet? Register.
- 11-29-2008 #1Just Joined!
- Join Date
- Nov 2008
- Posts
- 7
Uncompressing the Linux Kernel
Hello to all Linux Experts,
I am currently trying to boot arm linux on an ARM core processor board. But the kernel does not load and stops by displaying the message
" Uncompressing kernel............................................ ............."
It just does not proceed after this.
Can you please urgently tell the possible problem areas I have to look into ?
Thankyou
- 12-03-2008 #2Just Joined!
- Join Date
- Sep 2008
- Posts
- 25
hi,
what kind of boot loader do you use? maybe that you had loaded the kernel
on low address and overwrote it during uncompressing.
e.g. my kernel is 2.5MB if it is not compressed so I should load it at least on
address 0x280000.
HTH,
ndevel.
- 12-04-2008 #3Just Joined!
- Join Date
- Nov 2008
- Posts
- 7
Thanks
I am using UBoot boot loader. The address space after uncompressing is taken care of, also the adjustment required for the virtual memory with respect to the arm memory is done, still the problem persists. Should the interrupts and timer routines create a problem here if not properly written with respect to the hardware ?
Thanks
- 12-04-2008 #4Just Joined!
- Join Date
- Sep 2008
- Posts
- 25
hi,
I've never used uboot on ARM only on powerpc but I think uboot should care about things like timer
or interrupts.
One thing comes to my mind, maybe you could try uncompressed kernel first to check whether
uncompressing causes the problem.
regards,
ndevel.
- 12-04-2008 #5Just Joined!
- Join Date
- Nov 2008
- Posts
- 7
Thankyou For clearing one doubt
Thanks very much for clarifying the issue of interrupts and timers I was looking into these for the problem.
I checked; the kernel successfully uncompresses itself; but while it is booting it says
" booting kernel " and then just halts.
can't figure out the exact problem.
Thankyou
- 12-09-2008 #6Just Joined!
- Join Date
- Sep 2008
- Posts
- 25
How do your boot arguments look like? did you specify a correct console device?
Just for sure here is a very helpful manual for uboot.
Regards,
ndevel
- 12-10-2008 #7Just Joined!
- Join Date
- Nov 2008
- Posts
- 7
Hello,
Yes boot parameters are correct, console is set. When I debugged the kernel source code, I found that it hangs in mmu.c in the for loop where it is taking VMALLOC_END as the looping criteria in function devicemaps().
I strongly feel my memory setting is incorrect and should check again.
VMALLOC_END depends on machine architeccture right. Can you tell me the memory map of kernel and where exactly this setting has to be done. I know VMALLOC_END should be placed without interfering with static IO.
My IO starts at 0X80000000 and place VMALLOC_END just below this but problem persists.
Can you give me a clear picture of memory and the settings to be done inside this.
- 12-11-2008 #8Just Joined!
- Join Date
- Nov 2008
- Posts
- 7
Booting problem resolved
Hello the kernel now boots but does not fully proceed ofcourse
this is the message displayed.
Uncompressing Linux........................................... done, booting the
kernel.
abcdef<5>Linux version 2.6.19.2 (atrisys-l3@atrisys-l3) (gcc version 4.0.0 (DENX
ELDK 4.1 4.0.0)) #40 Wed Nov 12 06:18:27 IST 2008
setup_archCPU: ARM926EJ-S [41069264] revision 4 (ARMv5TEJ), cr=00053177
Machine: SiRF AtlasIII HHGPS
Memory policy: ECC disabled, Data cache writeback
<7>On node 0 totalpages: 8192
<7> DMA zone: 64 pages used for memmap
<7> DMA zone: 0 pages reserved
<7> DMA zone: 8128 pages, LIFO batch:0
<7> Normal zone: 0 pages used for memmap
<7> Normal zone: 0 pages used for memmap
<4>BUG: not creating mapping for 0x90000000 at 0x90000000 in user region
<4>BUG: not creating mapping for 0x80000000 at 0x80000000 in user region
stops at this any hint ? I know it is because of IO mapping which we have set to 0x8000000 has to be redefined one possibility any other ?



