help with bootup code

I am just learning Linux and hope to port it to a custom platform using older I/O chipsets(W83c553 and 87307). Basically the system consists of a 603/MPC106 with 64M sdram, PCI to ISA and IDE interface(W83c553), and I/O controller(87307). I believe the basic system is already supportted by Linux but support for I/O chipsets must be added. I am using LinuxPPC 4.1 while waiting for release 5.

I can load the compiled kernel into high memory and jump to head.S.

I have traced the boot code(arch/ppc/boot) and think I understand the relocate, vga_init, and decompress code but am unable to determine where the input is coming from when fill_inbuf() is called. It gets input from an array(input_data[]?) but I can't find where the data is assigned to that array. I don't see an open call, a pointer assignment, or a memcpy to draw data from. The kernel isn't running yet so Linux device drivers aren't supplying the data.

The code is something like "if ( data_ptr < buffer_size) { data_inbuf[data_ptr] = input_data[offset]; data_ptr++;}". I am sorry but I don't have the code accessible at the moment. If this is too confusing to answer, I will re-submit it tomorrow evening when the source code listings will be available. I check the list in the evening and am trying to get feedback as soon as possible.

I have been monitoring this list for a few weeks and have not seen these chips mentioned except once when someone, from Korea(?), indicated they were interested in porting Linux to a Motorola YellowKnife system which uses them. (if your that person let me know how its going!)

Thanks in advance for you help,

Bill Whitehead
[email protected]

[[ This message was sent via the linuxppc-dev mailing list. Replies are ]] [[ not forced back to the list, so be sure to Cc linuxppc-dev if your ]] [[ reply is of general interest. Please check http://lists.linuxppc.org/ ]]
[[ and http://www.linuxppc.org/ for useful information before posting. ]]

MPC106, W83c553 and NS87307 are all supported in one or more systems. Hence the basic code is there.

Greetings,

Geert

--
Geert Uytterhoeven [email protected]
Wavelets, Linux/{m68k~Amiga,PPC~CHRP} http://www.cs.kuleuven.ac.be/~geert/
Department of Computer Science -- Katholieke Universiteit Leuven -- Belgium

[[ This message was sent via the linuxppc-dev mailing list. Replies are ]] [[ not forced back to the list, so be sure to Cc linuxppc-dev if your ]] [[ reply is of general interest. Please check http://lists.linuxppc.org/ ]]
[[ and http://www.linuxppc.org/ for useful information before posting. ]]

This unpacking magic is aided by the kernel build process. The details differ depending on the kernel version (2.0.XX is different from 2.1.XX and 2.2.XX). The basic idea is that the top level makfile (in /usr/src/linux) builds a file 'vmlinux' which gets compressed and packed into that array. Then there is some 'make' magic that puts that together with the bootup code.

The code you're looking at above deals with the actual unpacking/uncompressing of the data. The kernel is uncompressed from the relocated image into low memory by this process. Once it's all uncompressed, the bootstrap passes some data to it and jumps into it to start the kernel.

Hope this helps.

------------------------------------------------------------------------

Gary Thomas                              |
email: [email protected]                  | "Fine wine is a necessity of
   ... opinions expressed here are mine  |        life for me"
       and no one else would claim them! |
                                         |      Thomas Jefferson

------------------------------------------------------------------------

[[ This message was sent via the linuxppc-dev mailing list. Replies are ]] [[ not forced back to the list, so be sure to Cc linuxppc-dev if your ]] [[ reply is of general interest. Please check http://lists.linuxppc.org/ ]]
[[ and http://www.linuxppc.org/ for useful information before posting. ]]