2007/4/9, Wolfgang Denk wd@denx.de:
Dear Ilya,
in message 8496f91a0704080420m1e72900cn70116c3dc7973b2b@mail.gmail.com you wrote:
This patch adds support for MC68360 based board for u-boot-1.1.6. Mail
me if
you have any questions :)
Please correct me if I'm wrong, but MC68360 is not the name of a board, but of the Freescale (resp. ex Motorola) QICC CPU. There are many different boards using this processor, so you will have to chose a more descriptive board name.
Yes, you are right. MC68360 is the name of the CPU. But the board I'm working at right now is my own design. That's why I can't do it.
Also, U-Boot 1.1.6 is outdated. Please submit your patch against a
recent version of U-Boot (i. e. top of tree in git repository, or at least release 1.2.0).
Ooops, I was not aware of it. As far as I know, there was no official information about it.
Then, there are some coding style issues with your patch (indentation
not by TAB, indentation not by multiple of 8 columns, trailing white space, C++ comments, too long lines, etc.) as well as other formal issues (missing Copyright entries, missing license headers, missing Signed-off-by: line).
Ok. Could you tell me where I can get exact information about it?
It seems you use a private flash driver for a device that looks as if
it was CFI conformant. Please explain why you cannot use the CFI driver instead.
This is because I have old non-CFI flash chips :(
I also wonder if you really need a new ethernet driver
(cpu/mc68360/enet.c) and a new serial driver (cpu/mc68360/serial.c) - IIRC the CPM on the 360 is mostly compatible (with very few minor deviations that can be handled easily) to the CPM on the PowerQICC I (= MPC8xx) processors. I think most of this could could (and should) be shared? The same holds for some other files like for example include/asm-m68k/arch-mc68360/commproc.h which is highly redundand with the MPC8xx commproc.h
I don't know this processors family well enough, but I see what I can do with it....
Some of your files contain version ID stuff like here:
include/asm-m68k/arch-mc68360/mc68360_enet.h:
*********************************** * $Id: m68360_enet.h,v 1.1.1.1 2001/05/18 17:10:11 hamilton Exp $ ***********************************Please remove this.
ok
I see some "#if 0" blocks in your code; please remove these.
ok
I see:
#define CONFIG_BOOTFILE vmlinuzThis looks broken to me. U-Boot does not boot a vmlinuz file, but U-Boot images created by mkimage.
Yes, this name is incorrect... Let it be uImage.
#define CFG_BAUDRATE_TABLE { 19200 }
This is *very* restrictive. I recommend to support all commonly used baudrates instead (i. e. at least the range from 9600 through 115200).
ok
#define CFG_HZ (unsigned long int)32768000
This is broken. CFG_HZ is required to be 1000 (i. e. millisecond ticks).
okkk..
#define CFG_SDRAM_SIZE 0x00800000
Please don't do that. U-Boot style is to allow auto-adjustment to the actual RAM size.
#define CFG_MALLOC_LEN (5*1024*1024)
You have 8 MB of RAM and reserve 5 MB for malloc()? This seems broken to me.
Why not? I really have it working :)
You comment out relevant parts of lib_m68k/board.c - you must not do
that!
Could you tell me the reason why I should not do it?
lib_m68k/time.c - your "implementation" of udelay() is broken. Please
replace by a working version.
I'll try.
Please clean up these issues and resubmit.
I'm grateful to you for your help and for what you are doing :)
I will take into consideration your advice and recommendations. And I'll try to correct my patch.
Best regards, Matvejchikov Ilya.