the first thing you should do is grab the latest U-Boot. The
version you're experimenting is quite old, I'm afraid.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
responding wrong person.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
: *** No rule to make target 'arm926ejs_config'. Stop.
That's correct, 'make' is unable to find such target. You must specify
the target corresponding to your board, for example make
davinci_dvevm_config, i.e. prefix preceding '_config' normally
indicate a platform's name.
--
Roman Mashak
DEBUG right in your board's configuration file. Will
provide you with lots of useful debug messages.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
for this.
Would it for example be in : arch/powerpc/boot and in one of the assembly
language files?
Look at $(uboot)/common/cmd_bootm.c, cmd_boot.c. On ARM the process
also involves $(uboot)/lib_arm/bootm.c, I guess should be something
similar on PPC.
--
Roman Mashak
that modifying platform's startup code is not encouraged. Is
there more correct approach?
Thanks.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
doing this (like
-fpic or -pie etc.). Perhaps I misunderstand something in a work of
U-Boot...
Thanks.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
of timer's interface?
- Now, what are reset_timer_masked() and other *_masked() for?
- Is it required to customize 'udelay()' routine fo revery ARM-based
platform or there's a basic skeleton to use?
Thanks.
--
Roman Mashak
___
U-Boot mailing list
U-Boot
that (_armboot_start - CFG_MALLOC_LEN - sizeof(gd_t))
is negative and converted to 0x which is certainly wrong.
Where's the mistake -- link address or something else?
Thanks in advance.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http
Hello
2008/12/11 Roman Mashak [EMAIL PROTECTED]:
Briefly about the memory map: before remap NOR flash is at 0x0 and
SDRAM 0x1000, after remap -- SDRAM is mapped at 0x0. I've built
a U-Boot image linked to 0x0, as I want to load it in RAM, so I linked
it to 0x0.
Loading this image via
* U-Boot-1.3.4
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
Hello
2008/12/12 Roman Mashak romez...@gmail.com:
Certainly.
* CPU is arm926ejs based processor with standard peripherals (UART,
I2C, timers etc.)
* our custom board
* U-Boot-1.3.4
Forgot to mention the serial settings in my board's configuration file:
#define CONFIG_BAUDRATE
Hello
2008/12/12 Roman Mashak romez...@gmail.com:
[skip]
So I believe 'default_environment[]' has correct baudrate
environment set and on startup it will be picked up by
'init_baudrate()' from $(UBOOT)/lib_arm/board.c
Problem is fixed, it was due to a bug in the function configuring
clock
2c00 cmd f0 00f0 16bit x 16 bit
fwc addr 2c000aaa cmd aa 00aa 16bit x 16 bit
fwc addr 2c000554 cmd 55 0055 16bit x 16 bit
fwc addr 2c000aaa cmd 90 0090 16bit x 16 bit
Does this indicate the problem with sectors number or this chip is not
supported by CFI drive?
Thanks.
--
Roman Mashak
OMAP identification is implemented in 'cpuinfo.c' and located in ARM926EJ-S
directory.
It makes sense to place this file in OMAP specific subdirectory, i.e.
cpu/arm926ejs/omap
Signed-off-by: Roman Mashak [EMAIL PROTECTED]
---
cpu/arm926ejs/Makefile |2 +-
cpu/arm926ejs/cpuinfo.c
Image
Double check these addresses. The convention is to load the Linux
kernel at the base of physical RAM plus an offset of 0x8000 (32K),
which allows enough room for the parameters area (typically placed at
offset 0x100), page exception vectors and page tables.
--
Roman
Removal of never used functions.
Signed-off-by: Roman Mashak [EMAIL PROTECTED]
---
board/versatile/versatile.c | 22 --
1 files changed, 0 insertions(+), 22 deletions(-)
diff --git a/board/versatile/versatile.c b/board/versatile/versatile.c
index 3b9b020..0f35caa 100644
OMAP identification is implemented in 'cpuinfo.c' and located in
ARM926EJ-S directory. It makes sense to place this file in OMAP
specific subdirectory, i.e. cpu/arm926ejs/omap
Signed-off-by: Roman Mashak [EMAIL PROTECTED]
---
cpu/arm926ejs/Makefile |2 +-
cpu/arm926ejs/cpuinfo.c
this before?
Thanks.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
precisely define what to build. Further,
$(uboot)/config.mk includes config_local.mk, and later on Makefile
organizes LIBS something like that:
...
ifdef BUILD_NET
LIBS +=drivers/net/libnet.a
endif
...
It may slightly speed up a building process.
--
Roman Mashak
if someone explains me this nuance. Thanks in advance!
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
not it involve remapping as well?
I didn't find how it's done in U-Boot (if it's done at all). Would
appreciate a lot for clarification.
I'm trying to understand how it's done on ARM --
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http
Hello
2008/10/2 Scott Wood [EMAIL PROTECTED]:
On Wed, Oct 01, 2008 at 06:04:02PM -0700, Roman Mashak wrote:
As far as I understand not all manufacturers adhere to CFI standards,
so this is flash chip's problem rather then board itself.
CFI is for both NAND and NOR flash devices (even though
);
break;
}
size += flash_info[i].size;
}
.
}
Regardless of CFG_MAX_FLASH_BANKS value, the only first bank is
configured. Was it done deliberately or it's logical flaw? I think
this code could cloned on a multiple of other platforms.
--
Roman Mashak
is a different file, care is taken to
see
that an incorrect bank is not considered for initialization
Yes, that makes sense then. Thank you.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
be taken as a reference?
Thanks.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
comments.
Thanks.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
Hello
2008/9/30 Roman Mashak [EMAIL PROTECTED]:
Flash ID is in $(UBOOT)/include/linux/mtd/nand_ids.h
Seems this chip is not CFI compliant (at least datasheet says nothing
of it), so the use of CFI driver falls away.
After some investigations I see that $(uboot)/drivers/mtd/nand/nand.c
this macro.
(2) do I understand right that when CFG_ENV_IS_NOWHERE defined,
environment variables are not changeable and always default values are
used.?
Thanks in advance!
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de
of
booting from flash. How will it look when u-boot is loaded in RAM by
preliminary boot code?
I specifically want to understand how memory is organized by U-Boot on
ARM platform.
Sorry if question sounds silly.
--
Roman Mashak
___
U-Boot mailing list
U-Boot
involves CRC checking and if it's faulty, then
restore configuration.
Well, something like that :) Correct me if I'm wrong.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
the arguments get put into registers r0, r1, r2 and are 0, the
machine ID, and a pointer to the ATAGS structure respectively.
Thanks a lot for such a useful and clear explanation.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http
platform
(provided that CONFIG_USE_IRQ is defined), then which one will be
actually compiled in resulting image?
Thanks.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
)/cpu/arm920t/interrupts.c.
Therefore I was wondering how it turns out that in the target image
one 'do_irq' function exists? Perhaps it's u-boot irrelevant question
and rather concern GNU toolchains behavior?
Thanks.
--
Roman Mashak
___
U-Boot mailing
Hello,
2008/9/24 Gururaja Hebbar K R [EMAIL PROTECTED]:
[skip]
So both are mutual exclusive and finally only one will be compiled.
Exactly! Really I didn't spot that subtle difference -- #ifdef, #ifndef.
Thanks for pointing out.
--
Roman Mashak
OS on it. IRQ/FIQ may be necessary to run
standalone code within bootlloader ?
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
. (AMBA bus-compliant),
but marked in U-Boot sources with different names or so ?
Thanks in advance.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
like PL19x are not supported by
default?
I developed few others my self, like pl18x mmc, rtc, gpio.
I recently sent a patch for RTC pl031.
Yes, I saw that patch. Are you planning to send patch for GPIO as well?
--
Roman Mashak
___
U-Boot mailing list
U
are for future ?
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
for other platforms?
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
platform; moreover README says about interrupts mainly in PPC or
PowerPC marked sections.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
on just a few targets.
3) The stack set up -- commonly startup code initializes stacks for
all modes, but it's not clear how it's done in 'start.S'. Seems like
somehow stacks are run-time configured, but I didn't perceive how.
Thanks in advance for comments.
--
Roman Mashak
doubts: it's valid to use ELDK
toolchain for both building bootloader and linux kernel? I ever
encountered cases where two distinct toolchains were used to compile
u-boot and kernel, target was ARM9. Does it make sense, and what might
be the reasons for that?
Thanks.
--
Roman Mashak
it and use it for, let's say,
ARM9 based platform ?
How necessary is it to install ELDK in order to be able to compile
U-Boot for non-PPC platform?
Thanks.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u
-initialize the RAM
in U-Boot, though.
This is my understanding that I have to define these optioins:
CONFIG_SKIP_LOWLEVEL_INIT and CONFIG_SKIP_RELOCATE_UBOOT in order to
avoid re-initialization of RAM and copying U-Boot image in to RAM.
--
Roman Mashak
___
U
.
Thanks.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
Hello,
I'm planning to make a port on a ARM926EJ-S based board. What is the
recommended way: take release or GIT version as a base ?
Thanks.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
was considering this, but thought there would probably be something
more specific. Anyway, I need to investigate H/W more thoroughly.
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
?
The current window for changes is (almost) closed: there will be a period of
stabilization. The ToT is generally stable despite the change
What is ToT ?
--
Roman Mashak
___
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot
50 matches
Mail list logo