June 2004 - U-Boot

RE: [U-Boot-Users] necessary fix for ARM u-boot 1.1.1 and GCC3.4.0.
by Woodruff, Richard 11 Jun '04

11 Jun '04

Yes, I am now able to compile and boot a kernel using 1.1.1 and 3.4.0 after a few changes. I think I've posted the main bits which needed changing (tool chain patch, examples/stubs tweak, lib_arm/board.c optimization stop, as you highlight out per-cpu make file tweak) it looks like it will be ok. Short-load-bytes is deprecated and needs to be replaced or removed. Per the manual it is only necessary for ARMv3 and below (Ie no half word load available). Regards, Richard W. > -----Original Message----- > From: Robert Schwebel [mailto:robert@pengutronix.de] On Behalf Of Robert > Schwebel > Sent: Friday, June 11, 2004 12:53 AM > To: Woodruff, Richard > Cc: u-boot-users > Subject: Re: [U-Boot-Users] necessary fix for ARM u-boot 1.1.1 and > GCC3.4.0. > > On Thu, Jun 10, 2004 at 04:53:13PM -0500, Woodruff, Richard wrote: > > I just found that lib_arm/board.c:start_armboot()'s GCC3.4.0 generated > > code sequence breaks ARM u-boot's. I am using the GCC3.4.0 which > > CodeSourcery put out. > > Hmm, that suggests that you have been able to compile it at all? I just > tried current CVS with the crosstool-0.28-rc23 gcc and it breaks: > > ----------8<----------8<----------8<----------8<----------8<---------- > arm-softfloat-linux-gnu-gcc -M -g -Os -fno-strict-aliasing -fno-common > -ffixed-r8 -mshort-load-bytes -msoft-float -D__KERNEL__ - > DTEXT_BASE=0xa1fe0000 -I/home/robert/work/u-boot-maintainance/include - > msoft-float -fno-builtin -ffreestanding -nostdinc -isystem > /ptx/opt/crosstool/arm-softfloat-linux-gnu/gcc-3.4.0-glibc- > 2.3.2/bin/../lib/gcc/arm-softfloat-linux-gnu/3.4.0/include -pipe - > DCONFIG_ARM -D__ARM__ -mapcs-32 -march=armv5 -mtune=xscale -Wall -Wstrict- > prototypes hello_world.c stubs.c > .depend > cc1: error: invalid option `short-load-bytes' > cc1: error: invalid option `short-load-bytes' > make[1]: *** [.depend] Error 1 > ----------8<----------8<----------8<----------8<----------8<---------- > > Robert > -- > Dipl.-Ing. Robert Schwebel | http://www.pengutronix.de > Pengutronix - Linux Solutions for Science and Industry > Handelsregister: Amtsgericht Hildesheim, HRA 2686 > Hornemannstraße 12, 31137 Hildesheim, Germany > Phone: +49-5121-28619-0 | Fax: +49-5121-28619-4

1 0

[U-Boot-Users] lib_arm/board.c calls eth_init() for smc91111...why?
by Paul Ruhland 11 Jun '04

11 Jun '04

I noticed when my board booted with the latest CVS there was an unusually long delay (3-4 seconds) before the prompt came up. I initially thought it had hung. Turns out a call to 'eth_init()' was recently added in 'lib_arm/board.c'. Since 'eth_init()' is called from 'Net_Loop()' why do we need the call in board.c? It adds quite a large delay to the bootup time that I think is unnecessary. -- Paul Ruhland <pruhland(a)microwavedata.com> Software Engineer Microwave Data Systems Rochester, NY 16420

2 1

[U-Boot-Users] How to reduce the size of u-boot image file
by Frank Young 11 Jun '04

11 Jun '04

Hi, I have two questions here: 1. Because of the size of the flash memory constraint, I need to strip the u-boot to its minimum size. Right now by executing "make all" in the u-boot directory, I got a image file u-boot at around 700K-800K. It's too big for my target. So I am thinking about getting rid of some functionalities which u-boot provides. I do need u-boot to initialize my hardware. But after that, I only need to use tftp to get the linux image, decompress the image, put the decompressed image in RAM and jump to the linux entry point to let the linux take over the control. Is there any FAQ or manual explaining how to strip down the u-boot size? 2. Since my flash memory address starts from 0xFFC00020 instead of 0xFFF80000, so I changed the TEXT_BASE in board/walnut405/config.mk to 0xFFC00020. When I use "make u-boot", everything is fine. But if I use "make u-boot.bin", it ends up with a u-boot.bin more than 4MB. I think I didn't get the TEXT_BASE set correctly. Isn't TEXT_BASE the start point of the flash? If it isn't, what does TEXT_BASE mean? And where to indicate the start of the flash memory? BTW, here is my environment: CPU: PPC 405 200MHz Flash: 4MB SDRAM: 16MB BOARD: walnut Thanks a lot! Frank _________________________________________________________________ FREE pop-up blocking with the new MSN Toolbar get it now! http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/

2 1

[U-Boot-Users] [PATCH 2/2] LPD7A40x flash cleanup
by Paul Ruhland 10 Jun '04

10 Jun '04

This patch cleans up some debug code leftover in lpd7a40x board support. -- Paul Ruhland <pruhland(a)microwavedata.com> Software Engineer Microwave Data Systems Rochester, NY 16420

2 1

[U-Boot-Users] [PATCH 1/2] LPD7A404 support
by Paul Ruhland 10 Jun '04

10 Jun '04

This patch fixes up support for the Logic SDK-LH7A404 board and cleans up the LH7A404 register macros. -- Paul Ruhland <pruhland(a)microwavedata.com> Software Engineer Microwave Data Systems Rochester, NY 16420

2 1

[U-Boot-Users] Sandpoint8245 serial clock patch
by Matthew McClintock 10 Jun '04

10 Jun '04

Hi Wolfgang, Since I have not been able to contact the maintainer of the Sandpoint8245 board I have decided to go ahead and send this patch in. The patch is very simple and simply selects the correct clock for the console based on which serial port you select to be used as the console. When you get a chance, could you please apply this patch? Thanks -- Matthew McClintock <mcclintock(a)freescale.com>

2 1

Re: [U-Boot-Users] [Patch] flashh
by Robert Schwebel 10 Jun '04

10 Jun '04

On Wed, Jun 09, 2004 at 02:21:24PM +0200, Wolfgang Denk wrote: > > @@ -340,6 +343,9 @@ > > #define FLASH_AMDL163T 0x00B2 /* AMD AM29DL163T (2M x 16 ) */ > > #define FLASH_AMDL163B 0x00B3 > > > > +#define FLASH_28F128K3 0x00A7 /* Intel 28F128K3 ( 128M = 8M x 16 ) */ > > This is IMHO wrong. The comment a few lines above reads: > > Be careful when adding new type! Odd numbers are "bottom boot sector" types! > > I don't think this is a bottom boot sector type. It's a "all sectors are created equal" type ;) > > +#define FLASH_28F256K3 0x00A8 /* Intel 28F256K3 ( 256M = 16M x 16 ) */ > > + > > Already included now. > > Please fix and resubmit if needed. I'm a little bit confused; I understand these numbers as internal U-Boot identifiers, arent' they? The list has duplicate entries. And, according to what you said above, how is this meant: #define FLASH_28F320J3A 0x007C /* INTEL 28F320J3A ( 32M = 128K x 32) */ #define FLASH_28F640J3A 0x007D /* INTEL 28F640J3A ( 64M = 128K x 64) */ #define FLASH_28F128J3A 0x007E /* INTEL 28F128J3A (128M = 128K x 128) */ These are equal sized as well but have ascending numbers. New patch is attached. Robert -- Dipl.-Ing. Robert Schwebel | http://www.pengutronix.de Pengutronix - Linux Solutions for Science and Industry Handelsregister: Amtsgericht Hildesheim, HRA 2686 Hornemannstraße 12, 31137 Hildesheim, Germany Phone: +49-5121-28619-0 | Fax: +49-5121-28619-4

2 1

[U-Boot-Users] RE: u-boot compile failure with gcc-3.4.x and fix.
by Woodruff, Richard 10 Jun '04

10 Jun '04

Or perhaps a gcc header like the kernel is needed and the value changed to __attributed_used__ . -------- * These definitions are for GCC v3.x. */ #include <linux/compiler-gcc.h> #if __GNUC_MINOR__ >= 1 # define inline __inline__ __attribute__((always_inline)) # define __inline__ __inline__ __attribute__((always_inline)) # define __inline __inline__ __attribute__((always_inline)) #endif #if __GNUC_MINOR__ > 0 # define __deprecated __attribute__((deprecated)) #endif #if __GNUC_MINOR__ >= 3 # define __attribute_used__ __attribute__((__used__)) #else # define __attribute_used__ __attribute__((__unused__)) #endif #define __attribute_pure__ __attribute__((pure)) #define __attribute_const__ __attribute__((__const__)) #if __GNUC_MINOR__ >= 1 #define noinline __attribute__((noinline)) #endif > -----Original Message----- > From: Woodruff, Richard > Sent: Thursday, June 10, 2004 2:59 PM > To: u-boot-users(a)lists.sourceforge.net > Subject: u-boot compile failure with gcc-3.4.x and fix. > > Hello, > > I'm noticing that the examples directory fails to compile with gcc-3.4.0. > The _exports.h symbols are not making it into stubs.o . > > I found that the compile works if you: > > Change: > static void __attribute__((unused)) dummy(void) > to: > void __attribute__((unused)) dummy(void) > > This would seem to be a new optimization or bug. I would expect that > unused should make it stick around, even if it is not used locally. > > The kernel uses this same construct in a few places mostly in debug code > and drivers I don't' use... > > Regards, > > Richard W.

1 0

[U-Boot-Users] srec download fail
by Thomas Brand 10 Jun '04

10 Jun '04

Hi, I'm trying to download the hello_world srec-sample to my Nios Stratix Board with cu. I get the following printout: DK1S10 > loads ## Ready for S-Record download ... ~>hello_world.srec 1 R R R R R R R R R [timed out looking for echo] [connected] my configuration: U-Boot 1.1.2 (May 24 2004 - 15:54:07) CPU: Nios-32 Rev. 3.3 (0x3038) Reg file size: 512 LO_LIMIT/HI_LIMIT: 1/30 Board: Altera Nios 1S10 Development Kit Conf.: Microtronix LDK 2.0 (LDK2) In: serial Out: serial Err: serial It does not work with kermit either. The loadb cmd with kermit works fine. Does anybody has a hint for me? Thanks in advance. Thomas Brand

3 9

[U-Boot-Users] RPX Classic Question
by Travis Sawyer 10 Jun '04

10 Jun '04

Greetings u-boot'ers: Slightly off topic: I have in my possession several Embedded Planet RPX Classic boards. (Currently orphaned according to the readme). They have the EP boot firmware loaded in the flash. I've tried running they're 'program flash' code in an attempt to get u-boot into the flash to no avail. The flash is soldered down and I have no bdi firmware for the 860. (this is a personal project, so no budget for debugger FW). Does anyone have instructions for burning u-boot into the flash using their program? It looks like I have to wrap the u-boot.bin with something, but it isn't intuitive from their documentation. TIA, travis

1 0