February 2007 - U-Boot

[U-Boot-Users] FW: AT91 NAND om AT91SAM9260EK
by Peter Pearse 09 Feb '07

09 Feb '07

> -----Original Message----- > From: u-boot-users-bounces(a)lists.sourceforge.net > [mailto:u-boot-users-bounces@lists.sourceforge.net] On Behalf > Of Ulf Samuelsson > Sent: 08 February 2007 19:26 > To: Michel Benoit > Cc: u-boot-users(a)lists.sourceforge.net; Stefan Roese > Subject: Re: [U-Boot-Users] AT91 NAND om AT91SAM9260EK > <---snip---> > Peter Pearse which will be responsible for maintaining the > ARM part is not up to speed yet, and noone else seems to be > interested. <---snip---> No, I'm going to be the gatekeeper, not the maintainer. I shall be relying on board maintainers for testing - I only have access to (and time/expertise for) ARM supplied development boards:- Integrator Platform baseboard Application baseboard Emulation baseboard. Regards Peter -- IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.

1 0

[U-Boot-Users] ethernet on FCC
by jie han 09 Feb '07

09 Feb '07

Hi, My mpc8270 board use three ehternet ports at FCC1,FCC2,FCC3.it is work well at FCC1.When I use port FCC2, I found out the u-boot is not receiving packets (It's sending correct packet when i use ethereal for sniffing). When trying "ping", it generated an ARP packet first, the other host replied with , but the board never seems to receive it(set printf at function fec_recv() /cpu/mpc8260/ether_fcc.c ).all three ethernet fcc ports initiailze is same.Do I miss something?thank you for your suggestion,my config file config FCC port parameter as follow: #undef CONFIG_ETHER_ON_SCC /* define if ether on SCC */ #define CONFIG_ETHER_ON_FCC /* define if ether on FCC */ #undef CONFIG_ETHER_NONE /* define if ether on something else */ #define CONFIG_NET_MULTI #define CONFIG_ETHER_ON_FCC1 #define CONFIG_ETHER_ON_FCC2 #define CONFIG_ETHER_ON_FCC3 #define CFG_CMXFCR_VALUE1 (CMXFCR_RF1CS_CLK11 | CMXFCR_TF1CS_CLK12) #define CFG_CMXFCR_MASK1 (CMXFCR_FC1 | CMXFCR_RF1CS_MSK | CMXFCR_TF1CS_MSK) #define CFG_CMXFCR_VALUE3 (CMXFCR_RF3CS_CLK13 | CMXFCR_TF3CS_CLK14) #define CFG_CMXFCR_MASK3 (CMXFCR_FC3 | CMXFCR_RF3CS_MSK | CMXFCR_TF3CS_MSK) #define CFG_CMXFCR_VALUE2 (CMXFCR_RF2CS_CLK15 | CMXFCR_TF2CS_CLK16) #define CFG_CMXFCR_MASK2 (CMXFCR_FC2 | CMXFCR_RF2CS_MSK | CMXFCR_TF2CS_MSK) #define CFG_RX_ETH_BUFFER 128 #define CONFIG_IPADDR 192.168.10.17 #define CONFIG_SERVERIP 192.168.10.101 #define CONFIG_HOSTNAME OceanSCC #define CONFIG_LOADADDR 1000 /* default location for tftp and bootm */ #define CONFIG_BOOTFILE u-boot.bin #define CONFIG_ETHADDR 00:13:09:11:22:33 #define CONFIG_HAS_ETH1 #define CONFIG_ETH1ADDR 00:13:09:04:05:36 #define CONFIG_HAS_ETH2 #define CONFIG_ETH2ADDR 00:13:09:07:08:39 #define CONFIG_EXTRA_ENV_SETTINGS "ethprime=FCC1 ETHERNET" #define CFG_CPMFCR_RAMTYPE 0 /* BDs and buffers on 60x bus */ #define CFG_FCC_PSMR (FCC_PSMR_FDE | FCC_PSMR_LPB) /* Full duplex */ #define CONFIG_MII /* MII PHY management */ #define CONFIG_BITBANGMII /* bit-bang MII PHY management */ Sincerely, Jie --------------------------------- Don't pick lemons. See all the new 2007 cars at Yahoo! Autos.

2 2

Re: [U-Boot-Users] AT91 NAND om AT91SAM9260EK
by Ivan Kuten 08 Feb '07

08 Feb '07

> > I tried to start the process of merging with mainstream U-boot > by sending in a patch which split the at45.c located in > board/at91rm9200dk into > * drivers/at45.c > * cpu/arm920t/at91rm9200/spi.c > but I have no feedback. > > Peter Pearse which will be responsible for maintaining the ARM part > is not up to speed yet, and noone else seems to be interested. > > There has been a discussion about how dataflash should be supported > but the things people do not like are not in this file, they > are in the "board/dataflash.c" file so I have no clue > why nothing is happening. > Hi Ulf, It's a good decision to move board/at91rm9200dk/at45.c into drivers/at45.c cause our custom RM9200 board uses at45.c too. We have dataflash only setup without NOR. Also may be as suggestion to make AT91C_SPI_PCS3_DATAFLASH_CARD #ifdefed because some boards such as ours may not use CS3 at all. Best regards, Ivan

1 0

Re: [U-Boot-Users] Changes to U-Boot Development Process
by Wolfgang Denk 08 Feb '07

08 Feb '07

Hello, it is more than time for a summary of the reactions to our proposal to split the U-Boot development into several subsystems and to share the responsibility for these in a group of "custodians". Here it comes: 1) The reaction was all positive. So we will take the next steps in that direction. 2) There was a general agreement that it would be beneficial to have all the custodian's (public) repositories located on the same server, so we willdo that. If nobody objects, we will host this on www.denx.de We intend to name these repositories after their function (rather, for example, after the name of the custodian), so we don't have to modify anything in case there should be a change of the custodians. 3) It was no problem to find custodians for all suggested areas. Thanks a lot to all of you who volunteered! You can find an overview about the current situation at http://www.denx.de/wiki/UBoot/Custodians This also includes the suggested names for the git repositories. Please feel free to comment. To avoid conflicts, I would like to establish a strict match between a repository and a custodian, i. e. each of the git repositories can be pushed into by exactly one custodian. As you can see, we already have mostly such matches, only for MIPS we have three instead of one volunteer. It's really great to receive so much voluntary help, but in the end we can have only one custodian for MIPS. However, for me it is impossible to decide who to chose of Robert, Rodolfo or Hans. Can you please try to sort this out among the three of you, eventually with feedback from the list? Custodians: can each of you please send me your public SSH keys? We will then set up the git repositories so we can get the whole thing rocking ASAP... [If you don't want to send it by e-mail, you can also use jabber: wd(a)jabber.denx.de]. Thanks in advance, Wolfgang Denk -- DENX Software Engineering GmbH, HRB 165235 Munich, CEO: Wolfgang Denk Office: Kirchenstr. 5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd(a)denx.de If the facts don't fit the theory, change the facts. -- Albert Einstein

2 1

Re: [U-Boot-Users] U-Boot-Users Digest, Vol 9, Issue 15
by 邹伟 08 Feb '07

08 Feb '07

Hi,everybody! I am new to u-boot.These days ,I am porting u-boot to a changed SMDK2410 .There is a dm9000a chip replacing the cs8900,and I don't use ROM for dm9000a.So I have used a static data struct(C langue)to replace reading MAC from the ROM.While it seems to wrong: when I not configed CONFIG_DM9000_DEBUG,there would be timeout,the serial printed something like this :##TTTTTTT##TTTTTT,but ,as long as configed it,the TFTP works very well,about 10k/ps. I am sorry to disturbed you,best wishes! Thank you,everyone! hongyualang

1 0

[U-Boot-Users] How to determine the values assigned to -a and -e options of mkimage tool?
by Jarod Wang 08 Feb '07

08 Feb '07

Hi all I'm porting u-boot-1.1.4 to an smdk2410-like custom board to boot Linux 2.4.18-rmk7 kernel by using cross-2.95.3 as cross-compiler on my CentOS 4.4 powered i386 box. The board got a 64MB SDRAM mapped into 0x3000 0000 - 0x3400 0000 in S3C2410A's address space. And yesterday I could see the output in minicom, but I was confused by the options of tools/mkimage when I tried to create an uImage file. According to the README file there are two options '-a' and '-e' which denote the 'load address' and 'entry point(address)' respectively. So I set both of them to 0x30008000 when I built my uImage file according to the following lines in arch/arm/boot/Makefile: ifeq ($(CONFIG_ARCH_S3C2410),y) ZTEXTADDR = 0x30008000 ZRELADDR = 0x30008000 endif My question is whether these two values should be assigned to -a and -e options, and what about the value of CFG_LOAD_ADDR in configuration header file? Any hints would be highly appreciate. Best regards, Jarod Wang

1 0

[U-Boot-Users] [PATCH] Rework flash_read_uchar(), flash_read_ushort(), flash_read_long() for reading flash safety.
by Zhang Wei 08 Feb '07

08 Feb '07

Signed-off-by: Zhang Wei <wei.zhang(a)freescale.com> --- drivers/cfi_flash.c | 107 +++++++++++++++++++++++---------------------------- 1 files changed, 48 insertions(+), 59 deletions(-) diff --git a/drivers/cfi_flash.c b/drivers/cfi_flash.c index 696f9a4..8289b21 100644 --- a/drivers/cfi_flash.c +++ b/drivers/cfi_flash.c @@ -250,83 +250,72 @@ static void flash_printqry (flash_info_t #endif -/*----------------------------------------------------------------------- - * read a character at a port width address - */ -inline uchar flash_read_uchar (flash_info_t * info, uint offset) +static inline ulong flash_read(flash_info_t * info, + flash_sect_t sect, uint offset, int data_size) { - uchar *cp; + volatile cfiptr_t addr; + unsigned long retval = 0; + int i; - cp = flash_make_addr (info, 0, offset); + addr.cp = flash_make_addr(info, sect, offset); + for (i = 0; i < data_size; i++) { + cfiword_t cw; + switch (info->portwidth) { + case FLASH_CFI_8BIT: + cw.c = addr.cp[i]; + break; + case FLASH_CFI_16BIT: + cw.w = addr.wp[i]; + break; + case FLASH_CFI_32BIT: + cw.l = addr.lp[i]; + break; + case FLASH_CFI_64BIT: + cw.ll = addr.llp[i]; + break; + default: + return 0; + } #if defined(__LITTLE_ENDIAN) || defined(CFG_WRITE_SWAPPED_DATA) - return (cp[0]); + retval |= (cw.c << (8 * i)); #else - return (cp[info->portwidth - 1]); + retval |= (((cw.ll >> (8 * (FLASH_CFI_64BIT - info->portwidth))) + & 0xff) << (8 * i)); #endif + } + + debug ("flash_read retval = 0x%x\n", retval); + return retval; } /*----------------------------------------------------------------------- - * read a short word by swapping for ppc format. + * read a character at a port width address */ -ushort flash_read_ushort (flash_info_t * info, flash_sect_t sect, uint offset) +inline uchar flash_read_uchar (flash_info_t * info, uint offset) { - uchar *addr; - ushort retval; - -#ifdef DEBUG - int x; -#endif - addr = flash_make_addr (info, sect, offset); - -#ifdef DEBUG - debug ("ushort addr is at %p info->portwidth = %d\n", addr, - info->portwidth); - for (x = 0; x < 2 * info->portwidth; x++) { - debug ("addr[%x] = 0x%x\n", x, addr[x]); - } -#endif -#if defined(__LITTLE_ENDIAN) || defined(CFG_WRITE_SWAPPED_DATA) - retval = ((addr[(info->portwidth)] << 8) | addr[0]); -#else - retval = ((addr[(2 * info->portwidth) - 1] << 8) | - addr[info->portwidth - 1]); -#endif + return (uchar)(flash_read(info, 0, offset, sizeof(uchar))); +} - debug ("retval = 0x%x\n", retval); - return retval; +/*----------------------------------------------------------------------- + * read a short word. + */ +ushort flash_read_ushort (flash_info_t * info, flash_sect_t sect, uint offset) +{ + return (ushort)(flash_read(info, sect, offset, sizeof(ushort))); } /*----------------------------------------------------------------------- * read a long word by picking the least significant byte of each maximum - * port size word. Swap for ppc format. + * port size word. */ ulong flash_read_long (flash_info_t * info, flash_sect_t sect, uint offset) { - uchar *addr; - ulong retval; - -#ifdef DEBUG - int x; -#endif - addr = flash_make_addr (info, sect, offset); - -#ifdef DEBUG - debug ("long addr is at %p info->portwidth = %d\n", addr, - info->portwidth); - for (x = 0; x < 4 * info->portwidth; x++) { - debug ("addr[%x] = 0x%x\n", x, addr[x]); - } -#endif -#if defined(__LITTLE_ENDIAN) || defined(CFG_WRITE_SWAPPED_DATA) - retval = (addr[0] << 16) | (addr[(info->portwidth)] << 24) | - (addr[(2 * info->portwidth)]) | (addr[(3 * info->portwidth)] << 8); -#else - retval = (addr[(2 * info->portwidth) - 1] << 24) | - (addr[(info->portwidth) - 1] << 16) | - (addr[(4 * info->portwidth) - 1] << 8) | - addr[(3 * info->portwidth) - 1]; -#endif - return retval; + ulong rval; + rval = (((flash_read(info, sect, offset, sizeof(ushort)) + & 0xffff) << 16 ) | + (flash_read(info, sect, offset + FLASH_CFI_16BIT, sizeof(ushort)) + & 0xffff)); + return rval; } -- 1.4.0

1 0

[U-Boot-Users] How to determine the values assigned to -a and -e options of mkimage tool?
by Jarod Wang 08 Feb '07

08 Feb '07

Hi all I'm porting u-boot-1.1.4 to an smdk2410-like custom board to boot Linux 2.4.18-rmk7 kernel by using cross-2.95.3 as cross-compiler on my CentOS 4.4 powered i386 box. The board got a 64MB SDRAM mapped into 0x3000 0000 - 0x3400 0000 in S3C2410A's address space. And yesterday I could see the output in minicom, but I was confused by the options of tools/mkimage when I tried to create an uImage file. According to the README file there are two options '-a' and '-e' which denote the 'load address' and 'entry point(address)' respectively. So I set both of them to 0x30008000 when I built my uImage file according to the following lines in arch/arm/boot/Makefile: ifeq ($(CONFIG_ARCH_S3C2410),y) ZTEXTADDR = 0x30008000 ZRELADDR = 0x30008000 endif My question is whether these two values should be assigned to -a and -e options, and what about the value of CFG_LOAD_ADDR in configuration header file? Any hints would be highly appreciate. Best regards, Jarod Wang ___________________________________________________________ 抢注雅虎免费邮箱-3.5G容量，20M附件！ http://cn.mail.yahoo.com

1 0

Re: [U-Boot-Users] [PATCH] Define FLASH_CMD_SYNC for MPC8641
by Tolunay Orkun 08 Feb '07

08 Feb '07

haiying.wang(a)freescale.com wrote: > > diff --git a/include/configs/MPC8641HPCN.h b/include/configs/MPC8641HPCN.h > index 0bc8324..cbc635f 100644 > --- a/include/configs/MPC8641HPCN.h > +++ b/include/configs/MPC8641HPCN.h > @@ -204,6 +204,7 @@ #define CFG_MONITOR_BASE TEXT_BASE / > #define CFG_FLASH_CFI_DRIVER > #define CFG_FLASH_CFI > #define CFG_FLASH_EMPTY_INFO > +#define FLASH_CMD_SYNC No, this is not going to be board specific option. See my other email for more details. Tolunay

1 0

Re: [U-Boot-Users] [PATCH] Add FLASH_CMD_SYNC macro for flash_write_cmd
by Tolunay Orkun 08 Feb '07

08 Feb '07

haiying.wang(a)freescale.com wrote: > +#ifdef FLASH_CMD_SYNC > + sync(); > +#endif > } > I think you misunderstood the last change request regarding your patch. Please re-read my earlier email. As discussed on the mailing list we do not need/want a FLASH_CMD_SYNC. 1) In your cfi_flash.c::flash_write_cmd() patch instead of #ifdef FLASH_CMD_SYNC sync(); #endif Use SYNC; 2) In a separate patch define SYNC for all CPU architectures (i.e. add the definition of SYNC to all CPU header files. For example, each PowerPC based CPU support file will have the following line added. #define SYNC asm("sync;") You will have to add this to PowerPC and BLACKFIN at this moment. Others will need to have a dummy define so builds must not break. #define SYNC /* stub */ Please re-do the patches and resubmit.

1 0