January 2009 - U-Boot

[U-Boot] [PATCH][u-boot-85xx]The 32bit errata fix for 8572 for DDR2 (fwd)
by Poonam_Aggrwal-b10812 19 Jan '09

19 Jan '09

This patch fixes the fix for 8572 SOC errata for 32bit DDR2. The earlier fix was trying to access the SDRAM_CFG register before initializing it. Now the SDRAM_CFG is initialized before. But the MEM_EN is set in the end. Signed-off-by: Poonam Aggrwal-b10812 <b10812(a)freescale.net> Acked-by: Dave Liu <daveliu(a)freescale.com> --- cpu/mpc85xx/ddr-gen3.c | 9 ++++++++- include/asm-ppc/fsl_ddr_sdram.h | 17 +++++++++++++++++ 2 files changed, 25 insertions(+), 1 deletions(-) diff --git a/cpu/mpc85xx/ddr-gen3.c b/cpu/mpc85xx/ddr-gen3.c index a2b45c5..b9cb4e6 100644 --- a/cpu/mpc85xx/ddr-gen3.c +++ b/cpu/mpc85xx/ddr-gen3.c @@ -19,6 +19,7 @@ void fsl_ddr_set_memctl_regs(const fsl_ddr_cfg_regs_t *regs, { unsigned int i; volatile ccsr_ddr_t *ddr; + u32 temp_sdram_cfg; switch (ctrl_num) { case 0: @@ -78,6 +79,10 @@ void fsl_ddr_set_memctl_regs(const fsl_ddr_cfg_regs_t *regs, out_be32(&ddr->ddr_sdram_rcw_1, regs->ddr_sdram_rcw_1); out_be32(&ddr->ddr_sdram_rcw_2, regs->ddr_sdram_rcw_2); + /* Do not enable the memory */ + temp_sdram_cfg = in_be32(&ddr->sdram_cfg); + temp_sdram_cfg &= ~(SDRAM_CFG_MEM_EN); + out_be32(&ddr->sdram_cfg, temp_sdram_cfg); /* * For 8572 DDR1 erratum - DDR controller may enter illegal state * when operatiing in 32-bit bus mode with 4-beat bursts, @@ -99,7 +104,9 @@ void fsl_ddr_set_memctl_regs(const fsl_ddr_cfg_regs_t *regs, udelay(200); asm volatile("sync;isync"); - out_be32(&ddr->sdram_cfg, regs->ddr_sdram_cfg); + /* Let the controller go */ + temp_sdram_cfg = in_be32(&ddr->sdram_cfg); + out_be32(&ddr->sdram_cfg, temp_sdram_cfg | SDRAM_CFG_MEM_EN); /* Poll DDR_SDRAM_CFG_2[D_INIT] bit until auto-data init is done. */ while (in_be32(&ddr->sdram_cfg_2) & 0x10) { diff --git a/include/asm-ppc/fsl_ddr_sdram.h b/include/asm-ppc/fsl_ddr_sdram.h index b213af3..6e3b255 100644 --- a/include/asm-ppc/fsl_ddr_sdram.h +++ b/include/asm-ppc/fsl_ddr_sdram.h @@ -51,6 +51,23 @@ typedef ddr3_spd_eeprom_t generic_spd_eeprom_t; #define FSL_DDR_BANK_INTERLEAVING 0x2 #define FSL_DDR_SUPERBANK_INTERLEAVING 0x3 +/* DDR_SDRAM_CFG - DDR SDRAM Control Configuration + */ +#define SDRAM_CFG_MEM_EN 0x80000000 +#define SDRAM_CFG_SREN 0x40000000 +#define SDRAM_CFG_ECC_EN 0x20000000 +#define SDRAM_CFG_RD_EN 0x10000000 +#define SDRAM_CFG_SDRAM_TYPE_DDR1 0x02000000 +#define SDRAM_CFG_SDRAM_TYPE_DDR2 0x03000000 +#define SDRAM_CFG_SDRAM_TYPE_MASK 0x07000000 +#define SDRAM_CFG_SDRAM_TYPE_SHIFT 24 +#define SDRAM_CFG_DYN_PWR 0x00200000 +#define SDRAM_CFG_32_BE 0x00080000 +#define SDRAM_CFG_8_BE 0x00040000 +#define SDRAM_CFG_NCAP 0x00020000 +#define SDRAM_CFG_2T_EN 0x00008000 +#define SDRAM_CFG_BI 0x00000001 + /* Record of register values computed */ typedef struct fsl_ddr_cfg_regs_s { struct { -- 1.5.6.5

3 2

[U-Boot] General Flash Programming Error
by Jerry Zhang 19 Jan '09

19 Jan '09

Hello, u-boot world! I have wrote u-boot (U-Boot 2009.01-rc1 ) into my smdk2410 board. But when i issued protect off all and then setenv command , i got error as following: SMDK2410 # saveenv Saving Environment to Flash... Un-Protected 1 sectors Erasing Flash...Erasing sector 10 ... ok. Erased 1 sectors Writing to Flash... General Flash Programming Error Protected 1 sectors Why could it be this? How to fix it? Another question, how to get one variable's Location in flash with its name? Thanks in advance! Jerry.Zhang

1 0

[U-Boot] [PATCH] NAND: cmd_nand.c: fix invalid pointers to static relocated chip names
by gvv＠horse.lst.loc 19 Jan '09

19 Jan '09

When U-boot starting from NAND the pointers to chip names are set before relocation. If the original memory block is overwritten after relocation that led to invalid string data in the nand utility. Signed-off-by: Valeriy Glushkov <gvv(a)lstec.com> --- diff --git a/common/cmd_nand.c b/common/cmd_nand.c index b94a2bf..14925cd 100644 --- a/common/cmd_nand.c +++ b/common/cmd_nand.c @@ -10,6 +10,7 @@ #include <common.h> +DECLARE_GLOBAL_DATA_PTR; #ifndef CONFIG_NAND_LEGACY /* @@ -193,7 +194,7 @@ int do_nand(cmd_tbl_t * cmdtp, int flag, int argc, char *argv[]) for (i = 0; i < CFG_MAX_NAND_DEVICE; i++) { if (nand_info[i].name) printf("Device %d: %s, sector size %u KiB\n", - i, nand_info[i].name, + i, nand_info[i].name + gd->reloc_off, nand_info[i].erasesize >> 10); } return 0; @@ -207,7 +208,7 @@ int do_nand(cmd_tbl_t * cmdtp, int flag, int argc, char *argv[]) puts("\nno devices available\n"); else printf("\nDevice %d: %s\n", nand_curr_device, - nand_info[nand_curr_device].name); + nand_info[nand_curr_device].name + gd->reloc_off); return 0; } dev = (int)simple_strtoul(argv[2], NULL, 10); @@ -215,7 +216,7 @@ int do_nand(cmd_tbl_t * cmdtp, int flag, int argc, char *argv[]) puts("No such device\n"); return 1; } - printf("Device %d: %s", dev, nand_info[dev].name); + printf("Device %d: %s", dev, nand_info[dev].name + gd->reloc_off); puts("... is now current device\n"); nand_curr_device = dev; @@ -993,8 +994,8 @@ int do_nandboot (cmd_tbl_t *cmdtp, int flag, int argc, char *argv[]) show_boot_progress (55); printf ("\nLoading from device %d: %s at 0x%lx (offset 0x%lx)\n", - dev, nand_dev_desc[dev].name, nand_dev_desc[dev].IO_ADDR, - offset); + dev, nand_dev_desc[dev].name + gd->reloc_off, + nand_dev_desc[dev].IO_ADDR, offset); if (nand_legacy_rw (nand_dev_desc + dev, NANDRW_READ, offset, SECTORSIZE, NULL, (u_char *)addr)) { -- 1.5.2.5

3 2

[U-Boot] [PATCH] NAND: cmd_nand.c: fix invalid pointers to static relocated chip names
by gvv＠horse.lst.loc 19 Jan '09

19 Jan '09

When U-boot starting from NAND the pointers to chip names are set before relocation. If the original memory block is overwritten after relocation that led to invalid string data in the nand utility. Signed-off-by: Valeriy Glushkov <gvv(a)lstec.com> --- diff --git a/common/cmd_nand.c b/common/cmd_nand.c index b94a2bf..14925cd 100644 --- a/common/cmd_nand.c +++ b/common/cmd_nand.c @@ -10,6 +10,7 @@ #include <common.h> +DECLARE_GLOBAL_DATA_PTR; #ifndef CONFIG_NAND_LEGACY /* @@ -193,7 +194,7 @@ int do_nand(cmd_tbl_t * cmdtp, int flag, int argc, char *argv[]) for (i = 0; i < CFG_MAX_NAND_DEVICE; i++) { if (nand_info[i].name) printf("Device %d: %s, sector size %u KiB\n", - i, nand_info[i].name, + i, nand_info[i].name + gd->reloc_off, nand_info[i].erasesize >> 10); } return 0; @@ -207,7 +208,7 @@ int do_nand(cmd_tbl_t * cmdtp, int flag, int argc, char *argv[]) puts("\nno devices available\n"); else printf("\nDevice %d: %s\n", nand_curr_device, - nand_info[nand_curr_device].name); + nand_info[nand_curr_device].name + gd->reloc_off); return 0; } dev = (int)simple_strtoul(argv[2], NULL, 10); @@ -215,7 +216,7 @@ int do_nand(cmd_tbl_t * cmdtp, int flag, int argc, char *argv[]) puts("No such device\n"); return 1; } - printf("Device %d: %s", dev, nand_info[dev].name); + printf("Device %d: %s", dev, nand_info[dev].name + gd->reloc_off); puts("... is now current device\n"); nand_curr_device = dev; @@ -993,8 +994,8 @@ int do_nandboot (cmd_tbl_t *cmdtp, int flag, int argc, char *argv[]) show_boot_progress (55); printf ("\nLoading from device %d: %s at 0x%lx (offset 0x%lx)\n", - dev, nand_dev_desc[dev].name, nand_dev_desc[dev].IO_ADDR, - offset); + dev, nand_dev_desc[dev].name + gd->reloc_off, + nand_dev_desc[dev].IO_ADDR, offset); if (nand_legacy_rw (nand_dev_desc + dev, NANDRW_READ, offset, SECTORSIZE, NULL, (u_char *)addr)) { -- 1.5.2.5

1 0

Re: [U-Boot] U-boot as a mass storage device and DFU support
by Jacky Lam 19 Jan '09

19 Jan '09

Jean-Christophe PLAGNIOL-VILLARD ??: > On 17:59 Fri 16 Jan , Jacky Lam wrote: >> Dear all, >> >> Could anyone kindly tell me if u-boot supports itself as a mass storage >> device? > no, maybe MTP will be simplest to implement >> Or if u-boot main stream support the a protocol so call USB >> DFU(Universal Serial Bus Device Firmware Upgrade)? > no >> I have first seen this on openmoko.org and seems they use u-boot as >> well. However, when I look at the newest u-boot (1.3.4). It seems no >> code is about it. I don't know if I have overlooked something. > I've ssen it in the u-boot dev tree of openmoko but they never do the > necessary to add it to the Mainline > > feel free to send patch > Would it be appropriate to send the patch if I am not the original author??? > Best Regards, > J. > > >

2 1

[U-Boot] [PATCH 00/15 v6] OMAP3: Add support for some of TIs ARM-Cortex A8 OMAP3 boards
by Dirk Behme 19 Jan '09

19 Jan '09

This patch series adds U-Boot v1 support for some of TI's ARM-Cortex A8 based OMAP3 boards. These are BeagleBoard [1][2], EVM [3], Overo [4], Pandora [5] and Zoom1 [6]. The patch series is based on U-Boot tar ball [7] for BeagleBoard and EVM done by several TI employees. To be able to easily add new boards, most of the code is common for all boards. After the header files (patches 1 - 2) and the common code for ARM Cortex A8 (patch 3) and OMAP3 SoC (patches 4 & 5), support and drivers for NAND, MMC and I2C are added (patches 6 - 8). Patches 9 - 13 introduce the individual board support for Beagle, EVM, Overo, Pandora and Zoom1. The last two patches 14 & 15 add the config files for these boards and enable them by adding main Makefile, Maintainers, MAKEALL and README. As discussed earlier on this list, we compile for armv5 to be compatible with existing toolchains. This patch adds ~300k of new code. Thanks to Steve Sakoman <sakoman at gmail.com> Pillai, Manikandan <mani.pillai at ti.com> Syed Mohammed, Khasim <khasim at ti.com> Nishanth Menon <nm at ti.com> Grazvydas Ignotas <notasas at gmail.com> Jason Kridner <jkridner at gmail.com> and all others who helped with this. Dirk This patch series is against U-Boot mainline, commit 89d56f5503eed351efe5ab0b4dd0f1e888fd2336 "Merge branch 'master' of git://git.denx.de/u-boot-ppc4xx" This OMAP3 v6 patch series makes all older OMAP3 patches obsolete. Changes in version v6: - Clean up coding style and whitespace - Replace hardcoded values by macros - Convert readx/writex register access to register struct style - Add L1 NEON HW workaround - Fix issues in NAND HW/SW ECC switch - Add Pandora board - Add Zoom1 board - Clean up config files and add/enable 'hush' scripting - Add check for MMC card - Remove BITx magic - In MAKEALL, move OMAP3 from ARM9 to Cortex A8 section - Update EVM maintainer - Update ARM's if then else logic in examples Makefile - Introduce CONFIG_OMAP3_MMC - Move common power initialization code to common board directory - Fix GPL headers [1] http://beagleboard.org/ [2] http://elinux.org/BeagleBoard [3] http://focus.ti.com/docs/toolsw/folders/print/tmdxevm3503.html [4] http://www.gumstix.net/Overo/ [5] http://openpandora.org/ [6] http://www.logicpd.com/products/devkit/ti/zoom_mobile_development_kit [7] http://beagleboard.googlecode.com/files/u-boot_beagle_revb.tar.gz

4 22

[U-Boot] configuring the board as a pci target in u-boot
by arun c 19 Jan '09

19 Jan '09

Hi all, I want to configure my board as a PCI target, its based on mcf548x processor. What are the things to be taken care to configure a board as a PCI agent in u-boot? My processor can act as a PCI master as well as a target. I don't want master interface at all. I want to make my board as a PCI slave and let it respond to other masters by simply sitting on a PCI bus. I have gone through the u-boot sources and I found that board/freescale/mpc8360emds/pci.c does something similar if CONFIG_PCISLAVE is defined. Its is setting up inbound windows and using pci_setup_indirect(), pci_hose_read_config_word(), pci_hose_write_config_word() functions to make the board a pci slave. I don't really understand the use of these functions, if i am able to read and write PCI config registers(256 bytes) through my internal bus why these are required? As per my understanding I just have to configure the PCI inbound windows and leave the rest to PCI master of the entire system. am I thinking wrong? Regards, Arun C

2 2

[U-Boot] [PATCH v3 0/5] usb : musb : Add MUSB contorller driver and enable Davinci MSC support
by Thomas Abraham 19 Jan '09

19 Jan '09

This patch series adds Mentor Graphics USB OTG host controller driver support and MSC support for DM6446 (TI DaVinci) platform. - [PATCH v3 1/5] usb : musb : Adding host controller driver for Mentor USB controller. - [PATCH v3 2/5] usb : musb : Adding USB VBUS enable functionality for DM6446 DVEVM. - [PATCH v3 3/5] usb : musb : Adding DM6446 (TI DaVinci) platform specific USB support. - [PATCH v3 4/5] usb : musb : Enabling DM6446 (TI DaVinci) USB module power. - [PATCH v3 5/5] usb : musb : Enabling USB MSC support for DM6446 (TI DaVinci) platform. Note: This patch series depends on the following patch submitted by Michael Trimarchi. - [U-Boot] [USB PATCH] USB change speed Changes since v2: - Addressed changes suggested by Dirk Behme - Added register overlay structure to access mentor USB module registers Changes since v1: - Rebased to U-Boot USB next branch - Removed high element from usb_device structure

6 8

[U-Boot] User supplied lowlevel initialization code on at91 plattforms
by Michael Roth 19 Jan '09

19 Jan '09

Hello, I'm hacking U-Boot to a new at91sam9261 board and would like to include a little bit low level init code (pll, sdram, etc.). Normally, in U-Boot it is assumed that on at91 plattforms the bootstrap code lives in dataflash which do the job. On the board I'm hacking there isn't any dataflash but only nor flash from that the board boots. So my question is: How can I supply an self written lowlevel_init.S in my board directoy that get linked instead the default cpu/arm926ejs/at91/lowlevel_init.S code? Are there any configuration variables to trick the U-Boot makefiles to use my lowlevel_init.S? Michael Roth

2 2

[U-Boot] v2009.01-rc3 is out - update of release schedule
by Wolfgang Denk 18 Jan '09

18 Jan '09

Hi everybody... v2009.01-rc3 has just been released. Please help testing. If no new bugs are found and/or other urgent stuff needs to get pulled I plan to release v2009.01 on Wednesday, January 21. Best regards, Wolfgang Denk -- DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd(a)denx.de Minds are like parachutes - they only function when open.

2 1