January 2013 - U-Boot

Re: [U-Boot] [PATCH 1/2] Davinci: Make MAC address offset in EEPROM configurable, CONFIG_MAC_EEPROM_ADDR
by Tom Rini 31 Jan '13

31 Jan '13

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 01/30/2013 03:10 AM, Kim Bøndergaard Poulsen wrote: > Hi Tom > > >> -----Original Message----- From: Tom Rini [mailto:trini@ti.com] >> Sent: 29. januar 2013 21:32 To: Kim Bøndergaard Poulsen Cc: >> u-boot(a)lists.denx.de; vapier(a)gentoo.org; manjunath.hadli(a)ti.com >> Subject: Re: [PATCH 1/2] Davinci: Make MAC address offset in >> EEPROM configurable, CONFIG_MAC_EEPROM_ADDR >> > On 01/29/2013 02:28 AM, Kim B￸ndergaard wrote: >>>> --- arch/arm/cpu/arm926ejs/davinci/misc.c | 5 ++++- 1 file >>>> changed, 4 insertions(+), 1 deletion(-) >>>> >>>> diff --git a/arch/arm/cpu/arm926ejs/davinci/misc.c >>>> b/arch/arm/cpu/arm926ejs/davinci/misc.c index >>>> c310c69..d0389a4 100644 --- >>>> a/arch/arm/cpu/arm926ejs/davinci/misc.c +++ >>>> b/arch/arm/cpu/arm926ejs/davinci/misc.c @@ -55,11 +55,14 @@ >>>> void dram_init_banksize(void) * Read ethernet MAC address >>>> from EEPROM for DVEVM compatible boards. * Returns 1 if >>>> found, 0 otherwise. */ +#ifndef CONFIG_MAC_EEPROM_ADDR >>>> +#define > CONFIG_MAC_EEPROM_ADDR >>>> 0x7F00 +#endif int dvevm_read_mac_address(uint8_t *buf) { >>>> #ifdef CONFIG_SYS_I2C_EEPROM_ADDR /* Read MAC address. */ - >>>> if (i2c_read(CONFIG_SYS_I2C_EEPROM_ADDR, 0x7F00, + if >>>> (i2c_read(CONFIG_SYS_I2C_EEPROM_ADDR, > CONFIG_MAC_EEPROM_ADDR, >>>> CONFIG_SYS_I2C_EEPROM_ADDR_LEN, (uint8_t *) &buf[0], 6)) >>>> goto i2cerr; > > Lets use CONFIG_SYS_I2C_EEPROM_MAC_OFFSET and just set it in all > of the davinci configs (yes, that's touching more files). Thanks! > > >> So what you suggest is to add a #define >> CONFIG_SYS_I2C_EEPROM_MAC_OFFSET 0x7F00 > >> in all DAVINCI related include/configs/*.h ? > >> Also I assume you want me to remove the #ifdef >> CONFIG_SYS_I2C_EEPROM_MAC_OFFSET in misc.c? > >> My only problem is that I don't know if it makes sense to define >> it for all boards > >> grep -l DAVINCI * in include/configs gives me the following >> files: calimain.h cam_enc_4xx.h da830evm.h da850evm.h >> davinci_dm355evm.h davinci_dm355leopard.h davinci_dm365evm.h >> davinci_dm6467evm.h davinci_dvevm.h davinci_schmoogie.h >> davinci_sffsdr.h davinci_sonata.h ea20.h enbw_cmc.h hawkboard.h >> tnetv107x_evm.h > >> Anyone who knows if they all have I2C eeproms The subset of these with CONFIG_SYS_I2C_EEPROM set to start with, should as they have been reading there already. Cc'ing the various board maintainers with the patch would be good. - -- Tom -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQIcBAEBAgAGBQJRCu/WAAoJENk4IS6UOR1WKBcP/3bE685l64rhKRGhsVp7zdsT 3XsD4oKMqIeFiHwiSiiS5PXuRibgt3h3aLgbbJnrhMGKBrXsDkEvb9Mjg3f/2yqg duuWIiER6utP2WW16PzjwR5Y/RNjjttzBHfDKsZCOtoZqA8v2+MmnMNrcBZmGI48 foimzTD2VikxO63UElzX3vB83JyFkBaicNfk06vjE73wU2dRR1dnPi5+ufbyMp8r ra9BI5++xVQHBmeZiGngRkO2tESll+L6L9HiTp3NNcNU3TWvV0675lwTnmXXHYy5 yoQHCI4wXQteBixPybp/903p3BhFbw8jRnAok0icSP2chcp8Dvjt8z8n29/JCxYZ 1ZOJdfbrPQXg68Un9+AQnXIoP1wby0ldkVnUfySVNL+ZzKXNA249idx8aY4s1V/K 8jBwP9xMhEaCVirP/DnLV3MetXNKy8hiE9pXqMUhuRn1xe3GwUGV4VUC0EgHJNfx iK8Ne/p68Oeol8WHrWqG3gG9UMZgSzUqWKMSeSIILclLoo8DXLNRM2HM630rfgdA E1MMzpBHPp4XB7RArlWXmibbh9A+ZLFraUZzCp75+OF3Ox0Tdn+vPKyMkf5qXV8w WMZSKZBQbFI2xAYzi6YlcqBP2jC/6e2zOwFXvehQErRI51H5gLMsv0MoPnPx54t4 sUbq0r0V0EdaH2LCS2iM =zDPr -----END PGP SIGNATURE-----

1 0

[U-Boot] [PATCH] vfat: Fix mkcksum argument sizes
by Marek Vasut 31 Jan '13

31 Jan '13

In case a function argument is known/fixed size array in C, the argument is still decoyed as pointer instead ( T f(U n[k]) ~= T fn(U *n) ) and therefore calling sizeof on the function argument will result in the size of the pointer, not the size of the array. The VFAT code contains such a bug, this patch fixes it. Reported-by: Aaron Williams <Aaron.Williams(a)cavium.com> Signed-off-by: Marek Vasut <marex(a)denx.de> Cc: Tom Rini <tom.rini(a)gmail.com> Cc: Aaron Williams <Aaron.Williams(a)cavium.com> --- fs/fat/fat.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/fat/fat.c b/fs/fat/fat.c index 393c378..25d3318 100644 --- a/fs/fat/fat.c +++ b/fs/fat/fat.c @@ -569,9 +569,9 @@ static __u8 mkcksum(const char name[8], const char ext[3]) __u8 ret = 0; - for (i = 0; i < sizeof(name); i++) + for (i = 0; i < 8; i++) ret = (((ret & 1) << 7) | ((ret & 0xfe) >> 1)) + name[i]; - for (i = 0; i < sizeof(ext); i++) + for (i = 0; i < 3; i++) ret = (((ret & 1) << 7) | ((ret & 0xfe) >> 1)) + ext[i]; return ret; -- 1.7.10.4

4 4

[U-Boot] anyone implemented a failsafe upgrade involving two kernel/rootfs pairs?
by Robert P. J. Day 31 Jan '13

31 Jan '13

i want to do pretty much what's described here: https://github.com/DFE/HidaV/wiki/Safe-Update-and-boot-fallbacks that is, a single u-boot and config in flash, followed by a pair of kernel/rootfs partitions, that is remotely upgradeable by, from userspace, determining the "older" (backup) of the pairs, overwriting that pair with a new pair, then flipping the switch in the u-boot configuration to always boot the newer kernel and rootfs (with an emergency revert to the backup pair under circumstances i haven't imagined yet). it *looks* fairly straightforward, but while that's a nice write-up, it does warn right up front that it hasn't been implemented so if anyone has done something like this they care to share, that would be delightful. FYI, i'm using this pdlogic zoom am1808 experimenter kit: http://www.logicpd.com/products/system-on-modules/zoom-am1808/ for testing and, at the very end, the only change i would have to make is modify the solution to handle NOR flash that will be on the production board. if there's a writeup on this out there somewhere, that would be great. if not, if i can get this done, i'll be happy to write it up and post it publicly on my wiki for others. at the moment, i'm using openembedded to build a bootable kernel/rootfs pair for this board, and starting by adding the u-boot-fw-utils package so i have fw_{printenv/setenv} onboard. thanks for any pointers. rday p.s. i do have the bootable SD card that came with the kit, it's based on the arago project so i know i have something that boots to userspace as a starting point -- always good to know. -- ======================================================================== Robert P. J. Day Ottawa, Ontario, CANADA http://crashcourse.ca Twitter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================

2 3

[U-Boot] Virtex 4 with XMD
by Frank Lombardo 31 Jan '13

31 Jan '13

I am trying to get u-boot to run on a custom board with a Virtex 4 on it. I have started with the latest code from the u-boot-ppc4xx.git repository. I have been trying to get a simple (just a serial port) configuration of u-boot to run without success. The board has the following memory map: Address Map for Processor ppc405_0 (0b0100000000-0b0111111111) ppc405_0 (0000000000-0x03ffffff) SDR_SDRAM_CUSTOM ppc405_0_dplb1 (0000000000-0x03ffffff) SDR_SDRAM_CUSTOM ppc405_0_iplb1 (0x80800000-0x8080ffff) xps_epc_0 plb (0x81000000-0x8100ffff) xps_ethernetlite_0 plb (0x81010000-0x8101ffff) xps_ethernetlite_1 plb (0x81020000-0x8102ffff) xps_ethernetlite_2 plb (0x81800000-0x8180ffff) xps_intc_0 plb (0x83600000-0x8360ffff) SysACE_CompactFlash plb (0x84000000-0x8400ffff) xps_uart_console plb (0x84010000-0x8401ffff) xps_uart_1 plb (0x84020000-0x8402ffff) xps_uart_2 plb (0x84030000-0x8403ffff) xps_uart_3 plb (0x84040000-0x8404ffff) xps_uart_4 plb (0x84050000-0x8405ffff) xps_uart_5 plb (0x84060000-0x8406ffff) xps_uart_6 plb (0x85e00000-0x85e0ffff) plbv46_pci_0 plb (0x90000000-0x9fffffff) plbv46_pci_0 plb (0xf0000000-0xf000ffff) plbv46_pci_0 plb (0xffff8000-0xffffffff) xps_bram_if_cntlr_1 plb I have based my board's code on the ppc405-generic code. My board specific configs header file contains only a few items #include "../board/xilinx/ppc405-custom/xparameters.h" #define XPAR_PLB_CLOCK_FREQ_HZ (XPAR_CPU_PPC405_CORE_CLOCK_FREQ_HZ) #define XPAR_CORE_CLOCK_FREQ_HZ (XPAR_CPU_PPC405_CORE_CLOCK_FREQ_HZ) /* sdram */ #define CONFIG_SYS_SDRAM_SIZE_MB 64 #define XPAR_DDR2_SDRAM_MEM_BASEADDR (XPAR_SDR_SDRAM_CUSTOM_MPMC_BASEADDR) /* environment */ #define CONFIG_ENV_SIZE 0x1000 /*Misc*/ #define CONFIG_SYS_PROMPT "xlx-ppc405:/# " /* Monitor Command Prompt */ #define CONFIG_PREBOOT "echo U-Boot is up and runnining;" #include <configs/xilinx-ppc405.h> In boards.cfg I set the following: SYS_TEXT_BASE=0x00800000,RESET_VECTOR_ADDRESS=0x00900000 when I run in XMD, noting seems to happen, and when I stop the processor, the program counter seems to be at an odd address. See the output below: XMD% dow u-boot System Reset .... DONE Downloading Program -- u-boot section, .text: 0x00800000-0x008165df section, .resetvec: 0x00900000-0x00900003 section, .rodata: 0x008165e0-0x0081b810 section, .reloc: 0x0081b900-0x0081cf4b section, .data: 0x0081cf4c-0x0081e17b section, .u_boot_list: 0x0081e17c-0x0081e7b7 section, .bss: 0x00900004-0x00902333 Setting PC with Program Start Address 0x00800000 XMD% run Info:Processor started. Type "stop" to stop processor RUNNING> XMD% stop XMD% Info:User Interrupt, Processor Stopped at 0x7c230700 I have also observed that single stepping gets me to that odd address immediately. See the output below: XMD% dow u-boot System Reset .... DONE Downloading Program -- u-boot section, .text: 0x00800000-0x008165df section, .resetvec: 0x00900000-0x00900003 section, .rodata: 0x008165e0-0x0081b810 section, .reloc: 0x0081b900-0x0081cf4b section, .data: 0x0081cf4c-0x0081e17b section, .u_boot_list: 0x0081e17c-0x0081e7b7 section, .bss: 0x00900004-0x00902333 Setting PC with Program Start Address 0x00800000 XMD% stp 00800000: XMD% stp 7c230700: XMD% The main thing that worries me with this board is that there is no flash memory on the board. I am hoping that won't prevent me from executing u-boot with XMD. So, that leaves me with the following questions: 1) Is the lack of flash memory causing THIS problem? 2) Am I missing some config items? 3) Does anybody have any advice on how I can debug this? Thanks, Frank

2 2

[U-Boot] [PATCH] MIPS: start.S: remove obsolete 64 bit handling in setup_c0_status
by daniel.schwierzeck＠gmail.com 31 Jan '13

31 Jan '13

From: Daniel Schwierzeck <daniel.schwierzeck(a)gmail.com> Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck(a)gmail.com> --- arch/mips/cpu/mips32/start.S | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/arch/mips/cpu/mips32/start.S b/arch/mips/cpu/mips32/start.S index 51ce914..65acf7d 100644 --- a/arch/mips/cpu/mips32/start.S +++ b/arch/mips/cpu/mips32/start.S @@ -47,14 +47,6 @@ .set pop .endm - .macro setup_c0_status_reset -#ifdef CONFIG_64BIT - setup_c0_status ST0_KX 0 -#else - setup_c0_status 0 0 -#endif - .endm - #define RVECENT(f,n) \ b f; nop #define XVECENT(f,bev) \ @@ -222,7 +214,7 @@ reset: /* WP(Watch Pending), SW0/1 should be cleared */ mtc0 zero, CP0_CAUSE - setup_c0_status_reset + setup_c0_status 0 0 /* Init Timer */ mtc0 zero, CP0_COUNT -- 1.8.1.1

2 1

[U-Boot] [PATCH v2] arm: fix CONFIG_DELAY_ENVIRONMENT to act like it claims in the README
by Lucas Stach 31 Jan '13

31 Jan '13

No one expects to end up in a delayed environment if CONFIG_DELAY_ENVIRONMENT isn't defined. Signed-off-by: Lucas Stach <dev(a)lynxeye.de> --- v2: keep preference of CONFIG_OF_CONTROL and just change default value --- arch/arm/lib/board.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/lib/board.c b/arch/arm/lib/board.c index cfe32cc..9f861cc 100644 --- a/arch/arm/lib/board.c +++ b/arch/arm/lib/board.c @@ -488,7 +488,7 @@ static char *failed = "*** failed ***\n"; static int should_load_env(void) { #ifdef CONFIG_OF_CONTROL - return fdtdec_get_config_int(gd->fdt_blob, "load-environment", 0); + return fdtdec_get_config_int(gd->fdt_blob, "load-environment", 1); #elif defined CONFIG_DELAY_ENVIRONMENT return 0; #else -- 1.8.0.2

5 6

[U-Boot] [GIT PULL] u-boot-mips/master
by Daniel Schwierzeck 31 Jan '13

31 Jan '13

The following changes since commit 54b08efcf2f4ff532ce99c53f341a59c193331a5: README.mips: update known issues and TODOs (2013-01-16 10:52:08 +0100) are available in the git repository at: git://git.denx.de/u-boot-mips.git master for you to fetch changes up to e93b98e216831d56b3fb0150f9723b1a15fb42a2: MIPS: qi_lb60: remove custom u-boot.lds script (2013-01-31 00:27:58 +0100) ---------------------------------------------------------------- Gabor Juhos (18): MIPS: use inline directive for __in*s functions MIPS: convert IO port accessor functions to 'static inline' MIPS: start{, 64}.S: fill branch delay slots with NOP instructions MIPS: start.S: fix boundary check in relocate_code MIPS: start.S: set sp register directly MIPS: start.S: save reused arguments earlier in relocate_code MIPS: start.S: simplify relocation offset calculation MIPS: start.S: don't save flush_cache parameters in advance MIPS: simplify relocated _G_O_T_ address calculation MIPS: xburst: simplify relocation offset calculation MIPS: remove OUTPUT_FORMAT from linker scripts MIPS: add unified u-boot.lds file MIPS: qemu-mips: use the unified u-boot.lds script MIPS: pb1x00: remove custom u-boot.lds script MIPS: vct: remove custom u-boot.lds script MIPS: incaip: remove custom u-boot.lds script MIPS: dbau1x00: remove custom u-boot.lds script MIPS: qi_lb60: remove custom u-boot.lds script arch/mips/cpu/mips32/config.mk | 6 ++++++ arch/mips/cpu/mips32/start.S | 41 +++++++++++++---------------------------- arch/mips/cpu/mips64/start.S | 40 ++++++++++++---------------------------- {board/micronas/vct => arch/mips/cpu}/u-boot.lds | 45 ++++++++++++++++++++++++++++++--------------- arch/mips/cpu/xburst/config.mk | 6 ++++++ arch/mips/cpu/xburst/start.S | 19 +++++-------------- arch/mips/include/asm/io.h | 8 ++++---- board/dbau1x00/u-boot.lds | 69 --------------------------------------------------------------------- board/incaip/u-boot.lds | 69 --------------------------------------------------------------------- board/pb1x00/u-boot.lds | 69 --------------------------------------------------------------------- board/qemu-mips/u-boot.lds | 78 ------------------------------------------------------------------------------ board/qi/qi_lb60/u-boot.lds | 63 --------------------------------------------------------------- 12 files changed, 76 insertions(+), 437 deletions(-) rename {board/micronas/vct => arch/mips/cpu}/u-boot.lds (73%) delete mode 100644 board/dbau1x00/u-boot.lds delete mode 100644 board/incaip/u-boot.lds delete mode 100644 board/pb1x00/u-boot.lds delete mode 100644 board/qemu-mips/u-boot.lds delete mode 100644 board/qi/qi_lb60/u-boot.lds

2 1

[U-Boot] [PATCH v2 0/7] MIPS: use an unified u-boot.lds script
by Gabor Juhos 31 Jan '13

31 Jan '13

This patch-set adds an unfied u-boot.lds file for the MIPS architecture, and removes the linker scripts of the individual boards. The series depends on the following patch: MIPS: remove OUTPUT_FORMAT from linker scripts http://patchwork.ozlabs.org/patch/216892/ Gabor Juhos (7): MIPS: add unified u-boot.lds file MIPS: qemu-mips: use the unified u-boot.lds script MIPS: pb1x00: remove custom u-boot.lds script MIPS: vct: remove custom u-boot.lds script MIPS: incaip: remove custom u-boot.lds script MIPS: dbau1x00: remove custom u-boot.lds script MIPS: qi_lb60: remove custom u-boot.lds script arch/mips/cpu/u-boot.lds | 84 +++++++++++++++++++++++++++++++++++++++++ board/dbau1x00/u-boot.lds | 65 ------------------------------- board/incaip/u-boot.lds | 65 ------------------------------- board/micronas/vct/u-boot.lds | 68 --------------------------------- board/pb1x00/u-boot.lds | 65 ------------------------------- board/qemu-mips/u-boot.lds | 70 ---------------------------------- board/qi/qi_lb60/u-boot.lds | 61 ------------------------------ 7 files changed, 84 insertions(+), 394 deletions(-) create mode 100644 arch/mips/cpu/u-boot.lds delete mode 100644 board/dbau1x00/u-boot.lds delete mode 100644 board/incaip/u-boot.lds delete mode 100644 board/micronas/vct/u-boot.lds delete mode 100644 board/pb1x00/u-boot.lds delete mode 100644 board/qemu-mips/u-boot.lds delete mode 100644 board/qi/qi_lb60/u-boot.lds -- 1.7.10

2 1

[U-Boot] [PATCH] MIPS: remove OUTPUT_FORMAT from linker scripts
by Gabor Juhos 31 Jan '13

31 Jan '13

The OUTPUT_FORMAT command in linker scripts was always misused due to some endianess and toolchain problems. Use GCC flags to ensure proper output format, and get rid of the OUTPUT_FORMAT commands in the board specific u-boot.lds files. Signed-off-by: Gabor Juhos <juhosg(a)openwrt.org> Cc: Daniel Schwierzeck <daniel.schwierzeck(a)googlemail.com> Cc: Stefan Roese <sr(a)denx.de> Cc: Wolfgang Denk <wd(a)denx.de> Cc: Xiangfu Liu <xiangfu(a)openmobilefree.net> --- arch/mips/cpu/mips32/config.mk | 6 ++++++ arch/mips/cpu/xburst/config.mk | 6 ++++++ board/dbau1x00/u-boot.lds | 4 ---- board/incaip/u-boot.lds | 4 ---- board/micronas/vct/u-boot.lds | 1 - board/pb1x00/u-boot.lds | 4 ---- board/qemu-mips/u-boot.lds | 8 -------- board/qi/qi_lb60/u-boot.lds | 2 -- 8 files changed, 12 insertions(+), 23 deletions(-) diff --git a/arch/mips/cpu/mips32/config.mk b/arch/mips/cpu/mips32/config.mk index 481e984..7399701 100644 --- a/arch/mips/cpu/mips32/config.mk +++ b/arch/mips/cpu/mips32/config.mk @@ -30,5 +30,11 @@ MIPSFLAGS := -march=mips32r2 PLATFORM_CPPFLAGS += $(MIPSFLAGS) +PLATFORM_CPPFLAGS += -mabi=32 -DCONFIG_32BIT +ifdef CONFIG_SYS_BIG_ENDIAN +PLATFORM_LDFLAGS += -m elf32btsmip +else +PLATFORM_LDFLAGS += -m elf32ltsmip +endif CONFIG_STANDALONE_LOAD_ADDR ?= 0x80200000 -T mips.lds diff --git a/arch/mips/cpu/xburst/config.mk b/arch/mips/cpu/xburst/config.mk index 1536746..cf5fa6a 100644 --- a/arch/mips/cpu/xburst/config.mk +++ b/arch/mips/cpu/xburst/config.mk @@ -21,5 +21,11 @@ # PLATFORM_CPPFLAGS += -march=mips32 +PLATFORM_CPPFLAGS += -mabi=32 -DCONFIG_32BIT +ifdef CONFIG_SYS_BIG_ENDIAN +PLATFORM_LDFLAGS += -m elf32btsmip +else +PLATFORM_LDFLAGS += -m elf32ltsmip +endif CONFIG_STANDALONE_LOAD_ADDR ?= 0x80200000 -T mips.lds diff --git a/board/dbau1x00/u-boot.lds b/board/dbau1x00/u-boot.lds index 8a871cf..d9c01af 100644 --- a/board/dbau1x00/u-boot.lds +++ b/board/dbau1x00/u-boot.lds @@ -21,10 +21,6 @@ * MA 02111-1307 USA */ -/* -OUTPUT_FORMAT("elf32-bigmips", "elf32-bigmips", "elf32-bigmips") -*/ -OUTPUT_FORMAT("elf32-tradbigmips", "elf32-tradbigmips", "elf32-tradlittlemips") OUTPUT_ARCH(mips) ENTRY(_start) SECTIONS diff --git a/board/incaip/u-boot.lds b/board/incaip/u-boot.lds index 8a871cf..d9c01af 100644 --- a/board/incaip/u-boot.lds +++ b/board/incaip/u-boot.lds @@ -21,10 +21,6 @@ * MA 02111-1307 USA */ -/* -OUTPUT_FORMAT("elf32-bigmips", "elf32-bigmips", "elf32-bigmips") -*/ -OUTPUT_FORMAT("elf32-tradbigmips", "elf32-tradbigmips", "elf32-tradlittlemips") OUTPUT_ARCH(mips) ENTRY(_start) SECTIONS diff --git a/board/micronas/vct/u-boot.lds b/board/micronas/vct/u-boot.lds index 2ce8d0e..cc7db1b 100644 --- a/board/micronas/vct/u-boot.lds +++ b/board/micronas/vct/u-boot.lds @@ -21,7 +21,6 @@ * MA 02111-1307 USA */ -OUTPUT_FORMAT("elf32-tradbigmips", "elf32-tradbigmips", "elf32-tradlittlemips") OUTPUT_ARCH(mips) ENTRY(_start) SECTIONS diff --git a/board/pb1x00/u-boot.lds b/board/pb1x00/u-boot.lds index 07ddd36..d9c01af 100644 --- a/board/pb1x00/u-boot.lds +++ b/board/pb1x00/u-boot.lds @@ -21,10 +21,6 @@ * MA 02111-1307 USA */ -/* -OUTPUT_FORMAT("elf32-bigmips", "elf32-bigmips", "elf32-bigmips") -*/ -OUTPUT_FORMAT("elf32-tradlittlemips", "elf32-tradbigmips", "elf32-tradlittlemips") OUTPUT_ARCH(mips) ENTRY(_start) SECTIONS diff --git a/board/qemu-mips/u-boot.lds b/board/qemu-mips/u-boot.lds index cb2356f..33b42f3 100644 --- a/board/qemu-mips/u-boot.lds +++ b/board/qemu-mips/u-boot.lds @@ -21,14 +21,6 @@ * MA 02111-1307 USA */ -/* -OUTPUT_FORMAT("elf32-bigmips", "elf32-bigmips", "elf32-bigmips") -*/ -#if defined(CONFIG_64BIT) -OUTPUT_FORMAT("elf64-tradbigmips", "elf64-tradbigmips", "elf64-tradlittlemips") -#else -OUTPUT_FORMAT("elf32-tradbigmips", "elf32-tradbigmips", "elf32-tradlittlemips") -#endif OUTPUT_ARCH(mips) ENTRY(_start) SECTIONS diff --git a/board/qi/qi_lb60/u-boot.lds b/board/qi/qi_lb60/u-boot.lds index b3cb869..d074a27 100644 --- a/board/qi/qi_lb60/u-boot.lds +++ b/board/qi/qi_lb60/u-boot.lds @@ -18,8 +18,6 @@ * MA 02111-1307 USA */ -OUTPUT_FORMAT("elf32-tradlittlemips", "elf32-tradlittlemips", "elf32-tradlittlemips") - OUTPUT_ARCH(mips) ENTRY(_start) SECTIONS -- 1.7.10

2 1

[U-Boot] [PATCH] MIPS: xburst: simplify relocation offset calculation
by Gabor Juhos 31 Jan '13

31 Jan '13

The current code uses four instructions and a temporary register to calculate the relocation offset and to adjust the gp register. The relocation offset can be calculated directly from the CONFIG_SYS_MONITOR_BASE constant and from the destination address. The resulting offset can be used to adjust the gp pointer. This approach makes the code a bit simpler because it needs two instructions only. Signed-off-by: Gabor Juhos <juhosg(a)openwrt.org> Cc: Daniel Schwierzeck <daniel.schwierzeck(a)googlemail.com> Cc: Xiangfu Liu <xiangfu(a)openmobilefree.net> --- arch/mips/cpu/xburst/start.S | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-) diff --git a/arch/mips/cpu/xburst/start.S b/arch/mips/cpu/xburst/start.S index 194d745..35728eb 100644 --- a/arch/mips/cpu/xburst/start.S +++ b/arch/mips/cpu/xburst/start.S @@ -64,19 +64,13 @@ relocate_code: move sp, a0 # set new stack pointer li t0, CONFIG_SYS_MONITOR_BASE + sub t6, a2, t0 # t6 <-- relocation offset + la t3, in_ram lw t2, -12(t3) # t2 <-- uboot_end_data move t1, a2 - /* - * Fix $gp: - * - * New $gp = (Old $gp - CONFIG_SYS_MONITOR_BASE) + Destination Address - */ - move t6, gp - sub gp, CONFIG_SYS_MONITOR_BASE - add gp, a2 # gp now adjusted - sub t6, gp, t6 # t6 <-- relocation offset + add gp, t6 # adjust gp /* * t0 = source address -- 1.7.10

2 1