2 # (C) Copyright 2000 - 2002
3 # Wolfgang Denk, DENX Software Engineering, wd@denx.de.
5 # See file CREDITS for list of people who contributed to this
8 # This program is free software; you can redistribute it and/or
9 # modify it under the terms of the GNU General Public License as
10 # published by the Free Software Foundation; either version 2 of
11 # the License, or (at your option) any later version.
13 # This program is distributed in the hope that it will be useful,
14 # but WITHOUT ANY WARRANTY; without even the implied warranty of
15 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 # GNU General Public License for more details.
18 # You should have received a copy of the GNU General Public License
19 # along with this program; if not, write to the Free Software
20 # Foundation, Inc., 59 Temple Place, Suite 330, Boston,
27 This directory contains the source code for U-Boot, a boot loader for
28 Embedded boards based on PowerPC and ARM processors, which can be
29 installed in a boot ROM and used to initialize and test the hardware
30 or to download and run application code.
32 The development of U-Boot is closely related to Linux: some parts of
33 the source code originate in the Linux source tree, we have some
34 header files in common, and special provision has been made to
35 support booting of Linux images.
37 Some attention has been paid to make this software easily
38 configurable and extendable. For instance, all monitor commands are
39 implemented with the same call interface, so that it's very easy to
40 add new commands. Also, instead of permanently adding rarely used
41 code (for instance hardware test utilities) to the monitor, you can
42 load and run it dynamically.
48 In general, all boards for which a configuration option exists in the
49 Makefile have been tested to some extent and can be considered
50 "working". In fact, many of them are used in production systems.
52 In case of problems see the CHANGELOG and CREDITS files to find out
53 who contributed the specific port.
59 In case you have questions about, problems with or contributions for
60 U-Boot you should send a message to the U-Boot mailing list at
61 <u-boot-users@lists.sourceforge.net>. There is also an archive of
62 previous traffic on the mailing list - please search the archive
63 before asking FAQ's. Please see
64 http://lists.sourceforge.net/lists/listinfo/u-boot-users/
70 - start from 8xxrom sources
71 - create PPCBoot project (http://sourceforge.net/projects/ppcboot)
73 - make it easier to add custom boards
74 - make it possible to add other [PowerPC] CPUs
75 - extend functions, especially:
76 * Provide extended interface to Linux boot loader
79 * PCMCIA / CompactFLash / ATA disk / SCSI ... boot
80 - create ARMBoot project (http://sourceforge.net/projects/armboot)
81 - add other CPU families (starting with ARM)
82 - create U-Boot project (http://sourceforge.net/projects/u-boot)
88 The "official" name of this project is "Das U-Boot". The spelling
89 "U-Boot" shall be used in all written text (documentation, comments
90 in source files etc.). Example:
92 This is the README file for the U-Boot project.
94 File names etc. shall be based on the string "u-boot". Examples:
96 include/asm-ppc/u-boot.h
98 #include <asm/u-boot.h>
100 Variable names, preprocessor constants etc. shall be either based on
101 the string "u_boot" or on "U_BOOT". Example:
103 U_BOOT_VERSION u_boot_logo
104 IH_OS_U_BOOT u_boot_hush_start
110 U-Boot uses a 3 level version number containing a version, a
111 sub-version, and a patchlevel: "U-Boot-2.34.5" means version "2",
112 sub-version "34", and patchlevel "4".
114 The patchlevel is used to indicate certain stages of development
115 between released versions, i. e. officially released versions of
116 U-Boot will always have a patchlevel of "0".
122 - board Board dependend files
123 - common Misc architecture independend functions
124 - cpu CPU specific files
125 - disk Code for disk drive partition handling
126 - doc Documentation (don't expect too much)
127 - drivers Common used device drivers
128 - dtt Digital Thermometer and Thermostat drivers
129 - examples Example code for standalone applications, etc.
130 - include Header Files
131 - disk Harddisk interface code
132 - net Networking code
133 - ppc Files generic to PowerPC architecture
134 - post Power On Self Test
135 - post/arch Symlink to architecture specific Power On Self Test
136 - post/arch-ppc PowerPC architecture specific Power On Self Test
137 - post/cpu/mpc8260 MPC8260 CPU specific Power On Self Test
138 - post/cpu/mpc8xx MPC8xx CPU specific Power On Self Test
139 - rtc Real Time Clock drivers
140 - tools Tools to build S-Record or U-Boot images, etc.
142 - cpu/74xx_7xx Files specific to Motorola MPC74xx and 7xx CPUs
143 - cpu/arm925t Files specific to ARM 925 CPUs
144 - cpu/mpc5xx Files specific to Motorola MPC5xx CPUs
145 - cpu/mpc8xx Files specific to Motorola MPC8xx CPUs
146 - cpu/mpc824x Files specific to Motorola MPC824x CPUs
147 - cpu/mpc8260 Files specific to Motorola MPC8260 CPU
148 - cpu/ppc4xx Files specific to IBM 4xx CPUs
151 - board/LEOX/ Files specific to boards manufactured by The LEOX team
152 - board/LEOX/elpt860 Files specific to ELPT860 boards
154 Files specific to RPXClassic boards
155 - board/RPXlite Files specific to RPXlite boards
156 - board/at91rm9200dk Files specific to AT91RM9200DK boards
157 - board/c2mon Files specific to c2mon boards
158 - board/cmi Files specific to cmi boards
159 - board/cogent Files specific to Cogent boards
160 (need further configuration)
161 Files specific to CPCIISER4 boards
162 - board/cpu86 Files specific to CPU86 boards
163 - board/cray/ Files specific to boards manufactured by Cray
164 - board/cray/L1 Files specific to L1 boards
165 - board/cu824 Files specific to CU824 boards
166 - board/ebony Files specific to IBM Ebony board
167 - board/eric Files specific to ERIC boards
168 - board/esd/ Files specific to boards manufactured by ESD
169 - board/esd/adciop Files specific to ADCIOP boards
170 - board/esd/ar405 Files specific to AR405 boards
171 - board/esd/canbt Files specific to CANBT boards
172 - board/esd/cpci405 Files specific to CPCI405 boards
173 - board/esd/cpciiser4 Files specific to CPCIISER4 boards
174 - board/esd/common Common files for ESD boards
175 - board/esd/dasa_sim Files specific to DASA_SIM boards
176 - board/esd/du405 Files specific to DU405 boards
177 - board/esd/ocrtc Files specific to OCRTC boards
178 - board/esd/pci405 Files specific to PCI405 boards
180 Files specific to ESTEEM192E boards
181 - board/etx094 Files specific to ETX_094 boards
183 Files specific to EVB64260 boards
184 - board/fads Files specific to FADS boards
185 - board/flagadm Files specific to FLAGADM boards
186 - board/gen860t Files specific to GEN860T and GEN860T_SC boards
187 - board/genietv Files specific to GENIETV boards
188 - board/gth Files specific to GTH boards
189 - board/hermes Files specific to HERMES boards
190 - board/hymod Files specific to HYMOD boards
191 - board/icu862 Files specific to ICU862 boards
192 - board/ip860 Files specific to IP860 boards
194 Files specific to Interphase4539 boards
195 - board/ivm Files specific to IVMS8/IVML24 boards
196 - board/lantec Files specific to LANTEC boards
197 - board/lwmon Files specific to LWMON boards
198 - board/mbx8xx Files specific to MBX boards
200 Files specific to MPC8260ADS and PQ2FADS-ZU boards
201 - board/mpl/ Files specific to boards manufactured by MPL
202 - board/mpl/common Common files for MPL boards
203 - board/mpl/pip405 Files specific to PIP405 boards
204 - board/mpl/mip405 Files specific to MIP405 boards
205 - board/musenki Files specific to MUSEKNI boards
206 - board/mvs1 Files specific to MVS1 boards
207 - board/nx823 Files specific to NX823 boards
208 - board/oxc Files specific to OXC boards
210 Files specific to OMAP 1510 Innovator boards
211 - board/pcippc2 Files specific to PCIPPC2/PCIPPC6 boards
212 - board/pm826 Files specific to PM826 boards
214 Files specific to PPMC8260 boards
216 Files specific to RPXsuper boards
218 Files specific to RSDproto boards
220 Files specific to Sandpoint boards
221 - board/sbc8260 Files specific to SBC8260 boards
222 - board/sacsng Files specific to SACSng boards
223 - board/siemens Files specific to boards manufactured by Siemens AG
224 - board/siemens/CCM Files specific to CCM boards
225 - board/siemens/IAD210 Files specific to IAD210 boards
226 - board/siemens/SCM Files specific to SCM boards
227 - board/siemens/pcu_e Files specific to PCU_E boards
228 - board/sixnet Files specific to SIXNET boards
229 - board/spd8xx Files specific to SPD8xxTS boards
230 - board/tqm8260 Files specific to TQM8260 boards
231 - board/tqm8xx Files specific to TQM8xxL boards
232 - board/w7o Files specific to W7O boards
234 Files specific to Walnut405 boards
235 - board/westel/ Files specific to boards manufactured by Westel Wireless
236 - board/westel/amx860 Files specific to AMX860 boards
237 - board/utx8245 Files specific to UTX8245 boards
239 Software Configuration:
240 =======================
242 Configuration is usually done using C preprocessor defines; the
243 rationale behind that is to avoid dead code whenever possible.
245 There are two classes of configuration variables:
247 * Configuration _OPTIONS_:
248 These are selectable by the user and have names beginning with
251 * Configuration _SETTINGS_:
252 These depend on the hardware etc. and should not be meddled with if
253 you don't know what you're doing; they have names beginning with
256 Later we will add a configuration tool - probably similar to or even
257 identical to what's used for the Linux kernel. Right now, we have to
258 do the configuration by hand, which means creating some symbolic
259 links and editing some configuration files. We use the TQM8xxL boards
263 Selection of Processor Architecture and Board Type:
264 ---------------------------------------------------
266 For all supported boards there are ready-to-use default
267 configurations available; just type "make <board_name>_config".
269 Example: For a TQM823L module type:
274 For the Cogent platform, you need to specify the cpu type as well;
275 e.g. "make cogent_mpc8xx_config". And also configure the cogent
276 directory according to the instructions in cogent/README.
279 Configuration Options:
280 ----------------------
282 Configuration depends on the combination of board and CPU type; all
283 such information is kept in a configuration file
284 "include/configs/<board_name>.h".
286 Example: For a TQM823L module, all configuration settings are in
287 "include/configs/TQM823L.h".
290 Many of the options are named exactly as the corresponding Linux
291 kernel configuration options. The intention is to make it easier to
292 build a config tool - later.
295 The following options need to be configured:
297 - CPU Type: Define exactly one of
301 CONFIG_MPC823, CONFIG_MPC850, CONFIG_MPC855, CONFIG_MPC860
303 or CONFIG_MPC824X, CONFIG_MPC8260
317 - Board Type: Define exactly one of
319 PowerPC based boards:
320 ---------------------
322 CONFIG_ADCIOP, CONFIG_ICU862 CONFIG_RPXsuper,
323 CONFIG_ADS860, CONFIG_IP860, CONFIG_SM850,
324 CONFIG_AMX860, CONFIG_IPHASE4539, CONFIG_SPD823TS,
325 CONFIG_AR405, CONFIG_IVML24, CONFIG_SXNI855T,
326 CONFIG_BAB7xx, CONFIG_IVML24_128, CONFIG_Sandpoint8240,
327 CONFIG_CANBT, CONFIG_IVML24_256, CONFIG_Sandpoint8245,
328 CONFIG_CCM, CONFIG_IVMS8, CONFIG_TQM823L,
329 CONFIG_CPCI405, CONFIG_IVMS8_128, CONFIG_TQM850L,
330 CONFIG_CPCI4052, CONFIG_IVMS8_256, CONFIG_TQM855L,
331 CONFIG_CPCIISER4, CONFIG_LANTEC, CONFIG_TQM860L,
332 CONFIG_CPU86, CONFIG_MBX, CONFIG_TQM8260,
333 CONFIG_CRAYL1, CONFIG_MBX860T, CONFIG_TTTech,
334 CONFIG_CU824, CONFIG_MHPC, CONFIG_UTX8245,
335 CONFIG_DASA_SIM, CONFIG_MIP405, CONFIG_W7OLMC,
336 CONFIG_DU405, CONFIG_MOUSSE, CONFIG_W7OLMG,
337 CONFIG_ELPPC, CONFIG_MPC8260ADS, CONFIG_WALNUT405,
338 CONFIG_ERIC, CONFIG_MUSENKI, CONFIG_ZUMA,
339 CONFIG_ESTEEM192E, CONFIG_MVS1, CONFIG_c2mon,
340 CONFIG_ETX094, CONFIG_NX823, CONFIG_cogent_mpc8260,
341 CONFIG_EVB64260, CONFIG_OCRTC, CONFIG_cogent_mpc8xx,
342 CONFIG_FADS823, CONFIG_ORSG, CONFIG_ep8260,
343 CONFIG_FADS850SAR, CONFIG_OXC, CONFIG_gw8260,
344 CONFIG_FADS860T, CONFIG_PCI405, CONFIG_hermes,
345 CONFIG_FLAGADM, CONFIG_PCIPPC2, CONFIG_hymod,
346 CONFIG_FPS850L, CONFIG_PCIPPC6, CONFIG_lwmon,
347 CONFIG_GEN860T, CONFIG_PIP405, CONFIG_pcu_e,
348 CONFIG_GENIETV, CONFIG_PM826, CONFIG_ppmc8260,
349 CONFIG_GTH, CONFIG_RPXClassic, CONFIG_rsdproto,
350 CONFIG_IAD210, CONFIG_RPXlite, CONFIG_sbc8260,
351 CONFIG_EBONY, CONFIG_sacsng, CONFIG_FPS860L,
352 CONFIG_V37, CONFIG_ELPT860, CONFIG_CMI,
353 CONFIG_NETVIA, CONFIG_RBC823
358 CONFIG_HHP_CRADLE, CONFIG_DNP1110, CONFIG_EP7312,
359 CONFIG_IMPA7, CONFIG_LART, CONFIG_LUBBOCK,
360 CONFIG_INNOVATOROMAP1510,
361 CONFIG_SHANNON, CONFIG_SMDK2400, CONFIG_SMDK2410,
362 CONFIG_TRAB, CONFIG_AT91RM9200DK
365 - CPU Module Type: (if CONFIG_COGENT is defined)
366 Define exactly one of
368 --- FIXME --- not tested yet:
369 CONFIG_CMA286_60, CONFIG_CMA286_21, CONFIG_CMA286_60P,
370 CONFIG_CMA287_23, CONFIG_CMA287_50
372 - Motherboard Type: (if CONFIG_COGENT is defined)
373 Define exactly one of
374 CONFIG_CMA101, CONFIG_CMA102
376 - Motherboard I/O Modules: (if CONFIG_COGENT is defined)
377 Define one or more of
380 - Motherboard Options: (if CONFIG_CMA101 or CONFIG_CMA102 are defined)
381 Define one or more of
382 CONFIG_LCD_HEARTBEAT - update a character position on
383 the lcd display every second with
386 - Board flavour: (if CONFIG_MPC8260ADS is defined)
389 CFG_8260ADS - original MPC8260ADS
390 CFG_8266ADS - MPC8266ADS (untested)
391 CFG_PQ2FADS - PQ2FADS-ZU
394 - MPC824X Family Member (if CONFIG_MPC824X is defined)
395 Define exactly one of
396 CONFIG_MPC8240, CONFIG_MPC8245
398 - 8xx CPU Options: (if using an 8xx cpu)
399 Define one or more of
400 CONFIG_8xx_GCLK_FREQ - if get_gclk_freq() can not work e.g.
401 no 32KHz reference PIT/RTC clock
406 U-Boot stores all clock information in Hz
407 internally. For binary compatibility with older Linux
408 kernels (which expect the clocks passed in the
409 bd_info data to be in MHz) the environment variable
410 "clocks_in_mhz" can be defined so that U-Boot
411 converts clock data to MHZ before passing it to the
414 When CONFIG_CLOCKS_IN_MHZ is defined, a definition of
415 "clocks_in_mhz=1" is automatically included in the
419 Depending on board, define exactly one serial port
420 (like CONFIG_8xx_CONS_SMC1, CONFIG_8xx_CONS_SMC2,
421 CONFIG_8xx_CONS_SCC1, ...), or switch off the serial
422 console by defining CONFIG_8xx_CONS_NONE
424 Note: if CONFIG_8xx_CONS_NONE is defined, the serial
425 port routines must be defined elsewhere
426 (i.e. serial_init(), serial_getc(), ...)
429 Enables console device for a color framebuffer. Needs following
430 defines (cf. smiLynxEM, i8042, board/eltec/bab7xx)
431 VIDEO_FB_LITTLE_ENDIAN graphic memory organisation
433 VIDEO_HW_RECTFILL graphic chip supports
436 VIDEO_HW_BITBLT graphic chip supports
437 bit-blit (cf. smiLynxEM)
438 VIDEO_VISIBLE_COLS visible pixel columns
440 VIDEO_VISIBLE_ROWS visible pixel rows
441 VIDEO_PIXEL_SIZE bytes per pixel
442 VIDEO_DATA_FORMAT graphic data format
443 (0-5, cf. cfb_console.c)
444 VIDEO_FB_ADRS framebuffer address
445 VIDEO_KBD_INIT_FCT keyboard int fct
446 (i.e. i8042_kbd_init())
447 VIDEO_TSTC_FCT test char fct
449 VIDEO_GETC_FCT get char fct
451 CONFIG_CONSOLE_CURSOR cursor drawing on/off
452 (requires blink timer
454 CFG_CONSOLE_BLINK_COUNT blink interval (cf. i8042.c)
455 CONFIG_CONSOLE_TIME display time/date info in
457 (requires CFG_CMD_DATE)
458 CONFIG_VIDEO_LOGO display Linux logo in
460 CONFIG_VIDEO_BMP_LOGO use bmp_logo.h instead of
461 linux_logo.h for logo.
462 Requires CONFIG_VIDEO_LOGO
463 CONFIG_CONSOLE_EXTRA_INFO
464 addional board info beside
467 When CONFIG_CFB_CONSOLE is defined, video console is
468 default i/o. Serial console can be forced with
469 environment 'console=serial'.
472 CONFIG_BAUDRATE - in bps
473 Select one of the baudrates listed in
474 CFG_BAUDRATE_TABLE, see below.
476 - Interrupt driven serial port input:
477 CONFIG_SERIAL_SOFTWARE_FIFO
480 Use an interrupt handler for receiving data on the
481 serial port. It also enables using hardware handshake
482 (RTS/CTS) and UART's built-in FIFO. Set the number of
483 bytes the interrupt driven input buffer should have.
485 Set to 0 to disable this feature (this is the default).
486 This will also disable hardware handshake.
488 - Console UART Number:
492 If defined internal UART1 (and not UART0) is used
493 as default U-Boot console.
495 - Boot Delay: CONFIG_BOOTDELAY - in seconds
496 Delay before automatically booting the default image;
497 set to -1 to disable autoboot.
499 See doc/README.autoboot for these options that
500 work with CONFIG_BOOTDELAY. None are required.
501 CONFIG_BOOT_RETRY_TIME
502 CONFIG_BOOT_RETRY_MIN
503 CONFIG_AUTOBOOT_KEYED
504 CONFIG_AUTOBOOT_PROMPT
505 CONFIG_AUTOBOOT_DELAY_STR
506 CONFIG_AUTOBOOT_STOP_STR
507 CONFIG_AUTOBOOT_DELAY_STR2
508 CONFIG_AUTOBOOT_STOP_STR2
509 CONFIG_ZERO_BOOTDELAY_CHECK
510 CONFIG_RESET_TO_RETRY
514 Only needed when CONFIG_BOOTDELAY is enabled;
515 define a command string that is automatically executed
516 when no character is read on the console interface
517 within "Boot Delay" after reset.
520 This can be used to pass arguments to the bootm
521 command. The value of CONFIG_BOOTARGS goes into the
522 environment value "bootargs".
524 CONFIG_RAMBOOT and CONFIG_NFSBOOT
525 The value of these goes into the environment as
526 "ramboot" and "nfsboot" respectively, and can be used
527 as a convenience, when switching between booting from
533 When this option is #defined, the existence of the
534 environment variable "preboot" will be checked
535 immediately before starting the CONFIG_BOOTDELAY
536 countdown and/or running the auto-boot command resp.
537 entering interactive mode.
539 This feature is especially useful when "preboot" is
540 automatically generated or modified. For an example
541 see the LWMON board specific code: here "preboot" is
542 modified when the user holds down a certain
543 combination of keys on the (special) keyboard when
546 - Serial Download Echo Mode:
548 If defined to 1, all characters received during a
549 serial download (using the "loads" command) are
550 echoed back. This might be needed by some terminal
551 emulations (like "cu"), but may as well just take
552 time on others. This setting #define's the initial
553 value of the "loads_echo" environment variable.
555 - Kgdb Serial Baudrate: (if CFG_CMD_KGDB is defined)
557 Select one of the baudrates listed in
558 CFG_BAUDRATE_TABLE, see below.
562 Most monitor functions can be selected (or
563 de-selected) by adjusting the definition of
564 CONFIG_COMMANDS; to select individual functions,
565 #define CONFIG_COMMANDS by "OR"ing any of the
568 #define enables commands:
569 -------------------------
570 CFG_CMD_ASKENV * ask for env variable
572 CFG_CMD_BEDBUG Include BedBug Debugger
574 CFG_CMD_CACHE icache, dcache
575 CFG_CMD_CONSOLE coninfo
576 CFG_CMD_DATE * support for RTC, date/time...
577 CFG_CMD_DHCP DHCP support
578 CFG_CMD_ECHO * echo arguments
579 CFG_CMD_EEPROM * EEPROM read/write support
580 CFG_CMD_ELF bootelf, bootvx
582 CFG_CMD_FDC * Floppy Disk Support
583 CFG_CMD_FAT FAT partition support
584 CFG_CMD_FDOS * Dos diskette Support
585 CFG_CMD_FLASH flinfo, erase, protect
586 CFG_CMD_FPGA FPGA device initialization support
587 CFG_CMD_I2C * I2C serial bus support
588 CFG_CMD_IDE * IDE harddisk support
590 CFG_CMD_IMMAP * IMMR dump support
591 CFG_CMD_IRQ * irqinfo
595 CFG_CMD_MEMORY md, mm, nm, mw, cp, cmp, crc, base,
597 CFG_CMD_MMC MMC memory mapped support
598 CFG_CMD_MII MII utility commands
599 CFG_CMD_NET bootp, tftpboot, rarpboot
600 CFG_CMD_PCI * pciinfo
601 CFG_CMD_PCMCIA * PCMCIA support
602 CFG_CMD_REGINFO * Register dump
603 CFG_CMD_RUN run command in env variable
604 CFG_CMD_SCSI * SCSI Support
605 CFG_CMD_SETGETDCR Support for DCR Register access (4xx only)
606 CFG_CMD_SPI * SPI serial bus support
607 CFG_CMD_USB * USB support
608 CFG_CMD_BSP * Board SPecific functions
609 -----------------------------------------------
612 CFG_CMD_DFL Default configuration; at the moment
613 this is includes all commands, except
614 the ones marked with "*" in the list
617 If you don't define CONFIG_COMMANDS it defaults to
618 CFG_CMD_DFL in include/cmd_confdefs.h. A board can
619 override the default settings in the respective
622 EXAMPLE: If you want all functions except of network
623 support you can write:
625 #define CONFIG_COMMANDS (CFG_CMD_ALL & ~CFG_CMD_NET)
628 Note: Don't enable the "icache" and "dcache" commands
629 (configuration option CFG_CMD_CACHE) unless you know
630 what you (and your U-Boot users) are doing. Data
631 cache cannot be enabled on systems like the 8xx or
632 8260 (where accesses to the IMMR region must be
633 uncached), and it cannot be disabled on all other
634 systems where we (mis-) use the data cache to hold an
635 initial stack and some data.
638 XXX - this list needs to get updated!
642 If this variable is defined, it enables watchdog
643 support. There must support in the platform specific
644 code for a watchdog. For the 8xx and 8260 CPUs, the
645 SIU Watchdog feature is enabled in the SYPCR
649 CONFIG_VERSION_VARIABLE
650 If this variable is defined, an environment variable
651 named "ver" is created by U-Boot showing the U-Boot
652 version as printed by the "version" command.
653 This variable is readonly.
657 When CFG_CMD_DATE is selected, the type of the RTC
658 has to be selected, too. Define exactly one of the
661 CONFIG_RTC_MPC8xx - use internal RTC of MPC8xx
662 CONFIG_RTC_PCF8563 - use Philips PCF8563 RTC
663 CONFIG_RTC_MC146818 - use MC146818 RTC
664 CONFIG_RTC_DS1307 - use Maxim, Inc. DS1307 RTC
665 CONFIG_RTC_DS1337 - use Maxim, Inc. DS1337 RTC
666 CONFIG_RTC_DS1338 - use Maxim, Inc. DS1338 RTC
667 CONFIG_RTC_DS164x - use Dallas DS164x RTC
669 Note that if the RTC uses I2C, then the I2C interface
670 must also be configured. See I2C Support, below.
674 When CONFIG_TIMESTAMP is selected, the timestamp
675 (date and time) of an image is printed by image
676 commands like bootm or iminfo. This option is
677 automatically enabled when you select CFG_CMD_DATE .
680 CONFIG_MAC_PARTITION and/or CONFIG_DOS_PARTITION
681 and/or CONFIG_ISO_PARTITION
683 If IDE or SCSI support is enabled (CFG_CMD_IDE or
684 CFG_CMD_SCSI) you must configure support for at least
685 one partition type as well.
688 CONFIG_IDE_RESET_ROUTINE
690 Set this to define that instead of a reset Pin, the
691 routine ide_set_reset(int idereset) will be used.
696 Set this to enable ATAPI support.
699 At the moment only there is only support for the
700 SYM53C8XX SCSI controller; define
701 CONFIG_SCSI_SYM53C8XX to enable it.
703 CFG_SCSI_MAX_LUN [8], CFG_SCSI_MAX_SCSI_ID [7] and
704 CFG_SCSI_MAX_DEVICE [CFG_SCSI_MAX_SCSI_ID *
705 CFG_SCSI_MAX_LUN] can be adjusted to define the
706 maximum numbers of LUNs, SCSI ID's and target
708 CFG_SCSI_SYM53C8XX_CCF to fix clock timing (80Mhz)
710 - NETWORK Support (PCI):
712 Support for Intel 8254x gigabit chips.
715 Support for Intel 82557/82559/82559ER chips.
716 Optional CONFIG_EEPRO100_SROM_WRITE enables eeprom
717 write routine for first time initialisation.
720 Support for Digital 2114x chips.
721 Optional CONFIG_TULIP_SELECT_MEDIA for board specific
722 modem chip initialisation (KS8761/QS6611).
725 Support for National dp83815 chips.
728 Support for National dp8382[01] gigabit chips.
730 - NETWORK Support (other):
732 CONFIG_DRIVER_LAN91C96
733 Support for SMSC's LAN91C96 chips.
736 Define this to hold the physical address
737 of the LAN91C96's I/O space
739 CONFIG_LAN91C96_USE_32_BIT
740 Define this to enable 32 bit addressing
743 At the moment only the UHCI host controller is
744 supported (PIP405, MIP405); define
745 CONFIG_USB_UHCI to enable it.
746 define CONFIG_USB_KEYBOARD to enable the USB Keyboard
747 end define CONFIG_USB_STORAGE to enable the USB
750 Supported are USB Keyboards and USB Floppy drives
754 The MMC controller on the Intel PXA is supported. To
755 enable this define CONFIG_MMC. The MMC can be
756 accessed from the boot prompt by mapping the device
757 to physical memory similar to flash. Command line is
758 enabled with CFG_CMD_MMC. The MMC driver also works with
759 the FAT fs. This is enabled with CFG_CMD_FAT.
764 Define this to enable standard (PC-Style) keyboard
768 Standard PC keyboard driver with US (is default) and
769 GERMAN key layout (switch via environment 'keymap=de') support.
770 Export function i8042_kbd_init, i8042_tstc and i8042_getc
771 for cfb_console. Supports cursor blinking.
776 Define this to enable video support (for output to
781 Enable Chips & Technologies 69000 Video chip
783 CONFIG_VIDEO_SMI_LYNXEM
784 Enable Silicon Motion SMI 712/710/810 Video chip
785 Videomode are selected via environment 'videomode' with
786 standard LiLo mode numbers.
787 Following modes are supported (* is default):
789 800x600 1024x768 1280x1024
790 256 (8bit) 303* 305 307
791 65536 (16bit) 314 317 31a
792 16,7 Mill (24bit) 315 318 31b
793 (i.e. setenv videomode 317; saveenv; reset;)
795 CONFIG_VIDEO_SED13806
796 Enable Epson SED13806 driver. This driver supports 8bpp
797 and 16bpp modes defined by CONFIG_VIDEO_SED13806_8BPP
798 or CONFIG_VIDEO_SED13806_16BPP
803 Define this to enable a custom keyboard support.
804 This simply calls drv_keyboard_init() which must be
805 defined in your board-specific files.
806 The only board using this so far is RBC823.
808 - LCD Support: CONFIG_LCD
810 Define this to enable LCD support (for output to LCD
811 display); also select one of the supported displays
812 by defining one of these:
814 CONFIG_NEC_NL6648AC33:
816 NEC NL6648AC33-18. Active, color, single scan.
818 CONFIG_NEC_NL6648BC20
820 NEC NL6648BC20-08. 6.5", 640x480.
821 Active, color, single scan.
825 Sharp 320x240. Active, color, single scan.
826 It isn't 16x9, and I am not sure what it is.
828 CONFIG_SHARP_LQ64D341
830 Sharp LQ64D341 display, 640x480.
831 Active, color, single scan.
835 HLD1045 display, 640x480.
836 Active, color, single scan.
840 Optrex CBL50840-2 NF-FW 99 22 M5
842 Hitachi LMG6912RPFC-00T
846 320x240. Black & white.
848 Normally display is black on white background; define
849 CFG_WHITE_ON_BLACK to get it inverted.
851 - Spash Screen Support: CONFIG_SPLASH_SCREEN
853 If this option is set, the environment is checked for
854 a variable "splashimage". If found, the usual display
855 of logo, copyright and system information on the LCD
856 is supressed and the BMP image at the address
857 specified in "splashimage" is loaded instead. The
858 console is redirected to the "nulldev", too. This
859 allows for a "silent" boot where a splash screen is
860 loaded very quickly after power-on.
862 - Compression support:
865 If this option is set, support for bzip2 compressed
866 images is included. If not, only uncompressed and gzip
867 compressed images are supported.
869 NOTE: the bzip2 algorithm requires a lot of RAM, so
870 the malloc area (as defined by CFG_MALLOC_LEN) should
878 Define a default value for ethernet address to use
879 for the respective ethernet interface, in case this
880 is not determined automatically.
885 Define a default value for the IP address to use for
886 the default ethernet interface, in case this is not
887 determined through e.g. bootp.
892 Defines a default value for theIP address of a TFTP
893 server to contact when using the "tftboot" command.
895 - BOOTP Recovery Mode:
896 CONFIG_BOOTP_RANDOM_DELAY
898 If you have many targets in a network that try to
899 boot using BOOTP, you may want to avoid that all
900 systems send out BOOTP requests at precisely the same
901 moment (which would happen for instance at recovery
902 from a power failure, when all systems will try to
903 boot, thus flooding the BOOTP server. Defining
904 CONFIG_BOOTP_RANDOM_DELAY causes a random delay to be
905 inserted before sending out BOOTP requests. The
906 following delays are insterted then:
908 1st BOOTP request: delay 0 ... 1 sec
909 2nd BOOTP request: delay 0 ... 2 sec
910 3rd BOOTP request: delay 0 ... 4 sec
912 BOOTP requests: delay 0 ... 8 sec
914 - DHCP Advanced Options:
917 You can fine tune the DHCP functionality by adding
918 these flags to the CONFIG_BOOTP_MASK define:
920 CONFIG_BOOTP_DNS2 - If a DHCP client requests the DNS
921 serverip from a DHCP server, it is possible that more
922 than one DNS serverip is offered to the client.
923 If CONFIG_BOOTP_DNS2 is enabled, the secondary DNS
924 serverip will be stored in the additional environment
925 variable "dnsip2". The first DNS serverip is always
926 stored in the variable "dnsip", when CONFIG_BOOTP_DNS
927 is added to the CONFIG_BOOTP_MASK.
929 CONFIG_BOOTP_SEND_HOSTNAME - Some DHCP servers are capable
930 to do a dynamic update of a DNS server. To do this, they
931 need the hostname of the DHCP requester.
932 If CONFIG_BOOP_SEND_HOSTNAME is added to the
933 CONFIG_BOOTP_MASK, the content of the "hostname"
934 environment variable is passed as option 12 to
937 - Status LED: CONFIG_STATUS_LED
939 Several configurations allow to display the current
940 status using a LED. For instance, the LED will blink
941 fast while running U-Boot code, stop blinking as
942 soon as a reply to a BOOTP request was received, and
943 start blinking slow once the Linux kernel is running
944 (supported by a status LED driver in the Linux
945 kernel). Defining CONFIG_STATUS_LED enables this
948 - CAN Support: CONFIG_CAN_DRIVER
950 Defining CONFIG_CAN_DRIVER enables CAN driver support
951 on those systems that support this (optional)
952 feature, like the TQM8xxL modules.
954 - I2C Support: CONFIG_HARD_I2C | CONFIG_SOFT_I2C
956 These enable I2C serial bus commands. Defining either of
957 (but not both of) CONFIG_HARD_I2C or CONFIG_SOFT_I2C will
958 include the appropriate I2C driver for the selected cpu.
960 This will allow you to use i2c commands at the u-boot
961 command line (as long as you set CFG_CMD_I2C in
962 CONFIG_COMMANDS) and communicate with i2c based realtime
963 clock chips. See common/cmd_i2c.c for a description of the
964 command line interface.
966 CONFIG_HARD_I2C selects the CPM hardware driver for I2C.
968 CONFIG_SOFT_I2C configures u-boot to use a software (aka
969 bit-banging) driver instead of CPM or similar hardware
972 There are several other quantities that must also be
973 defined when you define CONFIG_HARD_I2C or CONFIG_SOFT_I2C.
975 In both cases you will need to define CFG_I2C_SPEED
976 to be the frequency (in Hz) at which you wish your i2c bus
977 to run and CFG_I2C_SLAVE to be the address of this node (ie
978 the cpu's i2c node address).
980 Now, the u-boot i2c code for the mpc8xx (cpu/mpc8xx/i2c.c)
981 sets the cpu up as a master node and so its address should
982 therefore be cleared to 0 (See, eg, MPC823e User's Manual
983 p.16-473). So, set CFG_I2C_SLAVE to 0.
985 That's all that's required for CONFIG_HARD_I2C.
987 If you use the software i2c interface (CONFIG_SOFT_I2C)
988 then the following macros need to be defined (examples are
989 from include/configs/lwmon.h):
993 (Optional). Any commands necessary to enable the I2C
994 controller or configure ports.
996 eg: #define I2C_INIT (immr->im_cpm.cp_pbdir |= PB_SCL)
1000 (Only for MPC8260 CPU). The I/O port to use (the code
1001 assumes both bits are on the same port). Valid values
1002 are 0..3 for ports A..D.
1006 The code necessary to make the I2C data line active
1007 (driven). If the data line is open collector, this
1010 eg: #define I2C_ACTIVE (immr->im_cpm.cp_pbdir |= PB_SDA)
1014 The code necessary to make the I2C data line tri-stated
1015 (inactive). If the data line is open collector, this
1018 eg: #define I2C_TRISTATE (immr->im_cpm.cp_pbdir &= ~PB_SDA)
1022 Code that returns TRUE if the I2C data line is high,
1025 eg: #define I2C_READ ((immr->im_cpm.cp_pbdat & PB_SDA) != 0)
1029 If <bit> is TRUE, sets the I2C data line high. If it
1030 is FALSE, it clears it (low).
1032 eg: #define I2C_SDA(bit) \
1033 if(bit) immr->im_cpm.cp_pbdat |= PB_SDA; \
1034 else immr->im_cpm.cp_pbdat &= ~PB_SDA
1038 If <bit> is TRUE, sets the I2C clock line high. If it
1039 is FALSE, it clears it (low).
1041 eg: #define I2C_SCL(bit) \
1042 if(bit) immr->im_cpm.cp_pbdat |= PB_SCL; \
1043 else immr->im_cpm.cp_pbdat &= ~PB_SCL
1047 This delay is invoked four times per clock cycle so this
1048 controls the rate of data transfer. The data rate thus
1049 is 1 / (I2C_DELAY * 4). Often defined to be something
1052 #define I2C_DELAY udelay(2)
1056 When a board is reset during an i2c bus transfer
1057 chips might think that the current transfer is still
1058 in progress. On some boards it is possible to access
1059 the i2c SCLK line directly, either by using the
1060 processor pin as a GPIO or by having a second pin
1061 connected to the bus. If this option is defined a
1062 custom i2c_init_board() routine in boards/xxx/board.c
1063 is run early in the boot sequence.
1065 - SPI Support: CONFIG_SPI
1067 Enables SPI driver (so far only tested with
1068 SPI EEPROM, also an instance works with Crystal A/D and
1069 D/As on the SACSng board)
1073 Enables extended (16-bit) SPI EEPROM addressing.
1074 (symmetrical to CONFIG_I2C_X)
1078 Enables a software (bit-bang) SPI driver rather than
1079 using hardware support. This is a general purpose
1080 driver that only requires three general I/O port pins
1081 (two outputs, one input) to function. If this is
1082 defined, the board configuration must define several
1083 SPI configuration items (port pins to use, etc). For
1084 an example, see include/configs/sacsng.h.
1086 - FPGA Support: CONFIG_FPGA_COUNT
1088 Specify the number of FPGA devices to support.
1092 Used to specify the types of FPGA devices. For
1094 #define CONFIG_FPGA CFG_XILINX_VIRTEX2
1096 CFG_FPGA_PROG_FEEDBACK
1098 Enable printing of hash marks during FPGA
1103 Enable checks on FPGA configuration interface busy
1104 status by the configuration function. This option
1105 will require a board or device specific function to
1110 If defined, a function that provides delays in the
1111 FPGA configuration driver.
1113 CFG_FPGA_CHECK_CTRLC
1115 Allow Control-C to interrupt FPGA configuration
1117 CFG_FPGA_CHECK_ERROR
1119 Check for configuration errors during FPGA bitfile
1120 loading. For example, abort during Virtex II
1121 configuration if the INIT_B line goes low (which
1122 indicated a CRC error).
1126 Maximum time to wait for the INIT_B line to deassert
1127 after PROB_B has been deasserted during a Virtex II
1128 FPGA configuration sequence. The default time is 500 mS.
1132 Maximum time to wait for BUSY to deassert during
1133 Virtex II FPGA configuration. The default is 5 mS.
1135 CFG_FPGA_WAIT_CONFIG
1137 Time to wait after FPGA configuration. The default is
1140 - FPGA Support: CONFIG_FPGA_COUNT
1142 Specify the number of FPGA devices to support.
1146 Used to specify the types of FPGA devices. For example,
1147 #define CONFIG_FPGA CFG_XILINX_VIRTEX2
1149 CFG_FPGA_PROG_FEEDBACK
1151 Enable printing of hash marks during FPGA configuration.
1155 Enable checks on FPGA configuration interface busy
1156 status by the configuration function. This option
1157 will require a board or device specific function to
1162 If defined, a function that provides delays in the FPGA
1163 configuration driver.
1165 CFG_FPGA_CHECK_CTRLC
1166 Allow Control-C to interrupt FPGA configuration
1168 CFG_FPGA_CHECK_ERROR
1170 Check for configuration errors during FPGA bitfile
1171 loading. For example, abort during Virtex II
1172 configuration if the INIT_B line goes low (which
1173 indicated a CRC error).
1177 Maximum time to wait for the INIT_B line to deassert
1178 after PROB_B has been deasserted during a Virtex II
1179 FPGA configuration sequence. The default time is 500
1184 Maximum time to wait for BUSY to deassert during
1185 Virtex II FPGA configuration. The default is 5 mS.
1187 CFG_FPGA_WAIT_CONFIG
1189 Time to wait after FPGA configuration. The default is
1192 - Configuration Management:
1195 If defined, this string will be added to the U-Boot
1196 version information (U_BOOT_VERSION)
1198 - Vendor Parameter Protection:
1200 U-Boot considers the values of the environment
1201 variables "serial#" (Board Serial Number) and
1202 "ethaddr" (Ethernet Address) to bb parameters that
1203 are set once by the board vendor / manufacturer, and
1204 protects these variables from casual modification by
1205 the user. Once set, these variables are read-only,
1206 and write or delete attempts are rejected. You can
1207 change this behviour:
1209 If CONFIG_ENV_OVERWRITE is #defined in your config
1210 file, the write protection for vendor parameters is
1211 completely disabled. Anybody can change or delete
1214 Alternatively, if you #define _both_ CONFIG_ETHADDR
1215 _and_ CONFIG_OVERWRITE_ETHADDR_ONCE, a default
1216 ethernet address is installed in the environment,
1217 which can be changed exactly ONCE by the user. [The
1218 serial# is unaffected by this, i. e. it remains
1224 Define this variable to enable the reservation of
1225 "protected RAM", i. e. RAM which is not overwritten
1226 by U-Boot. Define CONFIG_PRAM to hold the number of
1227 kB you want to reserve for pRAM. You can overwrite
1228 this default value by defining an environment
1229 variable "pram" to the number of kB you want to
1230 reserve. Note that the board info structure will
1231 still show the full amount of RAM. If pRAM is
1232 reserved, a new environment variable "mem" will
1233 automatically be defined to hold the amount of
1234 remaining RAM in a form that can be passed as boot
1235 argument to Linux, for instance like that:
1237 setenv bootargs ... mem=\$(mem)
1240 This way you can tell Linux not to use this memory,
1241 either, which results in a memory region that will
1242 not be affected by reboots.
1244 *WARNING* If your board configuration uses automatic
1245 detection of the RAM size, you must make sure that
1246 this memory test is non-destructive. So far, the
1247 following board configurations are known to be
1250 ETX094, IVMS8, IVML24, SPD8xx, TQM8xxL,
1251 HERMES, IP860, RPXlite, LWMON, LANTEC,
1252 PCU_E, FLAGADM, TQM8260
1257 Define this variable to stop the system in case of a
1258 fatal error, so that you have to reset it manually.
1259 This is probably NOT a good idea for an embedded
1260 system where you want to system to reboot
1261 automatically as fast as possible, but it may be
1262 useful during development since you can try to debug
1263 the conditions that lead to the situation.
1265 CONFIG_NET_RETRY_COUNT
1267 This variable defines the number of retries for
1268 network operations like ARP, RARP, TFTP, or BOOTP
1269 before giving up the operation. If not defined, a
1270 default value of 5 is used.
1272 - Command Interpreter:
1275 Define this variable to enable the "hush" shell (from
1276 Busybox) as command line interpreter, thus enabling
1277 powerful command line syntax like
1278 if...then...else...fi conditionals or `&&' and '||'
1279 constructs ("shell scripts").
1281 If undefined, you get the old, much simpler behaviour
1282 with a somewhat smaller memory footprint.
1287 This defines the secondary prompt string, which is
1288 printed when the command interpreter needs more input
1289 to complete a command. Usually "> ".
1293 In the current implementation, the local variables
1294 space and global environment variables space are
1295 separated. Local variables are those you define by
1296 simply typing `name=value'. To access a local
1297 variable later on, you have write `$name' or
1298 `${name}'; to execute the contents of a variable
1299 directly type `$name' at the command prompt.
1301 Global environment variables are those you use
1302 setenv/printenv to work with. To run a command stored
1303 in such a variable, you need to use the run command,
1304 and you must not use the '$' sign to access them.
1306 To store commands and special characters in a
1307 variable, please use double quotation marks
1308 surrounding the whole text of the variable, instead
1309 of the backslashes before semicolons and special
1312 - Default Environment
1313 CONFIG_EXTRA_ENV_SETTINGS
1315 Define this to contain any number of null terminated
1316 strings (variable = value pairs) that will be part of
1317 the default enviroment compiled into the boot image.
1319 For example, place something like this in your
1320 board's config file:
1322 #define CONFIG_EXTRA_ENV_SETTINGS \
1326 Warning: This method is based on knowledge about the
1327 internal format how the environment is stored by the
1328 U-Boot code. This is NOT an official, exported
1329 interface! Although it is unlikely that this format
1330 will change soon, but there is no guarantee either.
1331 You better know what you are doing here.
1333 Note: overly (ab)use of the default environment is
1334 discouraged. Make sure to check other ways to preset
1335 the environment like the autoscript function or the
1339 CONFIG_HAS_DATAFLASH
1341 Defining this option enables DataFlash features and
1342 allows to read/write in Dataflash via the standard
1345 - Show boot progress
1346 CONFIG_SHOW_BOOT_PROGRESS
1348 Defining this option allows to add some board-
1349 specific code (calling a user-provided function
1350 "show_boot_progress(int)") that enables you to show
1351 the system's boot progress on some display (for
1352 example, some LED's) on your board. At the moment,
1353 the following checkpoints are implemented:
1356 1 common/cmd_bootm.c before attempting to boot an image
1357 -1 common/cmd_bootm.c Image header has bad magic number
1358 2 common/cmd_bootm.c Image header has correct magic number
1359 -2 common/cmd_bootm.c Image header has bad checksum
1360 3 common/cmd_bootm.c Image header has correct checksum
1361 -3 common/cmd_bootm.c Image data has bad checksum
1362 4 common/cmd_bootm.c Image data has correct checksum
1363 -4 common/cmd_bootm.c Image is for unsupported architecture
1364 5 common/cmd_bootm.c Architecture check OK
1365 -5 common/cmd_bootm.c Wrong Image Type (not kernel, multi, standalone)
1366 6 common/cmd_bootm.c Image Type check OK
1367 -6 common/cmd_bootm.c gunzip uncompression error
1368 -7 common/cmd_bootm.c Unimplemented compression type
1369 7 common/cmd_bootm.c Uncompression OK
1370 -8 common/cmd_bootm.c Wrong Image Type (not kernel, multi, standalone)
1371 8 common/cmd_bootm.c Image Type check OK
1372 -9 common/cmd_bootm.c Unsupported OS (not Linux, BSD, VxWorks, QNX)
1373 9 common/cmd_bootm.c Start initial ramdisk verification
1374 -10 common/cmd_bootm.c Ramdisk header has bad magic number
1375 -11 common/cmd_bootm.c Ramdisk header has bad checksum
1376 10 common/cmd_bootm.c Ramdisk header is OK
1377 -12 common/cmd_bootm.c Ramdisk data has bad checksum
1378 11 common/cmd_bootm.c Ramdisk data has correct checksum
1379 12 common/cmd_bootm.c Ramdisk verification complete, start loading
1380 -13 common/cmd_bootm.c Wrong Image Type (not PPC Linux Ramdisk)
1381 13 common/cmd_bootm.c Start multifile image verification
1382 14 common/cmd_bootm.c No initial ramdisk, no multifile, continue.
1383 15 common/cmd_bootm.c All preparation done, transferring control to OS
1385 -1 common/cmd_doc.c Bad usage of "doc" command
1386 -1 common/cmd_doc.c No boot device
1387 -1 common/cmd_doc.c Unknown Chip ID on boot device
1388 -1 common/cmd_doc.c Read Error on boot device
1389 -1 common/cmd_doc.c Image header has bad magic number
1391 -1 common/cmd_ide.c Bad usage of "ide" command
1392 -1 common/cmd_ide.c No boot device
1393 -1 common/cmd_ide.c Unknown boot device
1394 -1 common/cmd_ide.c Unknown partition table
1395 -1 common/cmd_ide.c Invalid partition type
1396 -1 common/cmd_ide.c Read Error on boot device
1397 -1 common/cmd_ide.c Image header has bad magic number
1399 -1 common/cmd_nvedit.c Environment not changable, but has bad CRC
1405 [so far only for SMDK2400 and TRAB boards]
1407 - Modem support endable:
1408 CONFIG_MODEM_SUPPORT
1410 - RTS/CTS Flow control enable:
1413 - Modem debug support:
1414 CONFIG_MODEM_SUPPORT_DEBUG
1416 Enables debugging stuff (char screen[1024], dbg())
1417 for modem support. Useful only with BDI2000.
1421 In the target system modem support is enabled when a
1422 specific key (key combination) is pressed during
1423 power-on. Otherwise U-Boot will boot normally
1424 (autoboot). The key_pressed() fuction is called from
1425 board_init(). Currently key_pressed() is a dummy
1426 function, returning 1 and thus enabling modem
1429 If there are no modem init strings in the
1430 environment, U-Boot proceed to autoboot; the
1431 previous output (banner, info printfs) will be
1434 See also: doc/README.Modem
1437 Configuration Settings:
1438 -----------------------
1440 - CFG_LONGHELP: Defined when you want long help messages included;
1441 undefine this when you're short of memory.
1443 - CFG_PROMPT: This is what U-Boot prints on the console to
1444 prompt for user input.
1446 - CFG_CBSIZE: Buffer size for input from the Console
1448 - CFG_PBSIZE: Buffer size for Console output
1450 - CFG_MAXARGS: max. Number of arguments accepted for monitor commands
1452 - CFG_BARGSIZE: Buffer size for Boot Arguments which are passed to
1453 the application (usually a Linux kernel) when it is
1456 - CFG_BAUDRATE_TABLE:
1457 List of legal baudrate settings for this board.
1459 - CFG_CONSOLE_INFO_QUIET
1460 Suppress display of console information at boot.
1462 - CFG_CONSOLE_IS_IN_ENV
1463 If the board specific function
1464 extern int overwrite_console (void);
1465 returns 1, the stdin, stderr and stdout are switched to the
1466 serial port, else the settings in the environment are used.
1468 - CFG_CONSOLE_OVERWRITE_ROUTINE
1469 Enable the call to overwrite_console().
1471 - CFG_CONSOLE_ENV_OVERWRITE
1472 Enable overwrite of previous console environment settings.
1474 - CFG_MEMTEST_START, CFG_MEMTEST_END:
1475 Begin and End addresses of the area used by the
1479 Enable an alternate, more extensive memory test.
1481 - CFG_TFTP_LOADADDR:
1482 Default load address for network file downloads
1484 - CFG_LOADS_BAUD_CHANGE:
1485 Enable temporary baudrate change while serial download
1488 Physical start address of SDRAM. _Must_ be 0 here.
1491 Physical start address of Motherboard I/O (if using a
1495 Physical start address of Flash memory.
1498 Physical start address of boot monitor code (set by
1499 make config files to be same as the text base address
1500 (TEXT_BASE) used when linking) - same as
1501 CFG_FLASH_BASE when booting from flash.
1504 Size of memory reserved for monitor code, used to
1505 determine _at_compile_time_ (!) if the environment is
1506 embedded within the U-Boot image, or in a separate
1510 Size of DRAM reserved for malloc() use.
1513 Maximum size of memory mapped by the startup code of
1514 the Linux kernel; all data that must be processed by
1515 the Linux kernel (bd_info, boot arguments, eventually
1516 initrd image) must be put below this limit.
1518 - CFG_MAX_FLASH_BANKS:
1519 Max number of Flash memory banks
1521 - CFG_MAX_FLASH_SECT:
1522 Max number of sectors on a Flash chip
1524 - CFG_FLASH_ERASE_TOUT:
1525 Timeout for Flash erase operations (in ms)
1527 - CFG_FLASH_WRITE_TOUT:
1528 Timeout for Flash write operations (in ms)
1530 - CFG_FLASH_LOCK_TOUT
1531 Timeout for Flash set sector lock bit operation (in ms)
1533 - CFG_FLASH_UNLOCK_TOUT
1534 Timeout for Flash clear lock bits operation (in ms)
1536 - CFG_FLASH_PROTECTION
1537 If defined, hardware flash sectors protection is used
1538 instead of U-Boot software protection.
1540 - CFG_DIRECT_FLASH_TFTP:
1542 Enable TFTP transfers directly to flash memory;
1543 without this option such a download has to be
1544 performed in two steps: (1) download to RAM, and (2)
1545 copy from RAM to flash.
1547 The two-step approach is usually more reliable, since
1548 you can check if the download worked before you erase
1549 the flash, but in some situations (when sytem RAM is
1550 too limited to allow for a tempory copy of the
1551 downloaded image) this option may be very useful.
1554 Define if the flash driver uses extra elements in the
1555 common flash structure for storing flash geometry
1557 - CFG_RX_ETH_BUFFER:
1558 Defines the number of ethernet receive buffers. On some
1559 ethernet controllers it is recommended to set this value
1560 to 8 or even higher (EEPRO100 or 405 EMAC), since all
1561 buffers can be full shortly after enabling the interface
1562 on high ethernet traffic.
1563 Defaults to 4 if not defined.
1565 The following definitions that deal with the placement and management
1566 of environment data (variable area); in general, we support the
1567 following configurations:
1569 - CFG_ENV_IS_IN_FLASH:
1571 Define this if the environment is in flash memory.
1573 a) The environment occupies one whole flash sector, which is
1574 "embedded" in the text segment with the U-Boot code. This
1575 happens usually with "bottom boot sector" or "top boot
1576 sector" type flash chips, which have several smaller
1577 sectors at the start or the end. For instance, such a
1578 layout can have sector sizes of 8, 2x4, 16, Nx32 kB. In
1579 such a case you would place the environment in one of the
1580 4 kB sectors - with U-Boot code before and after it. With
1581 "top boot sector" type flash chips, you would put the
1582 environment in one of the last sectors, leaving a gap
1583 between U-Boot and the environment.
1587 Offset of environment data (variable area) to the
1588 beginning of flash memory; for instance, with bottom boot
1589 type flash chips the second sector can be used: the offset
1590 for this sector is given here.
1592 CFG_ENV_OFFSET is used relative to CFG_FLASH_BASE.
1596 This is just another way to specify the start address of
1597 the flash sector containing the environment (instead of
1600 - CFG_ENV_SECT_SIZE:
1602 Size of the sector containing the environment.
1605 b) Sometimes flash chips have few, equal sized, BIG sectors.
1606 In such a case you don't want to spend a whole sector for
1611 If you use this in combination with CFG_ENV_IS_IN_FLASH
1612 and CFG_ENV_SECT_SIZE, you can specify to use only a part
1613 of this flash sector for the environment. This saves
1614 memory for the RAM copy of the environment.
1616 It may also save flash memory if you decide to use this
1617 when your environment is "embedded" within U-Boot code,
1618 since then the remainder of the flash sector could be used
1619 for U-Boot code. It should be pointed out that this is
1620 STRONGLY DISCOURAGED from a robustness point of view:
1621 updating the environment in flash makes it always
1622 necessary to erase the WHOLE sector. If something goes
1623 wrong before the contents has been restored from a copy in
1624 RAM, your target system will be dead.
1626 - CFG_ENV_ADDR_REDUND
1629 These settings describe a second storage area used to hold
1630 a redundand copy of the environment data, so that there is
1631 a valid backup copy in case there is a power failure during
1632 a "saveenv" operation.
1634 BE CAREFUL! Any changes to the flash layout, and some changes to the
1635 source code will make it necessary to adapt <board>/u-boot.lds*
1639 - CFG_ENV_IS_IN_NVRAM:
1641 Define this if you have some non-volatile memory device
1642 (NVRAM, battery buffered SRAM) which you want to use for the
1648 These two #defines are used to determin the memory area you
1649 want to use for environment. It is assumed that this memory
1650 can just be read and written to, without any special
1653 BE CAREFUL! The first access to the environment happens quite early
1654 in U-Boot initalization (when we try to get the setting of for the
1655 console baudrate). You *MUST* have mappend your NVRAM area then, or
1658 Please note that even with NVRAM we still use a copy of the
1659 environment in RAM: we could work on NVRAM directly, but we want to
1660 keep settings there always unmodified except somebody uses "saveenv"
1661 to save the current settings.
1664 - CFG_ENV_IS_IN_EEPROM:
1666 Use this if you have an EEPROM or similar serial access
1667 device and a driver for it.
1672 These two #defines specify the offset and size of the
1673 environment area within the total memory of your EEPROM.
1675 - CFG_I2C_EEPROM_ADDR:
1676 If defined, specified the chip address of the EEPROM device.
1677 The default address is zero.
1679 - CFG_EEPROM_PAGE_WRITE_BITS:
1680 If defined, the number of bits used to address bytes in a
1681 single page in the EEPROM device. A 64 byte page, for example
1682 would require six bits.
1684 - CFG_EEPROM_PAGE_WRITE_DELAY_MS:
1685 If defined, the number of milliseconds to delay between
1686 page writes. The default is zero milliseconds.
1688 - CFG_I2C_EEPROM_ADDR_LEN:
1689 The length in bytes of the EEPROM memory array address. Note
1690 that this is NOT the chip address length!
1693 The size in bytes of the EEPROM device.
1696 - CFG_SPI_INIT_OFFSET
1698 Defines offset to the initial SPI buffer area in DPRAM. The
1699 area is used at an early stage (ROM part) if the environment
1700 is configured to reside in the SPI EEPROM: We need a 520 byte
1701 scratch DPRAM area. It is used between the two initialization
1702 calls (spi_init_f() and spi_init_r()). A value of 0xB00 seems
1703 to be a good choice since it makes it far enough from the
1704 start of the data area as well as from the stack pointer.
1706 Please note that the environment is read-only as long as the monitor
1707 has been relocated to RAM and a RAM copy of the environment has been
1708 created; also, when using EEPROM you will have to use getenv_r()
1709 until then to read environment variables.
1711 The environment is protected by a CRC32 checksum. Before the monitor
1712 is relocated into RAM, as a result of a bad CRC you will be working
1713 with the compiled-in default environment - *silently*!!! [This is
1714 necessary, because the first environment variable we need is the
1715 "baudrate" setting for the console - if we have a bad CRC, we don't
1716 have any device yet where we could complain.]
1718 Note: once the monitor has been relocated, then it will complain if
1719 the default environment is used; a new CRC is computed as soon as you
1720 use the "saveenv" command to store a valid environment.
1723 Low Level (hardware related) configuration options:
1724 ---------------------------------------------------
1726 - CFG_CACHELINE_SIZE:
1727 Cache Line Size of the CPU.
1730 Default address of the IMMR after system reset.
1732 Needed on some 8260 systems (MPC8260ADS, PQ2FADS-ZU,
1733 and RPXsuper) to be able to adjust the position of
1734 the IMMR register after a reset.
1736 - Floppy Disk Support:
1737 CFG_FDC_DRIVE_NUMBER
1739 the default drive number (default value 0)
1743 defines the spacing between fdc chipset registers
1748 defines the offset of register from address. It
1749 depends on which part of the data bus is connected to
1750 the fdc chipset. (default value 0)
1752 If CFG_ISA_IO_STRIDE CFG_ISA_IO_OFFSET and
1753 CFG_FDC_DRIVE_NUMBER are undefined, they take their
1756 if CFG_FDC_HW_INIT is defined, then the function
1757 fdc_hw_init() is called at the beginning of the FDC
1758 setup. fdc_hw_init() must be provided by the board
1759 source code. It is used to make hardware dependant
1762 - CFG_IMMR: Physical address of the Internal Memory Mapped
1763 Register; DO NOT CHANGE! (11-4)
1764 [MPC8xx systems only]
1766 - CFG_INIT_RAM_ADDR:
1768 Start address of memory area tha can be used for
1769 initial data and stack; please note that this must be
1770 writable memory that is working WITHOUT special
1771 initialization, i. e. you CANNOT use normal RAM which
1772 will become available only after programming the
1773 memory controller and running certain initialization
1776 U-Boot uses the following memory types:
1777 - MPC8xx and MPC8260: IMMR (internal memory of the CPU)
1778 - MPC824X: data cache
1779 - PPC4xx: data cache
1781 - CFG_GBL_DATA_OFFSET:
1783 Offset of the initial data structure in the memory
1784 area defined by CFG_INIT_RAM_ADDR. Usually
1785 CFG_GBL_DATA_OFFSET is chosen such that the initial
1786 data is located at the end of the available space
1787 (sometimes written as (CFG_INIT_RAM_END -
1788 CFG_INIT_DATA_SIZE), and the initial stack is just
1789 below that area (growing from (CFG_INIT_RAM_ADDR +
1790 CFG_GBL_DATA_OFFSET) downward.
1793 On the MPC824X (or other systems that use the data
1794 cache for initial memory) the address chosen for
1795 CFG_INIT_RAM_ADDR is basically arbitrary - it must
1796 point to an otherwise UNUSED address space between
1797 the top of RAM and the start of the PCI space.
1799 - CFG_SIUMCR: SIU Module Configuration (11-6)
1801 - CFG_SYPCR: System Protection Control (11-9)
1803 - CFG_TBSCR: Time Base Status and Control (11-26)
1805 - CFG_PISCR: Periodic Interrupt Status and Control (11-31)
1807 - CFG_PLPRCR: PLL, Low-Power, and Reset Control Register (15-30)
1809 - CFG_SCCR: System Clock and reset Control Register (15-27)
1811 - CFG_OR_TIMING_SDRAM:
1815 periodic timer for refresh
1817 - CFG_DER: Debug Event Register (37-47)
1819 - FLASH_BASE0_PRELIM, FLASH_BASE1_PRELIM, CFG_REMAP_OR_AM,
1820 CFG_PRELIM_OR_AM, CFG_OR_TIMING_FLASH, CFG_OR0_REMAP,
1821 CFG_OR0_PRELIM, CFG_BR0_PRELIM, CFG_OR1_REMAP, CFG_OR1_PRELIM,
1823 Memory Controller Definitions: BR0/1 and OR0/1 (FLASH)
1825 - SDRAM_BASE2_PRELIM, SDRAM_BASE3_PRELIM, SDRAM_MAX_SIZE,
1826 CFG_OR_TIMING_SDRAM, CFG_OR2_PRELIM, CFG_BR2_PRELIM,
1827 CFG_OR3_PRELIM, CFG_BR3_PRELIM:
1828 Memory Controller Definitions: BR2/3 and OR2/3 (SDRAM)
1830 - CFG_MAMR_PTA, CFG_MPTPR_2BK_4K, CFG_MPTPR_1BK_4K, CFG_MPTPR_2BK_8K,
1831 CFG_MPTPR_1BK_8K, CFG_MAMR_8COL, CFG_MAMR_9COL:
1832 Machine Mode Register and Memory Periodic Timer
1833 Prescaler definitions (SDRAM timing)
1835 - CFG_I2C_UCODE_PATCH, CFG_I2C_DPMEM_OFFSET [0x1FC0]:
1836 enable I2C microcode relocation patch (MPC8xx);
1837 define relocation offset in DPRAM [DSP2]
1839 - CFG_SPI_UCODE_PATCH, CFG_SPI_DPMEM_OFFSET [0x1FC0]:
1840 enable SPI microcode relocation patch (MPC8xx);
1841 define relocation offset in DPRAM [SCC4]
1844 Use OSCM clock mode on MBX8xx board. Be careful,
1845 wrong setting might damage your board. Read
1846 doc/README.MBX before setting this variable!
1848 - CFG_CPM_POST_WORD_ADDR: (MPC8xx, MPC8260 only)
1849 Offset of the bootmode word in DPRAM used by post
1850 (Power On Self Tests). This definition overrides
1851 #define'd default value in commproc.h resp.
1854 - CFG_PCI_SLV_MEM_LOCAL, CFG_PCI_SLV_MEM_BUS, CFG_PICMR0_MASK_ATTRIB,
1855 CFG_PCI_MSTR0_LOCAL, CFG_PCIMSK0_MASK, CFG_PCI_MSTR1_LOCAL,
1856 CFG_PCIMSK1_MASK, CFG_PCI_MSTR_MEM_LOCAL, CFG_PCI_MSTR_MEM_BUS,
1857 CFG_CPU_PCI_MEM_START, CFG_PCI_MSTR_MEM_SIZE, CFG_POCMR0_MASK_ATTRIB,
1858 CFG_PCI_MSTR_MEMIO_LOCAL, CFG_PCI_MSTR_MEMIO_BUS, CPU_PCI_MEMIO_START,
1859 CFG_PCI_MSTR_MEMIO_SIZE, CFG_POCMR1_MASK_ATTRIB, CFG_PCI_MSTR_IO_LOCAL,
1860 CFG_PCI_MSTR_IO_BUS, CFG_CPU_PCI_IO_START, CFG_PCI_MSTR_IO_SIZE,
1861 CFG_POCMR2_MASK_ATTRIB: (MPC826x only)
1862 Overrides the default PCI memory map in cpu/mpc8260/pci.c if set.
1864 Building the Software:
1865 ======================
1867 Building U-Boot has been tested in native PPC environments (on a
1868 PowerBook G3 running LinuxPPC 2000) and in cross environments
1869 (running RedHat 6.x and 7.x Linux on x86, Solaris 2.6 on a SPARC, and
1872 If you are not using a native PPC environment, it is assumed that you
1873 have the GNU cross compiling tools available in your path and named
1874 with a prefix of "powerpc-linux-". If this is not the case, (e.g. if
1875 you are using Monta Vista's Hard Hat Linux CDK 1.2) you must change
1876 the definition of CROSS_COMPILE in Makefile. For HHL on a 4xx CPU,
1879 CROSS_COMPILE = ppc_4xx-
1882 U-Boot is intended to be simple to build. After installing the
1883 sources you must configure U-Boot for one specific board type. This
1888 where "NAME_config" is the name of one of the existing
1889 configurations; the following names are supported:
1891 ADCIOP_config GTH_config TQM850L_config
1892 ADS860_config IP860_config TQM855L_config
1893 AR405_config IVML24_config TQM860L_config
1894 CANBT_config IVMS8_config WALNUT405_config
1895 CPCI405_config LANTEC_config cogent_common_config
1896 CPCIISER4_config MBX_config cogent_mpc8260_config
1897 CU824_config MBX860T_config cogent_mpc8xx_config
1898 ESTEEM192E_config RPXlite_config hermes_config
1899 ETX094_config RPXsuper_config hymod_config
1900 FADS823_config SM850_config lwmon_config
1901 FADS850SAR_config SPD823TS_config pcu_e_config
1902 FADS860T_config SXNI855T_config rsdproto_config
1903 FPS850L_config Sandpoint8240_config sbc8260_config
1904 GENIETV_config TQM823L_config PIP405_config
1905 GEN860T_config EBONY_config FPS860L_config
1906 ELPT860_config cmi_mpc5xx_config NETVIA_config
1907 at91rm9200dk_config omap1510inn_config MPC8260ADS_config
1909 Note: for some board special configuration names may exist; check if
1910 additional information is available from the board vendor; for
1911 instance, the TQM8xxL systems run normally at 50 MHz and use a
1912 SCC for 10baseT ethernet; there are also systems with 80 MHz
1913 CPU clock, and an optional Fast Ethernet module is available
1914 for CPU's with FEC. You can select such additional "features"
1915 when chosing the configuration, i. e.
1918 - will configure for a plain TQM860L, i. e. 50MHz, no FEC
1920 make TQM860L_FEC_config
1921 - will configure for a TQM860L at 50MHz with FEC for ethernet
1923 make TQM860L_80MHz_config
1924 - will configure for a TQM860L at 80 MHz, with normal 10baseT
1927 make TQM860L_FEC_80MHz_config
1928 - will configure for a TQM860L at 80 MHz with FEC for ethernet
1930 make TQM823L_LCD_config
1931 - will configure for a TQM823L with U-Boot console on LCD
1933 make TQM823L_LCD_80MHz_config
1934 - will configure for a TQM823L at 80 MHz with U-Boot console on LCD
1939 Finally, type "make all", and you should get some working U-Boot
1940 images ready for downlod to / installation on your system:
1942 - "u-boot.bin" is a raw binary image
1943 - "u-boot" is an image in ELF binary format
1944 - "u-boot.srec" is in Motorola S-Record format
1947 Please be aware that the Makefiles assume you are using GNU make, so
1948 for instance on NetBSD you might need to use "gmake" instead of
1952 If the system board that you have is not listed, then you will need
1953 to port U-Boot to your hardware platform. To do this, follow these
1956 1. Add a new configuration option for your board to the toplevel
1957 "Makefile" and to the "MAKEALL" script, using the existing
1958 entries as examples. Note that here and at many other places
1959 boards and other names are listed alphabetically sorted. Please
1961 2. Create a new directory to hold your board specific code. Add any
1962 files you need. In your board directory, you will need at least
1963 the "Makefile", a "<board>.c", "flash.c" and "u-boot.lds".
1964 3. Create a new configuration file "include/configs/<board>.h" for
1966 3. If you're porting U-Boot to a new CPU, then also create a new
1967 directory to hold your CPU specific code. Add any files you need.
1968 4. Run "make <board>_config" with your new name.
1969 5. Type "make", and you should get a working "u-boot.srec" file
1970 to be installed on your target system.
1971 6. Debug and solve any problems that might arise.
1972 [Of course, this last step is much harder than it sounds.]
1975 Testing of U-Boot Modifications, Ports to New Hardware, etc.:
1976 ==============================================================
1978 If you have modified U-Boot sources (for instance added a new board
1979 or support for new devices, a new CPU, etc.) you are expected to
1980 provide feedback to the other developers. The feedback normally takes
1981 the form of a "patch", i. e. a context diff against a certain (latest
1982 official or latest in CVS) version of U-Boot sources.
1984 But before you submit such a patch, please verify that your modifi-
1985 cation did not break existing code. At least make sure that *ALL* of
1986 the supported boards compile WITHOUT ANY compiler warnings. To do so,
1987 just run the "MAKEALL" script, which will configure and build U-Boot
1988 for ALL supported system. Be warned, this will take a while. You can
1989 select which (cross) compiler to use py passing a `CROSS_COMPILE'
1990 environment variable to the script, i. e. to use the cross tools from
1991 MontaVista's Hard Hat Linux you can type
1993 CROSS_COMPILE=ppc_8xx- MAKEALL
1995 or to build on a native PowerPC system you can type
1997 CROSS_COMPILE=' ' MAKEALL
1999 See also "U-Boot Porting Guide" below.
2002 Monitor Commands - Overview:
2003 ============================
2005 go - start application at address 'addr'
2006 run - run commands in an environment variable
2007 bootm - boot application image from memory
2008 bootp - boot image via network using BootP/TFTP protocol
2009 tftpboot- boot image via network using TFTP protocol
2010 and env variables "ipaddr" and "serverip"
2011 (and eventually "gatewayip")
2012 rarpboot- boot image via network using RARP/TFTP protocol
2013 diskboot- boot from IDE devicebootd - boot default, i.e., run 'bootcmd'
2014 loads - load S-Record file over serial line
2015 loadb - load binary file over serial line (kermit mode)
2017 mm - memory modify (auto-incrementing)
2018 nm - memory modify (constant address)
2019 mw - memory write (fill)
2021 cmp - memory compare
2022 crc32 - checksum calculation
2023 imd - i2c memory display
2024 imm - i2c memory modify (auto-incrementing)
2025 inm - i2c memory modify (constant address)
2026 imw - i2c memory write (fill)
2027 icrc32 - i2c checksum calculation
2028 iprobe - probe to discover valid I2C chip addresses
2029 iloop - infinite loop on address range
2030 isdram - print SDRAM configuration information
2031 sspi - SPI utility commands
2032 base - print or set address offset
2033 printenv- print environment variables
2034 setenv - set environment variables
2035 saveenv - save environment variables to persistent storage
2036 protect - enable or disable FLASH write protection
2037 erase - erase FLASH memory
2038 flinfo - print FLASH memory information
2039 bdinfo - print Board Info structure
2040 iminfo - print header information for application image
2041 coninfo - print console devices and informations
2042 ide - IDE sub-system
2043 loop - infinite loop on address range
2044 mtest - simple RAM test
2045 icache - enable or disable instruction cache
2046 dcache - enable or disable data cache
2047 reset - Perform RESET of the CPU
2048 echo - echo args to console
2049 version - print monitor version
2050 help - print online help
2051 ? - alias for 'help'
2054 Monitor Commands - Detailed Description:
2055 ========================================
2059 For now: just type "help <command>".
2062 Environment Variables:
2063 ======================
2065 U-Boot supports user configuration using Environment Variables which
2066 can be made persistent by saving to Flash memory.
2068 Environment Variables are set using "setenv", printed using
2069 "printenv", and saved to Flash using "saveenv". Using "setenv"
2070 without a value can be used to delete a variable from the
2071 environment. As long as you don't save the environment you are
2072 working with an in-memory copy. In case the Flash area containing the
2073 environment is erased by accident, a default environment is provided.
2075 Some configuration options can be set using Environment Variables:
2077 baudrate - see CONFIG_BAUDRATE
2079 bootdelay - see CONFIG_BOOTDELAY
2081 bootcmd - see CONFIG_BOOTCOMMAND
2083 bootargs - Boot arguments when booting an RTOS image
2085 bootfile - Name of the image to load with TFTP
2087 autoload - if set to "no" (any string beginning with 'n'),
2088 "bootp" will just load perform a lookup of the
2089 configuration from the BOOTP server, but not try to
2090 load any image using TFTP
2092 autostart - if set to "yes", an image loaded using the "bootp",
2093 "rarpboot", "tftpboot" or "diskboot" commands will
2094 be automatically started (by internally calling
2097 If set to "no", a standalone image passed to the
2098 "bootm" command will be copied to the load address
2099 (and eventually uncompressed), but NOT be started.
2100 This can be used to load and uncompress arbitrary
2103 initrd_high - restrict positioning of initrd images:
2104 If this variable is not set, initrd images will be
2105 copied to the highest possible address in RAM; this
2106 is usually what you want since it allows for
2107 maximum initrd size. If for some reason you want to
2108 make sure that the initrd image is loaded below the
2109 CFG_BOOTMAPSZ limit, you can set this environment
2110 variable to a value of "no" or "off" or "0".
2111 Alternatively, you can set it to a maximum upper
2112 address to use (U-Boot will still check that it
2113 does not overwrite the U-Boot stack and data).
2115 For instance, when you have a system with 16 MB
2116 RAM, and want to reseve 4 MB from use by Linux,
2117 you can do this by adding "mem=12M" to the value of
2118 the "bootargs" variable. However, now you must make
2119 sure, that the initrd image is placed in the first
2120 12 MB as well - this can be done with
2122 setenv initrd_high 00c00000
2124 If you set initrd_high to 0xFFFFFFFF, this is an
2125 indication to U-Boot that all addresses are legal
2126 for the Linux kernel, including addresses in flash
2127 memory. In this case U-Boot will NOT COPY the
2128 ramdisk at all. This may be useful to reduce the
2129 boot time on your system, but requires that this
2130 feature is supported by your Linux kernel.
2132 ipaddr - IP address; needed for tftpboot command
2134 loadaddr - Default load address for commands like "bootp",
2135 "rarpboot", "tftpboot", "loadb" or "diskboot"
2137 loads_echo - see CONFIG_LOADS_ECHO
2139 serverip - TFTP server IP address; needed for tftpboot command
2141 bootretry - see CONFIG_BOOT_RETRY_TIME
2143 bootdelaykey - see CONFIG_AUTOBOOT_DELAY_STR
2145 bootstopkey - see CONFIG_AUTOBOOT_STOP_STR
2148 The following environment variables may be used and automatically
2149 updated by the network boot commands ("bootp" and "rarpboot"),
2150 depending the information provided by your boot server:
2152 bootfile - see above
2153 dnsip - IP address of your Domain Name Server
2154 dnsip2 - IP address of your secondary Domain Name Server
2155 gatewayip - IP address of the Gateway (Router) to use
2156 hostname - Target hostname
2158 netmask - Subnet Mask
2159 rootpath - Pathname of the root filesystem on the NFS server
2160 serverip - see above
2163 There are two special Environment Variables:
2165 serial# - contains hardware identification information such
2166 as type string and/or serial number
2167 ethaddr - Ethernet address
2169 These variables can be set only once (usually during manufacturing of
2170 the board). U-Boot refuses to delete or overwrite these variables
2171 once they have been set once.
2174 Further special Environment Variables:
2176 ver - Contains the U-Boot version string as printed
2177 with the "version" command. This variable is
2178 readonly (see CONFIG_VERSION_VARIABLE).
2181 Please note that changes to some configuration parameters may take
2182 only effect after the next boot (yes, that's just like Windoze :-).
2185 Command Line Parsing:
2186 =====================
2188 There are two different command line parsers available with U-Boot:
2189 the old "simple" one, and the much more pwerful "hush" shell:
2191 Old, simple command line parser:
2192 --------------------------------
2194 - supports environment variables (through setenv / saveenv commands)
2195 - several commands on one line, separated by ';'
2196 - variable substitution using "... $(name) ..." syntax
2197 - special characters ('$', ';') can be escaped by prefixing with '\',
2199 setenv bootcmd bootm \$(address)
2200 - You can also escape text by enclosing in single apostrophes, for example:
2201 setenv addip 'setenv bootargs $bootargs ip=$ipaddr:$serverip:$gatewayip:$netmask:$hostname::off'
2206 - similar to Bourne shell, with control structures like
2207 if...then...else...fi, for...do...done; while...do...done,
2208 until...do...done, ...
2209 - supports environment ("global") variables (through setenv / saveenv
2210 commands) and local shell variables (through standard shell syntax
2211 "name=value"); only environment variables can be used with "run"
2217 (1) If a command line (or an environment variable executed by a "run"
2218 command) contains several commands separated by semicolon, and
2219 one of these commands fails, then the remaining commands will be
2222 (2) If you execute several variables with one call to run (i. e.
2223 calling run with a list af variables as arguments), any failing
2224 command will cause "run" to terminate, i. e. the remaining
2225 variables are not executed.
2227 Note for Redundant Ethernet Interfaces:
2228 =======================================
2230 Some boards come with redundand ethernet interfaces; U-Boot supports
2231 such configurations and is capable of automatic selection of a
2232 "working" interface when needed. MAC assignemnt works as follows:
2234 Network interfaces are numbered eth0, eth1, eth2, ... Corresponding
2235 MAC addresses can be stored in the environment as "ethaddr" (=>eth0),
2236 "eth1addr" (=>eth1), "eth2addr", ...
2238 If the network interface stores some valid MAC address (for instance
2239 in SROM), this is used as default address if there is NO correspon-
2240 ding setting in the environment; if the corresponding environment
2241 variable is set, this overrides the settings in the card; that means:
2243 o If the SROM has a valid MAC address, and there is no address in the
2244 environment, the SROM's address is used.
2246 o If there is no valid address in the SROM, and a definition in the
2247 environment exists, then the value from the environment variable is
2250 o If both the SROM and the environment contain a MAC address, and
2251 both addresses are the same, this MAC address is used.
2253 o If both the SROM and the environment contain a MAC address, and the
2254 addresses differ, the value from the environment is used and a
2257 o If neither SROM nor the environment contain a MAC address, an error
2264 The "boot" commands of this monitor operate on "image" files which
2265 can be basicly anything, preceeded by a special header; see the
2266 definitions in include/image.h for details; basicly, the header
2267 defines the following image properties:
2269 * Target Operating System (Provisions for OpenBSD, NetBSD, FreeBSD,
2270 4.4BSD, Linux, SVR4, Esix, Solaris, Irix, SCO, Dell, NCR, VxWorks,
2271 LynxOS, pSOS, QNX, RTEMS, ARTOS;
2272 Currently supported: Linux, NetBSD, VxWorks, QNX, RTEMS, ARTOS, LynxOS).
2273 * Target CPU Architecture (Provisions for Alpha, ARM, Intel x86,
2274 IA64, MIPS, MIPS, PowerPC, IBM S390, SuperH, Sparc, Sparc 64 Bit;
2275 Currently supported: PowerPC).
2276 * Compression Type (uncompressed, gzip, bzip2)
2282 The header is marked by a special Magic Number, and both the header
2283 and the data portions of the image are secured against corruption by
2290 Although U-Boot should support any OS or standalone application
2291 easily, Linux has always been in the focus during the design of
2294 U-Boot includes many features that so far have been part of some
2295 special "boot loader" code within the Linux kernel. Also, any
2296 "initrd" images to be used are no longer part of one big Linux image;
2297 instead, kernel and "initrd" are separate images. This implementation
2298 serves serveral purposes:
2300 - the same features can be used for other OS or standalone
2301 applications (for instance: using compressed images to reduce the
2302 Flash memory footprint)
2304 - it becomes much easier to port new Linux kernel versions because
2305 lots of low-level, hardware dependend stuff are done by U-Boot
2307 - the same Linux kernel image can now be used with different "initrd"
2308 images; of course this also means that different kernel images can
2309 be run with the same "initrd". This makes testing easier (you don't
2310 have to build a new "zImage.initrd" Linux image when you just
2311 change a file in your "initrd"). Also, a field-upgrade of the
2312 software is easier now.
2318 Porting Linux to U-Boot based systems:
2319 ---------------------------------------
2321 U-Boot cannot save you from doing all the necessary modifications to
2322 configure the Linux device drivers for use with your target hardware
2323 (no, we don't intend to provide a full virtual machine interface to
2326 But now you can ignore ALL boot loader code (in arch/ppc/mbxboot).
2328 Just make sure your machine specific header file (for instance
2329 include/asm-ppc/tqm8xx.h) includes the same definition of the Board
2330 Information structure as we define in include/u-boot.h, and make
2331 sure that your definition of IMAP_ADDR uses the same value as your
2332 U-Boot configuration in CFG_IMMR.
2335 Configuring the Linux kernel:
2336 -----------------------------
2338 No specific requirements for U-Boot. Make sure you have some root
2339 device (initial ramdisk, NFS) for your target system.
2342 Building a Linux Image:
2343 -----------------------
2345 With U-Boot, "normal" build targets like "zImage" or "bzImage" are
2346 not used. If you use recent kernel source, a new build target
2347 "uImage" will exist which automatically builds an image usable by
2348 U-Boot. Most older kernels also have support for a "pImage" target,
2349 which was introduced for our predecessor project PPCBoot and uses a
2350 100% compatible format.
2359 The "uImage" build target uses a special tool (in 'tools/mkimage') to
2360 encapsulate a compressed Linux kernel image with header information,
2361 CRC32 checksum etc. for use with U-Boot. This is what we are doing:
2363 * build a standard "vmlinux" kernel image (in ELF binary format):
2365 * convert the kernel into a raw binary image:
2367 ${CROSS_COMPILE}-objcopy -O binary \
2368 -R .note -R .comment \
2369 -S vmlinux linux.bin
2371 * compress the binary image:
2375 * package compressed binary image for U-Boot:
2377 mkimage -A ppc -O linux -T kernel -C gzip \
2378 -a 0 -e 0 -n "Linux Kernel Image" \
2379 -d linux.bin.gz uImage
2382 The "mkimage" tool can also be used to create ramdisk images for use
2383 with U-Boot, either separated from the Linux kernel image, or
2384 combined into one file. "mkimage" encapsulates the images with a 64
2385 byte header containing information about target architecture,
2386 operating system, image type, compression method, entry points, time
2387 stamp, CRC32 checksums, etc.
2389 "mkimage" can be called in two ways: to verify existing images and
2390 print the header information, or to build new images.
2392 In the first form (with "-l" option) mkimage lists the information
2393 contained in the header of an existing U-Boot image; this includes
2394 checksum verification:
2396 tools/mkimage -l image
2397 -l ==> list image header information
2399 The second form (with "-d" option) is used to build a U-Boot image
2400 from a "data file" which is used as image payload:
2402 tools/mkimage -A arch -O os -T type -C comp -a addr -e ep \
2403 -n name -d data_file image
2404 -A ==> set architecture to 'arch'
2405 -O ==> set operating system to 'os'
2406 -T ==> set image type to 'type'
2407 -C ==> set compression type 'comp'
2408 -a ==> set load address to 'addr' (hex)
2409 -e ==> set entry point to 'ep' (hex)
2410 -n ==> set image name to 'name'
2411 -d ==> use image data from 'datafile'
2413 Right now, all Linux kernels use the same load address (0x00000000),
2414 but the entry point address depends on the kernel version:
2416 - 2.2.x kernels have the entry point at 0x0000000C,
2417 - 2.3.x and later kernels have the entry point at 0x00000000.
2419 So a typical call to build a U-Boot image would read:
2421 -> tools/mkimage -n '2.4.4 kernel for TQM850L' \
2422 > -A ppc -O linux -T kernel -C gzip -a 0 -e 0 \
2423 > -d /opt/elsk/ppc_8xx/usr/src/linux-2.4.4/arch/ppc/coffboot/vmlinux.gz \
2424 > examples/uImage.TQM850L
2425 Image Name: 2.4.4 kernel for TQM850L
2426 Created: Wed Jul 19 02:34:59 2000
2427 Image Type: PowerPC Linux Kernel Image (gzip compressed)
2428 Data Size: 335725 Bytes = 327.86 kB = 0.32 MB
2429 Load Address: 0x00000000
2430 Entry Point: 0x00000000
2432 To verify the contents of the image (or check for corruption):
2434 -> tools/mkimage -l examples/uImage.TQM850L
2435 Image Name: 2.4.4 kernel for TQM850L
2436 Created: Wed Jul 19 02:34:59 2000
2437 Image Type: PowerPC Linux Kernel Image (gzip compressed)
2438 Data Size: 335725 Bytes = 327.86 kB = 0.32 MB
2439 Load Address: 0x00000000
2440 Entry Point: 0x00000000
2442 NOTE: for embedded systems where boot time is critical you can trade
2443 speed for memory and install an UNCOMPRESSED image instead: this
2444 needs more space in Flash, but boots much faster since it does not
2445 need to be uncompressed:
2447 -> gunzip /opt/elsk/ppc_8xx/usr/src/linux-2.4.4/arch/ppc/coffboot/vmlinux.gz
2448 -> tools/mkimage -n '2.4.4 kernel for TQM850L' \
2449 > -A ppc -O linux -T kernel -C none -a 0 -e 0 \
2450 > -d /opt/elsk/ppc_8xx/usr/src/linux-2.4.4/arch/ppc/coffboot/vmlinux \
2451 > examples/uImage.TQM850L-uncompressed
2452 Image Name: 2.4.4 kernel for TQM850L
2453 Created: Wed Jul 19 02:34:59 2000
2454 Image Type: PowerPC Linux Kernel Image (uncompressed)
2455 Data Size: 792160 Bytes = 773.59 kB = 0.76 MB
2456 Load Address: 0x00000000
2457 Entry Point: 0x00000000
2460 Similar you can build U-Boot images from a 'ramdisk.image.gz' file
2461 when your kernel is intended to use an initial ramdisk:
2463 -> tools/mkimage -n 'Simple Ramdisk Image' \
2464 > -A ppc -O linux -T ramdisk -C gzip \
2465 > -d /LinuxPPC/images/SIMPLE-ramdisk.image.gz examples/simple-initrd
2466 Image Name: Simple Ramdisk Image
2467 Created: Wed Jan 12 14:01:50 2000
2468 Image Type: PowerPC Linux RAMDisk Image (gzip compressed)
2469 Data Size: 566530 Bytes = 553.25 kB = 0.54 MB
2470 Load Address: 0x00000000
2471 Entry Point: 0x00000000
2474 Installing a Linux Image:
2475 -------------------------
2477 To downloading a U-Boot image over the serial (console) interface,
2478 you must convert the image to S-Record format:
2480 objcopy -I binary -O srec examples/image examples/image.srec
2482 The 'objcopy' does not understand the information in the U-Boot
2483 image header, so the resulting S-Record file will be relative to
2484 address 0x00000000. To load it to a given address, you need to
2485 specify the target address as 'offset' parameter with the 'loads'
2488 Example: install the image to address 0x40100000 (which on the
2489 TQM8xxL is in the first Flash bank):
2491 => erase 40100000 401FFFFF
2497 ## Ready for S-Record download ...
2498 ~>examples/image.srec
2499 1 2 3 4 5 6 7 8 9 10 11 12 13 ...
2501 15989 15990 15991 15992
2502 [file transfer complete]
2504 ## Start Addr = 0x00000000
2507 You can check the success of the download using the 'iminfo' command;
2508 this includes a checksum verification so you can be sure no data
2509 corruption happened:
2513 ## Checking Image at 40100000 ...
2514 Image Name: 2.2.13 for initrd on TQM850L
2515 Image Type: PowerPC Linux Kernel Image (gzip compressed)
2516 Data Size: 335725 Bytes = 327 kB = 0 MB
2517 Load Address: 00000000
2518 Entry Point: 0000000c
2519 Verifying Checksum ... OK
2525 The "bootm" command is used to boot an application that is stored in
2526 memory (RAM or Flash). In case of a Linux kernel image, the contents
2527 of the "bootargs" environment variable is passed to the kernel as
2528 parameters. You can check and modify this variable using the
2529 "printenv" and "setenv" commands:
2532 => printenv bootargs
2533 bootargs=root=/dev/ram
2535 => setenv bootargs root=/dev/nfs rw nfsroot=10.0.0.2:/LinuxPPC nfsaddrs=10.0.0.99:10.0.0.2
2537 => printenv bootargs
2538 bootargs=root=/dev/nfs rw nfsroot=10.0.0.2:/LinuxPPC nfsaddrs=10.0.0.99:10.0.0.2
2541 ## Booting Linux kernel at 40020000 ...
2542 Image Name: 2.2.13 for NFS on TQM850L
2543 Image Type: PowerPC Linux Kernel Image (gzip compressed)
2544 Data Size: 381681 Bytes = 372 kB = 0 MB
2545 Load Address: 00000000
2546 Entry Point: 0000000c
2547 Verifying Checksum ... OK
2548 Uncompressing Kernel Image ... OK
2549 Linux version 2.2.13 (wd@denx.local.net) (gcc version 2.95.2 19991024 (release)) #1 Wed Jul 19 02:35:17 MEST 2000
2550 Boot arguments: root=/dev/nfs rw nfsroot=10.0.0.2:/LinuxPPC nfsaddrs=10.0.0.99:10.0.0.2
2551 time_init: decrementer frequency = 187500000/60
2552 Calibrating delay loop... 49.77 BogoMIPS
2553 Memory: 15208k available (700k kernel code, 444k data, 32k init) [c0000000,c1000000]
2556 If you want to boot a Linux kernel with initial ram disk, you pass
2557 the memory addreses of both the kernel and the initrd image (PPBCOOT
2558 format!) to the "bootm" command:
2560 => imi 40100000 40200000
2562 ## Checking Image at 40100000 ...
2563 Image Name: 2.2.13 for initrd on TQM850L
2564 Image Type: PowerPC Linux Kernel Image (gzip compressed)
2565 Data Size: 335725 Bytes = 327 kB = 0 MB
2566 Load Address: 00000000
2567 Entry Point: 0000000c
2568 Verifying Checksum ... OK
2570 ## Checking Image at 40200000 ...
2571 Image Name: Simple Ramdisk Image
2572 Image Type: PowerPC Linux RAMDisk Image (gzip compressed)
2573 Data Size: 566530 Bytes = 553 kB = 0 MB
2574 Load Address: 00000000
2575 Entry Point: 00000000
2576 Verifying Checksum ... OK
2578 => bootm 40100000 40200000
2579 ## Booting Linux kernel at 40100000 ...
2580 Image Name: 2.2.13 for initrd on TQM850L
2581 Image Type: PowerPC Linux Kernel Image (gzip compressed)
2582 Data Size: 335725 Bytes = 327 kB = 0 MB
2583 Load Address: 00000000
2584 Entry Point: 0000000c
2585 Verifying Checksum ... OK
2586 Uncompressing Kernel Image ... OK
2587 ## Loading RAMDisk Image at 40200000 ...
2588 Image Name: Simple Ramdisk Image
2589 Image Type: PowerPC Linux RAMDisk Image (gzip compressed)
2590 Data Size: 566530 Bytes = 553 kB = 0 MB
2591 Load Address: 00000000
2592 Entry Point: 00000000
2593 Verifying Checksum ... OK
2594 Loading Ramdisk ... OK
2595 Linux version 2.2.13 (wd@denx.local.net) (gcc version 2.95.2 19991024 (release)) #1 Wed Jul 19 02:32:08 MEST 2000
2596 Boot arguments: root=/dev/ram
2597 time_init: decrementer frequency = 187500000/60
2598 Calibrating delay loop... 49.77 BogoMIPS
2600 RAMDISK: Compressed image found at block 0
2601 VFS: Mounted root (ext2 filesystem).
2605 More About U-Boot Image Types:
2606 ------------------------------
2608 U-Boot supports the following image types:
2610 "Standalone Programs" are directly runnable in the environment
2611 provided by U-Boot; it is expected that (if they behave
2612 well) you can continue to work in U-Boot after return from
2613 the Standalone Program.
2614 "OS Kernel Images" are usually images of some Embedded OS which
2615 will take over control completely. Usually these programs
2616 will install their own set of exception handlers, device
2617 drivers, set up the MMU, etc. - this means, that you cannot
2618 expect to re-enter U-Boot except by resetting the CPU.
2619 "RAMDisk Images" are more or less just data blocks, and their
2620 parameters (address, size) are passed to an OS kernel that is
2622 "Multi-File Images" contain several images, typically an OS
2623 (Linux) kernel image and one or more data images like
2624 RAMDisks. This construct is useful for instance when you want
2625 to boot over the network using BOOTP etc., where the boot
2626 server provides just a single image file, but you want to get
2627 for instance an OS kernel and a RAMDisk image.
2629 "Multi-File Images" start with a list of image sizes, each
2630 image size (in bytes) specified by an "uint32_t" in network
2631 byte order. This list is terminated by an "(uint32_t)0".
2632 Immediately after the terminating 0 follow the images, one by
2633 one, all aligned on "uint32_t" boundaries (size rounded up to
2634 a multiple of 4 bytes).
2636 "Firmware Images" are binary images containing firmware (like
2637 U-Boot or FPGA images) which usually will be programmed to
2640 "Script files" are command sequences that will be executed by
2641 U-Boot's command interpreter; this feature is especially
2642 useful when you configure U-Boot to use a real shell (hush)
2643 as command interpreter.
2649 One of the features of U-Boot is that you can dynamically load and
2650 run "standalone" applications, which can use some resources of
2651 U-Boot like console I/O functions or interrupt services.
2653 Two simple examples are included with the sources:
2658 'examples/hello_world.c' contains a small "Hello World" Demo
2659 application; it is automatically compiled when you build U-Boot.
2660 It's configured to run at address 0x00040004, so you can play with it
2664 ## Ready for S-Record download ...
2665 ~>examples/hello_world.srec
2666 1 2 3 4 5 6 7 8 9 10 11 ...
2667 [file transfer complete]
2669 ## Start Addr = 0x00040004
2671 => go 40004 Hello World! This is a test.
2672 ## Starting application at 0x00040004 ...
2683 Hit any key to exit ...
2685 ## Application terminated, rc = 0x0
2687 Another example, which demonstrates how to register a CPM interrupt
2688 handler with the U-Boot code, can be found in 'examples/timer.c'.
2689 Here, a CPM timer is set up to generate an interrupt every second.
2690 The interrupt service routine is trivial, just printing a '.'
2691 character, but this is just a demo program. The application can be
2692 controlled by the following keys:
2694 ? - print current values og the CPM Timer registers
2695 b - enable interrupts and start timer
2696 e - stop timer and disable interrupts
2697 q - quit application
2700 ## Ready for S-Record download ...
2701 ~>examples/timer.srec
2702 1 2 3 4 5 6 7 8 9 10 11 ...
2703 [file transfer complete]
2705 ## Start Addr = 0x00040004
2708 ## Starting application at 0x00040004 ...
2711 tgcr @ 0xfff00980, tmr @ 0xfff00990, trr @ 0xfff00994, tcr @ 0xfff00998, tcn @ 0xfff0099c, ter @ 0xfff009b0
2714 [q, b, e, ?] Set interval 1000000 us
2717 [q, b, e, ?] ........
2718 tgcr=0x1, tmr=0xff1c, trr=0x3d09, tcr=0x0, tcn=0xef6, ter=0x0
2721 tgcr=0x1, tmr=0xff1c, trr=0x3d09, tcr=0x0, tcn=0x2ad4, ter=0x0
2724 tgcr=0x1, tmr=0xff1c, trr=0x3d09, tcr=0x0, tcn=0x1efc, ter=0x0
2727 tgcr=0x1, tmr=0xff1c, trr=0x3d09, tcr=0x0, tcn=0x169d, ter=0x0
2729 [q, b, e, ?] ...Stopping timer
2731 [q, b, e, ?] ## Application terminated, rc = 0x0
2737 Over time, many people have reported problems when trying to used the
2738 "minicom" terminal emulation program for serial download. I (wd)
2739 consider minicom to be broken, and recommend not to use it. Under
2740 Unix, I recommend to use C-Kermit for general purpose use (and
2741 especially for kermit binary protocol download ("loadb" command), and
2742 use "cu" for S-Record download ("loads" command).
2744 Nevertheless, if you absolutely want to use it try adding this
2745 configuration to your "File transfer protocols" section:
2747 Name Program Name U/D FullScr IO-Red. Multi
2748 X kermit /usr/bin/kermit -i -l %l -s Y U Y N N
2749 Y kermit /usr/bin/kermit -i -l %l -r N D Y N N
2755 Starting at version 0.9.2, U-Boot supports NetBSD both as host
2756 (build U-Boot) and target system (boots NetBSD/mpc8xx).
2758 Building requires a cross environment; it is known to work on
2759 NetBSD/i386 with the cross-powerpc-netbsd-1.3 package (you will also
2760 need gmake since the Makefiles are not compatible with BSD make).
2761 Note that the cross-powerpc package does not install include files;
2762 attempting to build U-Boot will fail because <machine/ansi.h> is
2763 missing. This file has to be installed and patched manually:
2765 # cd /usr/pkg/cross/powerpc-netbsd/include
2767 # ln -s powerpc machine
2768 # cp /usr/src/sys/arch/powerpc/include/ansi.h powerpc/ansi.h
2769 # ${EDIT} powerpc/ansi.h ## must remove __va_list, _BSD_VA_LIST
2771 Native builds *don't* work due to incompatibilities between native
2772 and U-Boot include files.
2774 Booting assumes that (the first part of) the image booted is a
2775 stage-2 loader which in turn loads and then invokes the kernel
2776 proper. Loader sources will eventually appear in the NetBSD source
2777 tree (probably in sys/arc/mpc8xx/stand/u-boot_stage2/); in the
2778 meantime, send mail to bruno@exet-ag.de and/or wd@denx.de for
2782 Implementation Internals:
2783 =========================
2785 The following is not intended to be a complete description of every
2786 implementation detail. However, it should help to understand the
2787 inner workings of U-Boot and make it easier to port it to custom
2791 Initial Stack, Global Data:
2792 ---------------------------
2794 The implementation of U-Boot is complicated by the fact that U-Boot
2795 starts running out of ROM (flash memory), usually without access to
2796 system RAM (because the memory controller is not initialized yet).
2797 This means that we don't have writable Data or BSS segments, and BSS
2798 is not initialized as zero. To be able to get a C environment working
2799 at all, we have to allocate at least a minimal stack. Implementation
2800 options for this are defined and restricted by the CPU used: Some CPU
2801 models provide on-chip memory (like the IMMR area on MPC8xx and
2802 MPC826x processors), on others (parts of) the data cache can be
2803 locked as (mis-) used as memory, etc.
2805 Chris Hallinan posted a good summy of these issues to the
2806 u-boot-users mailing list:
2808 Subject: RE: [U-Boot-Users] RE: More On Memory Bank x (nothingness)?
2809 From: "Chris Hallinan" <clh@net1plus.com>
2810 Date: Mon, 10 Feb 2003 16:43:46 -0500 (22:43 MET)
2813 Correct me if I'm wrong, folks, but the way I understand it
2814 is this: Using DCACHE as initial RAM for Stack, etc, does not
2815 require any physical RAM backing up the cache. The cleverness
2816 is that the cache is being used as a temporary supply of
2817 necessary storage before the SDRAM controller is setup. It's
2818 beyond the scope of this list to expain the details, but you
2819 can see how this works by studying the cache architecture and
2820 operation in the architecture and processor-specific manuals.
2822 OCM is On Chip Memory, which I believe the 405GP has 4K. It
2823 is another option for the system designer to use as an
2824 initial stack/ram area prior to SDRAM being available. Either
2825 option should work for you. Using CS 4 should be fine if your
2826 board designers haven't used it for something that would
2827 cause you grief during the initial boot! It is frequently not
2830 CFG_INIT_RAM_ADDR should be somewhere that won't interfere
2831 with your processor/board/system design. The default value
2832 you will find in any recent u-boot distribution in
2833 Walnut405.h should work for you. I'd set it to a value larger
2834 than your SDRAM module. If you have a 64MB SDRAM module, set
2835 it above 400_0000. Just make sure your board has no resources
2836 that are supposed to respond to that address! That code in
2837 start.S has been around a while and should work as is when
2838 you get the config right.
2843 It is essential to remember this, since it has some impact on the C
2844 code for the initialization procedures:
2846 * Initialized global data (data segment) is read-only. Do not attempt
2849 * Do not use any unitialized global data (or implicitely initialized
2850 as zero data - BSS segment) at all - this is undefined, initiali-
2851 zation is performed later (when relocationg to RAM).
2853 * Stack space is very limited. Avoid big data buffers or things like
2856 Having only the stack as writable memory limits means we cannot use
2857 normal global data to share information beween the code. But it
2858 turned out that the implementation of U-Boot can be greatly
2859 simplified by making a global data structure (gd_t) available to all
2860 functions. We could pass a pointer to this data as argument to _all_
2861 functions, but this would bloat the code. Instead we use a feature of
2862 the GCC compiler (Global Register Variables) to share the data: we
2863 place a pointer (gd) to the global data into a register which we
2864 reserve for this purpose.
2866 When chosing a register for such a purpose we are restricted by the
2867 relevant (E)ABI specifications for the current architecture, and by
2868 GCC's implementation.
2870 For PowerPC, the following registers have specific use:
2873 R3-R4: parameter passing and return values
2874 R5-R10: parameter passing
2875 R13: small data area pointer
2879 (U-Boot also uses R14 as internal GOT pointer.)
2881 ==> U-Boot will use R29 to hold a pointer to the global data
2883 Note: on PPC, we could use a static initializer (since the
2884 address of the global data structure is known at compile time),
2885 but it turned out that reserving a register results in somewhat
2886 smaller code - although the code savings are not that big (on
2887 average for all boards 752 bytes for the whole U-Boot image,
2888 624 text + 127 data).
2890 On ARM, the following registers are used:
2892 R0: function argument word/integer result
2893 R1-R3: function argument word
2895 R10: stack limit (used only if stack checking if enabled)
2896 R11: argument (frame) pointer
2897 R12: temporary workspace
2900 R15: program counter
2902 ==> U-Boot will use R8 to hold a pointer to the global data
2908 U-Boot runs in system state and uses physical addresses, i.e. the
2909 MMU is not used either for address mapping nor for memory protection.
2911 The available memory is mapped to fixed addresses using the memory
2912 controller. In this process, a contiguous block is formed for each
2913 memory type (Flash, SDRAM, SRAM), even when it consists of several
2914 physical memory banks.
2916 U-Boot is installed in the first 128 kB of the first Flash bank (on
2917 TQM8xxL modules this is the range 0x40000000 ... 0x4001FFFF). After
2918 booting and sizing and initializing DRAM, the code relocates itself
2919 to the upper end of DRAM. Immediately below the U-Boot code some
2920 memory is reserved for use by malloc() [see CFG_MALLOC_LEN
2921 configuration setting]. Below that, a structure with global Board
2922 Info data is placed, followed by the stack (growing downward).
2924 Additionally, some exception handler code is copied to the low 8 kB
2925 of DRAM (0x00000000 ... 0x00001FFF).
2927 So a typical memory configuration with 16 MB of DRAM could look like
2930 0x0000 0000 Exception Vector code
2933 0x0000 2000 Free for Application Use
2939 0x00FB FF20 Monitor Stack (Growing downward)
2940 0x00FB FFAC Board Info Data and permanent copy of global data
2941 0x00FC 0000 Malloc Arena
2944 0x00FE 0000 RAM Copy of Monitor Code
2945 ... eventually: LCD or video framebuffer
2946 ... eventually: pRAM (Protected RAM - unchanged by reset)
2947 0x00FF FFFF [End of RAM]
2950 System Initialization:
2951 ----------------------
2953 In the reset configuration, U-Boot starts at the reset entry point
2954 (on most PowerPC systens at address 0x00000100). Because of the reset
2955 configuration for CS0# this is a mirror of the onboard Flash memory.
2956 To be able to re-map memory U-Boot then jumps to it's link address.
2957 To be able to implement the initialization code in C, a (small!)
2958 initial stack is set up in the internal Dual Ported RAM (in case CPUs
2959 which provide such a feature like MPC8xx or MPC8260), or in a locked
2960 part of the data cache. After that, U-Boot initializes the CPU core,
2961 the caches and the SIU.
2963 Next, all (potentially) available memory banks are mapped using a
2964 preliminary mapping. For example, we put them on 512 MB boundaries
2965 (multiples of 0x20000000: SDRAM on 0x00000000 and 0x20000000, Flash
2966 on 0x40000000 and 0x60000000, SRAM on 0x80000000). Then UPM A is
2967 programmed for SDRAM access. Using the temporary configuration, a
2968 simple memory test is run that determines the size of the SDRAM
2971 When there is more than one SDRAM bank, and the banks are of
2972 different size, the larger is mapped first. For equal size, the first
2973 bank (CS2#) is mapped first. The first mapping is always for address
2974 0x00000000, with any additional banks following immediately to create
2975 contiguous memory starting from 0.
2977 Then, the monitor installs itself at the upper end of the SDRAM area
2978 and allocates memory for use by malloc() and for the global Board
2979 Info data; also, the exception vector code is copied to the low RAM
2980 pages, and the final stack is set up.
2982 Only after this relocation will you have a "normal" C environment;
2983 until that you are restricted in several ways, mostly because you are
2984 running from ROM, and because the code will have to be relocated to a
2988 U-Boot Porting Guide:
2989 ----------------------
2991 [Based on messages by Jerry Van Baren in the U-Boot-Users mailing
2995 int main (int argc, char *argv[])
2997 sighandler_t no_more_time;
2999 signal (SIGALRM, no_more_time);
3000 alarm (PROJECT_DEADLINE - toSec (3 * WEEK));
3002 if (available_money > available_manpower) {
3003 pay consultant to port U-Boot;
3007 Download latest U-Boot source;
3009 Subscribe to u-boot-users mailing list;
3012 email ("Hi, I am new to U-Boot, how do I get started?");
3016 Read the README file in the top level directory;
3017 Read http://www.denx.de/re/DPLG.html
3018 Read the source, Luke;
3021 if (available_money > toLocalCurrency ($2500)) {
3024 Add a lot of aggravation and time;
3027 Create your own board support subdirectory;
3029 Create your own board config file;
3033 Add / modify source code;
3037 email ("Hi, I am having problems...");
3039 Send patch file to Wolfgang;
3044 void no_more_time (int sig)
3053 All contributions to U-Boot should conform to the Linux kernel
3054 coding style; see the file "Documentation/CodingStyle" in your Linux
3055 kernel source directory.
3057 Please note that U-Boot is implemented in C (and to some small parts
3058 in Assembler); no C++ is used, so please do not use C++ style
3059 comments (//) in your code.
3061 Submissions which do not conform to the standards may be returned
3062 with a request to reformat the changes.
3068 Since the number of patches for U-Boot is growing, we need to
3069 establish some rules. Submissions which do not conform to these rules
3070 may be rejected, even when they contain important and valuable stuff.
3073 When you send a patch, please include the following information with
3076 * For bug fixes: a description of the bug and how your patch fixes
3077 this bug. Please try to include a way of demonstrating that the
3078 patch actually fixes something.
3080 * For new features: a description of the feature and your
3083 * A CHANGELOG entry as plaintext (separate from the patch)
3085 * For major contributions, your entry to the CREDITS file
3087 * When you add support for a new board, don't forget to add this
3088 board to the MAKEALL script, too.
3090 * If your patch adds new configuration options, don't forget to
3091 document these in the README file.
3093 * The patch itself. If you are accessing the CVS repository use "cvs
3094 update; cvs diff -puRN"; else, use "diff -purN OLD NEW". If your
3095 version of diff does not support these options, then get the latest
3096 version of GNU diff.
3098 The current directory when running this command shall be the top
3099 level directory of the U-Boot source tree, or it's parent directory
3100 (i. e. please make sure that your patch includes sufficient
3101 directory information for the affected files).
3103 We accept patches as plain text, MIME attachments or as uuencoded
3106 * If one logical set of modifications affects or creates several
3107 files, all these changes shall be submitted in a SINGLE patch file.
3109 * Changesets that contain different, unrelated modifications shall be
3110 submitted as SEPARATE patches, one patch per changeset.
3115 * Before sending the patch, run the MAKEALL script on your patched
3116 source tree and make sure that no errors or warnings are reported
3117 for any of the boards.
3119 * Keep your modifications to the necessary minimum: A patch
3120 containing several unrelated changes or arbitrary reformats will be
3121 returned with a request to re-formatting / split it.
3123 * If you modify existing code, make sure that your new code does not
3124 add to the memory footprint of the code ;-) Small is beautiful!
3125 When adding new features, these should compile conditionally only
3126 (using #ifdef), and the resulting code with the new feature
3127 disabled must not need more memory than the old code without your