April 2005 - U-Boot

[U-Boot-Users] Patch: allow a fixed port for TFTP (take 3)
by Jerry Van Baren 24 Sep '05

24 Sep '05

Hi Wolfgang, try three on the patch attached. The change here is to rename the env variable per Tolunay Orkun's recommendation. * Patch by Jerry Van Baren 11 January 2005 - Add a configuration option CONFIG_TFTP_PORT and optional env variable tftpsourceport to force TFTP to use a fixed UDP port. gvb Index: README =================================================================== RCS file: /cvsroot/u-boot/u-boot/README,v retrieving revision 1.117 diff -p -u -r1.117 README --- README 9 Jan 2005 23:16:42 -0000 1.117 +++ README 10 Jan 2005 04:13:50 -0000 @@ -1480,6 +1480,22 @@ The following options need to be configu When SystemACE support is added, the "ace" device type becomes available to the fat commands, i.e. fatls. +- TFTP Fixed UDP Port: + CONFIG_TFTP_PORT + + If this is defined, the environment variable tftpsourceport + is used to supply the TFTP UDP source port value. + If tftpsourceport isn't defined, the normal pseudo-random port + number generator is used. + + The purpose for this is to allow a TFTP server to + blindly start the TFTP transfer using the pre-configured + target IP address and UDP port. This has the effect of + "punching through" the (Windows XP) firewall, allowing + the remainder of the TFTP transfer to proceed normally. + A better solution is to properly configure the firewall, + but sometimes that is not allowed. + - Show boot progress: CONFIG_SHOW_BOOT_PROGRESS @@ -2432,6 +2448,9 @@ Some configuration options can be set us Useful on scripts which control the retry operation themselves. + tftpsourceport - If this is set, the value is used for TFTP's + UDP source port. + vlan - When set to a value < 4095 the traffic over ethernet is encapsulated/received over 802.1q VLAN tagged frames. Index: net/tftp.c =================================================================== RCS file: /cvsroot/u-boot/u-boot/net/tftp.c,v retrieving revision 1.9 diff -p -u -r1.9 tftp.c --- net/tftp.c 15 Apr 2004 21:48:55 -0000 1.9 +++ net/tftp.c 10 Jan 2005 04:13:51 -0000 @@ -301,6 +301,10 @@ TftpTimeout (void) void TftpStart (void) { +#ifdef CONFIG_TFTP_PORT + char *ep; /* Environment pointer */ +#endif + if (BootFile[0] == '\0') { IPaddr_t OurIP = ntohl(NetOurIP); @@ -354,7 +358,13 @@ TftpStart (void) TftpServerPort = WELL_KNOWN_PORT; TftpTimeoutCount = 0; TftpState = STATE_RRQ; + /* Use a pseudo-random port unless a specific port is set */ TftpOurPort = 1024 + (get_timer(0) % 3072); +#ifdef CONFIG_TFTP_PORT + if ((ep = getenv("tftpsourceport")) != NULL) { + TftpOurPort= simple_strtol(ep, NULL, 10); + } +#endif TftpBlock = 0; /* zero out server ether in case the server ip has changed */

3 10

[U-Boot-Users] Patch: allow a fixed port for TFTP
by Jerry Van Baren 24 Sep '05

24 Sep '05

Hi Wolfgang, try two on the patch attached. * Patch by Jerry Van Baren 10 January 2005 - Add a configuration option CONFIG_TFTP_PORT and optional env variable tftpport to force TFTP to use a fixed UDP port. gvb Index: README =================================================================== RCS file: /cvsroot/u-boot/u-boot/README,v retrieving revision 1.117 diff -p -u -r1.117 README --- README 9 Jan 2005 23:16:42 -0000 1.117 +++ README 10 Jan 2005 04:13:50 -0000 @@ -1480,6 +1480,22 @@ The following options need to be configu When SystemACE support is added, the "ace" device type becomes available to the fat commands, i.e. fatls. +- TFTP Fixed UDP Port: + CONFIG_TFTP_PORT + + If this is defined, the environment variable tftpport + is used to supply the TFTP UDP source port value. + If tftpport isn't defined, the normal pseudo-random port + number generator is used. + + The purpose for this is to allow a TFTP server to + blindly start the TFTP transfer using the pre-configured + target IP address and UDP port. This has the effect of + "punching through" the (Windows XP) firewall, allowing + the remainder of the TFTP transfer to proceed normally. + A better solution is to properly configure the firewall, + but sometimes that is not allowed. + - Show boot progress: CONFIG_SHOW_BOOT_PROGRESS @@ -2432,6 +2448,9 @@ Some configuration options can be set us Useful on scripts which control the retry operation themselves. + tftpport - If this is set, the value is used for TFTP's + UDP source port. + vlan - When set to a value < 4095 the traffic over ethernet is encapsulated/received over 802.1q VLAN tagged frames. Index: net/tftp.c =================================================================== RCS file: /cvsroot/u-boot/u-boot/net/tftp.c,v retrieving revision 1.9 diff -p -u -r1.9 tftp.c --- net/tftp.c 15 Apr 2004 21:48:55 -0000 1.9 +++ net/tftp.c 10 Jan 2005 04:13:51 -0000 @@ -301,6 +301,10 @@ TftpTimeout (void) void TftpStart (void) { +#ifdef CONFIG_TFTP_PORT + char *ep; /* Environment pointer */ +#endif + if (BootFile[0] == '\0') { IPaddr_t OurIP = ntohl(NetOurIP); @@ -354,7 +358,13 @@ TftpStart (void) TftpServerPort = WELL_KNOWN_PORT; TftpTimeoutCount = 0; TftpState = STATE_RRQ; + /* Use a pseudo-random port unless a specific port is set */ TftpOurPort = 1024 + (get_timer(0) % 3072); +#ifdef CONFIG_TFTP_PORT + if ((ep = getenv("tftpport")) != NULL) { + TftpOurPort= simple_strtol(ep, NULL, 10); + } +#endif TftpBlock = 0; /* zero out server ether in case the server ip has changed */

3 4

[U-Boot-Users] dbau1550 patch
by Andrew Dyer 24 Sep '05

24 Sep '05

A while back (Sep 23, 2004) someone named Leif posted a note about a patch to add support for the AMD DbAU1550 board to u-boot here https://sourceforge.net/mailarchive/forum.php?thread_id=5624545&forum_id=12… I can't seem to find the patch he referred to and I would like to get a hold of it. -- Hardware, n.: The parts of a computer system that can be kicked.

4 6

[U-Boot-Users] common/cmd_mii.c bug and patch
by Andrew Dyer 24 Sep '05

24 Sep '05

The mii info command (the !CONFIG_TERSE_MII version) has a bug in current cvs. If the phy address is passed to the command the code assigns the variable addr to start, instead of the values of addrlo which is extracted from the command line. At that point in the code addr is always 0. to demonstrate find a board with a phy at a non-zero address, compare the results of 'mii info' and mii info 1' (substitute the phy address on the test board for 1). Also in this file on lines 55 and 418 there are two places where a call to mii_init() is conditionally compiled in, but one condition includes defined(CONFIG_MCF52x2) and the other doesn't. Is this intentional? -- Hardware, n.: The parts of a computer system that can be kicked.

2 1

[U-Boot-Users] PATCH: Fix ppc 440 timer freq.
by Travis B. Sawyer 15 Aug '05

15 Aug '05

This patch fixes 440 timer frequency (move from 10ms to 1 ms), as found by Andrew Wozniak. Changelog * Patch by Travis Sawyer, 05 Apr 2005: - Change timer frequency for ppc 440 from 10 ms to 1 ms. Problem found by Andrew Wozniak. Andy, thank you for pointing this out. Best Regards, Travis diff -purN u-boot/cpu/ppc4xx/interrupts.c u-boot_modified/cpu/ppc4xx/interrupts.c --- u-boot/cpu/ppc4xx/interrupts.c 2004-03-14 10:20:56.000000000 -0500 +++ u-boot_modified/cpu/ppc4xx/interrupts.c 2005-04-05 19:56:27.793079640 -0400 @@ -134,7 +134,7 @@ int interrupt_init_cpu (unsigned *decrem mtspr( dec, 0 ); /* Prevent exception after TSR clear*/ mtspr( decar, 0 ); /* clear reload */ mtspr( tsr, 0x08000000 ); /* clear DEC status */ - val = gd->bd->bi_intfreq/100; /* 10 msec */ + val = gd->bd->bi_intfreq/1000; /* 1 msec */ mtspr( decar, val ); /* Set auto-reload value */ mtspr( dec, val ); /* Set inital val */ #else

2 1

[U-Boot-Users] [PATCH] support for CF/IDE on PCMCIA for PXA25X
by Christian Pell 13 Aug '05

13 Aug '05

Hi, I tried this with some different CFs with FAT file-systems on them. There is a README in the patch itself that goes in documentation directory. Bye, -- Christian Pellegrin <chri(a)ascensit.com>, <c.pellegrin(a)eurotech.it>, <chri(a)infis.univ.ts.it>, <c.pellegrin(a)exadron.com>

4 8

[U-Boot-Users] Patches for ARM IntegratorCP, ARM926EJS, and ARM946ES
by Shawn Jin 12 Aug '05

12 Aug '05

Hi, The patches were developed on an ARM IntegratorCP board with ARM946ES. Since the majority code of ARM946ES cpu was copied from ARM926EJS and there was bug about CFG_HZ found in ARM946ES later, I also fixed this CFG_HZ error on ARM926EJS. The major changes are 1. Mulitple core modules support (u-boot.multicm.patch) Following OMAP boards' methodology, I added mulitple core module configuration. For example, integratorcp_946es_config is for IntegratorCP board with ARM946ES, integratorcp_config is for IntegratorCP board with ARM926EJS by default, integratorap_946es_config is for IntegratorAP board with ARM946ES, and integratorap_config is for IntegratorAP with ARM926EJS. Now only two core modules are supported, ARM946ES and ARM926EJS. 2. ARM946ES support (u-boot.arm946es.patch) 3. ARM926EJS bug fix (u-boot.arm926ejs.patch) Fixed CFG_HZ problem. Set CFG_HZ to 1000. The solution is from AT91RM9200. 4. IntegratorCP bug fixes (u-boot.integratorcp.patch) Added SDRAM size information and FLASH support. Philippe, I don't have an ARM926EJS core module. Would you please test the CFG_HZ fix on your system? Thanks. Regards, -Shawn.

3 6

[U-Boot-Users] Minor cleanup
by Anders Larsen 12 Aug '05

12 Aug '05

Hi, Rick Bronson and the AT91RM9200DK is listed under PPC Systems. Since the AT91RM9200 is an ARM SoC, the attached patch against current top of CVS moves the entry to the appropriate position. Changelog: Move the AT91RM9200DK to the ARM Systems list. Cheers Anders

2 1

[U-Boot-Users] Eliminate calls of _divsi3 on ARM
by Anders Larsen 12 Aug '05

12 Aug '05

Hi, the enclosed patch agains current top of CVS eliminates two calls of the libgcc helper function _divsi3 and _modsi3. Rationale: Certain commercially available tool-chains provides b0rken libgcc.a (compiled with hardware FP) making these helper functions unusable (U-Boot already provides the unsigned counterparts in lib_arm/libarm.a). Changelog: Eliminate calls of ARM libgcc.a helper functions _divsi3 and _modsi3 Cheers Anders

3 4

[U-Boot-Users] UBoot cvs compile error.
by Bryan O'Donoghue 12 Aug '05

12 Aug '05

make Adder_config edit config.mk uncomment -DDEBUG fails to resolve measure_gclk Changing cpu/mpc8xx/speed.c #if !defined(CONFIG_8xx_CPUCLK_DEFAULT) || defined(CFG_MEASURE_CPUCLK) to #if !defined(CONFIG_8xx_CPUCLK_DEFAULT) || defined(CFG_MEASURE_CPUCLK)\ || defined(DEBUG) fixes. -- Best regards Bryan

2 1

[U-Boot-Users] Switch between PlanetCore and U-Boot freely
by Sam Song 08 Aug '05

08 Aug '05

Hi all, Well, somebody was trying to pull the cat out of my bag. Here it is:-) ------------------------------------------------- After several heart-struck failure, I got one workable way to program each other in FLASH between PlanetCore and U-Boot. Hardware Platform : RPXlite DW(EP 823 H1 DW) 1. From U-Boot to PlanetCore Utilities : PlanetCore Boot Loader - PCL200.mot [root@sam tftpboot]# ppc_8xx-objcopy -O ppcboot PCL200.mot pcl200.bin [Target Operation] u-boot>t 100000 pcl200.bin u-boot>go 0x100000 ## Starting application at 0x00100000 ... MPC8xx PlanetCore Flash Burner v2.00 Copyright 2001 Embedded Planet. All rights reserved. Construct Flash Device.....done. Program MPC8xx PlanetCore Boot Loader v2.00 Built Sep 19, 2001 at 14:34:42 Image located from FC000000 to FC01B5D1. (Skipping an image, only loading low boot image) Low boot board detected, skipping high boot image. Erasing, programming and verifying will start in 20 seconds Press P to start immediately or ESC to cancel Press Space or Enter for more options. .............. Erasing Programming FLASH programmed successfully! Press R to induce a hard reset MPC8xx PlanetCore Boot Loader v2.00 Copyright 2001 Embedded Planet. All rights reserved. DRAM available size = 64 MB wvCV DRAM OK > 2. From PlanetCore to U-Boot Utilities : PlanetCore FLASH Burner - PCB200.mot Use Flash Burner to finish the work: First, TFTP the U-Boot image file to RAM; For example, RPXlite_DW.bin to 0x400000 Second, TFTP FLASH Burner to RAM; For example, 0x100000 Third, run the FLASH Burner and Program the U-Boot image into the correct location in FLASH. [Target Operation] MPC8xx PlanetCore Boot Loader v2.00 Copyright 2001 Embedded Planet. All rights reserved. DRAM available size = 64 MB wvCV DRAM OK >t Load using tftp via Ethernet Enter server IP address <172.16.115.6> : Enter server filename <PCL200.mot> : RPXlite_DW.bin Enter (B)inary or (S)record input mode <S> : B Enter address offset : <00400000 hex> : Total bytes = 120096 in 232184 uSecs Loaded addresses 00400000 through 0041D51F. Start address = 00400000 >t Load using tftp via Ethernet Enter server IP address <172.16.115.6> : Enter server filename <RPXlite_DW.bin> : PCB200.mot Enter (B)inary or (S)record input mode <B> : S Enter address offset : <00000000 hex> : .512.1024..2048....4096..... Total bytes = 326280 in 2570249 uSecs Loaded addresses 00100000 through 0011BB51. Start address = 00100000 >go [Go 00100000] MPC8xx PlanetCore Flash Burner v2.00 Copyright 2001 Embedded Planet. All rights reserved. Construct Flash Device.....done. Bad start address Start = 0xFFFFFFFF, target = 0xFFFFFFFF, length = 0xFFFFFFFF Forcing Menu Interface h[elp] Show commands. c[ode] Show information on code to be loaded. di[splay] Display all flash sections. du[mp] Dump memory. d ? for more info. e[rase] Erase flash sections. f[ill] Fill flash sections. im[age] Toggle load high, low, or both flash images. in[fo] Show flash information. ma[p] Show memory map. mo[dify] Modify memory. m ? for more info. p[rogram] Erase, program, and verify now. reset Restart the loader. s[how] Show flash sections to erase and program. t[est] Test flash sections. q[uit] Quit without programming. #program 400000 ff000000 1D51F doProgram( 400000 ff000000 1D51F ) Start = 0x00400000, target = 0xFF000000, length = 0x0001D51F Erasing sector 0xFF000000, length 0x008000. Erasing sector 0xFF008000, length 0x008000. Erasing sector 0xFF010000, length 0x008000. Erasing sector 0xFF018000, length 0x008000. Programming FF000000 through FF01D51E FLASH programmed successfully! Press R to induce a hard reset Forcing Hard Reset by MachineCheck and ResetOnCheckstop... U-Boot 1.1.2 (Aug 29 2004 - 15:11:27) CPU: PPC823EZTnnB2 at 48 MHz: 16 kB I-Cache 8 kB D-Cache Board: RPXlite_DW DRAM: 64 MB FLASH: 16 MB *** Warning - bad CRC, using default environment In: serial Out: serial Err: serial Net: SCC ETHERNET u-boot> ------------------------------------------------- Well, sometimes network function of PlanetCore couldn't work when switching from U-Boot to PlanetCore. For example, you couldn't download a file from HOST PC via TFTP. Don't worry, just restart your HOST PC and everything would work as smooth as clockwork. I don't know the reason WHY:-) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Merry Christmas and Happy New Year! ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ===== Best regards, Sam _________________________________________________________ Do You Yahoo!? 150万曲MP3疯狂搜，带您闯入音乐殿堂 http://music.yisou.com/ 美女明星应有尽有，搜遍美图、艳图和酷图 http://image.yisou.com 1G就是1000兆，雅虎电邮自助扩容！ http://cn.rd.yahoo.com/mail_cn/tag/1g/*http://cn.mail.yahoo.com/event/mail_…

2 3

[U-Boot-Users] [patch]fixes for cmd_fpga.c
by Andrew Dyer 08 Aug '05

08 Aug '05

The attached patch: 1) fixes a bug with the 'fpga loadb' command The bug is in calculating the config data size - 4 bytes of length info were being read as signed chars, casted to long, shifted and added. The cast to long was doing sign extension and causing an incorrect result if a byte was greater than 0x80. For me this caused the fpga load to end early and time out waiting for the hardware completion. 2) removes compilation warnings about using __FUNCTION__ concatenated with strings 3) changes the fpga loadb to always print the file header info during a load. I realize this patch is a mishmash - hopefully it's helpful to someone. -- Hardware, n.: The parts of a computer system that can be kicked.

2 1

[U-Boot-Users] [PATCH] Fix SystemACE CF access errors
by Jeff Angielski 07 Aug '05

07 Aug '05

Attached is a patch that fixes the configuration controller errors that sometimes occur when accessing the SystemACE CF. The patch also fixes the locking routines so that they do not stomp on the other bits in the controller register. Finally, the patch will auto detect the partition type to prevent "Partition x not valid on device y" error messages. CHANGELONG: * Patch by Jeff Angielski 09 Jan 2005 Fix errors that occur when accessing SystemACE CF.

2 1

[U-Boot-Users] [PATCH] fix PowerQUICC II masks detection
by Eugene Surovegin 06 Aug '05

06 Aug '05

Hi! This trivial patch fixes typos in cpu/mpc8260/cpu.c:checkcpu(). * Patch by Eugene Surovegin, 20 Dec 2004: Fix PowerQUICC II mask detection.

2 1

[U-Boot-Users] Patch: Add support for Embedded Planet EP8248 board
by Yuli Barcohen 06 Aug '05

06 Aug '05

Hi, The attached patch adds support for Embedded Planet EP8248 board. It was tested on EP8248E revision 1.0. CHANGELOG: Add support for Embedded Planet EP8248 board -- ======================================================================== Yuli Barcohen | Phone +972-9-765-1788 | Software Project Leader yuli(a)arabellasw.com | Fax +972-9-765-7494 | Arabella Software, Israel ========================================================================

2 1

[U-Boot-Users] [PATCH] BUGFIX: serial setting for kgdb
by HoJin 06 Aug '05

06 Aug '05

Hi, KGDB_SMC_INDEX must be used as the index for brg_map, but it is used direclty. (compared with function serial_init) So it does not work properly. I have tested it and it works fine. I noticed the same problem with 'serial_scc.c', but I could not test it yet. Patch for only for 'serial_smc.c' is attached. Index: cpu/mpc8260/serial_smc.c =================================================================== --- cpu/mpc8260/serial_smc.c (revision 65) +++ cpu/mpc8260/serial_smc.c (working copy) @@ -360,10 +360,10 @@ /* Set up the baud rate generator. */ #if defined(CONFIG_KGDB_USE_EXTC) - m8260_cpm_extcbrg(KGDB_SMC_INDEX, speed, + m8260_cpm_extcbrg(brg_map[KGDB_SMC_INDEX], speed, CONFIG_KGDB_EXTC_RATE, CONFIG_KGDB_EXTC_PINSEL); #else - m8260_cpm_setbrg(KGDB_SMC_INDEX, speed); + m8260_cpm_setbrg(brg_map[KGDB_SMC_INDEX], speed); #endif /* Make the first buffer the only buffer. -- Hojin

2 1

[U-Boot-Users] [PATCH] MII help text formatting fix
by Cory Tusar 06 Aug '05

06 Aug '05

Attached is a patch fixing a minor glitch in the 'mii help' text formatting. Picky, picky, I know... Patch was generated against current CVS sources. CHANGELOG: * Patch by Cory Tusar, 10 Dec 2004: Fixed 'mii help' text formatting. -Cory -- Cory T. Tusar Embedded Systems Engineer Videon Central, Inc. 2171 Sandy Drive State College, PA 16801 (814) 235-1111 x316 (814) 235-1118 fax "Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it." --Brian W. Kernighan Index: common/cmd_mii.c =================================================================== RCS file: /home/cvs/u-boot/common/cmd_mii.c,v retrieving revision 1.1.1.1 retrieving revision 1.2 diff -u -r1.1.1.1 -r1.2 --- common/cmd_mii.c 7 Dec 2004 14:49:03 -0000 1.1.1.1 +++ common/cmd_mii.c 8 Dec 2004 17:42:39 -0000 1.2 @@ -549,7 +549,7 @@ U_BOOT_CMD( mii, 5, 1, do_mii, "mii - MII utility commands\n", - "info <addr> - display MII PHY info\n" + "info <addr> - display MII PHY info\n" "mii read <addr> <reg> - read MII PHY <addr> register <reg>\n" "mii write <addr> <reg> <data> - write MII PHY <addr> register <reg>\n" "mii dump <addr> <reg> - pretty-print <addr> <reg> (0-5 only)\n"

2 1

[U-Boot-Users] [Patch] bugfix:return value of NFS command
by Hiroshi Ito 06 Aug '05

06 Aug '05

Hello, 1.nfs command return code is depend on umount-all, but not state of downloading. 2.When fail to write flash, it doesn't change state to umount example1: ---nfs server--- [mldbox]# ls -l u-boot.bin ---------- 1 root root 204048 Dec 11 02:06 u-boot.bin --- u-boot --- # if nfs 80400000 192.168.3.98:/usr/src/mldbox/u-boot.bin; then echo ok; else echo error; fi Trying i82559#0 Using i82559#0 device File transfer via NFS from server 192.168.3.98; our IP address is 192.168.3.200 Filename '/usr/src/mldbox/u-boot.bin'. Load address: 0x80400000 Loading: * done ok <= Wrong!! example2: [root@tv mldbox]# rm u-boot.bin rm: remove regular file `u-boot.bin'? y [root@tv mldbox]# # if nfs 80400000 192.168.3.98:/usr/src/mldbox/u-boot.bin; then echo ok; else echo error; fi Trying i82559#0 Using i82559#0 device File transfer via NFS from server 192.168.3.98; our IP address is 192.168.3.200 Filename '/usr/src/mldbox/u-boot.bin'. Load address: 0x80400000 Loading: *** ERROR: File lookup fail done ok <= Wrong!! # attached patch fixes these problems. patch is for current CVS. -------- Hiroshi Ito Media Lab. Inc., URL http://www.mlb.co.jp ( Sorry, Japanese only. ) TEL +81-3-5294-7255 FAX +81-3-5294-7256

3 2

[U-Boot-Users] SCSI support for MIB405 board
by Wolfgang Denk 04 Aug '05

04 Aug '05

Dear Denis, can you tell me something about the SCSI support in the current versions of U-Boot? It seems the MIB405 board (maintained by you) has SCSI support enabled, but the whole code in common/cmd_scsi.c seemd to be unused - for example, there is no caller of do_scsi() or do_scsiboot() anywhere in the whole source tree? Best regards, Wolfgang Denk -- Software Engineering: Embedded and Realtime Systems, Embedded Linux Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd(a)denx.de In general, they do what you want, unless you want consistency. - Larry Wall in the perl man page

2 4

[U-Boot-Users] [Patch] default filename of tftp and nfs
by Hiroshi Ito 04 Aug '05

04 Aug '05

Hello, default filename of tftp and nfs has different value between little endian machine and big one. This patch fixes endian problem. Patch is for current cvs. -------- Hiroshi Ito Media Lab. Inc., URL http://www.mlb.co.jp ( Sorry, Japanese only. ) TEL +81-3-5294-7255 FAX +81-3-5294-7256

2 2

[U-Boot-Users] u-boot 1.1.1 and USB mass storage devices
by Shannon Holland 03 Aug '05

03 Aug '05

Hello, We are using u-boot on a Samsung S3C2410 based board and have USB mostly working using the arm920t USB driver. I say mostly working as some mass storage devices work fine, but most do not. Typically the behavior we see when they don't work is a "usb reset" reports the correct number of devices, a "usb tree" will return the correct device tree but "usb scan" will typically return a lot of CLI:TIMEOUT errors and seemingly never return. Sometimes it is able to find the storage device, but reading from it (it has a FAT16 filesystem) will return similar errors. I'm curious if these are known problems or likely to be something in the way we have (or have not) set up usb? Might anyone have some recommendations as to where to start looking? Thanks for any help! Shannon

5 8

[U-Boot-Users] Loading FPGA from an mkimage created image?
by Steven Scholz 10 Jun '05

10 Jun '05

Hi there, I know that fpga load [device number] [image address] [image size] can be used to init an FPGA. Now I want to compress and wrap the FPGA data with mkimage. Is there a way to pass the address of an Image to "fpga load". Or do I have to umcompress it first using "bootm" or something and then pass the address and size to "fpga load"? Thanks! -- Steven Scholz

2 7

[U-Boot-Users] mkimage for DOS
by Dave Westwood 10 Jun '05

10 Jun '05

long shot really, but does anyone have a working executable mkimage for DOS or a dos window in windows (don't flame me for using windows !!)

2 1

Re: [U-Boot-Users] Support for Samsung 1Giga bit NAND device
by Laurent Mohin 01 Jun '05

01 Jun '05

Andy, U-boot 1.1.0 doesn't provide support for NAND flash with a page size of 2K-byte instead of 512 byte (and you're right, theses devices require special handling). Pantelis Antoniou submitted a patch on 2004-04-30 which seems to manage these new chips, but I don't think it has been added to the latest CVS tree. By the way, Samsung also offer 1 Giga bit devices with a 512 byte page size (K9K1G08UOM - Manufacturer ID : 0xEC, Chip ID: 0x79). Hope this helps, Laurent Andy Hawkins <a.hawkins(a)cabletime.com> Sent by: u-boot-users-admin(a)lists.sourceforge.net 14/06/04 13:40 To: u-boot-users(a)lists.sourceforge.net cc: Subject: [U-Boot-Users] Support for Samsung 1Giga bit NAND device Hi, Can anyone tell me if U-Boot has support for the Samsung 1 Giga Bit NAND devices (Manufacturer ID: 0xec, Chip ID: 0xf1)? I understand these devices have an unusually large block size, that needs special handling. Is this correct? Thanks. Andy ------------------------------------------------------- This SF.Net email is sponsored by the new InstallShield X. >From Windows to Linux, servers to mobile, InstallShield X is the one installation-authoring solution that does it all. Learn more and evaluate today! http://www.installshield.com/Dev2Dev/0504 _______________________________________________ U-Boot-Users mailing list U-Boot-Users(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/u-boot-users

4 8

RE: [U-Boot-Users] Re: [PATCH] !CFG_MEMTEST_SCRATCH - do not dereference NULL ptr
by VanBaren, Gerald (AGRE) 30 May '05

30 May '05

> -----Original Message----- > From: u-boot-users-admin(a)lists.sourceforge.net > [mailto:u-boot-users-admin@lists.sourceforge.net] On Behalf > Of Ladislav Michl > Sent: Thursday, November 25, 2004 6:22 AM > To: u-boot-users(a)lists.sourceforge.net > Subject: [U-Boot-Users] Re: [PATCH] !CFG_MEMTEST_SCRATCH - do > not dereference NULL ptr > > On Mon, Nov 22, 2004 at 12:07:18PM +0100, Wolfgang Denk wrote: > > There is no explanation what it does or which problem it is > supposed > > to fix. > > When CFG_MEMTEST_SCRATCH is undefined alternate memory test > in do_mem_mtest (with CFG_ALT_MEMTEST defined) dereferences > null pointer. It defines: > vu_long *dummy = NULL; > and later does: > *dummy = ~val; > > > There is also no CHANGELOG entry. > > CHANGELOG > * Patch by Ladislav Michl, 22 November 2004 > - Fix NULL pointer dereference in alternate memory test > (CFG_ALT_MEMTEST) > when if no CFG_MEMTEST_SCRATCH area defined > > Index: common/cmd_mem.c > =================================================================== > RCS file: /cvsroot/u-boot/u-boot/common/cmd_mem.c,v > retrieving revision 1.19 > diff -p -u -r1.19 cmd_mem.c > --- common/cmd_mem.c 10 Oct 2004 23:27:33 -0000 1.19 > +++ common/cmd_mem.c 22 Nov 2004 11:21:43 -0000 > @@ -646,8 +646,9 @@ int do_mem_mtest (cmd_tbl_t *cmdtp, int > vu_long num_words; > #if defined(CFG_MEMTEST_SCRATCH) > vu_long *dummy = (vu_long*)CFG_MEMTEST_SCRATCH; > +#define write_dummy(val) do { *dummy = ~val; } while (0) > #else > - vu_long *dummy = NULL; > +#define write_dummy(val) do { } while (0) > #endif > int j; > int iterations = 1; > @@ -723,7 +724,7 @@ int do_mem_mtest (cmd_tbl_t *cmdtp, int > val = bitpattern[j]; > for(; val != 0; val <<= 1) { > *addr = val; > - *dummy = ~val; /* clear the test data > off of the bus */ > + write_dummy(~val); /* clear the test > data off of the bus */ > readback = *addr; > if(readback != val) { > printf ("FAILURE (data line): " > @@ -731,11 +732,11 @@ int do_mem_mtest (cmd_tbl_t *cmdtp, int > val, readback); > } > *addr = ~val; > - *dummy = val; > + write_dummy(val); > readback = *addr; > if(readback != ~val) { > printf ("FAILURE (data line): " > - "Is %08lx, should be %08lx\n", > + "is %08lx, should be %08lx\n", > readback, ~val); > } > } Ladislav: You are fixing the problem in the wrong way. You should point "dummy" at a location in RAM that is writable rather than "fixing" the code by suppressing the dummy write. Being pedantic here, dummy isn't NULL, it is a pointer to the RAM location 0x00000000 (which just happens to be the same as NULL ;-). Strictly speaking, the assignment should have been written "vu_long *dummy = 0x00000000;". The problem here is that the code assumes 0x00000000 is writable and apparently this isn't the case on your board. It obviously is a good assumption on the other boards supported by u-boot (at least the ones that use this POST code :-). The proper fix is, for your board, change the assignment "vu_long *dummy = 0x00000000;" to point to writable RAM on your board. Suppressing the write to 0x00000000 will make the memory test ineffective on all boards (bad, really bad). The purpose of the "*dummy = ~val" statement is to discharge the data bus lines. The problem this is addressing is when the previous line "*addr = val" is executed, the data bus has on it the value "val". Since there is inherent capacitance on the bus lines, that value will persist (float) on the bus for a finite time, longer than it takes for a processor to do the "readback = *addr". The result is that, if the memory doesn't respond and the bus drivers remain tristated, you will think the memory responded properly because you read the residual charge on the bus. The write of "~val" to dummy is a throw-away operation whose purpose is to put something other than "val" on the bus so that a read of the target location won't give a false positive if the memory isn't working properly. Where "dummy" points is immaterial as long as "~val" gets out on the data bus. Normally you would point it in your RAM space, but you actually could point it anywhere that causes a valid write cycle on the data bus. Regards, gvb ****************************************** The information contained in, or attached to, this e-mail, may contain confidential information and is intended solely for the use of the individual or entity to whom they are addressed and may be subject to legal privilege. If you have received this e-mail in error you should notify the sender immediately by reply e-mail, delete the message from your system and notify your system manager. Please do not copy it for any purpose, or disclose its contents to any other person. The views or opinions presented in this e-mail are solely those of the author and do not necessarily represent those of the company. The recipient should check this e-mail and any attachments for the presence of viruses. The company accepts no liability for any damage caused, directly or indirectly, by any virus transmitted in this email. ******************************************

4 4