August 2012 - U-Boot

[U-Boot] [PATCH 0/17]: Improve am33xx support
by Tom Rini 02 Aug '12

02 Aug '12

Hey all, The following series of patches improves am33xx support, and cleans up omap3/4/5 slightly. The slight cleanup to omap3/4/5 is that we can all share a single function to see if we are executing in SDRAM or not. The rest of the series cleans up the EMIF code for am33xx. While I had hoped to be able to re-use the omap4/5 code as the EMIF unit is the same block, it's simply wired up too differently to reuse the same codebase. We can however reuse the register structs so some of the commits here are about changing to use that. There's also bugfixes to the DDR2 init sequence and cleanups that will make adding DDR3 support easier. I also update the am335x_evm config to be like a "normal" board and support common commands and booting Linux. -- Tom

2 25

[U-Boot] [PATCH 1/2] GPIO: Extend the cmd_gpio API by gpio_{from, to}_string()
by Marek Vasut 02 Aug '12

02 Aug '12

The gpio_from_string() call shall parse the incoming GPIO name taken from the command line and return the GPIO number used within U-Boot or return -1 on error. The gpio_to_string() on the other hand allows nicer reporting of GPIO name in the output of cmd_gpio. Signed-off-by: Marek Vasut <marex(a)denx.de> Cc: Detlev Zundel <dzu(a)denx.de> Cc: Mike Frysinger <vapier(a)gentoo.org> Cc: Stefano Babic <sbabic(a)denx.de> --- common/cmd_gpio.c | 27 +++++++++++++++++++++++---- 1 file changed, 23 insertions(+), 4 deletions(-) diff --git a/common/cmd_gpio.c b/common/cmd_gpio.c index 47eee89..12870ef 100644 --- a/common/cmd_gpio.c +++ b/common/cmd_gpio.c @@ -15,6 +15,22 @@ #define name_to_gpio(name) simple_strtoul(name, NULL, 10) #endif +int __gpio_from_string(const char *name) +{ + return name_to_gpio(name); +} + +int gpio_from_string(const char *name) + __attribute__((weak, alias("__gpio_from_string"))); + +int __gpio_to_string(int gpio, char *buf, int buflen) +{ + return 0; +} + +int gpio_to_string(int gpio, char *buf, int buflen) + __attribute__((weak, alias("__gpio_from_string"))); + enum gpio_cmd { GPIO_INPUT, GPIO_SET, @@ -27,7 +43,9 @@ static int do_gpio(cmd_tbl_t *cmdtp, int flag, int argc, char * const argv[]) int gpio; enum gpio_cmd sub_cmd; ulong value; - const char *str_cmd, *str_gpio; + const char *str_cmd; + char str_gpio[16] = { 0 }; + int ret; #ifdef gpio_status if (argc == 2 && !strcmp(argv[1], "status")) { @@ -40,7 +58,6 @@ static int do_gpio(cmd_tbl_t *cmdtp, int flag, int argc, char * const argv[]) show_usage: return CMD_RET_USAGE; str_cmd = argv[1]; - str_gpio = argv[2]; /* parse the behavior */ switch (*str_cmd) { @@ -52,7 +69,7 @@ static int do_gpio(cmd_tbl_t *cmdtp, int flag, int argc, char * const argv[]) } /* turn the gpio name into a gpio number */ - gpio = name_to_gpio(str_gpio); + gpio = gpio_from_string(argv[2]); if (gpio < 0) goto show_usage; @@ -75,8 +92,10 @@ static int do_gpio(cmd_tbl_t *cmdtp, int flag, int argc, char * const argv[]) } gpio_direction_output(gpio, value); } + + ret = gpio_to_string(gpio, str_gpio, sizeof(str_gpio)); printf("gpio: pin %s (gpio %i) value is %lu\n", - str_gpio, gpio, value); + ret ? str_gpio : argv[2], gpio, value); gpio_free(gpio); -- 1.7.10

3 11

[U-Boot] Problem booting linux kernel due to garbled available machine support
by Dallas Clement 02 Aug '12

02 Aug '12

Hello All, I'm having a problem booting a linux kernel built with Marvell buildroot-sdk-6.0 for an Armada XP development board (DB-78460-BP). The kernel messages indicate that the machine id is not recognized or not supported, but I believe this is because the available machine support is all junk. Marvell>> bootm 0x2000000 ## Booting kernel from Legacy Image at 02000000 ... Image Name: linux-3.0 Created: 2012-08-02 14:04:49 UTC Image Type: ARM Linux Kernel Image (uncompressed) Data Size: 3303200 Bytes = 3.2 MB Load Address: 02008000 Entry Point: 02008000 Verifying Checksum ... OK Loading Kernel Image ... OK OK Using machid 0xbdc from environment Starting kernel ... Uncompressing Linux... done, booting the kernel. Error: unrecognized/unsupported machine ID (r1 = 0x00000bdc). Available machine support: ID (hex) NAME 00000000 (null) c000e150 (null) À á0Ø-é°LââßMâMëP1å ã c000e234 ¬õÀ À á0Ø-é°Lâ Please check your kernel config and/or bootloader. These are the machine types in my kernel config which should be showing up in the available machine support, but for some reason is garbled. [dallasc /development/dallasc/TS-3400/buildroot-2011.05-sdk6.0/output/build/linux-3.0.6] grep '^CONFIG_MACH' .config CONFIG_MACH_ARMADA_XP_DB=y CONFIG_MACH_ARMADA_XP_RDSRV=y CONFIG_MACH_ARMADA_XP_RD_NAS=y [dallasc /development/dallasc/TS-3400/buildroot-2011.05-sdk6.0/output/build/linux-3.0.6] grep MACH_ARMADA arch/arm/tools/mach-types armada_xp_fpga MACH_ARMADA_XP_FPGA ARMADA_XP_FPGA 528 armada_xp_db MACH_ARMADA_XP_DB ARMADA_XP_DB 3036 armada_xp_rd_nas MACH_ARMADA_XP_RD_NAS ARMADA_XP_RD_NAS 3500 armada_xp_rdsrv MACH_ARMADA_XP_RDSRV ARMADA_XP_RDSRV 3037 It looks like the DB-78460-BP board is identifying itself as 0x00000bdc = 3036, which corresponds to the armada_xp_db. Does anyone have any suggestions on how to debug the cause of the garbling? Thanks, Dallas

1 0

[U-Boot] [PATCH 0/7] km82xx and mgcogeXxx reworking
by Gerlando Falauto 02 Aug '12

02 Aug '12

Hello, this patch series does a bit of housekeeping on our mgcogeXxx boards. First we do some due code cleanups which just move code around, then we introduce dynamic SDRAM detection so the same u-boot binary can be used on different boards. Gerlando Falauto (6): powerpc/82xx: move mgcoge, mgcoge3ne defines to ease subsequent merge powerpc/82xx: merge mgcoge.h and mgcoge3ne.h into km82xx.h powerpc/82xx: move km/km82xx-common.h within km82xx.h powerpc/82xx: add SDRAM detection for km82xx powerpc/82xx: use SDRAM detection for mgcoge2ne powerpc/82xx: adapt SDRAM settings for mgcoge3ne Holger Brunck (1): powerpc/82xx: remove unused define for mgcoge3ne board/keymile/km82xx/km82xx.c | 51 ++++- boards.cfg | 4 +- include/configs/km/km82xx-common.h | 317 ------------------------- include/configs/km82xx.h | 448 ++++++++++++++++++++++++++++++++++++ include/configs/mgcoge.h | 93 -------- include/configs/mgcoge3ne.h | 96 -------- 6 files changed, 499 insertions(+), 510 deletions(-) delete mode 100644 include/configs/km/km82xx-common.h create mode 100644 include/configs/km82xx.h delete mode 100644 include/configs/mgcoge.h delete mode 100644 include/configs/mgcoge3ne.h

2 26

[U-Boot] [PATCH] mx6q: mx6qsabrelite: add GPIO_0__CCM_CLKO and GPIO_3__CCM_CLKO2 pin mux
by Dirk Behme 02 Aug '12

02 Aug '12

A recent Linux kernel (>= 3.5) has support for the SGTL 5000 sound on the SabreLite board. To make this work, U-Boot has to configure the pin mux for PAD_GPIO_0__CCM_CLKO and PAD_GPIO_3__CCM_CLKO2 correctly. Taken from Freescale's ER5 U-Boot for the SabreLite. Signed-off-by: Dirk Behme <dirk.behme(a)de.bosch.com> CC: Troy Kisky <troy.kisky(a)boundarydevices.com> CC: Stefano Babic <sbabic(a)denx.de> --- board/freescale/mx6qsabrelite/imximage.cfg | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/board/freescale/mx6qsabrelite/imximage.cfg b/board/freescale/mx6qsabrelite/imximage.cfg index 62498ab..2d7825a 100644 --- a/board/freescale/mx6qsabrelite/imximage.cfg +++ b/board/freescale/mx6qsabrelite/imximage.cfg @@ -168,3 +168,6 @@ DATA 4 0x020e0010 0xF00000CF # set IPU AXI-id0 Qos=0xf(bypass) AXI-id1 Qos=0x7 DATA 4 0x020e0018 0x007F007F DATA 4 0x020e001c 0x007F007F +# set PAD_GPIO_0__CCM_CLKO and PAD_GPIO_3__CCM_CLKO2 +DATA 4 0x020e0220 0x00000000 +DATA 4 0x020e022c 0x00000004 \ No newline at end of file -- 1.7.0.4

6 10

[U-Boot] [PATCH] doc/git-mailrc: Update 'ti' alias
by Tom Rini 02 Aug '12

02 Aug '12

Remove Sandeep, thanks for all the hard work! Signed-off-by: Tom Rini <trini(a)ti.com> --- doc/git-mailrc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/doc/git-mailrc b/doc/git-mailrc index c8a6390..5eac8a5 100644 --- a/doc/git-mailrc +++ b/doc/git-mailrc @@ -50,7 +50,7 @@ alias s5pc samsung alias samsung uboot, prom alias tegra uboot, Simon Glass <sjg(a)chromium.org>, Tom Warren <twarren(a)nvidia.com>, Stephen Warren <swarren(a)nvidia.com> alias tegra2 tegra -alias ti uboot, Sandeep Paulraj <s-paulraj(a)ti.com>, Tom Rini <trini(a)ti.com> +alias ti uboot, Tom Rini <trini(a)ti.com> alias avr32 uboot, Andreas Bießmann <andreas.devel(a)googlemail.com> -- 1.7.9.5

1 0

[U-Boot] [PATCH 1/2] sf: stmicro: support JEDEC standard two-byte signature
by Stephan Linz 01 Aug '12

01 Aug '12

There are more than the M25Pxx serial flashs that can be used with the stmicro driver, for example: the M25PXxx or N25Qxx serie. All these chips have burned in the original stmicro manufacture id 0x20 together with a standard two-byte signature. In preperation to support all these chips the stmicro driver have to decode the full two-byte signature. Signed-off-by: Stephan Linz <linz(a)li-pro.net> --- drivers/mtd/spi/stmicro.c | 24 ++++++++++++------------ 1 files changed, 12 insertions(+), 12 deletions(-) diff --git a/drivers/mtd/spi/stmicro.c b/drivers/mtd/spi/stmicro.c index a9b33cf..4bee9ce 100644 --- a/drivers/mtd/spi/stmicro.c +++ b/drivers/mtd/spi/stmicro.c @@ -47,7 +47,7 @@ #define CMD_M25PXX_RES 0xab /* Release from DP, and Read Signature */ struct stmicro_spi_flash_params { - u8 idcode1; + u16 id; u16 page_size; u16 pages_per_sector; u16 nr_sectors; @@ -56,56 +56,56 @@ struct stmicro_spi_flash_params { static const struct stmicro_spi_flash_params stmicro_spi_flash_table[] = { { - .idcode1 = 0x11, + .id = 0x2011, .page_size = 256, .pages_per_sector = 128, .nr_sectors = 4, .name = "M25P10", }, { - .idcode1 = 0x15, + .id = 0x2015, .page_size = 256, .pages_per_sector = 256, .nr_sectors = 32, .name = "M25P16", }, { - .idcode1 = 0x12, + .id = 0x2012, .page_size = 256, .pages_per_sector = 256, .nr_sectors = 4, .name = "M25P20", }, { - .idcode1 = 0x16, + .id = 0x2016, .page_size = 256, .pages_per_sector = 256, .nr_sectors = 64, .name = "M25P32", }, { - .idcode1 = 0x13, + .id = 0x2013, .page_size = 256, .pages_per_sector = 256, .nr_sectors = 8, .name = "M25P40", }, { - .idcode1 = 0x17, + .id = 0x2017, .page_size = 256, .pages_per_sector = 256, .nr_sectors = 128, .name = "M25P64", }, { - .idcode1 = 0x14, + .id = 0x2014, .page_size = 256, .pages_per_sector = 256, .nr_sectors = 16, .name = "M25P80", }, { - .idcode1 = 0x18, + .id = 0x2018, .page_size = 256, .pages_per_sector = 1024, .nr_sectors = 64, @@ -139,13 +139,13 @@ struct spi_flash *spi_flash_probe_stmicro(struct spi_slave *spi, u8 * idcode) for (i = 0; i < ARRAY_SIZE(stmicro_spi_flash_table); i++) { params = &stmicro_spi_flash_table[i]; - if (params->idcode1 == idcode[2]) { + if (params->id == ((idcode[1] << 8) | idcode[2])) break; - } } if (i == ARRAY_SIZE(stmicro_spi_flash_table)) { - debug("SF: Unsupported STMicro ID %02x\n", idcode[1]); + debug("SF: Unsupported STMicro ID %02x%02x\n", + idcode[1], idcode[2]); return NULL; } -- 1.7.0.4

2 4

[U-Boot] [PATCH] SPI: Remove superfluous semicolon
by Marek Vasut 01 Aug '12

01 Aug '12

This is from someone on IRC. Signed-off-by: Marek Vasut <marex(a)denx.de> Cc: Mike Frysinger <vapier(a)gentoo.org> Cc: Wolfgang Denk <wd(a)denx.de> --- common/cmd_spi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/common/cmd_spi.c b/common/cmd_spi.c index 8c623c9..eba5fb8 100644 --- a/common/cmd_spi.c +++ b/common/cmd_spi.c @@ -89,7 +89,7 @@ int do_spi (cmd_tbl_t *cmdtp, int flag, int argc, char * const argv[]) cs = bus; bus = CONFIG_DEFAULT_SPI_BUS; } - if (*cp == '.'); + if (*cp == '.') mode = simple_strtoul(cp+1, NULL, 10); } if (argc >= 3) -- 1.7.10.4

3 2

[U-Boot] [U-BOOT]: Add support for numonyx spi flash.
by jagan 01 Aug '12

01 Aug '12

Hi, We have a numonyx spi flash (N25Q128) on my target. I tried using u-boot stmicro.c driver as numonyx and stmicro have same manufacturing ID's. The driver detects M25P128 as below ID table { .idcode1 = 0x18, .page_size = 256, .pages_per_sector = 1024, .nr_sectors = 64, .name = "M25P128", }, I think this could be because we doesn't have a separate driver to support numonyx. But the problem is N25Q128 and M25P128 are identically different in-terms of number of sectors as well as pages per sector due to this I was unable to program the flash. So, I edited pages_per_sector and nr_sectors to make my flash read/write will work. and also added an N25Q128 ID table above M25P128 keeping M25P128 will not effect. Everything is fine up to now. But I think this process is wired when we have more number of numonyx flashes which identically same manufacturer id's that are supported in stmicro.c with differences in sectors. I have written a separate driver for numonyx which is identically same as stmicro with added numonyx ID tables, so it will not touch any thing on stmicro.c. The reason I am writing different driver for numonyx because it will be a bit messy If we keep on adding extra things(specific to numonyx) to stmicro.c Please let me know your views, if my suggestion is good I will send the patch. If am wrong with my understanding, please send an optimized suggestion. Regards, Jagan.

2 1

[U-Boot] [PATCH] net: Make netconsole src and dest ports configurable
by Joe Hershberger 01 Aug '12

01 Aug '12

It is desirable to use different port numbers for sending and receiving packets with netconsole in the case where you have more than one device on the local subnet with netconsole enabled for broadcast. When they use the same port for both, any output from one will look like input to the other. This is typlically not desirable. This patch allows the input and output ports to be specified separately in the environment. Signed-off-by: Joe Hershberger <joe.hershberger(a)ni.com> --- doc/README.NetConsole | 3 +++ drivers/net/netconsole.c | 29 ++++++++++++++++++++--------- tools/netconsole | 18 +++++++++++------- 3 files changed, 34 insertions(+), 16 deletions(-) diff --git a/doc/README.NetConsole b/doc/README.NetConsole index c8bcb90..070e86a 100644 --- a/doc/README.NetConsole +++ b/doc/README.NetConsole @@ -11,6 +11,9 @@ port of the destination. The format is <ip_addr>:<port>. If <port> is omitted, the value of 6666 is used. If the env var doesn't exist, the broadcast address and port 6666 are used. If it is set to an IP address of 0 (or 0.0.0.0) then no messages are sent to the network. +The source / listening port can be configured separately by setting +the 'ncinport' environment variable and the destination port can be +configured by setting the 'ncoutport' environment variable. For example, if your server IP is 192.168.1.1, you could use: diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c index 14243b8..86f5301 100644 --- a/drivers/net/netconsole.c +++ b/drivers/net/netconsole.c @@ -36,7 +36,8 @@ static int output_recursion; static int net_timeout; static uchar nc_ether[6]; /* server enet address */ static IPaddr_t nc_ip; /* server ip */ -static short nc_port; /* source/target port */ +static short nc_out_port; /* target output port */ +static short nc_in_port; /* source input port */ static const char *output_packet; /* used by first send udp */ static int output_packet_len; @@ -71,7 +72,7 @@ void NcStart(void) net_set_arp_handler(nc_wait_arp_handler); pkt = (uchar *)NetTxPacket + NetEthHdrSize() + IP_UDP_HDR_SIZE; memcpy(pkt, output_packet, output_packet_len); - NetSendUDPPacket(nc_ether, nc_ip, nc_port, nc_port, + NetSendUDPPacket(nc_ether, nc_ip, nc_out_port, nc_in_port, output_packet_len); } } @@ -80,7 +81,7 @@ int nc_input_packet(uchar *pkt, unsigned dest, unsigned src, unsigned len) { int end, chunk; - if (dest != nc_port || !len) + if (dest != nc_in_port || !len) return 0; /* not for us */ debug_cond(DEBUG_DEV_PKT, "input: \"%*.*s\"\n", len, len, pkt); @@ -139,7 +140,7 @@ static void nc_send_packet(const char *buf, int len) memcpy(pkt, buf, len); ether = nc_ether; ip = nc_ip; - NetSendUDPPacket(ether, ip, nc_port, nc_port, len); + NetSendUDPPacket(ether, ip, nc_out_port, nc_in_port, len); if (inited) eth_halt(); @@ -148,20 +149,30 @@ static void nc_send_packet(const char *buf, int len) static int nc_start(void) { int netmask, our_ip; + char *p; - nc_port = 6666; /* default port */ + nc_out_port = 6666; /* default port */ + nc_in_port = nc_out_port; if (getenv("ncip")) { - char *p; nc_ip = getenv_IPaddr("ncip"); if (!nc_ip) return -1; /* ncip is 0.0.0.0 */ p = strchr(getenv("ncip"), ':'); - if (p != NULL) - nc_port = simple_strtoul(p + 1, NULL, 10); + if (p != NULL) { + nc_out_port = simple_strtoul(p + 1, NULL, 10); + nc_in_port = nc_out_port; + } } else - nc_ip = ~0; /* ncip is not set */ + nc_ip = ~0; /* ncip is not set, so broadcast */ + + p = getenv("ncoutport"); + if (p != NULL) + nc_out_port = simple_strtoul(p, NULL, 10); + p = getenv("ncinport"); + if (p != NULL) + nc_in_port = simple_strtoul(p, NULL, 10); our_ip = getenv_IPaddr("ipaddr"); netmask = getenv_IPaddr("netmask"); diff --git a/tools/netconsole b/tools/netconsole index c8109bb..b2b6edb 100755 --- a/tools/netconsole +++ b/tools/netconsole @@ -2,7 +2,7 @@ usage() { ( - echo "Usage: $0 <board IP> [board port]" + echo "Usage: $0 <board-IP> [board-port [board-in-port]]" echo "" echo "If port is not specified, '6666' will be used" [ -z "$*" ] && exit 0 @@ -24,9 +24,13 @@ while [ -n "$1" ] ; do done ip=$1 -port=${2:-6666} +board_out_port=${2:-6666} +board_in_port=${3:-$board_out_port} -if [ -z "${ip}" ] || [ -n "$3" ] ; then +echo Board out port: $board_out_port +echo Board in port: $board_in_port + +if [ -z "${ip}" ] || [ -n "$4" ] ; then usage "Invalid number of arguments" fi @@ -41,19 +45,19 @@ stty -icanon -echo intr ^T ( if type ncb 2>/dev/null ; then # see if ncb is in $PATH - exec ncb ${port} + exec ncb ${board_out_port} elif [ -x ${0%/*}/ncb ] ; then # maybe it's in the same dir as the netconsole script - exec ${0%/*}/ncb ${port} + exec ${0%/*}/ncb ${board_out_port} else # blah, just use regular netcat - while ${nc} -u -l -p ${port} < /dev/null ; do + while ${nc} -u -l -p ${board_out_port} < /dev/null ; do : done fi ) & pid=$! -${nc} -u ${ip} ${port} +${nc} -u ${ip} ${board_in_port} kill ${pid} 2>/dev/null -- 1.6.0.2

2 3