[Buildroot] [PATCH v3 0/6] Add support for the snowball board

[Buildroot] [PATCH v3 0/6] Add support for the snowball board

[Gregory Hermant]

Snowball board is a low power, low cost Single Board Computer based
on the ST-Ericsson Nova A9500 processor (Dual Cortex-A9 + MALI 400 GPU).
More info at www.igloocommunity.org

The snowball board boots from eMMC, loads the linux kernel and mounts
its root filesystem either from eMMC or from a microSD card.
The user should prepare a microSD card with two partitions.
On the fist partition he will install the kernel image and on the second
one the root filesystem.

TO DO:

* Add the GPS support.
* Add the possibility to build an eMMC image directly from buildroot.
* Add the riff tool package

Gregory Hermant (6):
  snowball-hdmiservice: new package
  ux500-firmware: new package
  Add board files for snowball
  Add init script files for snowball board
  bluez-utils: add support for the ST-Ericsson cg2900 combo controller
  Add config file for the snowball board

 board/calao/snowball/linux-3.4.patch               |   45 +++++++++
 configs/calao_snowball_defconfig                   |   28 ++++++
 package/Config.in                                  |    3 +
 ...tils-4.101-enable_cg2900_on_upstream_4.91.patch |  106 ++++++++++++++++++++
 package/snowball-hdmiservice/Config.in             |   10 ++
 .../snowball-hdmiservice/snowball-hdmiservice.mk   |   41 ++++++++
 package/snowball-init/Config.in                    |   10 ++
 package/snowball-init/snowball-init.mk             |   15 +++
 package/ux500-firmware/Config.in                   |    7 ++
 package/ux500-firmware/ux500-firmware.mk           |   22 ++++
 10 files changed, 287 insertions(+)
 create mode 100644 board/calao/snowball/linux-3.4.patch
 create mode 100644 configs/calao_snowball_defconfig
 create mode 100644 package/bluez_utils/bluez_utils-4.101-enable_cg2900_on_upstream_4.91.patch
 create mode 100644 package/snowball-hdmiservice/Config.in
 create mode 100644 package/snowball-hdmiservice/snowball-hdmiservice.mk
 create mode 100644 package/snowball-init/Config.in
 create mode 100644 package/snowball-init/snowball-init.mk
 create mode 100644 package/ux500-firmware/Config.in
 create mode 100644 package/ux500-firmware/ux500-firmware.mk

-- 
1.7.9.5

[Buildroot] [PATCH v3 1/6] snowball-hdmiservice: new package

[Gregory Hermant]

Signed-off-by: Gregory Hermant <gregory.hermant@calao-systems.com>
---
 package/Config.in                                  |    1 +
 package/snowball-hdmiservice/Config.in             |   10 +++++
 .../snowball-hdmiservice/snowball-hdmiservice.mk   |   41 ++++++++++++++++++++
 3 files changed, 52 insertions(+)
 create mode 100644 package/snowball-hdmiservice/Config.in
 create mode 100644 package/snowball-hdmiservice/snowball-hdmiservice.mk

diff --git a/package/Config.in b/package/Config.in
index 074e9df..e9fd4a9 100644
--- a/package/Config.in
+++ b/package/Config.in
@@ -250,6 +250,7 @@ source "package/sane-backends/Config.in"
 source "package/sdparm/Config.in"
 source "package/setserial/Config.in"
 source "package/smartmontools/Config.in"
+source "package/snowball-hdmiservice/Config.in"
 source "package/sredird/Config.in"
 source "package/statserial/Config.in"
 source "package/sysstat/Config.in"
diff --git a/package/snowball-hdmiservice/Config.in b/package/snowball-hdmiservice/Config.in
new file mode 100644
index 0000000..0ea7c81
--- /dev/null
+++ b/package/snowball-hdmiservice/Config.in
@@ -0,0 +1,10 @@
+config BR2_PACKAGE_SNOWBALL_HDMISERVICE
+	bool "snowball-hdmiservice"
+	depends on BR2_TOOLCHAIN_HAS_THREADS	
+	help
+	  HDMI userspace control daemon
+
+	  This package contains the HDMI userspace control daemon for the 
+	  snowball board
+	  
+	  http://www.igloocommunity.org
diff --git a/package/snowball-hdmiservice/snowball-hdmiservice.mk b/package/snowball-hdmiservice/snowball-hdmiservice.mk
new file mode 100644
index 0000000..fefa95d
--- /dev/null
+++ b/package/snowball-hdmiservice/snowball-hdmiservice.mk
@@ -0,0 +1,41 @@
+#############################################################
+#
+# snowball-hdmiservice
+#
+#############################################################
+SNOWBALL_HDMISERVICE_VERSION = f75c99d1c52707240a78b4ba78e41d20d3aa3b08
+SNOWBALL_HDMISERVICE_SITE = git://igloocommunity.org/git/bsp/hdmiservice.git
+SNOWBALL_HDMISERVICE_LICENSE = MIT
+SNOWBALL_HDMISERVICE_LICENSE_FILES = debian/copyright
+SNOWBALL_HDMISERVICE_INSTALL_STAGING = YES
+
+define SNOWBALL_HDMISERVICE_BUILD_CMDS
+	$(MAKE) CC="$(TARGET_CC)" -C $(@D)
+endef
+
+define SNOWBALL_HDMISERVICE_INSTALL_STAGING_CMDS
+	$(MAKE) CC="$(TARGET_CC)" -C $(@D) DESTDIR=$(STAGING_DIR) install 
+endef
+
+define SNOWBALL_HDMISERVICE_INSTALL_TARGET_CMDS
+	$(MAKE) CC="$(TARGET_CC)" -C $(@D) DESTDIR=$(TARGET_DIR) install
+endef
+
+define SNOWBALL_HDMISERVICE_UNINSTALL_STAGING_CMDS
+	rm -f $(STAGING_DIR)/usr/lib/hdmiservice.so
+	rm -f $(STAGING_DIR)/usr/bin/hdmistart
+	rm -f $(STAGING_DIR)/usr/include/hdmi_service_api.h
+	rm -f $(STAGING_DIR)/usr/include/hdmi_service_local.h
+endef
+
+define SNOWBALL_HDMISERVICE_UNINSTALL_TARGET_CMDS
+	rm -f $(TARGET_DIR)/usr/lib/hdmiservice.so
+	rm -f $(TARGET_DIR)/usr/bin/hdmistart
+endef
+
+define SNOWBALL_HDMISERVICE_CLEAN_CMDS
+	$(MAKE) -C $(@D) clean
+endef
+
+
+$(eval $(generic-package))
-- 
1.7.9.5

[Buildroot] [PATCH v3 2/6] ux500-firmware: new package

[Gregory Hermant]

Signed-off-by: Gregory Hermant <gregory.hermant@calao-systems.com>
---
 package/Config.in                        |    1 +
 package/ux500-firmware/Config.in         |    7 +++++++
 package/ux500-firmware/ux500-firmware.mk |   22 ++++++++++++++++++++++
 3 files changed, 30 insertions(+)
 create mode 100644 package/ux500-firmware/Config.in
 create mode 100644 package/ux500-firmware/ux500-firmware.mk

diff --git a/package/Config.in b/package/Config.in
index e9fd4a9..4fa21ca 100644
--- a/package/Config.in
+++ b/package/Config.in
@@ -261,6 +261,7 @@ source "package/usb_modeswitch/Config.in"
 source "package/usb_modeswitch_data/Config.in"
 source "package/usbmount/Config.in"
 source "package/usbutils/Config.in"
+source "package/ux500-firmware/Config.in"
 source "package/wipe/Config.in"
 endmenu
 
diff --git a/package/ux500-firmware/Config.in b/package/ux500-firmware/Config.in
new file mode 100644
index 0000000..7e8a94b
--- /dev/null
+++ b/package/ux500-firmware/Config.in
@@ -0,0 +1,7 @@
+config BR2_PACKAGE_UX500_FIRMWARE
+	bool "ux500-firmware"
+	help
+	  This package provides various binary firmware files (closed binary blobs)
+	  for the Azurewave AW-NH580 combo module (wifi, bt, gps).
+
+	  https://launchpad.net/~igloocommunity-maintainers/+archive/snowball/+files/
diff --git a/package/ux500-firmware/ux500-firmware.mk b/package/ux500-firmware/ux500-firmware.mk
new file mode 100644
index 0000000..1fcb3b7
--- /dev/null
+++ b/package/ux500-firmware/ux500-firmware.mk
@@ -0,0 +1,22 @@
+#############################################################
+#
+# ux500-firmware
+#
+#############################################################
+UX500_FIRMWARE_VERSION = 1.1.3-6
+UX500_FIRMWARE_SOURCE = ux500-firmware_$(UX500_FIRMWARE_VERSION)linaro1.tar.gz
+UX500_FIRMWARE_SITE = https://launchpad.net/~igloocommunity-maintainers/+archive/snowball/+files/
+UX500_FIRMWARE_LICENSE = Snowball click-wrap license
+UX500_FIRMWARE_LICENSE_FILES = license.txt 
+UX500_FIRMWARE_REDISTRIBUTE = NO
+
+define UX500_FIRMWARE_INSTALL_TARGET_CMDS
+	$(MAKE) -C $(@D) DESTDIR=$(TARGET_DIR) install
+# The CG2900 linux driver has to load firmware named CG29XX_* but the firmware
+# filenames contained in this package are CG2900_* hence the code below
+	for f in $(TARGET_DIR)/lib/firmware/CG2900* ; do \
+		mv $$f $${f/CG2900/CG29XX}; \
+	done
+endef
+
+$(eval $(generic-package))
-- 
1.7.9.5

[Buildroot] [PATCH v3 3/6] Add board files for snowball

[Gregory Hermant]

Signed-off-by: Gregory Hermant <gregory.hermant@calao-systems.com>
---
 board/calao/snowball/linux-3.4.patch |   45 ++++++++++++++++++++++++++++++++++
 1 file changed, 45 insertions(+)
 create mode 100644 board/calao/snowball/linux-3.4.patch

diff --git a/board/calao/snowball/linux-3.4.patch b/board/calao/snowball/linux-3.4.patch
new file mode 100644
index 0000000..4e6fb51
--- /dev/null
+++ b/board/calao/snowball/linux-3.4.patch
@@ -0,0 +1,45 @@
+From 3d1a9b44b9dcd2c0f5d54c09596c96c0524d8340 Mon Sep 17 00:00:00 2001
+From: Gregory Hermant <gregory.hermant@calao-systems.com>
+Date: Tue, 13 Nov 2012 11:47:41 +0100
+Subject: [PATCH] trace-debug[stm]: Fix for BT/WLAN instability
+
+Pins under mop500_ske_pins array are used for different
+functionalities in snowball. Ex: GPIO161 pin configured
+for keypad here must stay configured as GPIO alternate
+function for operating PMU_EN pin of BT/WLAN chip in snowball.
+So forbid configuring these pins, for snowball
+
+Signed-off-by: Rajanikanth H.V <rajanikanth.hv@stericsson.com>
+Signed-off-by: Gregory Hermant <gregory.hermant@calao-systems.com>
+---
+ arch/arm/mach-ux500/board-mop500-stm.c |    7 +++++--
+ 1 file changed, 5 insertions(+), 2 deletions(-)
+
+diff --git a/arch/arm/mach-ux500/board-mop500-stm.c b/arch/arm/mach-ux500/board-mop500-stm.c
+index 1bef2a0..532fb9e 100644
+--- a/arch/arm/mach-ux500/board-mop500-stm.c
++++ b/arch/arm/mach-ux500/board-mop500-stm.c
+@@ -121,7 +121,9 @@ static int stm_ste_disable_ape_on_mipi60(void)
+ 	if (retval)
+ 		STM_ERR("Failed to disable MIPI60\n");
+ 	else {
+-		retval = nmk_config_pins(ARRAY_AND_SIZE(mop500_ske_pins));
++		if (!machine_is_snowball())
++			retval = nmk_config_pins(
++				ARRAY_AND_SIZE(mop500_ske_pins));
+ 		if (retval)
+ 			STM_ERR("Failed to enable SKE gpio\n");
+ 	}
+@@ -314,7 +316,8 @@ static int stm_ste_connection(enum stm_connection_type con_type)
+ 		/* Enable altC3 on GPIO70-74 (STMMOD) and GPIO75-76 (UARTMOD) */
+ 		prcmu_enable_stm_mod_uart();
+ 		/* Enable APE on MIPI60 */
+-		retval = nmk_config_pins_sleep(ARRAY_AND_SIZE(mop500_ske_pins));
++		if (!machine_is_snowball())
++			retval = nmk_config_pins_sleep(ARRAY_AND_SIZE(mop500_ske_pins));
+ 		if (retval)
+ 			STM_ERR("Failed to disable SKE GPIO\n");
+ 		else {
+-- 
+1.7.9.5
+
-- 
1.7.9.5

[Buildroot] [PATCH v3 4/6] Add init script files for snowball board

[Gregory Hermant]

Signed-off-by: Gregory Hermant <gregory.hermant@calao-systems.com>
---
 package/Config.in                      |    1 +
 package/snowball-init/Config.in        |   10 ++++++++++
 package/snowball-init/snowball-init.mk |   15 +++++++++++++++
 3 files changed, 26 insertions(+)
 create mode 100644 package/snowball-init/Config.in
 create mode 100644 package/snowball-init/snowball-init.mk

diff --git a/package/Config.in b/package/Config.in
index 4fa21ca..488d4be 100644
--- a/package/Config.in
+++ b/package/Config.in
@@ -550,6 +550,7 @@ source "package/empty/Config.in"
 source "package/googlefontdirectory/Config.in"
 source "package/mobile-broadband-provider-info/Config.in"
 source "package/shared-mime-info/Config.in"
+source "package/snowball-init/Config.in"
 source "package/sound-theme-borealis/Config.in"
 source "package/sound-theme-freedesktop/Config.in"
 endmenu
diff --git a/package/snowball-init/Config.in b/package/snowball-init/Config.in
new file mode 100644
index 0000000..906ff2d
--- /dev/null
+++ b/package/snowball-init/Config.in
@@ -0,0 +1,10 @@
+config BR2_PACKAGE_SNOWBALL_INIT
+	bool "snowball-init"
+	# Runtime dependency, needed by snowball startup script
+	select BR2_PACKAGE_BLUEZ_UTILS
+	select BR2_PACKAGE_UX500_FIRMWARE
+	help
+	  Snowball init scripts
+
+	  http://www.igloocommunity.org
+
diff --git a/package/snowball-init/snowball-init.mk b/package/snowball-init/snowball-init.mk
new file mode 100644
index 0000000..278cacb
--- /dev/null
+++ b/package/snowball-init/snowball-init.mk
@@ -0,0 +1,15 @@
+#############################################################
+#
+# snowball-init
+#
+#############################################################
+SNOWBALL_INIT_VERSION = b064be21de25729039e5e54037bbdd2e25cfd5b7
+SNOWBALL_INIT_SITE = git://igloocommunity.org/git/bsp/snowball-init.git
+SNOWBALL_INIT_LICENSE = BSD-4c
+SNOWBALL_INIT_LICENSE_FILES = debian/copyright
+
+define SNOWBALL_INIT_INSTALL_TARGET_CMDS
+	$(INSTALL) -D -m 0755 $(@D)/snowball $(TARGET_DIR)/etc/init.d/S50snowball
+endef
+
+$(eval $(generic-package))
-- 
1.7.9.5

[Buildroot] [PATCH v3 5/6] bluez-utils: add support for the ST-Ericsson cg2900 combo controller

[Gregory Hermant]

Signed-off-by: Gregory Hermant <gregory.hermant@calao-systems.com>
---
 ...tils-4.101-enable_cg2900_on_upstream_4.91.patch |  106 ++++++++++++++++++++
 1 file changed, 106 insertions(+)
 create mode 100644 package/bluez_utils/bluez_utils-4.101-enable_cg2900_on_upstream_4.91.patch

diff --git a/package/bluez_utils/bluez_utils-4.101-enable_cg2900_on_upstream_4.91.patch b/package/bluez_utils/bluez_utils-4.101-enable_cg2900_on_upstream_4.91.patch
new file mode 100644
index 0000000..e80af5f
--- /dev/null
+++ b/package/bluez_utils/bluez_utils-4.101-enable_cg2900_on_upstream_4.91.patch
@@ -0,0 +1,106 @@
+From ac7992081abba87627c9e91735b3309584b48585 Mon Sep 17 00:00:00 2001
+From: Gregory Hermant <gregory.hermant@calao-systems.com>
+Date: Wed, 14 Nov 2012 14:27:02 +0100
+Subject: [PATCH] add support for the ST-Ericsson CG2900 GPS FM Bluetooth
+ combo controller
+
+
+Signed-off-by: Gregory Hermant <gregory.hermant@calao-systems.com>
+---
+ tools/hciattach.c |   22 +++++++++++++++-------
+ tools/hciattach.h |    1 +
+ 2 files changed, 16 insertions(+), 7 deletions(-)
+
+diff --git a/tools/hciattach.c b/tools/hciattach.c
+index e4d5aa1..7f08243 100644
+--- a/tools/hciattach.c
++++ b/tools/hciattach.c
+@@ -1066,6 +1066,11 @@ struct uart_t uart[] = {
+ 	{ "texasalt",   0x0000, 0x0000, HCI_UART_LL,   115200, 115200,
+ 				FLOW_CTL, DISABLE_PM, NULL, texasalt, NULL   },
+ 
++	/* ST-Ericsson CG2900 GPS FM Bluetooth combo controller */
++	{ "cg2900",     0x0000, 0x0000, HCI_UART_STE,  115200, 115200,
++				FLOW_CTL, DISABLE_PM, NULL, NULL     },
++
++
+ 	/* ST Microelectronics minikits based on STLC2410/STLC2415 */
+ 	{ "st",         0x0000, 0x0000, HCI_UART_H4,    57600, 115200,
+ 				FLOW_CTL, DISABLE_PM,  NULL, st       },
+@@ -1157,10 +1162,10 @@ static struct uart_t * get_by_type(char *type)
+ }
+ 
+ /* Initialize UART driver */
+-static int init_uart(char *dev, struct uart_t *u, int send_break, int raw)
++static int init_uart(char *dev, struct uart_t *u, int send_break, int raw, int line_disc)
+ {
+ 	struct termios ti;
+-	int fd, i;
++	int fd;
+ 	unsigned long flags = 0;
+ 
+ 	if (raw)
+@@ -1217,8 +1222,7 @@ static int init_uart(char *dev, struct uart_t *u, int send_break, int raw)
+ 	}
+ 
+ 	/* Set TTY to N_HCI line discipline */
+-	i = N_HCI;
+-	if (ioctl(fd, TIOCSETD, &i) < 0) {
++	if (ioctl(fd, TIOCSETD, &line_disc) < 0) {
+ 		perror("Can't set line discipline");
+ 		return -1;
+ 	}
+@@ -1243,7 +1247,7 @@ static void usage(void)
+ {
+ 	printf("hciattach - HCI UART driver initialization utility\n");
+ 	printf("Usage:\n");
+-	printf("\thciattach [-n] [-p] [-b] [-r] [-t timeout] [-s initial_speed] <tty> <type | id> [speed] [flow|noflow] [bdaddr]\n");
++	printf("\thciattach [-n] [-p] [-a line_disc_nr] [-b] [-r] [-t timeout] [-s initial_speed] <tty> <type | id> [speed] [flow|noflow] [bdaddr]\n");
+ 	printf("\thciattach -l\n");
+ }
+ 
+@@ -1252,6 +1256,7 @@ int main(int argc, char *argv[])
+ 	struct uart_t *u = NULL;
+ 	int detach, printpid, raw, opt, i, n, ld, err;
+ 	int to = 10;
++	int line_disc = N_HCI;
+ 	int init_speed = 0;
+ 	int send_break = 0;
+ 	pid_t pid;
+@@ -1264,8 +1269,11 @@ int main(int argc, char *argv[])
+ 	printpid = 0;
+ 	raw = 0;
+ 
+-	while ((opt=getopt(argc, argv, "bnpt:s:lr")) != EOF) {
++	while ((opt=getopt(argc, argv, "bnpt:s:lra:")) != EOF) {
+ 		switch(opt) {
++		case 'a':
++                        line_disc = atoi(optarg);
++                        break;
+ 		case 'b':
+ 			send_break = 1;
+ 			break;
+@@ -1381,7 +1389,7 @@ int main(int argc, char *argv[])
+ 	alarm(to);
+ 	bcsp_max_retries = to;
+ 
+-	n = init_uart(dev, u, send_break, raw);
++	n = init_uart(dev, u, send_break, raw, line_disc);
+ 	if (n < 0) {
+ 		perror("Can't initialize device");
+ 		exit(1);
+diff --git a/tools/hciattach.h b/tools/hciattach.h
+index fed0d11..09b534d 100644
+--- a/tools/hciattach.h
++++ b/tools/hciattach.h
+@@ -39,6 +39,7 @@
+ #define HCI_UART_H4DS	3
+ #define HCI_UART_LL	4
+ #define HCI_UART_ATH3K  5
++#define HCI_UART_STE	6
+ 
+ #define HCI_UART_RAW_DEVICE	0
+ 
+-- 
+1.7.9.5
+
-- 
1.7.9.5

[Buildroot] [PATCH v3 6/6] Add config file for the snowball board

[Gregory Hermant]

Signed-off-by: Gregory Hermant <gregory.hermant@calao-systems.com>
---
 configs/calao_snowball_defconfig |   28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)
 create mode 100644 configs/calao_snowball_defconfig

diff --git a/configs/calao_snowball_defconfig b/configs/calao_snowball_defconfig
new file mode 100644
index 0000000..700be5d
--- /dev/null
+++ b/configs/calao_snowball_defconfig
@@ -0,0 +1,28 @@
+BR2_arm=y
+BR2_cortex_a9=y
+BR2_GCC_VERSION_4_7_X=y
+# BR2_SOFT_FLOAT is not set
+BR2_TOOLCHAIN_BUILDROOT_LARGEFILE=y
+BR2_TOOLCHAIN_BUILDROOT_INET_IPV6=y
+BR2_TOOLCHAIN_BUILDROOT_INET_RPC=y
+BR2_TOOLCHAIN_BUILDROOT_LOCALE=y
+BR2_TOOLCHAIN_BUILDROOT_CXX=y
+BR2_VFP_FLOAT=y
+BR2_ROOTFS_DEVICE_CREATION_DYNAMIC_MDEV=y
+BR2_TARGET_GENERIC_GETTY_PORT="ttyAMA2"
+BR2_PACKAGE_BUSYBOX_SHOW_OTHERS=y
+BR2_PACKAGE_FBSET=y
+BR2_PACKAGE_SNOWBALL_HDMISERVICE=y
+BR2_PACKAGE_LIBNL=y
+BR2_PACKAGE_SNOWBALL_INIT=y
+BR2_PACKAGE_WPA_SUPPLICANT=y
+BR2_TARGET_ROOTFS_TAR_GZIP=y
+BR2_TARGET_UBOOT=y
+BR2_TARGET_UBOOT_BOARDNAME="snowball"
+BR2_LINUX_KERNEL=y
+BR2_LINUX_KERNEL_CUSTOM_GIT=y
+BR2_LINUX_KERNEL_CUSTOM_GIT_REPO_URL="git://igloocommunity.org/git/kernel/igloo-kernel.git"
+BR2_LINUX_KERNEL_CUSTOM_GIT_VERSION="stable-linux-ux500-3.4"
+BR2_LINUX_KERNEL_PATCH="board/calao/snowball/"
+BR2_LINUX_KERNEL_DEFCONFIG="u8500"
+BR2_LINUX_KERNEL_INSTALL_TARGET=y
-- 
1.7.9.5

[Buildroot] [PATCH v3 1/6] snowball-hdmiservice: new package

[Arnout Vandecappelle]

On 15/11/12 09:25, Gregory Hermant wrote:
>
> Signed-off-by: Gregory Hermant<gregory.hermant@calao-systems.com>
[snip]
> +
> +define SNOWBALL_HDMISERVICE_BUILD_CMDS
> +	$(MAKE) CC="$(TARGET_CC)" -C $(@D)
> +endef

  I may have asked this before, but is there a reason to not pass buildroot's
CFLAGS and LDFLAGS?  We usually pass $(TARGET_CONFIGURE_OPTS) to sub-makes.

[snip]


  Regards,
  Arnout

-- 
Arnout Vandecappelle                               arnout at mind be
Senior Embedded Software Architect                 +32-16-286540
Essensium/Mind                                     http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium                BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7CB5 E4CC 6C2E EFD4 6E3D A754 F963 ECAB 2450 2F1F

[Buildroot] [PATCH v3 2/6] ux500-firmware: new package

[Arnout Vandecappelle]

On 15/11/12 09:25, Gregory Hermant wrote:
>
> Signed-off-by: Gregory Hermant<gregory.hermant@calao-systems.com>

Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

  Minor detail below.

[snip]

> +define UX500_FIRMWARE_INSTALL_TARGET_CMDS
> +	$(MAKE) -C $(@D) DESTDIR=$(TARGET_DIR) install
> +# The CG2900 linux driver has to load firmware named CG29XX_* but the firmware
> +# filenames contained in this package are CG2900_* hence the code below

  Comments inside the _CMDS macros don't work well: they are interpreted by
the shell.  It's better to put them before the macro definition.


  Regards,
  Arnout

> +	for f in $(TARGET_DIR)/lib/firmware/CG2900* ; do \
> +		mv $$f $${f/CG2900/CG29XX}; \
> +	done
> +endef
> +
> +$(eval $(generic-package))

-- 
Arnout Vandecappelle                               arnout at mind be
Senior Embedded Software Architect                 +32-16-286540
Essensium/Mind                                     http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium                BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7CB5 E4CC 6C2E EFD4 6E3D A754 F963 ECAB 2450 2F1F

[Buildroot] [PATCH v3 1/6] snowball-hdmiservice: new package

[Grégory Hermant]

Arnout,
Le 21/11/2012 07:51, Arnout Vandecappelle a ?crit :
> On 15/11/12 09:25, Gregory Hermant wrote:
>>
>> Signed-off-by: Gregory Hermant<gregory.hermant@calao-systems.com>
> [snip]
>> +
>> +define SNOWBALL_HDMISERVICE_BUILD_CMDS
>> +    $(MAKE) CC="$(TARGET_CC)" -C $(@D)
>> +endef
>
>  I may have asked this before, but is there a reason to not pass buildroot's
> CFLAGS and LDFLAGS?  We usually pass $(TARGET_CONFIGURE_OPTS) to sub-makes.
>
> [snip]
>
The build of the hdmiservice.so fails if i pass the buildroot's LDFLAGS.

>  Regards,
>  Arnout
>
BR,
gregory

[Buildroot] [PATCH v3 1/6] snowball-hdmiservice: new package

[Arnout Vandecappelle]

On 22/11/12 13:35, Gr?gory Hermant wrote:
> Arnout,
> Le 21/11/2012 07:51, Arnout Vandecappelle a ?crit :
>> On 15/11/12 09:25, Gregory Hermant wrote:
>>>
>>> Signed-off-by: Gregory Hermant<gregory.hermant@calao-systems.com>
>> [snip]
>>> +
>>> +define SNOWBALL_HDMISERVICE_BUILD_CMDS
>>> +    $(MAKE) CC="$(TARGET_CC)" -C $(@D)
>>> +endef
>>
>>   I may have asked this before, but is there a reason to not pass buildroot's
>> CFLAGS and LDFLAGS?  We usually pass $(TARGET_CONFIGURE_OPTS) to sub-makes.
>>
>> [snip]
>>
> The build of the hdmiservice.so fails if i pass the buildroot's LDFLAGS.

  It should be passed in the environment, not in the make argument list.
Otherwise it will override the definition in the Makefile itself.

  A good Makefile appends to LDFLAGS passed in from the environment, but this
is not always the case...  Anyway the worst that can happen is that the LDFLAGS
are just ignored.

  Regards,
  Arnout
-- 
Arnout Vandecappelle                               arnout at mind be
Senior Embedded Software Architect                 +32-16-286540
Essensium/Mind                                     http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium                BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7CB5 E4CC 6C2E EFD4 6E3D A754 F963 ECAB 2450 2F1F

[Buildroot] [PATCH v3 1/6] snowball-hdmiservice: new package

[Grégory Hermant]

Le 22/11/2012 15:00, Arnout Vandecappelle a ?crit :
> It should be passed in the environment, not in the make argument list.
> Otherwise it will override the definition in the Makefile itself.
>
>  A good Makefile appends to LDFLAGS passed in from the environment, but this
> is not always the case...  Anyway the worst that can happen is that the LDFLAGS
> are just ignored. 
You mean that way:

SNOWBALL_HDMISERVICE_MAKE_ENV = CC="$(TARGET_CC)" CCFLAGS="$(TARGET_CCFLAGS)" LDFLAGS="$(TARGET_LDFLAGS)"

$(SNOWBALL_HDMISERVICE_MAKE_ENV) $(MAKE) -C $(@D)

BR,
gregory

[Buildroot] [PATCH v3 1/6] snowball-hdmiservice: new package

[Thomas Petazzoni]

Dear Gr?gory Hermant,

On Thu, 22 Nov 2012 18:33:05 +0100, Gr?gory Hermant wrote:
> Le 22/11/2012 15:00, Arnout Vandecappelle a ?crit :
> > It should be passed in the environment, not in the make argument list.
> > Otherwise it will override the definition in the Makefile itself.
> >
> >  A good Makefile appends to LDFLAGS passed in from the environment, but this
> > is not always the case...  Anyway the worst that can happen is that the LDFLAGS
> > are just ignored. 
> You mean that way:
> 
> SNOWBALL_HDMISERVICE_MAKE_ENV = CC="$(TARGET_CC)" CCFLAGS="$(TARGET_CCFLAGS)" LDFLAGS="$(TARGET_LDFLAGS)"
> 
> $(SNOWBALL_HDMISERVICE_MAKE_ENV) $(MAKE) -C $(@D)

No, just:

$(TARGET_CONFIGURE_OPTS) $(MAKE) -C $(@D)

Thomas
-- 
Thomas Petazzoni, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com

[Buildroot] [PATCH v3 1/6] snowball-hdmiservice: new package

[Grégory Hermant]

Le 22/11/2012 18:45, Thomas Petazzoni a ?crit :
> Dear Gr?gory Hermant,
>
> On Thu, 22 Nov 2012 18:33:05 +0100, Gr?gory Hermant wrote:
>> Le 22/11/2012 15:00, Arnout Vandecappelle a ?crit :
>>> It should be passed in the environment, not in the make argument list.
>>> Otherwise it will override the definition in the Makefile itself.
>>>
>>>  A good Makefile appends to LDFLAGS passed in from the environment, but this
>>> is not always the case...  Anyway the worst that can happen is that the LDFLAGS
>>> are just ignored. 
>> You mean that way:
>>
>> SNOWBALL_HDMISERVICE_MAKE_ENV = CC="$(TARGET_CC)" CCFLAGS="$(TARGET_CCFLAGS)" LDFLAGS="$(TARGET_LDFLAGS)"
>>
>> $(SNOWBALL_HDMISERVICE_MAKE_ENV) $(MAKE) -C $(@D)
> No, just:
>
> $(TARGET_CONFIGURE_OPTS) $(MAKE) -C $(@D)
>
> Thomas

It doesn't work with this Makefile.

Indeed if i use $(TARGET_CONFIGURE_OPTS) $(MAKE) -C $(@D), CC is set to the host gcc compiler.

gregory

[Buildroot] [PATCH v3 1/6] snowball-hdmiservice: new package

[Arnout Vandecappelle]

On 23/11/12 08:54, Gr?gory Hermant wrote:
> Le 22/11/2012 18:45, Thomas Petazzoni a ?crit :
>> Dear Gr?gory Hermant,
>>
>> On Thu, 22 Nov 2012 18:33:05 +0100, Gr?gory Hermant wrote:
>>> Le 22/11/2012 15:00, Arnout Vandecappelle a ?crit :
>>>> It should be passed in the environment, not in the make argument list.
>>>> Otherwise it will override the definition in the Makefile itself.
>>>>
>>>>   A good Makefile appends to LDFLAGS passed in from the environment, but this
>>>> is not always the case...  Anyway the worst that can happen is that the LDFLAGS
>>>> are just ignored.
>>> You mean that way:
>>>
>>> SNOWBALL_HDMISERVICE_MAKE_ENV = CC="$(TARGET_CC)" CCFLAGS="$(TARGET_CCFLAGS)" LDFLAGS="$(TARGET_LDFLAGS)"
>>>
>>> $(SNOWBALL_HDMISERVICE_MAKE_ENV) $(MAKE) -C $(@D)
>> No, just:
>>
>> $(TARGET_CONFIGURE_OPTS) $(MAKE) -C $(@D)
>>
>> Thomas
>
> It doesn't work with this Makefile.
>
> Indeed if i use $(TARGET_CONFIGURE_OPTS) $(MAKE) -C $(@D), CC is set to the host gcc compiler.

  Silly Makefile...

  You can do either:

$(TARGET_CONFIGURE_OPTS) $(MAKE) -C $(@D) CC="$(TARGET_CC)"

or fix the Makefile in a patch.

  But if the Makefile overrides CC, it will most likely also override all the
other stuff so passing TARGET_CONFIGURE_OPTS in the environment doesn't help
much.


  What should be our philosophy for bad Makefiles?  Fix the Makefile and try
to upstream the patch, or just work around it in a package-specific way?

  We can actually work around it generically, using:

$(MAKE) -C $(@D) CC="$(TARGET_CC) $(TARGET_CPPFLAGS) $(TARGET_CFLAGS)"

but I don't like forcing flags into the CC command.


  Regards,
  Arnout

-- 
Arnout Vandecappelle                               arnout at mind be
Senior Embedded Software Architect                 +32-16-286540
Essensium/Mind                                     http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium                BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7CB5 E4CC 6C2E EFD4 6E3D A754 F963 ECAB 2450 2F1F