Add Alchemy build system and some related libs
cover Add Alchemy build system and some related libs
Hi,
This patch series introduces the Parrot Alchemy build system as a host tool (patch 1).
This tool is then used to build several libs (patch 2, 3 and 4). Finaly (patch 5), a test is provided to check that the binaries were correctly built.
Best regards,
Hervé Codina
Herve Codina (5):
package/alchemy: new host package package/ulog: new package
package/libfutils: new package package/libshdata: new package support/testing/tests/package/test_libshdata: new test
DEVELOPERS | 5 ++ package/Config.in | 3 + ...lchains-remove-hash-style-management.patch | 36 +++++++++ package/alchemy/alchemy.hash | 3 + package/alchemy/alchemy.mk | 73 +++++++++++++++++++ package/libfutils/Config.in | 14 ++++ package/libfutils/libfutils.hash | 3 + package/libfutils/libfutils.mk | 50 +++++++++++++ ...01-backend-Add-missing-include-files.patch | 44 +++++++++++ package/libshdata/Config.in | 23 ++++++ package/libshdata/libshdata.hash | 3 + package/libshdata/libshdata.mk | 59 +++++++++++++++ package/ulog/Config.in | 12 +++ package/ulog/ulog.hash | 3 + package/ulog/ulog.mk | 44 +++++++++++ .../testing/tests/package/test_libshdata.py | 25 +++++++ 16 files changed, 400 insertions(+) create mode 100644 package/alchemy/0001-toolchains-remove-hash-style-management.patch create mode 100644 package/alchemy/alchemy.hash create mode 100644 package/alchemy/alchemy.mk create mode 100644 package/libfutils/Config.in create mode 100644 package/libfutils/libfutils.hash create mode 100644 package/libfutils/libfutils.mk create mode 100644 package/libshdata/0001-backend-Add-missing-include-files.patch create mode 100644 package/libshdata/Config.in create mode 100644 package/libshdata/libshdata.hash create mode 100644 package/libshdata/libshdata.mk create mode 100644 package/ulog/Config.in create mode 100644 package/ulog/ulog.hash create mode 100644 package/ulog/ulog.mk create mode 100644 support/testing/tests/package/test_libshdata.py
Signature · 6 lines
2.31.1 _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
PATCH 1/5 package/alchemy: new host package
Alchemy is a build system developed by Parrot.
It is a new build system based on the one used in Android. A central makefile instance scans a workspace to find user makefiles, includes them and register modules to be built.
https://github.com/Parrot-Developers/alchemy
--- DEVELOPERS | 1 + ...lchains-remove-hash-style-management.patch | 36 +++++++++ package/alchemy/alchemy.hash | 3 + package/alchemy/alchemy.mk | 73 +++++++++++++++++++ 4 files changed, 113 insertions(+) create mode 100644 package/alchemy/0001-toolchains-remove-hash-style-management.patch create mode 100644 package/alchemy/alchemy.hash create mode 100644 package/alchemy/alchemy.mk
DEVELOPERS +1 -0
@@ -1147,6 +1147,7 @@ N: Henrique Camargo <[email protected]> F: package/json-glib/ N: Hervé Codina <[email protected]> +F: package/alchemy/ F: package/dtbocfg/ F: package/libdbi/ F: package/libdbi-drivers/
package/alchemy/0001-toolchains-remove-hash-style-management.patch new file +36 -0
@@ -0,0 +1,36 @@ +From 4cb9681ab99db3d8e330ac789e17ec69583aecd4 Mon Sep 17 00:00:00 2001 +From: Herve Codina <[email protected]> +Date: Thu, 28 Oct 2021 18:39:18 +0200 +Subject: [PATCH] toolchains: remove --hash-style management + +Using an external toolchain, with an external configuration given +using TARGET_GLOBAL_XXXX variables, having '--hash-style=both' +in Alchemy toolchains can lead to compilation failure. + +This patch simply removes all --hash-style settings. + +Signed-off-by: Herve Codina <[email protected]> +--- + toolchains/linux/eglibc/flags.mk | 9 +-------- + 1 file changed, 1 insertion(+), 8 deletions(-) + +diff --git a/toolchains/linux/eglibc/flags.mk b/toolchains/linux/eglibc/flags.mk +index 95c0a46..4474729 100644 +--- a/toolchains/linux/eglibc/flags.mk ++++ b/toolchains/linux/eglibc/flags.mk +@@ -11,11 +11,4 @@ TARGET_GLOBAL_LDLIBS += -pthread -lrt + TARGET_GLOBAL_CFLAGS += -funwind-tables + + # Enable link optimization for binutils's ld. +-# gnu hash not supported by mips ABI +-ifeq ("$(TARGET_ARCH)","mips") +- TARGET_GLOBAL_LDFLAGS += -Wl,-O1 +-else ifeq ("$(TARGET_ARCH)","mips64") +- TARGET_GLOBAL_LDFLAGS += -Wl,-O1 +-else +- TARGET_GLOBAL_LDFLAGS += -Wl,-O1,--hash-style=both +-endif ++TARGET_GLOBAL_LDFLAGS += -Wl,-O1 +-- +2.31.1 +
package/alchemy/alchemy.hash new file +3 -0
@@ -0,0 +1,3 @@ +# Locally computed: +sha256 346c06b9317691a9371d878b850e61f8bcec4f20132ac95340b9d2571be25ee8 alchemy-d95b3c38cd37814a1b98d0bbf813de7adaaecfbc.tar.gz +sha256 542f44eecd9f8806ec1ae285201e442ab547c82d9f8c70e44c977d6332a2bea0 COPYING
package/alchemy/alchemy.mk new file +73 -0
73 lines not on this page · raw message
Signature · 6 lines
2.31.1 _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
Re: [Buildroot] [PATCH 1/5] package/alchemy: new host package
Hello Hervé,
Quoted 9 lines · On Wed, 3 Nov 2021 08:36:52 +0100 Herve Codina <[email protected]>…
Alchemy is a build system developed by Parrot.
It is a new build system based on the one used in Android. A central makefile instance scans a workspace to find user makefiles, includes them and register modules to be built.
https://github.com/Parrot-Developers/alchemy
Signed-off-by Herve Codina <[email protected]>
Thanks for this. I think we should make it clear to other reviewers that we intentionally did not add a new package infrastructure for this new build system. Indeed, we do not expect to have a significant number of packages that will rely on the Alchemy build system.
Of course, should we have more of these packages in the future, we can consider adding a package infrastructure.
Quoted 6 lines
package/alchemy/alchemy.mk new file +0
-0
package/alchemy/alchemy.mk line 9
Quoted 9 lines
package/alchemy/alchemy.mk new file +9
-0@@ -0,0 +1,73 @@ +################################################################################ +# +# alchemy +# +################################################################################ + +ALCHEMY_VERSION = d95b3c38cd37814a1b98d0bbf813de7adaaecfbc +ALCHEMY_SITE = $(call github,Parrot-Developers,alchemy,$(ALCHEMY_VERSION)) +ALCHEMY_LICENSE = BSD-3-Clause
The kconfig code is under GPL-2.0.
package/alchemy/alchemy.mk line 18
Quoted 9 lines
package/alchemy/alchemy.mk new file +9
-0+ALCHEMY_LICENSE_FILES = COPYING +HOST_ALCHEMY_DEPENDENCIES = host-python3 + +ALCHEMY_SDK_BASEDIR = $(STAGING_DIR)/usr/lib/alchemy/sdk + +define HOST_ALCHEMY_INSTALL_CMDS + rm -rf $(HOST_DIR)/usr/bin/alchemy + mkdir -p $(HOST_DIR)/usr/bin/alchemy + cp -rf $(@D)/* $(HOST_DIR)/usr/bin/alchemy/
It feels really odd to have the entire Alchemy code base installed in $(HOST_DIR)/usr/bin/alchemy/. Perhaps better installed in $(HOST_DIR)/opt ?
Also $(HOST_DIR)/usr in fact no longer exists, it's a symlink to $(HOST_DIR) itself, so instead of using $(HOST_DIR)/usr/bin, we prefer using $(HOST_DIR)/bin directly.
package/alchemy/alchemy.mk line 19
Quoted 1 line
package/alchemy/alchemy.mk new file +1
-0+ mkdir -p $(ALCHEMY_SDK_BASEDIR)
Is this required to be done here? It's just that this directory is in the STAGING_DIR, so having it created during the installation of a host package isn't nice (but admittedly we do have a number of packages who already do this kind of things).
package/alchemy/alchemy.mk line 29
Quoted 8 lines
package/alchemy/alchemy.mk new file +8
-0+$(eval $(host-generic-package)) + +# Variables used by other packages + +ALCHEMY_HOME = $(HOST_DIR)/usr/bin/alchemy +ALCHEMY_MAKE = $(ALCHEMY_HOME)/scripts/alchemake + +ALCHEMY_TARGET_CONFIGURE_ENV = \
You never have any CONFIGURE_CMDS in alchemy based packages, and this variable gets used in BUILD_CMDS, so the name ALCHEMY_TARGET_CONFIGURE_ENV doesn't look very good. Perhaps just ALCHEMY_TARGET_ENV ?
package/alchemy/alchemy.mk line 37
Quoted 8 lines
package/alchemy/alchemy.mk new file +8
-0+ $(TARGET_MAKE_ENV) \ + ALCHEMY_HOME=$(ALCHEMY_HOME) \ + ALCHEMY_WORKSPACE_DIR="$(@D)" \ + ALCHEMY_TARGET_OUT=alchemy-out \ + TARGET_OS=linux \ + TARGET_OS_FLAVOUR=buildroot \ + TARGET_CROSS="$(TARGET_CROSS)" \ + TARGET_ARCH=xxx \
Is this intended ?
package/alchemy/alchemy.mk line 45
Quoted 8 lines
package/alchemy/alchemy.mk new file +8
-0+ TARGET_GLOBAL_CXXFLAGS="$(TARGET_CXXFLAGS)" \ + TARGET_GLOBAL_LDFLAGS="$(TARGET_LDFLAGS)" \ + TARGET_GLOBAL_FFLAGS="$(TARGET_FCFLAGS)" \ + TARGET_GLOBAL_FCFLAGS="$(TARGET_FCFLAGS)" + +ifeq ($(BR2_STATIC_LIBS),y) +ALCHEMY_TARGET_CONFIGURE_ENV += TARGET_FORCE_STATIC=1 +ALCHEMY_TARGET_CONFIGURE_ENV += TARGET_GLOBAL_CFLAGS="$(TARGET_CFLAGS)"
Just one assignment:
FOOBAR += \ THIS=1 \ THAT=2
package/alchemy/alchemy.mk line 55
Quoted 10 lines
package/alchemy/alchemy.mk new file +10
-0+else +ALCHEMY_TARGET_CONFIGURE_ENV += TARGET_GLOBAL_CFLAGS="$(TARGET_CFLAGS) -fPIC" +endif + +# Configure Alchemy package dependencies. +# This macro can be used by Alchemy packages +# $1: List of Buildroot alchemy package the caller depends on +define ALCHEMY_TARGET_CONFIGURE_SDKS + ALCHEMY_TARGET_SDK_DIRS="$(addprefix $(ALCHEMY_SDK_BASEDIR)/,$(1))" +endef
I don't really like this idiom, I don't think we have constructs like this in Buildroot so far, and I find it quite confusing. I think I would prefer to keep it explicit, i.e in libfutils:
+define LIBFUTILS_BUILD_CMDS
+ $(ALCHEMY_TARGET_CONFIGURE_ENV) \
+ ALCHEMY_TARGET_SDK_DIRS="$(addprefix $(ALCHEMY_SDK_BASEDIR)/,ulog)"
+ $(LIBFUTILS_CONF_ENV) \
+ $(ALCHEMY_MAKE) libfutils
+endef
Indeed, this macro which generates an assignment that should be added in the environment of ALCHEMY_MAKE feels too tricky.
Quoted 7 lines
package/alchemy/alchemy.mk new file +7
-0+# Install an Alchemy SDK file. +# This macro can be used by Alchemy packages +# $1: Buildroot package name +# $2: Alchemy module name +# $3: Alchemy module file name +define ALCHEMY_INSTALL_LIB_SDK_FILE + mkdir -p $(ALCHEMY_SDK_BASEDIR)/$(strip $(1))
package/alchemy/alchemy.mk line 73
Quoted 10 lines
package/alchemy/alchemy.mk new file +10
-0+ ( \ + echo 'LOCAL_PATH := $$(call my-dir)'; \ + echo 'include $$(CLEAR_VARS)'; \ + echo 'LOCAL_MODULE := $(strip $(2))'; \ + echo 'LOCAL_SDK := $(STAGING_DIR)'; \ + echo 'LOCAL_DESTDIR := usr/lib'; \ + echo 'LOCAL_MODULE_FILENAME := $(strip $(3))'; \ + echo 'include $$(BUILD_LIBRARY)'; \ + ) > $(ALCHEMY_SDK_BASEDIR)/$(strip $(1))/atom.mk +endef
Thanks!
Thomas
Signature · 7 lines
Thomas Petazzoni, co-owner and CEO, Bootlin Embedded Linux and Kernel engineering and training https://bootlin.com _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
Re: [Buildroot] [PATCH 1/5] package/alchemy: new host package
Hi,
Quoted 1 line · On Wed, 3 Nov 2021 16:44:31 +0100 Thomas Petazzoni…
Hello Hervé,
[...]
Quoted 5 lines
package/alchemy/alchemy.mk line 9
Quoted 3 lines
package/alchemy/alchemy.mk new file +3
-0+ALCHEMY_VERSION = d95b3c38cd37814a1b98d0bbf813de7adaaecfbc +ALCHEMY_SITE = $(call github,Parrot-Developers,alchemy,$(ALCHEMY_VERSION)) +ALCHEMY_LICENSE = BSD-3-ClauseThe kconfig code is under GPL-2.0.
I missed this one.
The GPL-2.0 license file is not provided by Alchemy. They mention 'kconfig source code is GPLv2' in their README file.
ALCHEMY_LICENSE = BSD-3-Clause (Alchemy), GPL-2.0 (kconfig) ALCHEMY_LICENSE_FILES = COPYING README
Is that correct from Buildroot even if the GLP-2.0 license file is not provided ?
Quoted 14 lines
package/alchemy/alchemy.mk line 18
Quoted 10 lines
package/alchemy/alchemy.mk new file +9
-0+ALCHEMY_LICENSE_FILES = COPYING +HOST_ALCHEMY_DEPENDENCIES = host-python3 + +ALCHEMY_SDK_BASEDIR = $(STAGING_DIR)/usr/lib/alchemy/sdk + +define HOST_ALCHEMY_INSTALL_CMDS + rm -rf $(HOST_DIR)/usr/bin/alchemy + mkdir -p $(HOST_DIR)/usr/bin/alchemy + cp -rf $(@D)/* $(HOST_DIR)/usr/bin/alchemy/It feels really odd to have the entire Alchemy code base installed in $(HOST_DIR)/usr/bin/alchemy/. Perhaps better installed in $(HOST_DIR)/opt ?
Ok, I will change in v2 to $(HOST_DIR)/opt/alchemy/
Quoted 11 lines
Also $(HOST_DIR)/usr in fact no longer exists, it's a symlink to $(HOST_DIR) itself, so instead of using $(HOST_DIR)/usr/bin, we prefer using $(HOST_DIR)/bin directly.
package/alchemy/alchemy.mk line 19
Quoted 1 line
package/alchemy/alchemy.mk new file +1
-0+ mkdir -p $(ALCHEMY_SDK_BASEDIR)Is this required to be done here? It's just that this directory is in the
STAGING_DIR, so having it created during the installation of a host package isn't nice (but admittedly we do have a number of packages who already do this kind of things).
I will simply remove this mkdir in v2 as it is done already done in ALCHEMY_INSTALL_LIB_SDK_FILE macro.
Quoted 14 lines
package/alchemy/alchemy.mk line 29
Quoted 9 lines
+$(eval $(host-generic-package))
package/alchemy/alchemy.mk new file +7
-0+ +# Variables used by other packages + +ALCHEMY_HOME = $(HOST_DIR)/usr/bin/alchemy +ALCHEMY_MAKE = $(ALCHEMY_HOME)/scripts/alchemake + +ALCHEMY_TARGET_CONFIGURE_ENV = \You never have any
CONFIGURE_CMDSin alchemy based packages, and this variable gets used inBUILD_CMDS, so the nameALCHEMY_TARGET_CONFIGURE_ENVdoesn't look very good. Perhaps justALCHEMY_TARGET_ENV?
I will change in v2 to ALCHEMY_TARGET_ENV.
Quoted 11 lines
package/alchemy/alchemy.mk line 37
Quoted 9 lines
package/alchemy/alchemy.mk new file +8
-0+ $(TARGET_MAKE_ENV) \ + ALCHEMY_HOME=$(ALCHEMY_HOME) \ + ALCHEMY_WORKSPACE_DIR="$(@D)" \ + ALCHEMY_TARGET_OUT=alchemy-out \ + TARGET_OS=linux \ + TARGET_OS_FLAVOUR=buildroot \ + TARGET_CROSS="$(TARGET_CROSS)" \ + TARGET_ARCH=xxx \Is this intended ?
Yes, it is to avoid the internal Alchemy arch management.
It can lead to incorrect results with the toolchain we provided in TARGET_CROSS. Indeed they add some CFLAGS depending on given TARGET_ARCH.
To avoid issues, this was the simplest way I found to say 'Hey Alchemy, do not care about the arch, I do it for you'.
Maybe this breaks some internal parts such as building a CMake or an autotools package but in the buildroot context, A CMake or an autotools package will not be built by Alchemy.
I will add a comment in v2.
Quoted 12 lines
package/alchemy/alchemy.mk line 45
Quoted 9 lines
package/alchemy/alchemy.mk new file +8
-0+ TARGET_GLOBAL_CXXFLAGS="$(TARGET_CXXFLAGS)" \ + TARGET_GLOBAL_LDFLAGS="$(TARGET_LDFLAGS)" \ + TARGET_GLOBAL_FFLAGS="$(TARGET_FCFLAGS)" \ + TARGET_GLOBAL_FCFLAGS="$(TARGET_FCFLAGS)" + +ifeq ($(BR2_STATIC_LIBS),y) +ALCHEMY_TARGET_CONFIGURE_ENV += TARGET_FORCE_STATIC=1 +ALCHEMY_TARGET_CONFIGURE_ENV += TARGET_GLOBAL_CFLAGS="$(TARGET_CFLAGS)"Just one assignment:
Quoted 3 lines
FOOBAR += \ THIS=1 \ THAT=2
Will be changed in v2.
Quoted 23 lines
package/alchemy/alchemy.mk line 55
Quoted 11 lines
package/alchemy/alchemy.mk new file +10
-0+else +ALCHEMY_TARGET_CONFIGURE_ENV += TARGET_GLOBAL_CFLAGS="$(TARGET_CFLAGS) -fPIC" +endif + +# Configure Alchemy package dependencies. +# This macro can be used by Alchemy packages +# $1: List of Buildroot alchemy package the caller depends on +define ALCHEMY_TARGET_CONFIGURE_SDKS + ALCHEMY_TARGET_SDK_DIRS="$(addprefix $(ALCHEMY_SDK_BASEDIR)/,$(1))" +endefI don't really like this idiom, I don't think we have constructs like this in Buildroot so far, and I find it quite confusing. I think I would prefer to keep it explicit, i.e in libfutils:
(quoted diff) +6
-0+define LIBFUTILS_BUILD_CMDS + $(ALCHEMY_TARGET_CONFIGURE_ENV) \ + ALCHEMY_TARGET_SDK_DIRS="$(addprefix $(ALCHEMY_SDK_BASEDIR)/,ulog)" + $(LIBFUTILS_CONF_ENV) \ + $(ALCHEMY_MAKE) libfutils +endef
Quoted 2 lines
Indeed, this macro which generates an assignment that should be added in the environment of
ALCHEMY_MAKEfeels too tricky.
I will remove the macro in v2 and make the explicit assignment in each packages.
Quoted 21 lines
Quoted 8 lines
package/alchemy/alchemy.mk new file +7
-0+# Install an Alchemy SDK file. +# This macro can be used by Alchemy packages +# $1: Buildroot package name +# $2: Alchemy module name +# $3: Alchemy module file name +define ALCHEMY_INSTALL_LIB_SDK_FILE + mkdir -p $(ALCHEMY_SDK_BASEDIR)/$(strip $(1))package/alchemy/alchemy.mk line 73
Quoted 10 lines
package/alchemy/alchemy.mk new file +10
-0+ ( \ + echo 'LOCAL_PATH := $$(call my-dir)'; \ + echo 'include $$(CLEAR_VARS)'; \ + echo 'LOCAL_MODULE := $(strip $(2))'; \ + echo 'LOCAL_SDK := $(STAGING_DIR)'; \ + echo 'LOCAL_DESTDIR := usr/lib'; \ + echo 'LOCAL_MODULE_FILENAME := $(strip $(3))'; \ + echo 'include $$(BUILD_LIBRARY)'; \ + ) > $(ALCHEMY_SDK_BASEDIR)/$(strip $(1))/atom.mk +endefThanks!
Quoted 1 line
Thomas
Thanks for the review.
Hervé
Signature · 7 lines
Hervé Codina, Bootlin Embedded Linux and Kernel engineering https://bootlin.com _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
Re: [Buildroot] [PATCH 1/5] package/alchemy: new host package
Hello Hervé,
Quoted 4 lines · On Wed, 3 Nov 2021 18:20:47 +0100 Herve Codina <[email protected]> wrote:
Quoted 1 line
The kconfig code is under GPL-2.0.
I missed this one.
Quoted 8 lines
The GPL-2.0 license file is not provided by Alchemy. They mention 'kconfig source code is GPLv2' in their README file.
ALCHEMY_LICENSE= BSD-3-Clause (Alchemy), GPL-2.0 (kconfig)ALCHEMY_LICENSE_FILES= COPYING READMEIs that correct from Buildroot even if the GLP-2.0 license file is not provided ?
Yes, that's fine.
Quoted 19 lines
Quoted 10 lines
package/alchemy/alchemy.mk line 37
Quoted 8 lines
package/alchemy/alchemy.mk new file +8
-0+ $(TARGET_MAKE_ENV) \ + ALCHEMY_HOME=$(ALCHEMY_HOME) \ + ALCHEMY_WORKSPACE_DIR="$(@D)" \ + ALCHEMY_TARGET_OUT=alchemy-out \ + TARGET_OS=linux \ + TARGET_OS_FLAVOUR=buildroot \ + TARGET_CROSS="$(TARGET_CROSS)" \ + TARGET_ARCH=xxx \Is this intended ?
Yes, it is to avoid the internal Alchemy arch management.
It can lead to incorrect results with the toolchain we provided in
TARGET_CROSS. Indeed they add some CFLAGS depending on givenTARGET_ARCH.
To avoid issues, this was the simplest way I found to say 'Hey Alchemy, do not care about the arch, I do it for you'.
Quoted 5 lines
Maybe this breaks some internal parts such as building a CMake or an autotools package but in the buildroot context, A CMake or an autotools package will not be built by Alchemy.
I will add a comment in v2.
Thanks for the explanation. Adding a comment in the code will definitely be useful to clear up what's happening with this TARGET_ARCH variable.
Thanks!
Thomas
Signature · 7 lines
Thomas Petazzoni, co-owner and CEO, Bootlin Embedded Linux and Kernel engineering and training https://bootlin.com _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
PATCH 2/5 package/ulog: new package
In reply to Herve CodinaThe ulog library is a minimalistic logging library derived from Android logger.
https://github.com/Parrot-Developers/ulog
--- DEVELOPERS | 1 + package/Config.in | 1 + package/ulog/Config.in | 12 ++++++++++++ package/ulog/ulog.hash | 3 +++ package/ulog/ulog.mk | 44 ++++++++++++++++++++++++++++++++++++++++++ 5 files changed, 61 insertions(+) create mode 100644 package/ulog/Config.in create mode 100644 package/ulog/ulog.hash create mode 100644 package/ulog/ulog.mk
DEVELOPERS +1 -0
@@ -1152,6 +1152,7 @@ F: package/dtbocfg/ F: package/libdbi/ F: package/libdbi-drivers/ F: package/lua-augeas/ +F: package/ulog/ F: support/testing/tests/package/test_dtbocfg.py F: support/testing/tests/package/test_lua_augeas.py
package/Config.in +1 -0
@@ -1704,6 +1704,7 @@ menu "Logging" source "package/log4qt/Config.in" source "package/opentracing-cpp/Config.in" source "package/spdlog/Config.in" + source "package/ulog/Config.in" source "package/zlog/Config.in" endmenu
package/ulog/Config.in new file +12 -0
@@ -0,0 +1,12 @@ +config BR2_PACKAGE_ULOG + bool "ulog" + depends on BR2_INSTALL_LIBSTDCPP + depends on BR2_TOOLCHAIN_HAS_THREADS + help + This is a minimalistic logging library derived from + Android logger. + + https://github.com/Parrot-Developers/ulog + +comment "ulog needs a toolchain w/ C++, threads" + depends on !BR2_INSTALL_LIBSTDCPP || !BR2_TOOLCHAIN_HAS_THREADS
package/ulog/ulog.hash new file +3 -0
@@ -0,0 +1,3 @@ +# Locally computed: +sha256 14e5773b32a79fa5380bdc7ac73a39a7cd3ab182830f57cf6f2994fb49cf38dc ulog-0389d243352255f6182326dccdae3d56dadc078f.tar.gz +sha256 cbb97dd2528af2aa2b9aee6c1b3ff1caed758044c17b2c811cf44b2528c496da COPYING
package/ulog/ulog.mk new file +44 -0
44 lines not on this page · raw message
Signature · 6 lines
2.31.1 _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
Re: [Buildroot] [PATCH 2/5] package/ulog: new package
package/ulog/ulog.mk line 16
Quoted 10 lines · On Wed, 3 Nov 2021 08:36:53 +0100 Herve Codina <[email protected]>…
package/ulog/ulog.mk new file +10
-0+ULOG_VERSION = 0389d243352255f6182326dccdae3d56dadc078f +ULOG_SITE = $(call github,Parrot-Developers,ulog,$(ULOG_VERSION)) +ULOG_LICENSE = Apache-2.0 +ULOG_LICENSE_FILES = COPYING +ULOG_DEPENDENCIES = host-alchemy +ULOG_INSTALL_STAGING = YES + +define ULOG_BUILD_CMDS + $(ALCHEMY_TARGET_CONFIGURE_ENV) \ + $(ALCHEMY_MAKE) libulog
Please indent the continuation line.
package/ulog/ulog.mk line 21
Quoted 3 lines
package/ulog/ulog.mk new file +3
-0+define ULOG_INSTALL_STATIC_LIBS + $(INSTALL) -m 644 $(@D)/alchemy-out/staging/usr/lib/libulog.a $(strip $(1))/usr/lib/ +endef
I am not entirely sure we need to $(strip ...) the $(1) argument here and in other macros. We have several other places in Buildroot for similar situations where we don't strip $(1). (Note: this is not a very important comment.)
Thanks,
Thomas
Signature · 7 lines
Thomas Petazzoni, co-owner and CEO, Bootlin Embedded Linux and Kernel engineering and training https://bootlin.com _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
Re: [Buildroot] [PATCH 2/5] package/ulog: new package
Hi,
Quoted 15 lines · On Wed, 3 Nov 2021 16:47:01 +0100 Thomas Petazzoni…
package/ulog/ulog.mk line 16
Quoted 10 lines · On Wed, 3 Nov 2021 08:36:53 +0100 Herve Codina <[email protected]>…
package/ulog/ulog.mk new file +10
-0+ULOG_VERSION = 0389d243352255f6182326dccdae3d56dadc078f +ULOG_SITE = $(call github,Parrot-Developers,ulog,$(ULOG_VERSION)) +ULOG_LICENSE = Apache-2.0 +ULOG_LICENSE_FILES = COPYING +ULOG_DEPENDENCIES = host-alchemy +ULOG_INSTALL_STAGING = YES + +define ULOG_BUILD_CMDS + $(ALCHEMY_TARGET_CONFIGURE_ENV) \ + $(ALCHEMY_MAKE) libulogPlease indent the continuation line.
Will be done in v2.
Quoted 9 lines
package/ulog/ulog.mk line 21
Quoted 4 lines
package/ulog/ulog.mk new file +3
-0+define ULOG_INSTALL_STATIC_LIBS + $(INSTALL) -m 644 $(@D)/alchemy-out/staging/usr/lib/libulog.a $(strip $(1))/usr/lib/ +endefI am not entirely sure we need to $(strip ...) the $(1) argument here and in other macros. We have several other places in Buildroot for similar situations where we don't strip $(1). (Note: this is not a very important comment.)
I will remove this kind of $(strip ...) in v2.
Quoted 4 lines
Thanks,
Thomas
Thanks fr the review.
Hervé
Signature · 7 lines
Hervé Codina, Bootlin Embedded Linux and Kernel engineering https://bootlin.com _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
PATCH 3/5 package/libfutils: new package
In reply to Herve CodinaThe libfutils library is a library that contains some common useful functions (list, hash, time).
https://github.com/Parrot-Developers/libfutils
--- DEVELOPERS | 1 + package/Config.in | 1 + package/libfutils/Config.in | 14 +++++++++ package/libfutils/libfutils.hash | 3 ++ package/libfutils/libfutils.mk | 50 ++++++++++++++++++++++++++++++++ 5 files changed, 69 insertions(+) create mode 100644 package/libfutils/Config.in create mode 100644 package/libfutils/libfutils.hash create mode 100644 package/libfutils/libfutils.mk
DEVELOPERS +1 -0
@@ -1151,6 +1151,7 @@ F: package/alchemy/ F: package/dtbocfg/ F: package/libdbi/ F: package/libdbi-drivers/ +F: package/libfutils/ F: package/lua-augeas/ F: package/ulog/ F: support/testing/tests/package/test_dtbocfg.py
package/Config.in +1 -0
@@ -1947,6 +1947,7 @@ menu "Other" source "package/libevdev/Config.in" source "package/libevent/Config.in" source "package/libffi/Config.in" + source "package/libfutils/Config.in" source "package/libgee/Config.in" source "package/libgeos/Config.in" source "package/libglib2/Config.in"
package/libfutils/Config.in new file +14 -0
@@ -0,0 +1,14 @@ +config BR2_PACKAGE_LIBFUTILS + bool "libfutils" + depends on BR2_TOOLCHAIN_HAS_ATOMIC + depends on BR2_INSTALL_LIBSTDCPP # ulog + depends on BR2_TOOLCHAIN_HAS_THREADS # ulog + select BR2_PACKAGE_ULOG + help + This library contains some common useful functions (list, + hash, time) + + https://github.com/Parrot-Developers/libfutils + +comment "libfutils needs a toolchain w/ C++, threads" + depends on !BR2_INSTALL_LIBSTDCPP || !BR2_TOOLCHAIN_HAS_THREADS
package/libfutils/libfutils.hash new file +3 -0
@@ -0,0 +1,3 @@ +# Locally computed: +sha256 b38d40c46b83e448d8c1716ddfb04b7ecd384f5cd2cb773db5c26c0589accdb3 libfutils-c326ce0cc2a7354381265a7664bb215be587fac6.tar.gz +sha256 eca11a1c62fae8d64a1247d93271375baecff8396141d7ff8d89671ee7e3ed0d COPYING
package/libfutils/libfutils.mk new file +50 -0
50 lines not on this page · raw message
Signature · 6 lines
2.31.1 _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
Re: [Buildroot] [PATCH 3/5] package/libfutils: new package
Hello,
On Wed, 3 Nov 2021 08:36:54 +0100
Herve Codina <[email protected]> wrote:
"
Quoted 9 lines
package/libfutils/Config.in new file +3
-0@@ -0,0 +1,14 @@ +config BR2_PACKAGE_LIBFUTILS + bool "libfutils" + depends on BR2_TOOLCHAIN_HAS_ATOMIC
package/libfutils/Config.in line 13
Quoted 10 lines
package/libfutils/Config.in new file +10
-0+ depends on BR2_INSTALL_LIBSTDCPP # ulog + depends on BR2_TOOLCHAIN_HAS_THREADS # ulog + select BR2_PACKAGE_ULOG + help + This library contains some common useful functions (list, + hash, time) + + https://github.com/Parrot-Developers/libfutils + +comment "libfutils needs a toolchain w/ C++, threads"
You need:
depends on BR2_TOOLCHAIN_HAS_ATOMIC
Quoted 28 lines
package/libfutils/Config.in new file +1
-0+ depends on !BR2_INSTALL_LIBSTDCPP || !BR2_TOOLCHAIN_HAS_THREADSpackage/libfutils/libfutils.hash new file +3
-0@@ -0,0 +1,3 @@ +# Locally computed: +sha256 b38d40c46b83e448d8c1716ddfb04b7ecd384f5cd2cb773db5c26c0589accdb3 libfutils-c326ce0cc2a7354381265a7664bb215be587fac6.tar.gz +sha256 eca11a1c62fae8d64a1247d93271375baecff8396141d7ff8d89671ee7e3ed0d COPYINGpackage/libfutils/libfutils.mk new file +12
-0@@ -0,0 +1,50 @@ +################################################################################ +# +# libfutils +# +################################################################################ + +LIBFUTILS_VERSION = c326ce0cc2a7354381265a7664bb215be587fac6 +LIBFUTILS_SITE = $(call github,Parrot-Developers,libfutils,$(LIBFUTILS_VERSION)) +LIBFUTILS_LICENSE = BSD-3-Clause +LIBFUTILS_LICENSE_FILES = COPYING +LIBFUTILS_DEPENDENCIES = ulog host-alchemy +LIBFUTILS_INSTALL_STAGING = YES
package/libfutils/libfutils.mk line 22
Quoted 10 lines
package/libfutils/libfutils.mk new file +10
-0+ +ifeq ($(BR2_TOOLCHAIN_HAS_LIBATOMIC),y) +LIBFUTILS_CONF_ENV += TARGET_GLOBAL_LDLIBS='-latomic' +endif + +define LIBFUTILS_BUILD_CMDS + $(ALCHEMY_TARGET_CONFIGURE_ENV) \ + $(call ALCHEMY_TARGET_CONFIGURE_SDKS, ulog) \ + $(LIBFUTILS_CONF_ENV) \ + $(ALCHEMY_MAKE) libfutils
Indentation of continuation lines.
Thomas
Signature · 7 lines
Thomas Petazzoni, co-owner and CEO, Bootlin Embedded Linux and Kernel engineering and training https://bootlin.com _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
Re: [Buildroot] [PATCH 3/5] package/libfutils: new package
Hi,
Quoted 27 lines · On Wed, 3 Nov 2021 16:47:56 +0100 Thomas Petazzoni…
Hello,
On Wed, 3 Nov 2021 08:36:54 +0100
Herve Codina <[email protected]> wrote:
"Quoted 9 lines
package/libfutils/Config.in new file +3
-0@@ -0,0 +1,14 @@ +config BR2_PACKAGE_LIBFUTILS + bool "libfutils" + depends on BR2_TOOLCHAIN_HAS_ATOMICpackage/libfutils/Config.in line 13
Quoted 10 lines
package/libfutils/Config.in new file +10
-0+ depends on BR2_INSTALL_LIBSTDCPP # ulog + depends on BR2_TOOLCHAIN_HAS_THREADS # ulog + select BR2_PACKAGE_ULOG + help + This library contains some common useful functions (list, + hash, time) + + https://github.com/Parrot-Developers/libfutils + +comment "libfutils needs a toolchain w/ C++, threads"You need:
Quoted 1 line
depends on
BR2_TOOLCHAIN_HAS_ATOMIC
Indeed.
Will be added in v2.
Quoted 33 lines
Quoted 33 lines
package/libfutils/Config.in new file +1
-0+ depends on !BR2_INSTALL_LIBSTDCPP || !BR2_TOOLCHAIN_HAS_THREADSpackage/libfutils/libfutils.hash new file +3
-0@@ -0,0 +1,3 @@ +# Locally computed: +sha256 b38d40c46b83e448d8c1716ddfb04b7ecd384f5cd2cb773db5c26c0589accdb3 libfutils-c326ce0cc2a7354381265a7664bb215be587fac6.tar.gz +sha256 eca11a1c62fae8d64a1247d93271375baecff8396141d7ff8d89671ee7e3ed0d COPYINGpackage/libfutils/libfutils.mk new file +16
-0@@ -0,0 +1,50 @@ +################################################################################ +# +# libfutils +# +################################################################################ + +LIBFUTILS_VERSION = c326ce0cc2a7354381265a7664bb215be587fac6 +LIBFUTILS_SITE = $(call github,Parrot-Developers,libfutils,$(LIBFUTILS_VERSION)) +LIBFUTILS_LICENSE = BSD-3-Clause +LIBFUTILS_LICENSE_FILES = COPYING +LIBFUTILS_DEPENDENCIES = ulog host-alchemy +LIBFUTILS_INSTALL_STAGING = YES + +ifeq ($(BR2_TOOLCHAIN_HAS_LIBATOMIC),y) +LIBFUTILS_CONF_ENV += TARGET_GLOBAL_LDLIBS='-latomic' +endif
Quoted 8 lines
package/libfutils/libfutils.mk line 22
Quoted 6 lines
package/libfutils/libfutils.mk new file +6
-0+ +define LIBFUTILS_BUILD_CMDS + $(ALCHEMY_TARGET_CONFIGURE_ENV) \ + $(call ALCHEMY_TARGET_CONFIGURE_SDKS, ulog) \ + $(LIBFUTILS_CONF_ENV) \ + $(ALCHEMY_MAKE) libfutilsIndentation of continuation lines.
Will be changed in v2.
Quoted 2 lines
Thomas
Thanks for the review.
Hervé
Signature · 7 lines
Hervé Codina, Bootlin Embedded Linux and Kernel engineering https://bootlin.com _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
PATCH 4/5 package/libshdata: new package
In reply to Herve CodinaThe libshdata library provides lock free shared-memory tools.
https://github.com/Parrot-Developers/libshdata
The provided patch was submitted upstream.
https://github.com/Parrot-Developers/libshdata/issues/1
--- DEVELOPERS | 1 + package/Config.in | 1 + ...01-backend-Add-missing-include-files.patch | 44 ++++++++++++++ package/libshdata/Config.in | 23 ++++++++ package/libshdata/libshdata.hash | 3 + package/libshdata/libshdata.mk | 59 +++++++++++++++++++ 6 files changed, 131 insertions(+) create mode 100644 package/libshdata/0001-backend-Add-missing-include-files.patch create mode 100644 package/libshdata/Config.in create mode 100644 package/libshdata/libshdata.hash create mode 100644 package/libshdata/libshdata.mk
DEVELOPERS +1 -0
@@ -1152,6 +1152,7 @@ F: package/dtbocfg/ F: package/libdbi/ F: package/libdbi-drivers/ F: package/libfutils/ +F: package/libshdata/ F: package/lua-augeas/ F: package/ulog/ F: support/testing/tests/package/test_dtbocfg.py
package/Config.in +1 -0
@@ -1988,6 +1988,7 @@ comment "linux-pam plugins" source "package/libpam-tacplus/Config.in" endif source "package/liquid-dsp/Config.in" + source "package/libshdata/Config.in" source "package/llvm/Config.in" source "package/lttng-libust/Config.in" source "package/matio/Config.in"
package/libshdata/0001-backend-Add-missing-include-files.patch new file +44 -0
44 lines not on this page · raw message
package/libshdata/Config.in new file +23 -0
@@ -0,0 +1,23 @@ +config BR2_PACKAGE_LIBSHDATA + bool "libshdata" + depends on BR2_TOOLCHAIN_HAS_SYNC_4 + depends on BR2_TOOLCHAIN_HAS_ATOMIC # libfutils + depends on BR2_INSTALL_LIBSTDCPP # libfutils + depends on BR2_TOOLCHAIN_HAS_THREADS # libfutils + select BR2_PACKAGE_LIBFUTILS + help + This library provides lock free shared-memory tools + + https://github.com/Parrot-Developers/libshdata + +comment "libshdata needs a toolchain w/ C++, threads" + depends on !BR2_INSTALL_LIBSTDCPP || !BR2_TOOLCHAIN_HAS_THREADS + +if BR2_PACKAGE_LIBSHDATA + +config BR2_PACKAGE_LIBSHDATA_STRESS + bool "libshdata-stress binary" + help + Install libshdata-stress binary as well + +endif
package/libshdata/libshdata.hash new file +3 -0
@@ -0,0 +1,3 @@ +# Locally computed: +sha256 1ae83ac73c2b5b17c726067737b8d446b99f190e16fe6198eb464368796ecce3 libshdata-d9ec4bdba834d8f3daf6bf9aa6da374bc462961f.tar.gz +sha256 eca11a1c62fae8d64a1247d93271375baecff8396141d7ff8d89671ee7e3ed0d COPYING
package/libshdata/libshdata.mk new file +59 -0
59 lines not on this page · raw message
Signature · 6 lines
2.31.1 _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
Re: [Buildroot] [PATCH 4/5] package/libshdata: new package
Hello,
Quoted 6 lines · On Wed, 3 Nov 2021 08:36:55 +0100 Herve Codina <[email protected]>…
The libshdata library provides lock free shared-memory tools.
https://github.com/Parrot-Developers/libshdata
The provided patch was submitted upstream.
https://github.com/Parrot-Developers/libshdata/issues/1
This should rather go in the patch itself.
Quoted 10 lines
package/libshdata/Config.in new file +4
-0@@ -0,0 +1,23 @@ +config BR2_PACKAGE_LIBSHDATA + bool "libshdata" + depends on BR2_TOOLCHAIN_HAS_SYNC_4 + depends on BR2_TOOLCHAIN_HAS_ATOMIC # libfutils
package/libshdata/Config.in line 14
Quoted 10 lines
package/libshdata/Config.in new file +10
-0+ depends on BR2_INSTALL_LIBSTDCPP # libfutils + depends on BR2_TOOLCHAIN_HAS_THREADS # libfutils + select BR2_PACKAGE_LIBFUTILS + help + This library provides lock free shared-memory tools + + https://github.com/Parrot-Developers/libshdata + +comment "libshdata needs a toolchain w/ C++, threads" + depends on !BR2_INSTALL_LIBSTDCPP || !BR2_TOOLCHAIN_HAS_THREADS
This needs to go *after* the if ... endif block, or the indentation of the if ... endif block is broken in menuconfig. Also, you need to account for the HAS_SYNC_4 and HAS_ATOMIC dependencies.
package/libshdata/Config.in line 19
Quoted 5 lines
package/libshdata/Config.in new file +5
-0+define LIBSHDATA_BUILD_CMDS + $(ALCHEMY_TARGET_CONFIGURE_ENV) \ + $(call ALCHEMY_TARGET_CONFIGURE_SDKS, libfutils) \ + $(LIBSHDATA_CONF_ENV) \ + $(ALCHEMY_MAKE) $(LIBSHDATA_TARGETS)
Indentation.
Quoted 4 lines
package/libshdata/Config.in new file +4
-0+endef + +define LIBSHDATA_INSTALL_STATIC_LIBS + $(INSTALL) -m 644 $(@D)/alchemy-out/staging/usr/lib/libshdata.a $(strip $(1))/usr/lib/
package/libshdata/Config.in
Quoted 10 lines
package/libshdata/Config.in new file +10
-0+ $(INSTALL) -m 644 $(@D)/alchemy-out/staging/usr/lib/libshdata-section-lookup.a $(strip $(1))/usr/lib/ +endef + +define LIBSHDATA_INSTALL_HEADERS + cp -Raf $(@D)/include/* $(strip $(1))/usr/include/ +endef + +ifeq ($(BR2_STATIC_LIBS),) +define LIBSHDATA_INSTALL_SHARED_LIBS + $(INSTALL) -m 755 $(@D)/alchemy-out/staging/usr/lib/libshdata.so* $(strip $(1))/usr/lib/
Sot here are two separate static libraries, but only one shared library ?
Thomas
Signature · 7 lines
Thomas Petazzoni, co-owner and CEO, Bootlin Embedded Linux and Kernel engineering and training https://bootlin.com _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
Re: [Buildroot] [PATCH 4/5] package/libshdata: new package
Hi,
Quoted 5 lines · On Wed, 3 Nov 2021 16:50:00 +0100 Thomas Petazzoni…
Hello,
On Wed, 3 Nov 2021 08:36:55 +0100
Herve Codina <[email protected]> wrote:
Quoted 8 lines
Quoted 6 lines
The libshdata library provides lock free shared-memory tools.
https://github.com/Parrot-Developers/libshdata
The provided patch was submitted upstream.
https://github.com/Parrot-Developers/libshdata/issues/1This should rather go in the patch itself.
Will be moved to the patch in v2.
Quoted 18 lines
Quoted 18 lines
package/libshdata/Config.in new file +10
-0@@ -0,0 +1,23 @@ +config BR2_PACKAGE_LIBSHDATA + bool "libshdata" + depends on BR2_TOOLCHAIN_HAS_SYNC_4 + depends on BR2_TOOLCHAIN_HAS_ATOMIC # libfutils + depends on BR2_INSTALL_LIBSTDCPP # libfutils + depends on BR2_TOOLCHAIN_HAS_THREADS # libfutils + select BR2_PACKAGE_LIBFUTILS + help + This library provides lock free shared-memory tools +
Quoted 8 lines
package/libshdata/Config.in line 14
Quoted 4 lines
package/libshdata/Config.in new file +4
-0+ https://github.com/Parrot-Developers/libshdata + +comment "libshdata needs a toolchain w/ C++, threads" + depends on !BR2_INSTALL_LIBSTDCPP || !BR2_TOOLCHAIN_HAS_THREADSThis needs to go *after* the if ... endif block, or the indentation of the if ... endif block is broken in menuconfig. Also, you need to account for the
HAS_SYNC_4andHAS_ATOMICdependencies.
Ok,
Will be changed in v2.
Quoted 9 lines
package/libshdata/Config.in line 19
Quoted 7 lines
package/libshdata/Config.in new file +5
-0+define LIBSHDATA_BUILD_CMDS + $(ALCHEMY_TARGET_CONFIGURE_ENV) \ + $(call ALCHEMY_TARGET_CONFIGURE_SDKS, libfutils) \ + $(LIBSHDATA_CONF_ENV) \ + $(ALCHEMY_MAKE) $(LIBSHDATA_TARGETS)Indentation.
Will be changed in v2.
Quoted 17 lines
Quoted 5 lines
package/libshdata/Config.in new file +4
-0+endef + +define LIBSHDATA_INSTALL_STATIC_LIBS + $(INSTALL) -m 644 $(@D)/alchemy-out/staging/usr/lib/libshdata.a $(strip $(1))/usr/lib/package/libshdata/Config.in
Quoted 10 lines
package/libshdata/Config.in new file +10
-0+ $(INSTALL) -m 644 $(@D)/alchemy-out/staging/usr/lib/libshdata-section-lookup.a $(strip $(1))/usr/lib/ +endef + +define LIBSHDATA_INSTALL_HEADERS + cp -Raf $(@D)/include/* $(strip $(1))/usr/include/ +endef + +ifeq ($(BR2_STATIC_LIBS),) +define LIBSHDATA_INSTALL_SHARED_LIBS + $(INSTALL) -m 755 $(@D)/alchemy-out/staging/usr/lib/libshdata.so* $(strip $(1))/usr/lib/Sot here are two separate static libraries, but only one shared library ?
Yes, that's what libshdata provides.
libshdata-section-lookup is only a static lib (include $(BUILD_STATIC_LIBRARY) in atom.mk).
Quoted 2 lines
Thomas
Thanks for the review.
Hervé
Signature · 7 lines
Hervé Codina, Bootlin Embedded Linux and Kernel engineering https://bootlin.com _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot
PATCH 5/5 support/testing/tests/package/test_libshdata: new test
In reply to Herve CodinaThis new test ensures that libraries and binaries generated using Parrot Alchemy build system are correct. Indeed, the test uses libshdata-stress.
This binary depends on libshdata.
libshdata depends on libfutils and libfutils depends on ulog. All of these binaries and libraries are built using Alchemy.
--- DEVELOPERS | 1 + .../testing/tests/package/test_libshdata.py | 25 +++++++++++++++++++ 2 files changed, 26 insertions(+) create mode 100644 support/testing/tests/package/test_libshdata.py
DEVELOPERS +1 -0
@@ -1156,6 +1156,7 @@ F: package/libshdata/ F: package/lua-augeas/ F: package/ulog/ F: support/testing/tests/package/test_dtbocfg.py +F: support/testing/tests/package/test_libshdata.py F: support/testing/tests/package/test_lua_augeas.py N: Hervé Codina <[email protected]>
support/testing/tests/package/test_libshdata.py new file +25 -0
@@ -0,0 +1,25 @@ +import os + +import infra.basetest + + +class TestLibshdata(infra.basetest.BRTest): + config = infra.basetest.BASIC_TOOLCHAIN_CONFIG + \ + """ + BR2_PACKAGE_LIBSHDATA=y + BR2_PACKAGE_LIBSHDATA_STRESS=y + BR2_TARGET_ROOTFS_CPIO=y + # BR2_TARGET_ROOTFS_TAR is not set + """ + + def test_run(self): + img = os.path.join(self.builddir, "images", "rootfs.cpio") + self.emulator.boot(arch="armv5", + kernel="builtin", + options=["-initrd", img]) + self.emulator.login() + + # Just run libshdata-stress. + # This ensures that library are well compiled and that all dependencies + # are met using Parrot Alchemy build system. + self.assertRunOk("libshdata-stress")
Signature · 6 lines
2.31.1 _______________________________________________ buildroot mailing list [email protected] https://lists.buildroot.org/mailman/listinfo/buildroot