nodejs: ensure to use correct compiler & flags always
PATCH 1/5 nodejs: ensure to use correct compiler & flags always
NodeJS comes with an embedded, patched, version of gyp.
Normally, gyp supports compiling for the build machine, e.g. native tools that need to be compiled to run during the build, and for the host, using different variables, e.g. CC and CC.host, etc.
Most of this has been patched out in the NodeJS version of gyp, and essentially it only supports compiling using one compiler - ${CC}. This modification excludes LDFLAGS for native tools, and those still evaluate LDFLAGS.host (only).
While this modified behaviour is OK for the OE use-case of building native and target tools separately, it means that this recipe can not work as-is with standard gyp, and wrong LDFLAGS are being used for some of the tools compiled (torque) in either case.
By setting the make variables that gyp-generated makefiles inspect, we support use of unpatched gyp, and we ensure that all tools are compiled with correct LDFLAGS in either case.
This now also allows us to drop the patch that had been applied to work-around this problem.
--- .../nodejs/0006-Use-target-ldflags.patch | 24 ------------------- .../recipes-devtools/nodejs/nodejs_10.17.0.bb | 13 +++++++++- 2 files changed, 12 insertions(+), 25 deletions(-) delete mode 100644 meta-oe/recipes-devtools/nodejs/nodejs/0006-Use-target-ldflags.patch
meta-oe/recipes-devtools/nodejs/nodejs/0006-Use-target-ldflags.patch deleted +0 -24
@@ -1,24 +0,0 @@ -The target LDFLAGS have been ignored. Tools like torque -have been loaded from system libraries, even if a native -one was the target. -|$ ldd torque -| libz.so.1 => /lib/x86_64-linux-gnu/libz.so.1 -| libcrypto.so.1.1 => /usr/lib/x86_64-linux-gnu/libcrypto.so.1.1 -| libssl.so.1.1 => /usr/lib/x86_64-linux-gnu/libssl.so.1.1 -| libicui18n.so.63 => not found -| libicuuc.so.63 => not found -... - -Signed-off-by: Andrej Valek <[email protected]> - -diff -Naur node-v10.15.1/deps/v8/gypfiles/toolchain.gypi node-v10.15.1/deps/v8/gypfiles/toolchain.gypi ---- node-v10.15.1/deps/v8/gypfiles/toolchain.gypi 2019-03-18 15:01:39.000000000 +0100 -+++ node-v10.15.1/deps/v8/gypfiles/toolchain.gypi 2019-03-18 15:04:08.628361308 +0100 -@@ -1106,6 +1106,7 @@ - 'cflags': [ '-fno-strict-aliasing' ], - }], - ], # conditions -+ 'ldflags+': [ '$(LDFLAGS)' ], - }], - ['OS=="solaris"', { - 'defines': [ '__C99FEATURES__=1' ], # isinf() etc.
meta-oe/recipes-devtools/nodejs/nodejs_10.17.0.bb +12 -1
@@ -19,7 +19,6 @@ SRC_URI = "http://nodejs.org/dist/v${PV}/node-v${PV}.tar.xz \ file://0001-Disable-running-gyp-files-for-bundled-deps.patch \ file://0004-Make-compatibility-with-gcc-4.8.patch \ file://0005-Link-atomic-library.patch \ - file://0006-Use-target-ldflags.patch \ " SRC_URI_append_class-target = " \ file://0002-Using-native-torque.patch \
@@ -58,6 +57,18 @@ PACKAGECONFIG[libuv] = "--shared-libuv,,libuv" PACKAGECONFIG[nghttp2] = "--shared-nghttp2,,nghttp2" PACKAGECONFIG[zlib] = "--shared-zlib,,zlib" +# We don't want to cross-compile during target compile, +# and we need to use the right flags during host compile, +# too. +EXTRA_OEMAKE_append = "\ + CC.host='${CC}' \ + CFLAGS.host='${CPPFLAGS} ${CFLAGS}' \ + CXX.host='${CXX}' \ + CXXFLAGS.host='${CPPFLAGS} ${CXXFLAGS}' \ + LDFLAGS.host='${LDFLAGS}' \ + AR.host='${AR}' \ +" + # Node is way too cool to use proper autotools, so we install two wrappers to forcefully inject proper arch cflags to workaround gypi do_configure () { rm -rf ${S}/deps/openssl
Signature · 1 line
2.23.0.rc1
PATCH 2/5 nodejs: delete all bundled deps in do_unpack() if needed
We can delete bundled deps where system-provided counterparts should be used instead.
Amongst others, this ensures they are not used accidentally.
--- .../recipes-devtools/nodejs/nodejs_10.17.0.bb | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-)
meta-oe/recipes-devtools/nodejs/nodejs_10.17.0.bb +16 -1
@@ -69,9 +69,24 @@ EXTRA_OEMAKE_append = "\ AR.host='${AR}' \ " +python do_unpack() { + import shutil + + bb.build.exec_func('base_do_unpack', d) + + shutil.rmtree(d.getVar('S') + '/deps/openssl', True) + if 'ares' in d.getVar('PACKAGECONFIG'): + shutil.rmtree(d.getVar('S') + '/deps/cares', True) + if 'libuv' in d.getVar('PACKAGECONFIG'): + shutil.rmtree(d.getVar('S') + '/deps/uv', True) + if 'nghttp2' in d.getVar('PACKAGECONFIG'): + shutil.rmtree(d.getVar('S') + '/deps/nghttp2', True) + if 'zlib' in d.getVar('PACKAGECONFIG'): + shutil.rmtree(d.getVar('S') + '/deps/zlib', True) +} + # Node is way too cool to use proper autotools, so we install two wrappers to forcefully inject proper arch cflags to workaround gypi do_configure () { - rm -rf ${S}/deps/openssl export LD="${CXX}" GYP_DEFINES="${GYP_DEFINES}" export GYP_DEFINES # $TARGET_ARCH settings don't match --dest-cpu settings
Signature · 1 line
2.23.0.rc1
PATCH 3/5 nodejs: allow use of system gyp
In reply to André DraszikNow that there is a gyp (python2) recipe in meta-python, allow its use instead of the bundled one for when meta-python is enabled.
At the same time, unconditionally inherit pythonnative as in either case gyp actually uses python, and the build machine's python and installed modules should never be used.
--- meta-oe/recipes-devtools/nodejs/nodejs_10.17.0.bb | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
meta-oe/recipes-devtools/nodejs/nodejs_10.17.0.bb +4 -1
@@ -6,7 +6,7 @@ LIC_FILES_CHKSUM = "file://LICENSE;md5=be980eb7ccafe287cb438076a65e888c" DEPENDS = "openssl" DEPENDS_append_class-target = " nodejs-native" -inherit pkgconfig +inherit pkgconfig pythonnative COMPATIBLE_MACHINE_armv4 = "(!.*armv4).*" COMPATIBLE_MACHINE_armv5 = "(!.*armv5).*"
@@ -52,6 +52,7 @@ ARCHFLAGS ?= "" PACKAGECONFIG ??= "ares icu libuv zlib" PACKAGECONFIG[ares] = "--shared-cares,,c-ares" +PACKAGECONFIG[gyp] = ",,gyp-py2-native" PACKAGECONFIG[icu] = "--with-intl=system-icu,--without-intl,icu" PACKAGECONFIG[libuv] = "--shared-libuv,,libuv" PACKAGECONFIG[nghttp2] = "--shared-nghttp2,,nghttp2"
@@ -77,6 +78,8 @@ python do_unpack() { shutil.rmtree(d.getVar('S') + '/deps/openssl', True) if 'ares' in d.getVar('PACKAGECONFIG'): shutil.rmtree(d.getVar('S') + '/deps/cares', True) + if 'gyp' in d.getVar('PACKAGECONFIG'): + shutil.rmtree(d.getVar('S') + '/tools/gyp', True) if 'libuv' in d.getVar('PACKAGECONFIG'): shutil.rmtree(d.getVar('S') + '/deps/uv', True) if 'nghttp2' in d.getVar('PACKAGECONFIG'):
Signature · 1 line
2.23.0.rc1
PATCH 4/5 nodejs: use OE-provided compiler flags (arm)
In reply to André DraszikThis overrides yocto-provided build flags with its own, e.g we get
arm-poky-linux-musleabi-g++ -mthumb -mfpu=neon -mfloat-abi=hard -mcpu=cortex-a7 \
... \
-march=armv7-a -mfpu=neon -mfloat-abi=hard -marmCausing the latter to override the former, and compiler warnings:
cc1plus: warning: switch '-mcpu=cortex-a7' conflicts with '-march=armv7-a' switch
Patch this out, so that yocto-provided flags take precedence. Note that in reality the same should probably be done for all the other supported architectures, too.
Note that this also switches to Thumb(2) mode (in my case). No obvious problems have been noted during compilation or runtime.
Upstream-Status: Inappropriate [oe-specific]
--- .../0007-v8-don-t-override-ARM-CFLAGS.patch | 102 ++++++++++++++++++ .../recipes-devtools/nodejs/nodejs_10.17.0.bb | 1 + 2 files changed, 103 insertions(+) create mode 100644 meta-oe/recipes-devtools/nodejs/nodejs/0007-v8-don-t-override-ARM-CFLAGS.patch
meta-oe/recipes-devtools/nodejs/nodejs/0007-v8-don-t-override-ARM-CFLAGS.patch new file +102 -0
@@ -0,0 +1,102 @@ +From 47ee5cc5501289205d3e8e9f27ea9daf18cebac1 Mon Sep 17 00:00:00 2001 +From: =?UTF-8?q?Andr=C3=A9=20Draszik?= <[email protected]> +Date: Sat, 9 Nov 2019 14:45:30 +0000 +Subject: [PATCH] v8: don't override ARM CFLAGS +MIME-Version: 1.0 +Content-Type: text/plain; charset=UTF-8 +Content-Transfer-Encoding: 8bit + +This overrides yocto-provided build flags with its own, e.g we get + arm-poky-linux-musleabi-g++ -mthumb -mfpu=neon -mfloat-abi=hard -mcpu=cortex-a7 \ + ... \ + -march=armv7-a -mfpu=neon -mfloat-abi=hard -marm + +Causing the latter to override the former, and compiler warnings: + cc1plus: warning: switch '-mcpu=cortex-a7' conflicts with '-march=armv7-a' switch + +Patch this out, so that yocto-provided flags take precedence. +Note that in reality the same should probably be done for all the other +supported architectures, too. + +Note that this also switches to Thumb(2) mode (in my case). No obvious +problems have been noted during compilation or runtime. + +Upstream-Status: Inappropriate [oe-specific] +Signed-off-by: André Draszik <[email protected]> +--- + deps/v8/gypfiles/toolchain.gypi | 52 ++------------------------------- + 1 file changed, 2 insertions(+), 50 deletions(-) + +diff --git a/deps/v8/gypfiles/toolchain.gypi b/deps/v8/gypfiles/toolchain.gypi +index 910a212..1390b15 100644 +--- a/deps/v8/gypfiles/toolchain.gypi ++++ b/deps/v8/gypfiles/toolchain.gypi +@@ -199,31 +199,7 @@ + 'target_conditions': [ + ['_toolset=="host"', { + 'conditions': [ +- ['v8_target_arch==host_arch', { +- # Host built with an Arm CXX compiler. +- 'conditions': [ +- [ 'arm_version==7', { +- 'cflags': ['-march=armv7-a',], +- }], +- [ 'arm_version==7 or arm_version=="default"', { +- 'conditions': [ +- [ 'arm_fpu!="default"', { +- 'cflags': ['-mfpu=<(arm_fpu)',], +- }], +- ], +- }], +- [ 'arm_float_abi!="default"', { +- 'cflags': ['-mfloat-abi=<(arm_float_abi)',], +- }], +- [ 'arm_thumb==1', { +- 'cflags': ['-mthumb',], +- }], +- [ 'arm_thumb==0', { +- 'cflags': ['-marm',], +- }], +- ], +- }, { +- # 'v8_target_arch!=host_arch' ++ ['v8_target_arch!=host_arch', { + # Host not built with an Arm CXX compiler (simulator build). + 'conditions': [ + [ 'arm_float_abi=="hard"', { +@@ -242,31 +218,7 @@ + }], # _toolset=="host" + ['_toolset=="target"', { + 'conditions': [ +- ['v8_target_arch==target_arch', { +- # Target built with an Arm CXX compiler. +- 'conditions': [ +- [ 'arm_version==7', { +- 'cflags': ['-march=armv7-a',], +- }], +- [ 'arm_version==7 or arm_version=="default"', { +- 'conditions': [ +- [ 'arm_fpu!="default"', { +- 'cflags': ['-mfpu=<(arm_fpu)',], +- }], +- ], +- }], +- [ 'arm_float_abi!="default"', { +- 'cflags': ['-mfloat-abi=<(arm_float_abi)',], +- }], +- [ 'arm_thumb==1', { +- 'cflags': ['-mthumb',], +- }], +- [ 'arm_thumb==0', { +- 'cflags': ['-marm',], +- }], +- ], +- }, { +- # 'v8_target_arch!=target_arch' ++ ['v8_target_arch!=target_arch', { + # Target not built with an Arm CXX compiler (simulator build). + 'conditions': [ + [ 'arm_float_abi=="hard"', { +-- +2.23.0.rc1 +
meta-oe/recipes-devtools/nodejs/nodejs_10.17.0.bb +1 -0
@@ -19,6 +19,7 @@ SRC_URI = "http://nodejs.org/dist/v${PV}/node-v${PV}.tar.xz \ file://0001-Disable-running-gyp-files-for-bundled-deps.patch \ file://0004-Make-compatibility-with-gcc-4.8.patch \ file://0005-Link-atomic-library.patch \ + file://0007-v8-don-t-override-ARM-CFLAGS.patch \ " SRC_URI_append_class-target = " \ file://0002-Using-native-torque.patch \
Signature · 1 line
2.23.0.rc1
PATCH 5/5 nodejs: support long directory names for ${B} / ${S}
In reply to André DraszikPart of the NodeJS build builds V8, which at some stage tries to call ar with all objects with *absolute* paths (and printf the command line first).
This will fail if the build path is too long:
make[1]: execvp: printf: Argument list too long
when trying to create Release/obj.target/deps/v8/gypfiles/libv8_base.a via below gyp-generated out/Makefile rule:
cmd_alink_thin = rm -f $@ && $(AR.$(TOOLSET)) crsT $@ $(filter %.o,$^)
i.e. something like
printf rm -f Release/obj.target/deps/v8/gypfiles/libv8_base.a && arm-poky-linux-musleabi-gcc-ar crsT Release/obj.target/deps/v8/gypfiles/libv8_base.a ...The above failure happened on a build-directory S with 204 characters on a Jenkins machine.
While one could probably increase the ulimit on that specific machine, that would be a pretty specific build machine fix which would need to be applied everywhere, or switch to non-verbose builds / compilation, but fortunately we can change all object references to be relative to the build directory itself by setting the builddir_name make variable and thus avoid the other two possible work-arounds.
--- meta-oe/recipes-devtools/nodejs/nodejs_10.17.0.bb | 2 ++ 1 file changed, 2 insertions(+)
meta-oe/recipes-devtools/nodejs/nodejs_10.17.0.bb +2 -0
@@ -69,6 +69,8 @@ EXTRA_OEMAKE_append = "\ CXXFLAGS.host='${CPPFLAGS} ${CXXFLAGS}' \ LDFLAGS.host='${LDFLAGS}' \ AR.host='${AR}' \ + \ + builddir_name=./ \ " python do_unpack() {
Signature · 1 line
2.23.0.rc1
PATCH 1/1 nodejs: ensure to use correct compiler & flags always
In reply to André DraszikNodeJS comes with an embedded, patched, version of gyp.
Normally, gyp supports compiling for the build machine, e.g. native tools that need to be compiled to run during the build, and for the host, using different variables, e.g. CC and CC.host, etc.
Most of this has been patched out in the NodeJS version of gyp, and essentially it only supports compiling using one compiler - ${CC}. This modification excludes LDFLAGS for native tools, and those still evaluate LDFLAGS.host (only).
While this modified behaviour is OK for the OE use-case of building native and target tools separately, it means that this recipe can not work as-is with standard gyp, and wrong LDFLAGS are being used for some of the tools compiled (torque) in either case.
By setting the make variables that gyp-generated makefiles inspect, we support use of unpatched gyp, and we ensure that all tools are compiled with correct LDFLAGS in either case.
This now also allows us to drop the patch that had been applied to work-around this problem.
---
v2: _append not needed for EXTRA_OEMAKE
--- .../nodejs/0006-Use-target-ldflags.patch | 24 ------------------- .../recipes-devtools/nodejs/nodejs_10.17.0.bb | 13 +++++++++- 2 files changed, 12 insertions(+), 25 deletions(-) delete mode 100644 meta-oe/recipes-devtools/nodejs/nodejs/0006-Use-target-ldflags.patch
meta-oe/recipes-devtools/nodejs/nodejs/0006-Use-target-ldflags.patch deleted +0 -24
@@ -1,24 +0,0 @@ -The target LDFLAGS have been ignored. Tools like torque -have been loaded from system libraries, even if a native -one was the target. -|$ ldd torque -| libz.so.1 => /lib/x86_64-linux-gnu/libz.so.1 -| libcrypto.so.1.1 => /usr/lib/x86_64-linux-gnu/libcrypto.so.1.1 -| libssl.so.1.1 => /usr/lib/x86_64-linux-gnu/libssl.so.1.1 -| libicui18n.so.63 => not found -| libicuuc.so.63 => not found -... - -Signed-off-by: Andrej Valek <[email protected]> - -diff -Naur node-v10.15.1/deps/v8/gypfiles/toolchain.gypi node-v10.15.1/deps/v8/gypfiles/toolchain.gypi ---- node-v10.15.1/deps/v8/gypfiles/toolchain.gypi 2019-03-18 15:01:39.000000000 +0100 -+++ node-v10.15.1/deps/v8/gypfiles/toolchain.gypi 2019-03-18 15:04:08.628361308 +0100 -@@ -1106,6 +1106,7 @@ - 'cflags': [ '-fno-strict-aliasing' ], - }], - ], # conditions -+ 'ldflags+': [ '$(LDFLAGS)' ], - }], - ['OS=="solaris"', { - 'defines': [ '__C99FEATURES__=1' ], # isinf() etc.
meta-oe/recipes-devtools/nodejs/nodejs_10.17.0.bb +12 -1
@@ -19,7 +19,6 @@ SRC_URI = "http://nodejs.org/dist/v${PV}/node-v${PV}.tar.xz \ file://0001-Disable-running-gyp-files-for-bundled-deps.patch \ file://0004-Make-compatibility-with-gcc-4.8.patch \ file://0005-Link-atomic-library.patch \ - file://0006-Use-target-ldflags.patch \ " SRC_URI_append_class-target = " \ file://0002-Using-native-torque.patch \
@@ -58,6 +57,18 @@ PACKAGECONFIG[libuv] = "--shared-libuv,,libuv" PACKAGECONFIG[nghttp2] = "--shared-nghttp2,,nghttp2" PACKAGECONFIG[zlib] = "--shared-zlib,,zlib" +# We don't want to cross-compile during target compile, +# and we need to use the right flags during host compile, +# too. +EXTRA_OEMAKE = "\ + CC.host='${CC}' \ + CFLAGS.host='${CPPFLAGS} ${CFLAGS}' \ + CXX.host='${CXX}' \ + CXXFLAGS.host='${CPPFLAGS} ${CXXFLAGS}' \ + LDFLAGS.host='${LDFLAGS}' \ + AR.host='${AR}' \ +" + # Node is way too cool to use proper autotools, so we install two wrappers to forcefully inject proper arch cflags to workaround gypi do_configure () { rm -rf ${S}/deps/openssl
Signature · 1 line
2.23.0.rc1
Khem,
This is a rev2, as v1 is still in master-next, only. I forgot to set the subject line right...
Cheers,
Andre'
Quoted 96 lines · André Draszik
NodeJS comes with an embedded, patched, version of gyp.
Normally, gyp supports compiling for the build machine, e.g. native tools that need to be compiled to run during the build, and for the host, using different variables, e.g. CC and CC.host, etc.
Most of this has been patched out in the NodeJS version of gyp, and essentially it only supports compiling using one compiler - ${CC}. This modification excludes LDFLAGS for native tools, and those still evaluate LDFLAGS.host (only).While this modified behaviour is OK for the OE use-case of building native and target tools separately, it means that this recipe can not work as-is with standard gyp, and wrong LDFLAGS are being used for some of the tools compiled (torque) in either case.
By setting the make variables that gyp-generated makefiles inspect, we support use of unpatched gyp, and we ensure that all tools are compiled with correct LDFLAGS in either case.
This now also allows us to drop the patch that had been applied to work-around this problem.
Signed-off-by André Draszik <[email protected]>---
v2:_appendnot needed forEXTRA_OEMAKE--- .../nodejs/0006-Use-target-ldflags.patch | 24 ------------------- .../recipes-devtools/nodejs/nodejs_10.17.0.bb | 13 +++++++++- 2 files changed, 12 insertions(+), 25 deletions(-) delete mode 100644 meta-oe/recipes-devtools/nodejs/nodejs/0006-Use-target-ldflags.patchmeta-oe/recipes- renamed from meta-oe/recipes-devtools/nodejs/nodejs/0006-Use-target-ldflags.patch +0
-0
devtools/nodejs/nodejs/0006-Use-target-ldflags.patchdeleted file mode 100644
index f6569cd57..000000000meta-oe/recipes-devtools/nodejs/nodejs/0006-Use-target-ldflags.patch deleted +0
-24@@ -1,24 +0,0 @@ -The target LDFLAGS have been ignored. Tools like torque -have been loaded from system libraries, even if a native -one was the target. -|$ ldd torque -| libz.so.1 => /lib/x86_64-linux-gnu/libz.so.1 -| libcrypto.so.1.1 => /usr/lib/x86_64-linux-gnu/libcrypto.so.1.1 -| libssl.so.1.1 => /usr/lib/x86_64-linux-gnu/libssl.so.1.1 -| libicui18n.so.63 => not found -| libicuuc.so.63 => not found -... - -Signed-off-by: Andrej Valek <[email protected]> - -diff -Naur node-v10.15.1/deps/v8/gypfiles/toolchain.gypi node-v10.15.1/deps/v8/gypfiles/toolchain.gypi ---- node-v10.15.1/deps/v8/gypfiles/toolchain.gypi 2019-03-18 15:01:39.000000000 +0100 -+++ node-v10.15.1/deps/v8/gypfiles/toolchain.gypi 2019-03-18 15:04:08.628361308 +0100 -@@ -1106,6 +1106,7 @@ - 'cflags': [ '-fno-strict-aliasing' ], - }], - ], # conditions -+ 'ldflags+': [ '$(LDFLAGS)' ], - }], - ['OS=="solaris"', { - 'defines': [ '__C99FEATURES__=1' ], # isinf() etc.meta-oe/recipes-devtools/nodejs/nodejs_10.17.0.bb +12
-1@@ -19,7 +19,6 @@ SRC_URI = "http://nodejs.org/dist/v${PV}/node-v${PV}.tar.xz \ file://0001-Disable-running-gyp-files-for-bundled-deps.patch \ file://0004-Make-compatibility-with-gcc-4.8.patch \ file://0005-Link-atomic-library.patch \ - file://0006-Use-target-ldflags.patch \ " SRC_URI_append_class-target = " \ file://0002-Using-native-torque.patch \@@ -58,6 +57,18 @@ PACKAGECONFIG[libuv] = "--shared-libuv,,libuv" PACKAGECONFIG[nghttp2] = "--shared-nghttp2,,nghttp2" PACKAGECONFIG[zlib] = "--shared-zlib,,zlib" +# We don't want to cross-compile during target compile, +# and we need to use the right flags during host compile, +# too. +EXTRA_OEMAKE = "\ + CC.host='${CC}' \ + CFLAGS.host='${CPPFLAGS} ${CFLAGS}' \ + CXX.host='${CXX}' \ + CXXFLAGS.host='${CPPFLAGS} ${CXXFLAGS}' \ + LDFLAGS.host='${LDFLAGS}' \ + AR.host='${AR}' \ +" + # Node is way too cool to use proper autotools, so we install two wrappers to forcefully inject proper arch cflags toworkaround gypi
do_configure () { rm -rf ${S}/deps/openssl
Quoted 6 lines · André Draszik
Khem,
This is a rev2, as v1 is still in master-next, only. I forgot to set the subject line right...
thanks for letting me know
Quoted 101 lines
Cheers,
Andre'Quoted 96 lines · André Draszik
NodeJS comes with an embedded, patched, version of gyp.
Normally, gyp supports compiling for the build machine, e.g. native tools that need to be compiled to run during the build, and for the host, using different variables, e.g. CC and CC.host, etc.
Most of this has been patched out in the NodeJS version of gyp, and essentially it only supports compiling using one compiler - ${CC}. This modification excludes LDFLAGS for native tools, and those still evaluate LDFLAGS.host (only).While this modified behaviour is OK for the OE use-case of building native and target tools separately, it means that this recipe can not work as-is with standard gyp, and wrong LDFLAGS are being used for some of the tools compiled (torque) in either case.
By setting the make variables that gyp-generated makefiles inspect, we support use of unpatched gyp, and we ensure that all tools are compiled with correct LDFLAGS in either case.
This now also allows us to drop the patch that had been applied to work-around this problem.
Signed-off-by André Draszik <[email protected]>---
v2:_appendnot needed forEXTRA_OEMAKE--- .../nodejs/0006-Use-target-ldflags.patch | 24 ------------------- .../recipes-devtools/nodejs/nodejs_10.17.0.bb | 13 +++++++++- 2 files changed, 12 insertions(+), 25 deletions(-) delete mode 100644 meta-oe/recipes-devtools/nodejs/nodejs/0006-Use-target-ldflags.patchmeta-oe/recipes- renamed from meta-oe/recipes-devtools/nodejs/nodejs/0006-Use-target-ldflags.patch +0
-0
devtools/nodejs/nodejs/0006-Use-target-ldflags.patchdeleted file mode 100644
index f6569cd57..000000000meta-oe/recipes-devtools/nodejs/nodejs/0006-Use-target-ldflags.patch deleted +0
-24@@ -1,24 +0,0 @@ -The target LDFLAGS have been ignored. Tools like torque -have been loaded from system libraries, even if a native -one was the target. -|$ ldd torque -| libz.so.1 => /lib/x86_64-linux-gnu/libz.so.1 -| libcrypto.so.1.1 => /usr/lib/x86_64-linux-gnu/libcrypto.so.1.1 -| libssl.so.1.1 => /usr/lib/x86_64-linux-gnu/libssl.so.1.1 -| libicui18n.so.63 => not found -| libicuuc.so.63 => not found -... - -Signed-off-by: Andrej Valek <[email protected]> - -diff -Naur node-v10.15.1/deps/v8/gypfiles/toolchain.gypi node-v10.15.1/deps/v8/gypfiles/toolchain.gypi ---- node-v10.15.1/deps/v8/gypfiles/toolchain.gypi 2019-03-18 15:01:39.000000000 +0100 -+++ node-v10.15.1/deps/v8/gypfiles/toolchain.gypi 2019-03-18 15:04:08.628361308 +0100 -@@ -1106,6 +1106,7 @@ - 'cflags': [ '-fno-strict-aliasing' ], - }], - ], # conditions -+ 'ldflags+': [ '$(LDFLAGS)' ], - }], - ['OS=="solaris"', { - 'defines': [ '__C99FEATURES__=1' ], # isinf() etc.meta-oe/recipes-devtools/nodejs/nodejs_10.17.0.bb +12
-1@@ -19,7 +19,6 @@ SRC_URI = "http://nodejs.org/dist/v${PV}/node-v${PV}.tar.xz \ file://0001-Disable-running-gyp-files-for-bundled-deps.patch \ file://0004-Make-compatibility-with-gcc-4.8.patch \ file://0005-Link-atomic-library.patch \ - file://0006-Use-target-ldflags.patch \ " SRC_URI_append_class-target = " \ file://0002-Using-native-torque.patch \@@ -58,6 +57,18 @@ PACKAGECONFIG[libuv] = "--shared-libuv,,libuv" PACKAGECONFIG[nghttp2] = "--shared-nghttp2,,nghttp2" PACKAGECONFIG[zlib] = "--shared-zlib,,zlib" +# We don't want to cross-compile during target compile, +# and we need to use the right flags during host compile, +# too. +EXTRA_OEMAKE = "\ + CC.host='${CC}' \ + CFLAGS.host='${CPPFLAGS} ${CFLAGS}' \ + CXX.host='${CXX}' \ + CXXFLAGS.host='${CPPFLAGS} ${CXXFLAGS}' \ + LDFLAGS.host='${LDFLAGS}' \ + AR.host='${AR}' \ +" + # Node is way too cool to use proper autotools, so we install two wrappers to forcefully inject proper arch cflags toworkaround gypi
do_configure () { rm -rf ${S}/deps/openssl