gcc: Upgrade GCC to 16.1 release
v3 cover gcc: Upgrade GCC to 16.1 release
From: Hemanth Kumar M D <[email protected]>
This patch series upgrades GCC to 16.1 and includes the required runtime and compatibility fixes identified during testing in OE-Core. The series also includes fixes for packages failing to build with GCC 16, updates for libatomic runtime handling, and packaging changes for newly generated libatomic_asneeded files.
Changes in v3:
- Fix missing Signed-off-by in gn and gcc inner patch files
Changes in v2:
- Add libatomic-dev to elfutils ptest RDEPENDS
- Add libatomic-staticdev to libc-test ptest RDEPENDS
- Package new libcaf_shmem.a static library in libgfortran
- Fix gn build with GCC 16 on musl
- Backport GCC i386 ICE fix with -ffloat-store and SSE [PR125308]
Summary of patches:
- Patch 01: Backport piglit fix for building with GCC 16.
- Patch 02: Backport rxvt-unicode fix for building with GCC 16.
- Patch 03: Disable automatic libatomic linking in glibc.
- Patch 04: Upgrade GCC to 16.1 release.
- Patch 05: Disable automatic libatomic linking during gcc-runtime build.
- Patch 06: Package new libatomic_asneeded runtime files.
- Patch 07: Avoid passing MULTIBUILDTOP to libatomic in OE builds.
- Patch 08: Package new libcaf_shmem.a static library in libgfortran.
- Patch 09: Fix gn build with GCC 16 on musl.
- Patch 10: Add libatomic-dev to elfutils ptest RDEPENDS.
- Patch 11: Add libatomic-staticdev to libc-test ptest RDEPENDS.
- Patch 12: Fix ICE on i386 with -ffloat-store and SSE [PR125308].
meta/conf/distro/include/maintainers.inc | 2 +-
meta/conf/distro/include/tcmode-default.inc | 2 +-
meta/recipes-core/glibc/glibc.inc | 3 +
meta/recipes-core/musl/libc-test_git.bb | 1 +
.../elfutils/elfutils_0.194.bb | 2 +-
.../gcc/{gcc-15.2.inc => gcc-16.1.inc} | 18 ++--
...ian_15.2.bb => gcc-cross-canadian_16.1.bb} | 0
.../{gcc-cross_15.2.bb => gcc-cross_16.1.bb} | 0
...-crosssdk_15.2.bb => gcc-crosssdk_16.1.bb} | 0
meta/recipes-devtools/gcc/gcc-runtime.inc | 18 +++-
...cc-runtime_15.2.bb => gcc-runtime_16.1.bb} | 0
...itizers_15.2.bb => gcc-sanitizers_16.1.bb} | 0
...{gcc-source_15.2.bb => gcc-source_16.1.bb} | 0
...AMIC_LINKER-and-UCLIBC_DYNAMIC_LINKE.patch | 2 +-
...sroot-gcc-version-specific-dirs-with.patch | 4 +-
...omic-Do-not-enforce-march-on-aarch64.patch | 18 ++--
...025-gcc-testsuite-tweaks-for-mips-OE.patch | 78 +++++++++-------
...6-fix-ICE-with-ffloat-store-PR125308.patch | 89 +++++++++++++++++++
.../gcc/{gcc_15.2.bb => gcc_16.1.bb} | 0
...initial_15.2.bb => libgcc-initial_16.1.bb} | 0
.../gcc/{libgcc_15.2.bb => libgcc_16.1.bb} | 0
...ibgfortran_15.2.bb => libgfortran_16.1.bb} | 1 +
...001-gn-fix-build-with-gcc-16-on-musl.patch | 33 +++++++
meta/recipes-devtools/gn/gn_git.bb | 4 +-
...h-rename-local-lerp-to-linear_interp.patch | 46 ++++++++++
meta/recipes-graphics/piglit/piglit_git.bb | 1 +
.../rxvt-unicode/rxvt-unicode.inc | 3 +-
.../0001-rxvt-unicode-fix-gcc16-c-20.patch | 61 +++++++++++++
28 files changed, 323 insertions(+), 63 deletions(-)
rename meta/recipes-devtools/gcc/{gcc-15.2.inc => gcc-16.1.inc} (88%)
rename meta/recipes-devtools/gcc/{gcc-cross-canadian_15.2.bb => gcc-cross-canadian_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc-cross_15.2.bb => gcc-cross_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc-crosssdk_15.2.bb => gcc-crosssdk_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc-runtime_15.2.bb => gcc-runtime_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc-sanitizers_15.2.bb => gcc-sanitizers_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc-source_15.2.bb => gcc-source_16.1.bb} (100%)
create mode 100644 meta/recipes-devtools/gcc/gcc/0026-gcc-i386-fix-ICE-with-ffloat-store-PR125308.patch
rename meta/recipes-devtools/gcc/{gcc_15.2.bb => gcc_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{libgcc-initial_15.2.bb => libgcc-initial_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{libgcc_15.2.bb => libgcc_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{libgfortran_15.2.bb => libgfortran_16.1.bb} (51%)
create mode 100644 meta/recipes-devtools/gn/files/0001-gn-fix-build-with-gcc-16-on-musl.patch
create mode 100644 meta/recipes-graphics/piglit/piglit/0001-fbo-blit-stretch-rename-local-lerp-to-linear_interp.patch
create mode 100644 meta/recipes-sato/rxvt-unicode/rxvt-unicode/0001-rxvt-unicode-fix-gcc16-c-20.patchSignature · 1 line
2.49.0
PATCH v3 1/12 piglit:backport a fix for building with gcc-16
From: Hemanth Kumar M D <[email protected]>
Avoid conflict with C++20 std::lerp from <cmath>, which causes redefinition and ambiguous call errors with newer toolchains.
Upstream-Status: Backport [https://gitlab.freedesktop.org/mesa/piglit/-/commit/ec0fc2fef0755b1595c3799ab25129174a4b2007]
--- ...h-rename-local-lerp-to-linear_interp.patch | 46 +++++++++++++++++++ meta/recipes-graphics/piglit/piglit_git.bb | 1 + 2 files changed, 47 insertions(+) create mode 100644 meta/recipes-graphics/piglit/piglit/0001-fbo-blit-stretch-rename-local-lerp-to-linear_interp.patch
meta/recipes-graphics/piglit/piglit/0001-fbo-blit-stretch-rename-local-lerp-to-linear_interp.patch new file +46 -0
@@ -0,0 +1,46 @@ +From ec0fc2fef0755b1595c3799ab25129174a4b2007 Mon Sep 17 00:00:00 2001 +From: Vinson Lee <[email protected]> +Date: Sat, 7 Feb 2026 18:39:39 -0800 +Subject: [PATCH] fbo-blit-stretch: rename local lerp to linear_interp + +Avoid conflict with C++20 std::lerp from <cmath>, which +causes redefinition and ambiguous call errors with newer +toolchains. + +Upstream-Status: Backport [https://gitlab.freedesktop.org/mesa/piglit/-/commit/ec0fc2fef0755b1595c3799ab25129174a4b2007] + +Signed-off-by: Vinson Lee <[email protected]> +Signed-off-by: Hemanth Kumar M D <[email protected]> +--- + tests/fbo/fbo-blit-stretch.cpp | 8 ++++---- + 1 file changed, 4 insertions(+), 4 deletions(-) + +diff --git a/tests/fbo/fbo-blit-stretch.cpp b/tests/fbo/fbo-blit-stretch.cpp +index a83288c0f..b0d362ecd 100644 +--- a/tests/fbo/fbo-blit-stretch.cpp ++++ b/tests/fbo/fbo-blit-stretch.cpp +@@ -173,7 +173,7 @@ clamp(GLint &x, GLint xmin, GLint xmax) + } + + static float +-lerp(float x0, float x1, float w) ++linear_interp(float x0, float x1, float w) + { + return x0 + (x1 - x0) * w; + } +@@ -181,9 +181,9 @@ lerp(float x0, float x1, float w) + static float + lerp2d(float xy00, float xy01, float xy10, float xy11, float wx, float wy) + { +- float y0 = lerp(xy00, xy01, wx); +- float y1 = lerp(xy10, xy11, wx); +- return lerp(y0, y1, wy); ++ float y0 = linear_interp(xy00, xy01, wx); ++ float y1 = linear_interp(xy10, xy11, wx); ++ return linear_interp(y0, y1, wy); + } + + static float clearColor[4] = { +-- +2.49.0 +
meta/recipes-graphics/piglit/piglit_git.bb +1 -0
@@ -14,6 +14,7 @@ SRC_URI = "git://gitlab.freedesktop.org/mesa/piglit.git;protocol=https;branch=ma file://0001-tests-egl-spec-make-egl_ext_surface_compression-cond.patch \ file://0001-tests-no_error.py-modify-_command-and-not-command.patch \ file://0001-generated_tests-use-shape-in-place-of-newshape-on-nu.patch \ + file://0001-fbo-blit-stretch-rename-local-lerp-to-linear_interp.patch \ " UPSTREAM_CHECK_COMMITS = "1"
Signature · 1 line
2.49.0
PATCH v3 2/12 rxvt-unicode: backport a fix for building with gcc-16
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
rename lerp to avoid conflict with the non-identical function std::lerp
Upstream-Status: Submitted [https://lists.schmorp.de/pipermail/rxvt-unicode/2026q2/002707.html]
--- .../rxvt-unicode/rxvt-unicode.inc | 3 +- .../0001-rxvt-unicode-fix-gcc16-c-20.patch | 61 +++++++++++++++++++ 2 files changed, 63 insertions(+), 1 deletion(-) create mode 100644 meta/recipes-sato/rxvt-unicode/rxvt-unicode/0001-rxvt-unicode-fix-gcc16-c-20.patch
meta/recipes-sato/rxvt-unicode/rxvt-unicode.inc +2 -1
@@ -11,7 +11,8 @@ DEPENDS = "virtual/libx11 libxt libxft gdk-pixbuf libxmu libptytty ncurses-nativ SRC_URI = "http://dist.schmorp.de/rxvt-unicode/Attic/rxvt-unicode-${PV}.tar.bz2 \ file://xwc.patch \ file://rxvt.desktop \ - file://rxvt.png" + file://rxvt.png \ + file://0001-rxvt-unicode-fix-gcc16-c-20.patch" inherit autotools features_check pkgconfig update-alternatives
meta/recipes-sato/rxvt-unicode/rxvt-unicode/0001-rxvt-unicode-fix-gcc16-c-20.patch new file +61 -0
61 lines not on this page · raw message
Signature · 1 line
2.49.0
PATCH v3 3/12 glibc: disable automatic libatomic linking
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
GCC 16 introduced -flink-libatomic which automatically injects -latomic_asneeded into link commands via LINK_LIBATOMIC_SPEC. When glibc is built with a GCC 16 cross-compiler, the compiler would attempt to link against libatomic which does not exist yet at this stage.(libatomic is built later as part of gcc-runtime)
--- meta/recipes-core/glibc/glibc.inc | 3 +++ 1 file changed, 3 insertions(+)
meta/recipes-core/glibc/glibc.inc +3 -0
@@ -46,6 +46,9 @@ EXTRA_OEMAKE += "SHELL=/bin/bash" # bash but it can be configured by setting KSHELL Makefile variable EXTRA_OEMAKE += "KSHELL=/bin/sh" +TARGET_CC_ARCH += "-fno-link-libatomic" +SDK_CC_ARCH += "-fno-link-libatomic" + do_configure:prepend() { sed -e "s#/bin/bash#/bin/sh#" -i ${S}/elf/ldd.bash.in }
Signature · 1 line
2.49.0
Re: [OE-core] [PATCH v3 03/12] glibc: disable automatic libatomic linking
Quoted 22 lines · Hemanth Kumar M D via lists.openembedded.org
From: Hemanth Kumar M D <[email protected]>
GCC 16 introduced -flink-libatomic which automatically injects -latomic_asneeded into link commands via
LINK_LIBATOMIC_SPEC. When glibc is built with a GCC 16 cross-compiler, the compiler would attempt to link against libatomic which does not exist yet at this stage.(libatomic is built later as part of gcc-runtime)Link:
https://github.com/gcc-mirror/gcc/commit/e63cf4b130b86dd7dde1bf499d3d40faca10ea2eSigned-off-by Hemanth Kumar M D <[email protected]>--- meta/recipes-core/glibc/glibc.inc | 3 +++ 1 file changed, 3 insertions(+)meta/recipes-core/glibc/glibc.inc +0
-0
meta/recipes-core/glibc/glibc.inc EXTRA_OEMAKE += "SHELL=/bin/bash" · line 51
Quoted 6 lines
meta/recipes-core/glibc/glibc.inc +2
-0@@ -46,6 +46,9 @@ EXTRA_OEMAKE += "SHELL=/bin/bash" # bash but it can be configured by setting KSHELL Makefile variable EXTRA_OEMAKE += "KSHELL=/bin/sh" +TARGET_CC_ARCH += "-fno-link-libatomic" +SDK_CC_ARCH += "-fno-link-libatomic"
This might be better controlled with toolchain-gcc override.
Quoted 19 lines
meta/recipes-core/glibc/glibc.inc +1
-0+ do_configure:prepend() { sed -e "s#/bin/bash#/bin/sh#" -i ${S}/elf/ldd.bash.in}
--
2.49.0-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#237276):
https://lists.openembedded.org/g/openembedded-core/message/237276
Mute This Topic: https://lists.openembedded.org/mt/119387074/1997914
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [ [email protected]]
-=-=-=-=-=-=-=-=-=-=-=-
Re: [OE-core] [PATCH v3 03/12] glibc: disable automatic libatomic linking
Quoted 43 lines · Khem Raj
**
*CAUTION: This email comes from a non Wind River email account!*
Do not click links or open attachments unless you recognize the sender and know the content is safe.On Tue, May 19, 2026 at 12:25 AM Hemanth Kumar M D via lists.openembedded.org <https://urldefense.com/v3/__http://lists.openembedded.org__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck6iGu8T68$> <[email protected]> wrote:
From: Hemanth Kumar M D <[email protected]> GCC 16 introduced -flink-libatomic which automatically injects -latomic_asneeded into link commands via LINK_LIBATOMIC_SPEC. When glibc is built with a GCC 16 cross-compiler, the compiler would attempt to link against libatomic which does not exist yet at this stage.(libatomic is built later as part of gcc-runtime) Link: https://github.com/gcc-mirror/gcc/commit/e63cf4b130b86dd7dde1bf499d3d40faca10ea2e <https://urldefense.com/v3/__https://github.com/gcc-mirror/gcc/commit/e63cf4b130b86dd7dde1bf499d3d40faca10ea2e__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck690KqPGs$> Signed-off-by: Hemanth Kumar M D <[email protected]> --- meta/recipes-core/glibc/glibc.inc <https://urldefense.com/v3/__http://glibc.inc__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck6_ujwH0A$> | 3 +++ 1 file changed, 3 insertions(+) diff --git a/meta/recipes-core/glibc/glibc.inc <https://urldefense.com/v3/__http://glibc.inc__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck6_ujwH0A$> b/meta/recipes-core/glibc/glibc.inc <https://urldefense.com/v3/__http://glibc.inc__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck6_ujwH0A$> index dae087d53f..751cc0c572 100644 --- a/meta/recipes-core/glibc/glibc.inc <https://urldefense.com/v3/__http://glibc.inc__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck6_ujwH0A$> +++ b/meta/recipes-core/glibc/glibc.inc <https://urldefense.com/v3/__http://glibc.inc__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck6_ujwH0A$> @@ -46,6 +46,9 @@ EXTRA_OEMAKE += "SHELL=/bin/bash" # bash but it can be configured by setting KSHELL Makefile variable EXTRA_OEMAKE += "KSHELL=/bin/sh" +TARGET_CC_ARCH += "-fno-link-libatomic" +SDK_CC_ARCH += "-fno-link-libatomic"
Quoted 1 line
This might be better controlled with toolchain-gcc override.
Hi Khem,
Updated this to use the toolchain-gcc override:
TARGET_CC_ARCH:toolchain-gcc += "-fno-link-libatomic" SDK_CC_ARCH:toolchain-gcc += "-fno-link-libatomic"
Will be sending the fix in the coming series.
Quoted 25 lines
+
do_configure:prepend() {
sed -e "s#/bin/bash#/bin/sh#" -i ${S}/elf/ldd.bash.in
<https://urldefense.com/v3/__http://ldd.bash.in__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck6yiEC9_0$>
}
--
2.49.0
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#237276):
https://lists.openembedded.org/g/openembedded-core/message/237276
<https://urldefense.com/v3/__https://lists.openembedded.org/g/openembedded-core/message/237276__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck6I4oFiHo$>
Mute This Topic:
https://lists.openembedded.org/mt/119387074/1997914
<https://urldefense.com/v3/__https://lists.openembedded.org/mt/119387074/1997914__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck68YPSyXM$>
Group Owner: [email protected]
<mailto:openembedded-core%[email protected]>
Unsubscribe:
https://lists.openembedded.org/g/openembedded-core/unsub
<https://urldefense.com/v3/__https://lists.openembedded.org/g/openembedded-core/unsub__;!!AjveYdw8EvQ!ep_WpaZmir4Aw1K_c2XMB01Y8AuFDouyrw1F8TpQr78EckKFe8ujYgJ9slTPRbu3p6yeUrHqulvlTck6TBdEFcI$>
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-Signature · 2 lines
Regards, Hemanth Kumar M D
PATCH v3 4/12 gcc: Upgrade GCC to 16.1 release
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
Notable changes in GCC 16 include:
- Improved C23/C++23 support and initial C++26 features
- Enhanced diagnostics and static analyzer improvements
- Optimizer and code generation enhancements
- Target-specific updates for AArch64, RISC-V, x86, and others
- Runtime library updates (libstdc++, libatomic, libgomp, libsanitizer)
- Multiple bug fixes and performance improvements
* Rebased existing OE-Core patches for GCC 16.1 and dropped patches merged upstream.
For detailed info: https://gcc.gnu.org/gcc-16/changes.html
---
meta/conf/distro/include/maintainers.inc | 2 +-
meta/conf/distro/include/tcmode-default.inc | 2 +-
.../gcc/{gcc-15.2.inc => gcc-16.1.inc} | 17 ++--
...ian_15.2.bb => gcc-cross-canadian_16.1.bb} | 0
.../{gcc-cross_15.2.bb => gcc-cross_16.1.bb} | 0
...-crosssdk_15.2.bb => gcc-crosssdk_16.1.bb} | 0
...cc-runtime_15.2.bb => gcc-runtime_16.1.bb} | 0
...itizers_15.2.bb => gcc-sanitizers_16.1.bb} | 0
...{gcc-source_15.2.bb => gcc-source_16.1.bb} | 0
...AMIC_LINKER-and-UCLIBC_DYNAMIC_LINKE.patch | 2 +-
...sroot-gcc-version-specific-dirs-with.patch | 4 +-
...omic-Do-not-enforce-march-on-aarch64.patch | 18 ++---
...025-gcc-testsuite-tweaks-for-mips-OE.patch | 78 +++++++++++--------
.../gcc/{gcc_15.2.bb => gcc_16.1.bb} | 0
...initial_15.2.bb => libgcc-initial_16.1.bb} | 0
.../gcc/{libgcc_15.2.bb => libgcc_16.1.bb} | 0
...ibgfortran_15.2.bb => libgfortran_16.1.bb} | 0
17 files changed, 66 insertions(+), 57 deletions(-)
rename meta/recipes-devtools/gcc/{gcc-15.2.inc => gcc-16.1.inc} (88%)
rename meta/recipes-devtools/gcc/{gcc-cross-canadian_15.2.bb => gcc-cross-canadian_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc-cross_15.2.bb => gcc-cross_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc-crosssdk_15.2.bb => gcc-crosssdk_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc-runtime_15.2.bb => gcc-runtime_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc-sanitizers_15.2.bb => gcc-sanitizers_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc-source_15.2.bb => gcc-source_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{gcc_15.2.bb => gcc_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{libgcc-initial_15.2.bb => libgcc-initial_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{libgcc_15.2.bb => libgcc_16.1.bb} (100%)
rename meta/recipes-devtools/gcc/{libgfortran_15.2.bb => libgfortran_16.1.bb} (100%)meta/conf/distro/include/maintainers.inc +1 -1
@@ -200,7 +200,7 @@ RECIPE_MAINTAINER:pn-gcc-cross-canadian-${TRANSLATED_TARGET_ARCH} = "Khem Raj <r RECIPE_MAINTAINER:pn-gcc-crosssdk-${SDK_SYS} = "Khem Raj <[email protected]>" RECIPE_MAINTAINER:pn-gcc-runtime = "Khem Raj <[email protected]>" RECIPE_MAINTAINER:pn-gcc-sanitizers = "Khem Raj <[email protected]>" -RECIPE_MAINTAINER:pn-gcc-source-15.2.0 = "Khem Raj <[email protected]>" +RECIPE_MAINTAINER:pn-gcc-source-16.1.0 = "Khem Raj <[email protected]>" RECIPE_MAINTAINER:pn-gcompat = "Khem Raj <[email protected]>" RECIPE_MAINTAINER:pn-gconf = "Ross Burton <[email protected]>" RECIPE_MAINTAINER:pn-gcr = "Unassigned <[email protected]>"
meta/conf/distro/include/tcmode-default.inc +1 -1
@@ -9,7 +9,7 @@ PREFERRED_PROVIDER_virtual/nativesdk-cross-binutils ?= "binutils-crosssdk-${SDK_ # Default libc config PREFERRED_PROVIDER_virtual/gettext ??= "gettext" -GCCVERSION ?= "15.%" +GCCVERSION ?= "16.%" SDKGCCVERSION ?= "${GCCVERSION}" GLIBCVERSION ?= "2.43%" RUSTVERSION ?= "1.94.1%"
meta/recipes-devtools/gcc/gcc-16.1.inc renamed from meta/recipes-devtools/gcc/gcc-15.2.inc +6 -11
@@ -2,11 +2,11 @@ require gcc-common.inc # Third digit in PV should be incremented after a minor release -PV = "15.2.0" +PV = "16.1.0" # BINV should be incremented to a revision after a minor gcc release -BINV = "15.2.0" +BINV = "16.1.0" FILESEXTRAPATHS =. "${FILE_DIRNAME}/gcc:${FILE_DIRNAME}/gcc/backport:"
@@ -34,17 +34,17 @@ LIC_FILES_CHKSUM = "\ #SRC_URI[sha256sum] = "41bf7be5dc029112a9df625266e7de030ffc36ff3638f270b180ae8c91fe6449" # from snapshot -#RELEASE ?= "15-20250420" -#RELEASE ?= "15.1.0-RC-20250418" +#RELEASE ?= "16-20260502" +#RELEASE ?= "16.1.0-RC-20260424" #BASEURI ?= "https://gcc.gnu.org/pub/gcc/snapshots/${RELEASE}/gcc-${RELEASE}.tar.xz" #SOURCEDIR ?= "gcc-${RELEASE}" -#SRC_URI[sha256sum] = "fe18624317c6c662977eea8b5a506ec6c132fc11c7a8d53e2f6d21b139b0af3c" +#SRC_URI[sha256sum] = "e8edc2aaf0affce7020658b0be9a3fc820aa33dc610640d3fe1ba1aee360051b" # official release RELEASE ?= "${PV}" BASEURI ?= "${GNU_MIRROR}/gcc/gcc-${PV}/gcc-${PV}.tar.xz" SOURCEDIR ?= "gcc-${PV}" -SRC_URI[sha256sum] = "438fd996826b0c82485a29da03a72d71d6e3541a83ec702df4271f6fe025d24e" +SRC_URI[sha256sum] = "50efb4d94c3397aff3b0d61a5abd748b4dd31d9d3f2ab7be05b171d36a510f79" SRC_URI = "${BASEURI} \ file://0001-gcc-4.3.1-ARCH_FLAGS_FOR_TARGET.patch \
@@ -72,11 +72,6 @@ SRC_URI = "${BASEURI} \ file://0023-Fix-install-path-of-linux64.h.patch \ file://0024-Avoid-hardcoded-build-paths-into-ppc-libgcc.patch \ file://0025-gcc-testsuite-tweaks-for-mips-OE.patch \ - file://0026-fix-pr90579-testcases.patch \ - file://0027-gcc15-pr122394.patch \ - file://0028-fix-gcc-libitm-false-postives-01.patch \ - file://0028-fix-gcc-libitm-false-postives-02.patch \ - file://0028-libcody-Make-it-buildable-by-C-11-to-C-26.patch \ file://0001-mapper-localhost-might-not-be-known.patch \ "
meta/recipes-devtools/gcc/gcc-cross-canadian_16.1.bb renamed from meta/recipes-devtools/gcc/gcc-cross-canadian_15.2.bb +0 -0
meta/recipes-devtools/gcc/gcc-cross_16.1.bb renamed from meta/recipes-devtools/gcc/gcc-cross_15.2.bb +0 -0
meta/recipes-devtools/gcc/gcc-crosssdk_16.1.bb renamed from meta/recipes-devtools/gcc/gcc-crosssdk_15.2.bb +0 -0
meta/recipes-devtools/gcc/gcc-runtime_16.1.bb renamed from meta/recipes-devtools/gcc/gcc-runtime_15.2.bb +0 -0
meta/recipes-devtools/gcc/gcc-sanitizers_16.1.bb renamed from meta/recipes-devtools/gcc/gcc-sanitizers_15.2.bb +0 -0
meta/recipes-devtools/gcc/gcc-source_16.1.bb renamed from meta/recipes-devtools/gcc/gcc-source_15.2.bb +0 -0
meta/recipes-devtools/gcc/gcc/0007-Define-GLIBC_DYNAMIC_LINKER-and-UCLIBC_DYNAMIC_LINKE.patch +1 -1
@@ -172,7 +172,7 @@ index fbc75a90ad5..38c239940bd 100644 + SYSTEMLIBS_DIR "ld-linux-loongarch-" ABI_SPEC ".so.1" #define MUSL_ABI_SPEC \ - "%{mabi=lp64d:}" \ + "%{mabi=ilp32d:}" \ @@ -40,7 +40,7 @@ along with GCC; see the file COPYING3. If not see #undef MUSL_DYNAMIC_LINKER
meta/recipes-devtools/gcc/gcc/0017-Search-target-sysroot-gcc-version-specific-dirs-with.patch +2 -2
@@ -89,12 +89,12 @@ index 007c91780f6..22eb17bc099 100644 + { + memcpy (path + len, this_multi, this_multi_len + 1); + memcpy (path + len + this_multi_len, multi_suffix, suffix_len + 1); -+ ret = callback (path, callback_info); ++ ret = callback (path); + if (ret) + break; + } + } + memcpy (path + len, multi_suffix, suffix_len + 1); - ret = callback (path, callback_info); + ret = callback (path); if (ret)
meta/recipes-devtools/gcc/gcc/0022-libatomic-Do-not-enforce-march-on-aarch64.patch +9 -9
@@ -29,14 +29,14 @@ index 0f1a7156084..255fe448e63 100644 endif diff --git a/libatomic/Makefile.in b/libatomic/Makefile.in -index 9798e7c09e9..3b3027eb77d 100644 +index 887ae41..b3d5e6f 100644 --- a/libatomic/Makefile.in +++ b/libatomic/Makefile.in -@@ -452,7 +452,6 @@ M_SRC = $(firstword $(filter %/$(M_FILE), $(all_c_files))) - libatomic_la_LIBADD = $(foreach s,$(SIZES),$(addsuffix \ - _$(s)_.lo,$(SIZEOBJS))) $(am__append_1) $(am__append_2) \ - $(am__append_3) $(am__append_4) --@ARCH_AARCH64_LINUX_TRUE@@HAVE_IFUNC_TRUE@IFUNC_OPTIONS = -march=armv8-a+lse - @ARCH_ARM_LINUX_TRUE@@HAVE_IFUNC_TRUE@IFUNC_OPTIONS = -march=armv7-a+fp -DHAVE_KERNEL64 - @ARCH_I386_TRUE@@HAVE_IFUNC_TRUE@IFUNC_OPTIONS = -march=i586 - @ARCH_X86_64_TRUE@@HAVE_IFUNC_TRUE@IFUNC_OPTIONS = -mcx16 -mcx16 +@@ -470,7 +470,6 @@ libatomic_la_LDFLAGS = $(libatomic_version_info) $(libatomic_version_script) \ + @PARTIAL_VXWORKS_FALSE@ _$(s)_.lo,$(SIZEOBJS))) $(am__append_1) \ + @PARTIAL_VXWORKS_FALSE@ $(am__append_2) $(am__append_3) \ + @PARTIAL_VXWORKS_FALSE@ $(am__append_4) $(am__append_5) +-@ARCH_AARCH64_LINUX_TRUE@@HAVE_IFUNC_TRUE@@PARTIAL_VXWORKS_FALSE@IFUNC_OPTIONS = -march=armv8-a+lse + @ARCH_ARM_LINUX_TRUE@@HAVE_IFUNC_TRUE@@PARTIAL_VXWORKS_FALSE@IFUNC_OPTIONS = -march=armv7-a+fp -DHAVE_KERNEL64 + @ARCH_I386_TRUE@@HAVE_IFUNC_TRUE@@PARTIAL_VXWORKS_FALSE@IFUNC_OPTIONS = -march=i586 + @ARCH_LOONGARCH_TRUE@@HAVE_IFUNC_TRUE@@PARTIAL_VXWORKS_FALSE@IFUNC_OPTIONS = -mlsx|-mscq
meta/recipes-devtools/gcc/gcc/0025-gcc-testsuite-tweaks-for-mips-OE.patch +46 -32
156 lines not on this page · raw message
meta/recipes-devtools/gcc/gcc_16.1.bb renamed from meta/recipes-devtools/gcc/gcc_15.2.bb +0 -0
meta/recipes-devtools/gcc/libgcc-initial_16.1.bb renamed from meta/recipes-devtools/gcc/libgcc-initial_15.2.bb +0 -0
meta/recipes-devtools/gcc/libgcc_16.1.bb renamed from meta/recipes-devtools/gcc/libgcc_15.2.bb +0 -0
meta/recipes-devtools/gcc/libgfortran_16.1.bb renamed from meta/recipes-devtools/gcc/libgfortran_15.2.bb +0 -0
Signature · 1 line
2.49.0
Quoted 2 lines · Hemanth Kumar M D via lists.openembedded.org
* Rebased existing OE-Core patches for GCC 16.1 and dropped patches merged upstream.
The patches are removed from SRC_URI, but have not been deleted.
Ross
PATCH v3 5/12 gcc-runtime: disable automatic libatomic linking during build
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
GCC 16 introduced -flink-libatomic which automatically injects -latomic_asneeded into link commands via LINK_LIBATOMIC_SPEC. During gcc-runtime do_compile, OE uses a pre-built GCC 16 cross-compiler to build target runtime libraries before libatomic itself is ready.
--- meta/recipes-devtools/gcc/gcc-runtime.inc | 3 +++ 1 file changed, 3 insertions(+)
meta/recipes-devtools/gcc/gcc-runtime.inc +3 -0
@@ -15,6 +15,9 @@ EXTRA_OECONF_PATHS = "\ --with-build-sysroot=${STAGING_DIR_TARGET} \ " +TARGET_CC_ARCH += "-fno-link-libatomic" +SDK_CC_ARCH += "-fno-link-libatomic" + EXTRA_OECONF:append:linuxstdbase = " --enable-clocale=gnu" EXTRA_OECONF:append = " --cache-file=${B}/config.cache" EXTRA_OECONF:append:libc-newlib = " --with-newlib --with-target-subdir"
Signature · 1 line
2.49.0
PATCH v3 6/12 gcc-runtime: package new libatomic_asneeded files
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
GCC 16 generates libatomic_asneeded.so and libatomic_asneeded.a during the build as part of the fix for unnecessary DT_NEEDED entries on libatomic in GCC runtime libraries (PR libstdc++/123396). Add them to avoid unpackaged files QA errors.
--- meta/recipes-devtools/gcc/gcc-runtime.inc | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
meta/recipes-devtools/gcc/gcc-runtime.inc +2 -1
@@ -268,9 +268,10 @@ SUMMARY:libatomic = "GNU C++11 atomics support library" FILES:libatomic-dev = "\ ${libdir}/libatomic.so \ ${libdir}/libatomic.la \ + ${libdir}/libatomic_asneeded.so \ " SUMMARY:libatomic-dev = "GNU C++11 atomics support library - development files" -FILES:libatomic-staticdev = "${libdir}/libatomic.a" +FILES:libatomic-staticdev = "${libdir}/libatomic.a ${libdir}/libatomic_asneeded.a" SUMMARY:libatomic-staticdev = "GNU C++11 atomics support library - static development files" FILES:libitm = "${libdir}/libitm.so.*"
Signature · 1 line
2.49.0
PATCH v3 7/12 gcc-runtime: avoid passing MULTIBUILDTOP to libatomic
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
libatomic constructs gcc_objdir using pwd and assumes MULTIBUILDTOP is relative. Passing an absolute MULTIBUILDTOP from OE results in malformed paths during libtool install operations in both compile and install stages.
Avoid passing MULTIBUILDTOP for libatomic and use the upstream default handling instead.
--- meta/recipes-devtools/gcc/gcc-runtime.inc | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-)
meta/recipes-devtools/gcc/gcc-runtime.inc +10 -2
@@ -83,14 +83,22 @@ do_configure[depends] += "${COMPILERDEP}" do_compile () { for d in libgcc ${RUNTIMETARGET}; do cd ${B}/${TARGET_SYS}/$d/ - oe_runmake MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ + if [ "$d" = "libatomic" ]; then + oe_runmake + else + oe_runmake MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ + fi done } do_install () { for d in ${RUNTIMETARGET}; do cd ${B}/${TARGET_SYS}/$d/ - oe_runmake 'DESTDIR=${D}' MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ install + if [ "$d" = "libatomic" ]; then + oe_runmake 'DESTDIR=${D}' install + else + oe_runmake 'DESTDIR=${D}' MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ install + fi done install -d ${D}${datadir}/gdb/auto-load/${libdir} mv ${D}${libdir}/libstdc++*-gdb.py ${D}${datadir}/gdb/auto-load/${libdir}
Signature · 1 line
2.49.0
Re: [OE-core] [PATCH v3 07/12] gcc-runtime: avoid passing MULTIBUILDTOP to libatomic
Quoted 7 lines · Hemanth Kumar M D via lists.openembedded.org
From: Hemanth Kumar M D <[email protected]>
libatomic constructs
gcc_objdirusingpwdand assumes MULTIBUILDTOP is relative. Passing an absolute MULTIBUILDTOP from OE results in malformed paths during libtool install operations in both compile and install stages.
Quoted 7 lines
Avoid passing MULTIBUILDTOP for libatomic and use the upstream default handling instead.
Link:
https://github.com/gcc-mirror/gcc/commit/e63cf4b130b86dd7dde1bf499d3d40faca10ea2e
It would be good to start a discussion/bug in gcc about it, I think using pwd relative paths is perhaps
less than ideal to use here. We can accept this additional logic here for OE in the meantime.
Quoted 53 lines
Signed-off-by Hemanth Kumar M D <[email protected]>--- meta/recipes-devtools/gcc/gcc-runtime.inc | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-)meta/recipes-devtools/gcc/gcc-runtime.inc +5
-2@@ -83,14 +83,22 @@ do_configure[depends] += "${COMPILERDEP}" do_compile () { for d in libgcc ${RUNTIMETARGET}; do cd ${B}/${TARGET_SYS}/$d/ - oe_runmake MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ + if [ "$d" = "libatomic" ]; then + oe_runmake + else + oe_runmake MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ + fi done } do_install () { for d in ${RUNTIMETARGET}; do cd ${B}/${TARGET_SYS}/$d/ - oe_runmake 'DESTDIR=${D}'MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ install
meta/recipes-devtools/gcc/gcc-runtime.inc +4
-0+ if [ "$d" = "libatomic" ]; then + oe_runmake 'DESTDIR=${D}' install + else + oe_runmake 'DESTDIR=${D}'MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ install
+ fidone install -d ${D}${datadir}/gdb/auto-load/${libdir} mv ${D}${libdir}/libstdc++*-gdb.py${D}${datadir}/gdb/auto-load/${libdir}
--
2.49.0-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#237280):
https://lists.openembedded.org/g/openembedded-core/message/237280
Mute This Topic: https://lists.openembedded.org/mt/119387078/1997914
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [ [email protected]]
-=-=-=-=-=-=-=-=-=-=-=-
Re: [OE-core] [PATCH v3 07/12] gcc-runtime: avoid passing MULTIBUILDTOP to libatomic
Quoted 17 lines · Khem Raj
**
*CAUTION: This email comes from a non Wind River email account!*
Do not click links or open attachments unless you recognize the sender and know the content is safe.On Tue, May 19, 2026 at 12:26 AM Hemanth Kumar M D via lists.openembedded.org <https://urldefense.com/v3/__http://lists.openembedded.org__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIU6FfjAXI$> <[email protected]> wrote:
From: Hemanth Kumar M D <[email protected]>
libatomic constructs
gcc_objdirusingpwdand assumes MULTIBUILDTOP is relative. Passing an absolute MULTIBUILDTOP from OE results in malformed paths during libtool install operations in both compile and install stages.Avoid passing MULTIBUILDTOP for libatomic and use the upstream default handling instead.
Quoted 7 lines
Link:
https://github.com/gcc-mirror/gcc/commit/e63cf4b130b86dd7dde1bf499d3d40faca10ea2e
<https://urldefense.com/v3/__https://github.com/gcc-mirror/gcc/commit/e63cf4b130b86dd7dde1bf499d3d40faca10ea2e__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIUeqxA3yQ$>It would be good to start a discussion/bug in gcc about it, I think using pwd relative paths is perhaps less than ideal to use here. We can accept this additional logic here for OE in the meantime.
Hi Khem,
I’ve filed a GCC bug for this issue here:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125449 <https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125449>
Quoted 66 lines
Signed-off-by: Hemanth Kumar M D <[email protected]> --- meta/recipes-devtools/gcc/gcc-runtime.inc <https://urldefense.com/v3/__http://gcc-runtime.inc__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIUzaMJWEM$> | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/meta/recipes-devtools/gcc/gcc-runtime.inc <https://urldefense.com/v3/__http://gcc-runtime.inc__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIUzaMJWEM$> b/meta/recipes-devtools/gcc/gcc-runtime.inc <https://urldefense.com/v3/__http://gcc-runtime.inc__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIUzaMJWEM$> index 30b2832c82..e305180c64 100644 --- a/meta/recipes-devtools/gcc/gcc-runtime.inc <https://urldefense.com/v3/__http://gcc-runtime.inc__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIUzaMJWEM$> +++ b/meta/recipes-devtools/gcc/gcc-runtime.inc <https://urldefense.com/v3/__http://gcc-runtime.inc__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIUzaMJWEM$> @@ -83,14 +83,22 @@ do_configure[depends] += "${COMPILERDEP}" do_compile () { for d in libgcc ${RUNTIMETARGET}; do cd ${B}/${TARGET_SYS}/$d/ - oe_runmake MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ + if [ "$d" = "libatomic" ]; then + oe_runmake + else + oe_runmake MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ + fi done } do_install () { for d in ${RUNTIMETARGET}; do cd ${B}/${TARGET_SYS}/$d/ - oe_runmake 'DESTDIR=${D}' MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ install + if [ "$d" = "libatomic" ]; then + oe_runmake 'DESTDIR=${D}' install + else + oe_runmake 'DESTDIR=${D}' MULTIBUILDTOP=${B}/${TARGET_SYS}/$d/ install + fi done install -d ${D}${datadir}/gdb/auto-load/${libdir} mv ${D}${libdir}/libstdc++*-gdb.py <https://urldefense.com/v3/__http://-gdb.py__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIU-UrTfKo$> ${D}${datadir}/gdb/auto-load/${libdir} -- 2.49.0 -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#237280): https://lists.openembedded.org/g/openembedded-core/message/237280 <https://urldefense.com/v3/__https://lists.openembedded.org/g/openembedded-core/message/237280__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIUNSXecxk$> Mute This Topic: https://lists.openembedded.org/mt/119387078/1997914 <https://urldefense.com/v3/__https://lists.openembedded.org/mt/119387078/1997914__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIUa_Sd8Ug$> Group Owner: [email protected] <mailto:openembedded-core%[email protected]> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub <https://urldefense.com/v3/__https://lists.openembedded.org/g/openembedded-core/unsub__;!!AjveYdw8EvQ!fIjDCZOqeIvU5tBeDszltF5-mgerauWqLOnwdmO_dVbVBxKuk8pb1o8dHC4hYiKpJWfgElDScNKN2kIUdN2hPAk$> [[email protected]] -=-=-=-=-=-=-=-=-=-=-=-
Signature · 2 lines
Regards, Hemanth Kumar M D
PATCH v3 8/12 gcc/libgfortran: package libcaf libraries
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
Package libcaf static libraries installed by GCC 16 to avoid installed-vs-shipped QA errors.
--- meta/recipes-devtools/gcc/libgfortran_16.1.bb | 1 + 1 file changed, 1 insertion(+)
meta/recipes-devtools/gcc/libgfortran_16.1.bb +1 -0
@@ -1,3 +1,4 @@ require recipes-devtools/gcc/gcc-${PV}.inc require libgfortran.inc +FILES:${PN}-dev += "${libdir}/gcc/${TARGET_SYS}/${BINV}/libcaf*"
Signature · 1 line
2.49.0
PATCH v3 9/12 gn: fix build with GCC 16 on musl
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
Add missing <cstdint> include in src/gn/pool.h for int64_t.
Fixes:
| ../sources/gn-0+git/src/gn/pool.h:26:3: error: 'int64_t' does not name a type
Upstream-Status: Submitted [https://gn-review.googlesource.com/c/gn/+/22740]
--- ...001-gn-fix-build-with-gcc-16-on-musl.patch | 33 +++++++++++++++++++ meta/recipes-devtools/gn/gn_git.bb | 4 ++- 2 files changed, 36 insertions(+), 1 deletion(-) create mode 100644 meta/recipes-devtools/gn/files/0001-gn-fix-build-with-gcc-16-on-musl.patch
meta/recipes-devtools/gn/files/0001-gn-fix-build-with-gcc-16-on-musl.patch new file +33 -0
@@ -0,0 +1,33 @@ +From 6d22f5c9e96811f1abcedd8fb1849fe72747b5c4 Mon Sep 17 00:00:00 2001 +From: Hemanth Kumar M D <[email protected]> +Date: Mon, 11 May 2026 02:43:10 -0700 +Subject: [PATCH] gn: fix build with GCC 16 on musl + +Add missing <cstdint> include in src/gn/pool.h for int64_t. + +Fixes: +| ../sources/gn-0+git/src/gn/pool.h:26:3: error: 'int64_t' does not name a type + +Upstream-Status: Submitted [https://gn-review.googlesource.com/c/gn/+/22740] + +Signed-off-by: Hemanth Kumar M D <[email protected]> + +--- + src/gn/pool.h | 1 + + 1 file changed, 1 insertion(+) + +diff --git a/src/gn/pool.h b/src/gn/pool.h +index e03113608..247cb6425 100644 +--- a/src/gn/pool.h ++++ b/src/gn/pool.h +@@ -6,6 +6,7 @@ + #define TOOLS_GN_POOL_H_ + + #include <string> ++#include <cstdint> + + #include "gn/item.h" + +-- +2.49.0 +
meta/recipes-devtools/gn/gn_git.bb +3 -1
@@ -6,7 +6,9 @@ LIC_FILES_CHKSUM = "file://LICENSE;md5=0fca02217a5d49a14dfe2d11837bb34d" DEPENDS += "ninja-native" UPSTREAM_CHECK_COMMITS = "1" -SRC_URI = "git://gn.googlesource.com/gn;protocol=https;branch=main" +SRC_URI = "git://gn.googlesource.com/gn;protocol=https;branch=main \ + file://0001-gn-fix-build-with-gcc-16-on-musl.patch \ +" SRCREV = "eab8a9f92dca9b8548a89d9e5eb6aeb8ac6bba77" PV = "0+git"
Signature · 1 line
2.49.0
PATCH v3 10/12 elfutils: add libatomic-dev to ptest RDEPENDS
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
GCC 16 adds libatomic_asneeded.so which is referenced by the GCC spec at link time. Add libatomic-dev to ptest RDEPENDS so it is available on the target when ptests compile code, following the same pattern as libgcc-dev.
--- meta/recipes-devtools/elfutils/elfutils_0.194.bb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
meta/recipes-devtools/elfutils/elfutils_0.194.bb +1 -1
@@ -46,7 +46,7 @@ PACKAGECONFIG[zstd] = "--with-zstd,--without-zstd,zstd" PACKAGECONFIG[libdebuginfod] = "--enable-libdebuginfod,--disable-libdebuginfod,curl json-c" PACKAGECONFIG[debuginfod] = "--enable-debuginfod,--disable-debuginfod,libarchive sqlite3 libmicrohttpd" -RDEPENDS:${PN}-ptest += "libasm libelf bash make coreutils ${PN}-binutils iproute2-ss bsdtar gcc-symlinks binutils-symlinks libgcc-dev" +RDEPENDS:${PN}-ptest += "libasm libelf bash make coreutils ${PN}-binutils iproute2-ss bsdtar gcc-symlinks binutils-symlinks libgcc-dev libatomic-dev" EXTRA_OECONF:append:class-target = " --disable-tests-rpath"
Signature · 1 line
2.49.0
PATCH v3 11/12 libc-test: add libatomic-staticdev to ptest RDEPENDS
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
GCC 16 adds libatomic_asneeded.a which is required when statically linking. The libc-test static tests fail on musl without it. Add libatomic-staticdev alongside musl-staticdev for the same reason.
--- meta/recipes-core/musl/libc-test_git.bb | 1 + 1 file changed, 1 insertion(+)
meta/recipes-core/musl/libc-test_git.bb +1 -0
@@ -36,6 +36,7 @@ RDEPENDS:${PN} = " \ RDEPENDS:${PN}-ptest = " \ ${PN} \ + libatomic-staticdev \ locale-base-en-us \ musl-staticdev \ sed \
Signature · 1 line
2.49.0
Re: [OE-core] [PATCH v3 11/12] libc-test: add libatomic-staticdev to ptest RDEPENDS
Quoted 17 lines · Hemanth Kumar M D via lists.openembedded.org
From: Hemanth Kumar M D <[email protected]>
GCC 16 adds libatomic_asneeded.a which is required when statically linking. The libc-test static tests fail on musl without it. Add libatomic-staticdev alongside musl-staticdev for the same reason.
Signed-off-by Hemanth Kumar M D <[email protected]>--- meta/recipes-core/musl/libc-test_git.bb | 1 + 1 file changed, 1 insertion(+)meta/recipes-core/musl/libc-test_git.bb +0
-0
meta/recipes-core/musl/libc-test_git.bb RDEPENDS:${PN} = " \ · line 40
Quoted 5 lines
meta/recipes-core/musl/libc-test_git.bb +1
-0@@ -36,6 +36,7 @@ RDEPENDS:${PN} = " \ RDEPENDS:${PN}-ptest = " \ ${PN} \ + libatomic-staticdev \
clang wont need it.
Quoted 18 lines
meta/recipes-core/musl/libc-test_git.bb +0
-0locale-base-en-us \ musl-staticdev \sed \--
2.49.0-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#237286):
https://lists.openembedded.org/g/openembedded-core/message/237286
Mute This Topic: https://lists.openembedded.org/mt/119387084/1997914
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [ [email protected]]
-=-=-=-=-=-=-=-=-=-=-=-
Re: [OE-core] [PATCH v3 11/12] libc-test: add libatomic-staticdev to ptest RDEPENDS
Quoted 36 lines · Khem Raj
**
*CAUTION: This email comes from a non Wind River email account!*
Do not click links or open attachments unless you recognize the sender and know the content is safe.On Tue, May 19, 2026 at 12:26 AM Hemanth Kumar M D via lists.openembedded.org <https://urldefense.com/v3/__http://lists.openembedded.org__;!!AjveYdw8EvQ!a6TtOzEO1jD8FfLw2fWpo-yZ_JjSr-W7G_kpsYcQ8udyue5aM3U7JhR4Db8pt0jPv4RjP0fip0S1BqgAkHeIvgw$> <[email protected]> wrote:
From: Hemanth Kumar M D <[email protected]> GCC 16 adds libatomic_asneeded.a which is required when statically linking. The libc-test static tests fail on musl without it. Add libatomic-staticdev alongside musl-staticdev for the same reason. Signed-off-by: Hemanth Kumar M D <[email protected]> --- meta/recipes-core/musl/libc-test_git.bb <https://urldefense.com/v3/__http://libc-test_git.bb__;!!AjveYdw8EvQ!a6TtOzEO1jD8FfLw2fWpo-yZ_JjSr-W7G_kpsYcQ8udyue5aM3U7JhR4Db8pt0jPv4RjP0fip0S1BqgA4FIHuj0$> | 1 + 1 file changed, 1 insertion(+) diff --git a/meta/recipes-core/musl/libc-test_git.bb <https://urldefense.com/v3/__http://libc-test_git.bb__;!!AjveYdw8EvQ!a6TtOzEO1jD8FfLw2fWpo-yZ_JjSr-W7G_kpsYcQ8udyue5aM3U7JhR4Db8pt0jPv4RjP0fip0S1BqgA4FIHuj0$> b/meta/recipes-core/musl/libc-test_git.bb <https://urldefense.com/v3/__http://libc-test_git.bb__;!!AjveYdw8EvQ!a6TtOzEO1jD8FfLw2fWpo-yZ_JjSr-W7G_kpsYcQ8udyue5aM3U7JhR4Db8pt0jPv4RjP0fip0S1BqgA4FIHuj0$> index 9d159ea9d2..d9999ec377 100644 --- a/meta/recipes-core/musl/libc-test_git.bb <https://urldefense.com/v3/__http://libc-test_git.bb__;!!AjveYdw8EvQ!a6TtOzEO1jD8FfLw2fWpo-yZ_JjSr-W7G_kpsYcQ8udyue5aM3U7JhR4Db8pt0jPv4RjP0fip0S1BqgA4FIHuj0$> +++ b/meta/recipes-core/musl/libc-test_git.bb <https://urldefense.com/v3/__http://libc-test_git.bb__;!!AjveYdw8EvQ!a6TtOzEO1jD8FfLw2fWpo-yZ_JjSr-W7G_kpsYcQ8udyue5aM3U7JhR4Db8pt0jPv4RjP0fip0S1BqgA4FIHuj0$> @@ -36,6 +36,7 @@ RDEPENDS:${PN} = " \ RDEPENDS:${PN}-ptest = " \ ${PN} \ + libatomic-staticdev \
Quoted 2 lines
clang wont need it.
Hi Khem,
Updated this to only add it for GCC:
RDEPENDS:${PN}-ptest:append:toolchain-gcc = " libatomic-staticdev"
Will be sending the fix in the coming series.
Quoted 24 lines
locale-base-en-us \
musl-staticdev \
sed \
--
2.49.0
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#237286):
https://lists.openembedded.org/g/openembedded-core/message/237286
<https://urldefense.com/v3/__https://lists.openembedded.org/g/openembedded-core/message/237286__;!!AjveYdw8EvQ!a6TtOzEO1jD8FfLw2fWpo-yZ_JjSr-W7G_kpsYcQ8udyue5aM3U7JhR4Db8pt0jPv4RjP0fip0S1BqgAcPvrfzM$>
Mute This Topic:
https://lists.openembedded.org/mt/119387084/1997914
<https://urldefense.com/v3/__https://lists.openembedded.org/mt/119387084/1997914__;!!AjveYdw8EvQ!a6TtOzEO1jD8FfLw2fWpo-yZ_JjSr-W7G_kpsYcQ8udyue5aM3U7JhR4Db8pt0jPv4RjP0fip0S1BqgAJBrJVrs$>
Group Owner: [email protected]
<mailto:openembedded-core%[email protected]>
Unsubscribe:
https://lists.openembedded.org/g/openembedded-core/unsub
<https://urldefense.com/v3/__https://lists.openembedded.org/g/openembedded-core/unsub__;!!AjveYdw8EvQ!a6TtOzEO1jD8FfLw2fWpo-yZ_JjSr-W7G_kpsYcQ8udyue5aM3U7JhR4Db8pt0jPv4RjP0fip0S1BqgAWq8fP6c$>
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-Signature · 2 lines
Regards, Hemanth Kumar M D
PATCH v3 12/12 gcc: i386: fix ICE with -ffloat-store and SSE [PR125308]
In reply to [email protected]From: Hemanth Kumar M D <[email protected]>
Backport upstream commit c96589414eeb0e46d9fe9cc1d562e048230b5f67 to fix an ICE on i386 seen while building piglit with GCC 16. The failure happens during the IRA pass in extract_insn() when handling UNSPEC_IEEE_MAX patterns generated with -ffloat-store and SSE enabled.
--- meta/recipes-devtools/gcc/gcc-16.1.inc | 1 + ...6-fix-ICE-with-ffloat-store-PR125308.patch | 89 +++++++++++++++++++ 2 files changed, 90 insertions(+) create mode 100644 meta/recipes-devtools/gcc/gcc/0026-gcc-i386-fix-ICE-with-ffloat-store-PR125308.patch
meta/recipes-devtools/gcc/gcc-16.1.inc +1 -0
@@ -73,6 +73,7 @@ SRC_URI = "${BASEURI} \ file://0024-Avoid-hardcoded-build-paths-into-ppc-libgcc.patch \ file://0025-gcc-testsuite-tweaks-for-mips-OE.patch \ file://0001-mapper-localhost-might-not-be-known.patch \ + file://0026-gcc-i386-fix-ICE-with-ffloat-store-PR125308.patch \ " UNPACKDIR = "${TMPDIR}/work-shared/gcc-${PV}-${PR}/sources"
meta/recipes-devtools/gcc/gcc/0026-gcc-i386-fix-ICE-with-ffloat-store-PR125308.patch new file +89 -0
89 lines not on this page · raw message
Signature · 1 line
2.49.0
Quoted 8 lines · Hemanth Kumar M D via lists.openembedded.org
From: Hemanth Kumar M D <[email protected]>
This patch series upgrades GCC to 16.1 and includes the required runtime and compatibility fixes identified during testing in OE-Core. The series also includes fixes for packages failing to build with GCC 16, updates for libatomic runtime handling, and packaging changes for newly generated
libatomic_asneededfiles.
A new GCC is quite invasive, would you be able to produce buildhistory-diff reports of before/after so we can verify that there’s nothing else that needs resolving (like debuginfo that doesn’t get stripped out correctly).
Cheers,
Ross
Quoted 1 line · Ross Burton
A new GCC is quite invasive, would you be able to produce buildhistory-diff reports of before/after so we can verify that there’s nothing else that needs resolving (like debuginfo that doesn’t get stripped out correctly).
I just did this because my machine was idle and whilst it’s mostly what you’d expect, some oddities were present:
packages/cortexa57-poky-linux/glib-networking/glib-networking: PKGSIZE changed from 200832 to 266368 (+33%) packages/cortexa57-poky-linux/pulseaudio/pulseaudio-lib-protocol-native: PKGSIZE changed from 198680 to 133144 (-33%) packages/cortexa57-poky-linux/pulseaudio/pulseaudio-module-echo-cancel: PKGSIZE changed from 133144 to 67608 (-49%)
Across all of core-image-sato those were the main changes that were statistically noticeable (I think the threshold is 5%).
So I’m happy enough that there’s no glaringly obvious problems.
Ross
I have some comments for compiler portability; it almost looks ok.
Quoted 105 lines · Hemanth Kumar M D via lists.openembedded.org
From: Hemanth Kumar M D <[email protected]>
This patch series upgrades GCC to 16.1 and includes the required runtime and compatibility fixes identified during testing in OE-Core. The series also includes fixes for packages failing to build with GCC 16, updates for libatomic runtime handling, and packaging changes for newly generated
libatomic_asneededfiles.Changes in v3:
- Fix missing Signed-off-by in gn and gcc inner patch filesChanges in v2:
- Add libatomic-dev to elfutils ptest RDEPENDS
- Add libatomic-staticdev to libc-test ptest RDEPENDS
- Package new libcaf_shmem.a static library in libgfortran
- Fix gn build with GCC 16 on musl
- Backport GCC i386 ICE fix with -ffloat-store and SSE [PR125308]Summary of patches:
- Patch 01: Backport piglit fix for building with GCC 16.
- Patch 02: Backport rxvt-unicode fix for building with GCC 16.
- Patch 03: Disable automatic libatomic linking in glibc.
- Patch 04: Upgrade GCC to 16.1 release.
- Patch 05: Disable automatic libatomic linking during gcc-runtime build.
- Patch 06: Package newlibatomic_asneededruntime files.
- Patch 07: Avoid passing MULTIBUILDTOP to libatomic in OE builds.
- Patch 08: Package new libcaf_shmem.a static library in libgfortran.
- Patch 09: Fix gn build with GCC 16 on musl.
- Patch 10: Add libatomic-dev to elfutils ptest RDEPENDS.
- Patch 11: Add libatomic-staticdev to libc-test ptest RDEPENDS.
- Patch 12: Fix ICE on i386 with -ffloat-store and SSE [PR125308].meta/conf/distro/include/maintainers.inc | 2 +- meta/conf/distro/include/tcmode-default.inc | 2 +- meta/recipes-core/glibc/glibc.inc | 3 + meta/recipes-core/musl/libc-test_git.bb | 1 + .../elfutils/elfutils_0.194.bb | 2 +- .../gcc/{gcc-15.2.inc => gcc-16.1.inc} | 18 ++-- ...ian_15.2.bb => gcc-cross-canadian_16.1.bb} | 0 .../{gcc-cross_15.2.bb => gcc-cross_16.1.bb} | 0 ...-crosssdk_15.2.bb => gcc-crosssdk_16.1.bb} | 0 meta/recipes-devtools/gcc/gcc-runtime.inc | 18 +++- ...cc-runtime_15.2.bb => gcc-runtime_16.1.bb} | 0 ...itizers_15.2.bb => gcc-sanitizers_16.1.bb} | 0 ...{gcc-source_15.2.bb => gcc-source_16.1.bb} | 0 ...AMIC_LINKER-and-UCLIBC_DYNAMIC_LINKE.patch | 2 +- ...sroot-gcc-version-specific-dirs-with.patch | 4 +- ...omic-Do-not-enforce-march-on-aarch64.patch | 18 ++-- ...025-gcc-testsuite-tweaks-for-mips-OE.patch | 78 +++++++++------- ...6-fix-ICE-with-ffloat-store-PR125308.patch | 89 +++++++++++++++++++ .../gcc/{gcc_15.2.bb => gcc_16.1.bb} | 0 ...initial_15.2.bb => libgcc-initial_16.1.bb} | 0 .../gcc/{libgcc_15.2.bb => libgcc_16.1.bb} | 0 ...ibgfortran_15.2.bb => libgfortran_16.1.bb} | 1 + ...001-gn-fix-build-with-gcc-16-on-musl.patch | 33 +++++++ meta/recipes-devtools/gn/gn_git.bb | 4 +- ...h-rename-local-lerp-to-linear_interp.patch | 46 ++++++++++ meta/recipes-graphics/piglit/piglit_git.bb | 1 + .../rxvt-unicode/rxvt-unicode.inc | 3 +- .../0001-rxvt-unicode-fix-gcc16-c-20.patch | 61 +++++++++++++ 28 files changed, 323 insertions(+), 63 deletions(-) rename meta/recipes-devtools/gcc/{gcc-15.2.inc => gcc-16.1.inc} (88%) rename meta/recipes-devtools/gcc/{gcc-cross-canadian_15.2.bb => gcc-cross-canadian_16.1.bb} (100%) rename meta/recipes-devtools/gcc/{gcc-cross_15.2.bb => gcc-cross_16.1.bb} (100%) rename meta/recipes-devtools/gcc/{gcc-crosssdk_15.2.bb => gcc-crosssdk_16.1.bb} (100%) rename meta/recipes-devtools/gcc/{gcc-runtime_15.2.bb => gcc-runtime_16.1.bb} (100%) rename meta/recipes-devtools/gcc/{gcc-sanitizers_15.2.bb => gcc-sanitizers_16.1.bb} (100%) rename meta/recipes-devtools/gcc/{gcc-source_15.2.bb => gcc-source_16.1.bb} (100%) create mode 100644 meta/recipes-devtools/gcc/gcc/0026-gcc-i386-fix-ICE-with-ffloat-store-PR125308.patch rename meta/recipes-devtools/gcc/{gcc_15.2.bb => gcc_16.1.bb} (100%) rename meta/recipes-devtools/gcc/{libgcc-initial_15.2.bb => libgcc-initial_16.1.bb} (100%) rename meta/recipes-devtools/gcc/{libgcc_15.2.bb => libgcc_16.1.bb} (100%) rename meta/recipes-devtools/gcc/{libgfortran_15.2.bb => libgfortran_16.1.bb} (51%) create mode 100644 meta/recipes-devtools/gn/files/0001-gn-fix-build-with-gcc-16-on-musl.patch create mode 100644 meta/recipes-graphics/piglit/piglit/0001-fbo-blit-stretch-rename-local-lerp-to-linear_interp.patch create mode 100644
meta/recipes-sato/rxvt-unicode/rxvt-unicode/0001-rxvt-unicode-fix-gcc16-c-20.patch--
2.49.0-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#237275):
https://lists.openembedded.org/g/openembedded-core/message/237275
Mute This Topic: https://lists.openembedded.org/mt/119387073/1997914
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [ [email protected]]
-=-=-=-=-=-=-=-=-=-=-=-