flex: fix stage1flex build under a C23 default
PATCH 1/1 flex: fix stage1flex build under a C23 default
lib/malloc.c declares "void *malloc ();" with an empty parameter list and calls it with one argument. That meant "unspecified" in C89 and means "none" in C23, so the call is rejected:
lib/malloc.c:16:15: error: too many arguments to function 'malloc';
expected 0, have 1
It is the language mode that decides, not the compiler version: the same GCC 16 compiles the file with -std=gnu17 and rejects it with -std=gnu23. GCC 15 made gnu23 the default, so every host from that release on hits it while GCC 14 does not.
The file is compiled by stage1flex -- the bootstrap scanner flex builds with the host compiler before it can build itself -- so neither CFLAGS nor BUILD_CFLAGS reaches that command line and no flag in the recipe can silence it.
stdlib.h has the right declaration and the file already includes sys/types.h for size_t. lib/realloc.c includes stdlib.h already and needs no change.
--- ...b-malloc-declare-malloc-via-stdlib.h.patch | 38 +++++++++++++++++++ meta/recipes-devtools/flex/flex_2.6.4.bb | 1 + 2 files changed, 39 insertions(+) create mode 100644 meta/recipes-devtools/flex/flex/0001-lib-malloc-declare-malloc-via-stdlib.h.patch
meta/recipes-devtools/flex/flex/0001-lib-malloc-declare-malloc-via-stdlib.h.patch new file +38 -0
@@ -0,0 +1,38 @@ +From: Aurelien Desbrieres <[email protected]> +Date: Sat, 13 Sep 2026 00:00:00 +0200 +Subject: [PATCH] lib/malloc.c: declare malloc via stdlib.h + +The gnulib fallback declares "void *malloc ();" with an empty parameter +list and calls it with one argument. That meant "unspecified" in C89 and +means "none" in C23, which GCC 14 and later implement by default, so the +call is rejected: + + lib/malloc.c:16:15: error: too many arguments to function 'malloc'; + expected 0, have 1 + +stdlib.h has the right declaration and the file already includes +sys/types.h for size_t, so the local one has nothing to add. +lib/realloc.c includes stdlib.h already and needs no change. + +The file is dead code wherever malloc(0) returns non-NULL -- glibc +included -- since AC_FUNC_MALLOC substitutes rpl_malloc only where it +does not, but it is compiled regardless and the build stops there. + +Upstream-Status: Inappropriate [flex 2.6.4 is the last release, 2017] + +Signed-off-by: Aurelien Desbrieres <[email protected]> +--- + lib/malloc.c | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +--- a/lib/malloc.c ++++ b/lib/malloc.c +@@ -3,7 +3,7 @@ + + #include <sys/types.h> + +- void *malloc (); ++#include <stdlib.h> + + /* Allocate an N-byte block of memory from the heap. + If N is zero, allocate a 1-byte block. */
meta/recipes-devtools/flex/flex_2.6.4.bb +1 -0
@@ -19,6 +19,7 @@ SRC_URI = "${GITHUB_BASE_URI}/download/v${PV}/flex-${PV}.tar.gz \ file://check-funcs.patch \ file://0001-Emit-no-line-directives-if-gen_line_dirs-is-false.patch \ file://0001-build-tests-add-missing-parser-scanner-dependencies.patch \ + file://0001-lib-malloc-declare-malloc-via-stdlib.h.patch \ " SRC_URI[md5sum] = "2882e3179748cc9f9c23ec593d6adc8d"
Signature · 1 line
2.55.0
Quoted 45 lines · Aurelien DESBRIERES via lists.openembedded.org
lib/malloc.cdeclares "void *malloc ();" with an empty parameter list and calls it with one argument. That meant "unspecified" in C89 and means "none" in C23, so the call is rejected:
lib/malloc.c:16:15: error: too many arguments to function 'malloc'; expected 0, have 1It is the language mode that decides, not the compiler version: the same GCC 16 compiles the file with -std=gnu17 and rejects it with -std=gnu23. GCC 15 made gnu23 the default, so every host from that release on hits it while GCC 14 does not.
The file is compiled by stage1flex -- the bootstrap scanner flex builds with the host compiler before it can build itself -- so neither CFLAGS nor
BUILD_CFLAGSreaches that command line and no flag in the recipe can silence it.
stdlib.hhas the right declaration and the file already includessys/types.hforsize_t.lib/realloc.cincludesstdlib.halready and needs no change.Signed-off-by Aurelien DESBRIERES <[email protected]>---
...b-malloc-declare-malloc-via-stdlib.h.patch | 38 +++++++++++++++++++ meta/recipes-devtools/flex/flex_2.6.4.bb | 1 +2 files changed, 39 insertions(+)create mode 100644
meta/recipes-devtools/flex/flex/0001-lib-malloc-declare-malloc-via-stdlib.h.patchmeta/recipes-devtools/flex/flex/0001-lib-malloc-declare-malloc-via-stdlib.h.patch new file +11
-0@@ -0,0 +1,38 @@ +From: Aurelien Desbrieres <[email protected]> +Date: Sat, 13 Sep 2026 00:00:00 +0200 +Subject: [PATCH] lib/malloc.c: declare malloc via stdlib.h + +The gnulib fallback declares "void *malloc ();" with an empty parameter +list and calls it with one argument. That meant "unspecified" in C89 and +means "none" in C23, which GCC 14 and later implement by default, so the +call is rejected: + + lib/malloc.c:16:15: error: too many arguments to function 'malloc'; + expected 0, have 1
meta/recipes-devtools/flex/flex/0001-lib-malloc-declare-malloc-via-stdlib.h.patch line 21
Quoted 10 lines
meta/recipes-devtools/flex/flex/0001-lib-malloc-declare-malloc-via-stdlib.h.patch new file +10
-0+ +stdlib.h has the right declaration and the file already includes +sys/types.h for size_t, so the local one has nothing to add. +lib/realloc.c includes stdlib.h already and needs no change. + +The file is dead code wherever malloc(0) returns non-NULL -- glibc +included -- since AC_FUNC_MALLOC substitutes rpl_malloc only where it +does not, but it is compiled regardless and the build stops there. + +Upstream-Status: Inappropriate [flex 2.6.4 is the last release, 2017]
I'm not sure I follow that reasoning. It might be better to follow what upstream did:
https://github.com/westes/flex/commit/bf254c75b1e0d2641ebbd7fc85fb183f36a62ea7
so this patch is then a backport and will fall out if/as/when we do see another release of flex?
Cheers,
Richard
PATCH v2 1/1 flex: backport the upstream malloc prototype fix
lib/malloc.c declares "void *malloc ();" with an empty parameter list and calls it with one argument. That meant "unspecified" in C89 and means "none" in C23, so a host compiler defaulting to C23 -- GCC 15 and later -- rejects the call:
lib/malloc.c:16:15: error: too many arguments to function 'malloc';
expected 0, have 1
The file is built by stage1flex with the build host's compiler before flex can build itself, so neither CFLAGS nor BUILD_CFLAGS reaches that command line and no flag in the recipe can silence it.
Upstream fixed it in bf254c75b1e0 and the patch falls out when flex next releases. Gentoo carries the same backport as flex-2.6.4-fix-malloc-prototype.patch.
--- ....c-match-malloc-signature-to-its-use.patch | 37 +++++++++++++++++++ meta/recipes-devtools/flex/flex_2.6.4.bb | 1 + 2 files changed, 38 insertions(+) create mode 100644 meta/recipes-devtools/flex/flex/0001-lib-malloc.c-match-malloc-signature-to-its-use.patch
meta/recipes-devtools/flex/flex/0001-lib-malloc.c-match-malloc-signature-to-its-use.patch new file +37 -0
@@ -0,0 +1,37 @@ +From bf254c75b1e0d2641ebbd7fc85fb183f36a62ea7 Mon Sep 17 00:00:00 2001 +From: Richard Barnes <[email protected]> +Date: Wed, 2 Oct 2024 10:35:09 -0700 +Subject: [PATCH] Match `malloc` signature to its use + +The gnulib fallback declares "void *malloc ();" with an empty parameter +list and calls it with one argument. That meant "unspecified" in C89 and +means "none" in C23, so a compiler defaulting to C23 -- GCC 15 and later +-- rejects the call: + + lib/malloc.c:16:15: error: too many arguments to function 'malloc'; + expected 0, have 1 + +The file is built by stage1flex with the build host's compiler before +flex can build itself, so neither CFLAGS nor BUILD_CFLAGS reaches that +command line and no flag in the recipe can silence it. + +Upstream-Status: Backport [https://github.com/westes/flex/commit/bf254c75b1e0d2641ebbd7fc85fb183f36a62ea7] + +Signed-off-by: Aurelien DESBRIERES <[email protected]> +--- + lib/malloc.c | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/lib/malloc.c b/lib/malloc.c +index 75e8ef97c..701b9b39d 100755 +--- a/lib/malloc.c ++++ b/lib/malloc.c +@@ -3,7 +3,7 @@ + + #include <sys/types.h> + +- void *malloc (); ++ void *malloc (size_t n); + + /* Allocate an N-byte block of memory from the heap. + If N is zero, allocate a 1-byte block. */
meta/recipes-devtools/flex/flex_2.6.4.bb +1 -0
@@ -19,6 +19,7 @@ SRC_URI = "${GITHUB_BASE_URI}/download/v${PV}/flex-${PV}.tar.gz \ file://check-funcs.patch \ file://0001-Emit-no-line-directives-if-gen_line_dirs-is-false.patch \ file://0001-build-tests-add-missing-parser-scanner-dependencies.patch \ + file://0001-lib-malloc.c-match-malloc-signature-to-its-use.patch \ " SRC_URI[md5sum] = "2882e3179748cc9f9c23ec593d6adc8d"
Signature · 1 line
2.55.0
Re: [OE-core] [PATCH v2] flex: backport the upstream malloc prototype fix
Quoted 8 lines · Aurelien DESBRIERES
lib/malloc.cdeclares "void *malloc ();" with an empty parameter list and calls it with one argument. That meant "unspecified" in C89 and means "none" in C23, so a host compiler defaulting to C23 -- GCC 15 and later -- rejects the call:
lib/malloc.c:16:15: error: too many arguments to function 'malloc';
expected 0, have 1
Quoted 7 lines
The file is built by stage1flex with the build host's compiler before flex can build itself, so neither CFLAGS nor
BUILD_CFLAGSreaches that command line and no flag in the recipe can silence it.Upstream fixed it in
bf254c75b1e0and the patch falls out when flex next releases. Gentoo carries the same backport as flex-2.6.4-fix-malloc-prototype.patch.
This patch is already applied to flex on the master branch, from commit e1832f02d70b ("flex: fix build with gcc-15 on host"). Maybe you need to rebase?
Antonin
Re: [OE-core] [PATCH v2] flex: backport the upstream malloc prototype fix
Thanks, you are right -- 98f2376d82 carries the same upstream commit and I was working from walnascar. Please drop this one.
Aurelien