Hi, we have to revert this patches back to 9.1.4 because it still not 
work
with the kernel-4.14.x headers. On i686 it always crash with "Bad System 
Call".
Looks like not all Systemcalls are defined in the headers.

This version build with kernel-5.10 works but i have an other problem:
On x86_64 i often got no IP at boot but if i rerun via connscheduler
reconnect it works.

I think this need intensive testing on all platforms.

Arne


Am 2020-12-29 13:36, schrieb Matthias Fischer:
> For details see:
> https://roy.marples.name/archives/dhcpcd-discuss/0003420.html
> 
> Former patch for Bug #12552 is now included.
> 
> Signed-off-by: Matthias Fischer <matthias.fischer(a)ipfire.org>
> ---
>  lfs/dhcpcd                                    |  7 ++--
>  ...r_SECCOMP_as_it_just_uses_socketcall.patch | 36 -------------------
>  2 files changed, 2 insertions(+), 41 deletions(-)
>  delete mode 100644
> src/patches/dhcpcd/01_Fix_Linux_i386_for_SECCOMP_as_it_just_uses_socketcall.patch
> 
> diff --git a/lfs/dhcpcd b/lfs/dhcpcd
> index 4e34e19d5..352308692 100644
> --- a/lfs/dhcpcd
> +++ b/lfs/dhcpcd
> @@ -24,7 +24,7 @@
> 
>  include Config
> 
> -VER        = 9.3.4
> +VER        = 9.4.0
> 
>  THISAPP    = dhcpcd-$(VER)
>  DL_FILE    = $(THISAPP).tar.xz
> @@ -40,7 +40,7 @@ objects = $(DL_FILE)
> 
>  $(DL_FILE) = $(DL_FROM)/$(DL_FILE)
> 
> -$(DL_FILE)_MD5 = badb02dfc69fe9bbeec35a02efcdb4db
> +$(DL_FILE)_MD5 = c36715fc629bc40aa94aae06fa1724c2
> 
>  install : $(TARGET)
> 
> @@ -70,9 +70,6 @@ $(subst %,%_MD5,$(objects)) :
>  $(TARGET) : $(patsubst %,$(DIR_DL)/%,$(objects))
>  	@$(PREBUILD)
>  	@rm -rf $(DIR_APP) && cd $(DIR_SRC) && tar axf $(DIR_DL)/$(DL_FILE)
> -
> -	cd $(DIR_APP) && patch -Np1 -i
> $(DIR_SRC)/src/patches/dhcpcd/01_Fix_Linux_i386_for_SECCOMP_as_it_just_uses_socketcall.patch
> -
>  	cd $(DIR_APP) && ./configure --prefix="" 
> --sysconfdir=/var/ipfire/dhcpc \
>  			--dbdir=/var/ipfire/dhcpc \
>  			--libexecdir=/var/ipfire/dhcpc \
> diff --git
> a/src/patches/dhcpcd/01_Fix_Linux_i386_for_SECCOMP_as_it_just_uses_socketcall.patch
> b/src/patches/dhcpcd/01_Fix_Linux_i386_for_SECCOMP_as_it_just_uses_socketcall.patch
> deleted file mode 100644
> index 9efcde219..000000000
> ---
> a/src/patches/dhcpcd/01_Fix_Linux_i386_for_SECCOMP_as_it_just_uses_socketcall.patch
> +++ /dev/null
> @@ -1,36 +0,0 @@
> -diff --git a/src/privsep-linux.c b/src/privsep-linux.c
> -index 050a30cf..d31d720d 100644
> ---- a/src/privsep-linux.c
> -+++ b/src/privsep-linux.c
> -@@ -32,6 +32,7 @@
> -
> - #include <linux/audit.h>
> - #include <linux/filter.h>
> -+#include <linux/net.h>
> - #include <linux/seccomp.h>
> - #include <linux/sockios.h>
> -
> -@@ -304,6 +305,23 @@ static struct sock_filter ps_seccomp_filter[] = {
> - #ifdef __NR_sendto
> - 	SECCOMP_ALLOW(__NR_sendto),
> - #endif
> -+#ifdef __NR_socketcall
> -+	/* i386 needs this and demonstrates why SECCOMP
> -+	 * is poor compared to OpenBSD pledge(2) and FreeBSD capsicum(4)
> -+	 * as this is soooo tied to the kernel API which changes per arch
> -+	 * and likely libc as well. */
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_ACCEPT),
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_ACCEPT4),
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_LISTEN),
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_GETSOCKOPT),	/* overflow 
> */
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_RECV),
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_RECVFROM),
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_RECVMSG),
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_SEND),
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_SENDMSG),
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_SENDTO),
> -+	SECCOMP_ALLOW_ARG(__NR_socketcall, 0, SYS_SHUTDOWN),
> -+#endif
> - #ifdef __NR_shutdown
> - 	SECCOMP_ALLOW(__NR_shutdown),
> - #endif