Please pull this fix for a build regression in seccomp.
The following changes since commit 7b55851367136b1efd84d98fea81ba57a98304cf:
fork: record start_time late (2019-01-08 09:40:53 -0800)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security.git fixes-v5.0-rc1
for you to fetch changes up to cba54b44d0be4eb66dbc7709e1f3f0d65e851f69:
samples/seccomp: fix 32-bit build (2019-01-08 13:00:00 -0800)
----------------------------------------------------------------
Tycho Andersen (1):
samples/seccomp: fix 32-bit build
samples/seccomp/Makefile | 1 +
1 file changed, 1 insertion(+)
---
commit cba54b44d0be4eb66dbc7709e1f3f0d65e851f69
Author: Tycho Andersen <[email protected]>
Date: Mon Jan 7 14:46:34 2019 -0700
samples/seccomp: fix 32-bit build
Both the .o and the actual executable need to be built with -m32 in order
to link correctly.
Fixes: fec7b6690541 ("samples: add an example of seccomp user trap")
Signed-off-by: Tycho Andersen <[email protected]>
Reported-by: Ingo Molnar <[email protected]>
Signed-off-by: James Morris <[email protected]>
diff --git a/samples/seccomp/Makefile b/samples/seccomp/Makefile
index 4920903c8009..a5607668a5c7 100644
--- a/samples/seccomp/Makefile
+++ b/samples/seccomp/Makefile
@@ -37,6 +37,7 @@ HOSTCFLAGS_bpf-fancy.o += $(MFLAG)
HOSTLDLIBS_bpf-direct += $(MFLAG)
HOSTLDLIBS_bpf-fancy += $(MFLAG)
HOSTLDLIBS_dropper += $(MFLAG)
+HOSTLDLIBS_user-trap.o += $(MFLAG)
HOSTLDLIBS_user-trap += $(MFLAG)
endif
always := $(hostprogs-m)
This was already picked up by x86-urgent...
-Kees
On Tue, Jan 8, 2019 at 1:04 PM James Morris <[email protected]> wrote:
>
> Please pull this fix for a build regression in seccomp.
>
> The following changes since commit 7b55851367136b1efd84d98fea81ba57a98304cf:
>
> fork: record start_time late (2019-01-08 09:40:53 -0800)
>
> are available in the Git repository at:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security.git fixes-v5.0-rc1
>
> for you to fetch changes up to cba54b44d0be4eb66dbc7709e1f3f0d65e851f69:
>
> samples/seccomp: fix 32-bit build (2019-01-08 13:00:00 -0800)
>
> ----------------------------------------------------------------
>
> Tycho Andersen (1):
> samples/seccomp: fix 32-bit build
>
> samples/seccomp/Makefile | 1 +
> 1 file changed, 1 insertion(+)
>
> ---
>
> commit cba54b44d0be4eb66dbc7709e1f3f0d65e851f69
> Author: Tycho Andersen <[email protected]>
> Date: Mon Jan 7 14:46:34 2019 -0700
>
> samples/seccomp: fix 32-bit build
>
> Both the .o and the actual executable need to be built with -m32 in order
> to link correctly.
>
> Fixes: fec7b6690541 ("samples: add an example of seccomp user trap")
>
> Signed-off-by: Tycho Andersen <[email protected]>
> Reported-by: Ingo Molnar <[email protected]>
> Signed-off-by: James Morris <[email protected]>
>
> diff --git a/samples/seccomp/Makefile b/samples/seccomp/Makefile
> index 4920903c8009..a5607668a5c7 100644
> --- a/samples/seccomp/Makefile
> +++ b/samples/seccomp/Makefile
> @@ -37,6 +37,7 @@ HOSTCFLAGS_bpf-fancy.o += $(MFLAG)
> HOSTLDLIBS_bpf-direct += $(MFLAG)
> HOSTLDLIBS_bpf-fancy += $(MFLAG)
> HOSTLDLIBS_dropper += $(MFLAG)
> +HOSTLDLIBS_user-trap.o += $(MFLAG)
> HOSTLDLIBS_user-trap += $(MFLAG)
> endif
> always := $(hostprogs-m)
--
Kees Cook
* Kees Cook <[email protected]> wrote:
> This was already picked up by x86-urgent...
>
> -Kees
I'm fine with both routes - if Linus pulls this I'll zap the x86/urgent
one.
Thanks,
Ingo
On Wed, Jan 09, 2019 at 07:59:16AM +0100, Ingo Molnar wrote:
>
> * Kees Cook <[email protected]> wrote:
>
> > This was already picked up by x86-urgent...
> >
> > -Kees
>
> I'm fine with both routes - if Linus pulls this I'll zap the x86/urgent
> one.
The patches are different actually. I've confirmed the one in
x86/urgent is correct, and this one is from the initial thread where I
was just guessing. The difference is HOSTCFLAGS vs. HOSTLDFLAGS for .o
files. So I think we should drop this and just keep the one in
x86/urgent.
Thanks,
Tycho
* Tycho Andersen <[email protected]> wrote:
> On Wed, Jan 09, 2019 at 07:59:16AM +0100, Ingo Molnar wrote:
> >
> > * Kees Cook <[email protected]> wrote:
> >
> > > This was already picked up by x86-urgent...
> > >
> > > -Kees
> >
> > I'm fine with both routes - if Linus pulls this I'll zap the x86/urgent
> > one.
>
> The patches are different actually. I've confirmed the one in
> x86/urgent is correct, and this one is from the initial thread where I
> was just guessing. The difference is HOSTCFLAGS vs. HOSTLDFLAGS for .o
> files. So I think we should drop this and just keep the one in
> x86/urgent.
Ok - will send this to Linus now.
Thanks,
Ingo