[PATCH] arch: um: Mark the stack non-executable to fix a binutils warning

692 views
Skip to first unread message

David Gow

unread,
Aug 22, 2022, 9:09:01 PM8/22/22
to Richard Weinberger, Anton Ivanov, Johannes Berg, Nick Desaulniers, David Gow, Thomas Gleixner, Ingo Molnar, Dave Hansen, Linus Torvalds, Brendan Higgins, Daniel Latypov, linu...@lists.infradead.org, linux-...@vger.kernel.org, kuni...@googlegroups.com, x...@kernel.org
Since binutils 2.39, ld will print a warning if any stack section is
executable, which is the default for stack sections on files without a
.note.GNU-stack section.

This was fixed for x86 in commit ffcf9c5700e4 ("x86: link vdso and boot with -z noexecstack --no-warn-rwx-segments"),
but remained broken for UML, resulting in several warnings:

/usr/bin/ld: warning: arch/x86/um/vdso/vdso.o: missing .note.GNU-stack section implies executable stack
/usr/bin/ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker
/usr/bin/ld: warning: .tmp_vmlinux.kallsyms1 has a LOAD segment with RWX permissions
/usr/bin/ld: warning: .tmp_vmlinux.kallsyms1.o: missing .note.GNU-stack section implies executable stack
/usr/bin/ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker
/usr/bin/ld: warning: .tmp_vmlinux.kallsyms2 has a LOAD segment with RWX permissions
/usr/bin/ld: warning: .tmp_vmlinux.kallsyms2.o: missing .note.GNU-stack section implies executable stack
/usr/bin/ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker
/usr/bin/ld: warning: vmlinux has a LOAD segment with RWX permissions

Link both the VDSO and vmlinux with -z noexecstack, fixing the warnings
about .note.GNU-stack sections. In addition, pass --no-warn-rwx-segments
to dodge the remaining warnings about LOAD segments with RWX permissions
in the kallsyms objects. (Note that this flag is apparently not
available on lld, so hide it behind a test for BFD, which is what the
x86 patch does.)

Link: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=ffcf9c5700e49c0aee42dcba9a12ba21338e8136
Link: https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=ba951afb99912da01a6e8434126b8fac7aa75107
Signed-off-by: David Gow <davi...@google.com>
---
arch/um/Makefile | 9 ++++++++-
arch/x86/um/vdso/Makefile | 2 +-
2 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/arch/um/Makefile b/arch/um/Makefile
index f2fe63bfd819..75d5c704b3a8 100644
--- a/arch/um/Makefile
+++ b/arch/um/Makefile
@@ -132,10 +132,17 @@ export LDS_ELF_FORMAT := $(ELF_FORMAT)
# The wrappers will select whether using "malloc" or the kernel allocator.
LINK_WRAPS = -Wl,--wrap,malloc -Wl,--wrap,free -Wl,--wrap,calloc

+# Avoid binutils 2.39+ warnings by marking the stack non-executable and
+# ignorning warnings for the kallsyms sections.
+LINK_RWXSECTION = -Wl,-z,noexecstack
+ifeq ($(CONFIG_LD_IS_BFD),y)
+LINK_RWXSECTION += -Wl,--no-warn-rwx-segments
+endif
+
LD_FLAGS_CMDLINE = $(foreach opt,$(KBUILD_LDFLAGS),-Wl,$(opt))

# Used by link-vmlinux.sh which has special support for um link
-export CFLAGS_vmlinux := $(LINK-y) $(LINK_WRAPS) $(LD_FLAGS_CMDLINE)
+export CFLAGS_vmlinux := $(LINK-y) $(LINK_WRAPS) $(LINK_RWXSECTION) $(LD_FLAGS_CMDLINE)

# When cleaning we don't include .config, so we don't include
# TT or skas makefiles and don't clean skas_ptregs.h.
diff --git a/arch/x86/um/vdso/Makefile b/arch/x86/um/vdso/Makefile
index 8c0396fd0e6f..6fbe97c52c99 100644
--- a/arch/x86/um/vdso/Makefile
+++ b/arch/x86/um/vdso/Makefile
@@ -65,7 +65,7 @@ quiet_cmd_vdso = VDSO $@
-Wl,-T,$(filter %.lds,$^) $(filter %.o,$^) && \
sh $(srctree)/$(src)/checkundef.sh '$(NM)' '$@'

-VDSO_LDFLAGS = -fPIC -shared -Wl,--hash-style=sysv
+VDSO_LDFLAGS = -fPIC -shared -Wl,--hash-style=sysv -z noexecstack
GCOV_PROFILE := n

#
--
2.37.1.595.g718a3a8f04-goog

Randy Dunlap

unread,
Aug 26, 2022, 4:35:38 PM8/26/22
to David Gow, Richard Weinberger, Anton Ivanov, Johannes Berg, Nick Desaulniers, Thomas Gleixner, Ingo Molnar, Dave Hansen, Linus Torvalds, Brendan Higgins, Daniel Latypov, linu...@lists.infradead.org, linux-...@vger.kernel.org, kuni...@googlegroups.com, x...@kernel.org


On 8/22/22 18:08, David Gow wrote:
> Since binutils 2.39, ld will print a warning if any stack section is
> executable, which is the default for stack sections on files without a
> .note.GNU-stack section.
>
> This was fixed for x86 in commit ffcf9c5700e4 ("x86: link vdso and boot with -z noexecstack --no-warn-rwx-segments"),
> but remained broken for UML, resulting in several warnings:
>
> /usr/bin/ld: warning: arch/x86/um/vdso/vdso.o: missing .note.GNU-stack section implies executable stack
> /usr/bin/ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker
> /usr/bin/ld: warning: .tmp_vmlinux.kallsyms1 has a LOAD segment with RWX permissions
> /usr/bin/ld: warning: .tmp_vmlinux.kallsyms1.o: missing .note.GNU-stack section implies executable stack
> /usr/bin/ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker
> /usr/bin/ld: warning: .tmp_vmlinux.kallsyms2 has a LOAD segment with RWX permissions
> /usr/bin/ld: warning: .tmp_vmlinux.kallsyms2.o: missing .note.GNU-stack section implies executable stack
> /usr/bin/ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker
> /usr/bin/ld: warning: vmlinux has a LOAD segment with RWX permissions
>
> Link both the VDSO and vmlinux with -z noexecstack, fixing the warnings
> about .note.GNU-stack sections. In addition, pass --no-warn-rwx-segments
> to dodge the remaining warnings about LOAD segments with RWX permissions
> in the kallsyms objects. (Note that this flag is apparently not
> available on lld, so hide it behind a test for BFD, which is what the
> x86 patch does.)
>
> Link: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=ffcf9c5700e49c0aee42dcba9a12ba21338e8136
> Link: https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=ba951afb99912da01a6e8434126b8fac7aa75107
> Signed-off-by: David Gow <davi...@google.com>

Acked-by: Randy Dunlap <rdu...@infradead.org> # build-tested

Thanks.

> ---
> arch/um/Makefile | 9 ++++++++-
> arch/x86/um/vdso/Makefile | 2 +-
> 2 files changed, 9 insertions(+), 2 deletions(-)


--
~Randy

Richard Weinberger

unread,
Sep 19, 2022, 4:56:48 PM9/19/22
to Randy Dunlap, davidgow, anton ivanov, Johannes Berg, Nick Desaulniers, tglx, mingo, dave hansen, torvalds, Brendan Higgins, Daniel Latypov, linux-um, linux-kernel, kuni...@googlegroups.com, x86
----- Ursprüngliche Mail -----
> Von: "Randy Dunlap" <rdu...@infradead.org>
This patch causes a build error on one of my systems:

/usr/lib64/gcc/x86_64-suse-linux/7/../../../../x86_64-suse-linux/bin/ld: unrecognized option '--no-warn-rwx-segments'
/usr/lib64/gcc/x86_64-suse-linux/7/../../../../x86_64-suse-linux/bin/ld: use the --help option for usage information
collect2: error: ld returned 1 exit status

Just like in commit ffcf9c5700e4 ("x86: link vdso and boot with -z noexecstack --no-warn-rwx-segments") you need to
test for it.

Thanks,
//richard

David Gow

unread,
Sep 21, 2022, 2:49:03 AM9/21/22
to Richard Weinberger, Anton Ivanov, Johannes Berg, Nick Desaulniers, David Gow, Thomas Gleixner, Ingo Molnar, Dave Hansen, Linus Torvalds, Brendan Higgins, Daniel Latypov, linu...@lists.infradead.org, linux-...@vger.kernel.org, kuni...@googlegroups.com, x...@kernel.org, Lukas Straub, Randy Dunlap
Since binutils 2.39, ld will print a warning if any stack section is
executable, which is the default for stack sections on files without a
.note.GNU-stack section.

This was fixed for x86 in commit ffcf9c5700e4 ("x86: link vdso and boot with -z noexecstack --no-warn-rwx-segments"),
but remained broken for UML, resulting in several warnings:

/usr/bin/ld: warning: arch/x86/um/vdso/vdso.o: missing .note.GNU-stack section implies executable stack
/usr/bin/ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker
/usr/bin/ld: warning: .tmp_vmlinux.kallsyms1 has a LOAD segment with RWX permissions
/usr/bin/ld: warning: .tmp_vmlinux.kallsyms1.o: missing .note.GNU-stack section implies executable stack
/usr/bin/ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker
/usr/bin/ld: warning: .tmp_vmlinux.kallsyms2 has a LOAD segment with RWX permissions
/usr/bin/ld: warning: .tmp_vmlinux.kallsyms2.o: missing .note.GNU-stack section implies executable stack
/usr/bin/ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker
/usr/bin/ld: warning: vmlinux has a LOAD segment with RWX permissions

Link both the VDSO and vmlinux with -z noexecstack, fixing the warnings
about .note.GNU-stack sections. In addition, pass --no-warn-rwx-segments
to dodge the remaining warnings about LOAD segments with RWX permissions
in the kallsyms objects. (Note that this flag is apparently not
available on lld, so hide it behind a test for BFD, which is what the
x86 patch does.)

Reviewed-by: Lukas Straub <lukass...@web.de>
Tested-by: Lukas Straub <lukass...@web.de>
Acked-by: Randy Dunlap <rdu...@infradead.org> # build-tested
---

Note that this still doesn't seem to be working properly with make
LLVM=1. It doesn't appear to break anything, and still is an improvement
for gcc, so seems worthwhile anyway...

Changes since v1:
http://lists.infradead.org/pipermail/linux-um/2022-August/004234.html
- Pass the -z noexecstack and --no-warn-rwx-segments flags as LDFLAGS,
rather than as CFLAGS via -Wl
- Check that --no-warn-rwx-segments exists with the ld-option function
(Thanks Richard)
- Add Lukas and Randy's tags.


arch/um/Makefile | 8 ++++++++
arch/x86/um/vdso/Makefile | 2 +-
2 files changed, 9 insertions(+), 1 deletion(-)

diff --git a/arch/um/Makefile b/arch/um/Makefile
index f2fe63bfd819..f1d4d67157be 100644
--- a/arch/um/Makefile
+++ b/arch/um/Makefile
@@ -132,10 +132,18 @@ export LDS_ELF_FORMAT := $(ELF_FORMAT)
# The wrappers will select whether using "malloc" or the kernel allocator.
LINK_WRAPS = -Wl,--wrap,malloc -Wl,--wrap,free -Wl,--wrap,calloc

+# Avoid binutils 2.39+ warnings by marking the stack non-executable and
+# ignorning warnings for the kallsyms sections.
+LDFLAGS_EXECSTACK = -z noexecstack
+ifeq ($(CONFIG_LD_IS_BFD),y)
+LDFLAGS_EXECSTACK += $(call ld-option,--no-warn-rwx-segments)
+endif
+
LD_FLAGS_CMDLINE = $(foreach opt,$(KBUILD_LDFLAGS),-Wl,$(opt))

# Used by link-vmlinux.sh which has special support for um link
export CFLAGS_vmlinux := $(LINK-y) $(LINK_WRAPS) $(LD_FLAGS_CMDLINE)
+export LDFLAGS_vmlinux := $(LDFLAGS_EXECSTACK)

# When cleaning we don't include .config, so we don't include
# TT or skas makefiles and don't clean skas_ptregs.h.
diff --git a/arch/x86/um/vdso/Makefile b/arch/x86/um/vdso/Makefile
index 8c0396fd0e6f..6fbe97c52c99 100644
--- a/arch/x86/um/vdso/Makefile
+++ b/arch/x86/um/vdso/Makefile
@@ -65,7 +65,7 @@ quiet_cmd_vdso = VDSO $@
-Wl,-T,$(filter %.lds,$^) $(filter %.o,$^) && \
sh $(srctree)/$(src)/checkundef.sh '$(NM)' '$@'

-VDSO_LDFLAGS = -fPIC -shared -Wl,--hash-style=sysv
+VDSO_LDFLAGS = -fPIC -shared -Wl,--hash-style=sysv -z noexecstack
GCOV_PROFILE := n

#
--
2.37.3.968.ga6b4b080e4-goog

Richard Weinberger

unread,
Sep 21, 2022, 3:14:25 AM9/21/22
to davidgow, anton ivanov, Johannes Berg, Nick Desaulniers, tglx, mingo, dave hansen, torvalds, Brendan Higgins, Daniel Latypov, linux-um, linux-kernel, kunit-dev, x86, Lukas Straub, Randy Dunlap
----- Ursprüngliche Mail -----
> Von: "davidgow" <davi...@google.com>
>
> Changes since v1:
> http://lists.infradead.org/pipermail/linux-um/2022-August/004234.html
> - Pass the -z noexecstack and --no-warn-rwx-segments flags as LDFLAGS,
> rather than as CFLAGS via -Wl
> - Check that --no-warn-rwx-segments exists with the ld-option function
> (Thanks Richard)
> - Add Lukas and Randy's tags.

Now it passes the test. :)
Applied.

Thanks,
//richard

Nick Desaulniers

unread,
Sep 21, 2022, 1:47:30 PM9/21/22
to David Gow, Richard Weinberger, Anton Ivanov, Johannes Berg, Thomas Gleixner, Ingo Molnar, Dave Hansen, Linus Torvalds, Brendan Higgins, Daniel Latypov, linu...@lists.infradead.org, linux-...@vger.kernel.org, kuni...@googlegroups.com, x...@kernel.org, Lukas Straub, Randy Dunlap, clang-built-linux
Hi David,
Do you have more info about this comment? Perhaps there's more
hermiticity issues with ARCH=um when a linker is explicitly specified
via LD= or HOSTLD=, or implied via LLVM=1.

Looking at our CI for ARCH=um builds...
https://github.com/ClangBuiltLinux/continuous-integration2/actions/runs/3095779516/jobs/5012260390
warnings from /usr/bin/ld. That's...unexpected.
I've filed https://github.com/ClangBuiltLinux/linux/issues/1715 to
follow up on this.
--
Thanks,
~Nick Desaulniers

David Gow

unread,
Sep 22, 2022, 12:43:10 AM9/22/22
to Nick Desaulniers, Richard Weinberger, Anton Ivanov, Johannes Berg, Thomas Gleixner, Ingo Molnar, Dave Hansen, Linus Torvalds, Brendan Higgins, Daniel Latypov, linux-um, Linux Kernel Mailing List, KUnit Development, x...@kernel.org, Lukas Straub, Randy Dunlap, clang-built-linux
Thanks for following up on this.

Yeah: what I'm seeing seems to match it does look like ld is being
used instead of lld.

I tried -fuse-ld=lld via the following patch, though, and got a whole
bunch of "error: relocation R_X86_64_64 cannot be used against local
symbol; recompile with -fPIC" errors, so I think there's probably more
issues with UML and LLD that need working out. (I had no luck with
CONFIG_STATIC_LINK=y, nor with an attempt to use $(LD) directly,
instead of via $(CC)).

---
diff --git a/arch/um/Makefile b/arch/um/Makefile
index f1d4d67157be..01d9eae736be 100644
--- a/arch/um/Makefile
+++ b/arch/um/Makefile
@@ -139,6 +139,9 @@ ifeq ($(CONFIG_LD_IS_BFD),y)
LDFLAGS_EXECSTACK += $(call ld-option,--no-warn-rwx-segments)
endif

+# Since we're using CC as the driver, we need to force LLD if it is requested.
+LINK-$(CONFIG_LD_IS_LLD) += $(call cc-option,--ld-path=$(LD),-fuse-ld=lld)
+
LD_FLAGS_CMDLINE = $(foreach opt,$(KBUILD_LDFLAGS),-Wl,$(opt))

# Used by link-vmlinux.sh which has special support for um link

---

I'll update the GitHub issue with these notes as well, but I
definitely think it'd be best not to make UML use lld here until it
actually builds something: even with the warnings, ld is at least
latting us run at all.

Cheers,
-- David
> --
> You received this message because you are subscribed to the Google Groups "KUnit Development" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to kunit-dev+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/kunit-dev/CAKwvOdnLaDn%3DEfGFhjhcZe4EzvjU7Cws3z3KjLGwW%3DxnqnmHyw%40mail.gmail.com.
Reply all
Reply to author
Forward
0 new messages