LKML Archive on lore.kernel.org
 help / color / Atom feed
* [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
@ 2021-01-12 20:55 Nathan Chancellor
  2021-01-12 21:15 ` Nick Desaulniers
                   ` (3 more replies)
  0 siblings, 4 replies; 14+ messages in thread
From: Nathan Chancellor @ 2021-01-12 20:55 UTC (permalink / raw)
  To: Kees Cook, Andrew Morton
  Cc: Nick Desaulniers, linux-kernel, clang-built-linux, Nathan Chancellor

When building ARCH=mips 32r2el_defconfig with CONFIG_UBSAN_ALIGNMENT:

ld.lld: error: undefined symbol: __ubsan_handle_alignment_assumption
>>> referenced by slab.h:557 (include/linux/slab.h:557)
>>>               main.o:(do_initcalls) in archive init/built-in.a
>>> referenced by slab.h:448 (include/linux/slab.h:448)
>>>               do_mounts_rd.o:(rd_load_image) in archive init/built-in.a
>>> referenced by slab.h:448 (include/linux/slab.h:448)
>>>               do_mounts_rd.o:(identify_ramdisk_image) in archive init/built-in.a
>>> referenced 1579 more times

Implement this for the kernel based on LLVM's
handleAlignmentAssumptionImpl because the kernel is not linked against
the compiler runtime.

Link: https://github.com/ClangBuiltLinux/linux/issues/1245
Link: https://github.com/llvm/llvm-project/blob/llvmorg-11.0.1/compiler-rt/lib/ubsan/ubsan_handlers.cpp#L151-L190
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
---
 lib/ubsan.c | 28 ++++++++++++++++++++++++++++
 lib/ubsan.h |  6 ++++++
 2 files changed, 34 insertions(+)

diff --git a/lib/ubsan.c b/lib/ubsan.c
index 3e3352f3d0da..a1e6cc9993f8 100644
--- a/lib/ubsan.c
+++ b/lib/ubsan.c
@@ -427,3 +427,31 @@ void __ubsan_handle_load_invalid_value(void *_data, void *val)
 	ubsan_epilogue();
 }
 EXPORT_SYMBOL(__ubsan_handle_load_invalid_value);
+
+void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
+					 unsigned long align,
+					 unsigned long offset)
+{
+	struct alignment_assumption_data *data = _data;
+	unsigned long real_ptr;
+
+	if (suppress_report(&data->location))
+		return;
+
+	ubsan_prologue(&data->location, "alignment-assumption");
+
+	if (offset)
+		pr_err("assumption of %lu byte alignment (with offset of %lu byte) for pointer of type %s failed",
+		       align, offset, data->type->type_name);
+	else
+		pr_err("assumption of %lu byte alignment for pointer of type %s failed",
+		       align, data->type->type_name);
+
+	real_ptr = ptr - offset;
+	pr_err("%saddress is %lu aligned, misalignment offset is %lu bytes",
+	       offset ? "offset " : "", BIT(ffs(real_ptr)),
+	       real_ptr & (align - 1));
+
+	ubsan_epilogue();
+}
+EXPORT_SYMBOL(__ubsan_handle_alignment_assumption);
diff --git a/lib/ubsan.h b/lib/ubsan.h
index 7b56c09473a9..9a0b71c5ff9f 100644
--- a/lib/ubsan.h
+++ b/lib/ubsan.h
@@ -78,6 +78,12 @@ struct invalid_value_data {
 	struct type_descriptor *type;
 };
 
+struct alignment_assumption_data {
+	struct source_location location;
+	struct source_location assumption_location;
+	struct type_descriptor *type;
+};
+
 #if defined(CONFIG_ARCH_SUPPORTS_INT128)
 typedef __int128 s_max;
 typedef unsigned __int128 u_max;

base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837
-- 
2.30.0


^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-12 20:55 [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption Nathan Chancellor
@ 2021-01-12 21:15 ` Nick Desaulniers
  2021-01-12 21:37   ` Nathan Chancellor
  2021-01-13  0:12 ` [PATCH v2] " Nathan Chancellor
                   ` (2 subsequent siblings)
  3 siblings, 1 reply; 14+ messages in thread
From: Nick Desaulniers @ 2021-01-12 21:15 UTC (permalink / raw)
  To: Nathan Chancellor; +Cc: Kees Cook, Andrew Morton, LKML, clang-built-linux

On Tue, Jan 12, 2021 at 12:55 PM Nathan Chancellor
<natechancellor@gmail.com> wrote:
>
> When building ARCH=mips 32r2el_defconfig with CONFIG_UBSAN_ALIGNMENT:
>
> ld.lld: error: undefined symbol: __ubsan_handle_alignment_assumption
> >>> referenced by slab.h:557 (include/linux/slab.h:557)
> >>>               main.o:(do_initcalls) in archive init/built-in.a
> >>> referenced by slab.h:448 (include/linux/slab.h:448)
> >>>               do_mounts_rd.o:(rd_load_image) in archive init/built-in.a
> >>> referenced by slab.h:448 (include/linux/slab.h:448)
> >>>               do_mounts_rd.o:(identify_ramdisk_image) in archive init/built-in.a
> >>> referenced 1579 more times
>
> Implement this for the kernel based on LLVM's
> handleAlignmentAssumptionImpl because the kernel is not linked against
> the compiler runtime.
>
> Link: https://github.com/ClangBuiltLinux/linux/issues/1245
> Link: https://github.com/llvm/llvm-project/blob/llvmorg-11.0.1/compiler-rt/lib/ubsan/ubsan_handlers.cpp#L151-L190
> Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
> ---
>  lib/ubsan.c | 28 ++++++++++++++++++++++++++++
>  lib/ubsan.h |  6 ++++++
>  2 files changed, 34 insertions(+)
>
> diff --git a/lib/ubsan.c b/lib/ubsan.c
> index 3e3352f3d0da..a1e6cc9993f8 100644
> --- a/lib/ubsan.c
> +++ b/lib/ubsan.c
> @@ -427,3 +427,31 @@ void __ubsan_handle_load_invalid_value(void *_data, void *val)
>         ubsan_epilogue();
>  }
>  EXPORT_SYMBOL(__ubsan_handle_load_invalid_value);
> +
> +void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
> +                                        unsigned long align,
> +                                        unsigned long offset)
> +{
> +       struct alignment_assumption_data *data = _data;
> +       unsigned long real_ptr;
> +
> +       if (suppress_report(&data->location))
> +               return;
> +
> +       ubsan_prologue(&data->location, "alignment-assumption");
> +
> +       if (offset)
> +               pr_err("assumption of %lu byte alignment (with offset of %lu byte) for pointer of type %s failed",
> +                      align, offset, data->type->type_name);
> +       else
> +               pr_err("assumption of %lu byte alignment for pointer of type %s failed",
> +                      align, data->type->type_name);
> +
> +       real_ptr = ptr - offset;
> +       pr_err("%saddress is %lu aligned, misalignment offset is %lu bytes",
> +              offset ? "offset " : "", BIT(ffs(real_ptr)),

if real_ptr is an unsigned long, do we want to use `__ffs(real_ptr) +
1` here rather than ffs which takes an int?  It seems the kernel is
missing a definition of ffsl. :(

Otherwise code LGTM.

> +              real_ptr & (align - 1));
> +
> +       ubsan_epilogue();
> +}
> +EXPORT_SYMBOL(__ubsan_handle_alignment_assumption);
> diff --git a/lib/ubsan.h b/lib/ubsan.h
> index 7b56c09473a9..9a0b71c5ff9f 100644
> --- a/lib/ubsan.h
> +++ b/lib/ubsan.h
> @@ -78,6 +78,12 @@ struct invalid_value_data {
>         struct type_descriptor *type;
>  };
>
> +struct alignment_assumption_data {
> +       struct source_location location;
> +       struct source_location assumption_location;
> +       struct type_descriptor *type;
> +};
> +
>  #if defined(CONFIG_ARCH_SUPPORTS_INT128)
>  typedef __int128 s_max;
>  typedef unsigned __int128 u_max;
>
> base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837
> --
> 2.30.0
>


-- 
Thanks,
~Nick Desaulniers

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-12 21:15 ` Nick Desaulniers
@ 2021-01-12 21:37   ` Nathan Chancellor
  2021-01-12 21:53     ` Nick Desaulniers
  0 siblings, 1 reply; 14+ messages in thread
From: Nathan Chancellor @ 2021-01-12 21:37 UTC (permalink / raw)
  To: Nick Desaulniers; +Cc: Kees Cook, Andrew Morton, LKML, clang-built-linux

On Tue, Jan 12, 2021 at 01:15:42PM -0800, Nick Desaulniers wrote:
> On Tue, Jan 12, 2021 at 12:55 PM Nathan Chancellor
> <natechancellor@gmail.com> wrote:
> >
> > When building ARCH=mips 32r2el_defconfig with CONFIG_UBSAN_ALIGNMENT:
> >
> > ld.lld: error: undefined symbol: __ubsan_handle_alignment_assumption
> > >>> referenced by slab.h:557 (include/linux/slab.h:557)
> > >>>               main.o:(do_initcalls) in archive init/built-in.a
> > >>> referenced by slab.h:448 (include/linux/slab.h:448)
> > >>>               do_mounts_rd.o:(rd_load_image) in archive init/built-in.a
> > >>> referenced by slab.h:448 (include/linux/slab.h:448)
> > >>>               do_mounts_rd.o:(identify_ramdisk_image) in archive init/built-in.a
> > >>> referenced 1579 more times
> >
> > Implement this for the kernel based on LLVM's
> > handleAlignmentAssumptionImpl because the kernel is not linked against
> > the compiler runtime.
> >
> > Link: https://github.com/ClangBuiltLinux/linux/issues/1245
> > Link: https://github.com/llvm/llvm-project/blob/llvmorg-11.0.1/compiler-rt/lib/ubsan/ubsan_handlers.cpp#L151-L190
> > Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
> > ---
> >  lib/ubsan.c | 28 ++++++++++++++++++++++++++++
> >  lib/ubsan.h |  6 ++++++
> >  2 files changed, 34 insertions(+)
> >
> > diff --git a/lib/ubsan.c b/lib/ubsan.c
> > index 3e3352f3d0da..a1e6cc9993f8 100644
> > --- a/lib/ubsan.c
> > +++ b/lib/ubsan.c
> > @@ -427,3 +427,31 @@ void __ubsan_handle_load_invalid_value(void *_data, void *val)
> >         ubsan_epilogue();
> >  }
> >  EXPORT_SYMBOL(__ubsan_handle_load_invalid_value);
> > +
> > +void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
> > +                                        unsigned long align,
> > +                                        unsigned long offset)
> > +{
> > +       struct alignment_assumption_data *data = _data;
> > +       unsigned long real_ptr;
> > +
> > +       if (suppress_report(&data->location))
> > +               return;
> > +
> > +       ubsan_prologue(&data->location, "alignment-assumption");
> > +
> > +       if (offset)
> > +               pr_err("assumption of %lu byte alignment (with offset of %lu byte) for pointer of type %s failed",
> > +                      align, offset, data->type->type_name);
> > +       else
> > +               pr_err("assumption of %lu byte alignment for pointer of type %s failed",
> > +                      align, data->type->type_name);
> > +
> > +       real_ptr = ptr - offset;
> > +       pr_err("%saddress is %lu aligned, misalignment offset is %lu bytes",
> > +              offset ? "offset " : "", BIT(ffs(real_ptr)),
> 
> if real_ptr is an unsigned long, do we want to use `__ffs(real_ptr) +
> 1` here rather than ffs which takes an int?  It seems the kernel is
> missing a definition of ffsl. :(

Why the + 1? I think if we use __ffs (which it seems like we should), I
think that needs to become

BIT(real_ptr ? __ffs(real_ptr) : 0)

I have made that change locally and will send it for v2 in a day or so
to give Kees some time to check it out.

Thanks for the review!
Nathan

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-12 21:37   ` Nathan Chancellor
@ 2021-01-12 21:53     ` Nick Desaulniers
  2021-01-12 22:06       ` Nathan Chancellor
  0 siblings, 1 reply; 14+ messages in thread
From: Nick Desaulniers @ 2021-01-12 21:53 UTC (permalink / raw)
  To: Nathan Chancellor; +Cc: Kees Cook, Andrew Morton, LKML, clang-built-linux

On Tue, Jan 12, 2021 at 1:37 PM Nathan Chancellor
<natechancellor@gmail.com> wrote:
>
> > if real_ptr is an unsigned long, do we want to use `__ffs(real_ptr) +
> > 1` here rather than ffs which takes an int?  It seems the kernel is
> > missing a definition of ffsl. :(
>
> Why the + 1? I think if we use __ffs (which it seems like we should), I
> think that needs to become

This came up recently in an internal code review; ffs and __ffs differ
in output by one.  See also the definition of ffs for alpha in
arch/alpha/include/asm/bitops.h.

Also, I just confirmed that:
```
#include <stdio.h>

// include/asm-generic/bitops/ffs.h
static inline int ffs(int x)
{
        int r = 1;

        if (!x)
                return 0;
        if (!(x & 0xffff)) {
                x >>= 16;
                r += 16;
        }
        if (!(x & 0xff)) {
                x >>= 8;
                r += 8;
        }
        if (!(x & 0xf)) {
                x >>= 4;
                r += 4;
        }
        if (!(x & 3)) {
                x >>= 2;
                r += 2;
        }
        if (!(x & 1)) {
                x >>= 1;
                r += 1;
        }
        return r;
}

// include/asm-generic/bitops/__ffs.h
static __always_inline unsigned long __ffs(unsigned long word)
{
        int num = 0;

        if ((word & 0xffffffff) == 0) {
                num += 32;
                word >>= 32;
        }
        if ((word & 0xffff) == 0) {
                num += 16;
                word >>= 16;
        }
        if ((word & 0xff) == 0) {
                num += 8;
                word >>= 8;
        }
        if ((word & 0xf) == 0) {
                num += 4;
                word >>= 4;
        }
        if ((word & 0x3) == 0) {
                num += 2;
                word >>= 2;
        }
        if ((word & 0x1) == 0)
                num += 1;
        return num;
}

int main() {
    int x = 3;
    unsigned long y = 3;
    printf("%d\n%lu\n", ffs(x), __ffs(y));
    return 0;
}
```
will print:
1
0
-- 
Thanks,
~Nick Desaulniers

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-12 21:53     ` Nick Desaulniers
@ 2021-01-12 22:06       ` Nathan Chancellor
  2021-01-12 23:56         ` Kees Cook
  0 siblings, 1 reply; 14+ messages in thread
From: Nathan Chancellor @ 2021-01-12 22:06 UTC (permalink / raw)
  To: Nick Desaulniers; +Cc: Kees Cook, Andrew Morton, LKML, clang-built-linux

On Tue, Jan 12, 2021 at 01:53:30PM -0800, Nick Desaulniers wrote:
> On Tue, Jan 12, 2021 at 1:37 PM Nathan Chancellor
> <natechancellor@gmail.com> wrote:
> >
> > > if real_ptr is an unsigned long, do we want to use `__ffs(real_ptr) +
> > > 1` here rather than ffs which takes an int?  It seems the kernel is
> > > missing a definition of ffsl. :(
> >
> > Why the + 1? I think if we use __ffs (which it seems like we should), I
> > think that needs to become
> 
> This came up recently in an internal code review; ffs and __ffs differ
> in output by one.  See also the definition of ffs for alpha in
> arch/alpha/include/asm/bitops.h.

Interesting, thanks for bringing it up! Looks like ffs returns 1-32 and
__ffs returns 0-31. I think that we want __ffs here because we are
shifting (1UL << 32 overflows on 32-bit architectures) and the code in
LLVM appears to agree. LeastSignificantSetBitIndex evaluates to
__builtin_ctzl, which is the asm-generic implementation of __ffs.

Cheers,
NAthan

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-12 22:06       ` Nathan Chancellor
@ 2021-01-12 23:56         ` Kees Cook
  0 siblings, 0 replies; 14+ messages in thread
From: Kees Cook @ 2021-01-12 23:56 UTC (permalink / raw)
  To: Nathan Chancellor
  Cc: Nick Desaulniers, Andrew Morton, LKML, clang-built-linux

On Tue, Jan 12, 2021 at 03:06:34PM -0700, Nathan Chancellor wrote:
> On Tue, Jan 12, 2021 at 01:53:30PM -0800, Nick Desaulniers wrote:
> > On Tue, Jan 12, 2021 at 1:37 PM Nathan Chancellor
> > <natechancellor@gmail.com> wrote:
> > >
> > > > if real_ptr is an unsigned long, do we want to use `__ffs(real_ptr) +
> > > > 1` here rather than ffs which takes an int?  It seems the kernel is
> > > > missing a definition of ffsl. :(
> > >
> > > Why the + 1? I think if we use __ffs (which it seems like we should), I
> > > think that needs to become
> > 
> > This came up recently in an internal code review; ffs and __ffs differ
> > in output by one.  See also the definition of ffs for alpha in
> > arch/alpha/include/asm/bitops.h.
> 
> Interesting, thanks for bringing it up! Looks like ffs returns 1-32 and
> __ffs returns 0-31. I think that we want __ffs here because we are
> shifting (1UL << 32 overflows on 32-bit architectures) and the code in
> LLVM appears to agree. LeastSignificantSetBitIndex evaluates to
> __builtin_ctzl, which is the asm-generic implementation of __ffs.

Sounds good. With __ffs, consider your v2:

Acked-by: Kees Cook <keescook@chromium.org>

-- 
Kees Cook

^ permalink raw reply	[flat|nested] 14+ messages in thread

* [PATCH v2] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-12 20:55 [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption Nathan Chancellor
  2021-01-12 21:15 ` Nick Desaulniers
@ 2021-01-13  0:12 ` Nathan Chancellor
  2021-01-27 22:44   ` [PATCH v3] " Nathan Chancellor
  2021-01-13  0:18 ` [PATCH] " kernel test robot
  2021-01-13  0:39 ` kernel test robot
  3 siblings, 1 reply; 14+ messages in thread
From: Nathan Chancellor @ 2021-01-13  0:12 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Kees Cook, Nick Desaulniers, linux-kernel, clang-built-linux,
	Nathan Chancellor

When building ARCH=mips 32r2el_defconfig with CONFIG_UBSAN_ALIGNMENT:

ld.lld: error: undefined symbol: __ubsan_handle_alignment_assumption
>>> referenced by slab.h:557 (include/linux/slab.h:557)
>>>               main.o:(do_initcalls) in archive init/built-in.a
>>> referenced by slab.h:448 (include/linux/slab.h:448)
>>>               do_mounts_rd.o:(rd_load_image) in archive init/built-in.a
>>> referenced by slab.h:448 (include/linux/slab.h:448)
>>>               do_mounts_rd.o:(identify_ramdisk_image) in archive init/built-in.a
>>> referenced 1579 more times

Implement this for the kernel based on LLVM's
handleAlignmentAssumptionImpl because the kernel is not linked against
the compiler runtime.

Link: https://github.com/ClangBuiltLinux/linux/issues/1245
Link: https://github.com/llvm/llvm-project/blob/llvmorg-11.0.1/compiler-rt/lib/ubsan/ubsan_handlers.cpp#L151-L190
Acked-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
---

v1 -> v2:

* Use __ffs instead of ffs because due to size of input (unsigned long
  vs int) and we want a zero based index (Nick Desaulniers).

* Pick up Kees's ack.

 lib/ubsan.c | 28 ++++++++++++++++++++++++++++
 lib/ubsan.h |  6 ++++++
 2 files changed, 34 insertions(+)

diff --git a/lib/ubsan.c b/lib/ubsan.c
index 3e3352f3d0da..1432a8645224 100644
--- a/lib/ubsan.c
+++ b/lib/ubsan.c
@@ -427,3 +427,31 @@ void __ubsan_handle_load_invalid_value(void *_data, void *val)
 	ubsan_epilogue();
 }
 EXPORT_SYMBOL(__ubsan_handle_load_invalid_value);
+
+void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
+					 unsigned long align,
+					 unsigned long offset)
+{
+	struct alignment_assumption_data *data = _data;
+	unsigned long real_ptr;
+
+	if (suppress_report(&data->location))
+		return;
+
+	ubsan_prologue(&data->location, "alignment-assumption");
+
+	if (offset)
+		pr_err("assumption of %lu byte alignment (with offset of %lu byte) for pointer of type %s failed",
+		       align, offset, data->type->type_name);
+	else
+		pr_err("assumption of %lu byte alignment for pointer of type %s failed",
+		       align, data->type->type_name);
+
+	real_ptr = ptr - offset;
+	pr_err("%saddress is %lu aligned, misalignment offset is %lu bytes",
+	       offset ? "offset " : "", BIT(real_ptr ? __ffs(real_ptr) : 0),
+	       real_ptr & (align - 1));
+
+	ubsan_epilogue();
+}
+EXPORT_SYMBOL(__ubsan_handle_alignment_assumption);
diff --git a/lib/ubsan.h b/lib/ubsan.h
index 7b56c09473a9..9a0b71c5ff9f 100644
--- a/lib/ubsan.h
+++ b/lib/ubsan.h
@@ -78,6 +78,12 @@ struct invalid_value_data {
 	struct type_descriptor *type;
 };
 
+struct alignment_assumption_data {
+	struct source_location location;
+	struct source_location assumption_location;
+	struct type_descriptor *type;
+};
+
 #if defined(CONFIG_ARCH_SUPPORTS_INT128)
 typedef __int128 s_max;
 typedef unsigned __int128 u_max;

base-commit: 7c53f6b671f4aba70ff15e1b05148b10d58c2837
-- 
2.30.0


^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-12 20:55 [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption Nathan Chancellor
  2021-01-12 21:15 ` Nick Desaulniers
  2021-01-13  0:12 ` [PATCH v2] " Nathan Chancellor
@ 2021-01-13  0:18 ` kernel test robot
  2021-01-13  0:39 ` kernel test robot
  3 siblings, 0 replies; 14+ messages in thread
From: kernel test robot @ 2021-01-13  0:18 UTC (permalink / raw)
  To: Nathan Chancellor, Kees Cook, Andrew Morton
  Cc: kbuild-all, Linux Memory Management List, Nick Desaulniers,
	linux-kernel, clang-built-linux, Nathan Chancellor


[-- Attachment #1: Type: text/plain, Size: 6745 bytes --]

Hi Nathan,

I love your patch! Perhaps something to improve:

[auto build test WARNING on 7c53f6b671f4aba70ff15e1b05148b10d58c2837]

url:    https://github.com/0day-ci/linux/commits/Nathan-Chancellor/ubsan-Implement-__ubsan_handle_alignment_assumption/20210113-055714
base:    7c53f6b671f4aba70ff15e1b05148b10d58c2837
config: m68k-allmodconfig (attached as .config)
compiler: m68k-linux-gcc (GCC) 9.3.0
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # https://github.com/0day-ci/linux/commit/775adad26a60878926c0ee6cd460a1375bbe51e6
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Nathan-Chancellor/ubsan-Implement-__ubsan_handle_alignment_assumption/20210113-055714
        git checkout 775adad26a60878926c0ee6cd460a1375bbe51e6
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=m68k 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All warnings (new ones prefixed by >>):

   lib/ubsan.c:192:6: warning: no previous prototype for '__ubsan_handle_add_overflow' [-Wmissing-prototypes]
     192 | void __ubsan_handle_add_overflow(void *data,
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~
   lib/ubsan.c:200:6: warning: no previous prototype for '__ubsan_handle_sub_overflow' [-Wmissing-prototypes]
     200 | void __ubsan_handle_sub_overflow(void *data,
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~
   lib/ubsan.c:207:6: warning: no previous prototype for '__ubsan_handle_mul_overflow' [-Wmissing-prototypes]
     207 | void __ubsan_handle_mul_overflow(void *data,
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~
   lib/ubsan.c:214:6: warning: no previous prototype for '__ubsan_handle_negate_overflow' [-Wmissing-prototypes]
     214 | void __ubsan_handle_negate_overflow(void *_data, void *old_val)
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   lib/ubsan.c:234:6: warning: no previous prototype for '__ubsan_handle_divrem_overflow' [-Wmissing-prototypes]
     234 | void __ubsan_handle_divrem_overflow(void *_data, void *lhs, void *rhs)
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   lib/ubsan.c:315:6: warning: no previous prototype for '__ubsan_handle_type_mismatch' [-Wmissing-prototypes]
     315 | void __ubsan_handle_type_mismatch(struct type_mismatch_data *data,
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
   lib/ubsan.c:329:6: warning: no previous prototype for '__ubsan_handle_type_mismatch_v1' [-Wmissing-prototypes]
     329 | void __ubsan_handle_type_mismatch_v1(void *_data, void *ptr)
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   lib/ubsan.c:343:6: warning: no previous prototype for '__ubsan_handle_out_of_bounds' [-Wmissing-prototypes]
     343 | void __ubsan_handle_out_of_bounds(void *_data, void *index)
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
   lib/ubsan.c:360:6: warning: no previous prototype for '__ubsan_handle_shift_out_of_bounds' [-Wmissing-prototypes]
     360 | void __ubsan_handle_shift_out_of_bounds(void *_data, void *lhs, void *rhs)
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   lib/ubsan.c:402:6: warning: no previous prototype for '__ubsan_handle_builtin_unreachable' [-Wmissing-prototypes]
     402 | void __ubsan_handle_builtin_unreachable(void *_data)
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   lib/ubsan.c:412:6: warning: no previous prototype for '__ubsan_handle_load_invalid_value' [-Wmissing-prototypes]
     412 | void __ubsan_handle_load_invalid_value(void *_data, void *val)
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> lib/ubsan.c:431:6: warning: no previous prototype for '__ubsan_handle_alignment_assumption' [-Wmissing-prototypes]
     431 | void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
         |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


vim +/__ubsan_handle_alignment_assumption +431 lib/ubsan.c

   359	
 > 360	void __ubsan_handle_shift_out_of_bounds(void *_data, void *lhs, void *rhs)
   361	{
   362		struct shift_out_of_bounds_data *data = _data;
   363		struct type_descriptor *rhs_type = data->rhs_type;
   364		struct type_descriptor *lhs_type = data->lhs_type;
   365		char rhs_str[VALUE_LENGTH];
   366		char lhs_str[VALUE_LENGTH];
   367		unsigned long ua_flags = user_access_save();
   368	
   369		if (suppress_report(&data->location))
   370			goto out;
   371	
   372		ubsan_prologue(&data->location, "shift-out-of-bounds");
   373	
   374		val_to_string(rhs_str, sizeof(rhs_str), rhs_type, rhs);
   375		val_to_string(lhs_str, sizeof(lhs_str), lhs_type, lhs);
   376	
   377		if (val_is_negative(rhs_type, rhs))
   378			pr_err("shift exponent %s is negative\n", rhs_str);
   379	
   380		else if (get_unsigned_val(rhs_type, rhs) >=
   381			type_bit_width(lhs_type))
   382			pr_err("shift exponent %s is too large for %u-bit type %s\n",
   383				rhs_str,
   384				type_bit_width(lhs_type),
   385				lhs_type->type_name);
   386		else if (val_is_negative(lhs_type, lhs))
   387			pr_err("left shift of negative value %s\n",
   388				lhs_str);
   389		else
   390			pr_err("left shift of %s by %s places cannot be"
   391				" represented in type %s\n",
   392				lhs_str, rhs_str,
   393				lhs_type->type_name);
   394	
   395		ubsan_epilogue();
   396	out:
   397		user_access_restore(ua_flags);
   398	}
   399	EXPORT_SYMBOL(__ubsan_handle_shift_out_of_bounds);
   400	
   401	
   402	void __ubsan_handle_builtin_unreachable(void *_data)
   403	{
   404		struct unreachable_data *data = _data;
   405		ubsan_prologue(&data->location, "unreachable");
   406		pr_err("calling __builtin_unreachable()\n");
   407		ubsan_epilogue();
   408		panic("can't return from __builtin_unreachable()");
   409	}
   410	EXPORT_SYMBOL(__ubsan_handle_builtin_unreachable);
   411	
   412	void __ubsan_handle_load_invalid_value(void *_data, void *val)
   413	{
   414		struct invalid_value_data *data = _data;
   415		char val_str[VALUE_LENGTH];
   416	
   417		if (suppress_report(&data->location))
   418			return;
   419	
   420		ubsan_prologue(&data->location, "invalid-load");
   421	
   422		val_to_string(val_str, sizeof(val_str), data->type, val);
   423	
   424		pr_err("load of value %s is not a valid value for type %s\n",
   425			val_str, data->type->type_name);
   426	
   427		ubsan_epilogue();
   428	}
   429	EXPORT_SYMBOL(__ubsan_handle_load_invalid_value);
   430	
 > 431	void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 59573 bytes --]

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-12 20:55 [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption Nathan Chancellor
                   ` (2 preceding siblings ...)
  2021-01-13  0:18 ` [PATCH] " kernel test robot
@ 2021-01-13  0:39 ` kernel test robot
  2021-01-13  1:31   ` Nathan Chancellor
  3 siblings, 1 reply; 14+ messages in thread
From: kernel test robot @ 2021-01-13  0:39 UTC (permalink / raw)
  To: Nathan Chancellor, Kees Cook, Andrew Morton
  Cc: kbuild-all, clang-built-linux, Linux Memory Management List,
	Nick Desaulniers, linux-kernel, Nathan Chancellor


[-- Attachment #1: Type: text/plain, Size: 7053 bytes --]

Hi Nathan,

I love your patch! Perhaps something to improve:

[auto build test WARNING on 7c53f6b671f4aba70ff15e1b05148b10d58c2837]

url:    https://github.com/0day-ci/linux/commits/Nathan-Chancellor/ubsan-Implement-__ubsan_handle_alignment_assumption/20210113-055714
base:    7c53f6b671f4aba70ff15e1b05148b10d58c2837
config: arm64-randconfig-r031-20210112 (attached as .config)
compiler: clang version 12.0.0 (https://github.com/llvm/llvm-project 32bcfcda4e28375e5a85268d2acfabcfcc011abf)
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # install arm64 cross compiling tool for clang build
        # apt-get install binutils-aarch64-linux-gnu
        # https://github.com/0day-ci/linux/commit/775adad26a60878926c0ee6cd460a1375bbe51e6
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Nathan-Chancellor/ubsan-Implement-__ubsan_handle_alignment_assumption/20210113-055714
        git checkout 775adad26a60878926c0ee6cd460a1375bbe51e6
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=arm64 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All warnings (new ones prefixed by >>):

   lib/ubsan.c:192:6: warning: no previous prototype for function '__ubsan_handle_add_overflow' [-Wmissing-prototypes]
   void __ubsan_handle_add_overflow(void *data,
        ^
   lib/ubsan.c:192:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_add_overflow(void *data,
   ^
   static 
   lib/ubsan.c:200:6: warning: no previous prototype for function '__ubsan_handle_sub_overflow' [-Wmissing-prototypes]
   void __ubsan_handle_sub_overflow(void *data,
        ^
   lib/ubsan.c:200:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_sub_overflow(void *data,
   ^
   static 
   lib/ubsan.c:207:6: warning: no previous prototype for function '__ubsan_handle_mul_overflow' [-Wmissing-prototypes]
   void __ubsan_handle_mul_overflow(void *data,
        ^
   lib/ubsan.c:207:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_mul_overflow(void *data,
   ^
   static 
   lib/ubsan.c:214:6: warning: no previous prototype for function '__ubsan_handle_negate_overflow' [-Wmissing-prototypes]
   void __ubsan_handle_negate_overflow(void *_data, void *old_val)
        ^
   lib/ubsan.c:214:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_negate_overflow(void *_data, void *old_val)
   ^
   static 
   lib/ubsan.c:234:6: warning: no previous prototype for function '__ubsan_handle_divrem_overflow' [-Wmissing-prototypes]
   void __ubsan_handle_divrem_overflow(void *_data, void *lhs, void *rhs)
        ^
   lib/ubsan.c:234:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_divrem_overflow(void *_data, void *lhs, void *rhs)
   ^
   static 
   lib/ubsan.c:315:6: warning: no previous prototype for function '__ubsan_handle_type_mismatch' [-Wmissing-prototypes]
   void __ubsan_handle_type_mismatch(struct type_mismatch_data *data,
        ^
   lib/ubsan.c:315:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_type_mismatch(struct type_mismatch_data *data,
   ^
   static 
   lib/ubsan.c:329:6: warning: no previous prototype for function '__ubsan_handle_type_mismatch_v1' [-Wmissing-prototypes]
   void __ubsan_handle_type_mismatch_v1(void *_data, void *ptr)
        ^
   lib/ubsan.c:329:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_type_mismatch_v1(void *_data, void *ptr)
   ^
   static 
   lib/ubsan.c:343:6: warning: no previous prototype for function '__ubsan_handle_out_of_bounds' [-Wmissing-prototypes]
   void __ubsan_handle_out_of_bounds(void *_data, void *index)
        ^
   lib/ubsan.c:343:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_out_of_bounds(void *_data, void *index)
   ^
   static 
   lib/ubsan.c:360:6: warning: no previous prototype for function '__ubsan_handle_shift_out_of_bounds' [-Wmissing-prototypes]
   void __ubsan_handle_shift_out_of_bounds(void *_data, void *lhs, void *rhs)
        ^
   lib/ubsan.c:360:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_shift_out_of_bounds(void *_data, void *lhs, void *rhs)
   ^
   static 
   lib/ubsan.c:402:6: warning: no previous prototype for function '__ubsan_handle_builtin_unreachable' [-Wmissing-prototypes]
   void __ubsan_handle_builtin_unreachable(void *_data)
        ^
   lib/ubsan.c:402:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_builtin_unreachable(void *_data)
   ^
   static 
   lib/ubsan.c:412:6: warning: no previous prototype for function '__ubsan_handle_load_invalid_value' [-Wmissing-prototypes]
   void __ubsan_handle_load_invalid_value(void *_data, void *val)
        ^
   lib/ubsan.c:412:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_load_invalid_value(void *_data, void *val)
   ^
   static 
>> lib/ubsan.c:431:6: warning: no previous prototype for function '__ubsan_handle_alignment_assumption' [-Wmissing-prototypes]
   void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
        ^
   lib/ubsan.c:431:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
   ^
   static 
   12 warnings generated.


vim +/__ubsan_handle_alignment_assumption +431 lib/ubsan.c

   411	
 > 412	void __ubsan_handle_load_invalid_value(void *_data, void *val)
   413	{
   414		struct invalid_value_data *data = _data;
   415		char val_str[VALUE_LENGTH];
   416	
   417		if (suppress_report(&data->location))
   418			return;
   419	
   420		ubsan_prologue(&data->location, "invalid-load");
   421	
   422		val_to_string(val_str, sizeof(val_str), data->type, val);
   423	
   424		pr_err("load of value %s is not a valid value for type %s\n",
   425			val_str, data->type->type_name);
   426	
   427		ubsan_epilogue();
   428	}
   429	EXPORT_SYMBOL(__ubsan_handle_load_invalid_value);
   430	
 > 431	void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 31723 bytes --]

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-13  0:39 ` kernel test robot
@ 2021-01-13  1:31   ` Nathan Chancellor
  2021-01-13  1:39     ` Nick Desaulniers
  0 siblings, 1 reply; 14+ messages in thread
From: Nathan Chancellor @ 2021-01-13  1:31 UTC (permalink / raw)
  To: kernel test robot
  Cc: Kees Cook, Andrew Morton, kbuild-all, clang-built-linux,
	Linux Memory Management List, Nick Desaulniers, linux-kernel

On Wed, Jan 13, 2021 at 08:39:52AM +0800, kernel test robot wrote:
> Hi Nathan,
> 
> I love your patch! Perhaps something to improve:
> 
> [auto build test WARNING on 7c53f6b671f4aba70ff15e1b05148b10d58c2837]
> 
> url:    https://github.com/0day-ci/linux/commits/Nathan-Chancellor/ubsan-Implement-__ubsan_handle_alignment_assumption/20210113-055714
> base:    7c53f6b671f4aba70ff15e1b05148b10d58c2837
> config: arm64-randconfig-r031-20210112 (attached as .config)
> compiler: clang version 12.0.0 (https://github.com/llvm/llvm-project 32bcfcda4e28375e5a85268d2acfabcfcc011abf)
> reproduce (this is a W=1 build):
>         wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
>         chmod +x ~/bin/make.cross
>         # install arm64 cross compiling tool for clang build
>         # apt-get install binutils-aarch64-linux-gnu
>         # https://github.com/0day-ci/linux/commit/775adad26a60878926c0ee6cd460a1375bbe51e6
>         git remote add linux-review https://github.com/0day-ci/linux
>         git fetch --no-tags linux-review Nathan-Chancellor/ubsan-Implement-__ubsan_handle_alignment_assumption/20210113-055714
>         git checkout 775adad26a60878926c0ee6cd460a1375bbe51e6
>         # save the attached .config to linux build tree
>         COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=arm64 
> 
> If you fix the issue, kindly add following tag as appropriate
> Reported-by: kernel test robot <lkp@intel.com>
> 
> All warnings (new ones prefixed by >>):
> 
>    lib/ubsan.c:192:6: warning: no previous prototype for function '__ubsan_handle_add_overflow' [-Wmissing-prototypes]
>    void __ubsan_handle_add_overflow(void *data,
>         ^
>    lib/ubsan.c:192:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
>    void __ubsan_handle_add_overflow(void *data,
>    ^
>    static 
>    lib/ubsan.c:200:6: warning: no previous prototype for function '__ubsan_handle_sub_overflow' [-Wmissing-prototypes]
>    void __ubsan_handle_sub_overflow(void *data,
>         ^
>    lib/ubsan.c:200:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
>    void __ubsan_handle_sub_overflow(void *data,
>    ^

Given that these are compiler inserted functions, there is not much of a
point to having prototypes to them. If people feel shutting these
warnings up is worthwhile, we can just add the prototypes right above
the function definition in a follow up patch.

Cheers,
Nathan

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-13  1:31   ` Nathan Chancellor
@ 2021-01-13  1:39     ` Nick Desaulniers
  2021-01-27 22:26       ` Nick Desaulniers
  0 siblings, 1 reply; 14+ messages in thread
From: Nick Desaulniers @ 2021-01-13  1:39 UTC (permalink / raw)
  To: Nathan Chancellor
  Cc: kernel test robot, Kees Cook, Andrew Morton, kbuild-all,
	clang-built-linux, Linux Memory Management List, LKML

On Tue, Jan 12, 2021 at 5:31 PM Nathan Chancellor
<natechancellor@gmail.com> wrote:
>
> On Wed, Jan 13, 2021 at 08:39:52AM +0800, kernel test robot wrote:
> > Hi Nathan,
> >
> > I love your patch! Perhaps something to improve:
> >
> > [auto build test WARNING on 7c53f6b671f4aba70ff15e1b05148b10d58c2837]
> >
> > url:    https://github.com/0day-ci/linux/commits/Nathan-Chancellor/ubsan-Implement-__ubsan_handle_alignment_assumption/20210113-055714
> > base:    7c53f6b671f4aba70ff15e1b05148b10d58c2837
> > config: arm64-randconfig-r031-20210112 (attached as .config)
> > compiler: clang version 12.0.0 (https://github.com/llvm/llvm-project 32bcfcda4e28375e5a85268d2acfabcfcc011abf)
> > reproduce (this is a W=1 build):
> >         wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
> >         chmod +x ~/bin/make.cross
> >         # install arm64 cross compiling tool for clang build
> >         # apt-get install binutils-aarch64-linux-gnu
> >         # https://github.com/0day-ci/linux/commit/775adad26a60878926c0ee6cd460a1375bbe51e6
> >         git remote add linux-review https://github.com/0day-ci/linux
> >         git fetch --no-tags linux-review Nathan-Chancellor/ubsan-Implement-__ubsan_handle_alignment_assumption/20210113-055714
> >         git checkout 775adad26a60878926c0ee6cd460a1375bbe51e6
> >         # save the attached .config to linux build tree
> >         COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=arm64
> >
> > If you fix the issue, kindly add following tag as appropriate
> > Reported-by: kernel test robot <lkp@intel.com>
> >
> > All warnings (new ones prefixed by >>):
> >
> >    lib/ubsan.c:192:6: warning: no previous prototype for function '__ubsan_handle_add_overflow' [-Wmissing-prototypes]
> >    void __ubsan_handle_add_overflow(void *data,
> >         ^
> >    lib/ubsan.c:192:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
> >    void __ubsan_handle_add_overflow(void *data,
> >    ^
> >    static
> >    lib/ubsan.c:200:6: warning: no previous prototype for function '__ubsan_handle_sub_overflow' [-Wmissing-prototypes]
> >    void __ubsan_handle_sub_overflow(void *data,
> >         ^
> >    lib/ubsan.c:200:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
> >    void __ubsan_handle_sub_overflow(void *data,
> >    ^
>
> Given that these are compiler inserted functions, there is not much of a
> point to having prototypes to them. If people feel shutting these
> warnings up is worthwhile, we can just add the prototypes right above
> the function definition in a follow up patch.

Same as stpcpy; it would be nice though. ;)
-- 
Thanks,
~Nick Desaulniers

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-13  1:39     ` Nick Desaulniers
@ 2021-01-27 22:26       ` Nick Desaulniers
  0 siblings, 0 replies; 14+ messages in thread
From: Nick Desaulniers @ 2021-01-27 22:26 UTC (permalink / raw)
  To: Nathan Chancellor
  Cc: kernel test robot, Kees Cook, Andrew Morton, kbuild-all,
	clang-built-linux, Linux Memory Management List, LKML

On Tue, Jan 12, 2021 at 5:39 PM Nick Desaulniers
<ndesaulniers@google.com> wrote:
>
> On Tue, Jan 12, 2021 at 5:31 PM Nathan Chancellor
> <natechancellor@gmail.com> wrote:
> >
> > On Wed, Jan 13, 2021 at 08:39:52AM +0800, kernel test robot wrote:
> > > Hi Nathan,
> > >
> > > I love your patch! Perhaps something to improve:
> > >
> > > [auto build test WARNING on 7c53f6b671f4aba70ff15e1b05148b10d58c2837]
> > >
> > > url:    https://github.com/0day-ci/linux/commits/Nathan-Chancellor/ubsan-Implement-__ubsan_handle_alignment_assumption/20210113-055714
> > > base:    7c53f6b671f4aba70ff15e1b05148b10d58c2837
> > > config: arm64-randconfig-r031-20210112 (attached as .config)
> > > compiler: clang version 12.0.0 (https://github.com/llvm/llvm-project 32bcfcda4e28375e5a85268d2acfabcfcc011abf)
> > > reproduce (this is a W=1 build):
> > >         wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
> > >         chmod +x ~/bin/make.cross
> > >         # install arm64 cross compiling tool for clang build
> > >         # apt-get install binutils-aarch64-linux-gnu
> > >         # https://github.com/0day-ci/linux/commit/775adad26a60878926c0ee6cd460a1375bbe51e6
> > >         git remote add linux-review https://github.com/0day-ci/linux
> > >         git fetch --no-tags linux-review Nathan-Chancellor/ubsan-Implement-__ubsan_handle_alignment_assumption/20210113-055714
> > >         git checkout 775adad26a60878926c0ee6cd460a1375bbe51e6
> > >         # save the attached .config to linux build tree
> > >         COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=arm64
> > >
> > > If you fix the issue, kindly add following tag as appropriate
> > > Reported-by: kernel test robot <lkp@intel.com>
> > >
> > > All warnings (new ones prefixed by >>):
> > >
> > >    lib/ubsan.c:192:6: warning: no previous prototype for function '__ubsan_handle_add_overflow' [-Wmissing-prototypes]
> > >    void __ubsan_handle_add_overflow(void *data,
> > >         ^
> > >    lib/ubsan.c:192:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
> > >    void __ubsan_handle_add_overflow(void *data,
> > >    ^
> > >    static
> > >    lib/ubsan.c:200:6: warning: no previous prototype for function '__ubsan_handle_sub_overflow' [-Wmissing-prototypes]
> > >    void __ubsan_handle_sub_overflow(void *data,
> > >         ^
> > >    lib/ubsan.c:200:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
> > >    void __ubsan_handle_sub_overflow(void *data,
> > >    ^
> >
> > Given that these are compiler inserted functions, there is not much of a
> > point to having prototypes to them. If people feel shutting these
> > warnings up is worthwhile, we can just add the prototypes right above
> > the function definition in a follow up patch.
>
> Same as stpcpy; it would be nice though. ;)

If you would be so kind, I'd be happy to sign off on such a v3.

> --
> Thanks,
> ~Nick Desaulniers



-- 
Thanks,
~Nick Desaulniers

^ permalink raw reply	[flat|nested] 14+ messages in thread

* [PATCH v3] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-13  0:12 ` [PATCH v2] " Nathan Chancellor
@ 2021-01-27 22:44   ` Nathan Chancellor
  2021-01-27 22:54     ` Nick Desaulniers
  0 siblings, 1 reply; 14+ messages in thread
From: Nathan Chancellor @ 2021-01-27 22:44 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Kees Cook, Nick Desaulniers, linux-kernel, clang-built-linux,
	Nathan Chancellor

When building ARCH=mips 32r2el_defconfig with CONFIG_UBSAN_ALIGNMENT:

ld.lld: error: undefined symbol: __ubsan_handle_alignment_assumption
>>> referenced by slab.h:557 (include/linux/slab.h:557)
>>>               main.o:(do_initcalls) in archive init/built-in.a
>>> referenced by slab.h:448 (include/linux/slab.h:448)
>>>               do_mounts_rd.o:(rd_load_image) in archive init/built-in.a
>>> referenced by slab.h:448 (include/linux/slab.h:448)
>>>               do_mounts_rd.o:(identify_ramdisk_image) in archive init/built-in.a
>>> referenced 1579 more times

Implement this for the kernel based on LLVM's
handleAlignmentAssumptionImpl because the kernel is not linked against
the compiler runtime.

Link: https://github.com/ClangBuiltLinux/linux/issues/1245
Link: https://github.com/llvm/llvm-project/blob/llvmorg-11.0.1/compiler-rt/lib/ubsan/ubsan_handlers.cpp#L151-L190
Acked-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
---

v2 -> v3:

* Add prototype right above definition to avoid introducing a warning
  with W=1.

v1 -> v2:

* Use __ffs instead of ffs because due to size of input (unsigned long
  vs int) and we want a zero based index (Nick Desaulniers).

* Pick up Kees's ack.

 lib/ubsan.c | 31 +++++++++++++++++++++++++++++++
 lib/ubsan.h |  6 ++++++
 2 files changed, 37 insertions(+)

diff --git a/lib/ubsan.c b/lib/ubsan.c
index 3e3352f3d0da..bec38c64d6a6 100644
--- a/lib/ubsan.c
+++ b/lib/ubsan.c
@@ -427,3 +427,34 @@ void __ubsan_handle_load_invalid_value(void *_data, void *val)
 	ubsan_epilogue();
 }
 EXPORT_SYMBOL(__ubsan_handle_load_invalid_value);
+
+void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
+					 unsigned long align,
+					 unsigned long offset);
+void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
+					 unsigned long align,
+					 unsigned long offset)
+{
+	struct alignment_assumption_data *data = _data;
+	unsigned long real_ptr;
+
+	if (suppress_report(&data->location))
+		return;
+
+	ubsan_prologue(&data->location, "alignment-assumption");
+
+	if (offset)
+		pr_err("assumption of %lu byte alignment (with offset of %lu byte) for pointer of type %s failed",
+		       align, offset, data->type->type_name);
+	else
+		pr_err("assumption of %lu byte alignment for pointer of type %s failed",
+		       align, data->type->type_name);
+
+	real_ptr = ptr - offset;
+	pr_err("%saddress is %lu aligned, misalignment offset is %lu bytes",
+	       offset ? "offset " : "", BIT(real_ptr ? __ffs(real_ptr) : 0),
+	       real_ptr & (align - 1));
+
+	ubsan_epilogue();
+}
+EXPORT_SYMBOL(__ubsan_handle_alignment_assumption);
diff --git a/lib/ubsan.h b/lib/ubsan.h
index 7b56c09473a9..9a0b71c5ff9f 100644
--- a/lib/ubsan.h
+++ b/lib/ubsan.h
@@ -78,6 +78,12 @@ struct invalid_value_data {
 	struct type_descriptor *type;
 };
 
+struct alignment_assumption_data {
+	struct source_location location;
+	struct source_location assumption_location;
+	struct type_descriptor *type;
+};
+
 #if defined(CONFIG_ARCH_SUPPORTS_INT128)
 typedef __int128 s_max;
 typedef unsigned __int128 u_max;

base-commit: 6ee1d745b7c9fd573fba142a2efdad76a9f1cb04
-- 
2.30.0


^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH v3] ubsan: Implement __ubsan_handle_alignment_assumption
  2021-01-27 22:44   ` [PATCH v3] " Nathan Chancellor
@ 2021-01-27 22:54     ` Nick Desaulniers
  0 siblings, 0 replies; 14+ messages in thread
From: Nick Desaulniers @ 2021-01-27 22:54 UTC (permalink / raw)
  To: Nathan Chancellor, Andrew Morton; +Cc: Kees Cook, LKML, clang-built-linux

On Wed, Jan 27, 2021 at 2:46 PM Nathan Chancellor <nathan@kernel.org> wrote:
>
> When building ARCH=mips 32r2el_defconfig with CONFIG_UBSAN_ALIGNMENT:
>
> ld.lld: error: undefined symbol: __ubsan_handle_alignment_assumption
> >>> referenced by slab.h:557 (include/linux/slab.h:557)
> >>>               main.o:(do_initcalls) in archive init/built-in.a
> >>> referenced by slab.h:448 (include/linux/slab.h:448)
> >>>               do_mounts_rd.o:(rd_load_image) in archive init/built-in.a
> >>> referenced by slab.h:448 (include/linux/slab.h:448)
> >>>               do_mounts_rd.o:(identify_ramdisk_image) in archive init/built-in.a
> >>> referenced 1579 more times
>
> Implement this for the kernel based on LLVM's
> handleAlignmentAssumptionImpl because the kernel is not linked against
> the compiler runtime.
>
> Link: https://github.com/ClangBuiltLinux/linux/issues/1245
> Link: https://github.com/llvm/llvm-project/blob/llvmorg-11.0.1/compiler-rt/lib/ubsan/ubsan_handlers.cpp#L151-L190
> Acked-by: Kees Cook <keescook@chromium.org>
> Signed-off-by: Nathan Chancellor <nathan@kernel.org>

Thanks for the patch!

Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>

> ---
>
> v2 -> v3:
>
> * Add prototype right above definition to avoid introducing a warning
>   with W=1.
>
> v1 -> v2:
>
> * Use __ffs instead of ffs because due to size of input (unsigned long
>   vs int) and we want a zero based index (Nick Desaulniers).
>
> * Pick up Kees's ack.
>
>  lib/ubsan.c | 31 +++++++++++++++++++++++++++++++
>  lib/ubsan.h |  6 ++++++
>  2 files changed, 37 insertions(+)
>
> diff --git a/lib/ubsan.c b/lib/ubsan.c
> index 3e3352f3d0da..bec38c64d6a6 100644
> --- a/lib/ubsan.c
> +++ b/lib/ubsan.c
> @@ -427,3 +427,34 @@ void __ubsan_handle_load_invalid_value(void *_data, void *val)
>         ubsan_epilogue();
>  }
>  EXPORT_SYMBOL(__ubsan_handle_load_invalid_value);
> +
> +void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
> +                                        unsigned long align,
> +                                        unsigned long offset);
> +void __ubsan_handle_alignment_assumption(void *_data, unsigned long ptr,
> +                                        unsigned long align,
> +                                        unsigned long offset)
> +{
> +       struct alignment_assumption_data *data = _data;
> +       unsigned long real_ptr;
> +
> +       if (suppress_report(&data->location))
> +               return;
> +
> +       ubsan_prologue(&data->location, "alignment-assumption");
> +
> +       if (offset)
> +               pr_err("assumption of %lu byte alignment (with offset of %lu byte) for pointer of type %s failed",
> +                      align, offset, data->type->type_name);
> +       else
> +               pr_err("assumption of %lu byte alignment for pointer of type %s failed",
> +                      align, data->type->type_name);
> +
> +       real_ptr = ptr - offset;
> +       pr_err("%saddress is %lu aligned, misalignment offset is %lu bytes",
> +              offset ? "offset " : "", BIT(real_ptr ? __ffs(real_ptr) : 0),
> +              real_ptr & (align - 1));
> +
> +       ubsan_epilogue();
> +}
> +EXPORT_SYMBOL(__ubsan_handle_alignment_assumption);
> diff --git a/lib/ubsan.h b/lib/ubsan.h
> index 7b56c09473a9..9a0b71c5ff9f 100644
> --- a/lib/ubsan.h
> +++ b/lib/ubsan.h
> @@ -78,6 +78,12 @@ struct invalid_value_data {
>         struct type_descriptor *type;
>  };
>
> +struct alignment_assumption_data {
> +       struct source_location location;
> +       struct source_location assumption_location;
> +       struct type_descriptor *type;
> +};
> +
>  #if defined(CONFIG_ARCH_SUPPORTS_INT128)
>  typedef __int128 s_max;
>  typedef unsigned __int128 u_max;
>
> base-commit: 6ee1d745b7c9fd573fba142a2efdad76a9f1cb04
> --
> 2.30.0
>


-- 
Thanks,
~Nick Desaulniers

^ permalink raw reply	[flat|nested] 14+ messages in thread

end of thread, back to index

Thread overview: 14+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-01-12 20:55 [PATCH] ubsan: Implement __ubsan_handle_alignment_assumption Nathan Chancellor
2021-01-12 21:15 ` Nick Desaulniers
2021-01-12 21:37   ` Nathan Chancellor
2021-01-12 21:53     ` Nick Desaulniers
2021-01-12 22:06       ` Nathan Chancellor
2021-01-12 23:56         ` Kees Cook
2021-01-13  0:12 ` [PATCH v2] " Nathan Chancellor
2021-01-27 22:44   ` [PATCH v3] " Nathan Chancellor
2021-01-27 22:54     ` Nick Desaulniers
2021-01-13  0:18 ` [PATCH] " kernel test robot
2021-01-13  0:39 ` kernel test robot
2021-01-13  1:31   ` Nathan Chancellor
2021-01-13  1:39     ` Nick Desaulniers
2021-01-27 22:26       ` Nick Desaulniers

LKML Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/lkml/0 lkml/git/0.git
	git clone --mirror https://lore.kernel.org/lkml/1 lkml/git/1.git
	git clone --mirror https://lore.kernel.org/lkml/2 lkml/git/2.git
	git clone --mirror https://lore.kernel.org/lkml/3 lkml/git/3.git
	git clone --mirror https://lore.kernel.org/lkml/4 lkml/git/4.git
	git clone --mirror https://lore.kernel.org/lkml/5 lkml/git/5.git
	git clone --mirror https://lore.kernel.org/lkml/6 lkml/git/6.git
	git clone --mirror https://lore.kernel.org/lkml/7 lkml/git/7.git
	git clone --mirror https://lore.kernel.org/lkml/8 lkml/git/8.git
	git clone --mirror https://lore.kernel.org/lkml/9 lkml/git/9.git
	git clone --mirror https://lore.kernel.org/lkml/10 lkml/git/10.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 lkml lkml/ https://lore.kernel.org/lkml \
		linux-kernel@vger.kernel.org
	public-inbox-index lkml

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.linux-kernel


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git