When building with sanitizers on Fedora we get a wierd error
message

In file included from /usr/include/string.h:519,
                 from ../src/internal.h:28,
                 from ../src/util/virsocket.h:21,
                 from ../src/util/virsocketaddr.h:21,
                 from ../src/util/virnetdevip.h:21,
                 from ../src/util/virnetdevip.c:21:
In function ‘memcpy’,
    inlined from ‘virNetDevGetifaddrsAddress’ at ../src/util/virnetdevip.c:702:13,
    inlined from ‘virNetDevIPAddrGet’ at ../src/util/virnetdevip.c:754:16:
/usr/include/bits/string_fortified.h:29:10: error: ‘__builtin_memcpy’ offset [2, 27] from the object at ‘addr’ is out of the bounds of referenced subobject ‘ss_family’ with type ‘short unsigned int’ at offset 0 [-Werror=array-bounds]
   29 |   return __builtin___memcpy_chk (__dest, __src, __len,
      |          ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   30 |                                  __glibc_objsize0 (__dest));
      |                                  ~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from /usr/include/bits/socket.h:175,
                 from /usr/include/sys/socket.h:33,
                 from ../src/util/virsocket.h:66,
                 from ../src/util/virsocketaddr.h:21,
                 from ../src/util/virnetdevip.h:21,
                 from ../src/util/virnetdevip.c:21:
../src/util/virnetdevip.c: In function ‘virNetDevIPAddrGet’:
/usr/include/bits/socket.h:193:5: note: subobject ‘ss_family’ declared here
  193 |     __SOCKADDR_COMMON (ss_);    /* Address family, etc.  */
      |     ^~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors

The code is correct, and this only happens when building at -O2.

The docs for -Warray-bounds say that a value of "2" is known to
be liable to generate false positives. Rather than downgrade the
check everywhere, we do it selectively for sanitizers.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
---
 meson.build | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/meson.build b/meson.build
index ca4291e37a..d4c142eebb 100644
--- a/meson.build
+++ b/meson.build
@@ -227,6 +227,11 @@ alloc_max = run_command(
 # sanitizer instrumentation may enlarge stack frames
 stack_frame_size = get_option('b_sanitize') == 'none' ? 4096 : 8192
 
+# array_bounds=2 check triggers false positive on some GCC
+# versions when using sanitizers. Seen on Fedora 34 with
+# GCC 11.1.1
+array_bounds = get_option('b_sanitize') == 'none' ? 2 : 1
+
 cc_flags += [
   '-fasynchronous-unwind-tables',
   '-fexceptions',
@@ -238,7 +243,7 @@ cc_flags += [
   '-Waggressive-loop-optimizations',
   '-Walloc-size-larger-than=@0@'.format(alloc_max.stdout().strip()),
   '-Walloca',
-  '-Warray-bounds=2',
+  '-Warray-bounds=@0@'.format(array_bounds),
   '-Wattribute-alias=2',
   '-Wattribute-warning',
   '-Wattributes',
-- 
2.31.1

On Mon, 2021-07-19 at 19:19 +0100, Daniel P. Berrangé wrote:
> When building with sanitizers on Fedora we get a wierd error

weird

> message
> 
> In file included from /usr/include/string.h:519,
>                  from ../src/internal.h:28,
>                  from ../src/util/virsocket.h:21,
>                  from ../src/util/virsocketaddr.h:21,
>                  from ../src/util/virnetdevip.h:21,
>                  from ../src/util/virnetdevip.c:21:
> In function ‘memcpy’,
>     inlined from ‘virNetDevGetifaddrsAddress’ at
> ../src/util/virnetdevip.c:702:13,
>     inlined from ‘virNetDevIPAddrGet’ at
> ../src/util/virnetdevip.c:754:16:
> /usr/include/bits/string_fortified.h:29:10: error: ‘__builtin_memcpy’
> offset [2, 27] from the object at ‘addr’ is out of the bounds of
> referenced subobject ‘ss_family’ with type ‘short unsigned int’ at
> offset 0 [-Werror=array-bounds]
>    29 |   return __builtin___memcpy_chk (__dest, __src, __len,
>       |          ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>    30 |                                  __glibc_objsize0 (__dest));
>       |                                  ~~~~~~~~~~~~~~~~~~~~~~~~~~
> In file included from /usr/include/bits/socket.h:175,
>                  from /usr/include/sys/socket.h:33,
>                  from ../src/util/virsocket.h:66,
>                  from ../src/util/virsocketaddr.h:21,
>                  from ../src/util/virnetdevip.h:21,
>                  from ../src/util/virnetdevip.c:21:
> ../src/util/virnetdevip.c: In function ‘virNetDevIPAddrGet’:
> /usr/include/bits/socket.h:193:5: note: subobject ‘ss_family’
> declared here
>   193 |     __SOCKADDR_COMMON (ss_);    /* Address family, etc.  */
>       |     ^~~~~~~~~~~~~~~~~
> cc1: all warnings being treated as errors
> 
> The code is correct, and this only happens when building at -O2.
> 
> The docs for -Warray-bounds say that a value of "2" is known to
> be liable to generate false positives. Rather than downgrade the
> check everywhere, we do it selectively for sanitizers.
> 
> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>

Reviewed-by: Tim Wiederhake <twiederh@redhat.com>

> ---
>  meson.build | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
> 
> diff --git a/meson.build b/meson.build
> index ca4291e37a..d4c142eebb 100644
> --- a/meson.build
> +++ b/meson.build
> @@ -227,6 +227,11 @@ alloc_max = run_command(
>  # sanitizer instrumentation may enlarge stack frames
>  stack_frame_size = get_option('b_sanitize') == 'none' ? 4096 : 8192
>  
> +# array_bounds=2 check triggers false positive on some GCC
> +# versions when using sanitizers. Seen on Fedora 34 with
> +# GCC 11.1.1
> +array_bounds = get_option('b_sanitize') == 'none' ? 2 : 1
> +
>  cc_flags += [
>    '-fasynchronous-unwind-tables',
>    '-fexceptions',
> @@ -238,7 +243,7 @@ cc_flags += [
>    '-Waggressive-loop-optimizations',
>    '-Walloc-size-larger-than=@0@'.format(alloc_max.stdout().strip()),
>    '-Walloca',
> -  '-Warray-bounds=2',
> +  '-Warray-bounds=@0@'.format(array_bounds),
>    '-Wattribute-alias=2',
>    '-Wattribute-warning',
>    '-Wattributes',