| Message ID | 1485888144-16531-6-git-send-email-mst@redhat.com |
|---|---|
| State | New |
| Headers | show |
On 01/31/2017 12:42 PM, Michael S. Tsirkin wrote: > There are theoretical concerns that some compilers might not trigger > build failures on attempts to define an array of size (x ? -1 : 1) where > x is a variable and make it a variable sized array instead. Let rewrite s/Let/Let's/ > using a struct with a negative bit field size instead as there are no > dynamic bit field sizes. This is similar to what Linux does. On one of the earlier review spins, I suggested making it explicit that we are now requiring that the condition be a compile-time constant (you got that wording in 2/23, but not here) - food for thought if you have to respin this. > > Signed-off-by: Michael S. Tsirkin <mst@redhat.com> > Reviewed-by: Eric Blake <eblake@redhat.com> > Reviewed-by: Markus Armbruster <armbru@redhat.com> > --- > include/qemu/compiler.h | 10 +++++++--- > 1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/include/qemu/compiler.h b/include/qemu/compiler.h index 77b9ce3..057639a 100644 --- a/include/qemu/compiler.h +++ b/include/qemu/compiler.h @@ -85,10 +85,14 @@ #define typeof_field(type, field) typeof(((type *)0)->field) #define type_check(t1,t2) ((t1*)0 - (t2*)0) +#define QEMU_BUILD_BUG_ON_STRUCT(x) \ + struct { \ + int:(x) ? -1 : 1; \ + } + #ifdef __COUNTER__ -#define QEMU_BUILD_BUG_ON(x) \ - typedef char glue(qemu_build_bug_on__, __COUNTER__)[(x) ? -1 : 1] \ - __attribute__((unused)) +#define QEMU_BUILD_BUG_ON(x) typedef QEMU_BUILD_BUG_ON_STRUCT(x) \ + glue(qemu_build_bug_on__, __COUNTER__) __attribute__((unused)) #else #define QEMU_BUILD_BUG_ON(x) #endif