On 11/11, Eduard Zingerman wrote:
A fix for the LLVM compilation error while building bpftool.
Replaces the expression:
_Static_assert((p) == NULL || ...)
by expression:
_Static_assert((__builtin_constant_p((p)) ? (p) == NULL : 0) || ...)
IIUC, when __builtin_constant_p(p) returns false, we just ignore the NULL
check?
Do we have cases like that? If no, maybe it's safer to fail?
s/(p) == NULL : 0/(p) == NULL : 1/ ?
When "p" is not a constant the former is not considered to be a
constant expression by LLVM 14.
The error was introduced in the following patch-set: [1].
The error was reported here: [2].
Reported-by: kernel test robot <lkp@xxxxxxxxx>
Signed-off-by: Eduard Zingerman <eddyz87@xxxxxxxxx>
[1] https://lore.kernel.org/bpf/20221109142611.879983-1-eddyz87@xxxxxxxxx/
[2] https://lore.kernel.org/all/202211110355.BcGcbZxP-lkp@xxxxxxxxx/
---
tools/lib/bpf/hashmap.h | 3 ++-
tools/perf/util/hashmap.h | 3 ++-
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/tools/lib/bpf/hashmap.h b/tools/lib/bpf/hashmap.h
index 3fe647477bad..0a5bf1937a7c 100644
--- a/tools/lib/bpf/hashmap.h
+++ b/tools/lib/bpf/hashmap.h
@@ -123,7 +123,8 @@ enum hashmap_insert_strategy {
};
#define hashmap_cast_ptr(p) ({ \
- _Static_assert((p) == NULL || sizeof(*(p)) == sizeof(long), \
+ _Static_assert((__builtin_constant_p((p)) ? (p) == NULL : 0) || \
+ sizeof(*(p)) == sizeof(long), \
#p " pointee should be a long-sized integer or a pointer"); \
(long *)(p); \
})
diff --git a/tools/perf/util/hashmap.h b/tools/perf/util/hashmap.h
index 3fe647477bad..0a5bf1937a7c 100644
--- a/tools/perf/util/hashmap.h
+++ b/tools/perf/util/hashmap.h
@@ -123,7 +123,8 @@ enum hashmap_insert_strategy {
};
#define hashmap_cast_ptr(p) ({ \
- _Static_assert((p) == NULL || sizeof(*(p)) == sizeof(long), \
+ _Static_assert((__builtin_constant_p((p)) ? (p) == NULL : 0) || \
+ sizeof(*(p)) == sizeof(long), \
#p " pointee should be a long-sized integer or a pointer"); \
(long *)(p); \
})
--
2.34.1
