On 2/18/22 9:39 PM, Mauricio Vásquez wrote:
This commit fixes a compilation error on systems with glibc < 2.26 [0]:
```
In file included from main.h:14:0,
from gen.c:24:
linux/tools/include/tools/libc_compat.h:11:21: error: attempt to use poisoned "reallocarray"
static inline void *reallocarray(void *ptr, size_t nmemb, size_t size)
```
This happens because gen.c pulls <bpf/libbpf_internal.h>, and then
<tools/libc_compat.h> (through main.h). When
COMPAT_NEED_REALLOCARRAY is set, libc_compat.h defines reallocarray()
which libbpf_internal.h poisons with a GCC pragma.
This follows the same approach of libbpf in commit
029258d7b228 ("libbpf: Remove any use of reallocarray() in libbpf").
Reported-by: Quentin Monnet <quentin@xxxxxxxxxxxxx>
Signed-off-by: Mauricio Vásquez <mauricio@xxxxxxxxxx>
[0]: https://lore.kernel.org/bpf/3bf2bd49-9f2d-a2df-5536-bc0dde70a83b@xxxxxxxxxxxxx/
[...]
+ * Copied from tools/lib/bpf/libbpf_internal.h
+ */
+static inline void *bpftool_reallocarray(void *ptr, size_t nmemb, size_t size)
+{
+ size_t total;
+
+#if __has_builtin(__builtin_mul_overflow)
+ if (unlikely(__builtin_mul_overflow(nmemb, size, &total)))
+ return NULL;
+#else
+ if (size == 0 || nmemb > ULONG_MAX / size)
+ return NULL;
+ total = nmemb * size;
+#endif
+ return realloc(ptr, total);
+}
Can't we just reuse libbpf_reallocarray() given we copy over libbpf_internal.h
anyway via a9caaba399f9 ("bpftool: Implement "gen min_core_btf" logic")?
Thanks,
Daniel
