On 8/6/25 4:33 PM, Namhyung Kim wrote:
Hello,
On Wed, Aug 06, 2025 at 01:03:01AM +0100, Sam James wrote:
When exploring building bpf_skel with GCC's BPF support, there was a
buid failure because of bpf_core_field_exists vs the mem_hops bitfield:
```
In file included from util/bpf_skel/sample_filter.bpf.c:6:
util/bpf_skel/sample_filter.bpf.c: In function 'perf_get_sample':
tools/perf/libbpf/include/bpf/bpf_core_read.h:169:42: error: cannot take address of bit-field 'mem_hops'
169 | #define ___bpf_field_ref1(field) (&(field))
| ^
tools/perf/libbpf/include/bpf/bpf_helpers.h:222:29: note: in expansion of macro '___bpf_field_ref1'
222 | #define ___bpf_concat(a, b) a ## b
| ^
tools/perf/libbpf/include/bpf/bpf_helpers.h:225:29: note: in expansion of macro '___bpf_concat'
225 | #define ___bpf_apply(fn, n) ___bpf_concat(fn, n)
| ^~~~~~~~~~~~~
tools/perf/libbpf/include/bpf/bpf_core_read.h:173:9: note: in expansion of macro '___bpf_apply'
173 | ___bpf_apply(___bpf_field_ref, ___bpf_narg(args))(args)
| ^~~~~~~~~~~~
tools/perf/libbpf/include/bpf/bpf_core_read.h:188:39: note: in expansion of macro '___bpf_field_ref'
188 | __builtin_preserve_field_info(___bpf_field_ref(field), BPF_FIELD_EXISTS)
| ^~~~~~~~~~~~~~~~
util/bpf_skel/sample_filter.bpf.c:167:29: note: in expansion of macro 'bpf_core_field_exists'
167 | if (bpf_core_field_exists(data->mem_hops))
| ^~~~~~~~~~~~~~~~~~~~~
cc1: error: argument is not a field access
```
___bpf_field_ref1 was adapted for GCC in 12bbcf8e840f40b82b02981e96e0a5fbb0703ea9
but the trick added for compatibility in 3a8b8fc3174891c4c12f5766d82184a82d4b2e3e
isn't compatible with that as an address is used as an argument.
Workaround this by calling __builtin_preserve_field_info directly as the
bpf_core_field_exists macro does, but without the ___bpf_field_ref use.
IIUC GCC doesn't support bpf_core_fields_exists() for bitfield members,
right? Is it gonna change in the future?
Link: https://gcc.gnu.org/PR121420
Co-authored-by: Andrew Pinski <quic_apinski@xxxxxxxxxxx>
Signed-off-by: Sam James <sam@xxxxxxxxxx>
---
tools/perf/util/bpf_skel/sample_filter.bpf.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/perf/util/bpf_skel/sample_filter.bpf.c b/tools/perf/util/bpf_skel/sample_filter.bpf.c
index b195e6efeb8be..e5666d4c17228 100644
--- a/tools/perf/util/bpf_skel/sample_filter.bpf.c
+++ b/tools/perf/util/bpf_skel/sample_filter.bpf.c
@@ -164,7 +164,7 @@ static inline __u64 perf_get_sample(struct bpf_perf_event_data_kern *kctx,
if (entry->part == 8) {
union perf_mem_data_src___new *data = (void *)&kctx->data->data_src;
- if (bpf_core_field_exists(data->mem_hops))
+ if (__builtin_preserve_field_info(data->mem_hops, BPF_FIELD_EXISTS))
I believe those two are equivalent (maybe worth a comment?). But it'd
be great if BPF/clang folks can review if it's ok.
Yes, from clang side, they are almost equnivalent. See tools/lib/bpf/bpf_core_read.h.
#define bpf_core_field_exists(field...) \
__builtin_preserve_field_info(___bpf_field_ref(field), BPF_FIELD_EXISTS)
bpf_core_field_exists actually relies on clang builtin function
__builtin_preserve_field_info(). This builtin is handled in frontend and
also at early IR stage.
So your above code is okay to me although bpf_core_field_exists() is much
easy to understand the intent.
Anyway, I can build it with clang.
Tested-by: Namhyung Kim <namhyung@xxxxxxxxxx>
Thanks,
Namhyung
return data->mem_hops;
return 0;
--
2.50.1
