Hi,
On 9/17/2021 7:16 AM, Yonghong Song wrote:
>
>
> On 9/16/21 6:55 AM, Hou Tao wrote:
>> Commit 9df1c28bb752 ("bpf: add writable context for raw tracepoints")
>> supports writable context for tracepoint, but it misses the support
>> for bare tracepoint which has no associated trace event.
>>
>> Bare tracepoint is defined by DECLARE_TRACE(), so adding a corresponding
>> DECLARE_TRACE_WRITABLE() macro to generate a definition in __bpf_raw_tp_map
>> section for bare tracepoint in a similar way to DEFINE_TRACE_WRITABLE().
>>
>> Signed-off-by: Hou Tao <houtao1@xxxxxxxxxx>
>> ---
>> include/trace/bpf_probe.h | 19 +++++++++++++++----
>> 1 file changed, 15 insertions(+), 4 deletions(-)
>>
>> diff --git a/include/trace/bpf_probe.h b/include/trace/bpf_probe.h
>> index a23be89119aa..d08ee1060d82 100644
>> --- a/include/trace/bpf_probe.h
>> +++ b/include/trace/bpf_probe.h
>> @@ -93,8 +93,7 @@ __section("__bpf_raw_tp_map") = { \
>> #define FIRST(x, ...) x
>> -#undef DEFINE_EVENT_WRITABLE
>> -#define DEFINE_EVENT_WRITABLE(template, call, proto, args, size) \
>> +#define __CHECK_WRITABLE_BUF_SIZE(call, proto, args, size) \
>> static inline void bpf_test_buffer_##call(void) \
>> { \
>> /* BUILD_BUG_ON() is ignored if the code is completely eliminated, but \
>> @@ -103,8 +102,12 @@ static inline void
>> bpf_test_buffer_##call(void) \
>> */ \
>> FIRST(proto); \
>> (void)BUILD_BUG_ON_ZERO(size != sizeof(*FIRST(args))); \
>> -} \
>> -__DEFINE_EVENT(template, call, PARAMS(proto), PARAMS(args), size)
>> +}
>> +
>> +#undef DEFINE_EVENT_WRITABLE
>> +#define DEFINE_EVENT_WRITABLE(template, call, proto, args, size) \
>> + __CHECK_WRITABLE_BUF_SIZE(call, PARAMS(proto), PARAMS(args), size) \
>> + __DEFINE_EVENT(template, call, PARAMS(proto), PARAMS(args), size)
>> #undef DEFINE_EVENT
>> #define DEFINE_EVENT(template, call, proto, args) \
>> @@ -119,10 +122,18 @@ __DEFINE_EVENT(template, call, PARAMS(proto),
>> PARAMS(args), size)
>> __BPF_DECLARE_TRACE(call, PARAMS(proto), PARAMS(args)) \
>> __DEFINE_EVENT(call, call, PARAMS(proto), PARAMS(args), 0)
>> +#undef DECLARE_TRACE_WRITABLE
>> +#define DECLARE_TRACE_WRITABLE(call, proto, args, size) \
>> + __CHECK_WRITABLE_BUF_SIZE(call, PARAMS(proto), PARAMS(args), size) \
>> + __BPF_DECLARE_TRACE(call, PARAMS(proto), PARAMS(args)) \
>> + __DEFINE_EVENT(call, call, PARAMS(proto), PARAMS(args), size)
>> +
>> #include TRACE_INCLUDE(TRACE_INCLUDE_FILE)
>> #undef DEFINE_EVENT_WRITABLE
>> +#undef DECLARE_TRACE_WRITABLE
>> #undef __DEFINE_EVENT
>> +#undef __CHECK_WRITABLE_BUF_SIZE
>
> Put "#undef __CHECK_WRITABLE_BUF_SIZE" right after "#undef
> DECLARE_TRACE_WRITABLE" since they are related to each other
> and also they are in correct reverse order w.r.t. __DEFINE_EVENT?
If considering __CHECK_WRITABLE_BUF_SIZE is used in both DECLARE_TRACE_WRITABLE and
DEFINE_EVENT_WRITABLE and the order of definitions, is the following order better ?
#undef DECLARE_TRACE_WRITABLE
#undef DEFINE_EVENT_WRITABLE
#undef __CHECK_WRITABLE_BUF_SIZE
>
>> #undef FIRST
>> #endif /* CONFIG_BPF_EVENTS */
>>
> .
