2019-08-23 14:49:09

by Alakesh Haloi

Subject: [PATCH] seccomp: fix compilation errors in seccomp-bpf kselftest

Without this patch we see following error while building and kselftest
for secccomp_bpf fails.

seccomp_bpf.c:1787:20: error: ‘PTRACE_EVENTMSG_SYSCALL_ENTRY’ undeclared (first use in this function);
seccomp_bpf.c:1788:6: error: ‘PTRACE_EVENTMSG_SYSCALL_EXIT’ undeclared (first use in this function);

Signed-off-by: Alakesh Haloi <[email protected]>
---
tools/testing/selftests/seccomp/seccomp_bpf.c | 8 ++++++++
1 file changed, 8 insertions(+)

diff --git a/tools/testing/selftests/seccomp/seccomp_bpf.c b/tools/testing/selftests/seccomp/seccomp_bpf.c
index 6ef7f16c4cf5..2e619760fc3e 100644
--- a/tools/testing/selftests/seccomp/seccomp_bpf.c
+++ b/tools/testing/selftests/seccomp/seccomp_bpf.c
@@ -1353,6 +1353,14 @@ TEST_F(precedence, log_is_fifth_in_any_order)
#define PTRACE_EVENT_SECCOMP 7
#endif

+#ifndef PTRACE_EVENTMSG_SYSCALL_ENTRY
+#define PTRACE_EVENTMSG_SYSCALL_ENTRY 1
+#endif
+
+#ifndef PTRACE_EVENTMSG_SYSCALL_EXIT
+#define PTRACE_EVENTMSG_SYSCALL_EXIT 2
+#endif
+
#define IS_SECCOMP_EVENT(status) ((status >> 16) == PTRACE_EVENT_SECCOMP)
bool tracer_running;
void tracer_stop(int sig)
--
2.17.1

2019-08-30 14:12:00

by Shuah Khan

Subject: Re: [PATCH] seccomp: fix compilation errors in seccomp-bpf kselftest

On 8/22/19 3:58 PM, Alakesh Haloi wrote:
> Without this patch we see following error while building and kselftest
> for secccomp_bpf fails.
>
> seccomp_bpf.c:1787:20: error: ‘PTRACE_EVENTMSG_SYSCALL_ENTRY’ undeclared (first use in this function);
> seccomp_bpf.c:1788:6: error: ‘PTRACE_EVENTMSG_SYSCALL_EXIT’ undeclared (first use in this function);
>
> Signed-off-by: Alakesh Haloi <[email protected]>
> ---
> tools/testing/selftests/seccomp/seccomp_bpf.c | 8 ++++++++
> 1 file changed, 8 insertions(+)
>
> diff --git a/tools/testing/selftests/seccomp/seccomp_bpf.c b/tools/testing/selftests/seccomp/seccomp_bpf.c
> index 6ef7f16c4cf5..2e619760fc3e 100644
> --- a/tools/testing/selftests/seccomp/seccomp_bpf.c
> +++ b/tools/testing/selftests/seccomp/seccomp_bpf.c
> @@ -1353,6 +1353,14 @@ TEST_F(precedence, log_is_fifth_in_any_order)
> #define PTRACE_EVENT_SECCOMP 7
> #endif
>
> +#ifndef PTRACE_EVENTMSG_SYSCALL_ENTRY
> +#define PTRACE_EVENTMSG_SYSCALL_ENTRY 1
> +#endif
> +
> +#ifndef PTRACE_EVENTMSG_SYSCALL_EXIT
> +#define PTRACE_EVENTMSG_SYSCALL_EXIT 2
> +#endif
> +
> #define IS_SECCOMP_EVENT(status) ((status >> 16) == PTRACE_EVENT_SECCOMP)
> bool tracer_running;
> void tracer_stop(int sig)
>

Hi Kees,

Okay to apply this one for 5.4-rc1. Or is this going through bpf tree?
If it is going through bpf tree:

Acked-by: Shuah Khan <[email protected]>

thanks,
-- Shuah

2019-08-30 15:00:48

by Shuah Khan

Subject: Re: [PATCH] seccomp: fix compilation errors in seccomp-bpf kselftest

On 8/30/19 8:09 AM, shuah wrote:
> On 8/22/19 3:58 PM, Alakesh Haloi wrote:
>> Without this patch we see following error while building and kselftest
>> for secccomp_bpf fails.
>>
>> seccomp_bpf.c:1787:20: error: ‘PTRACE_EVENTMSG_SYSCALL_ENTRY’
>> undeclared (first use in this function);
>> seccomp_bpf.c:1788:6: error: ‘PTRACE_EVENTMSG_SYSCALL_EXIT’ undeclared
>> (first use in this function);
>>
>> Signed-off-by: Alakesh Haloi <[email protected]>
>> ---
>>   tools/testing/selftests/seccomp/seccomp_bpf.c | 8 ++++++++
>>   1 file changed, 8 insertions(+)
>>
>> diff --git a/tools/testing/selftests/seccomp/seccomp_bpf.c
>> b/tools/testing/selftests/seccomp/seccomp_bpf.c
>> index 6ef7f16c4cf5..2e619760fc3e 100644
>> --- a/tools/testing/selftests/seccomp/seccomp_bpf.c
>> +++ b/tools/testing/selftests/seccomp/seccomp_bpf.c
>> @@ -1353,6 +1353,14 @@ TEST_F(precedence, log_is_fifth_in_any_order)
>>   #define PTRACE_EVENT_SECCOMP 7
>>   #endif
>> +#ifndef PTRACE_EVENTMSG_SYSCALL_ENTRY
>> +#define PTRACE_EVENTMSG_SYSCALL_ENTRY 1
>> +#endif
>> +
>> +#ifndef PTRACE_EVENTMSG_SYSCALL_EXIT
>> +#define PTRACE_EVENTMSG_SYSCALL_EXIT 2
>> +#endif
>> +
>>   #define IS_SECCOMP_EVENT(status) ((status >> 16) ==
>> PTRACE_EVENT_SECCOMP)
>>   bool tracer_running;
>>   void tracer_stop(int sig)
>>
>
> Hi Kees,
>
> Okay to apply this one for 5.4-rc1. Or is this going through bpf tree?
> If it is going through bpf tree:
>
> Acked-by: Shuah Khan <[email protected]>
>
> thanks,
> -- Shuah
>

I saw your mail about Tycho's solution to be your preferred. Ignore this
message. I am applying Tycho's patch.

thanks,
-- Shuah