summary |
shortlog |
log |
commit | commitdiff |
tree
raw |
patch |
inline | side by side (from parent 1:
3097761)
The current lttng-modules ctf2 implementation produce the following
"no field" event record:
{
"type": "event-record-class",
"id": 0,
"name": "syscall_entry_gettid"
"data-stream-class-id": 0,
}
Which is valid and simply result in an output without payload.
In ctf1 the same "event record" is:
event {
name = "syscall_entry_gettid";
id = 0;
stream_id = 0;
fields := struct {
};
};
Resulting in an empty but present payload -> " { } "
The tests looked for the presence of a payload ("{.*}"),
For now I decided to modify the regex since I have no feedback from eepp
for the last 5 hours... And ultimately I'm not sure what we should
prioritize in term of preventing "regression".
A ctf2 event record yielding the same output would be:
{
"type": "event-record-class",
"id": 0,
"name": "syscall_entry_gettid"
"data-stream-class-id": 0,
"payload-field-class": {
"type": "structure"
}
}
Signed-off-by: Jonathan Rajotte <jonathan.rajotte-julien@efficios.com>
Change-Id: I42a38d15a2968b278d6ab6625799e5fdcb2d7e53
return recorded_callstack
# Regex capturing the callstack_user and callstack_kernel context
return recorded_callstack
# Regex capturing the callstack_user and callstack_kernel context
-user_cs_rexp='.*callstack_user\ \=\ \[(.*)\]\ .*\}, \{.*\}'
-kernel_cs_rexp='.*callstack_kernel\ \=\ \[(.*)\]\ .*\}, \{.*\}'
+user_cs_rexp='.*callstack_user\ \=\ \[(.*)\]\ .*\}'
+kernel_cs_rexp='.*callstack_kernel\ \=\ \[(.*)\]\ .*\}'