This is the mail archive of the gdb-patches@sourceware.org mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] Modify trace-status output to always output stop-notes.


On 10/10/12 4:17 AM, Dmitry Kozlov wrote:
Hi,
accoriding to discussion in gdb mailing list I have added stop-notes as mandatory part of MI trace-status command. This makes processing of stop-notes on IDE side more uniform with trace-notes processing. Previously Yao mentioned that he implemented notifications that can be used to monitor trace-status, but putting stop-notes to trace-status looks more clear way because processing one extra trace-status parameter defferently from others.


Stan, is it ok?

Yes, this just needs a couple tweaks:


+ * gdbserver/traicepoint.c (cmd_qtstatus): Modify trace-status output to


Extra 'i' ...plus gdbserver has its own ChangeLog where this piece of the entry should go.


@@ -2164,6 +2164,7 @@ trace_status_mi (int on_stop)
ui_out_field_string (uiout, "user-name", ts->user_name);
ui_out_field_string (uiout, "notes", ts->notes);
+ ui_out_field_string (uiout, "stop-notes", ts->stop_desc);



It would be useful to add a brief note to the effect that this is unconditional for the MI command, so that some future version of ourselves doesn't assume that there is a missing if-statement.


We really should have a testsuite test for this. Yao's recently-added mi-* tests in gdb.trace look like they would be good candidates to crib from.

Stan
stan@codesourcery.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]