This is the mail archive of the 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: [RFA] Make the prec support signal better[0/4]

>>>>> "Joel" == Joel Brobecker <> writes:

>> if [target_info exists gdb,nosignals] {
>> verbose "Skipping sigall-reverse.exp because of nosignals."
>> continue
>> }

Joel> I wonder why we do a continue here, whereas we do a return elsewhere:

>> if ![target_info exists gdb,can_reverse] {
>> return
>> }

Joel> I wish we had a cookbook for writing testcases, I always forget what
Joel> we're supposed to do :-(. Anyone knows if this is significant?

Internally, all Tcl functions return a result code.  It has been a
while, but ISTR the codes are something like: OK, ERROR, CONTINUE,
BREAK, RETURN.  This allows execution control without the use of
longjmp... the "return" function returns RETURN, continue returns
CONTINUE, etc; then a surrounding loop function examines the code to
decide what to do next.

I presume, without looking, that the "source" command probably treats
CONTINUE, BREAK, and RETURN equivalently.

That is, there is likely no difference.  "return" is clearer, though.

I agree it would be good to have a test case cookbook, and guidelines.


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