Re: help: dejagnu stumps me again

[Elena Zannoni <ezannoni@redhat.com>]

Daniel Jacobowitz writes:
 > On Thu, Jan 10, 2002 at 12:54:46PM -0500, Andrew Cagney wrote:
 > > >RCS file: /cvs/src/src/gdb/testsuite/gdb.c++/demangle.exp,v
 > > >retrieving revision 1.1.1.4
 > > >diff -u -r1.1.1.4 demangle.exp
 > > >--- demangle.exp	1999/09/09 00:00:27	1.1.1.4
 > > >+++ demangle.exp	2002/01/10 03:42:29
 > > >@@ -531,7 +531,18 @@
 > > >     test_demangling_exact "gnu: 
 > > >     foo__I_4000000000000000000000000000000000000000000000000000000000000000000000000" "Can't demangle \"foo__I_4000000000000000000000000000000000000000000000000000000000000000000000000\""
 > > > 
 > > >     ## 1999-04-19: "Fix from Dale Hawkins".  Shouldn't segfault.
 > > >-    test_demangling_exact "gnu: 
 > > >__thunk_64__0RL__list__Q29CosNaming20_proxy_NamingContextUlRPt25_CORBA_Unbounded_Sequence1ZQ29CosNaming7BindingRPQ29CosNaming15BindingIterator" "Can't demangle \"__thunk_64__0RL__list__Q29CosNaming20_proxy_NamingContextUlRPt25_CORBA_Unbounded_Sequence1ZQ29CosNaming7BindingRPQ29CosNaming15BindingIterator\""
 > > >+    # Accept even a dubious demangling; the string is ambiguous.
 > > >+    send_gdb "maintenance demangle 
 > > >__thunk_64__0RL__list__Q29CosNaming20_proxy_NamingContextUlRPt25_CORBA_Unbounded_Sequence1ZQ29CosNaming7BindingRPQ29CosNaming15BindingIterator\n"
 > > >+    gdb_expect {
 > > >+	-ex "virtual function thunk (delta:-64) for 
 > > >CosNaming::_proxy_NamingContext::_0RL__list(unsigned long, 
 > > >_CORBA_Unbounded_Sequence<CosNaming::Binding> *&, 
 > > >CosNaming::BindingIterator *&)"
 > > >+		{ pass "gnu: 
 > > >__thunk_64__0RL__list__Q29CosNaming20_proxy_NamingContextUlRPt25_CORBA_Unbounded_Sequence1ZQ29CosNaming7BindingRPQ29CosNaming15BindingIterator" }
 > > >+	-ex "Can't demangle 
 > > >\"__thunk_64__0RL__list__Q29CosNaming20_proxy_NamingContextUlRPt25_CORBA_Unbounded_Sequence1ZQ29CosNaming7BindingRPQ29CosNaming15BindingIterator\""
 > > >+		{ pass "gnu: 
 > > >__thunk_64__0RL__list__Q29CosNaming20_proxy_NamingContextUlRPt25_CORBA_Unbounded_Sequence1ZQ29CosNaming7BindingRPQ29CosNaming15BindingIterator" }
 > > >+	-re ".*$gdb_prompt $"
 > > >+		{ fail "gnu: 
 > > >__thunk_64__0RL__list__Q29CosNaming20_proxy_NamingContextUlRPt25_CORBA_Unbounded_Sequence1ZQ29CosNaming7BindingRPQ29CosNaming15BindingIterator" }
 > > >+	timeout
 > > >+		{ fail "gnu: 
 > > >__thunk_64__0RL__list__Q29CosNaming20_proxy_NamingContextUlRPt25_CORBA_Unbounded_Sequence1ZQ29CosNaming7BindingRPQ29CosNaming15BindingIterator (timeout)" }
 > > >+    }
 > > > }
 > > >
 > > 
 > > An un anchored pattern such as ".*...." is bad.  It sends expect into 
 > > drug induced bliss while it tries to match everything.
 > > 
 > > TCL is fun.  expect well, deajgnu er ...
 > 
 > I don't see why it's a problem - that's how we match the prompt on all
 > the other gdb_expect blocks.  I'll keep poking at it.

Dan,
I usually wrap the clause that gives me a hard time with a 
exp_internal 1
....
exp_internal 0

That spews out all the expect buffers as it tries to do the matching,
so you can see what actually goes wrong.

Elena


 > 
 > -- 
 > Daniel Jacobowitz                           Carnegie Mellon University
 > MontaVista Software                         Debian GNU/Linux Developer