What a suspens ^^

2012/4/27 Markus Klotzbuecher <markus [dot] klotzbuecher [..] ...>:
> On Tue, Apr 24, 2012 at 04:05:09PM +0200, Markus Klotzbuecher wrote:
>> On Thu, Apr 19, 2012 at 10:34:27AM +0200, Markus Klotzbuecher wrote:
>> > On Wed, Apr 18, 2012 at 04:34:11PM +0000, Peter Soetens wrote:
>> > >
>> > > http://bugs.orocos.org/show_bug.cgi?id=945
>> > >
>> > >            Summary: Can't self-configure/start a LuaComponent
>> > >            Product: Toolchain
>> > >            Version: master
>> > >           Platform: All
>> > >         OS/Version: All
>> > >             Status: NEW
>> > >           Severity: major
>> > >           Priority: P3
>> > >          Component: Lua
>> > >         AssignedTo: orocos-dev [..] ...
>> > >         ReportedBy: peter [..] ...
>> > >                 CC: orocos-dev [..] ...
>> > >    Estimated Hours: 0.0
>> > >
>> > >
>> > > When loading a .lua file you can't start 'tc'. For example:
>> > >
>> > > require "rttlib"
>> > >
>> > > tc=rtt.getTC()
>> > >
>> > > tc:configure() -- hangs here
>> > >
>> > > The reason is that a plain RTT::os::Mutex is used in LuaComponent instead of a
>> > > RTT::os::MutexRecursive
>> > >
>> > > I think it should be able to self-configure and self-start, just like you can
>> > > do in a C++ component.
>> >
>> > Ack! Patch welcome...
>>
>> Nack! I just remembered why this is not allowed: if you call yourself
>> your are going from Lua to C++ and back into the same Lua. However the
>> interpreter virtual stack and state is setup to expect the C++ call to
>> return and not to receive a new call. Strange things would happen...
>
> Un-Nack, I was wrong! Recursive self-calls are legal and even common
> in Lua (just confirmed it on the Lua ML), (which also makes sense as
> by calling into C we cleanly leave the Lua state). So you patch
> switching to recusive mutexes would be welcome again :-)
>
> Markus
> --
> Orocos-Dev mailing list
> Orocos-Dev [..] ...
> http://lists.mech.kuleuven.be/mailman/listinfo/orocos-dev