On Mon, 2008-01-07 at 11:51 -0800, chromatic via RT wrote: > On Monday 07 January 2008 08:18:24 Geoffrey Broadwell wrote: > > It seems like this should be available from interpinfo in PIR ... and > > since 'parrot -v' displays optimization information, it should probably > > display the runcore chosen as well. > > I'm not sure half of this is entirely useful. The way I've always selected a > runcore is on the command line, so you'd have to write: > > parrot --runcore=LOOKHERE -v > > ... to display the runcore you're using with -v.
The problem is the -O and -j options, and their interaction with other things. -O (== -O1) and -O2 are both supposed to alter the default runcore selection (according to the running doc, at least). Turns out they don't, but it's not obvious how to tell that. But in any case, just like -v shows parrot's interpretation of the optimization flags, it should show parrot's choice of runcore, especially if it got there by choosing a non-default default. > Being able to detect a particular runcore from within a program is much more > useful. This works with particle's patch, as demonstrated by the PIR program I nopasted to #parrot. I'm not sure what particle was planning next. -'f
