From derek.m.graham at gmail.com Wed Jan 16 07:10:03 2008 From: derek.m.graham at gmail.com (Derek) Date: Wed, 16 Jan 2008 12:10:03 +0000 (UTC) Subject: Different usage Message-ID: Hi Suraj I would like to replace existing C/VPI code that interacts with the simulator with Ruby using ruby-vpi. This would involve loading/running some verilog, setting callbacks, and accessing the design hierarchy. How could I do this that doesn't involve writing specifications, test runners, etc as given in the current documentation? Thanks, Derek From snk at gna.org Wed Jan 16 13:12:55 2008 From: snk at gna.org (Suraj N. Kurapati) Date: Wed, 16 Jan 2008 10:12:55 -0800 (PST) Subject: Different usage In-Reply-To: Message-ID: <618611.84735.qm@web83313.mail.sp1.yahoo.com> Hi Derek, Derek wrote: > I would like to replace existing C/VPI code that > interacts with the simulator with Ruby using > ruby-vpi. This would involve loading/running some > verilog, setting callbacks, and accessing the design > hierarchy. > > How could I do this that doesn't involve writing > specifications, test runners, etc as given in the > current documentation? With the 19.0.0 release this is not possible. However, I did implement part of this feature a while back (I just need to write the documentation now): http://rubyforge.org/pipermail/ruby-vpi-discuss/2007-September/000077.html http://rubyforge.org/pipermail/ruby-vpi-discuss/2007-October/000081.html The basic idea is that the test runner invokes a user-defined Ruby script, which then does whatever it wants with Ruby-VPI. However, if you want to get rid of the test runner as well, I will need to provide two separate .so files: one containing vlog_startup_routines and one without. You could then use the .so file w/o that symbol to embed Ruby-VPI into an existing C/VPI application. I think this is a good approach to pursue. I will make a 20.0.0 release this week containing whatever I implemented so far. The following release, I'll provide the two separate .so files. From snk at gna.org Sun Jan 27 14:41:16 2008 From: snk at gna.org (Suraj N. Kurapati) Date: Sun, 27 Jan 2008 11:41:16 -0800 Subject: Different usage In-Reply-To: <618611.84735.qm@web83313.mail.sp1.yahoo.com> References: <618611.84735.qm@web83313.mail.sp1.yahoo.com> Message-ID: <479CDE5C.6000602@gna.org> Suraj N. Kurapati wrote: > I will make a 20.0.0 release this week containing > whatever I implemented so far. The following release, > I'll provide the two separate .so files. Sorry for the delay, I fell behind schedule due to other projects. I'll try to publish whatever code I have to the ruby-vpi code repository tonight and make an official release by next weekend. From snk at gna.org Mon Jan 28 02:13:38 2008 From: snk at gna.org (Suraj N. Kurapati) Date: Sun, 27 Jan 2008 23:13:38 -0800 Subject: Different usage In-Reply-To: <479CDE5C.6000602@gna.org> References: <618611.84735.qm@web83313.mail.sp1.yahoo.com> <479CDE5C.6000602@gna.org> Message-ID: <479D80A2.6050103@gna.org> Suraj N. Kurapati wrote: > Sorry for the delay, I fell behind schedule due to other projects. > I'll try to publish whatever code I have to the ruby-vpi code > repository tonight and make an official release by next weekend. Alright, I spent my Sunday afternoon and evening to make this release for you. :-) See the complete release announcement here: http://snk.tuxfamily.org/web/2008-01-27-ruby-vpi-20-0-0.html Be sure to read the new "Theory of Operation" section, which explains the new way tests are run and how to load multiple DUTs into a single test: http://ruby-vpi.rubyforge.org/#usage.theory Cheers.