Comments on: The S4D Debug Conference http://jakob.engbloms.se/archives/942?&owa_medium=feed&owa_sid= Computer Technology: Simulation, Virtualization, Virtual Platforms, Embedded, Multicore and Multiprocessing (by Jakob Engblom) Thu, 12 Jan 2012 21:54:16 -0800 hourly 1 http://wordpress.org/?v=3.3.1 By: Jakob http://jakob.engbloms.se/archives/942/comment-page-1#comment-2555 Jakob Wed, 14 Oct 2009 04:30:41 +0000 http://jakob.engbloms.se/?p=942#comment-2555 You are right. You did talk about greater-than-SoC debug. The main issue with the hardware-synchronized approach is that you need a system small enough in physical size that you can keep it synchronized. But in essence, you are chasing the right thing: debug across multiple SoC, using extended on-chip logic that lets you do synchronized commands across the entire set. You are right. You did talk about greater-than-SoC debug. The main issue with the hardware-synchronized approach is that you need a system small enough in physical size that you can keep it synchronized. But in essence, you are chasing the right thing: debug across multiple SoC, using extended on-chip logic that lets you do synchronized commands across the entire set.

]]> By: Roland Höller http://jakob.engbloms.se/archives/942/comment-page-1#comment-2554 Roland Höller Tue, 13 Oct 2009 14:17:42 +0000 http://jakob.engbloms.se/?p=942#comment-2554 Hi! I just wanted to add a comment on your statemant that "system-level debug was not in any talk". Unfortunately you seem to have missed my talk on the second day, where I presented our approach to coordinate debug and test activities of a multitude of nodes of a distributed system. The approach is based on precisely synchronized time-of-day clocks used to trigger debug and test action, operating with little hardware overhead in the CPU chips. I add a link to one of our project websites in case a reader is interested in some more information: http://embsys.technikum-wien.at/projects/coordes/index.html Best regards, Roland Höller Hi!

I just wanted to add a comment on your statemant that “system-level debug was not in any talk”. Unfortunately you seem to have missed my talk on the second day, where I presented our approach to coordinate debug and test activities of a multitude of nodes of a distributed system. The approach is based on precisely synchronized time-of-day clocks used to trigger debug and test action, operating with little hardware overhead in the CPU chips.

I add a link to one of our project websites in case a reader is interested in some more information: http://embsys.technikum-wien.at/projects/coordes/index.html

Best regards,

Roland Höller

]]>