Common objective for real time executive.
With the sophistication of microcontroller and the advance of modern communication stack and bus technology the environment a microcontroller is used now is ever more complex. The increasingly complex environment is fuelling the need for a context switching executive to handle background tasks.
In most embedded control environment, time-critical functions are likely to be shifted to hardware or interrupt driven routines, leaving the executive to do non-realtime tasks, nonetheless the response time is still the emphasis, the trade-off between response time and throughput is carefully considered.
The consideration of plugging in a real time executive to existing loop based task handling is taken into account.
Resource monitoring and instrumentation of the RTX is a key component of the design. Intregrating these functionalities to existing open source tool will be preferred over non-standard tools.
The inclusion of a standard IP stack, USB bus will form part of the DRTX framework.
A collection of random thoughts and materials that might prove enlightening to me and my friends.
MathJax
Sunday, July 31, 2005
Friday, July 22, 2005
OLS 2005 - day 3
A late start for today since the topics aren't truely relevant to NMX.
But Ian Pratt's talk on Xen is the most spectacular our of them all.
Bumped into Marcelo today and introduce myself to him. He was chatting with Andrew Morton at the time.
But Ian Pratt's talk on Xen is the most spectacular our of them all.
Bumped into Marcelo today and introduce myself to him. He was chatting with Andrew Morton at the time.
Thursday, July 21, 2005
OLS 2005 - day 2
Today is an extremely productive day. I got to meet pretty much all the ppc embedded people. They are a great bunch. However I missed the CE forum's embedded BOF, it happened at the same time as the ppc BOF, but it was well worth the time.
Matt Porter's talk on RapidIO is excellent as well.
Rusty Russell's talk on nfsim is cool, he is truely another great figure.
Greg K-H is also an amazing driver person.
Murphy's embedded system seems to be a good idea. Should check out his paper when I have the time.
Profiling Java and pretty much anything else in Linux is done using OProfile now.
Matt Porter's talk on RapidIO is excellent as well.
Rusty Russell's talk on nfsim is cool, he is truely another great figure.
Greg K-H is also an amazing driver person.
Murphy's embedded system seems to be a good idea. Should check out his paper when I have the time.
Profiling Java and pretty much anything else in Linux is done using OProfile now.
Wednesday, July 20, 2005
OLS 2005 - day 1
The spotlight of the first day is J. Corbet's Road map for the 2.6 kernel. It has interesting git talk and new features for the kernel. Very Informative.
The CELinux forum is holding an embedded BOF tomorrow sounds very interesting. Will also be meeting up with the ppc group tomorrow. XIP, and Power management will be excellent topics to be explored tomorrow.
Timesys is not doing particularly well as the real-time module is not making any sales and the notable exit of Jason M.
The bluetooth stack is a humongous beast to be honest. Intel's ACPI spec is another beast. Those two topics really make me feel like a small tiny potato.
Noted from today's display booth is that CELinux fixed the ARM compiler in thumb mode among other things. C++ is nowhere in sight for embedded computing.
The CELinux forum is holding an embedded BOF tomorrow sounds very interesting. Will also be meeting up with the ppc group tomorrow. XIP, and Power management will be excellent topics to be explored tomorrow.
Timesys is not doing particularly well as the real-time module is not making any sales and the notable exit of Jason M.
The bluetooth stack is a humongous beast to be honest. Intel's ACPI spec is another beast. Those two topics really make me feel like a small tiny potato.
Noted from today's display booth is that CELinux fixed the ARM compiler in thumb mode among other things. C++ is nowhere in sight for embedded computing.
Subscribe to:
Posts (Atom)