From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me X-Spam-Level: X-Spam-Status: No, score=-1.3 required=5.0 tests=BAYES_00,INVALID_MSGID autolearn=no autolearn_force=no version=3.4.4 X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 103376,afca8f81aa244179 X-Google-Attributes: gid103376,public From: Tucker Taft Subject: Re: Compiler able to map Tasks on MP-Systems? Date: 2000/08/09 Message-ID: <3991DF1D.98316124@averstar.com>#1/1 X-Deja-AN: 656538705 Content-Transfer-Encoding: 7bit References: <398A67D6.2E8FEE8C@sbox.tu-graz.ac.at> X-Accept-Language: en Content-Type: text/plain; charset=us-ascii X-Complaints-To: usenet@inmet2.burl.averstar.com X-Trace: inmet2.burl.averstar.com 965861150 12217 141.199.8.77 (9 Aug 2000 22:45:50 GMT) Organization: AverStar (formerly Intermetrics) Burlington, MA USA Mime-Version: 1.0 NNTP-Posting-Date: 9 Aug 2000 22:45:50 GMT Newsgroups: comp.lang.ada Date: 2000-08-09T22:45:50+00:00 List-Id: Manfred Muecke wrote: > > We do research in the field of distributed computing. > We are looking for an Ada Compiler which is able to map an Ada program > consisting of > several tasks on a Multiprocessor system (currently we are using the > TMS320C40, which holds 4 TI C40 DSPs). > > Is anyone of you aware of such a Compiler or has even already used it? Many Ada compilers map tasks across multiple processors. The most frequent way to do this is to map Ada tasks to operating system threads, and then rely on the underlying operating system to map threads across multiple processors. However, for "bare machine" run-times, where there is no underlying operating system, the Ada run-time has to do it itself (and there are many that do so). As far as the TI C40, I don't know which vendors are supporting that platform currently. We are supporting the Analog Devices SHARC, but the multiple processor configuration we are using does not have any shared memory, and the Ada run-time generally requires that all tasks of a single Ada program (partition) be in the same address space. Hence, we map separate Ada programs to separate SHARC processors. > Thanks in advance, Manfred > > Manfred Muecke > moskito@sbox.tu-graz.ac.at -- -Tucker Taft stt@averstar.com http://www.averstar.com/~stt/ Technical Director, Commercial Division, AverStar (formerly Intermetrics) (http://www.averstar.com/services/IT_consulting.html) Burlington, MA USA