Re: [Tinyos-help] errors casting int64_t to float on MSP430
On Wed, Oct 3, 2012 at 12:49 PM, Flemming Nyboe flemm...@rocketscience.euwrote: Update III: On Wed, Oct 3, 2012 at 1:00 PM, Flemming Nyboe flemm...@rocketscience.eu wrote: Update II: On Wed, Oct 3, 2012 at 12:16 PM, Flemming Nyboe flemm...@rocketscience.eu wrote: How does this reproduce on 4.6.3? It seems the answer is here: https://www.millennium.berkeley.edu/pipermail/tinyos-help/2012-October/055752.html It has been fixed in 4.6.3. I found that 4.6.3 20120301 (mspgcc LTS 20120406 unpatched) is actually available from http://sourceware.org/cygwinports/ That's cool one word of caution though... The mcu and libc packages from the above cygwinports are date stamped 20120606, the official LTS-20120406 (current 4.6.3 release) uses the mcu and libc packages available as of 20120406. It is probably okay but I wanted to give you a heads up. eric result: While 4.6.3 does the int64_t sum correctly, as reported by Marcin (link above), the conversion to float and back still fails in the same way as 4.5.3 did: clips to +/- 2^32. On that basis, FTSP skew calculation still fails, if the time points in its table are spaced far enough for localSum to exceed 2^32. Regards Flemming ___ Tinyos-help mailing list Tinyos-help@millennium.berkeley.edu https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help -- Eric B. Decker Senior (over 50 :-) Researcher ___ Tinyos-help mailing list Tinyos-help@millennium.berkeley.edu https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
Re: [Tinyos-help] errors casting int64_t to float on MSP430
On Fri, Oct 5, 2012 at 9:16 AM, Eric Decker one word of caution though... The mcu and libc packages from the above cygwinports are date stamped 20120606, the official LTS-20120406 (current 4.6.3 release) uses the mcu and libc packages available as of 20120406. It is probably okay but I wanted to give you a heads up. eric Thank you. I realize cygwin is not supported, and I should be using vmware. http://www.mail-archive.com/shimmer-users@eecs.harvard.edu/msg00413.html Currently the tinyos.net instructions http://docs.tinyos.net/tinywiki/index.php/Installing_TinyOS_2.1.1#Manual_installation_on_your_host_OS_with_RPMs are a bit misleading in describing a cygwin setup which is way outdated. I would suggest simply deleting it. Regards Flemming ___ Tinyos-help mailing list Tinyos-help@millennium.berkeley.edu https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
Re: [Tinyos-help] errors casting int64_t to float on MSP430
Update III: On Wed, Oct 3, 2012 at 1:00 PM, Flemming Nyboe flemm...@rocketscience.eu wrote: Update II: On Wed, Oct 3, 2012 at 12:16 PM, Flemming Nyboe flemm...@rocketscience.eu wrote: How does this reproduce on 4.6.3? It seems the answer is here: https://www.millennium.berkeley.edu/pipermail/tinyos-help/2012-October/055752.html It has been fixed in 4.6.3. I found that 4.6.3 20120301 (mspgcc LTS 20120406 unpatched) is actually available from http://sourceware.org/cygwinports/ result: While 4.6.3 does the int64_t sum correctly, as reported by Marcin (link above), the conversion to float and back still fails in the same way as 4.5.3 did: clips to +/- 2^32. On that basis, FTSP skew calculation still fails, if the time points in its table are spaced far enough for localSum to exceed 2^32. Regards Flemming ___ Tinyos-help mailing list Tinyos-help@millennium.berkeley.edu https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help