Re: vulcano math problem
Sebastian Kuzminsky <[email protected]>
| Newsgroups | gmane.linux.real-time.rtai |
|---|---|
| Message-ID | <[email protected]> |
On 11/26/2015 03:27 AM, Paolo Mantegazza wrote: > On 11/26/2015 12:05 AM, Sebastian Kuzminsky wrote: >> FYI: I switched from glibc's libm.a to a hand-built newlib libm.a >> specifically for the kernel (using the configure lines in rtai's >> README.KLIBM), and everything is working well for me now. >> >> Thanks! > Thanks for the feedback info. > I still do not get why the standard 32 bits libm works for me, while you > have missing entries that should not exist. Yeah, that's strange. But i think building the math code from source is cleaner in every way, and newlib seems like a good collection of math source, so this is a good outcome and I'm happy. > I know that the need of choosing an option among newlib-uclib-libm is a > bit annoying for a user. > Nevertheless, the working one eventually provides a more comprehensive > math kernel support than before. Yes, and it's also great that the new RTAI math source (newlib) is maintained by a different project that focuses on that, and that it's free of the licensing encumberance that troubled the old one. All in all, a very positive move forward for the RTAI project, and a big improvement for downstream projects like LinuxCNC, so thank you!! -- Sebastian Kuzminsky _______________________________________________ Rtai mailing list [email protected] https://mail.rtai.org/cgi-bin/mailman/listinfo/rtai