> From: Christophe Coutand > > I doubt there is any limitation that prevent LIBM to be used > in driver code.
On the x86, how does eCos deal with the FP context? Does it include it as part of every thread and DSR context, or just thread context? Does it rely on the chip mechanism to defer FP context swapping until an actual FP instruction is made following a thread switch? I can imagine a kernel implementation that would assume that no DSR (let alone ISR) would ever use FP, and which would crash and burn if one ever did. But I've never looked at the x86 HAL. -- Ciao, Paul D. DeRocco Paul mailto:pdero...@ix.netcom.com -- Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss
