Victor Yodaiken wrote: > > On Thu, Dec 27, 2001 at 12:48:24PM +0100, Paolo Mantegazza wrote: > > A hold story. > > Silly assumption for recent CPUs, fully violated in 2.4.xx, as shown by > > the Stanford code checker. On Athlons I think the kernel is using the > > FPU even more badly (not so sure). > > Standard FPU save/restore is in the range of 200 clock cycles, since > > every kid is now playing with 1G Celeron/Duron are you willing to mess > > If everything is a PC for you, that's true. We work a lot with little > boards and with dedicated RT systems. > > > things up just to save .2 us while any interrupt can cause much more > > Assuming no cache miss. > > > latency? I think that even on any 486 still sold there is no use in > > avoiding the FPU in the kernel. > > Many CPUs still have no fpu or are slow. the AMD Elan 133Mhz is a > perfectly good processor for many applications for example. >
We also work with smaller systems, down to embedded PC-104 with 386SX 33 MHz, but that has nothing to do with this posting thread. You are definitely missing the aim, nobody was talking about using FP in the kernel without a hard FP support. Paolo. -- [rtl] --- To unsubscribe: echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED] -- For more information on Real-Time Linux see: http://www.rtlinux.org/
