Re: MMX for kernel

Harald Koenig (koenig@tat.physik.uni-tuebingen.de)
Fri, 7 Feb 1997 15:06:28 +0100 (MET)

> If there were many, many proceses which used the FPU, one could imagine
> an advisory bit which meant "when task switching to this process,
> immediately restore my FPU registers, since I'm an intensive FPU
> user".

wouldn't only 2 FP-intensive (compute) processes be enough for this scenario ?

> This would avoid the hardware fault after each task switch to
> that process if there were many processes using the FPU. However, this
> hasn't been implemented yet, and in practice it hasn't seemed to have
> been a problem.

and what about runnung 2+ FP cmopute jobs on a dual-PPro200 (SMP) machine ?
how often do processes switch from one CPU to another needing to save/reload
FPU registers and hardware faults ?

any rough estimation how much benefit it might be to add such a flag
for FP-bound processes if Linux is only used as "compute server" ?

Harald

--
All SCSI disks will from now on                     ___       _____
be required to send an email notice                0--,|    /OOOOOOO\
24 hours prior to complete hardware failure!      <_/  /  /OOOOOOOOOOO\
                                                    \  \/OOOOOOOOOOOOOOO\
                                                      \ OOOOOOOOOOOOOOOOO|//
Harald Koenig,                                         \/\/\/\/\/\/\/\/\/
Inst.f.Theoret.Astrophysik                              //  /     \\  \
koenig@tat.physik.uni-tuebingen.de                     ^^^^^       ^^^^^