jack@cwi.nl (Jack Jansen) (11/29/90)
I'm currently porting Amoeba, our distributed OS, to the R[23]000. Currently, I'm using the DEC3100 and the SGI 3D/25 as the first machines to port to. Everything runs fine, with one exception: the floating point unit. I keep getting 'Unimplemented Operation' traps at quite unexpected times. After looking at the Ultrix sources I'm getting the impression that the kernel should be ready to execute any FP op, not just the unimplemented ones like sine or sqrt. Is this correct? If so, is there some document that tells me when the FPU fails, and what I can do about it? Or, even better, would someone happen to have one lying around? I'm not really thrilled by the idea of having to implement a full fp emulator.... -- -- Een volk dat voor tirannen zwicht | Oral: Jack Jansen zal meer dan lijf en goed verliezen | Internet: jack@cwi.nl dan dooft het licht | Uucp: hp4nl!cwi.nl!jack
msc@ramoth.esd.sgi.com (Mark Callow) (12/04/90)
In article <2628@charon.cwi.nl>, jack@cwi.nl (Jack Jansen) writes: |> Currently, I'm using the DEC3100 and the SGI 3D/25 as the first |> machines to port to. |> |> I keep getting 'Unimplemented Operation' traps at quite unexpected |> times. After looking at the Ultrix sources I'm getting the impression |> that the kernel should be ready to execute any FP op, not just the |> unimplemented ones like sine or sqrt. Is this correct? |> All IRIS 4D's have FPU's. Are you having this problem on the IRIS as well as the 3100 as your article implies or only on the 3100? -- From the TARDIS of Mark Callow msc@ramoth.sgi.com, ...{ames,decwrl}!sgi!msc "Spirits of genius are always opposed by mediocre minds" - Albert Einstein