Russel wrote: >Oops...I've got either a 2.2.10, 2.2.12, and 2.2.14. Tried em all. > >I've got the binutils and gcc versions required in Documentation/Changes >( I even checked using the commands listed in that file) > >If the file builds just fine on your computer...has the file changed >between kernel versions? Has it changed at all in years? Would something >else have changed? Naturally, I must be missing something. I just wish I >knew what. ugh.
I think this might be your problem: --- kernel-source-2.2.10/arch/m68k/fpsp040.orig/bindec.S Fri Apr 26 05:12:35 1996 +++ kernel-source-2.2.10/arch/m68k/fpsp040/bindec.S Tue Oct 12 17:36:55 1999 @@ -484,7 +484,7 @@ fmovex (%a0),%fp0 |load X from memory fabsx %fp0 |use abs(X) tstw %d5 |LAMBDA is in lower word of d5 - bnes sc_mul |if neg (LAMBDA = 1), scale by mul + jne sc_mul |if neg (LAMBDA = 1), scale by mul fdivx %fp1,%fp0 |calculate X / SCALE -> Y to fp0 bras A10_st |branch to A10 The line # matches the one in your error message: >>Assembler Messages: >>Bindec.S :487 >>/usr/src/linux/arch/m68k/fpsp040 >>[Bindec.o] error 1 >>value of -512 too large for field of 1 bytes at 511 Judging from the error message I'd say that the source has changed to the extent that sc_mul is now out of reach of a short (8 bit) branch from that point and now requires a (32? bit) jump (sorry, my m68k assembler is a bit rusty these days). That fragment comes from a patch named binutils295-fpsp040-2.2.10.diff.gz included in kernel-patch-2.2.10-m68k_2.2.10-6.deb. I presume then that you are rolling your own kernel from the upstream sources. Looks to me like you might have better luck with the Debian kernel-source* and kernel-patch* packages. Frank