In article <3E225192.48CEA9B5@cs.oberlin.edu>
hbent@cs.oberlin.edu wrote:

> This implies that GCC is generating 68030 code for my 68020 system, which is
> presumably being emulated?  You'll forgive my relative ignorance, as I'm mostly
> a MIPS person.  This seems like an awfully slow way to do things, though I've
> not yet had a chance to cross-build my own copy of GCC for speed comparison. 
> Can anyone offer any insight?

According to src/gnu/dist/toolchain/gcc/config/m68k/m68k.h,
it seems -m68030 is same with -m[c]68020:
---
#define TARGET_SWITCHES  \
  { { "68020", - (MASK_5200|MASK_68060|MASK_68040|MASK_68040_ONLY)},    \
    { "c68020", - (MASK_5200|MASK_68060|MASK_68040|MASK_68040_ONLY)},   \
    { "68020", (MASK_68020|MASK_BITFIELD)},                             \
    { "c68020", (MASK_68020|MASK_BITFIELD)},                            \
 :
    { "68030", - (MASK_5200|MASK_68060|MASK_68040|MASK_68040_ONLY)},    \
    { "68030", (MASK_68020|MASK_BITFIELD)},                             \
---
Izumi Tsutsui
tsutsui@ceres.dti.ne.jp