Thiemo Seufer <[EMAIL PROTECTED]> writes: (in May 2007)
> CVSROOT: /sources/qemu > Module name: qemu > Changes by: Thiemo Seufer <ths> 07/05/13 16:35:36 > > Modified files: > target-sh4 : op.c > > Log message: > Remove unnecessary pointer magic in shift operations, by Magnus Damm. > > CVSWeb URLs: > http://cvs.savannah.gnu.org/viewcvs/qemu/target-sh4/op.c?cvsroot=qemu&r1=1.4&r2=1.5 > Hi, I'm just playing with qemu-sh4, and found that 'shar' seems to be doing a logical right shift, rather than an arithmetic right shift. I think that pointer magic was necessary after all. In the current version of target-sh4/op.c, op_shar_Rn() and op_shlr_Rn() are identical, which is surely wrong. void OPPROTO op_shar_Rn(void) { cond_t(env->gregs[PARAM1] & 1); env->gregs[PARAM1] >>= 1; RETURN(); } void OPPROTO op_shlr_Rn(void) { cond_t(env->gregs[PARAM1] & 1); env->gregs[PARAM1] >>= 1; RETURN(); } The behaviour is consistent with env->gregs[] being an array of unsigned ints, (I'm not familiar with qemu details...) and so to get an arithmetic shift, some kind of cast is necessary. dd -- Dave Denholm <[EMAIL PROTECTED]> http://www.esmertec.com
