This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH, rtl-optimization]: Fix PR rtl-optimization/33846, ICE in trunc_int_for_mode, at explow.c:55
On 10/23/07, Eric Botcazou <ebotcazou@libertysurf.fr> wrote:
> > IIUC, this code also converts NOT (x) into XOR (x, -1).
>
> OK, but is this a win if you cannot further combine?
Actually, the code below is _not_ executed for either scalar or vector case:
/* We can't do this if we have (ashiftrt (xor)) and the
constant has its sign bit set in shift_mode. */
It looks that "[...] the case below" refers to some other part of the
code. Instead of tracking RTX through a maze of a couple of thousand
lines in combine.c or disabling an obscure simplification due to early
return, I simply fixed what looks like obvious oversight.
Uros.
- References:
- Re: [PATCH, rtl-optimization]: Fix PR rtl-optimization/33846, ICE in trunc_int_for_mode, at explow.c:55
- Re: [PATCH, rtl-optimization]: Fix PR rtl-optimization/33846, ICE in trunc_int_for_mode, at explow.c:55
- Re: [PATCH, rtl-optimization]: Fix PR rtl-optimization/33846, ICE in trunc_int_for_mode, at explow.c:55
- Re: [PATCH, rtl-optimization]: Fix PR rtl-optimization/33846, ICE in trunc_int_for_mode, at explow.c:55