On 06/13/2018 02:55 AM, Alex Bennée wrote:
> +static void sve_dump_preg_diff(FILE *f, int vq,
> + uint16_t const (*p1)[SVE_VQ_MAX],
> + uint16_t const (*p2)[SVE_VQ_MAX])
> +{
> + int q;
> +
> + for (q = 0; q < vq; q++) {
> + fprintf(f, "%#04x", *p1[q]);
> + }
> + fprintf(f, " vs ");
> + for (q = 0; q < vq; q++) {
> + fprintf(f, "%#04x", *p2[q]);
%# adds 0x into every 16-bit unit, so for vq=2 we get
0xffff0xffff
Emit the 0x separately to start?
> + for (i = 0; i < SVE_NUM_ZREGS; i++) {
> + if (!sve_zreg_is_eq(m, a, i)) {
> + int q;
> + char *pad="";
> + fprintf(f, " Z%2d : ", i);
%-2d? %02d?
> + for (q = 0; q < sve_vq_from_vl(ms->vl); q++) {
> + if (ms->zregs[i][q] != as->zregs[i][q]) {
> + fprintf(f, "%sq%02d: %016" PRIx64 "%016" PRIx64
> + " vs %016" PRIx64 "%016" PRIx64"\n", pad, q,
Actually, another thing that has annoyed me in the past,
but apparently not quite enough to actually fix, is the
fact that reginfo_dump and reginfo_dump_mismatch have a
slightly different format for Zregs.
It's probably worth splitting those bits out to helper
functions so that they must match.
r~
