[Bug lto/83201] [7/8 Regression] SPEC CPU2017 505.mcf_r produces incorrect output when built with -flto and FDO

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83201

--- Comment #13 from Richard Biener <rguenth at gcc dot gnu.org> ---
Likely invalid.  spec_qsort is full of alias violations.  We sort

typedef struct basket
{
    arc_t *a;
    cost_t cost;
    cost_t abs_cost;
    LONG number;
} BASKET;

and spec_qsort does stuff like

        if (n < 7) {
                for (pm = (char *)a + es; pm < (char *)a + n * es; pm += es)
                        for (pl = pm;
                             pl > (char *)a && cmp(pl - es, pl) > 0;
                             pl -= es)
                                swap(pl, pl - es);
                return;

with

#define swap(a, b)                              \
        if (swaptype_long == 0) {               \
                long t = *(long *)(a);          \
                *(long *)(a) = *(long *)(b);    \
                *(long *)(b) = t;               \
        } else if (swaptype_int == 0) {         \
                int t = *(int *)(a);            \
                *(int *)(a) = *(int *)(b);      \
                *(int *)(b) = t;                \
        } else                                  \
                swapfunc((char *)a, (char *)b, es, swaptype_long, swaptype_int)

eh... (no, swapfunc isn't any better).

Anybody up to reporting this to SPEC?