On Wed, 21 Oct 2026 at 15:45, Richard Henderson <[email protected]> wrote: > > On 10/19/26 18:22, Peter Maydell wrote: > > In exynos4210_combiner_write(), we rule out "reg_n > 1" before we > > start doing anything as being an invalid guest access. The default > > case in the later "switch (reg_n)" is therefore unreachable, because > > it has explicit cases for 0 and 1. Use g_assert_not_reached() > > instead of a LOG_GUEST_ERROR. > > > > Signed-off-by: Peter Maydell <[email protected]> > > --- > > hw/intc/exynos4210_combiner.c | 4 +--- > > 1 file changed, 1 insertion(+), 3 deletions(-) > > > > diff --git a/hw/intc/exynos4210_combiner.c b/hw/intc/exynos4210_combiner.c > > index 2bdb9a0152c..80295ae6402 100644 > > --- a/hw/intc/exynos4210_combiner.c > > +++ b/hw/intc/exynos4210_combiner.c > > @@ -242,9 +242,7 @@ static void exynos4210_combiner_write(void *opaque, > > hwaddr offset, > > exynos4210_combiner_update(s, grp_quad_base_n + 3); > > break; > > default: > > - qemu_log_mask(LOG_GUEST_ERROR, > > - "exynos4210.combiner: unallowed write access at > > offset 0x" > > - HWADDR_FMT_plx "\n", offset); > > + g_assert_not_reached(); > > break; > > } > > } > > You could remove the break as well as unreachable.
Yes, good idea. -- PMM
