Alex Bennée <alex.ben...@linaro.org> writes:
> Peter Maydell <peter.mayd...@linaro.org> writes: > >> On Mon, 14 Oct 2019 at 12:38, Alex Bennée <alex.ben...@linaro.org> wrote: >>> >>> From: "Emilio G. Cota" <c...@braap.org> >>> >>> We don't bother with replicating the fast path (tlb_hit) of the old >>> cpu_ldst helpers as it has no measurable effect on performance. This >>> probably indicates we should consider flattening the whole set of >>> helpers but that is out of scope for this change. >>> >>> Suggested-by: Richard Henderson <richard.hender...@linaro.org> >>> Signed-off-by: Emilio G. Cota <c...@braap.org> >>> [AJB: directly plumb into softmmu/user helpers] >>> Signed-off-by: Alex Bennée <alex.ben...@linaro.org> >>> >>> diff --git a/tcg/tcg.h b/tcg/tcg.h >>> index a38659ea5b..302533b463 100644 >>> --- a/tcg/tcg.h >>> +++ b/tcg/tcg.h >>> @@ -1317,6 +1317,7 @@ uint64_t helper_be_ldq_cmmu(CPUArchState *env, >>> target_ulong addr, >>> # define helper_ret_stl_mmu helper_be_stl_mmu >>> # define helper_ret_stq_mmu helper_be_stq_mmu >>> # define helper_ret_ldw_cmmu helper_be_ldw_cmmu >>> +# define helper_ret_lduw_cmmu helper_be_ldw_cmmu >>> # define helper_ret_ldl_cmmu helper_be_ldl_cmmu >>> # define helper_ret_ldq_cmmu helper_be_ldq_cmmu >>> #else >>> @@ -1330,6 +1331,7 @@ uint64_t helper_be_ldq_cmmu(CPUArchState *env, >>> target_ulong addr, >>> # define helper_ret_stl_mmu helper_le_stl_mmu >>> # define helper_ret_stq_mmu helper_le_stq_mmu >>> # define helper_ret_ldw_cmmu helper_le_ldw_cmmu >>> +# define helper_ret_lduw_cmmu helper_le_ldw_cmmu >>> # define helper_ret_ldl_cmmu helper_le_ldl_cmmu >>> # define helper_ret_ldq_cmmu helper_le_ldq_cmmu >>> #endif >> >> This looks odd. Why is it ok to define a 'lduw' helper >> as the 'ldw' cmmu helper ? One ought to be sign >> extending and the other not... > > This was attempting to make things line up between the softmmu helpers > and the user-mode ld*_p helpers that we need to expand to. I'm not sure > a sign extending loader even makes sense for code load anyway. That last bit is not true as sign extending helpers are used for loading sign-extended immediate values. > >> >> thanks >> -- PMM -- Alex Bennée
