Re: [Qemu-devel] [RFC v6 10/14] softmmu: Simplify helper_*_st_name, wrap unaligned code

["Alex Bennée" <alex.bennee@linaro.org>]

Alvise Rigo <a.rigo@virtualopensystems.com> writes:

> Attempting to simplify the helper_*_st_name, wrap the
> do_unaligned_access code into an inline function.
> Remove also the goto statement.
>
> Suggested-by: Jani Kokkonen <jani.kokkonen@huawei.com>
> Suggested-by: Claudio Fontana <claudio.fontana@huawei.com>
> Signed-off-by: Alvise Rigo <a.rigo@virtualopensystems.com>
> ---
>  softmmu_template.h | 96 ++++++++++++++++++++++++++++++++++--------------------
>  1 file changed, 60 insertions(+), 36 deletions(-)
>
> diff --git a/softmmu_template.h b/softmmu_template.h
> index d3d5902..92f92b1 100644
> --- a/softmmu_template.h
> +++ b/softmmu_template.h
> @@ -370,6 +370,32 @@ static inline void glue(io_write, SUFFIX)(CPUArchState *env,
>                                   iotlbentry->attrs);
>  }
>
> +static inline void glue(helper_le_st_name, _do_unl_access)(CPUArchState *env,
> +                                                           DATA_TYPE val,
> +                                                           target_ulong addr,
> +                                                           TCGMemOpIdx oi,
> +                                                           unsigned mmu_idx,
> +                                                           uintptr_t retaddr)
> +{
> +    int i;
> +
> +    if ((get_memop(oi) & MO_AMASK) == MO_ALIGN) {
> +        cpu_unaligned_access(ENV_GET_CPU(env), addr, MMU_DATA_STORE,
> +                             mmu_idx, retaddr);
> +    }
> +    /* XXX: not efficient, but simple */
> +    /* Note: relies on the fact that tlb_fill() does not remove the
> +     * previous page from the TLB cache.  */
> +    for (i = DATA_SIZE - 1; i >= 0; i--) {
> +        /* Little-endian extract.  */
> +        uint8_t val8 = val >> (i * 8);
> +        /* Note the adjustment at the beginning of the function.
> +           Undo that for the recursion.  */
> +        glue(helper_ret_stb, MMUSUFFIX)(env, addr + i, val8,
> +                                        oi, retaddr + GETPC_ADJ);
> +    }
> +}
> +
>  void helper_le_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
>                         TCGMemOpIdx oi, uintptr_t retaddr)
>  {
> @@ -433,7 +459,8 @@ void helper_le_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
>              return;
>          } else {
>              if ((addr & (DATA_SIZE - 1)) != 0) {
> -                goto do_unaligned_access;
> +                glue(helper_le_st_name, _do_unl_access)(env, val, addr, mmu_idx,
> +                                                        oi, retaddr);

I've just noticed this drops and implicit return. However I'm seeing if
I can put together an RFC cleanup patch set for softmmu based on these
plus a few other clean-ups. I'll CC when done.

>              }
>              iotlbentry = &env->iotlb[mmu_idx][index];
>
> @@ -449,23 +476,8 @@ void helper_le_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
>      if (DATA_SIZE > 1
>          && unlikely((addr & ~TARGET_PAGE_MASK) + DATA_SIZE - 1
>                       >= TARGET_PAGE_SIZE)) {
> -        int i;
> -    do_unaligned_access:
> -        if ((get_memop(oi) & MO_AMASK) == MO_ALIGN) {
> -            cpu_unaligned_access(ENV_GET_CPU(env), addr, MMU_DATA_STORE,
> -                                 mmu_idx, retaddr);
> -        }
> -        /* XXX: not efficient, but simple */
> -        /* Note: relies on the fact that tlb_fill() does not remove the
> -         * previous page from the TLB cache.  */
> -        for (i = DATA_SIZE - 1; i >= 0; i--) {
> -            /* Little-endian extract.  */
> -            uint8_t val8 = val >> (i * 8);
> -            /* Note the adjustment at the beginning of the function.
> -               Undo that for the recursion.  */
> -            glue(helper_ret_stb, MMUSUFFIX)(env, addr + i, val8,
> -                                            oi, retaddr + GETPC_ADJ);
> -        }
> +        glue(helper_le_st_name, _do_unl_access)(env, val, addr, oi, mmu_idx,
> +                                                retaddr);
>          return;
>      }
>
> @@ -485,6 +497,32 @@ void helper_le_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
>  }
>
>  #if DATA_SIZE > 1
> +static inline void glue(helper_be_st_name, _do_unl_access)(CPUArchState *env,
> +                                                           DATA_TYPE val,
> +                                                           target_ulong addr,
> +                                                           TCGMemOpIdx oi,
> +                                                           unsigned mmu_idx,
> +                                                           uintptr_t retaddr)
> +{
> +    int i;
> +
> +    if ((get_memop(oi) & MO_AMASK) == MO_ALIGN) {
> +        cpu_unaligned_access(ENV_GET_CPU(env), addr, MMU_DATA_STORE,
> +                             mmu_idx, retaddr);
> +    }
> +    /* XXX: not efficient, but simple */
> +    /* Note: relies on the fact that tlb_fill() does not remove the
> +     * previous page from the TLB cache.  */
> +    for (i = DATA_SIZE - 1; i >= 0; i--) {
> +        /* Big-endian extract.  */
> +        uint8_t val8 = val >> (((DATA_SIZE - 1) * 8) - (i * 8));
> +        /* Note the adjustment at the beginning of the function.
> +           Undo that for the recursion.  */
> +        glue(helper_ret_stb, MMUSUFFIX)(env, addr + i, val8,
> +                                        oi, retaddr + GETPC_ADJ);
> +    }
> +}
> +
>  void helper_be_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
>                         TCGMemOpIdx oi, uintptr_t retaddr)
>  {
> @@ -548,7 +586,8 @@ void helper_be_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
>              return;
>          } else {
>              if ((addr & (DATA_SIZE - 1)) != 0) {
> -                goto do_unaligned_access;
> +                glue(helper_be_st_name, _do_unl_access)(env, val, addr, mmu_idx,
> +                                                        oi, retaddr);
>              }
>              iotlbentry = &env->iotlb[mmu_idx][index];
>
> @@ -564,23 +603,8 @@ void helper_be_st_name(CPUArchState *env, target_ulong addr, DATA_TYPE val,
>      if (DATA_SIZE > 1
>          && unlikely((addr & ~TARGET_PAGE_MASK) + DATA_SIZE - 1
>                       >= TARGET_PAGE_SIZE)) {
> -        int i;
> -    do_unaligned_access:
> -        if ((get_memop(oi) & MO_AMASK) == MO_ALIGN) {
> -            cpu_unaligned_access(ENV_GET_CPU(env), addr, MMU_DATA_STORE,
> -                                 mmu_idx, retaddr);
> -        }
> -        /* XXX: not efficient, but simple */
> -        /* Note: relies on the fact that tlb_fill() does not remove the
> -         * previous page from the TLB cache.  */
> -        for (i = DATA_SIZE - 1; i >= 0; i--) {
> -            /* Big-endian extract.  */
> -            uint8_t val8 = val >> (((DATA_SIZE - 1) * 8) - (i * 8));
> -            /* Note the adjustment at the beginning of the function.
> -               Undo that for the recursion.  */
> -            glue(helper_ret_stb, MMUSUFFIX)(env, addr + i, val8,
> -                                            oi, retaddr + GETPC_ADJ);
> -        }
> +        glue(helper_be_st_name, _do_unl_access)(env, val, addr, oi, mmu_idx,
> +                                                retaddr);
>          return;
>      }


--
Alex Bennée