qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Alex Bennée" <alex.bennee@linaro.org>
To: Richard Henderson <rth@twiddle.net>
Cc: qemu-devel@nongnu.org, aurelien@aurel32.net
Subject: Re: [Qemu-devel] [PATCH 04/10] tcg: Use helper-gen.h in tcg-op.h
Date: Tue, 13 May 2014 10:34:19 +0100	[thread overview]
Message-ID: <8738ge81k4.fsf@linaro.org> (raw)
In-Reply-To: <1399938424-6703-5-git-send-email-rth@twiddle.net>


Richard Henderson <rth@twiddle.net> writes:

> No need to open-code the setup of the builtin helpers.
>
> Signed-off-by: Richard Henderson <rth@twiddle.net>
> ---
>  tcg/tcg-op.h | 175 +++++++++--------------------------------------------------
>  1 file changed, 25 insertions(+), 150 deletions(-)
>
> diff --git a/tcg/tcg-op.h b/tcg/tcg-op.h
> index 8560695..f4367a0 100644
> --- a/tcg/tcg-op.h
> +++ b/tcg/tcg-op.h
> @@ -24,6 +24,14 @@
>  #include "tcg.h"
>  #include "exec/helper-proto.h"
>  
> +static inline void tcg_gen_helperN(void *func, int flags, int sizemask,
> +                                   TCGArg ret, int nargs, TCGArg *args)
> +{
> +    tcg_gen_callN(&tcg_ctx, func, flags, sizemask, ret, nargs, args);
> +}
> +
> +#include "exec/helper-gen.h"
> +
<snip>
>  
> -/* A version of dh_sizemask from def-helper.h that doesn't rely on
> -   preprocessor magic.  */
> -static inline int tcg_gen_sizemask(int n, int is_64bit, int is_signed)
> -{
> -    return (is_64bit << n*2) | (is_signed << (n*2 + 1));
> -}
> -
> -/* helper calls */
> -static inline void tcg_gen_helperN(void *func, int flags, int sizemask,
> -                                   TCGArg ret, int nargs, TCGArg *args)
> -{
> -    tcg_gen_callN(&tcg_ctx, func, flags, sizemask, ret, nargs, args);
> -}
> -
<snip>
>  
>  static inline void tcg_gen_ld8u_i32(TCGv_i32 ret, TCGv_ptr arg2, tcg_target_long offset)
> @@ -708,12 +675,7 @@ static inline void tcg_gen_div_i32(TCGv_i32 ret, TCGv_i32 arg1, TCGv_i32 arg2)
>          tcg_gen_op5_i32(INDEX_op_div2_i32, ret, t0, arg1, t0, arg2);
>          tcg_temp_free_i32(t0);
>      } else {
> -        int sizemask = 0;
> -        /* Return value and both arguments are 32-bit and signed.  */
> -        sizemask |= tcg_gen_sizemask(0, 0, 1);
> -        sizemask |= tcg_gen_sizemask(1, 0, 1);
> -        sizemask |= tcg_gen_sizemask(2, 0, 1);
> -        tcg_gen_helper32(helper_div_i32, sizemask, ret, arg1, arg2);
> +        gen_helper_div_i32(ret, arg1, arg2);
>      }
>  }

I'm having trouble figuring out what the sizemask stuff was trying to
achieve and if it's required/covered after the change. Is the final
helper now the one defined as?

DEF(div_i32, 1, 2, 0, IMPL(TCG_TARGET_HAS_div_i32))

it's hard to follow the parameters of this compared to the target
helpers where the the argument sizes are a lot more explicit.

-- 
Alex Bennée

  reply	other threads:[~2014-05-13  9:34 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-05-12 23:46 [Qemu-devel] [PATCH 00/10] tcg cleanup and optimization Richard Henderson
2014-05-12 23:46 ` [Qemu-devel] [PATCH 01/10] tcg: Optimize brcond2 and setcond2 ne/eq Richard Henderson
2014-05-12 23:46 ` [Qemu-devel] [PATCH 02/10] tcg: Invert the inclusion of helper.h Richard Henderson
2014-05-13  9:10   ` Alex Bennée
2014-05-12 23:46 ` [Qemu-devel] [PATCH 03/10] tcg: Push tcg-runtime routines into exec/helper-* Richard Henderson
2014-05-13  9:20   ` Alex Bennée
2014-05-13 13:39     ` Richard Henderson
2014-05-12 23:46 ` [Qemu-devel] [PATCH 04/10] tcg: Use helper-gen.h in tcg-op.h Richard Henderson
2014-05-13  9:34   ` Alex Bennée [this message]
2014-05-13 13:44     ` Richard Henderson
2014-05-12 23:46 ` [Qemu-devel] [PATCH 05/10] tcg: Inline tcg_gen_helperN Richard Henderson
2014-05-13  9:34   ` Alex Bennée
2014-05-12 23:47 ` [Qemu-devel] [PATCH 06/10] tcg: Move side effects out of dh_sizemask Richard Henderson
2014-05-12 23:47 ` [Qemu-devel] [PATCH 07/10] tcg: Register the helper info struct rather than the name Richard Henderson
2014-05-13  9:36   ` Alex Bennée
2014-05-12 23:47 ` [Qemu-devel] [PATCH 08/10] tcg: Save flags and computed sizemask in TCGHelperInfo Richard Henderson
2014-05-13  9:38   ` Alex Bennée
2014-05-13 13:54     ` Richard Henderson
2014-05-12 23:47 ` [Qemu-devel] [PATCH 09/10] tcg: Remove sizemask and flags arguments to tcg_gen_callN Richard Henderson
2014-05-12 23:47 ` [Qemu-devel] [PATCH 10/10] tcg: Move size effects out of dh_arg Richard Henderson
2014-05-13  9:42   ` Alex Bennée

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=8738ge81k4.fsf@linaro.org \
    --to=alex.bennee@linaro.org \
    --cc=aurelien@aurel32.net \
    --cc=qemu-devel@nongnu.org \
    --cc=rth@twiddle.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).