qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Alexander Graf <agraf@suse.de>
To: "Andreas Färber" <afaerber@suse.de>
Cc: qemu-ppc@nongnu.org, qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [PATCH v2 for-1.1 4/4] tcg/ppc: Fix CONFIG_TCG_PASS_AREG0 mode
Date: Wed, 09 May 2012 12:27:49 +0200	[thread overview]
Message-ID: <4FAA46A5.8080505@suse.de> (raw)
In-Reply-To: <1336523290-6899-5-git-send-email-afaerber@suse.de>

On 05/09/2012 02:28 AM, Andreas Färber wrote:
> Adjust the tcg_out_qemu_{ld,st}() slow paths to pass AREG0 in r3,
> based on patches by malc.
>
> Also adjust the registers clobbered, based on patch by Alex.
>
> Signed-off-by: Andreas Färber<afaerber@suse.de>
> ---
>   tcg/ppc/tcg-target.c |   37 ++++++++++++++++++++++++++++++++++++-
>   1 files changed, 36 insertions(+), 1 deletions(-)
>
> diff --git a/tcg/ppc/tcg-target.c b/tcg/ppc/tcg-target.c
> index 20888e2..ca84aba 100644
> --- a/tcg/ppc/tcg-target.c
> +++ b/tcg/ppc/tcg-target.c
> @@ -244,9 +244,19 @@ static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str)
>           tcg_regset_set32(ct->u.regs, 0, 0xffffffff);
>           tcg_regset_reset_reg(ct->u.regs, TCG_REG_R3);
>           tcg_regset_reset_reg(ct->u.regs, TCG_REG_R4);
> +#ifdef CONFIG_TCG_PASS_AREG0
> +        tcg_regset_reset_reg(ct->u.regs, TCG_REG_R5);
> +#if TARGET_LONG_BITS == 64
> +        tcg_regset_reset_reg(ct->u.regs, TCG_REG_R6);
> +#ifdef TCG_TARGET_CALL_ALIGN_ARGS
> +        tcg_regset_reset_reg(ct->u.regs, TCG_REG_R7);
> +#endif
> +#endif
> +#else /* !AREG0 */
>   #if TARGET_LONG_BITS == 64
>           tcg_regset_reset_reg(ct->u.regs, TCG_REG_R5);
>   #endif
> +#endif
>           break;
>       case 'K':                   /* qemu_st[8..32] constraint */
>           ct->ct |= TCG_CT_REG;
> @@ -254,9 +264,19 @@ static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str)
>           tcg_regset_reset_reg(ct->u.regs, TCG_REG_R3);
>           tcg_regset_reset_reg(ct->u.regs, TCG_REG_R4);
>           tcg_regset_reset_reg(ct->u.regs, TCG_REG_R5);
> +#ifdef CONFIG_TCG_PASS_AREG0
> +        tcg_regset_reset_reg(ct->u.regs, TCG_REG_R6);
> +#if TARGET_LONG_BITS == 64
> +        tcg_regset_reset_reg(ct->u.regs, TCG_REG_R7);
> +#ifdef TCG_TARGET_CALL_ALIGN_ARGS
> +        tcg_regset_reset_reg(ct->u.regs, TCG_REG_R8);
> +#endif
> +#endif
> +#else /* !AREG0 */
>   #if TARGET_LONG_BITS == 64
>           tcg_regset_reset_reg(ct->u.regs, TCG_REG_R6);
>   #endif
> +#endif
>           break;
>       case 'M':                   /* qemu_st64 constraint */
>           ct->ct |= TCG_CT_REG;
> @@ -266,6 +286,12 @@ static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str)
>           tcg_regset_reset_reg(ct->u.regs, TCG_REG_R5);
>           tcg_regset_reset_reg(ct->u.regs, TCG_REG_R6);
>           tcg_regset_reset_reg(ct->u.regs, TCG_REG_R7);
> +#if defined(CONFIG_TCG_PASS_AREG0)
> +        tcg_regset_reset_reg(ct->u.regs, TCG_REG_R8);
> +#ifdef TCG_TARGET_CALL_ALIGN_ARGS
> +        tcg_regset_reset_reg(ct->u.regs, TCG_REG_R9);
> +#endif
> +#endif
>           break;
>   #else
>       case 'L':
> @@ -512,7 +538,6 @@ static void tcg_out_call (TCGContext *s, tcg_target_long arg, int const_arg)
>   #include "../../softmmu_defs.h"
>
>   #ifdef CONFIG_TCG_PASS_AREG0
> -#error CONFIG_TCG_PASS_AREG0 is not supported
>   /* helper signature: helper_ld_mmu(CPUState *env, target_ulong addr,
>      int mmu_idx) */
>   static const void * const qemu_ld_helpers[4] = {
> @@ -617,7 +642,12 @@ static void tcg_out_qemu_ld (TCGContext *s, const TCGArg *args, int opc)
>   #endif
>
>       /* slow path */
> +#ifdef CONFIG_TCG_PASS_AREG0
> +    tcg_out_mov (s, TCG_TYPE_I32, 3, TCG_AREG0);
> +    ir = 4;
> +#else
>       ir = 3;
> +#endif

I liked your "start ir with 3 and only ir++ from then on" way of doing 
this better.

ir = 3;
#ifdef ...
    tcg_out_mov(,,, ir++);
#endif

>   #if TARGET_LONG_BITS == 32
>       tcg_out_mov (s, TCG_TYPE_I32, ir++, addr_reg);
>   #else
> @@ -816,7 +846,12 @@ static void tcg_out_qemu_st (TCGContext *s, const TCGArg *args, int opc)
>   #endif
>
>       /* slow path */
> +#ifdef CONFIG_TCG_PASS_AREG0
> +    tcg_out_mov (s, TCG_TYPE_I32, 3, TCG_AREG0);
> +    ir = 4;
> +#else
>       ir = 3;
> +#endif

Same here


Alex

  parent reply	other threads:[~2012-05-09 10:27 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-05-09  0:28 [Qemu-devel] [PATCH v2 for-1.1 0/4] tcg/ppc: AREG0 support and Darwin fixes Andreas Färber
2012-05-09  0:28 ` [Qemu-devel] [PATCH v2 for-1.1 1/4] tcg/ppc: Do not overwrite lower address word on Darwin and AIX Andreas Färber
2012-05-09  0:28 ` [Qemu-devel] [PATCH v2 for-1.1 2/4] tcg/ppc: Don't hardcode register numbers Andreas Färber
2012-05-09  0:28 ` [Qemu-devel] [PATCH v2 for-1.1 3/4] tcg/ppc: Clobber r5 for 64-bit qemu_ld Andreas Färber
2012-05-09  0:28 ` [Qemu-devel] [PATCH v2 for-1.1 4/4] tcg/ppc: Fix CONFIG_TCG_PASS_AREG0 mode Andreas Färber
2012-05-09  0:43   ` malc
2012-05-09  1:03     ` Andreas Färber
2012-05-09 10:30     ` Alexander Graf
2012-05-09 10:27   ` Alexander Graf [this message]
2012-05-09 10:46     ` Andreas Färber
2012-05-09  0:33 ` [Qemu-devel] [PATCH v2 for-1.1 0/4] tcg/ppc: AREG0 support and Darwin fixes malc
2012-05-09  0:55   ` Andreas Färber
2012-05-09 10:31   ` Alexander Graf

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=4FAA46A5.8080505@suse.de \
    --to=agraf@suse.de \
    --cc=afaerber@suse.de \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-ppc@nongnu.org \
    /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).