public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Andreas Schwab <schwab@suse.de>
To: Eric Botcazou <ebotcazou@adacore.com>
Cc: gcc-patches@gcc.gnu.org
Subject: Re: [patch v2] Get rid of stack trampolines for nested functions (1/4)
Date: Thu, 23 Mar 2017 16:48:00 -0000	[thread overview]
Message-ID: <mvm37e4rlcm.fsf@suse.de> (raw)
In-Reply-To: <8467344.X3JSntxAug@arcturus.home> (Eric Botcazou's message of	"Sun, 04 Sep 2016 22:10:41 +0200")

On Sep 04 2016, Eric Botcazou <ebotcazou@adacore.com> wrote:

> Index: calls.c
> ===================================================================
> --- calls.c	(revision 239944)
> +++ calls.c	(working copy)
> @@ -183,17 +183,76 @@ static void restore_fixed_argument_area (rtx, rtx,
>  
>  rtx
>  prepare_call_address (tree fndecl_or_type, rtx funexp, rtx static_chain_value,
> -		      rtx *call_fusage, int reg_parm_seen, int sibcallp)
> +		      rtx *call_fusage, int reg_parm_seen, int flags)
>  {
>    /* Make a valid memory address and copy constants through pseudo-regs,
>       but not for a constant address if -fno-function-cse.  */
>    if (GET_CODE (funexp) != SYMBOL_REF)
> -    /* If we are using registers for parameters, force the
> -       function address into a register now.  */
> -    funexp = ((reg_parm_seen
> -	       && targetm.small_register_classes_for_mode_p (FUNCTION_MODE))
> -	      ? force_not_mem (memory_address (FUNCTION_MODE, funexp))
> -	      : memory_address (FUNCTION_MODE, funexp));
> +    {
> +      /* If it's an indirect call by descriptor, generate code to perform
> +	 runtime identification of the pointer and load the descriptor.  */
> +      if ((flags & ECF_BY_DESCRIPTOR) && !flag_trampolines)
> +	{
> +	  const int bit_val = targetm.calls.custom_function_descriptors;
> +	  rtx call_lab = gen_label_rtx ();
> +
> +	  gcc_assert (fndecl_or_type && TYPE_P (fndecl_or_type));
> +	  fndecl_or_type
> +	    = build_decl (UNKNOWN_LOCATION, FUNCTION_DECL, NULL_TREE,
> +			  fndecl_or_type);
> +	  DECL_STATIC_CHAIN (fndecl_or_type) = 1;
> +	  rtx chain = targetm.calls.static_chain (fndecl_or_type, false);
> +
> +	  /* Avoid long live ranges around function calls.  */
> +	  funexp = copy_to_mode_reg (Pmode, funexp);

That needs to use ptr_mode, not Pmode.

Andreas.

-- 
Andreas Schwab, SUSE Labs, schwab@suse.de
GPG Key fingerprint = 0196 BAD8 1CE9 1970 F4BE  1748 E4D4 88E3 0EEA B9D7
"And now for something completely different."

  parent reply	other threads:[~2017-03-23 16:47 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-09-04 20:10 [patch v2] Get rid of stack trampolines for nested functions (0/4) Eric Botcazou
2016-09-04 20:12 ` [patch v2] Get rid of stack trampolines for nested functions (1/4) Eric Botcazou
2016-09-12 19:41   ` Jeff Law
2016-09-12 19:45   ` Jeff Law
2016-12-05 20:52   ` Ian Lance Taylor
2016-12-05 21:29     ` Lynn A. Boger
2016-12-05 22:41       ` Ian Lance Taylor
2016-12-05 22:12     ` Eric Botcazou
2016-12-06 17:52       ` Eric Botcazou
2016-12-06 20:18         ` Ian Lance Taylor
2016-12-06 21:59           ` Lynn A. Boger
2016-12-06 22:26             ` Eric Botcazou
2016-12-07 13:38               ` Lynn A. Boger
2016-12-07 13:56                 ` Eric Botcazou
2016-12-07  7:23           ` Eric Botcazou
2017-03-23 16:48   ` Andreas Schwab [this message]
2017-03-28 17:01     ` Eric Botcazou
2017-03-29 10:05       ` Andreas Schwab
2017-03-29 14:05         ` Eric Botcazou
2017-03-29 14:15           ` Andreas Schwab
2016-09-04 20:14 ` [patch v2] Get rid of stack trampolines for nested functions (2/4) Eric Botcazou
2016-09-04 20:15 ` [patch v2] Get rid of stack trampolines for nested functions (3/4) Eric Botcazou
2016-09-05 10:52   ` Segher Boessenkool
2016-09-12 19:56   ` Jeff Law
2016-09-04 21:31 ` [patch v2] Get rid of stack trampolines for nested functions (4/4) Eric Botcazou
2016-10-16 20:29 ` [patch v2] Get rid of stack trampolines for nested functions (0/4) Eric Botcazou
2016-10-17 10:40   ` Andreas Schwab
2016-10-17 11:14     ` Eric Botcazou
2016-10-17 22:35     ` Eric Botcazou

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=mvm37e4rlcm.fsf@suse.de \
    --to=schwab@suse.de \
    --cc=ebotcazou@adacore.com \
    --cc=gcc-patches@gcc.gnu.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 read-only IMAP folder(s) and NNTP newsgroup(s).