public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Eric Botcazou <ebotcazou@adacore.com>
To: gcc-patches@gcc.gnu.org
Subject: Small cleanup in RTL expansion
Date: Tue, 02 Jul 2019 09:45:00 -0000	[thread overview]
Message-ID: <3413974.qYVSK5bheN@polaris> (raw)

[-- Attachment #1: Type: text/plain, Size: 665 bytes --]

pass_expand::execute does the same job as commit_edge_insertions but manually, 
except for a special case if the entry block has a single successor.  But we 
can extract the special case and just call commit_edge_insertions afterwards.

Tested on x86_64-suse-linux, applied on the mainline.


2019-07-02  Eric Botcazou  <ebotcazou@adacore.com>

	* cfgexpand.c (pass_expand::execute): Deal specially with instructions
 	to be inserted on single successor edge of the entry block.  Then call
	commit_edge_insertions instead of inserting the instructions manually.
	* cfgrtl.c (commit_edge_insertions): Do not verify flow info during
	RTL expansion.

-- 
Eric Botcazou

[-- Attachment #2: p.diff --]
[-- Type: text/x-patch, Size: 2455 bytes --]

Index: cfgexpand.c
===================================================================
--- cfgexpand.c	(revision 272891)
+++ cfgexpand.c	(working copy)
@@ -6576,37 +6576,27 @@ pass_expand::execute (function *fun)
      split edges which edge insertions might do.  */
   rebuild_jump_labels (get_insns ());
 
-  FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (fun),
-		  EXIT_BLOCK_PTR_FOR_FN (fun), next_bb)
+  /* If we have a single successor to the entry block, put the pending insns
+     after parm birth, but before NOTE_INSNS_FUNCTION_BEG.  */
+  if (single_succ_p (ENTRY_BLOCK_PTR_FOR_FN (fun)))
     {
-      edge e;
-      edge_iterator ei;
-      for (ei = ei_start (bb->succs); (e = ei_safe_edge (ei)); )
+      edge e = single_succ_edge (ENTRY_BLOCK_PTR_FOR_FN (fun));
+      if (e->insns.r)
 	{
-	  if (e->insns.r)
-	    {
-	      rebuild_jump_labels_chain (e->insns.r);
-	      /* Put insns after parm birth, but before
-		 NOTE_INSNS_FUNCTION_BEG.  */
-	      if (e->src == ENTRY_BLOCK_PTR_FOR_FN (fun)
-		  && single_succ_p (ENTRY_BLOCK_PTR_FOR_FN (fun)))
-		{
-		  rtx_insn *insns = e->insns.r;
-		  e->insns.r = NULL;
-		  if (NOTE_P (parm_birth_insn)
-		      && NOTE_KIND (parm_birth_insn) == NOTE_INSN_FUNCTION_BEG)
-		    emit_insn_before_noloc (insns, parm_birth_insn, e->dest);
-		  else
-		    emit_insn_after_noloc (insns, parm_birth_insn, e->dest);
-		}
-	      else
-		commit_one_edge_insertion (e);
-	    }
+	  rtx_insn *insns = e->insns.r;
+	  e->insns.r = NULL;
+	  rebuild_jump_labels_chain (insns);
+	  if (NOTE_P (parm_birth_insn)
+	      && NOTE_KIND (parm_birth_insn) == NOTE_INSN_FUNCTION_BEG)
+	    emit_insn_before_noloc (insns, parm_birth_insn, e->dest);
 	  else
-	    ei_next (&ei);
+	    emit_insn_after_noloc (insns, parm_birth_insn, e->dest);
 	}
     }
 
+  /* Otherwise, as well as for other edges, take the usual way.  */
+  commit_edge_insertions ();
+
   /* We're done expanding trees to RTL.  */
   currently_expanding_to_rtl = 0;
 
Index: cfgrtl.c
===================================================================
--- cfgrtl.c	(revision 272891)
+++ cfgrtl.c	(working copy)
@@ -2105,7 +2105,8 @@ commit_edge_insertions (void)
      which will be done by fixup_partitions.  */
   fixup_partitions ();
 
-  checking_verify_flow_info ();
+  if (!currently_expanding_to_rtl)
+    checking_verify_flow_info ();
 
   FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun),
 		  EXIT_BLOCK_PTR_FOR_FN (cfun), next_bb)

                 reply	other threads:[~2019-07-02  9:45 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=3413974.qYVSK5bheN@polaris \
    --to=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).