public inbox for gcc-cvs@sourceware.org
help / color / mirror / Atom feed
From: Richard Biener <rguenth@gcc.gnu.org>
To: gcc-cvs@gcc.gnu.org
Subject: [gcc r13-2263] Use reachability analysis to improve uninit diagnostic
Date: Tue, 30 Aug 2022 08:37:10 +0000 (GMT)	[thread overview]
Message-ID: <20220830083710.516B73AA8008@sourceware.org> (raw)

https://gcc.gnu.org/g:f71abacfed170852ab25485e5c52f8d5aad7c2a0

commit r13-2263-gf71abacfed170852ab25485e5c52f8d5aad7c2a0
Author: Richard Biener <rguenther@suse.de>
Date:   Fri Aug 19 15:11:14 2022 +0200

    Use reachability analysis to improve uninit diagnostic
    
    This patch does what the comment in uninit diagnostic suggests.
    When the value-numbering run done without optimizing figures there's
    a fallthru path, consider blocks on it as always executed.
    
            * tree-ssa-uninit.cc (warn_uninitialized_vars): Pre-compute
            the set of fallthru reachable blocks from function entry
            and use that to determine wlims.always_executed.

Diff:
---
 gcc/tree-ssa-uninit.cc | 47 ++++++++++++++++++++++++++++++++++++++---------
 1 file changed, 38 insertions(+), 9 deletions(-)

diff --git a/gcc/tree-ssa-uninit.cc b/gcc/tree-ssa-uninit.cc
index 0bd567f237c..c25fbe6381e 100644
--- a/gcc/tree-ssa-uninit.cc
+++ b/gcc/tree-ssa-uninit.cc
@@ -988,10 +988,43 @@ warn_uninitialized_vars (bool wmaybe_uninit)
   wlimits wlims = { };
   wlims.wmaybe_uninit = wmaybe_uninit;
 
-  gimple_stmt_iterator gsi;
-  basic_block bb;
+  auto_bb_flag ft_reachable (cfun);
+
+  /* Mark blocks that are always executed when we ignore provably
+     not executed edges.  */
+  basic_block bb = single_succ (ENTRY_BLOCK_PTR_FOR_FN (cfun));
+  while (!(bb->flags & ft_reachable))
+    {
+      bb->flags |= ft_reachable;
+      /* Find a single executable edge.  */
+      edge_iterator ei;
+      edge e, ee = NULL;
+      FOR_EACH_EDGE (e, ei, bb->succs)
+	if (e->flags & EDGE_EXECUTABLE)
+	  {
+	    if (!ee)
+	      ee = e;
+	    else
+	      {
+		ee = NULL;
+		break;
+	      }
+	  }
+      if (ee)
+	bb = ee->dest;
+      else
+	{
+	  bb = get_immediate_dominator (CDI_POST_DOMINATORS, bb);
+	  if (!bb || bb->index == EXIT_BLOCK)
+	    break;
+	}
+    }
+
   FOR_EACH_BB_FN (bb, cfun)
     {
+      wlims.always_executed = (bb->flags & ft_reachable);
+      bb->flags &= ~ft_reachable;
+
       edge_iterator ei;
       edge e;
       FOR_EACH_EDGE (e, ei, bb->preds)
@@ -1002,14 +1035,10 @@ warn_uninitialized_vars (bool wmaybe_uninit)
       if (!e)
 	continue;
 
-      basic_block succ = single_succ (ENTRY_BLOCK_PTR_FOR_FN (cfun));
-      /* ???  This could be improved when we use a greedy walk and have
-	 some edges marked as not executable.  */
-      wlims.always_executed = dominated_by_p (CDI_POST_DOMINATORS, succ, bb);
-
       if (wlims.always_executed)
 	warn_uninit_phi_uses (bb);
 
+      gimple_stmt_iterator gsi;
       for (gsi = gsi_start_bb (bb); !gsi_end_p (gsi); gsi_next (&gsi))
 	{
 	  gimple *stmt = gsi_stmt (gsi);
@@ -1030,7 +1059,7 @@ warn_uninitialized_vars (bool wmaybe_uninit)
 	  FOR_EACH_SSA_USE_OPERAND (use_p, stmt, op_iter, SSA_OP_USE)
 	    {
 	      /* BIT_INSERT_EXPR first operand should not be considered
-	         a use for the purpose of uninit warnings.  */
+		 a use for the purpose of uninit warnings.  */
 	      if (gassign *ass = dyn_cast <gassign *> (stmt))
 		{
 		  if (gimple_assign_rhs_code (ass) == BIT_INSERT_EXPR
@@ -1041,7 +1070,7 @@ warn_uninitialized_vars (bool wmaybe_uninit)
 	      if (wlims.always_executed)
 		warn_uninit (OPT_Wuninitialized, use,
 			     SSA_NAME_VAR (use), stmt);
-	      else if (wmaybe_uninit)
+	      else if (wlims.wmaybe_uninit)
 		warn_uninit (OPT_Wmaybe_uninitialized, use,
 			     SSA_NAME_VAR (use), stmt);
 	    }

                 reply	other threads:[~2022-08-30  8:37 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=20220830083710.516B73AA8008@sourceware.org \
    --to=rguenth@gcc.gnu.org \
    --cc=gcc-cvs@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).