public inbox for gas2@sourceware.org
 help / color / mirror / Atom feed
From: Ian Lance Taylor <ian@cygnus.com>
To: dm@sgi.com
Cc: gas2@cygnus.com
Subject: Re: BFD_ASSERT failure in elf32-mips.c with possible fix
Date: Fri, 09 Aug 1996 09:44:00 -0000	[thread overview]
Message-ID: <199608091644.MAA07050@sanguine.cygnus.com> (raw)
In-Reply-To: <199608091345.GAA14227@neteng.engr.sgi.com>

   Date: Fri, 9 Aug 1996 06:45:58 -0700
   From: "David S. Miller" <dm@neteng.engr.sgi.com>

   The following c-torture case:

   extern int abort();
   typedef int (*frob)();
   frob f[] = {abort};
   main()
   {
     exit(0);
   }

Thanks for the note.  I think this patch is more nearly correct, since
it won't waste space on an unnecessary .rel.dyn section.

I don't consider the dynamic linking code in elf32-mips.c to be
particularly robust.  It doesn't even pass the ld testsuite on Irix 5.

Ian

Index: elf32-mips.c
===================================================================
RCS file: /cvs/cvsfiles/devo/bfd/elf32-mips.c,v
retrieving revision 1.84
diff -u -r1.84 elf32-mips.c
--- elf32-mips.c	1996/07/24 23:11:13	1.84
+++ elf32-mips.c	1996/08/09 16:42:21
@@ -5107,37 +5107,38 @@
 	  if ((info->shared || h != NULL)
 	      && (sec->flags & SEC_ALLOC) != 0)
 	    {
+	      if (sreloc == NULL)
+		{
+		  const char *name = ".rel.dyn";
+
+		  sreloc = bfd_get_section_by_name (dynobj, name);
+		  if (sreloc == NULL)
+		    {
+		      sreloc = bfd_make_section (dynobj, name);
+		      if (sreloc == NULL
+			  || ! bfd_set_section_flags (dynobj, sreloc,
+						      (SEC_ALLOC
+						       | SEC_LOAD
+						       | SEC_HAS_CONTENTS
+						       | SEC_IN_MEMORY
+						       | SEC_READONLY))
+			  || ! bfd_set_section_alignment (dynobj, sreloc,
+							  4))
+			return false;
+		    }
+		}
 	      if (info->shared)
 		{
 		  /* When creating a shared object, we must copy these
 		     reloc types into the output file as R_MIPS_REL32
-		     relocs.  We create the .rel.dyn reloc section in
-		     dynobj and make room for this reloc.  */
-		  if (sreloc == NULL)
+		     relocs.  We make room for this reloc in the
+		     .rel.dyn reloc section */
+		  if (sreloc->_raw_size == 0)
 		    {
-		      const char *name = ".rel.dyn";
-
-		      sreloc = bfd_get_section_by_name (dynobj, name);
-		      if (sreloc == NULL)
-			{
-			  sreloc = bfd_make_section (dynobj, name);
-			  if (sreloc == NULL
-			      || ! bfd_set_section_flags (dynobj, sreloc,
-							  (SEC_ALLOC
-							   | SEC_LOAD
-							   | SEC_HAS_CONTENTS
-							   | SEC_IN_MEMORY
-							   | SEC_READONLY))
-			      || ! bfd_set_section_alignment (dynobj, sreloc,
-							      4))
-			    return false;
-
-			  /* Add a null element. */
-			  sreloc->_raw_size += sizeof (Elf32_External_Rel);
-			  ++sreloc->reloc_count;
-			}
+		      /* Add a null element. */
+		      sreloc->_raw_size += sizeof (Elf32_External_Rel);
+		      ++sreloc->reloc_count;
 		    }
-
 		  sreloc->_raw_size += sizeof (Elf32_External_Rel);
 		}
 	      else
@@ -5213,6 +5214,12 @@
       s = bfd_get_section_by_name (dynobj, ".rel.dyn");
       BFD_ASSERT (s != NULL);
 
+      if (s->_raw_size == 0)
+	{
+	  /* Make room for a null element. */
+	  s->_raw_size += sizeof (Elf32_External_Rel);
+	  ++s->reloc_count;
+	}
       s->_raw_size += hmips->mips_32_relocs * sizeof (Elf32_External_Rel);
     }
 
@@ -6085,7 +6092,7 @@
 
     /* Clean up a first relocation in .rel.dyn.  */
     s = bfd_get_section_by_name (dynobj, ".rel.dyn");
-    if (s != NULL)
+    if (s != NULL && s->_raw_size > 0)
       memset (s->contents, 0, sizeof (Elf32_External_Rel));
   }
 



      reply	other threads:[~1996-08-09  9:44 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
1996-08-09  6:46 David S. Miller
1996-08-09  9:44 ` Ian Lance Taylor [this message]

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=199608091644.MAA07050@sanguine.cygnus.com \
    --to=ian@cygnus.com \
    --cc=dm@sgi.com \
    --cc=gas2@cygnus.com \
    /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).