public inbox for binutils@sourceware.org
 help / color / mirror / Atom feed
From: Alan Modra <amodra@gmail.com>
To: binutils@sourceware.org
Subject: setup_archive parameter mismatch
Date: Thu, 19 Mar 2020 13:55:37 +1030	[thread overview]
Message-ID: <20200319032537.GA4583@bubble.grove.modra.org> (raw)

readelf.c and elfedit.c both define BFD64 in order to support
processing of ELFCLASS64 objects.  Or at least that's what the comment
says.  This means I should not have used bfd_size_type to pass file
size to a function in elfcomm.c which also includes bfd.h but doesn't
define BFD64.  When you build and test on a 64-bit host, the parameter
mismatch doesn't show up, but it does if bfd/ is compiled on a 32-bit
host without 64-bit BFD support.

	* elfcomm.c (setup_archive): Make file_size an off_t.  Comment.
	* elfcomm.h (setup_archive): Update prototype.

diff --git a/binutils/elfcomm.c b/binutils/elfcomm.c
index e5c6a32c6d..b0fbf57e1f 100644
--- a/binutils/elfcomm.c
+++ b/binutils/elfcomm.c
@@ -21,6 +21,11 @@
    Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston, MA
    02110-1301, USA.  */
 
+/* Do not use BFD types in this file that differ in size depending on
+   whether BFD64 is defined.  Functions in this file are used by
+   readelf.c and elfedit.c which define BFD64, and by objdump.c which
+   doesn't.  */
+
 #include "sysdep.h"
 #include "libiberty.h"
 #include "filenames.h"
@@ -607,7 +612,7 @@ process_archive_index_and_symbols (struct archive_info *  arch,
 
 int
 setup_archive (struct archive_info *arch, const char *file_name,
-	       FILE *file, bfd_size_type file_size,
+	       FILE *file, off_t file_size,
 	       bfd_boolean is_thin_archive, bfd_boolean read_symbols)
 {
   size_t got;
@@ -671,7 +676,7 @@ setup_archive (struct archive_info *arch, const char *file_name,
 	  return 1;
 	}
       /* PR 17531: file: 639d6a26.  */
-      if (arch->longnames_size > file_size
+      if ((off_t) arch->longnames_size > file_size
 	  || (signed long) arch->longnames_size < 0)
 	{
 	  error (_("%s: long name table is too big, (size = 0x%lx)\n"),
diff --git a/binutils/elfcomm.h b/binutils/elfcomm.h
index 731d3dbd34..8d962f2731 100644
--- a/binutils/elfcomm.h
+++ b/binutils/elfcomm.h
@@ -74,7 +74,7 @@ extern char *adjust_relative_path (const char *, const char *, unsigned long);
 
 /* Read the symbol table and long-name table from an archive.  */
 extern int setup_archive (struct archive_info *, const char *, FILE *,
-			  bfd_size_type, bfd_boolean, bfd_boolean);
+			  off_t, bfd_boolean, bfd_boolean);
 
 /* Open and setup a nested archive, if not already open.  */
 extern int setup_nested_archive (struct archive_info *, const char *);

-- 
Alan Modra
Australia Development Lab, IBM

             reply	other threads:[~2020-03-19  3:25 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-03-19  3:25 Alan Modra [this message]
2020-03-19  5:40 ` Alan Modra

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=20200319032537.GA4583@bubble.grove.modra.org \
    --to=amodra@gmail.com \
    --cc=binutils@sourceware.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).