[PATCH] ld: Restore file offset after a plugin fails to claim a file

[Andrew Burgess <andrew dot burgess at embecosm dot com>]

When using the plugin interface to claim an input file the claim method
from (possible) many plugins can be called on an input file.  If these
claim methods read content from the input file then the file offset
stored in the underlying file descriptor will change.

As we share a file descriptor between the plugin interface (created with
dup in ld/plugin.c:plugin_object_p) and the input bfd object, then any
changes to the file offset in the file descriptor will effect the bfd
object.  Also, as the changes to the file offset did not originate from
calls through the bfd interface, but instead came from the plugin
directly, then the bfd will not be aware that the file offset has
changed.  This is a problem as the bfd library caches the file offset.
If the plugin decides not to claim an input file then, currently, we
leave the bfd in a state where the actual file offset is out of sync
with the cached file offset.

This problem came to light after a recent commit
7d0b9ebc1e0079271a7c7737b53bc026525eab64 (Don't include libbfd.h outside
of bfd, part 6) however, I don't believe that commit actual introduces
the bug, it just exposed the existing issue.

This commit solves the problem by backing up and restoring the file
offset for the file descriptor of the input file.  The restore is only
done if the plugin does not claim the input file, as it is in this case
that the bfd library might be used again to try and identify the
unclaimed file.

ld/ChangeLog:

	* plugin.c (plugin_call_claim_file): Restore the file offset after
	an unsuccessful attempt to claim a file.
	* testplug.c (bytes_to_read_before_claim): New global.
	(record_read_length): New function, sets new global
	bytes_to_read_before_claim.
	(parse_option): Handle 'read:<NUMBER>' option.
	(onclaim_file): Read file content before checking for claim.
	* testsuite/ld-plugin/plugin-30.d: New file.
	* testsuite/ld-plugin/plugin.exp: Add new test.
---
 ld/ChangeLog                       | 12 ++++++++++++
 ld/plugin.c                        |  5 +++++
 ld/testplug.c                      | 33 +++++++++++++++++++++++++++++++++
 ld/testsuite/ld-plugin/plugin-30.d | 26 ++++++++++++++++++++++++++
 ld/testsuite/ld-plugin/plugin.exp  |  7 ++++++-
 5 files changed, 82 insertions(+), 1 deletion(-)
 create mode 100644 ld/testsuite/ld-plugin/plugin-30.d

diff --git a/ld/plugin.c b/ld/plugin.c
index ffa0dd3..36094dd 100644
--- a/ld/plugin.c
+++ b/ld/plugin.c
@@ -1049,9 +1049,14 @@ plugin_call_claim_file (const struct ld_plugin_input_file *file, int *claimed)
     {
       if (curplug->claim_file_handler)
 	{
+	  off_t cur_offset;
 	  enum ld_plugin_status rv;
+
 	  called_plugin = curplug;
+	  cur_offset = lseek (file->fd, 0, SEEK_CUR);
 	  rv = (*curplug->claim_file_handler) (file, claimed);
+	  if (!*claimed)
+	    lseek (file->fd, cur_offset, SEEK_SET);
 	  called_plugin = NULL;
 	  if (rv != LDPS_OK)
 	    set_plugin_error (curplug->name);
diff --git a/ld/testplug.c b/ld/testplug.c
index cec947c..7eccd64 100644
--- a/ld/testplug.c
+++ b/ld/testplug.c
@@ -139,6 +139,10 @@ static add_file_t *addfiles_list = NULL;
 /* We keep a tail pointer for easy linking on the end.  */
 static add_file_t **addfiles_tail_chain_ptr = &addfiles_list;
 
+/* Number of bytes read in claim file before deciding if the file can be
+   claimed.  */
+static int bytes_to_read_before_claim = 0;
+
 /* Add a new claimfile on the end of the chain.  */
 static enum ld_plugin_status
 record_claim_file (const char *file)
@@ -159,6 +163,25 @@ record_claim_file (const char *file)
   return LDPS_OK;
 }
 
+/* How many bytes to read before claiming (or not) an input file.  */
+static enum ld_plugin_status
+record_read_length (const char *length)
+{
+  const char *tmp;
+
+  tmp = length;
+  while (*tmp != '\0' && isdigit (*tmp))
+    ++tmp;
+  if (*tmp != '\0' || *length == '\0')
+    {
+      fprintf (stderr, "APB: Bad length string: %s\n", tmp);
+      return LDPS_ERR;
+    }
+
+  bytes_to_read_before_claim = atoi (length);
+  return LDPS_OK;
+}
+
 /* Add a new addfile on the end of the chain.  */
 static enum ld_plugin_status
 record_add_file (const char *file, addfile_enum_t type)
@@ -325,6 +348,8 @@ parse_option (const char *opt)
     return set_register_hook (opt + 10, FALSE);
   else if (!strncmp ("claim:", opt, 6))
     return record_claim_file (opt + 6);
+  else if (!strncmp ("read:", opt, 5))
+    return record_read_length (opt + 5);
   else if (!strncmp ("sym:", opt, 4))
     return record_claimed_file_symbol (opt + 4);
   else if (!strncmp ("add:", opt, 4))
@@ -527,6 +552,14 @@ onload (struct ld_plugin_tv *tv)
 static enum ld_plugin_status
 onclaim_file (const struct ld_plugin_input_file *file, int *claimed)
 {
+  if (bytes_to_read_before_claim > 0)
+    {
+      char *buffer = malloc (bytes_to_read_before_claim);
+
+      read (file->fd, buffer, bytes_to_read_before_claim);
+      free (buffer);
+    }
+
   /* Let's see if we want to claim this file.  */
   claim_file_t *claimfile = claimfiles_list;
   while (claimfile)
diff --git a/ld/testsuite/ld-plugin/plugin-30.d b/ld/testsuite/ld-plugin/plugin-30.d
new file mode 100644
index 0000000..eb9d424
--- /dev/null
+++ b/ld/testsuite/ld-plugin/plugin-30.d
@@ -0,0 +1,26 @@
+Hello from testplugin.
+.*: LDPT_MESSAGE func@0x.*
+.*: LDPT_API_VERSION value        0x1 \(1\)
+.*: LDPT_GNU_LD_VERSION value       0x.*
+.*: LDPT_LINKER_OUTPUT value        0x1 \(1\)
+.*: LDPT_OUTPUT_NAME 'tmpdir/main.x'
+.*: LDPT_REGISTER_CLAIM_FILE_HOOK func@0x.*
+.*: LDPT_REGISTER_ALL_SYMBOLS_READ_HOOK func@0x.*
+.*: LDPT_REGISTER_CLEANUP_HOOK func@0x.*
+.*: LDPT_ADD_SYMBOLS func@0x.*
+.*: LDPT_GET_INPUT_FILE func@0x.*
+.*: LDPT_GET_VIEW func@0x.*
+.*: LDPT_RELEASE_INPUT_FILE func@0x.*
+.*: LDPT_GET_SYMBOLS func@0x.*
+.*: LDPT_GET_SYMBOLS_V2 func@0x.*
+.*: LDPT_ADD_INPUT_FILE func@0x.*
+.*: LDPT_ADD_INPUT_LIBRARY func@0x.*
+.*: LDPT_SET_EXTRA_LIBRARY_PATH func@0x.*
+.*: LDPT_OPTION 'registerclaimfile'
+.*: LDPT_OPTION 'read:8'
+.*: LDPT_NULL value        0x0 \(0\)
+#...
+hook called: claim_file tmpdir/main.o \[@0/.* not claimed
+hook called: claim_file tmpdir/func.o \[@0/.* not claimed
+hook called: claim_file tmpdir/text.o \[@0/.* not claimed
+hook called: claim_file tmpdir/libempty.a \[@.* not claimed
diff --git a/ld/testsuite/ld-plugin/plugin.exp b/ld/testsuite/ld-plugin/plugin.exp
index 4c9116b..a6946c4 100644
--- a/ld/testsuite/ld-plugin/plugin.exp
+++ b/ld/testsuite/ld-plugin/plugin.exp
@@ -236,6 +236,10 @@ set plugin_lib_tests [list \
 			-plugin-opt sym:${_}text::0:0:0 \
 			-plugin-opt add:tmpdir/text.o \
     $testsrcfiles_notext -Ltmpdir -ltext $libs" "" "" "" {{ld plugin-19.d}} "main.x" ] \
+    [list "plugin with empty archive" \
+	               "-plugin $plugin_path $regclm \
+			-plugin-opt read:8 \
+    $testobjfiles tmpdir/libempty.a $libs" "" "" "" {{ld plugin-30.d}} "main.x" ] \
 ]
 
 set plugin_extra_elf_tests [list \
@@ -289,7 +293,8 @@ if { [is_elf_format] \
     run_ld_link_tests $plugin_extra_elf_tests
 }
 
-if ![ar_simple_create $ar "" "tmpdir/libtext.a" "tmpdir/text.o"] {
+if {![ar_simple_create $ar "" "tmpdir/libtext.a" "tmpdir/text.o"] || \
+    ![ar_simple_create $ar "" "tmpdir/libempty.a" ""]} {
     foreach testitem $plugin_lib_tests {
 	unresolved [lindex $testitem 0]
     }
-- 
2.5.1