Fix reads for sizes larger than INT_MAX in AF_INET lookup

Currently for AF_INET lookups from the hosts file, buffer sizes larger
than INT_MAX silently overflow and may result in access beyond bounds
of a buffer.  This happens when the number of results in an AF_INET
lookup in /etc/hosts are very large.

There are two aspects to the problem.  One problem is that the size
computed from the buffer size is stored into an int, which results in
overflow for large sizes.  Additionally, even if this size was
expanded, the function used to read content into the buffer (fgets)
accepts only int sizes.  As a result, the fix is to have a function
wrap around fgets that calls it multiple times with int sizes if
necessary.

1 files changed, 51 insertions, 8 deletions

diff --git a/nss/nss_files/files-XXX.c b/nss/nss_files/files-XXX.c
index 082d1ea..b62208c 100644
--- a/nss/nss_files/files-XXX.c
+++ b/nss/nss_files/files-XXX.c
@@ -179,8 +179,51 @@ CONCAT(_nss_files_end,ENTNAME) (void)
   return NSS_STATUS_SUCCESS;
 }
 
-/* Parsing the database file into `struct STRUCTURE' data structures.  */
 
+typedef enum
+{
+  gcr_ok = 0,
+  gcr_error = -1,
+  gcr_overflow = -2
+} get_contents_ret;
+
+/* Hack around the fact that fgets only accepts int sizes.  */
+static get_contents_ret
+get_contents (char *linebuf, size_t len, FILE *stream)
+{
+  size_t remaining_len = len;
+  char *curbuf = linebuf;
+
+  do
+    {
+      int curlen = ((remaining_len > (size_t) INT_MAX) ? INT_MAX
+		    : remaining_len);
+      char *p = fgets_unlocked (curbuf, curlen, stream);
+
+      ((unsigned char *) curbuf)[curlen - 1] = 0xff;
+
+      /* EOF or read error.  */
+      if (p == NULL)
+        return gcr_error;
+
+      /* Done reading in the line.  */
+      if (((unsigned char *) curbuf)[curlen - 1] == 0xff)
+        return gcr_ok;
+
+      /* Drop the terminating '\0'.  */
+      remaining_len -= curlen - 1;
+      curbuf += curlen - 1;
+    }
+  /* fgets copies one less than the input length.  Our last iteration is of
+     REMAINING_LEN and once that is done, REMAINING_LEN is decremented by
+     REMAINING_LEN - 1, leaving the result as 1.  */
+  while (remaining_len > 1);
+
+  /* This means that the current buffer was not large enough.  */
+  return gcr_overflow;
+}
+
+/* Parsing the database file into `struct STRUCTURE' data structures.  */
 static enum nss_status
 internal_getent (struct STRUCTURE *result,
 		 char *buffer, size_t buflen, int *errnop H_ERRNO_PROTO
@@ -188,7 +231,7 @@ internal_getent (struct STRUCTURE *result,
 {
   char *p;
   struct parser_data *data = (void *) buffer;
-  int linebuflen = buffer + buflen - data->linebuffer;
+  size_t linebuflen = buffer + buflen - data->linebuffer;
   int parse_result;
 
   if (buflen < sizeof *data + 2)
@@ -200,17 +243,16 @@ internal_getent (struct STRUCTURE *result,
 
   do
     {
-      /* Terminate the line so that we can test for overflow.  */
-      ((unsigned char *) data->linebuffer)[linebuflen - 1] = '\xff';
+      get_contents_ret r = get_contents (data->linebuffer, linebuflen, stream);
 
-      p = fgets_unlocked (data->linebuffer, linebuflen, stream);
-      if (p == NULL)
+      if (r == gcr_error)
 	{
 	  /* End of file or read error.  */
 	  H_ERRNO_SET (HOST_NOT_FOUND);
 	  return NSS_STATUS_NOTFOUND;
 	}
-      else if (((unsigned char *) data->linebuffer)[linebuflen - 1] != 0xff)
+
+      if (r == gcr_overflow)
 	{
 	  /* The line is too long.  Give the user the opportunity to
 	     enlarge the buffer.  */
@@ -219,7 +261,8 @@ internal_getent (struct STRUCTURE *result,
 	  return NSS_STATUS_TRYAGAIN;
 	}
 
-      /* Skip leading blanks.  */
+      /* Everything OK.  Now skip leading blanks.  */
+      p = data->linebuffer;
       while (isspace (*p))
 	++p;
     }