PERFORCE change 84545 for review

Todd Miller millert at FreeBSD.org
Fri Sep 30 13:04:45 GMT 2005


http://perforce.freebsd.org/chv.cgi?CH=84545

Change 84545 by millert at millert_g4tower on 2005/09/30 13:03:55

	Further clarification of MACH_RCV_TRAILER_AV and MACH_RCV_TRAILER_LABELS
	gotchas.

Affected files ...

.. //depot/projects/trustedbsd/sedarwin7/src/darwin/xnu/osfmk/mach/message.h#7 edit

Differences ...

==== //depot/projects/trustedbsd/sedarwin7/src/darwin/xnu/osfmk/mach/message.h#7 (text+ko) ====

@@ -519,6 +519,11 @@
  * NOTE: a 0x00------ RCV mask implies to ask for
  * a MACH_MSG_TRAILER_FORMAT_0 with 0 Elements, 
  * which is equivalent to a mach_msg_trailer_t.
+ *
+ * XXXMAC: unlike the rest of the MACH_RCV_* flags, MACH_RCV_TRAILER_LABELS
+ * and MACH_RCV_TRAILER_AV need their own private bit since we only calculate
+ * their fields when absolutely required.  This will cause us problems if
+ * Apple adds new trailers.
  */
 #define MACH_RCV_TRAILER_NULL   0
 #define MACH_RCV_TRAILER_SEQNO  1
@@ -534,12 +539,13 @@
 #define GET_RCV_ELEMENTS(y) (((y) >> 24) & 0xf)
 
 /* 
-   It should be noted here that in the case of MACH_RCV_TRAILER_AV,
-   we just fall through to mach_msg_max_trailer_t. This is correct
-   behavior since mach_msg_max_trailer_t is defined as mac_msg_mac_trailer_t
-   which is used for the AV trailer. At some point this code should be 
-   updated to clarify this case. -d 
-*/
+ * XXXMAC: note that in the case of MACH_RCV_TRAILER_AV and
+ * MACH_RCV_TRAILER_LABELS, we just fall through to mach_msg_max_trailer_t.
+ * This is correct behavior since mach_msg_max_trailer_t is defined as
+ * mac_msg_mac_trailer_t which is used for the LABELS and AV trailers.
+ * It also makes things work properly if MACH_RCV_TRAILER_AV or
+ * MACH_RCV_TRAILER_LABELS are ORed with one of the other options.
+ */
 #define REQUESTED_TRAILER_SIZE(y) 				\
 	((mach_msg_trailer_size_t)				\
 	 ((GET_RCV_ELEMENTS(y) == MACH_RCV_TRAILER_NULL) ?	\
To Unsubscribe: send mail to majordomo at trustedbsd.org
with "unsubscribe trustedbsd-cvs" in the body of the message



More information about the trustedbsd-cvs mailing list