[RTEMS Project] #2369: [PowerPC Book E] Invalid mftb instruction in _CPU_Counter_read()

RTEMS trac trac at rtems.org
Thu Jul 9 06:14:47 UTC 2015


#2369: [PowerPC Book E] Invalid mftb instruction in _CPU_Counter_read()
--------------------------+---------------------
 Reporter:  nick.withers  |       Owner:
     Type:  defect        |      Status:  new
 Priority:  normal        |   Milestone:  4.11.1
Component:  General       |     Version:  4.11
 Severity:  normal        |  Resolution:
 Keywords:                |
--------------------------+---------------------

Comment (by nick.withers):

 Replying to [comment:1 sebastian.huber]:
 > The ppc8540 is the multilib intended for the e500 processors, why is
 this not used?

 No idea - where's that picked up from? I've been wondering for a quite a
 while why I get "-mcpu=powerpc" when building RTEMS and not
 "-mcpu=8540"... is that intentional / anything to do with it?

 I'm thoroughly confused by PowerPC versioning in general... According to
 Figure 4 from
 http://www.freescale.com/files/32bit/doc/white_paper/POWRPCARCPRMRM.pdf
 there's no alternate time base on Book E, so the ''mfspr %0, 526'' doesn't
 seem correct here... But then going off
 http://cache.freescale.com/files/32bit/doc/ref_manual/E500CORERM.pdf Table
 3.44, there are a bunch of instructions that are valid for Book E but
 invalid for the e500, so presumably the e500 doesn't actually conform to
 Book E... or something?!

 The marketing guff for the MVME3100 board says "System-on-chip Freescale
 MPC8540 with PowerPC® e500 processor core" - does that imply that not all
 MPC8540s use e500 cores?

--
Ticket URL: <http://devel.rtems.org/ticket/2369#comment:2>
RTEMS Project <http://www.rtems.org/>
RTEMS Project


More information about the bugs mailing list