I'm not on that list either - I don't do enough with the kernel to deal with the message volume. :-)

Are any of the gumstix software engineers able to post the question?

    Steve

Chris Dollar wrote:
I just compiled the 2.6.20 kernel with 'no forced preemption' and get
the same results as you, Steve.

I've never posted to the kernel list.... are you planning on doing
that? Or does any one else have any other ideas?

Chris

On 2/24/07, Steven A. Falco <safalco@optonline.net> wrote:
  
 I tried compiling with kernel feature "preemption mode" set to "no forced
preemption", and with no frame buffer.  I get this bug:

 BUG: at kernel/sched.c:4035 __schedule()
 [<c001cbc8>] (dump_stack+0x0/0x14) from [<c01379ac>]
(__schedule+0x62c/0x678)
 [<c0137380>] (__schedule+0x0/0x678) from [<c0137af0>] (schedule+0xd0/0x118)
 [<c0137a20>] (schedule+0x0/0x118) from [<c0137efc>]
(wait_for_completion+0x98/0xfc)
  r4 = C0263ED0
 [<c0137e64>] (wait_for_completion+0x0/0xfc) from [<c003adf0>]
(keventd_create_kthread+0x3c/0x60)
  r6 = C0263F1C  r5 = 00000002  r4 = C0263F40
 [<c003adb4>] (keventd_create_kthread+0x0/0x60) from
[<c003ae94>] (kthread_create+0x80/0xcc)
  r6 = C0262000  r5 = C0263F1C  r4 = 00000000
 [<c003ae14>] (kthread_create+0x0/0xcc) from [<c003c074>]
(posix_cpu_thread_call+0x40/0xb8)
  r3 = 00000000  r2 = C0167308
  r5 = C00164B0  r4 = 00000000
 [<c003c034>] (posix_cpu_thread_call+0x0/0xb8) from
[<c000f6b4>] (posix_cpu_thread_init+0x24/0x40)
  r5 = C00164B0  r4 = C018DFA8
 [<c000f690>] (posix_cpu_thread_init+0x0/0x40) from
[<c0018068>] (init+0x3c/0x284)
  r4 = 00000000
 [<c001802c>] (init+0x0/0x284) from [<c0029d34>] (do_exit+0x0/0x804)
  r7 = 00000000  r6 = 00000000  r5 = 00000000  r4 = 00000000

 With preemption set to "realtime" I get:

 BUG: at kernel/sched.c:4035 __schedule()
 [<c001cd48>] (dump_stack+0x0/0x24) from [<c0143888>]
(__schedule+0x6c8/0x718)
 [<c01431c0>] (__schedule+0x0/0x718) from [<c0143ac8>] (schedule+0xd0/0x118)
 [<c01439f8>] (schedule+0x0/0x118) from [<c003d2ac>] (kthread+0xc8/0x128)
  r4 = 00000000
 [<c003d1e4>] (kthread+0x0/0x128) from [<c002b1b8>] (do_exit+0x0/0x85c)
  r7 = 00000000  r6 = 00000000  r5 = 00000000  r4 = 00000000
 ---------------------------
 | preempt count: 00000000 ]
 | 0-level deep critical section nesting:
 ----------------------------------------

 So, if the RT patch is relevant, then just changing that configuration
setting is insufficient.

 It looks like this bug is triggered by interrupts being enabled when they
should not be.  I'm wondering if this discussion should be moved over to the
linux-kernel mailing list where a larger group of developers can be
consulted.

     Steve

 Dave Hylands wrote:
 Hi guys,



 But the crash is more important. If it is happening first, then you have to
fix it before you have any hope of getting the error -12 fixed.

 Th crash seems to happen on kernels with no fb too. I see it on mine
as well. It seems to be related to the RT stuff (at leat that was what
I concluded when I looked at the problem). So you might want to try
rebuilding with the RT stuff turned off and at least that crash will
go away.



-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
gumstix-users mailing list
gumstix-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/gumstix-users


    

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
gumstix-users mailing list
gumstix-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/gumstix-users