[okl4-developer] booting fails on pxa270

Carl van Schaik carl at ok-labs.com
Mon Sep 24 21:58:18 EST 2007


Hi Kashin,

You can try put some of the same debug in the timer driver and timer server/

iguana/timer/
drivers/pxa250_timer/

I suspect its the serial driver causing the problem (but I could be wrong).

regards,
Carl

kashin Lin wrote:
> Hi,
>
> i pepper some "L4_KDB_Enter" in "linux/kernel/arch/l4/kernel/main.c"
> and the details are:
> ==================================================
> int
> main(int argc, char **argv)
> {
>       int r;
> *L4_KDB_Enter("In main");*
>       /* Start a new thread */
>       assert(argc > 0 && argv[0] != NULL);
>       strlcpy(saved_command_line, argv[1], COMMAND_LINE_SIZE);
> * L4_KDB_Enter("Before SetThreadName");*
>       L4_KDB_SetThreadName(L4_Myself(), "L_timer");
> *L4_KDB_Enter("Before create_priority");*
>       r = thread_create_priority(99, &main_thread);
>       assert (r != 0);
> *L4_KDB_Enter("Before SetThreadName for main thread");*
>       timer_thread = L4_Myself();
>       L4_KDB_SetThreadName(main_thread, "L_syscall");
> *L4_KDB_Enter("Before L4_Start_SpIp");*
>       /* Setup our TLS as well */
>       setup_tls(0);
>       L4_Start_SpIp(main_thread,
>                            (L4_Word_t) &main_stack[STACK_SIZE-1],
>                            (L4_Word_t) start_kernel);
> *L4_KDB_Enter("Before interrupt_loop");*
>       /* Now we go and do the timer stuff */
>       interrupt_loop();
>       return 0;
> }
>
>
> and each time when i get in l4_kdb, i show the schedule
> queue to know the threads' information.
>
> the following is the message i got:
> ========================
> *PXA25X OS Timer driver starting...
> PXA25X wrap handler installed
> Using FFUART--- KD# User: In main ---
> > showqueue
>
> [255]: (roottask) (IRQ 28) (IRQ 27)
> [242]: (tmr_wrap)
> [240]: (ig_timer)
> [200]: (ig_namin)
> [100]: ig_seria vmlinux
> idle : idle
>
> > go
>
> --- KD# User: Before SetThreadName ---
> > showqueue
>
> [255]: (roottask) (IRQ 28) (IRQ 27)
> [242]: (tmr_wrap)
> [240]: (ig_timer)
> [200]: (ig_namin)
> [100]: ig_seria vmlinux
> idle : idle
>
> > go
> .
>
> Backed --- KD# User: Before create_priority ---
> > showqueue
>
> [255]: (roottask) (IRQ 28) (IRQ 27) (IRQ 22)
> [242]: (tmr_wrap)
> [240]: (ig_timer)
> [200]: (ig_namin)
> [100]: ig_seria L_timer
> idle : idle
> > go
> (STOP HERE)*
>
> but i am not sure this stopping is due to which thread or it's just
> hardware error...
> can you give me some suggestions about this situation?
> or what should i do to bring you more information?
>
> thanks for your answering.
>
>
> 2007/9/20, Carl van Schaik < carl at ok-labs.com <mailto:carl at ok-labs.com>>:
>
>     To me it looks like the UART server is starting up and
>     reconfigures the
>     UART etc. At this point, the uart fifo empties and it seems that no
>     further writes happen. This is probably because you are not getting
>     interrupts. Either the serial driver registers the wrong interrupt
>     number or the driver is misconfiguring the UART.
>
>     regards,
>     Carl
>
>     Ben Leslie wrote:
>     > On Wed Sep 05, 2007 at 17:15:09 +0800, kashin Lin wrote:
>     >
>     >> Hi,
>     >>
>     >> i build okl4 for pxa270 using following machine configuration:
>     >> vbase is 0x1302000, vend is 0xcfffffff
>     >> addr is 0x1302000, vbase is 0x1302000
>     >> PXA25X OS Timer driver starting...
>     >> PXA25X wrap handler installed
>     >> Using FFUA *(STOP HERE)
>     >> ========================================
>     >>
>     >> can anyone give me some suggestions about what may be the cause
>     of this
>     >> status
>     >> and how can i fix it?
>     >>
>     >
>     > I'm not sure what the cause is. Are you able to enter the kernel
>     > debugger at this point to try and get some further information?
>     >
>     > Cheers,
>     >
>     > Ben
>     >
>     > _______________________________________________
>     > Developer mailing list
>     > Developer at okl4.org <mailto:Developer at okl4.org>
>     > https://lists.okl4.org/mailman/listinfo/developer
>     >
>     >
>
>




More information about the Developer mailing list