[LinuxBIOS] problems with reboot on K8

Rudolf Marek r.marek at assembler.cz
Tue Sep 4 21:47:30 CEST 2007


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

Sorry for the delay, I moved to flat and got no internet for a while.

LinuxBIOS-2.0.0_Fallback Po zář  3 22:17:08 CEST 2007 starting...

now booting... fallback





LinuxBIOS-2.0.0_Normal Po zář  3 22:16:19 CEST 2007 starting...

now booting... real_main







INIT detected from  ---- {APICID = 00 NODEID = 00 COREID = 00} ---



Issuing SOFT_RESET...

soft reset

soft reset done

now booting... after init_cpus

now booting... after udelay

now booting... bist

000c1044 <-00000000

000c104c <-00000001

000c1054 <-



This is my boot log when I type "reboot" from Linux. It seems chipset just sends
INIT# to CPU. I could not force it to generate proper CPURST#. It seems there is
some bit for that in SB but documentation is unclear (VT8237) I asked VIA
recently, but still no answer.

To catch the log with my code:


       w83627ehg_enable_serial(SERIAL_DEV, TTYS0_BASE);

        uart_init();

        console_init();



       print_info("now booting... real_main\r\n");



        if (bist == 0) {

               //init_cpus(cpu_init_detectedx);

               bsp_apicid = init_cpus(cpu_init_detectedx, sysinfo);

        }

       // enable_lapic();

        init_timer();



       print_info("now booting... after init_cpus\r\n");



        mdelay(1000);

       print_info("now booting... after udelay\r\n");



       /* Halt if there was a built in self test failure */

       report_bist_failure(bist);
      print_info("now booting... bist\r\n");



       // setup_a8v_resource_map();

        setup_default_resource_map();



       print_info("now booting... resource map\r\n");



       setup_coherent_ht_domain();



       print_info("now booting... HT domain\r\n");





        wait_all_core0_started();



       print_info("now booting... Core0 started\r\n");

So the problem is with setup_default resource map... With the delay(1000) it
will hang on the third write, without the delay it will hang in random place in
the very same routine.

What is interesting if I reboot through undocumented 0xcf9 (already send some
questions about this to VIA) I can go on at least to setup_coherent_ht_domain();

I dont know what might be wrong :/ My ROM strap for VIA chipset is not working
(all registers gets filled with FF) correctly right now, but I dont think this
should be a problem while rebooting??? I think HT Link is set up bit later....

My soft_reset routine just triggers PCI_RESET signal from SB.

I put online some log file of full boot and reboot tries at
http://assembler.cz/lb there is also my mainboard specific code.

Thanks,

Rudolf
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFG3bZS3J9wPJqZRNURAgg3AKCGzybE9IxEDt1bBLLp5QqUQPQU/QCgmUML
lQQX8wblhLJGZsU+BxyXM4Y=
=O5GO
-----END PGP SIGNATURE-----




More information about the coreboot mailing list