<p><br>
On Jul 22, 2012 11:08 AM, "Will Deacon" <<a href="mailto:will.deacon@arm.com">will.deacon@arm.com</a>> wrote:<br>
><br>
> On Sun, Jul 22, 2012 at 02:09:23PM +0100, Rob Herring wrote:<br>
> > On Jul 21, 2012 5:56 PM, "Arnd Bergmann" <<a href="mailto:arnd@arndb.de">arnd@arndb.de</a><mailto:<a href="mailto:arnd@arndb.de">arnd@arndb.de</a>>> wrote:<br>
> > > On Saturday 21 July 2012, Will Deacon wrote:<br>
> > > > I dusted off the integrator, but I'm failing to boot at all if I build from<br>
> > > > that branch:<br>
> > > ><br>
> > > > Uncompressing Linux... done, booting the kernel.<br>
> > > > <silence><br>
> > > ><br>
> > > > Using the same .config, I can boot v3.5-rc7 just fine (I even rebased your<br>
> > > > branch onto that in case something had been fixed in mainline, but it made<br>
> > > > no difference).<br>
> > ><br>
> > > I've looked at integrator_defconfig, and could not find any code that<br>
> > > actually uses the PIO accessors. Is your configuration different to that?<br>
> > > Do you actually have PCI enabled and present on the machine? Do things<br>
> > > change if you turn PCI off?<br>
><br>
> I have PCI up and running, yes, but all I use it for is an Intel e100<br>
> ethernet card which isn't required for booting. Disabling PCI makes no<br>
> difference, but see below.<br>
><br>
> > Could be overlapping static mappings. I manually checked that, but may have missed something.<br>
> ><br>
> > Can you turn on earlyprintk?<br>
><br>
> I had that turned on already... after banging my head against a wall, I<br>
> realised that increasing the baudrate in u-boot (I have to use ymodem to<br>
> transfer the kernel image...) kills the serial console completely when<br>
> booting Linux if I forget to change it back.<br>
><br>
> With that observation, I see booting get stuck:<br>
><br>
><br>
> Uncompressing Linux... done, booting the kernel.<br>
> [ 0.000000] Booting Linux on physical CPU 0<br>
> [ 0.000000] Initializing cgroup subsys cpu<br>
> [ 0.000000] Linux version 3.5.0-rc6-00017-g52f1412 (will@tiny-lites) (gcc version 4.5.3 (Gent2<br>
> [ 0.000000] CPU: ARM926EJ-S [41069263] revision 3 (ARMv5TEJ), cr=00053177<br>
> [ 0.000000] CPU: VIVT data cache, VIVT instruction cache<br>
> [ 0.000000] Machine: ARM-Integrator<br>
> [ 0.000000] bootconsole [earlycon0] enabled<br>
> [ 0.000000] Memory policy: ECC disabled, Data cache writeback<br>
> [ 0.000000] On node 0 totalpages: 32768<br>
> [ 0.000000] free_area_init_node: node 0, pgdat c04ad1f0, node_mem_map c04c9000<br>
> [ 0.000000] Normal zone: 256 pages used for memmap<br>
> [ 0.000000] Normal zone: 0 pages reserved<br>
> [ 0.000000] Normal zone: 32512 pages, LIFO batch:7<br>
> [ 0.000000] CPU: found DTCM0 32k @ 00000000, not enabled<br>
> [ 0.000000] CPU: moved DTCM0 32k to fffe8000, enabled<br>
> [ 0.000000] CPU: found ITCM0 32k @ 00000000, not enabled<br>
> [ 0.000000] CPU: moved ITCM0 32k to fffe0000, enabled<br>
> [ 0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768<br>
> [ 0.000000] pcpu-alloc: [0] 0<br>
> [ 0.000000] Built 1 zonelists in Zone order, mobility grouping on. Total pages: 32512<br>
> [ 0.000000] Kernel command line: console=ttyAM0 mem=128M earlyprintk debug user_debug=31 logl9<br>
> [ 0.000000] PID hash table entries: 512 (order: -1, 2048 bytes)<br>
> [ 0.000000] Dentry cache hash table entries: 16384 (order: 4, 65536 bytes)<br>
> [ 0.000000] Inode-cache hash table entries: 8192 (order: 3, 32768 bytes)<br>
> [ 0.000000] Memory: 128MB = 128MB total<br>
> [ 0.000000] Memory: 124968k/124968k available, 6104k reserved, 0K highmem<br>
> [ 0.000000] Virtual kernel memory layout:<br>
> [ 0.000000] vector : 0xffff0000 - 0xffff1000 ( 4 kB)<br>
> [ 0.000000] DTCM : 0xfffe8000 - 0xffff0000 ( 32 kB)<br>
> [ 0.000000] ITCM : 0xfffe0000 - 0xfffe8000 ( 32 kB)<br>
> [ 0.000000] fixmap : 0xfff00000 - 0xfffe0000 ( 896 kB)<br>
> [ 0.000000] vmalloc : 0xc8800000 - 0xff000000 ( 872 MB)<br>
> [ 0.000000] lowmem : 0xc0000000 - 0xc8000000 ( 128 MB)<br>
> [ 0.000000] modules : 0xbf000000 - 0xc0000000 ( 16 MB)<br>
> [ 0.000000] .text : 0xc0008000 - 0xc036e9b0 (3483 kB)<br>
> [ 0.000000] .init : 0xc036f000 - 0xc048e7a4 (1150 kB)<br>
> [ 0.000000] .data : 0xc0490000 - 0xc04ad920 ( 119 kB)<br>
> [ 0.000000] .bss : 0xc04ae024 - 0xc04c86b4 ( 106 kB)<br>
> [ 0.000000] SLUB: Genslabs=13, HWalign=32, Order=0-3, MinObjects=0, CPUs=1, Nodes=1<br>
> [ 0.000000] NR_IRQS:16 nr_irqs:34 34<br>
> [ 0.000000] FPGA IRQ chip 0 "SC" @ f1400000, 22 irqs<br>
> [ 0.000000] sched_clock: 16 bits at 1500kHz, resolution 666ns, wraps every 43ms<br>
><br>
><br>
> Now, the next line is usually when the VGA text console is poked. Sure<br>
> enough, disabling that (CONFIG_VGA_CONSOLE) is enough to boot with your<br>
> patches and PCI appears to work correctly (I can do basic networking).<br>
><br>
> I guess there's some mapping race with the VGA code since vga_base =<br>
> PCI_MEMORY_VADDR, but you left the static mapping alone for that region,<br>
> so I'm not sure. Any ideas?</p>
<p>Perhaps pcibios_min_io changing from 6000 to default of 1000 causes probing for vga?</p>
<p>Rob<br>
</p>