newest "bk pull" of 8xx tree still builds a kernel that hangs

newest "bk pull" of 8xx tree still builds a kernel that hangs

[Robert P. J. Day]

  having just got back from vacation, i may have missed some of the latest
developments.  i just did a "bk pull" from the linuxppc-2.5 tree, rebuilt
the kernel for my 8xx board and, just like before i left, the kernel hung
almost immediately upon booting:

Linux/PPC load: console=ttyS0,9600 rw root=/dev/ram0 init=/sbin/minit
Uncompressing Linux...done.
Now booting the kernel
(hang)

should i be expecting this?  once upon a time, of course, i got all the
way to user space before the hang.  have there been some recent patches to
cause this latest behaviour?  i haven't messed with the config file in
quite some time.  should i?

thanks for any guidance.

rday


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: newest "bk pull" of 8xx tree still builds a kernel that hangs

[Tom Rini]

On Mon, Jun 07, 2004 at 11:55:33AM -0400, Robert P. J. Day wrote:

>
>   having just got back from vacation, i may have missed some of the latest
> developments.  i just did a "bk pull" from the linuxppc-2.5 tree, rebuilt
> the kernel for my 8xx board and, just like before i left, the kernel hung
> almost immediately upon booting:
>
> Linux/PPC load: console=ttyS0,9600 rw root=/dev/ram0 init=/sbin/minit
> Uncompressing Linux...done.
> Now booting the kernel
> (hang)
>
> should i be expecting this?  once upon a time, of course, i got all the
> way to user space before the hang.  have there been some recent patches to
> cause this latest behaviour?  i haven't messed with the config file in
> quite some time.  should i?

What probably happened is that you don't have a UART driver selected.
I've had my best luck with CONFIG_SERIAL_CPM, so you might want to try
that.

--
Tom Rini
http://gate.crashing.org/~trini/

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: newest "bk pull" of 8xx tree still builds a kernel that hangs

[Robert P. J. Day]

On Mon, 7 Jun 2004, Tom Rini wrote:

>
> On Mon, Jun 07, 2004 at 11:55:33AM -0400, Robert P. J. Day wrote:
>
> >
> >   having just got back from vacation, i may have missed some of the latest
> > developments.  i just did a "bk pull" from the linuxppc-2.5 tree, rebuilt
> > the kernel for my 8xx board and, just like before i left, the kernel hung
> > almost immediately upon booting:
> >
> > Linux/PPC load: console=ttyS0,9600 rw root=/dev/ram0 init=/sbin/minit
> > Uncompressing Linux...done.
> > Now booting the kernel
> > (hang)
> >
> > should i be expecting this?  once upon a time, of course, i got all the
> > way to user space before the hang.  have there been some recent patches to
> > cause this latest behaviour?  i haven't messed with the config file in
> > quite some time.  should i?
>
> What probably happened is that you don't have a UART driver selected.
> I've had my best luck with CONFIG_SERIAL_CPM, so you might want to try
> that.

i selected that, but still no luck -- kernel still hangs almost
immediately -- but there may be other issues here.

i'm working with a custom board that, despite the initial contents of the
file arch/ppc/8xx_io/uart.c, supports both ethernet on SCC3 and console on
SMC1.  for our 2.4 kernel, i have the MPC8xx options (among others):

  [*] SMC1/SCC3/I2C/SPI Microcode Patch
  [*] CPM SCC Ethernet
  (SCC3) SCC used for Etherhet
  [*] Use SMC1 for UART

so you can definitely see that we've been hacking the kernel source for
our 2.4 build on this board.  i was *hoping* to get a minimal 2.6 config
to boot just to prove that it could be done, before starting to make any
necessary mods from there.  now i'm wondering if that's even possible.

since i'd rather not start mucking with the source code just yet (it being
in a state of flux), i think what i want to try is to get an absolutely
minimal config to see if it boots.  so, one would think, i could turn off
stuff like JFFS2 support, tmpfs, even networking, just to see if i can get
any further.

if anyone wants to suggest the bare minimum config that should build a
bootable kernel, i'd be interested.  but i'm starting to get this
unpleasant sensation that i really might need to hack uart.c as we did
with the 2.4 kernel just for our board.

rday


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: newest "bk pull" of 8xx tree still builds a kernel that hangs

[Tom Rini]

On Mon, Jun 07, 2004 at 08:04:44PM -0400, Robert P. J. Day wrote:
> On Mon, 7 Jun 2004, Tom Rini wrote:
> > On Mon, Jun 07, 2004 at 11:55:33AM -0400, Robert P. J. Day wrote:
> >
> > >
> > >   having just got back from vacation, i may have missed some of the latest
> > > developments.  i just did a "bk pull" from the linuxppc-2.5 tree, rebuilt
> > > the kernel for my 8xx board and, just like before i left, the kernel hung
> > > almost immediately upon booting:
> > >
> > > Linux/PPC load: console=ttyS0,9600 rw root=/dev/ram0 init=/sbin/minit
> > > Uncompressing Linux...done.
> > > Now booting the kernel
> > > (hang)
> > >
> > > should i be expecting this?  once upon a time, of course, i got all the
> > > way to user space before the hang.  have there been some recent patches to
> > > cause this latest behaviour?  i haven't messed with the config file in
> > > quite some time.  should i?
> >
> > What probably happened is that you don't have a UART driver selected.
> > I've had my best luck with CONFIG_SERIAL_CPM, so you might want to try
> > that.
>
> i selected that, but still no luck -- kernel still hangs almost
> immediately -- but there may be other issues here.
[snip]

I haven't had time to look into this yet, in part because I just don't
know how well to trust my h/w, but to get SMC1 console working (RPX
Lite) I need to enable both SCC1 and SMC1.  For your board, I would
recommend enabling all of the consoles.

And in case others were wondering, the old driver has a different set of
issues, at least on my hw, it doesn't see input (and/or it causes a
livelock at boot, I haven't tried again in a while).

--
Tom Rini
http://gate.crashing.org/~trini/

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: newest "bk pull" of 8xx tree still builds a kernel that hangs

[Robert P. J. Day]

On Mon, 7 Jun 2004, Tom Rini wrote:

...
> I haven't had time to look into this yet, in part because I just don't
> know how well to trust my h/w, but to get SMC1 console working (RPX
> Lite) I need to enable both SCC1 and SMC1.  For your board, I would
> recommend enabling all of the consoles.

ok, i just turned them all on, it's compiling as we speak.  i'll let you
know what happens.  my .config file now looks like (in part):

...
CONFIG_VT=y
CONFIG_VT_CONSOLE=y
CONFIG_HW_CONSOLE=y
...
CONFIG_SERIAL_8250=y
CONFIG_SERIAL_8250_CONSOLE=y
CONFIG_SERIAL_8250_NR_UARTS=4
CONFIG_SERIAL_8250_EXTENDED=y
...
CONFIG_SERIAL_CORE=y
CONFIG_SERIAL_CORE_CONSOLE=y
CONFIG_SERIAL_CPM=y
CONFIG_SERIAL_CPM_CONSOLE=y
CONFIG_SERIAL_CPM_SCC1=y
CONFIG_SERIAL_CPM_SCC2=y
CONFIG_SERIAL_CPM_SCC3=y
CONFIG_SERIAL_CPM_SCC4=y
CONFIG_SERIAL_CPM_SMC1=y
CONFIG_SERIAL_CPM_SMC2=y
CONFIG_UNIX98_PTYS=y
...

i'm pretty sure there's not much more i can turn on related to this. :-)

rday

\x01\x01\x01

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: newest "bk pull" of 8xx tree still builds a kernel that hangs

[Tom Rini]

On Mon, Jun 07, 2004 at 08:28:26PM -0400, Robert P. J. Day wrote:

>
>
> On Mon, 7 Jun 2004, Tom Rini wrote:
>
> ...
> > I haven't had time to look into this yet, in part because I just don't
> > know how well to trust my h/w, but to get SMC1 console working (RPX
> > Lite) I need to enable both SCC1 and SMC1.  For your board, I would
> > recommend enabling all of the consoles.
>
> ok, i just turned them all on, it's compiling as we speak.  i'll let you
> know what happens.  my .config file now looks like (in part):
>
> ...
> CONFIG_VT=y
> CONFIG_VT_CONSOLE=y
> CONFIG_HW_CONSOLE=y
> ...
> CONFIG_SERIAL_8250=y
> CONFIG_SERIAL_8250_CONSOLE=y
> CONFIG_SERIAL_8250_NR_UARTS=4
> CONFIG_SERIAL_8250_EXTENDED=y
> ...
> CONFIG_SERIAL_CORE=y
> CONFIG_SERIAL_CORE_CONSOLE=y
> CONFIG_SERIAL_CPM=y
> CONFIG_SERIAL_CPM_CONSOLE=y
> CONFIG_SERIAL_CPM_SCC1=y
> CONFIG_SERIAL_CPM_SCC2=y
> CONFIG_SERIAL_CPM_SCC3=y
> CONFIG_SERIAL_CPM_SCC4=y
> CONFIG_SERIAL_CPM_SMC1=y
> CONFIG_SERIAL_CPM_SMC2=y
> CONFIG_UNIX98_PTYS=y

Then don't forget a console=ttySX (or ttyCPMX) if you have CONFIG_VT
set.  This early on, I'd turn off 8250 myself (and VT and INPUT).

--
Tom Rini
http://gate.crashing.org/~trini/

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: newest "bk pull" of 8xx tree still builds a kernel that hangs

[Robert P. J. Day]

On Mon, 7 Jun 2004, Tom Rini wrote:

> Then don't forget a console=ttySX (or ttyCPMX) if you have CONFIG_VT
> set.  This early on, I'd turn off 8250 myself (and VT and INPUT).

you da man.  i de-selected the 8250 stuff, and VT, and my load line was
nothing more than:

Linux/PPC load: console=ttyS0,9600        [EP's PlanetCore loader]

(which would, of course, crash and burn as soon as it tried to find an
init-like program, but who cares?)

the results:

Linux/PPC load: console=ttyS0,9600
Uncompressing Linux...done.
Now booting the kernel
Linux version 2.6.7-rc2 (rpjday@localhost.localdomain) (gcc version 3.2.2
200304
On node 0 totalpages: 8192
  DMA zone: 8192 pages, LIFO batch:2
  Normal zone: 0 pages, LIFO batch:1
  HighMem zone: 0 pages, LIFO batch:1
Built 1 zonelists
Kernel command line: console=ttyS0,9600
PID hash table entries: 256 (order 8: 2048 bytes)
Decrementer Frequency = 187500000/60
Memory: 29584k available (764k kernel code, 200k data, 52k init, 0k
highmem)
Calibrating delay loop... 46.72 BogoMIPS
Dentry cache hash table entries: 4096 (order: 2, 16384 bytes)
Inode-cache hash table entries: 2048 (order: 1, 8192 bytes)
Mount-cache hash table entries: 512 (order: 0, 4096 bytes)
checking if image is initramfs...it isn't (no cpio magic); looks like an
initrd
Freeing initrd memory: 1801k freed
Linux NoNET1.0 for Linux 2.6
devfs: 2004-01-31 Richard Gooch (rgooch@atnf.csiro.au)
devfs: boot_options: 0x1
Generic RTC Driver v1.07
Serial: CPM driver $Revision: 0.01 $.


at which point it hung, but at least that's progress.  i suspect i'm
getting closer and closer to the point where i really do have to take
into account the custom nature of this board, but first, i'll see if i
can get any farther without descending into that kind of madness.

rday

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/