linux.git/arch/powerpc/kernel/rtas.c, branch v2.6.24

Revert "[POWERPC] Fix RTAS os-term usage on kernel panic"

2007-12-02T22:39:45+00:00

This reverts commit a2b51812a4dc5db09ab4d4638d4d8ed456e2457e.

It turns out that this change caused some machines to fail to come
back up when being rebooted, and generated an error in the hypervisor
error log on some machines.  The platform architecture (PAPR) is a
little unclear on exactly when the RTAS ibm,os-term function should be
called.  Until that is clarified I'm reverting this commit.

Signed-off-by: Paul Mackerras

[POWERPC] Fix RTAS os-term usage on kernel panic

2007-11-20T05:10:09+00:00

The rtas_os_term() routine was being called at the wrong time.
The actual rtas call "os-term" will not ever return, and so
calling it from the panic notifier is too early.  Instead,
call it from the machine_reset() call.

This splits the rtas_os_term() routine into two: one part to capture
the kernel panic message, invoked during the panic notifier, and
another part that is invoked during machine_reset().

Prior to this patch, the os-term call was never being made,
because panic_timeout was always non-zero.  Calling os-term
helps keep the hypervisor happy!  We have to keep the hypervisor
happy to avoid service, dump and error reporting problems.

Signed-off-by: Linas Vepstas 
Signed-off-by: Paul Mackerras

[POWERPC] Fix multiple bugs in rtas_ibm_suspend_me code

2007-11-19T04:11:30+00:00

There are several issues with the rtas_ibm_suspend_me code, which
enables platform-assisted suspension of an LPAR as covered in PAPR
2.2.

1.) rtas_ibm_suspend_me uses on_each_cpu() to invoke
rtas_percpu_suspend_me on all cpus via IPI:

if (on_each_cpu(rtas_percpu_suspend_me, &data, 1, 0))
...

'data' is on the calling task's stack, but rtas_ibm_suspend_me takes
no measures to ensure that all instances of rtas_percpu_suspend_me are
finished accessing 'data' before returning.  This can result in the
IPI'd cpus accessing random stack data and getting stuck in H_JOIN.

This is addressed by using an atomic count of workers and a completion
on the stack.

2.) rtas_percpu_suspend_me is needlessly calling H_JOIN in a loop.
The only event that can cause a cpu to return from H_JOIN is an H_PROD
from another cpu or a NMI/system reset.  Each cpu need call H_JOIN
only once per suspend operation.

Remove the loop and the now unnecessary 'waiting' state variable.

3.) H_JOIN must be called with MSR[EE] off, but lazy interrupt
disabling may cause the caller of rtas_ibm_suspend_me to call H_JOIN
with it on; the local_irq_disable() in on_each_cpu() is not
sufficient.

Fix this by explicitly saving the MSR and clearing the EE bit before
calling H_JOIN.

4.) H_PROD is being called with the Linux logical cpu number as the
parameter, not the platform interrupt server value.  (It's also being
called for all possible cpus, which is harmless, but unnecessary.)

This is fixed by calling H_PROD for each online cpu using
get_hard_smp_processor_id(cpu) for the argument.

Signed-off-by: Nathan Lynch 
Signed-off-by: Paul Mackerras

[POWERPC] Remove old interface find_path_device

2007-04-24T12:08:59+00:00

Replaced by of_find_node_by_path.

Signed-off-by: Stephen Rothwell 
Signed-off-by: Paul Mackerras

[POWERPC] Rename get_property to of_get_property: arch/powerpc

2007-04-12T17:55:19+00:00

Signed-off-by: Stephen Rothwell 
Acked-by: Benjamin Herrenschmidt 
Signed-off-by: Paul Mackerras

Storage class should be before const qualifier

2007-02-17T19:11:19+00:00

The C99 specification states in section 6.11.5:

The placement of a storage-class specifier other than at the
beginning of the declaration specifiers in a declaration is an
obsolescent feature.

Signed-off-by: Tobias Klauser 
Signed-off-by: Adrian Bunk

[POWERPC] Add rtas_service_present() helper

2006-12-08T06:10:22+00:00

To test for the existence of an RTAS function, we typically do:

   foo_token = rtas_token("foo");
   if (foo_token == RTAS_UNKNOWN_SERVICE)
      return;

Add a rtas_service_present method, which provides a more conventional
boolean interface for testing the existence of an RTAS method.

Signed-off-by: Nathan Lynch 
Signed-off-by: Paul Mackerras

[POWERPC] Move rtas_stop_self() into platforms/pseries/hotplug-cpu.c

2006-12-08T04:55:54+00:00

As the first step in consolidating the pseries hotplug cpu code,
create platforms/pseries/hotplug-cpu.c and move rtas_stop_self()
into it.  Do the rtas token initialisation in a new initcall, rather
than rtas_initialize().

Signed-off-by: Michael Ellerman 
Acked-by: Linas Vepstas 
Signed-off-by: Paul Mackerras

[POWERPC] Wrap cpu_die() with CONFIG_HOTPLUG_CPU

2006-12-04T09:39:29+00:00

Per email discussion, it appears that rtas_stop_self()
and pSeries_mach_cpu_die() should not be compiled if
CONFIG_HOTPLUG_CPU is not defined. This patch adds
#ifdefs around these bits of code.

Signed-off-by: Linas Vepstas 
Signed-off-by: Paul Mackerras

[POWERPC] reboot when panic_timout is set

2006-08-25T03:30:26+00:00

Only call into RTAS when booted with panic=0 because the RTAS call
does not return.  The system has to be rebooted via the HMC or via the
management console right now.  This is cumbersome and not what the
default panic=180 is supposed to do.

Signed-off-by: Olaf Hering 
Signed-off-by: Paul Mackerras