Commit Graph

1264 Commits

Author SHA1 Message Date
David S. Miller
44b50e5a1a sparc64: Fix missing devices due to PCI bridge test in of_create_pci_dev().
Just like in the arch/sparc64/kernel/of_device.c code fix commit
071d7f4c3b411beae08d27656e958070c43b78b4 ("sparc64: Fix SMP bootup
with CONFIG_STACK_DEBUG or ftrace.") we have to check the OF device
node name for "pci" instead of relying upon the 'device_type' property
being there on all PCI bridges.

Tested by Meelis Roos, and confirmed to make the PCI QFE devices
reappear on the E3500 system.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-22 15:51:54 -07:00
David S. Miller
7ee766d8fb sparc64: Fix disappearing PCI devices on e3500.
Based upon a bug report by Meelis Roos.

The OF device layer builds properties by matching bus types and
applying 'range' properties as appropriate, up to the root.

The match for "PCI" busses is looking at the 'device_type' property,
and this does work %99 of the time.

But on an E3500 system with a PCI QFE card, the DEC 21153 bridge
sitting above the QFE network interface devices has a 'name' of "pci",
but it completely lacks a 'device_type' property.  So we don't match
it as a PCI bus, and subsequently we end up with no resource values at
all for the devices sitting under that DEC bridge.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-20 22:00:40 -07:00
David S. Miller
9843099ff4 sparc64: Fix SMP bootup with CONFIG_STACK_DEBUG or ftrace.
Based upon a report by Meelis Roos.

Any function call can try to access the current
thread register via the _mcount hooks when the kernel
is built with -pg (via ftrace or STACK_DEBUG).

That can't be setup properly very early on during
the bootup of other cpus for sun4u and some early
sun4v systems.

So add notrace markers to these specific functions, so
that _mcount doesn't get invoked too early.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-16 11:44:00 -07:00
David S. Miller
f948cc6ab9 sparc64: Fix OOPS in psycho_pcierr_intr_other().
We no longer put the top-level PCI controller device into the
PCI layer device list.  So pbm->pci_bus->self is always NULL.

Therefore, use direct PCI config space accesses to get at
the PCI controller's PCI_STATUS register.

Tested by Meelis Roos.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-16 09:53:42 -07:00
David S. Miller
7d4ee289d1 sparc: Fix user_regset 'n' field values.
As noticed by Russell King, we were not setting this properly
to the number of entries, but rather the total size.

This results in the core dumping code allocating waayyyy too
much memory.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-12 15:55:44 -07:00
David S. Miller
80a56ab626 sparc64: Fix PCI error interrupt registry on PSYCHO.
We need to pass IRQF_SHARED, otherwise we get things like:

IRQ handler type mismatch for IRQ 33
current handler: PSYCHO_UE
Call Trace:
 [000000000048394c] request_irq+0xac/0x120
 [00000000007c5f6c] psycho_scan_bus+0x98/0x158
 [00000000007c2bc0] pcibios_init+0xdc/0x12c
 [0000000000426a5c] do_one_initcall+0x1c/0x160
 [00000000007c0180] kernel_init+0x9c/0xfc
 [0000000000427050] kernel_thread+0x30/0x60
 [00000000006ae1d0] rest_init+0x10/0x60

on e3500 and similar systems.

On a single board, the UE interrupts of two Psycho nodes
are funneled through the same interrupt, from of_debug=3
dump:

/pci@b,4000: direct translate 2ee --> 21
 ...
/pci@b,2000: direct translate 2ee --> 21

Decimal "33" mentioned above is the hex "21" mentioned here.

Thanks to Meelis Roos for dumps and testing.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-12 15:13:15 -07:00
David S. Miller
ebfb2c6340 sparc64: Fix interrupt register calculations on Psycho and Sabre.
Use the IMAP offset calculation for OBIO devices as documented in the
programmer's manual.  Which is "0x10000 + ((ino & 0x1f) << 3)"

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-10 14:08:27 -07:00
David S. Miller
2eb2f77900 sparc64: Disable timer interrupts in fixup_irqs().
When a CPU is offlined, we leave the timer interrupts disabled
because fixup_irqs() does not explicitly take care of that case.

Fix this by invoking tick_ops->disable_irq().

Based upon analysis done by Paul E. McKenney.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-08 17:21:07 -07:00
Paul E. McKenney
4d084617fb sparc64: Prevent sparc64 from invoking irq handlers on offline CPUs
Make sparc64 refrain from clearing a given to-be-offlined CPU's bit in the
cpu_online_mask until it has processed pending irqs.  This change
prevents other CPUs from being blindsided by an apparently offline CPU
nevertheless changing globally visible state.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-03 02:15:30 -07:00
David S. Miller
e5bd1c3fdd sparc64: Fix IPI call locking.
When I switched sparc64 over to the generic helpers for
smp_call_function(), I didn't convert the dinky call_lock
we had.

Use ipi_call_lock() and ipi_call_unlock().

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-09-03 02:14:39 -07:00
David S. Miller
dbb8c35d90 sparc64: setup_valid_addr_bitmap_from_pavail() should be __init
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-30 02:04:45 -07:00
David S. Miller
e3c71a3291 sparc: Fix resource flags for PCI children in OF device tree.
When a device is under an EBUS or ISA bus, the resource flags
don't get set properly.

Fix this by re-evaluating the resource flags at each level of
bus as we apply ranges on the way to the root.  And let PCI
override any existing flags setting, but don't let the
default flags calculator make such overrides.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 22:59:10 -07:00
David S. Miller
a39f2f466f sparc64: Make NUMA depend upon SMP.
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-24 19:41:43 -07:00
David S. Miller
f2b6079464 sparc64: Fix cmdline_memory_size handling bugs.
First, lmb_enforce_memory_limit() interprets it's argument
(mostly, heh) as a size limit not an address limit.  So pass
the raw cmdline_memory_size value into it.  And we don't
need to check it against zero, lmb_enforce_memory_limit() does
that for us.

Next, free_initmem() needs special handling when the kernel
command line trims the available memory.  The problem case is
if the trimmed out memory is where the kernel image itself
resides.

When that memory is trimmed out, we don't add those physical
ram areas to the sparsemem active ranges, amongst other things.
Which means that this free_initmem() code will free up invalid
page structs, resulting in either crashes or hangs.

Just quick fix this by not freeing initmem at all if "mem="
was given on the boot command line.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-14 01:45:41 -07:00
David S. Miller
c918dcce92 sparc64: Fix overshoot in nid_range().
If 'start' does not begin on a page boundary, we can overshoot
past 'end'.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-14 01:41:39 -07:00
David S. Miller
6f63e781ea sparc64: Handle stack trace attempts before irqstacks are setup.
Things like lockdep can try to do stack backtraces before
the irqstack blocks have been setup.  So don't try to match
their ranges so early on.

Also, remove unused variable in save_stack_trace().

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-13 17:20:04 -07:00
David S. Miller
4f70f7a91b sparc64: Implement IRQ stacks.
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-12 18:33:56 -07:00
David S. Miller
c7498081a6 sparc64: Fix recursion in stack overflow detection handling.
The calls down into prom_printf() when we detect an overflowed stack
can recurse again since the overflow stack will be "below" the current
kernel stack limit.

Prevent this by just returning straight if we are on the stack
overflow safe stack already.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-12 02:03:49 -07:00
David S. Miller
b6b7922fbd sparc64: Don't MAGIC_SYSRQ ifdef smp_fetch_global_regs and support code.
Based upon a report and initial patch by Friedrich Oslage.

The intention is to provide this facility for
__trigger_all_cpu_backtrace even if MAGIC_SYSRQ is not set.

The only part that should have MAGIC_SYSRQ ifdef protection is the
sparc_globalreg_op sysrq regitration and immediate code.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-09 16:25:26 -07:00
David S. Miller
433c5f7068 sparc64: Fix end-of-stack checking in save_stack_trace().
Bug reported by Alexander Beregalov.

Before we dereference the stack frame or try to peek at the
pt_regs magic value, make sure the entire object is within
the kernel stack bounds.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-07 23:04:37 -07:00
Stephen Rothwell
764f2579d9 sparc: don't use asm/of_device.h
Use linux/of_device.h instead.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-07 15:33:36 -07:00
David S. Miller
ea771bd51c sparc64: Use kernel/uid16.c helpers instead of own copy.
Noticed by Adrian Bunk.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-06 23:11:08 -07:00
David S. Miller
ae583885bf sparc64: Remove all cpumask_t local variables in xcall dispatch.
All of the xcall delivery implementation is cpumask agnostic, so
we can pass around pointers to const cpumask_t objects everywhere.

The sad remaining case is the argument to arch_send_call_function_ipi().

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 16:56:15 -07:00
David S. Miller
ed4d9c66eb sparc64: Kill error_mask from hypervisor_xcall_deliver().
It can eat up a lot of stack space when NR_CPUS is large.
We retain some of it's functionality by reporting at least one
of the cpu's which are seen in error state.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 16:47:57 -07:00
David S. Miller
90f7ae8a55 sparc64: Build cpu list and mondo block at top-level xcall_deliver().
Then modify all of the xcall dispatch implementations get passed and
use this information.

Now all of the xcall dispatch implementations do not need to be mindful
of details such as "is current cpu in the list?" and "is cpu online?"

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 16:42:58 -07:00
David S. Miller
c02a5119e8 sparc64: Disable local interrupts around xcall_deliver_impl() invocation.
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 16:18:40 -07:00
David S. Miller
deb16999e4 sparc64: Make all xcall_deliver's go through common helper function.
This just facilitates the next changeset where we'll be building
the cpu list and mondo block in this helper function.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 16:16:20 -07:00
David S. Miller
43f589235e sparc64: Always allocate the send mondo blocks, even on non-sun4v.
The idea is that we'll use this cpu list array and mondo block
even for non-hypervisor platforms.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 16:13:51 -07:00
David S. Miller
91a4231cc2 sparc64: Make smp_cross_call_masked() take a cpumask_t pointer.
Ideally this could be simplified further such that we could pass
the pointer down directly into the xcall_deliver() implementation.

But if we do that we need to do the "cpu_online(cpu)" and
"cpu != self" checks down in those functions.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 13:51:40 -07:00
David S. Miller
24445a4ac9 sparc64: Directly call xcall_deliver() in smp_start_sync_tick_client.
We know the cpu is online and not the current cpu here.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 13:51:40 -07:00
David S. Miller
1992663053 sparc64: Call xcall_deliver() directly in some cases.
For these cases the callers make sure:

1) The cpus indicated are online.

2) The current cpu is not in the list of indicated cpus.

Therefore we can pass a pointer to the mask directly.

One of the motivations in this transformation is to make use of
"&cpumask_of_cpu(cpu)" which evaluates to a pointer to constant
data in the kernel and thus takes up no stack space.

Hopefully someone in the future will change the interface of
arch_send_call_function_ipi() such that it passes a const cpumask_t
pointer so that this will optimize ever further.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 13:51:39 -07:00
David S. Miller
cd5bc89deb sparc64: Use cpumask_t pointers and for_each_cpu_mask_nr() in xcall_deliver.
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 13:51:38 -07:00
David S. Miller
622824dbb5 sparc64: Use xcall_deliver() consistently.
There remained some spots still vectoring to the appropriate
*_xcall_deliver() function manually.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 13:51:37 -07:00
David S. Miller
5e0797e5b8 sparc64: Use function pointer for cross-call sending.
Initialize it using the smp_setup_processor_id() hook.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 13:51:37 -07:00
Huang Weiyi
abd9e69828 arch/sparc64/kernel/signal.c: removed duplicated #include
Removed duplicated #include <linux/tracehook.h> in
arch/sparc64/kernel/signal.c.

Signed-off-by: Huang Weiyi <weiyi.huang@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 13:51:36 -07:00
David S. Miller
a3cf5e6b6f sparc64: Need to disable preemption around smp_tsb_sync().
Based upon a bug report by Mariusz Kozlowski

It uses smp_call_function_masked() now, which has a preemption-disabled
requirement.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-04 13:51:35 -07:00
David S. Miller
0a4949c441 sparc64: Do not clobber %g7 in setcontext() trap.
That's the userland thread register, so we should never try to change
it like this.

Based upon glibc bug nptl/6577 and suggestions by Jakub Jelinek.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-31 20:40:46 -07:00
David S. Miller
dbf3e95067 sparc64: Kill __show_regs().
The story is that what we used to do when we actually used
smp_report_regs() is that if you specifically only wanted to have the
current cpu's registers dumped you would call "__show_regs()"
otherwise you would call show_regs() which also invoked
smp_report_regs().

Now that we killed off smp_report_regs() there is no longer any
reason to have these two routines, just show_regs() is sufficient.

Also kill off a stray declaration of show_regs() in sparc64_ksym.c

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-31 20:33:43 -07:00
David S. Miller
9c636e30a3 sparc64: Kill smp_report_regs().
All the call sites are #if 0'd out and we have a much more
useful global cpu dumping facility these days.  smp_report_regs()
is way too verbose to be usable.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-31 01:06:02 -07:00
David S. Miller
a014821340 sparc64: Kill VERBOSE_SHOWREGS code.
It just clutters everything up and even though I wrote that hack I
can't remember having used it in the last 5 years or so.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-31 00:58:35 -07:00
David S. Miller
09ee167cbf sparc64: Hook up trigger_all_cpu_backtrace().
We already have code that does this, but it is only currently attached
to sysrq-'y'.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-30 22:35:00 -07:00
David S. Miller
5afe27380b sparc64: Make global reg dumping even more useful.
Record one more level of stack frame program counter.

Particularly when lockdep and all sorts of spinlock debugging is
enabled, figuring out the caller of spin_lock() is difficult when the
cpu is stuck on the lock.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-30 21:57:59 -07:00
David S. Miller
71fc324b5b sparc64: Kill isa_bus_type.
I forgot to delete this when I removed the ISA bus layer
from the sparc ports.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-29 23:47:17 -07:00
David S. Miller
17b6f586b8 sparc64: Fix global reg snapshotting on self-cpu.
We were picking %i7 out of the wrong register window
stack slot.

Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-28 00:44:29 -07:00
Roland McGrath
768225868c sparc64: tracehook: CONFIG_HAVE_ARCH_TRACEHOOK
The sparc64 arch code has all the prerequisites, so set HAVE_ARCH_TRACEHOOK.

Signed-off-by: Roland McGrath <roland@redhat.com>
2008-07-27 17:33:07 -07:00
Roland McGrath
95698466cf sparc64: tracehook_signal_handler
Call the standard hook after setting up signal handlers.

Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-27 17:32:35 -07:00
Roland McGrath
e35a8925e0 sparc64: tracehook: TIF_NOTIFY_RESUME
This adds TIF_NOTIFY_RESUME support for sparc64.
When set, we call tracehook_notify_resume() on the way to user mode.

Signed-off-by: Roland McGrath <roland@redhat.com>
2008-07-27 17:32:19 -07:00
Roland McGrath
73ccefab8a sparc64: tracehook syscall
This changes sparc64 syscall tracing to use the new tracehook.h entry
points.

[ Add assembly changes to force an immediate -ENOSYS return from
  the system call when syscall_trace() returns non-zero at syscall
  entry.  -DaveM ]

Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-27 17:28:55 -07:00
Sam Ravnborg
a439fe51a1 sparc, sparc64: use arch/sparc/include
The majority of this patch was created by the following script:

***
ASM=arch/sparc/include/asm
mkdir -p $ASM
git mv include/asm-sparc64/ftrace.h $ASM
git rm include/asm-sparc64/*
git mv include/asm-sparc/* $ASM
sed -ie 's/asm-sparc64/asm/g' $ASM/*
sed -ie 's/asm-sparc/asm/g' $ASM/*
***

The rest was an update of the top-level Makefile to use sparc
for header files when sparc64 is being build.
And a small fixlet to pick up the correct unistd.h from
sparc64 code.

Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
2008-07-27 23:00:59 +02:00
Johannes Weiner
471a54239a sparc64: use generic show_mem()
Remove arch-specific show_mem() in favor of the generic version.

This also removes the following redundant information display:

	- free swap pages, printed by show_swap_cache_info()
	- pages in swapcache, printed by show_swap_cache_info()
	- dirty pages, writeback pages, mapped pages, slab pages,
	  pagetables pages, printed by show_free_areas()

where show_mem() calls show_free_areas(), which calls
show_swap_cache_info().

Signed-off-by: Johannes Weiner <hannes@saeurebad.de>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-26 12:00:10 -07:00