The operation for releasing a remapping entry is iommu
specific too.
Signed-off-by: NJoerg Roedel <joerg.roedel@amd.com>
Acked-by: NYinghai Lu <yinghai@kernel.org>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: NJoerg Roedel <joerg.roedel@amd.com>

The function to set interrupt affinity with interrupt
remapping enabled is Intel specific too. So move it to the
irq_remap_ops too.
Signed-off-by: NJoerg Roedel <joerg.roedel@amd.com>
Acked-by: NYinghai Lu <yinghai@kernel.org>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: NJoerg Roedel <joerg.roedel@amd.com>

The IOAPIC setup routine for interrupt remapping is VT-d
specific. Move it to the irq_remap_ops and add a call helper
function.
Signed-off-by: NJoerg Roedel <joerg.roedel@amd.com>
Acked-by: NYinghai Lu <yinghai@kernel.org>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: NJoerg Roedel <joerg.roedel@amd.com>

This patch introduces irq_remap_ops to hold implementation
specific function pointer to handle interrupt remapping. As
the first part the initialization functions for VT-d are
converted to these ops.
Signed-off-by: NJoerg Roedel <joerg.roedel@amd.com>
Acked-by: NYinghai Lu <yinghai@kernel.org>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: NJoerg Roedel <joerg.roedel@amd.com>

Xen dom0 needs to paravirtualize IO operations to the IO APIC,
so add a io_apic_ops for it to intercept.  Do this as ops
structure because there's at least some chance that another
paravirtualized environment may want to intercept these.
Signed-off-by: NJeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Acked-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: jwboyer@redhat.com
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/1332385090-18056-2-git-send-email-konrad.wilk@oracle.com
[ Made all the affected code easier on the eyes ]
Signed-off-by: NIngo Molnar <mingo@kernel.org>

This patch removes dead code from certain .config variations.

When CONFIG_GENERIC_PENDING_IRQ=n irq move and reenable code is
never get executed, nor do_unmask_irq variable updates its init
value. Move the code under CONFIG_GENERIC_PENDING_IRQ macro.
Signed-off-by: NAlexander Gordeev <agordeev@redhat.com>
Link: http://lkml.kernel.org/r/20120320141935.GA24806@dhcp-26-207.brq.redhat.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

With the recent changes to clear_IO_APIC_pin() which tries to
clear remoteIRR bit explicitly, some of the users started to see
"Unable to reset IRR for apic .." messages.

Close look shows that these are related to bogus IO-APIC entries
which return's all 1's for their io-apic registers. And the
above mentioned error messages are benign. But kernel should
have ignored such io-apic's in the first place.

Check if register 0, 1, 2 of the listed io-apic are all 1's and
ignore such io-apic.
Reported-by: NÁlvaro Castillo <midgoon@gmail.com>
Tested-by: NJon Dufresne <jon@jondufresne.org>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: yinghai@kernel.org
Cc: kernel-team@fedoraproject.org
Cc: Josh Boyer <jwboyer@redhat.com>
Cc: <stable@kernel.org>
Link: http://lkml.kernel.org/r/1331577393.31585.94.camel@sbsiddha-desk.sc.intel.com
[ Performed minor cleanup of affected code. ]
Signed-off-by: NIngo Molnar <mingo@elte.hu>

We use MP IRQs for SFI presented timer interrupts, we should
also set mp_bus_not_pci for MP_ISA_BUS so that pin_2_irq mapping
is correct.
Signed-off-by: NJacob Pan <jacob.jun.pan@linux.intel.com>
Signed-off-by: NDirk Brandewie <dirk.brandewie@gmail.com>
Link: http://lkml.kernel.org/n/tip-8h3rc1igpp8ir94aas69qmhk@git.kernel.orgSigned-off-by: NIngo Molnar <mingo@elte.hu>

Using compile time NR_LEGACY_IRQS causes the wrong gsi-irq
mapping on non-PC platforms, such as Moorestown. This patch uses
legacy_pic abstraction to set the correct number of legacy
interrupts at runtime. For Moorestown, nr_legacy_irqs = 0. We
have 1:1 mapping for gsi-irq even within the legacy irq range.
Signed-off-by: NJacob Pan <jacob.jun.pan@linux.intel.com>
Signed-off-by: NDirk Brandewie <dirk.brandewie@gmail.com>
Link: http://lkml.kernel.org/n/tip-kzvj4xp9tmicuoqoh2w05iay@git.kernel.orgSigned-off-by: NIngo Molnar <mingo@elte.hu>

On some of the recent Intel SNB platforms, by default bios is pre-enabling
x2apic mode in the cpu with out setting up interrupt-remapping.
This case was resulting in the kernel to panic as the cpu is already in
x2apic mode but the OS was not able to enable interrupt-remapping (which
is a pre-req for using x2apic capability).

On these platforms all the apic-ids are < 255 and the kernel can fallback to
xapic mode if the bios has not enabled interrupt-remapping (which is
mostly the case if the bios has not exported interrupt-remapping tables to the
OS).
Reported-by: NBerck E. Nash <flyboy@gmail.com>
Signed-off-by: NYinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/20111222014632.600418637@sbsiddha-desk.sc.intel.comSigned-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: NH. Peter Anvin <hpa@linux.intel.com>

Interrupts notify the idle exit state before calling irq_enter().
But the notifier code calls rcu_read_lock() and this is not
allowed while rcu is in an extended quiescent state. We need
to wait for irq_enter() -> rcu_idle_exit() to be called before
doing so otherwise this results in a grumpy RCU:

[    0.099991] WARNING: at include/linux/rcupdate.h:194 __atomic_notifier_call_chain+0xd2/0x110()
[    0.099991] Hardware name: AMD690VM-FMH
[    0.099991] Modules linked in:
[    0.099991] Pid: 0, comm: swapper Not tainted 3.0.0-rc6+ #255
[    0.099991] Call Trace:
[    0.099991]  <IRQ>  [<ffffffff81051c8a>] warn_slowpath_common+0x7a/0xb0
[    0.099991]  [<ffffffff81051cd5>] warn_slowpath_null+0x15/0x20
[    0.099991]  [<ffffffff817d6fa2>] __atomic_notifier_call_chain+0xd2/0x110
[    0.099991]  [<ffffffff817d6ff1>] atomic_notifier_call_chain+0x11/0x20
[    0.099991]  [<ffffffff81001873>] exit_idle+0x43/0x50
[    0.099991]  [<ffffffff81020439>] smp_apic_timer_interrupt+0x39/0xa0
[    0.099991]  [<ffffffff817da253>] apic_timer_interrupt+0x13/0x20
[    0.099991]  <EOI>  [<ffffffff8100ae67>] ? default_idle+0xa7/0x350
[    0.099991]  [<ffffffff8100ae65>] ? default_idle+0xa5/0x350
[    0.099991]  [<ffffffff8100b19b>] amd_e400_idle+0x8b/0x110
[    0.099991]  [<ffffffff810cb01f>] ? rcu_enter_nohz+0x8f/0x160
[    0.099991]  [<ffffffff810019a0>] cpu_idle+0xb0/0x110
[    0.099991]  [<ffffffff817a7505>] rest_init+0xe5/0x140
[    0.099991]  [<ffffffff817a7468>] ? rest_init+0x48/0x140
[    0.099991]  [<ffffffff81cc5ca3>] start_kernel+0x3d1/0x3dc
[    0.099991]  [<ffffffff81cc5321>] x86_64_start_reservations+0x131/0x135
[    0.099991]  [<ffffffff81cc5412>] x86_64_start_kernel+0xed/0xf4
Signed-off-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Andy Henroid <andrew.d.henroid@intel.com>
Signed-off-by: NPaul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: NJosh Triplett <josh@joshtriplett.org>

with "apic=verbose" the print_IO_APIC() function tries to print
IRQ to pin mappings for every active irq. It assumes chip_data
is of type irq_cfg and may cause an oops if not.

As the print_IO_APIC() is called from a late_initcall other
chained irq chips may already be registered with custom
chip_data information, causing an oops. This is the case with
intel MID SoC devices with gpio demuxers registered as irq_chips.
Signed-off-by: NMathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: NAlan Cox <alan@linux.intel.com>
[ -v2: fixed build failure ]
Signed-off-by: NIngo Molnar <mingo@elte.hu>

nr_legacy_irqs is set in probe_nr_irqs_gsi, we should not clear
it after that. Otherwise, the result is that MSI irqs will be
allocated from the wrong range for the systems without legacy
PIC.
Signed-off-by: NJacob Pan <jacob.jun.pan@linux.intel.com>
Signed-off-by: NDirk Brandewie <dirk.brandewie@gmail.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

Sparseirq got introduced in v2.6.28 and Thomas did a huge cleanup
around v2.6.38 that eliminated basically all disadvantages
of it.

So we can remove non-sparseirq support now and simplify
our IRQ degrees of freedom a bit.
Suggested-and-acked-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NYinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/4E95E21D.6090200@oracle.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

While looking at the code, apic_id sometime is referred to index
of ioapic, but sometime is used for phys apic id. and some even
use apic for real apic id. It is very confusing.

So try to limit apic_id or ioapic_id to be real apic id for
ioapic, and use ioapic_idx for ioapic index in the array.

-v2: Suggested by Ingo, use ioapic_idx consistently, instead of ioapic
Signed-off-by: NYinghai Lu <yinghai@kernel.org>
Cc: Naga Chumbalkar <nagananda.chumbalkar@hp.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/4E9542DC.3090509@oracle.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

It is getting too big after the interrupt remaping entries debug
print out was added.

Original print_IO_APIC() becomes print_IO_APICs().
New print_IO_APIC() will only print one ioapic's registers

As a side-effect this clean-up also made checkpatch.pl happier.
Signed-off-by: NYinghai Lu <yinghai@kernel.org>
Cc: Naga Chumbalkar <nagananda.chumbalkar@hp.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/4E9542D3.5000008@oracle.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

While checking irte dump in dmesg, the print out is confusing
ioapic index with real io apic id:

IOAPIC[0]: Set routing entry (1-1 -> 0x31 -> IRQ 1 Mode:0
Active:0 Dest:1) IOAPIC[1]: Set IRTE entry (P:1 FPD:0 Dst_Mode:1
Redir_hint:1 Trig_Mode:0 Dlvry_Mode:1 Avail:0 Vector:31
Dest:00000001 SID:00FF SQ:0 SVT:1) IOAPIC[0]: Set routing entry
(1-2 -> 0x30 -> IRQ 0 Mode:0 Active:0 Dest:1) IOAPIC[1]: Set IRTE entry (P:1 FPD:0 Dst_Mode:1 Redir_hint:1 Trig_Mode:0 Dlvry_Mode:1 Avail:0 Vector:30 Dest:00000001 SID:00FF SQ:0 SVT:1)

The system's first ioapic id is 1.

This commit:

| commit 3040db92
| Author: Naga Chumbalkar <nagananda.chumbalkar@hp.com>
| Date:   Tue Jul 12 21:17:41 2011 +0000
|
|    x86, ioapic: Print IRTE when IR is enabled

Confused apic_id with the ioapic ID - fix it.
Signed-off-by: NYinghai Lu <yinghai@kernel.org>
Cc: Naga Chumbalkar <nagananda.chumbalkar@hp.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/4E9542C8.8040209@oracle.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Ingo pointed out that setup_ioapic_entry() is way too big now.

Split the intr-remap code out into setup_ir_ioapic_entry().

Also pass struct io_apic_irq_attr * instead of 5 parameters
in those two functions.

At last in setup_ir_ioapic_entry() we don't need to panic.
Signed-off-by: NYinghai Lu <yinghai@kernel.org>
Cc: Naga Chumbalkar <nagananda.chumbalkar@hp.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/4E9542BB.4070807@oracle.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Do not expand that struct, and just pass pointer to reduce the
number of parameters in related functions.
Signed-off-by: NYinghai Lu <yinghai@kernel.org>
Cc: Naga Chumbalkar <nagananda.chumbalkar@hp.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/4E9542B1.7050800@oracle.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Consolidate the io-apic EOI code in clear_IO_APIC_pin() and
eoi_ioapic_irq().
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: Thomas Renninger <trenn@suse.de>
Cc: Rafael Wysocki <rjw@novell.com>
Cc: Maciej W. Rozycki <macro@linux-mips.org>
Cc: lchiquitto@novell.com
Cc: jbeulich@novell.com
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110825190657.259696697@sbsiddha-desk.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

For older IO-APIC's, we were clearing the remote-IRR by changing
the RTE trigger mode to edge and then back to level. We wanted
to mask the RTE during this process, so we were essentially
doing mask+edge and then to unmask+level.

As part of the commit ca64c47c,
we moved this EOI process earlier where the IO-APIC RTE is
masked. So we were wrongly unmasking it in the eoi_ioapic_irq().

So change the remote-IRR clear sequence in eoi_ioapic_irq() to
mask + edge and then restore the previous RTE entry which will
restore the mask status as well as the level trigger.
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: Maciej W. Rozycki <macro@linux-mips.org>
Cc: Thomas Renninger <trenn@suse.de>
Cc: Rafael Wysocki <rjw@novell.com>
Cc: lchiquitto@novell.com
Cc: jbeulich@novell.com
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110825190657.210286410@sbsiddha-desk.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

In the kdump scenario mentioned below, we can have a case where
the device using level triggered interrupt will not generate any
interrupts in the kdump kernel.

1. IO-APIC sends a level triggered interrupt to the CPU's local APIC.

2. Kernel crashed before the CPU services this interrupt, leaving
   the remote-IRR in the IO-APIC set.

3. kdump kernel boot sequence does clear_IO_APIC() as part of IO-APIC
   initialization. But this fails to reset remote-IRR bit of the
   IO-APIC RTE as the remote-IRR bit is read-only.

4. Device using that level triggered entry can't generate any
   more interrupts because of the remote-IRR bit.

In clear_IO_APIC_pin(), check if the remote-IRR bit is set and if
so do an explicit attempt to clear it (by doing EOI write on
modern io-apic's and changing trigger mode to edge/level on
older io-apic's). Also before doing the explicit EOI to the
io-apic, ensure that the trigger mode is indeed set to level.
This will enable the explicit EOI to the io-apic to reset the
remote-IRR bit.
Tested-by: NLeonardo Chiquitto <lchiquitto@novell.com>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Fixes: https://bugzilla.novell.com/show_bug.cgi?id=701686
Cc: Rafael Wysocki <rjw@novell.com>
Cc: Maciej W. Rozycki <macro@linux-mips.org>
Cc: Thomas Renninger <trenn@suse.de>
Cc: jbeulich@novell.com
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110825190657.157502602@sbsiddha-desk.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Change the CONFIG_DMAR to CONFIG_INTEL_IOMMU to be consistent
with the other IOMMU options.

Rename the CONFIG_INTR_REMAP to CONFIG_IRQ_REMAP to match the
irq subsystem name.

And define the CONFIG_DMAR_TABLE for the common ACPI DMAR
routines shared by both CONFIG_INTEL_IOMMU and CONFIG_IRQ_REMAP.
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: yinghai@kernel.org
Cc: youquan.song@intel.com
Cc: joerg.roedel@amd.com
Cc: tony.luck@intel.com
Cc: dwmw2@infradead.org
Link: http://lkml.kernel.org/r/20110824001456.558630224@sbsiddha-desk.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Define irq_remap_modify_chip_defaults() and remove the duplicate
code, cleanup the unnecessary ifdefs.
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: yinghai@kernel.org
Cc: youquan.song@intel.com
Cc: joerg.roedel@amd.com
Cc: tony.luck@intel.com
Cc: dwmw2@infradead.org
Link: http://lkml.kernel.org/r/20110824001456.499225692@sbsiddha-desk.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

IRQ set affinity routine is same for the IO-APIC IRQ's aswell as
the MSI IRQ's in the presence of interrupt-remapping. This is
because we modify the interrupt-remapping table entry and
doesn't touch the IO-APIC RTE or the MSI entry.

So remove the ir_msi_set_affinity() and re-use the
ir_ioapic_set_affinity()
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: yinghai@kernel.org
Cc: youquan.song@intel.com
Cc: joerg.roedel@amd.com
Cc: tony.luck@intel.com
Cc: dwmw2@infradead.org
Link: http://lkml.kernel.org/r/20110824001456.452760446@sbsiddha-desk.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

When IR (interrupt remapping) is enabled print_IO_APIC() displays output according
to legacy RTE (redirection table entry) definitons:

 NR Dst Mask Trig IRR Pol Stat Dmod Deli Vect:
 00 00  1    0    0   0   0    0    0    00
 01 00  0    0    0   0   0    0    0    01
 02 00  0    0    0   0   0    0    0    02
 03 00  1    0    0   0   0    0    0    03
 04 00  1    0    0   0   0    0    0    04
 05 00  1    0    0   0   0    0    0    05
 06 00  1    0    0   0   0    0    0    06
...

The above output is as per Sec 3.2.4 of the IOAPIC datasheet:
82093AA I/O Advanced Programmable Interrupt Controller (IOAPIC):
http://download.intel.com/design/chipsets/datashts/29056601.pdf

Instead the output should display the fields as discussed in Sec 5.5.1
of the VT-d specification:

(Intel Virtualization Technology for Directed I/O:
http://download.intel.com/technology/computing/vptech/Intel(r)_VT_for_Direct_IO.pdf)

After the fix:
 NR Indx Fmt Mask Trig IRR Pol Stat Indx2 Zero Vect:
 00 0000 0   1    0    0   0   0    0     0    00
 01 000F 1   0    0    0   0   0    0     0    01
 02 0001 1   0    0    0   0   0    0     0    02
 03 0002 1   1    0    0   0   0    0     0    03
 04 0011 1   1    0    0   0   0    0     0    04
 05 0004 1   1    0    0   0   0    0     0    05
 06 0005 1   1    0    0   0   0    0     0    06
...
Signed-off-by: NNaga Chumbalkar <nagananda.chumbalkar@hp.com>
Link: http://lkml.kernel.org/r/20110712211658.2939.93123.sendpatchset@nchumbalkar.americas.cpqcorp.netSigned-off-by: NH. Peter Anvin <hpa@linux.intel.com>

When "apic=debug" is used as a boot parameter, Linux prints the IOAPIC routing
entries in "dmesg". Below is output from IOAPIC whose apic_id is 8:

# dmesg | grep "routing entry"
IOAPIC[8]: Set routing entry (8-1 -> 0x31 -> IRQ 1 Mode:0 Active:0 Dest:0)
IOAPIC[8]: Set routing entry (8-2 -> 0x30 -> IRQ 0 Mode:0 Active:0 Dest:0)
IOAPIC[8]: Set routing entry (8-3 -> 0x33 -> IRQ 3 Mode:0 Active:0 Dest:0)
...

Similarly, when IR (interrupt remapping) is enabled, and the IRTE
(interrupt remapping table entry) is set up we should display it.

After the fix:

# dmesg | grep IRTE
IOAPIC[8]: Set IRTE entry (P:1 FPD:0 Dst_Mode:0 Redir_hint:1 Trig_Mode:0 Dlvry_Mode:0 Avail:0 Vector:31 Dest:00000000 SID:00F1 SQ:0 SVT:1)
IOAPIC[8]: Set IRTE entry (P:1 FPD:0 Dst_Mode:0 Redir_hint:1 Trig_Mode:0 Dlvry_Mode:0 Avail:0 Vector:30 Dest:00000000 SID:00F1 SQ:0 SVT:1)
IOAPIC[8]: Set IRTE entry (P:1 FPD:0 Dst_Mode:0 Redir_hint:1 Trig_Mode:0 Dlvry_Mode:0 Avail:0 Vector:33 Dest:00000000 SID:00F1 SQ:0 SVT:1)
...

The IRTE is defined in Sec 9.5 of the Intel VT-d Specification.
Signed-off-by: NNaga Chumbalkar <nagananda.chumbalkar@hp.com>
Link: http://lkml.kernel.org/r/20110712211704.2939.71291.sendpatchset@nchumbalkar.americas.cpqcorp.netSigned-off-by: NH. Peter Anvin <hpa@linux.intel.com>

The code in setup_ioapic_irq() determines the Destination Field,
so why not also include it in the debug printk output that gets
displayed when the boot parameter "apic=debug" is used.

Before the change, "dmesg" will show:

 IOAPIC[0]: Set routing entry (8-1 -> 0x31 -> IRQ 1 Mode:0 Active:0)
 IOAPIC[0]: Set routing entry (8-2 -> 0x30 -> IRQ 0 Mode:0 Active:0)
 IOAPIC[0]: Set routing entry (8-3 -> 0x33 -> IRQ 3 Mode:0 Active:0) ...

After the change, you will see:

 IOAPIC[0]: Set routing entry (8-1 -> 0x31 -> IRQ 1 Mode:0 Active:0 Dest:0)
 IOAPIC[0]: Set routing entry (8-2 -> 0x30 -> IRQ 0 Mode:0 Active:0 Dest:0)
 IOAPIC[0]: Set routing entry (8-3 -> 0x33 -> IRQ 3 Mode:0 Active:0 Dest:0) ...
Signed-off-by: NNaga Chumbalkar <nagananda.chumbalkar@hp.com>
Link: http://lkml.kernel.org/r/20110708184603.2734.91071.sendpatchset@nchumbalkar.americas.cpqcorp.netSigned-off-by: NIngo Molnar <mingo@elte.hu>

When IOAPIC data is displayed in "dmesg" with the help of the
boot parameter "apic=debug" certain values are not formatted
correctly wrt their size.

In the "dmesg" snippet below, note that the output for "max
redirection entries", and "IO APIC version" which are each
defined to be just 8-bits long are displayed as 2 bytes in
length. Similarly, "Dst" under the "IRQ redirection table"
should only be 8-bits long.

IO APIC #0......
...
...
.... register #01: 00170020
.......     : max redirection entries: 0017
.......     : PRQ implemented: 0
.......     : IO APIC version: 0020
...
...
.... IRQ redirection table:
 NR Dst Mask Trig IRR Pol Stat Dmod Deli Vect:
 00 000 1    0    0   0   0    0    0    00
 01 000 0    0    0   0   0    0    0    31
 02 000 0    0    0   0   0    0    0    30
 03 000 1    0    0   0   0    0    0    33
...
...

Do some formatting clean up, so you will see output like below:

IO APIC #0......
...
...
.... register #01: 00170020
.......     : max redirection entries: 17
.......     : PRQ implemented: 0
.......     : IO APIC version: 20
...
...
.... IRQ redirection table:
 NR Dst Mask Trig IRR Pol Stat Dmod Deli Vect:
 00 00  1    0    0   0   0    0    0    00
 01 00  0    0    0   0   0    0    0    31
 02 00  0    0    0   0   0    0    0    30
 03 00  1    0    0   0   0    0    0    33
...
...
Signed-off-by: NNaga Chumbalkar <nagananda.chumbalkar@hp.com>
Link: http://lkml.kernel.org/r/20110708184557.2734.61830.sendpatchset@nchumbalkar.americas.cpqcorp.netSigned-off-by: NIngo Molnar <mingo@elte.hu>

To view IOAPIC data you could boot with "apic=debug".

When booting in such a way then the kernel will dump the
IO-APIC's registers, for example:

NR Dst Mask Trig IRR Pol Stat Dmod Deli Vect:
 00 000 1    0    0   0   0    0    0    00
 01 000 0    0    0   0   0    0    0    31
 02 000 0    0    0   0   0    0    0    30
 03 000 0    0    0   0   0    0    0    33
 04 000 0    0    0   0   0    0    0    34
 05 000 0    0    0   0   0    0    0    35
 06 000 0    0    0   0   0    0    0    36
 07 000 0    0    0   0   0    0    0    37
 08 000 0    0    0   0   0    0    0    38
 09 000 0    1    0   0   0    0    0    39
 0a 000 0    0    0   0   0    0    0    3A
 0b 000 0    0    0   0   0    0    0    3B
 0c 000 0    0    0   0   0    0    0    3C
 0d 000 0    0    0   0   0    0    0    3D
 0e 000 0    0    0   0   0    0    0    3E
 0f 000 0    0    0   0   0    0    0    3F
 10 000 1    0    0   0   0    0    0    00
 11 000 1    0    0   0   0    0    0    00
 12 000 1    0    0   0   0    0    0    00
 13 000 1    0    0   0   0    0    0    00
 14 000 1    0    0   0   0    0    0    00
 15 000 1    0    0   0   0    0    0    00
 16 000 1    0    0   0   0    0    0    00
 17 000 1    0    0   0   0    0    0    00

Delaying the call to print_ICs() gives better results:

NR Dst Mask Trig IRR Pol Stat Dmod Deli Vect:
 00 000 1    0    0   0   0    0    0    00
 01 000 0    0    0   0   0    0    0    31
 02 000 0    0    0   0   0    0    0    30
 03 000 1    0    0   0   0    0    0    33
 04 000 1    0    0   0   0    0    0    34
 05 000 1    0    0   0   0    0    0    35
 06 000 1    0    0   0   0    0    0    36
 07 000 1    0    0   0   0    0    0    37
 08 000 0    0    0   0   0    0    0    38
 09 000 0    1    0   0   0    0    0    39
 0a 000 1    0    0   0   0    0    0    3A
 0b 000 1    0    0   0   0    0    0    3B
 0c 000 0    0    0   0   0    0    0    3C
 0d 000 1    0    0   0   0    0    0    3D
 0e 000 1    0    0   0   0    0    0    3E
 0f 000 1    0    0   0   0    0    0    3F
 10 000 1    1    0   1   0    0    0    29
 11 000 1    0    0   0   0    0    0    00
 12 000 1    0    0   0   0    0    0    00
 13 000 1    0    0   0   0    0    0    00
 14 000 0    1    0   1   0    0    0    51
 15 000 1    0    0   0   0    0    0    00
 16 000 0    1    0   1   0    0    0    61
 17 000 0    1    0   1   0    0    0    59

Notice that the entries beyond interrupt input signal 0x0f also
get populated and arent just the hw-initialization default of
all zeroes.
Signed-off-by: NNaga Chumbalkar <nagananda.chumbalkar@hp.com>
Link: http://lkml.kernel.org/r/20110708083555.2598.42216.sendpatchset@nchumbalkar.americas.hpqcorp.netSigned-off-by: NIngo Molnar <mingo@elte.hu>

In mask/restore_ioapic_entries() we should be restoring ioapic
entries when ioapics[apic].saved_registers is not NULL.

Fix the typo and address the resume hang regression reported by
Linus.

This was not found sooner because the systems where these
changes were tested on kept the IO-APIC entries intact over
resume.
Reported-and-tested-by: NLinus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: Daniel J Blueman <daniel.blueman@gmail.com>
Link: http://lkml.kernel.org/r/1306259131.7171.7.camel@sbsiddha-MOBL3.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: daniel.blueman@gmail.com
Link: http://lkml.kernel.org/r/20110518233158.089978277@sbsiddha-MOBL3.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: daniel.blueman@gmail.com
Link: http://lkml.kernel.org/r/20110518233157.994002011@sbsiddha-MOBL3.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: daniel.blueman@gmail.com
Link: http://lkml.kernel.org/r/20110518233157.909013179@sbsiddha-MOBL3.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: daniel.blueman@gmail.com
Link: http://lkml.kernel.org/r/20110518233157.830697056@sbsiddha-MOBL3.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Introduce struct ioapic with nr_registers field.

This will pave way for consolidating different MAX_IO_APICS
arrays into it.
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: daniel.blueman@gmail.com
Link: http://lkml.kernel.org/r/20110518233157.744315519@sbsiddha-MOBL3.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Code flow for enabling interrupt-remapping has its own routines
for saving and restoring io-apic RTE's. ioapic suspend/resume
code flow also has similar routines. Remove the duplicate code.
Tested-by: NDaniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/20110518233157.673130611@sbsiddha-MOBL3.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Code flow for enabling interrupt-remapping was
allocating/freeing buffers for saving/restoring io-apic RTE's.
ioapic suspend/resume code uses boot time allocated
ioapic_saved_data that is a perfect match for reuse here.

This will remove the unnecessary allocation/free of the
temporary buffers during suspend/resume of interrupt-remapping
enabled platforms aswell as paving the way for further code
consolidation.
Tested-by: NDaniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/20110518233157.574469296@sbsiddha-MOBL3.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

This allows re-using this buffer for enabling
interrupt-remapping during boot and resume. And thus allow for
consolidating the code between ioapic suspend/resume and
interrupt-remapping.
Tested-by: NDaniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/20110518233157.481404505@sbsiddha-MOBL3.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

Fix a potential deadlock when resuming; here the calling
function has disabled interrupts, so we cannot sleep.

Change the memory allocation flag from GFP_KERNEL to GFP_ATOMIC.

TODO: We can do away with this memory allocation during resume
      by reusing the ioapic suspend/resume code that uses boot time
      allocated buffers, but we want to keep this -stable patch
      simple.
Signed-off-by: NDaniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: NSuresh Siddha <suresh.b.siddha@intel.com>
Cc: <stable@kernel.org> # v2.6.38/39
Link: http://lkml.kernel.org/r/20110518233157.385970138@sbsiddha-MOBL3.sc.intel.comSigned-off-by: NIngo Molnar <mingo@elte.hu>