There's a bug in the IOAPIC code for level-triggered interrupts. Its
relatively easy to trigger by sharing (virtio-blk + usbtablet was the
testcase, initially reported by Gerd von Egidy).

The "remote_irr" variable is used to indicate accepted but not yet acked
interrupts. Its cleared from the EOI handler.

Problem is that the EOI handler clears remote_irr unconditionally, even
if it reinjected another pending interrupt.

In that case, kvm_ioapic_set_irq() proceeds to ioapic_service() which
sets remote_irr even if it failed to inject (since the IRR was high due
to EOI reinjection).

Since the TMR bit has been cleared by the first EOI, the second one
fails to clear remote_irr.

End result is interrupt line dead.

Fix it by setting remote_irr only if a new pending interrupt has been
generated (and the TMR bit for vector in question set).
Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

Shadows for large guests can take a long time to tear down, so reschedule
occasionally to avoid softlockup warnings.
Signed-off-by: NAvi Kivity <avi@qumranet.com>

Clear CR4.VMXE in hardware_disable. There's no reason to leave it set
after doing a VMXOFF.

VMware Workstation 6.5 checks CR4.VMXE as a proxy for whether the CPU is
in VMX mode, so leaving VMXE set means we'll refuse to power on. With this
change the user can power on after unloading the kvm-intel module. I
tested on kvm-67 and kvm-69.
Signed-off-by: NEli Collins <ecollins@vmware.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

Migrate the PIT timer to the physical CPU which vcpu0 is scheduled on,
similarly to what is done for the LAPIC timers, otherwise PIT interrupts
will be delayed until an unrelated event causes an exit.
Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

This code shouldn't be hit anyways, but when it is, it's useful to have a
little more information about the failure.
Signed-off-by: NHollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

gfn_to_page() and kvm_release_page_clean() are called from other contexts with
mmap_sem locked only for reading.
Signed-off-by: NHollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

We're not calling kmap() now, so we shouldn't call kunmap() either. This has no
practical effect in the non-highmem case, which is why it hasn't caused more
obvious problems.

Pointed out by Anthony Liguori.
Signed-off-by: NHollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

Somehow these load/store instructions got missed before, but weren't used by
the guest so didn't break anything.
Signed-off-by: NHollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: NChristian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

This was left behind from some code movement.
Signed-off-by: NHollis Blanchard <hollisb@us.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

The call to add_timer was issued before local_int.lock was taken and before
timer_due was set to 0. If the timer expires before the lock is being taken,
the timer function will set timer_due to 1 and exit before the vcpu falls
asleep. Depending on other external events, the vcpu might sleep forever.
This fix pulls setting timer_due to the beginning of the function before
add_timer, which ensures correct behavior.
Signed-off-by: NCarsten Otte <cotte@de.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

If the guest accesses non-existing memory, the sie64a function returns
-EFAULT. We must check the return value and send a program check to the
guest if the sie instruction faulted, otherwise the guest will loop at
the faulting code.
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NCarsten Otte <cotte@de.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

The current code delivers pending interrupts before it checks for
need_resched. On a busy host, this can lead to a longer interrupt
latency if the interrupt is injected while the process is scheduled
away. This patch moves delivering the interrupt _after_ schedule(),
which makes more sense.
Signed-off-by: NCarsten Otte <cotte@de.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

The low-level interrupt handler on s390 checks for _TIF_WORK_INT and
exits the guest context, if work is pending.
TIF_WORK_INT is defined as_TIF_SIGPENDING | _TIF_NEED_RESCHED |
 _TIF_MCCK_PENDING. Currently the sie loop checks for signals and
reschedule, but it does not check for machine checks. That means that
we exit the guest context if a machine check is pending, but we do not
handle the machine check.
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>
CC: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: NCarsten Otte <cotte@de.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

There are potential locking problem in enable_sie. We take the task_lock
and the mmap_sem. As exit_mm uses the same locks vice versa, this triggers
a lockdep warning.
The second problem is that dup_mm and mmput might sleep, so we must not
hold the task_lock at that moment.

The solution is to dup the mm unconditional and use the task_lock before and
afterwards to check  if we can use the new mm. dup_mm and mmput are called
outside the task_lock, but we run update_mm while holding the task_lock,
protection us against ptrace.
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NCarsten Otte <cotte@de.ibm.com>
Acked-by: NMartin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

diag 0x44 is the common way on s390 to yield the cpu to the hypervisor.
It is called by the guest in cpu_relax and in the spinlock code to
yield to other guest cpus.

This semantic is similar to yield. Lets replace the call to schedule with
yield to make sure that current is really yielding.
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NCarsten Otte <cotte@de.ibm.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

The hypercall instructions on Intel and AMD are different.  KVM allows the
guest to choose one or the other (the default is Intel), and if the guest
chooses incorrectly, KVM will patch it at runtime to select the correct
instruction.  This allows live migration between Intel and AMD machines.

This patching occurs in the x86 emulator.  The current code also executes
the hypercall.  Unfortunately, the tail end of the x86 emulator code also
executes, overwriting the return value of the hypercall with the original
contents of rax (which happens to be the hypercall number).

Fix not by executing the hypercall in the emulator context; instead let the
guest reissue the patched instruction and execute the hypercall via the
normal path.
Signed-off-by: NAvi Kivity <avi@qumranet.com>

Only copy in the data actually requested by the instruction emulation
and zero pad the destination register first. This avoids the problem
where emulated mmio access got garbled data from ld2.acq instructions
in the vga console driver.
Signed-off-by: NJes Sorensen <jes@sgi.com>
Acked-by: NXiantao Zhang <xiantao.zhang@intel.com>
Signed-off-by: NAvi Kivity <avi@qumranet.com>

* master.kernel.org:/home/rmk/linux-2.6-arm:
  [ARM] pxa: fix tosa.c build error
  [ARM] 5067/1: _raw_write_can_lock macro bugfix
  [ARM] 5070/1: pxa: add GPIO104_PSKTSEL to pxa27x MFP configuration
  [ARM] 5068/1: PXA2xx Additional gpio definitions
  [ARM] 5066/2: EM-X270: Fix DM9000 IRQ flags initialisation
  [ARM] 5065/2: CM-X270: Fix DM9000 IRQ flags initialisation
  [ARM] 5062/1: pxa: remove unused definition of CONFIG_ARCH_COTULLA_IDP
  [ARM] 5060/1: remove unnecessary include of asm/io.h
  [ARM] fix AT91 include loops

Fairly simple. "dev_use" was being allocated as a zero length array
because of bad math on 64-bit systems, causing a crash in
find_first_zero_bit(). One-liner follows:
Signed-off-by: NBen Collins <ben.collins@canonical.com>
Acked-by: NPierre Ossman <drzeus@drzeus.cx>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

* git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/v4l-dvb: (48 commits)
  V4L/DVB (8001): dib0070: fix dib0070_attach when !CONFIG_DVB_TUNER_DIB0070
  V4L/DVB (8000): tda827x: fix NULL pointer in tda827xa_lna_gain
  V4L/DVB (7990): Fix entry for PowerColor RA 330 and make it run with firmware version 2.7
  V4L/DVB (7983): tda18271_calc_rf_cal must return the return value of tda18271_lookup_map
  V4L/DVB (7978): cx18: explicitly test for XC2028 tuner
  V4L/DVB (7977): cx18: fix init order and remove duplicate open_on_first_use.
  V4L/DVB (7975): saa7134_empress
  V4L/DVB (7974): fix MEDIA_TUNER && FW_LOADER build error
  V4L/DVB (7972): or51132.c: unaligned
  V4L/DVB (7971): usb: unaligned
  V4L/DVB (7970): mix trivial endianness annotations
  V4L/DVB (7969): m920x: unaligned access
  V4L/DVB (7968): zoran: endianness annotations
  V4L/DVB (7967): bt8xx: unaligned access
  V4L/DVB (7966): cx18: direct dereferencing of iomem
  V4L/DVB (7965): annotate bcx_riscmem
  V4L/DVB (7964): cx18 iomem annotations
  V4L/DVB (7963): ivtv: trivial annotations
  V4L/DVB (7962): ttusb endianness annotations and fixes
  V4L/DVB (7961): fix endianness bug in dib0700_devices.c
  ...

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  Fix divide by zero error in build_clear_page() and build_copy_page()
  [MIPS] Fix typo in header guard
  [MIPS] Fix build error - Delete debugging crap that crept in with CMP
  [MIPS] Add accessors for random register.
  [MIPS] IP27: misc fixes
  [MIPS] IP27: Fix clockevent setup
  [MIPS] IP27: Fix bootmem memory setup
  [MIPS] remove CONFIG_CPU_R4000 line from Makefile
  [MIPS] Fix check for valid stack pointer during backtrace
  [MIPS] Add missing braces to pte_mkyoung
  [MIPS] R4700: Fix build_tlb_probe_entry
  [MIPS] Alchemy: dbdma: add API to delete custom DDMA device ids.
  [MIPS] Alchemy: export get_au1x00_speed for modules

* git://git.infradead.org/~dwmw2/mtd-2.6.26:
  [MTD] [MAPS] Fix cmdlineparse handling in mapping files
  [MTD] [NAND] pxa: fix incorrect calling of pxa3xx_nand_config() on resume path

Suggest how to deal with patch modifications caused by
merging or back-porting when you're a maintainer.
Signed-off-by: NWilly Tarreau <w@1wt.eu>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

- Don't trust a length which is greater than the working buffer.
  An invalid length could cause overflow when calculating buffer size
  for decoding oid.

- An oid length of zero is invalid and allows for an off-by-one error when
  decoding oid because the first subid actually encodes first 2 subids.

- A primitive encoding may not have an indefinite length.

Thanks to Wei Wang from McAfee for report.

Cc: Steven French <sfrench@us.ibm.com>
Cc: stable@kernel.org
Acked-by: NPatrick McHardy <kaber@trash.net>
Signed-off-by: NChris Wright <chrisw@sous-sol.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Alias brd to rd in the hope of helping legacy users. Suggested by Jan.
Signed-off-by: NNick Piggin <npiggin@suse.de>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Here is an attempt to translate the prompt and help text into something
which is legible and, as a bonus, correct.
Signed-off-by: NStefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

ip_fast_csum() requires a memory clobber on its inline asm as it accesses
memory in a fashion that gcc can't predict.

The GCC manual says:

 If your assembler instructions access memory in an unpredictable
 fashion, add `memory' to the list of clobbered registers.  This will
 cause GCC to not keep memory values cached in registers across the
 assembler instruction and not optimize stores or loads to that memory.

The bug hasn't been noticed in FRV, but it has been seen in PA-RISC.
Signed-off-by: NDavid Howells <dhowells@redhat.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

We don't need to reserve "unset" resources.  Trying to reserve
them results in messages like this, which are ugly but harmless:

    system 00:08: iomem range 0x0-0x0 could not be reserved

Future PNP patches will remove use of IORESOURCE_UNSET, but
we still need it for now.
Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Signed-off-by: NYoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NVegard Nossum <vegard.nossum@gmail.com>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

- fix PCI interrupt assignment by emulating ioc3 interrupt pin register
- use pci_probe_only mode
- select correct page size in bridge
- remove no longer needed ioc3_sio_init() code

[Ralf: Fix for 64kB or larger pagesizes]
Signed-off-by: NThomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Fix breakage introduced by converting hub_rt to clockevent.
Signed-off-by: NThomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Changes in the generic bootmem code broke memory setup for IP27. This
patch fixes this by replacing lots of special IP27 code with generic
bootmon code. This has been tested only on a single node.
Signed-off-by: NThomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

The existing options are named CONFIG_CPU_R4300 and CONFIG_CPU_R4X00,
and they are directly below.
Reported-by: NRobert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: NAdrian Bunk <bunk@kernel.org>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

The newly added check for valid stack pointer address breaks at least for
64bit kernels.  Use __get_user() for accessing stack content to avoid crashes,
when doing the backtrace.
Signed-off-by: NThomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Only the version pte_mkyoung for 36-bit pagetables on 32-bit hw was
affected and with this bug being around since November 29, 2004 there
is evidence to suport the assumption it was benign ;-)
Signed-off-by: NIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Treat R4700 like R4600 in build_tlb_probe_entry. Without this fix kernel
will lock up.
Signed-off-by: NThomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Add API to delete custom DDMA device ids create with
au1xxx_ddma_device_add().
Signed-off-by: NManuel Lauss <mano@roarinelk.homelinux.net>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>