| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
commit 816afe4ff98ee10b1d30fd66361be132a0a5cee6 upstream.
We still patch SMP instructions to UP variants if we boot with a
single CPU, but not at any other time. In particular, not if we
unplug CPUs to return to a single cpu.
Paul McKenney points out:
mean offline overhead is 6251/48=130.2 milliseconds.
If I remove the alternatives_smp_switch() from the offline
path [...] the mean offline overhead is 550/42=13.1 milliseconds
Basically, we're never going to get those 120ms back, and the
code is pretty messy.
We get rid of:
1) The "smp-alt-once" boot option. It's actually "smp-alt-boot", the
documentation is wrong. It's now the default.
2) The skip_smp_alternatives flag used by suspend.
3) arch_disable_nonboot_cpus_begin() and arch_disable_nonboot_cpus_end()
which were only used to set this one flag.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Paul McKenney <paul.mckenney@us.ibm.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/87vcgwwive.fsf@rustcorp.com.au
Signed-off-by: Ingo Molnar <mingo@kernel.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 660da7c9228f685b2ebe664f9fd69aaddcc420b5 upstream.
We can use PCID if the CPU has PCID and PGE and we're not on Xen.
By itself, this has no effect. A followup patch will start using PCID.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Reviewed-by: Nadav Amit <nadav.amit@gmail.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: linux-mm@kvack.org
Link: http://lkml.kernel.org/r/6327ecd907b32f79d5aa0d466f04503bbec5df88.1498751203.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
[Hugh Dickins: Backported to 3.2:
- arch/x86/xen/enlighten_pv.c (not in this tree)
- arch/x86/xen/enlighten.c (patched instead of that)]
Signed-off-by: Hugh Dickins <hughd@google.com>
[Borislav Petkov: Fix bad backport to disable PCID on Xen]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit b7a584598aea7ca73140cb87b40319944dd3393f upstream.
On Xen PV, regs->flags doesn't reliably reflect IOPL and the
exit-to-userspace code doesn't change IOPL. We need to context
switch it manually.
I'm doing this without going through paravirt because this is
specific to Xen PV. After the dust settles, we can merge this with
the 32-bit code, tidy up the iopl syscall implementation, and remove
the set_iopl pvop entirely.
Fixes XSA-171.
Reviewewd-by: Jan Beulich <JBeulich@suse.com>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andrew Cooper <andrew.cooper3@citrix.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Jan Beulich <JBeulich@suse.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/693c3bd7aeb4d3c27c92c622b7d0f554a458173c.1458162709.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
[bwh: Backported to 3.2:
- Use xen_pv_domain() directly as X86_FEATURE_XENPV is not defined
- Adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 6a1f513776b78c994045287073e55bae44ed9f8c upstream.
On a cancelled suspend the vcpu_info location does not change (it's
still in the per-cpu area registered by xen_vcpu_setup()). So do not
call xen_hvm_init_shared_info() which would make the kernel think its
back in the shared info. With the wrong vcpu_info, events cannot be
received and the domain will hang after a cancelled suspend.
Signed-off-by: Charles Ouyang <ouyangzhaowei@huawei.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 64c98e7f49100b637cd20a6c63508caed6bbba7a upstream.
Sanitizing the e820 map may produce extra E820 entries which would result in
the topmost E820 entries being removed. The removed entries would typically
include the top E820 usable RAM region and thus result in the domain having
signicantly less RAM available to it.
Fix by allowing sanitize_e820_map to use the full size of the allocated E820
array.
Signed-off-by: Malcolm Crossley <malcolm.crossley@citrix.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
[bwh: Backported to 3.2:
s/xen_e820_map_entries/memmap.nr_entries/; s/xen_e820_map/map/g]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit aa1acff356bbedfd03b544051f5b371746735d89 upstream.
The update_va_mapping hypercall can fail if the VA isn't present
in the guest's page tables. Under certain loads, this can
result in an OOPS when the target address is in unpopulated vmap
space.
While we're at it, add comments to help explain what's going on.
This isn't a great long-term fix. This code should probably be
changed to use something like set_memory_ro.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andrew Cooper <andrew.cooper3@citrix.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: David Vrabel <dvrabel@cantab.net>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Jan Beulich <jbeulich@suse.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: security@kernel.org <security@kernel.org>
Cc: xen-devel <xen-devel@lists.xen.org>
Link: http://lkml.kernel.org/r/0b0e55b995cda11e7829f140b833ef932fcabe3a.1438291540.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 8d5999df35314607c38fbd6bdd709e25c3a4eeab upstream.
If the timer irqs are resumed during device resume it is possible in
certain circumstances for the resume to hang early on, before device
interrupts are resumed. For an Ubuntu 14.04 PVHVM guest this would
occur in ~0.5% of resume attempts.
It is not entirely clear what is occuring the point of the hang but I
think a task necessary for the resume calls schedule_timeout(),
waiting for a timer interrupt (which never arrives). This failure may
require specific tasks to be running on the other VCPUs to trigger
(processes are not frozen during a suspend/resume if PREEMPT is
disabled).
Add IRQF_EARLY_RESUME to the timer interrupts so they are resumed in
syscore_resume().
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 3bc38cbceb85881a8eb789ee1aa56678038b1909 upstream.
If there are UNUSABLE regions in the machine memory map, dom0 will
attempt to map them 1:1 which is not permitted by Xen and the kernel
will crash.
There isn't anything interesting in the UNUSABLE region that the dom0
kernel needs access to so we can avoid making the 1:1 mapping and
treat it as RAM.
We only do this for dom0, as that is where tboot case shows up.
A PV domU could have an UNUSABLE region in its pseudo-physical map
and would need to be handled in another patch.
This fixes a boot failure on hosts with tboot.
tboot marks a region in the e820 map as unusable and the dom0 kernel
would attempt to map this region and Xen does not permit unusable
regions to be mapped by guests.
(XEN) 0000000000000000 - 0000000000060000 (usable)
(XEN) 0000000000060000 - 0000000000068000 (reserved)
(XEN) 0000000000068000 - 000000000009e000 (usable)
(XEN) 0000000000100000 - 0000000000800000 (usable)
(XEN) 0000000000800000 - 0000000000972000 (unusable)
tboot marked this region as unusable.
(XEN) 0000000000972000 - 00000000cf200000 (usable)
(XEN) 00000000cf200000 - 00000000cf38f000 (reserved)
(XEN) 00000000cf38f000 - 00000000cf3ce000 (ACPI data)
(XEN) 00000000cf3ce000 - 00000000d0000000 (reserved)
(XEN) 00000000e0000000 - 00000000f0000000 (reserved)
(XEN) 00000000fe000000 - 0000000100000000 (reserved)
(XEN) 0000000100000000 - 0000000630000000 (usable)
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
[v1: Altered the patch and description with domU's with UNUSABLE regions]
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 0b0c002c340e78173789f8afaa508070d838cf3d upstream.
... because the "clock_event_device framework" already accounts for idle
time through the "event_handler" function pointer in
xen_timer_interrupt().
The patch is intended as the completion of [1]. It should fix the double
idle times seen in PV guests' /proc/stat [2]. It should be orthogonal to
stolen time accounting (the removed code seems to be isolated).
The approach may be completely misguided.
[1] https://lkml.org/lkml/2011/10/6/10
[2] http://lists.xensource.com/archives/html/xen-devel/2010-08/msg01068.html
John took the time to retest this patch on top of v3.10 and reported:
"idle time is correctly incremented for pv and hvm for the normal
case, nohz=off and nohz=idle." so lets put this patch in.
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: John Haxby <john.haxby@oracle.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 7f1fc268c47491fd5e63548f6415fc8604e13003 upstream.
If a user did:
echo 0 > /sys/devices/system/cpu/cpu1/online
echo 1 > /sys/devices/system/cpu/cpu1/online
we would (this a build with DEBUG enabled) get to:
smpboot: ++++++++++++++++++++=_---CPU UP 1
.. snip..
smpboot: Stack at about ffff880074c0ff44
smpboot: CPU1: has booted.
and hang. The RCU mechanism would kick in an try to IPI the CPU1
but the IPIs (and all other interrupts) would never arrive at the
CPU1. At first glance at least. A bit digging in the hypervisor
trace shows that (using xenanalyze):
[vla] d4v1 vec 243 injecting
0.043163027 --|x d4v1 intr_window vec 243 src 5(vector) intr f3
] 0.043163639 --|x d4v1 vmentry cycles 1468
] 0.043164913 --|x d4v1 vmexit exit_reason PENDING_INTERRUPT eip ffffffff81673254
0.043164913 --|x d4v1 inj_virq vec 243 real
[vla] d4v1 vec 243 injecting
0.043164913 --|x d4v1 intr_window vec 243 src 5(vector) intr f3
] 0.043165526 --|x d4v1 vmentry cycles 1472
] 0.043166800 --|x d4v1 vmexit exit_reason PENDING_INTERRUPT eip ffffffff81673254
0.043166800 --|x d4v1 inj_virq vec 243 real
[vla] d4v1 vec 243 injecting
there is a pending event (subsequent debugging shows it is the IPI
from the VCPU0 when smpboot.c on VCPU1 has done
"set_cpu_online(smp_processor_id(), true)") and the guest VCPU1 is
interrupted with the callback IPI (0xf3 aka 243) which ends up calling
__xen_evtchn_do_upcall.
The __xen_evtchn_do_upcall seems to do *something* but not acknowledge
the pending events. And the moment the guest does a 'cli' (that is the
ffffffff81673254 in the log above) the hypervisor is invoked again to
inject the IPI (0xf3) to tell the guest it has pending interrupts.
This repeats itself forever.
The culprit was the per_cpu(xen_vcpu, cpu) pointer. At the bootup
we set each per_cpu(xen_vcpu, cpu) to point to the
shared_info->vcpu_info[vcpu] but later on use the VCPUOP_register_vcpu_info
to register per-CPU structures (xen_vcpu_setup).
This is used to allow events for more than 32 VCPUs and for performance
optimizations reasons.
When the user performs the VCPU hotplug we end up calling the
the xen_vcpu_setup once more. We make the hypercall which returns
-EINVAL as it does not allow multiple registration calls (and
already has re-assigned where the events are being set). We pick
the fallback case and set per_cpu(xen_vcpu, cpu) to point to the
shared_info->vcpu_info[vcpu] (which is a good fallback during bootup).
However the hypervisor is still setting events in the register
per-cpu structure (per_cpu(xen_vcpu_info, cpu)).
As such when the events are set by the hypervisor (such as timer one),
and when we iterate in __xen_evtchn_do_upcall we end up reading stale
events from the shared_info->vcpu_info[vcpu] instead of the
per_cpu(xen_vcpu_info, cpu) structures. Hence we never acknowledge the
events that the hypervisor has set and the hypervisor keeps on reminding
us to ack the events which we never do.
The fix is simple. Don't on the second time when xen_vcpu_setup is
called over-write the per_cpu(xen_vcpu, cpu) if it points to
per_cpu(xen_vcpu_info).
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 7918c92ae9638eb8a6ec18e2b4a0de84557cccc8 upstream.
When we online the CPU, we get this splat:
smpboot: Booting Node 0 Processor 1 APIC 0x2
installing Xen timer for CPU 1
BUG: sleeping function called from invalid context at /home/konrad/ssd/konrad/linux/mm/slab.c:3179
in_atomic(): 1, irqs_disabled(): 0, pid: 0, name: swapper/1
Pid: 0, comm: swapper/1 Not tainted 3.9.0-rc6upstream-00001-g3884fad #1
Call Trace:
[<ffffffff810c1fea>] __might_sleep+0xda/0x100
[<ffffffff81194617>] __kmalloc_track_caller+0x1e7/0x2c0
[<ffffffff81303758>] ? kasprintf+0x38/0x40
[<ffffffff813036eb>] kvasprintf+0x5b/0x90
[<ffffffff81303758>] kasprintf+0x38/0x40
[<ffffffff81044510>] xen_setup_timer+0x30/0xb0
[<ffffffff810445af>] xen_hvm_setup_cpu_clockevents+0x1f/0x30
[<ffffffff81666d0a>] start_secondary+0x19c/0x1a8
The solution to that is use kasprintf in the CPU hotplug path
that 'online's the CPU. That is, do it in in xen_hvm_cpu_notify,
and remove the call to in xen_hvm_setup_cpu_clockevents.
Unfortunatly the later is not a good idea as the bootup path
does not use xen_hvm_cpu_notify so we would end up never allocating
timer%d interrupt lines when booting. As such add the check for
atomic() to continue.
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
online/offline
commit 66ff0fe9e7bda8aec99985b24daad03652f7304e upstream.
While we don't use the spinlock interrupt line (see for details
commit f10cd522c5fbfec9ae3cc01967868c9c2401ed23 -
xen: disable PV spinlocks on HVM) - we should still do the proper
init / deinit sequence. We did not do that correctly and for the
CPU init for PVHVM guest we would allocate an interrupt line - but
failed to deallocate the old interrupt line.
This resulted in leakage of an irq_desc but more importantly this splat
as we online an offlined CPU:
genirq: Flags mismatch irq 71. 0002cc20 (spinlock1) vs. 0002cc20 (spinlock1)
Pid: 2542, comm: init.late Not tainted 3.9.0-rc6upstream #1
Call Trace:
[<ffffffff811156de>] __setup_irq+0x23e/0x4a0
[<ffffffff81194191>] ? kmem_cache_alloc_trace+0x221/0x250
[<ffffffff811161bb>] request_threaded_irq+0xfb/0x160
[<ffffffff8104c6f0>] ? xen_spin_trylock+0x20/0x20
[<ffffffff813a8423>] bind_ipi_to_irqhandler+0xa3/0x160
[<ffffffff81303758>] ? kasprintf+0x38/0x40
[<ffffffff8104c6f0>] ? xen_spin_trylock+0x20/0x20
[<ffffffff810cad35>] ? update_max_interval+0x15/0x40
[<ffffffff816605db>] xen_init_lock_cpu+0x3c/0x78
[<ffffffff81660029>] xen_hvm_cpu_notify+0x29/0x33
[<ffffffff81676bdd>] notifier_call_chain+0x4d/0x70
[<ffffffff810bb2a9>] __raw_notifier_call_chain+0x9/0x10
[<ffffffff8109402b>] __cpu_notify+0x1b/0x30
[<ffffffff8166834a>] _cpu_up+0xa0/0x14b
[<ffffffff816684ce>] cpu_up+0xd9/0xec
[<ffffffff8165f754>] store_online+0x94/0xd0
[<ffffffff8141d15b>] dev_attr_store+0x1b/0x20
[<ffffffff81218f44>] sysfs_write_file+0xf4/0x170
[<ffffffff811a2864>] vfs_write+0xb4/0x130
[<ffffffff811a302a>] sys_write+0x5a/0xa0
[<ffffffff8167ada9>] system_call_fastpath+0x16/0x1b
cpu 1 spinlock event irq -16
smpboot: Booting Node 0 Processor 1 APIC 0x2
And if one looks at the /proc/interrupts right after
offlining (CPU1):
70: 0 0 xen-percpu-ipi spinlock0
71: 0 0 xen-percpu-ipi spinlock1
77: 0 0 xen-percpu-ipi spinlock2
There is the oddity of the 'spinlock1' still being present.
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 888b65b4bc5e7fcbbb967023300cd5d44dba1950 upstream.
In the PVHVM path when we do CPU online/offline path we would
leak the timer%d IRQ line everytime we do a offline event. The
online path (xen_hvm_setup_cpu_clockevents via
x86_cpuinit.setup_percpu_clockev) would allocate a new interrupt
line for the timer%d.
But we would still use the old interrupt line leading to:
kernel BUG at /home/konrad/ssd/konrad/linux/kernel/hrtimer.c:1261!
invalid opcode: 0000 [#1] SMP
RIP: 0010:[<ffffffff810b9e21>] [<ffffffff810b9e21>] hrtimer_interrupt+0x261/0x270
.. snip..
<IRQ>
[<ffffffff810445ef>] xen_timer_interrupt+0x2f/0x1b0
[<ffffffff81104825>] ? stop_machine_cpu_stop+0xb5/0xf0
[<ffffffff8111434c>] handle_irq_event_percpu+0x7c/0x240
[<ffffffff811175b9>] handle_percpu_irq+0x49/0x70
[<ffffffff813a74a3>] __xen_evtchn_do_upcall+0x1c3/0x2f0
[<ffffffff813a760a>] xen_evtchn_do_upcall+0x2a/0x40
[<ffffffff8167c26d>] xen_hvm_callback_vector+0x6d/0x80
<EOI>
[<ffffffff81666d01>] ? start_secondary+0x193/0x1a8
[<ffffffff81666cfd>] ? start_secondary+0x18f/0x1a8
There is also the oddity (timer1) in the /proc/interrupts after
offlining CPU1:
64: 1121 0 xen-percpu-virq timer0
78: 0 0 xen-percpu-virq timer1
84: 0 2483 xen-percpu-virq timer2
This patch fixes it.
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 511ba86e1d386f671084b5d0e6f110bb30b8eeb2 upstream.
Invoking arch_flush_lazy_mmu_mode() results in calls to
preempt_enable()/disable() which may have performance impact.
Since lazy MMU is not used on bare metal we can patch away
arch_flush_lazy_mmu_mode() so that it is never called in such
environment.
[ hpa: the previous patch "Fix vmalloc_fault oops during lazy MMU
updates" may cause a minor performance regression on
bare metal. This patch resolves that performance regression. It is
somewhat unclear to me if this is a good -stable candidate. ]
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Link: http://lkml.kernel.org/r/1364045796-10720-2-git-send-email-konrad.wilk@oracle.com
Tested-by: Josh Boyer <jwboyer@redhat.com>
Tested-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Acked-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 76eaca031f0af2bb303e405986f637811956a422 upstream.
There is a loophole between Xen's current implementation of
pv-spinlocks and the scheduler. This was triggerable through
a testcase until v3.6 changed the TLB flushing code. The
problem potentially is still there just not observable in the
same way.
What could happen was (is):
1. CPU n tries to schedule task x away and goes into a slow
wait for the runq lock of CPU n-# (must be one with a lower
number).
2. CPU n-#, while processing softirqs, tries to balance domains
and goes into a slow wait for its own runq lock (for updating
some records). Since this is a spin_lock_irqsave in softirq
context, interrupts will be re-enabled for the duration of
the poll_irq hypercall used by Xen.
3. Before the runq lock of CPU n-# is unlocked, CPU n-1 receives
an interrupt (e.g. endio) and when processing the interrupt,
tries to wake up task x. But that is in schedule and still
on_cpu, so try_to_wake_up goes into a tight loop.
4. The runq lock of CPU n-# gets unlocked, but the message only
gets sent to the first waiter, which is CPU n-# and that is
busily stuck.
5. CPU n-# never returns from the nested interruption to take and
release the lock because the scheduler uses a busy wait.
And CPU n never finishes the task migration because the unlock
notification only went to CPU n-#.
To avoid this and since the unlocking code has no real sense of
which waiter is best suited to grab the lock, just send the IPI
to all of them. This causes the waiters to return from the hyper-
call (those not interrupted at least) and do active spinlocking.
BugLink: http://bugs.launchpad.net/bugs/1011792
Acked-by: Jan Beulich <JBeulich@suse.com>
Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 13d2b4d11d69a92574a55bfd985cfb0ca77aebdc upstream.
This fixes CVE-2013-0228 / XSA-42
Drew Jones while working on CVE-2013-0190 found that that unprivileged guest user
in 32bit PV guest can use to crash the > guest with the panic like this:
-------------
general protection fault: 0000 [#1] SMP
last sysfs file: /sys/devices/vbd-51712/block/xvda/dev
Modules linked in: sunrpc ipt_REJECT nf_conntrack_ipv4 nf_defrag_ipv4
iptable_filter ip_tables ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6
xt_state nf_conntrack ip6table_filter ip6_tables ipv6 xen_netfront ext4
mbcache jbd2 xen_blkfront dm_mirror dm_region_hash dm_log dm_mod [last
unloaded: scsi_wait_scan]
Pid: 1250, comm: r Not tainted 2.6.32-356.el6.i686 #1
EIP: 0061:[<c0407462>] EFLAGS: 00010086 CPU: 0
EIP is at xen_iret+0x12/0x2b
EAX: eb8d0000 EBX: 00000001 ECX: 08049860 EDX: 00000010
ESI: 00000000 EDI: 003d0f00 EBP: b77f8388 ESP: eb8d1fe0
DS: 0000 ES: 007b FS: 0000 GS: 00e0 SS: 0069
Process r (pid: 1250, ti=eb8d0000 task=c2953550 task.ti=eb8d0000)
Stack:
00000000 0027f416 00000073 00000206 b77f8364 0000007b 00000000 00000000
Call Trace:
Code: c3 8b 44 24 18 81 4c 24 38 00 02 00 00 8d 64 24 30 e9 03 00 00 00
8d 76 00 f7 44 24 08 00 00 02 80 75 33 50 b8 00 e0 ff ff 21 e0 <8b> 40
10 8b 04 85 a0 f6 ab c0 8b 80 0c b0 b3 c0 f6 44 24 0d 02
EIP: [<c0407462>] xen_iret+0x12/0x2b SS:ESP 0069:eb8d1fe0
general protection fault: 0000 [#2]
---[ end trace ab0d29a492dcd330 ]---
Kernel panic - not syncing: Fatal exception
Pid: 1250, comm: r Tainted: G D ---------------
2.6.32-356.el6.i686 #1
Call Trace:
[<c08476df>] ? panic+0x6e/0x122
[<c084b63c>] ? oops_end+0xbc/0xd0
[<c084b260>] ? do_general_protection+0x0/0x210
[<c084a9b7>] ? error_code+0x73/
-------------
Petr says: "
I've analysed the bug and I think that xen_iret() cannot cope with
mangled DS, in this case zeroed out (null selector/descriptor) by either
xen_failsafe_callback() or RESTORE_REGS because the corresponding LDT
entry was invalidated by the reproducer. "
Jan took a look at the preliminary patch and came up a fix that solves
this problem:
"This code gets called after all registers other than those handled by
IRET got already restored, hence a null selector in %ds or a non-null
one that got loaded from a code or read-only data descriptor would
cause a kernel mode fault (with the potential of crashing the kernel
as a whole, if panic_on_oops is set)."
The way to fix this is to realize that the we can only relay on the
registers that IRET restores. The two that are guaranteed are the
%cs and %ss as they are always fixed GDT selectors. Also they are
inaccessible from user mode - so they cannot be altered. This is
the approach taken in this patch.
Another alternative option suggested by Jan would be to relay on
the subtle realization that using the %ebp or %esp relative references uses
the %ss segment. In which case we could switch from using %eax to %ebp and
would not need the %ss over-rides. That would also require one extra
instruction to compensate for the one place where the register is used
as scaled index. However Andrew pointed out that is too subtle and if
further work was to be done in this code-path it could escape folks attention
and lead to accidents.
Reviewed-by: Petr Matousek <pmatouse@redhat.com>
Reported-by: Petr Matousek <pmatouse@redhat.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 95a7d76897c1e7243d4137037c66d15cbf2cce76 upstream.
As Mukesh explained it, the MMUEXT_TLB_FLUSH_ALL allows the
hypervisor to do a TLB flush on all active vCPUs. If instead
we were using the generic one (which ends up being xen_flush_tlb)
we end up making the MMUEXT_TLB_FLUSH_LOCAL hypercall. But
before we make that hypercall the kernel will IPI all of the
vCPUs (even those that were asleep from the hypervisor
perspective). The end result is that we needlessly wake them
up and do a TLB flush when we can just let the hypervisor
do it correctly.
This patch gives around 50% speed improvement when migrating
idle guest's from one host to another.
Oracle-bug: 14630170
Tested-by: Jingjie Jiang <jingjie.jiang@oracle.com>
Suggested-by: Mukesh Rathor <mukesh.rathor@oracle.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 1a7bbda5b1ab0e02622761305a32dc38735b90b2 upstream.
We actually do not do anything about it. Just return a default
value of zero and if the kernel tries to write anything but 0
we BUG_ON.
This fixes the case when an user tries to suspend the machine
and it blows up in save_processor_state b/c 'read_cr8' is set
to NULL and we get:
kernel BUG at /home/konrad/ssd/linux/arch/x86/include/asm/paravirt.h:100!
invalid opcode: 0000 [#1] SMP
Pid: 2687, comm: init.late Tainted: G O 3.6.0upstream-00002-gac264ac-dirty #4 Bochs Bochs
RIP: e030:[<ffffffff814d5f42>] [<ffffffff814d5f42>] save_processor_state+0x212/0x270
.. snip..
Call Trace:
[<ffffffff810733bf>] do_suspend_lowlevel+0xf/0xac
[<ffffffff8107330c>] ? x86_acpi_suspend_lowlevel+0x10c/0x150
[<ffffffff81342ee2>] acpi_suspend_enter+0x57/0xd5
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit cd0608e71e9757f4dae35bcfb4e88f4d1a03a8ab upstream.
The hypervisor will trap it. However without this patch,
we would crash as the .read_tscp is set to NULL. This patch
fixes it and sets it to the native_read_tscp call.
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 8d54db795dfb1049d45dc34f0dddbc5347ec5642 upstream.
The hypervisor is in charge of allocating the proper "NUMA" memory
and dealing with the CPU scheduler to keep them bound to the proper
NUMA node. The PV guests (and PVHVM) have no inkling of where they
run and do not need to know that right now. In the future we will
need to inject NUMA configuration data (if a guest spans two or more
NUMA nodes) so that the kernel can make the right choices. But those
patches are not yet present.
In the meantime, disable the NUMA capability in the PV guest, which
also fixes a bootup issue. Andre says:
"we see Dom0 crashes due to the kernel detecting the NUMA topology not
by ACPI, but directly from the northbridge (CONFIG_AMD_NUMA).
This will detect the actual NUMA config of the physical machine, but
will crash about the mismatch with Dom0's virtual memory. Variation of
the theme: Dom0 sees what it's not supposed to see.
This happens with the said config option enabled and on a machine where
this scanning is still enabled (K8 and Fam10h, not Bulldozer class)
We have this dump then:
NUMA: Warning: node ids are out of bound, from=-1 to=-1 distance=10
Scanning NUMA topology in Northbridge 24
Number of physical nodes 4
Node 0 MemBase 0000000000000000 Limit 0000000040000000
Node 1 MemBase 0000000040000000 Limit 0000000138000000
Node 2 MemBase 0000000138000000 Limit 00000001f8000000
Node 3 MemBase 00000001f8000000 Limit 0000000238000000
Initmem setup node 0 0000000000000000-0000000040000000
NODE_DATA [000000003ffd9000 - 000000003fffffff]
Initmem setup node 1 0000000040000000-0000000138000000
NODE_DATA [0000000137fd9000 - 0000000137ffffff]
Initmem setup node 2 0000000138000000-00000001f8000000
NODE_DATA [00000001f095e000 - 00000001f0984fff]
Initmem setup node 3 00000001f8000000-0000000238000000
Cannot find 159744 bytes in node 3
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: [<ffffffff81d220e6>] __alloc_bootmem_node+0x43/0x96
Pid: 0, comm: swapper Not tainted 3.3.6 #1 AMD Dinar/Dinar
RIP: e030:[<ffffffff81d220e6>] [<ffffffff81d220e6>] __alloc_bootmem_node+0x43/0x96
.. snip..
[<ffffffff81d23024>] sparse_early_usemaps_alloc_node+0x64/0x178
[<ffffffff81d23348>] sparse_init+0xe4/0x25a
[<ffffffff81d16840>] paging_init+0x13/0x22
[<ffffffff81d07fbb>] setup_arch+0x9c6/0xa9b
[<ffffffff81683954>] ? printk+0x3c/0x3e
[<ffffffff81d01a38>] start_kernel+0xe5/0x468
[<ffffffff81d012cf>] x86_64_start_reservations+0xba/0xc1
[<ffffffff81007153>] ? xen_setup_runstate_info+0x2c/0x36
[<ffffffff81d050ee>] xen_start_kernel+0x565/0x56c
"
so we just disable NUMA scanning by setting numa_off=1.
Reported-and-Tested-by: Andre Przywara <andre.przywara@amd.com>
Acked-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit bd49940a35ec7d488ae63bd625639893b3385b97 upstream.
As the initial domain we are able to search/map certain regions
of memory to harvest configuration data. For all low-level we
use ACPI tables - for interrupts we use exclusively ACPI _PRT
(so DSDT) and MADT for INT_SRC_OVR.
The SMP MP table is not used at all. As a matter of fact we do
not even support machines that only have SMP MP but no ACPI tables.
Lets follow how Moorestown does it and just disable searching
for BIOS SMP tables.
This also fixes an issue on HP Proliant BL680c G5 and DL380 G6:
9f->100 for 1:1 PTE
Freeing 9f-100 pfn range: 97 pages freed
1-1 mapping on 9f->100
.. snip..
e820: BIOS-provided physical RAM map:
Xen: [mem 0x0000000000000000-0x000000000009efff] usable
Xen: [mem 0x000000000009f400-0x00000000000fffff] reserved
Xen: [mem 0x0000000000100000-0x00000000cfd1dfff] usable
.. snip..
Scan for SMP in [mem 0x00000000-0x000003ff]
Scan for SMP in [mem 0x0009fc00-0x0009ffff]
Scan for SMP in [mem 0x000f0000-0x000fffff]
found SMP MP-table at [mem 0x000f4fa0-0x000f4faf] mapped at [ffff8800000f4fa0]
(XEN) mm.c:908:d0 Error getting mfn 100 (pfn 5555555555555555) from L1 entry 0000000000100461 for l1e_owner=0, pg_owner=0
(XEN) mm.c:4995:d0 ptwr_emulate: could not get_page_from_l1e()
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: [<ffffffff81ac07e2>] xen_set_pte_init+0x66/0x71
. snip..
Pid: 0, comm: swapper Not tainted 3.6.0-rc6upstream-00188-gb6fb969-dirty #2 HP ProLiant BL680c G5
.. snip..
Call Trace:
[<ffffffff81ad31c6>] __early_ioremap+0x18a/0x248
[<ffffffff81624731>] ? printk+0x48/0x4a
[<ffffffff81ad32ac>] early_ioremap+0x13/0x15
[<ffffffff81acc140>] get_mpc_size+0x2f/0x67
[<ffffffff81acc284>] smp_scan_config+0x10c/0x136
[<ffffffff81acc2e4>] default_find_smp_config+0x36/0x5a
[<ffffffff81ac3085>] setup_arch+0x5b3/0xb5b
[<ffffffff81624731>] ? printk+0x48/0x4a
[<ffffffff81abca7f>] start_kernel+0x90/0x390
[<ffffffff81abc356>] x86_64_start_reservations+0x131/0x136
[<ffffffff81abfa83>] xen_start_kernel+0x65f/0x661
(XEN) Domain 0 crashed: 'noreboot' set - not rebooting.
which is that ioremap would end up mapping 0xff using _PAGE_IOMAP
(which is what early_ioremap sticks as a flag) - which meant
we would get MFN 0xFF (pte ff461, which is OK), and then it would
also map 0x100 (b/c ioremap tries to get page aligned request, and
it was trying to map 0xf4fa0 + PAGE_SIZE - so it mapped the next page)
as _PAGE_IOMAP. Since 0x100 is actually a RAM page, and the _PAGE_IOMAP
bypasses the P2M lookup we would happily set the PTE to 1000461.
Xen would deny the request since we do not have access to the
Machine Frame Number (MFN) of 0x100. The P2M[0x100] is for example
0x80140.
Fixes-Oracle-Bugzilla: https://bugzilla.oracle.com/bugzilla/show_bug.cgi?id=13665
Acked-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit c96aae1f7f393387d160211f60398d58463a7e65 upstream.
When we are finished with return PFNs to the hypervisor, then
populate it back, and also mark the E820 MMIO and E820 gaps
as IDENTITY_FRAMEs, we then call P2M to set areas that can
be used for ballooning. We were off by one, and ended up
over-writting a P2M entry that most likely was an IDENTITY_FRAME.
For example:
1-1 mapping on 40000->40200
1-1 mapping on bc558->bc5ac
1-1 mapping on bc5b4->bc8c5
1-1 mapping on bc8c6->bcb7c
1-1 mapping on bcd00->100000
Released 614 pages of unused memory
Set 277889 page(s) to 1-1 mapping
Populating 40200-40466 pfn range: 614 pages added
=> here we set from 40466 up to bc559 P2M tree to be
INVALID_P2M_ENTRY. We should have done it up to bc558.
The end result is that if anybody is trying to construct
a PTE for PFN bc558 they end up with ~PAGE_PRESENT.
Reported-by-and-Tested-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit b9e0d95c041ca2d7ad297ee37c2e9cfab67a188f upstream.
When the frontend and the backend reside on the same domain, even if we
add pages to the m2p_override, these pages will never be returned by
mfn_to_pfn because the check "get_phys_to_machine(pfn) != mfn" will
always fail, so the pfn of the frontend will be returned instead
(resulting in a deadlock because the frontend pages are already locked).
INFO: task qemu-system-i38:1085 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
qemu-system-i38 D ffff8800cfc137c0 0 1085 1 0x00000000
ffff8800c47ed898 0000000000000282 ffff8800be4596b0 00000000000137c0
ffff8800c47edfd8 ffff8800c47ec010 00000000000137c0 00000000000137c0
ffff8800c47edfd8 00000000000137c0 ffffffff82213020 ffff8800be4596b0
Call Trace:
[<ffffffff81101ee0>] ? __lock_page+0x70/0x70
[<ffffffff81a0fdd9>] schedule+0x29/0x70
[<ffffffff81a0fe80>] io_schedule+0x60/0x80
[<ffffffff81101eee>] sleep_on_page+0xe/0x20
[<ffffffff81a0e1ca>] __wait_on_bit_lock+0x5a/0xc0
[<ffffffff81101ed7>] __lock_page+0x67/0x70
[<ffffffff8106f750>] ? autoremove_wake_function+0x40/0x40
[<ffffffff811867e6>] ? bio_add_page+0x36/0x40
[<ffffffff8110b692>] set_page_dirty_lock+0x52/0x60
[<ffffffff81186021>] bio_set_pages_dirty+0x51/0x70
[<ffffffff8118c6b4>] do_blockdev_direct_IO+0xb24/0xeb0
[<ffffffff811e71a0>] ? ext3_get_blocks_handle+0xe00/0xe00
[<ffffffff8118ca95>] __blockdev_direct_IO+0x55/0x60
[<ffffffff811e71a0>] ? ext3_get_blocks_handle+0xe00/0xe00
[<ffffffff811e91c8>] ext3_direct_IO+0xf8/0x390
[<ffffffff811e71a0>] ? ext3_get_blocks_handle+0xe00/0xe00
[<ffffffff81004b60>] ? xen_mc_flush+0xb0/0x1b0
[<ffffffff81104027>] generic_file_aio_read+0x737/0x780
[<ffffffff813bedeb>] ? gnttab_map_refs+0x15b/0x1e0
[<ffffffff811038f0>] ? find_get_pages+0x150/0x150
[<ffffffff8119736c>] aio_rw_vect_retry+0x7c/0x1d0
[<ffffffff811972f0>] ? lookup_ioctx+0x90/0x90
[<ffffffff81198856>] aio_run_iocb+0x66/0x1a0
[<ffffffff811998b8>] do_io_submit+0x708/0xb90
[<ffffffff81199d50>] sys_io_submit+0x10/0x20
[<ffffffff81a18d69>] system_call_fastpath+0x16/0x1b
The explanation is in the comment within the code:
We need to do this because the pages shared by the frontend
(xen-blkfront) can be already locked (lock_page, called by
do_read_cache_page); when the userspace backend tries to use them
with direct_IO, mfn_to_pfn returns the pfn of the frontend, so
do_blockdev_direct_IO is going to try to lock the same pages
again resulting in a deadlock.
A simplified call graph looks like this:
pygrub QEMU
-----------------------------------------------
do_read_cache_page io_submit
| |
lock_page ext3_direct_IO
|
bio_add_page
|
lock_page
Internally the xen-blkback uses m2p_add_override to swizzle (temporarily)
a 'struct page' to have a different MFN (so that it can point to another
guest). It also can easily find out whether another pfn corresponding
to the mfn exists in the m2p, and can set the FOREIGN bit
in the p2m, making sure that mfn_to_pfn returns the pfn of the backend.
This allows the backend to perform direct_IO on these pages, but as a
side effect prevents the frontend from using get_user_pages_fast on
them while they are being shared with the backend.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 5e626254206a709c6e937f3dda69bf26c7344f6f upstream.
Xen PV kernels allow access to the APERF/MPERF registers to read the
effective frequency. Access to the MSRs is however redirected to the
currently scheduled physical CPU, making consecutive read and
compares unreliable. In addition each rdmsr traps into the hypervisor.
So to avoid bogus readouts and expensive traps, disable the kernel
internal feature flag for APERF/MPERF if running under Xen.
This will
a) remove the aperfmperf flag from /proc/cpuinfo
b) not mislead the power scheduler (arch/x86/kernel/cpu/sched.c) to
use the feature to improve scheduling (by default disabled)
c) not mislead the cpufreq driver to use the MSRs
This does not cover userland programs which access the MSRs via the
device file interface, but this will be addressed separately.
Signed-off-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 1ab46fd319bcf1fcd9fb6311727d532b580e4eba upstream.
Stub out MSR methods that aren't actually needed. This fixes a crash
as Xen Dom0 on AMD Trinity systems. A bigger patch should be added to
remove the paravirt machinery completely for the methods which
apparently have no users!
Reported-by: Andre Przywara <andre.przywara@amd.com>
Link: http://lkml.kernel.org/r/20120530222356.GA28417@andromeda.dapyr.net
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 76a8df7b49168509df02461f83fab117a4a86e08 upstream.
The accessing PCI configuration space with the PCI BIOS32 service does
not work in PV guests.
On systems without MMCONFIG or where the BIOS hasn't marked the
MMCONFIG region as reserved in the e820 map, the BIOS service is
probed (even though direct access is preferred) and this hangs.
Acked-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
[v1: Fixed compile error when CONFIG_PCI is not set]
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit b7e5ffe5d83fa40d702976d77452004abbe35791 upstream.
If I try to do "cat /sys/kernel/debug/kernel_page_tables"
I end up with:
BUG: unable to handle kernel paging request at ffffc7fffffff000
IP: [<ffffffff8106aa51>] ptdump_show+0x221/0x480
PGD 0
Oops: 0000 [#1] SMP
CPU 0
.. snip..
RAX: 0000000000000000 RBX: ffffc00000000fff RCX: 0000000000000000
RDX: 0000800000000000 RSI: 0000000000000000 RDI: ffffc7fffffff000
which is due to the fact we are trying to access a PFN that is not
accessible to us. The reason (at least in this case) was that
PGD[256] is set to __HYPERVISOR_VIRT_START which was setup (by the
hypervisor) to point to a read-only linear map of the MFN->PFN array.
During our parsing we would get the MFN (a valid one), try to look
it up in the MFN->PFN tree and find it invalid and return ~0 as PFN.
Then pte_mfn_to_pfn would happilly feed that in, attach the flags
and return it back to the caller. 'ptdump_show' bitshifts it and
gets and invalid value that it tries to dereference.
Instead of doing all of that, we detect the ~0 case and just
return !_PAGE_PRESENT.
This bug has been in existence .. at least until 2.6.37 (yikes!)
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit cf405ae612b0f7e2358db7ff594c0e94846137aa upstream.
When we boot on a machine that can hotplug CPUs and we
are using 'dom0_max_vcpus=X' on the Xen hypervisor line
to clip the amount of CPUs available to the initial domain,
we get this:
(XEN) Command line: com1=115200,8n1 dom0_mem=8G noreboot dom0_max_vcpus=8 sync_console mce_verbosity=verbose console=com1,vga loglvl=all guest_loglvl=all
.. snip..
DMI: Intel Corporation S2600CP/S2600CP, BIOS SE5C600.86B.99.99.x032.072520111118 07/25/2011
.. snip.
SMP: Allowing 64 CPUs, 32 hotplug CPUs
installing Xen timer for CPU 7
cpu 7 spinlock event irq 361
NMI watchdog: disabled (cpu7): hardware events not enabled
Brought up 8 CPUs
.. snip..
[acpi processor finds the CPUs are not initialized and starts calling
arch_register_cpu, which creates /sys/devices/system/cpu/cpu8/online]
CPU 8 got hotplugged
CPU 9 got hotplugged
CPU 10 got hotplugged
.. snip..
initcall 1_acpi_battery_init_async+0x0/0x1b returned 0 after 406 usecs
calling erst_init+0x0/0x2bb @ 1
[and the scheduler sticks newly started tasks on the new CPUs, but
said CPUs cannot be initialized b/c the hypervisor has limited the
amount of vCPUS to 8 - as per the dom0_max_vcpus=8 flag.
The spinlock tries to kick the other CPU, but the structure for that
is not initialized and we crash.]
BUG: unable to handle kernel paging request at fffffffffffffed8
IP: [<ffffffff81035289>] xen_spin_lock+0x29/0x60
PGD 180d067 PUD 180e067 PMD 0
Oops: 0002 [#1] SMP
CPU 7
Modules linked in:
Pid: 1, comm: swapper/0 Not tainted 3.4.0-rc2upstream-00001-gf5154e8 #1 Intel Corporation S2600CP/S2600CP
RIP: e030:[<ffffffff81035289>] [<ffffffff81035289>] xen_spin_lock+0x29/0x60
RSP: e02b:ffff8801fb9b3a70 EFLAGS: 00010282
With this patch, we cap the amount of vCPUS that the initial domain
can run, to exactly what dom0_max_vcpus=X has specified.
In the future, if there is a hypercall that will allow a running
domain to expand past its initial set of vCPUS, this patch should
be re-evaluated.
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 7eb7ce4d2e8991aff4ecb71a81949a907ca755ac upstream.
In xen_restore_fl_direct(), xen_force_evtchn_callback() was being
called even if no events were pending. This resulted in (depending on
workload) about a 100 times as many xen_version hypercalls as
necessary.
Fix this by correcting the sense of the conditional jump.
This seems to give a significant performance benefit for some
workloads.
There is some subtle tricksy "..since the check here is trying to
check both pending and masked in a single cmpw, but I think this is
correct. It will call check_events now only when the combined
mask+pending word is 0x0001 (aka unmasked, pending)." (Ian)
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 7a7546b377bdaa25ac77f33d9433c59f259b9688 upstream.
If NR_CPUS < 256 then arch_spinlock_t is only 16 bits wide but struct
xen_spinlock is 32 bits. When a spin lock is contended and
xl->spinners is modified the two bytes immediately after the spin lock
would be corrupted.
This is a regression caused by 84eb950db13ca40a0572ce9957e14723500943d6
(x86, ticketlock: Clean up types and accessors) which reduced the size
of arch_spinlock_t.
Fix this by making xl->spinners a u8 if NR_CPUS < 256. A
BUILD_BUG_ON() is also added to check the sizes of the two structures
are compatible.
In many cases this was not noticable as there would often be padding
bytes after the lock (e.g., if any of CONFIG_GENERIC_LOCKBREAK,
CONFIG_DEBUG_SPINLOCK, or CONFIG_DEBUG_LOCK_ALLOC were enabled).
The bnx2 driver is affected. In struct bnx2, phy_lock and
indirect_lock may have no padding after them. Contention on phy_lock
would corrupt indirect_lock making it appear locked and the driver
would deadlock.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy@goop.org>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen
* 'stable/for-linus-fixes-3.2' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
xen/swiotlb: Use page alignment for early buffer allocation.
xen: only limit memory map to maximum reservation for domain 0.
|
|
d312ae878b6a "xen: use maximum reservation to limit amount of usable RAM"
clamped the total amount of RAM to the current maximum reservation. This is
correct for dom0 but is not correct for guest domains. In order to boot a guest
"pre-ballooned" (e.g. with memory=1G but maxmem=2G) in order to allow for
future memory expansion the guest must derive max_pfn from the e820 provided by
the toolstack and not the current maximum reservation (which can reflect only
the current maximum, not the guest lifetime max). The existing algorithm
already behaves this correctly if we do not artificially limit the maximum
number of pages for the guest case.
For a guest booted with maxmem=512, memory=128 this results in:
[ 0.000000] BIOS-provided physical RAM map:
[ 0.000000] Xen: 0000000000000000 - 00000000000a0000 (usable)
[ 0.000000] Xen: 00000000000a0000 - 0000000000100000 (reserved)
-[ 0.000000] Xen: 0000000000100000 - 0000000008100000 (usable)
-[ 0.000000] Xen: 0000000008100000 - 0000000020800000 (unusable)
+[ 0.000000] Xen: 0000000000100000 - 0000000020800000 (usable)
...
[ 0.000000] NX (Execute Disable) protection: active
[ 0.000000] DMI not present or invalid.
[ 0.000000] e820 update range: 0000000000000000 - 0000000000010000 (usable) ==> (reserved)
[ 0.000000] e820 remove range: 00000000000a0000 - 0000000000100000 (usable)
-[ 0.000000] last_pfn = 0x8100 max_arch_pfn = 0x1000000
+[ 0.000000] last_pfn = 0x20800 max_arch_pfn = 0x1000000
[ 0.000000] initial memory mapped : 0 - 027ff000
[ 0.000000] Base memory trampoline at [c009f000] 9f000 size 4096
-[ 0.000000] init_memory_mapping: 0000000000000000-0000000008100000
-[ 0.000000] 0000000000 - 0008100000 page 4k
-[ 0.000000] kernel direct mapping tables up to 8100000 @ 27bb000-27ff000
+[ 0.000000] init_memory_mapping: 0000000000000000-0000000020800000
+[ 0.000000] 0000000000 - 0020800000 page 4k
+[ 0.000000] kernel direct mapping tables up to 20800000 @ 26f8000-27ff000
[ 0.000000] xen: setting RW the range 27e8000 - 27ff000
[ 0.000000] 0MB HIGHMEM available.
-[ 0.000000] 129MB LOWMEM available.
-[ 0.000000] mapped low ram: 0 - 08100000
-[ 0.000000] low ram: 0 - 08100000
+[ 0.000000] 520MB LOWMEM available.
+[ 0.000000] mapped low ram: 0 - 20800000
+[ 0.000000] low ram: 0 - 20800000
With this change "xl mem-set <domain> 512M" will successfully increase the
guest RAM (by reducing the balloon).
There is no change for dom0.
Reported-and-Tested-by: George Shuklin <george.shuklin@gmail.com>
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Cc: stable@kernel.org
Reviewed-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
The idea behind commit d91ee5863b71 ("cpuidle: replace xen access to x86
pm_idle and default_idle") was to have one call - disable_cpuidle()
which would make pm_idle not be molested by other code. It disallows
cpuidle_idle_call to be set to pm_idle (which is excellent).
But in the select_idle_routine() and idle_setup(), the pm_idle can still
be set to either: amd_e400_idle, mwait_idle or default_idle. This
depends on some CPU flags (MWAIT) and in AMD case on the type of CPU.
In case of mwait_idle we can hit some instances where the hypervisor
(Amazon EC2 specifically) sets the MWAIT and we get:
Brought up 2 CPUs
invalid opcode: 0000 [#1] SMP
Pid: 0, comm: swapper Not tainted 3.1.0-0.rc6.git0.3.fc16.x86_64 #1
RIP: e030:[<ffffffff81015d1d>] [<ffffffff81015d1d>] mwait_idle+0x6f/0xb4
...
Call Trace:
[<ffffffff8100e2ed>] cpu_idle+0xae/0xe8
[<ffffffff8149ee78>] cpu_bringup_and_idle+0xe/0x10
RIP [<ffffffff81015d1d>] mwait_idle+0x6f/0xb4
RSP <ffff8801d28ddf10>
In the case of amd_e400_idle we don't get so spectacular crashes, but we
do end up making an MSR which is trapped in the hypervisor, and then
follow it up with a yield hypercall. Meaning we end up going to
hypervisor twice instead of just once.
The previous behavior before v3.0 was that pm_idle was set to
default_idle regardless of select_idle_routine/idle_setup.
We want to do that, but only for one specific case: Xen. This patch
does that.
Fixes RH BZ #739499 and Ubuntu #881076
Reported-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
PVHVM running with more than 32 vcpus and pv_irq/pv_time enabled
need VCPU placement to work, or else it will softlockup.
CC: stable@kernel.org
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Zhenzhong Duan <zhenzhong.duan@oracle.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
When mapping a foreign page with xenbus_map_ring_valloc() with the
GNTTABOP_map_grant_ref hypercall, set the GNTMAP_contains_pte flag and
pass a pointer to the PTE (in init_mm).
After the page is mapped, the usual fault mechanism can be used to
update additional MMs. This allows the vmalloc_sync_all() to be
removed from alloc_vm_area().
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Acked-by: Andrew Morton <akpm@linux-foundation.org>
[v1: Squashed fix by Michal for no-mmu case]
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Michal Simek <monstr@monstr.eu>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen
* 'upstream/xen-settime' of git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen:
xen/dom0: set wallclock time in Xen
xen: add dom0_op hypercall
xen/acpi: Domain0 acpi parser related platform hypercall
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen
* 'stable/vmalloc-3.2' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
net: xen-netback: use API provided by xenbus module to map rings
block: xen-blkback: use API provided by xenbus module to map rings
xen: use generic functions instead of xen_{alloc, free}_vm_area()
|
|
'stable/pci.fixes-3.2' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen
* 'stable/drivers-3.2' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
xenbus: don't rely on xen_initial_domain to detect local xenstore
xenbus: Fix loopback event channel assuming domain 0
xen/pv-on-hvm:kexec: Fix implicit declaration of function 'xen_hvm_domain'
xen/pv-on-hvm kexec: add xs_reset_watches to shutdown watches from old kernel
xen/pv-on-hvm kexec: update xs_wire.h:xsd_sockmsg_type from xen-unstable
xen/pv-on-hvm kexec+kdump: reset PV devices in kexec or crash kernel
xen/pv-on-hvm kexec: rebind virqs to existing eventchannel ports
xen/pv-on-hvm kexec: prevent crash in xenwatch_thread() when stale watch events arrive
* 'stable/drivers.bugfixes-3.2' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
xen/pciback: Check if the device is found instead of blindly assuming so.
xen/pciback: Do not dereference psdev during printk when it is NULL.
xen: remove XEN_PLATFORM_PCI config option
xen: XEN_PVHVM depends on PCI
xen/pciback: double lock typo
xen/pciback: use mutex rather than spinlock in vpci backend
xen/pciback: Use mutexes when working with Xenbus state transitions.
xen/pciback: miscellaneous adjustments
xen/pciback: use mutex rather than spinlock in passthrough backend
xen/pciback: use resource_size()
* 'stable/pci.fixes-3.2' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
xen/pci: support multi-segment systems
xen-swiotlb: When doing coherent alloc/dealloc check before swizzling the MFNs.
xen/pci: make bus notifier handler return sane values
xen-swiotlb: fix printk and panic args
xen-swiotlb: Fix wrong panic.
xen-swiotlb: Retry up three times to allocate Xen-SWIOTLB
xen-pcifront: Update warning comment to use 'e820_host' option.
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen
* 'stable/bug.fixes-3.2' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
xen/p2m/debugfs: Make type_name more obvious.
xen/p2m/debugfs: Fix potential pointer exception.
xen/enlighten: Fix compile warnings and set cx to known value.
xen/xenbus: Remove the unnecessary check.
xen/irq: If we fail during msi_capability_init return proper error code.
xen/events: Don't check the info for NULL as it is already done.
xen/events: BUG() when we can't allocate our event->irq array.
* 'stable/mmu.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
xen: Fix selfballooning and ensure it doesn't go too far
xen/gntdev: Fix sleep-inside-spinlock
xen: modify kernel mappings corresponding to granted pages
xen: add an "highmem" parameter to alloc_xenballooned_pages
xen/p2m: Use SetPagePrivate and its friends for M2P overrides.
xen/p2m: Make debug/xen/mmu/p2m visible again.
Revert "xen/debug: WARN_ON when identity PFN has no _PAGE_IOMAP flag set."
|
|
Per Ian Campbell suggestion to defend against future breakage
in case we expand the P2M values, incorporate the defines
in the string array.
Suggested-by: Ian Campbell <Ian.Campbell@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
We could be referencing the last + 1 element of level_name[]
array which would cause a pointer exception, because of the
initial setup of lvl=4.
[v1: No need to do this for type_name, pointed out by Ian Campbell]
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
We get:
linux/arch/x86/xen/enlighten.c: In function ‘xen_start_kernel’:
linux/arch/x86/xen/enlighten.c:226: warning: ‘cx’ may be used uninitialized in this function
linux/arch/x86/xen/enlighten.c:240: note: ‘cx’ was declared here
and the cx is really not set but passed in the xen_cpuid instruction
which masks the value with returned masked_ecx from cpuid. This
can potentially lead to invalid data being stored in cx.
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
Replace calls to the Xen-specific xen_alloc_vm_area() and
xen_free_vm_area() functions with the generic equivalent
(alloc_vm_area() and free_vm_area()).
On x86, these were identical already.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
In xen_memory_setup() all reserved regions and gaps are set to an
identity (1-1) p2m mapping. If an available page has a PFN within one
of these 1-1 mappings it will become inaccessible (as it MFN is lost)
so release them before setting up the mapping.
This can make an additional 256 MiB or more of RAM available
(depending on the size of the reserved regions in the memory map) if
the initial pages overlap with reserved regions.
The 1:1 p2m mappings are also extended to cover partial pages. This
fixes an issue with (for example) systems with a BIOS that puts the
DMI tables in a reserved region that begins on a non-page boundary.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
Allow the extra memory (used by the balloon driver) to be in multiple
regions (typically two regions, one for low memory and one for high
memory). This allows the balloon driver to increase the number of
available low pages (if the initial number if pages is small).
As a side effect, the algorithm for building the e820 memory map is
simpler and more obviously correct as the map supplied by the
hypervisor is (almost) used as is (in particular, all reserved regions
and gaps are preserved). Only RAM regions are altered and RAM regions
above max_pfn + extra_pages are marked as unused (the region is split
in two if necessary).
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
Allow the xen balloon driver to populate its list of extra pages from
more than one region of memory. This will allow platforms to provide
(for example) a region of low memory and a region of high memory.
The maximum possible number of extra regions is 128 (== E820MAX) which
is quite large so xen_extra_mem is placed in __initdata. This is safe
as both xen_memory_setup() and balloon_init() are in __init.
The balloon regions themselves are not altered (i.e., there is still
only the one region).
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
In xen_memory_setup() pages that occur in gaps in the memory map are
released back to Xen. This reduces the domain's current page count in
the hypervisor. The Xen balloon driver does not correctly decrease
its initial current_pages count to reflect this. If 'delta' pages are
released and the target is adjusted the resulting reservation is
always 'delta' less than the requested target.
This affects dom0 if the initial allocation of pages overlaps the PCI
memory region but won't affect most domU guests that have been setup
with pseudo-physical memory maps that don't have gaps.
Fix this by accouting for the released pages when starting the balloon
driver.
If the domain's targets are managed by xapi, the domain may eventually
run out of memory and die because xapi currently gets its target
calculations wrong and whenever it is restarted it always reduces the
target by 'delta'.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
Xen PV on HVM guests require PCI support because they need the
xen-platform-pci driver in order to initialize xenbus.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
If we want to use granted pages for AIO, changing the mappings of a user
vma and the corresponding p2m is not enough, we also need to update the
kernel mappings accordingly.
Currently this is only needed for pages that are created for user usages
through /dev/xen/gntdev. As in, pages that have been in use by the
kernel and use the P2M will not need this special mapping.
However there are no guarantees that in the future the kernel won't
start accessing pages through the 1:1 even for internal usage.
In order to avoid the complexity of dealing with highmem, we allocated
the pages lowmem.
We issue a HYPERVISOR_grant_table_op right away in
m2p_add_override and we remove the mappings using another
HYPERVISOR_grant_table_op in m2p_remove_override.
Considering that m2p_add_override and m2p_remove_override are called
once per page we use multicalls and hypercall batching.
Use the kmap_op pointer directly as argument to do the mapping as it is
guaranteed to be present up until the unmapping is done.
Before issuing any unmapping multicalls, we need to make sure that the
mapping has already being done, because we need the kmap->handle to be
set correctly.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
[v1: Removed GRANT_FRAME_BIT usage]
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
|
|
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
|
