openwrt/staging/blogic.git
8 years agox86/irq: Plug vector cleanup race
Thomas Gleixner [Thu, 31 Dec 2015 16:30:54 +0000 (16:30 +0000)]
x86/irq: Plug vector cleanup race

We still can end up with a stale vector due to the following:

CPU0                          CPU1                      CPU2
lock_vector()
data->move_in_progress=0
sendIPI()
unlock_vector()
                              set_affinity()
                              assign_irq_vector()
                              lock_vector()             handle_IPI
                              move_in_progress = 1      lock_vector()
                              unlock_vector()
                                                        move_in_progress == 1

So we need to serialize the vector assignment against a pending cleanup. The
solution is rather simple now. We not only check for the move_in_progress flag
in assign_irq_vector(), we also check whether there is still a cleanup pending
in the old_domain cpumask. If so, we return -EBUSY to the caller and let him
deal with it. Though we have to be careful in the cpu unplug case. If the
cleanout has not yet completed then the following setaffinity() call would
return -EBUSY. Add code which prevents this.

Full context is here: http://lkml.kernel.org/r/5653B688.4050809@stratus.com

Reported-and-tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160107.207265407@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Call irq_force_move_complete with irq descriptor
Thomas Gleixner [Thu, 31 Dec 2015 16:30:53 +0000 (16:30 +0000)]
x86/irq: Call irq_force_move_complete with irq descriptor

First of all there is no point in looking up the irq descriptor again, but we
also need the descriptor for the final cleanup race fix in the next
patch. Make that change seperate. No functional difference.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160107.125211743@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Remove outgoing CPU from vector cleanup mask
Thomas Gleixner [Thu, 31 Dec 2015 16:30:52 +0000 (16:30 +0000)]
x86/irq: Remove outgoing CPU from vector cleanup mask

We want to synchronize new vector assignments with a pending cleanup. Remove a
dying cpu from a pending cleanup mask.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160107.045961667@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Remove the cpumask allocation from send_cleanup_vector()
Thomas Gleixner [Thu, 31 Dec 2015 16:30:52 +0000 (16:30 +0000)]
x86/irq: Remove the cpumask allocation from send_cleanup_vector()

There is no need to allocate a new cpumask for sending the cleanup vector. The
old_domain mask is now protected by the vector_lock, so we can safely remove
the offline cpus from it and send the IPI with the resulting mask.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160106.967993932@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Clear move_in_progress before sending cleanup IPI
Thomas Gleixner [Thu, 31 Dec 2015 16:30:51 +0000 (16:30 +0000)]
x86/irq: Clear move_in_progress before sending cleanup IPI

send_cleanup_vector() fiddles with the old_domain mask unprotected because it
relies on the protection by the move_in_progress flag. But this is fatal, as
the flag is reset after the IPI has been sent. So a cpu which receives the IPI
can still see the flag set and therefor ignores the cleanup request. If no
other cleanup request happens then the vector stays stale on that cpu and in
case of an irq removal the vector still persists. That can lead to use after
free when the next cleanup IPI happens.

Protect the code with vector_lock and clear move_in_progress before sending
the IPI.

This does not plug the race which Joe reported because:

CPU0                          CPU1                      CPU2
lock_vector()
data->move_in_progress=0
sendIPI()
unlock_vector()
                              set_affinity()
                              assign_irq_vector()
                              lock_vector()             handle_IPI
                              move_in_progress = 1      lock_vector()
                              unlock_vector()
                                                        move_in_progress == 1

The full fix comes with a later patch.

Reported-and-tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160106.892412198@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Remove offline cpus from vector cleanup
Thomas Gleixner [Thu, 31 Dec 2015 16:30:50 +0000 (16:30 +0000)]
x86/irq: Remove offline cpus from vector cleanup

No point of keeping offline cpus in the cleanup mask.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160106.808642683@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Get rid of code duplication
Thomas Gleixner [Thu, 31 Dec 2015 16:30:49 +0000 (16:30 +0000)]
x86/irq: Get rid of code duplication

Reusing an existing vector and assigning a new vector has duplicated
code. Consolidate it.

This is also a preparatory patch for finally plugging the cleanup race.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160106.721599216@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Copy vectormask instead of an AND operation
Thomas Gleixner [Thu, 31 Dec 2015 16:30:49 +0000 (16:30 +0000)]
x86/irq: Copy vectormask instead of an AND operation

In the case that the new vector mask is a subset of the existing mask there is
no point to do a AND operation of currentmask & newmask. The result is
newmask. So we can simply copy the new mask to the current mask and be done
with it. Preparatory patch for further consolidation.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160106.640253454@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Check vector allocation early
Thomas Gleixner [Thu, 31 Dec 2015 16:30:48 +0000 (16:30 +0000)]
x86/irq: Check vector allocation early

__assign_irq_vector() uses the vector_cpumask which is assigned by
apic->vector_allocation_domain() without doing basic sanity checks. That can
result in a situation where the final assignement of a newly found vector
fails in apic->cpu_mask_to_apicid_and(). So we have to do rollbacks for no
reason.

apic->cpu_mask_to_apicid_and() only fails if

  vector_cpumask & requested_cpumask & cpu_online_mask

is empty.

Check for this condition right away and if the result is empty try immediately
the next possible cpu in the requested mask. So in case of a failure the old
setting is unchanged and we can remove the rollback code.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160106.561877324@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Reorganize the search in assign_irq_vector
Thomas Gleixner [Thu, 31 Dec 2015 16:30:47 +0000 (16:30 +0000)]
x86/irq: Reorganize the search in assign_irq_vector

Split out the code which advances the target cpu for the search so we can
reuse it for the next patch which adds an early validation check for the
vectormask which we get from the apic.

Add comments while at it.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160106.484562040@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Reorganize the return path in assign_irq_vector
Thomas Gleixner [Thu, 31 Dec 2015 16:30:46 +0000 (16:30 +0000)]
x86/irq: Reorganize the return path in assign_irq_vector

Use an explicit goto for the cases where we have success in the search/update
and return -ENOSPC if the search loop ends due to no space.

Preparatory patch for fixes. No functional change.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160106.403491024@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Do not use apic_chip_data.old_domain as temporary buffer
Jiang Liu [Thu, 31 Dec 2015 16:30:46 +0000 (16:30 +0000)]
x86/irq: Do not use apic_chip_data.old_domain as temporary buffer

Function __assign_irq_vector() makes use of apic_chip_data.old_domain as a
temporary buffer, which is in the way of using apic_chip_data.old_domain for
synchronizing the vector cleanup with the vector assignement code.

Use a proper temporary cpumask for this.

[ tglx: Renamed the mask to searched_cpumask for clarity ]

Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/1450880014-11741-1-git-send-email-jiang.liu@linux.intel.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Validate that irq descriptor is still active
Thomas Gleixner [Thu, 31 Dec 2015 16:30:45 +0000 (16:30 +0000)]
x86/irq: Validate that irq descriptor is still active

In fixup_irqs() we unconditionally dereference the irq chip of an irq
descriptor. The descriptor might still be valid, but already cleaned up,
i.e. the chip removed. Add a check for this condition.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/20151231160106.236423282@linutronix.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Fix a race in x86_vector_free_irqs()
Jiang Liu [Thu, 31 Dec 2015 16:30:44 +0000 (16:30 +0000)]
x86/irq: Fix a race in x86_vector_free_irqs()

There's a race condition between

x86_vector_free_irqs()
{
free_apic_chip_data(irq_data->chip_data);
xxxxx //irq_data->chip_data has been freed, but the pointer
//hasn't been reset yet
irq_domain_reset_irq_data(irq_data);
}

and

smp_irq_move_cleanup_interrupt()
{
raw_spin_lock(&vector_lock);
data = apic_chip_data(irq_desc_get_irq_data(desc));
access data->xxxx // may access freed memory
raw_spin_unlock(&desc->lock);
}

which may cause smp_irq_move_cleanup_interrupt() to access freed memory.

Call irq_domain_reset_irq_data(), which clears the pointer with vector lock
held.

[ tglx: Free memory outside of lock held region. ]

Signed-off-by: Jiang Liu <jiang.liu@linux.intel.com>
Tested-by: Borislav Petkov <bp@alien8.de>
Tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org #4.3+
Link: http://lkml.kernel.org/r/1450880014-11741-3-git-send-email-jiang.liu@linux.intel.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/irq: Call chip->irq_set_affinity in proper context
Thomas Gleixner [Thu, 14 Jan 2016 07:43:38 +0000 (08:43 +0100)]
x86/irq: Call chip->irq_set_affinity in proper context

setup_ioapic_dest() calls irqchip->irq_set_affinity() completely
unprotected. That's wrong in several aspects:

 - it opens a race window where irq_set_affinity() can be interrupted and the
   irq chip left in unconsistent state.

 - it triggers a lockdep splat when we fix the vector race for 4.3+ because
   vector lock is taken with interrupts enabled.

The proper calling convention is irq descriptor lock held and interrupts
disabled.

Reported-and-tested-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
Cc: Jeremiah Mahler <jmmahler@gmail.com>
Cc: andy.shevchenko@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Joe Lawrence <joe.lawrence@stratus.com>
Cc: stable@vger.kernel.org
Link: http://lkml.kernel.org/r/alpine.DEB.2.11.1601140919420.3575@nanos
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agox86/cpu/amd: Remove an unneeded condition in srat_detect_node()
Dan Carpenter [Wed, 13 Jan 2016 12:39:40 +0000 (15:39 +0300)]
x86/cpu/amd: Remove an unneeded condition in srat_detect_node()

Originally we calculated ht_nodeid as "ht_nodeid = apicid -
boot_cpu_id;" so presumably it could be negative.

But after commit:

  01aaea1afbcd ('x86: introduce initial apicid')

we use c->initial_apicid which is an unsigned short and thus always >= 0.

It causes a static checker warning to test for impossible
conditions so let's remove it.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Hector Marco-Gisbert <hecmargi@upv.es>
Cc: Huang Rui <ray.huang@amd.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Yinghai Lu <yhlu.kernel@gmail.com>
Link: http://lkml.kernel.org/r/20160113123940.GE19993@mwanda
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/vdso/pvclock: Protect STABLE check with the seqcount
Andy Lutomirski [Mon, 4 Jan 2016 23:14:28 +0000 (15:14 -0800)]
x86/vdso/pvclock: Protect STABLE check with the seqcount

If the clock becomes unstable while we're reading it, we need to
bail.  We can do this by simply moving the check into the
seqcount loop.

Reported-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Alexander Graf <agraf@suse.de>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Radim Krcmar <rkrcmar@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/755dcedb17269e1d7ce12a9a713dea303835137e.1451949191.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/mm: Improve switch_mm() barrier comments
Andy Lutomirski [Tue, 12 Jan 2016 20:47:40 +0000 (12:47 -0800)]
x86/mm: Improve switch_mm() barrier comments

My previous comments were still a bit confusing and there was a
typo. Fix it up.

Reported-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Fixes: 71b3c126e611 ("x86/mm: Add barriers and document switch_mm()-vs-flush synchronization")
Link: http://lkml.kernel.org/r/0a0b43cdcdd241c5faaaecfbcc91a155ddedc9a1.1452631609.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agoselftests/x86: Test __kernel_sigreturn and __kernel_rt_sigreturn
Andy Lutomirski [Tue, 12 Jan 2016 19:56:15 +0000 (11:56 -0800)]
selftests/x86: Test __kernel_sigreturn and __kernel_rt_sigreturn

The vdso-based sigreturn mechanism is fragile and isn't used by
modern glibc so, if we break it, we'll only notice when someone
tests an unusual libc.

Add an explicit selftest.

[ I wrote this while debugging a Bionic breakage -- my first guess
  was that I had somehow messed up sigreturn.  I've caused problems in
  that code before, and it's really easy to fail to notice it because
  there's nothing on a modern distro that needs vdso-based sigreturn. ]

Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Shuah Khan <shuahkh@osg.samsung.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/32946d714156879cd8e5d8eab044cd07557ed558.1452628504.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/reboot/quirks: Add iMac10,1 to pci_reboot_dmi_table[]
Mario Kleiner [Fri, 18 Dec 2015 19:24:06 +0000 (20:24 +0100)]
x86/reboot/quirks: Add iMac10,1 to pci_reboot_dmi_table[]

Without the reboot=pci method, the iMac 10,1 simply
hangs after printing "Restarting system" at the point
when it should reboot. This fixes it.

Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Cc: <stable@vger.kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dave Jones <davej@codemonkey.org.uk>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.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/1450466646-26663-1-git-send-email-mario.kleiner.de@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agolguest: Map switcher text R/O
Rusty Russell [Tue, 22 Dec 2015 00:09:54 +0000 (10:39 +1030)]
lguest: Map switcher text R/O

Pavel noted that lguest maps the switcher code executable and
read-write.  This is a bad idea for any kernel text, but
particularly for text mapped at a fixed address.

Create two vmas, one for the text (PAGE_KERNEL_RX) and another
for the stacks (PAGE_KERNEL).  Use VM_NO_GUARD to map them
adjacent (as expected by the rest of the code).

Reported-by: Pavel Machek <pavel@ucw.cz>
Tested-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/boot: Hide local labels in verify_cpu()
Borislav Petkov [Sun, 3 Jan 2016 22:38:53 +0000 (23:38 +0100)]
x86/boot: Hide local labels in verify_cpu()

... from the final ELF image's symbol table as they're not
really needed there.

Before:

$ readelf -a vmlinux | grep verify_cpu
    43: ffffffff810001a9     0 NOTYPE  LOCAL  DEFAULT    1 verify_cpu
    45: ffffffff8100028f     0 NOTYPE  LOCAL  DEFAULT    1 verify_cpu_no_longmode
    46: ffffffff810001de     0 NOTYPE  LOCAL  DEFAULT    1 verify_cpu_noamd
    47: ffffffff8100022b     0 NOTYPE  LOCAL  DEFAULT    1 verify_cpu_check
    48: ffffffff8100021c     0 NOTYPE  LOCAL  DEFAULT    1 verify_cpu_clear_xd
    49: ffffffff81000263     0 NOTYPE  LOCAL  DEFAULT    1 verify_cpu_sse_test
    50: ffffffff81000296     0 NOTYPE  LOCAL  DEFAULT    1 verify_cpu_sse_ok

After:

$ readelf -a vmlinux | grep verify_cpu
    43: ffffffff810001a9     0 NOTYPE  LOCAL  DEFAULT    1 verify_cpu

No functionality change.

Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Quentin Casasnovas <quentin.casasnovas@oracle.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1451860733-21163-1-git-send-email-bp@alien8.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/fpu: Disable AVX when eagerfpu is off
yu-cheng yu [Wed, 6 Jan 2016 22:24:54 +0000 (14:24 -0800)]
x86/fpu: Disable AVX when eagerfpu is off

When "eagerfpu=off" is given as a command-line input, the kernel
should disable AVX support.

The Task Switched bit used for lazy context switching does not
support AVX. If AVX is enabled without eagerfpu context
switching, one task's AVX state could become corrupted or leak
to other tasks. This is a bug and has bad security implications.

This only affects systems that have AVX/AVX2/AVX512 and this
issue will be found only when one actually uses AVX/AVX2/AVX512
_AND_ does eagerfpu=off.

Reference: Intel Software Developer's Manual Vol. 3A

Sec. 2.5 Control Registers:
TS Task Switched bit (bit 3 of CR0) -- Allows the saving of the
x87 FPU/ MMX/SSE/SSE2/SSE3/SSSE3/SSE4 context on a task switch
to be delayed until an x87 FPU/MMX/SSE/SSE2/SSE3/SSSE3/SSE4
instruction is actually executed by the new task.

Sec. 13.4.1 Using the TS Flag to Control the Saving of the X87
FPU and SSE State
When the TS flag is set, the processor monitors the instruction
stream for x87 FPU, MMX, SSE instructions. When the processor
detects one of these instructions, it raises a
device-not-available exeception (#NM) prior to executing the
instruction.

Signed-off-by: Yu-cheng Yu <yu-cheng.yu@intel.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Borislav Petkov <bp@suse.de>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Quentin Casasnovas <quentin.casasnovas@oracle.com>
Cc: Ravi V. Shankar <ravi.v.shankar@intel.com>
Cc: Sai Praneeth Prakhya <sai.praneeth.prakhya@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: yu-cheng yu <yu-cheng.yu@intel.com>
Link: http://lkml.kernel.org/r/1452119094-7252-5-git-send-email-yu-cheng.yu@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/fpu: Disable MPX when eagerfpu is off
yu-cheng yu [Wed, 6 Jan 2016 22:24:53 +0000 (14:24 -0800)]
x86/fpu: Disable MPX when eagerfpu is off

This issue is a fallout from the command-line parsing move.

When "eagerfpu=off" is given as a command-line input, the kernel
should disable MPX support. The decision for turning off MPX was
made in fpu__init_system_ctx_switch(), which is after the
selection of the XSAVE format. This patch fixes it by getting
that decision done earlier in fpu__init_system_xstate().

Signed-off-by: Yu-cheng Yu <yu-cheng.yu@intel.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Borislav Petkov <bp@suse.de>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Quentin Casasnovas <quentin.casasnovas@oracle.com>
Cc: Ravi V. Shankar <ravi.v.shankar@intel.com>
Cc: Sai Praneeth Prakhya <sai.praneeth.prakhya@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: yu-cheng yu <yu-cheng.yu@intel.com>
Link: http://lkml.kernel.org/r/1452119094-7252-4-git-send-email-yu-cheng.yu@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/fpu: Disable XGETBV1 when no XSAVE
yu-cheng yu [Wed, 6 Jan 2016 22:24:52 +0000 (14:24 -0800)]
x86/fpu: Disable XGETBV1 when no XSAVE

When "noxsave" is given as a command-line input, the kernel
should disable XGETBV1. This issue currently does not cause any
actual problems. XGETBV1 is only useful if we have something
using the 'init optimization' (i.e. xsaveopt, xsaves). We
already clear both of those in fpu__xstate_clear_all_cpu_caps().
But this is good for completeness.

Signed-off-by: Yu-cheng Yu <yu-cheng.yu@intel.com>
Reviewed-by: Dave Hansen <dave.hansen@intel.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Borislav Petkov <bp@suse.de>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Quentin Casasnovas <quentin.casasnovas@oracle.com>
Cc: Ravi V. Shankar <ravi.v.shankar@intel.com>
Cc: Sai Praneeth Prakhya <sai.praneeth.prakhya@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: yu-cheng yu <yu-cheng.yu@intel.com>
Link: http://lkml.kernel.org/r/1452119094-7252-3-git-send-email-yu-cheng.yu@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/fpu: Fix early FPU command-line parsing
yu-cheng yu [Wed, 6 Jan 2016 22:24:51 +0000 (14:24 -0800)]
x86/fpu: Fix early FPU command-line parsing

The function fpu__init_system() is executed before
parse_early_param(). This causes wrong FPU configuration. This
patch fixes this issue by parsing boot_command_line in the
beginning of fpu__init_system().

With all four patches in this series, each parameter disables
features as the following:

eagerfpu=off: eagerfpu, avx, avx2, avx512, mpx
no387: fpu
nofxsr: fxsr, fxsropt, xmm
noxsave: xsave, xsaveopt, xsaves, xsavec, avx, avx2, avx512,
mpx, xgetbv1 noxsaveopt: xsaveopt
noxsaves: xsaves

Signed-off-by: Yu-cheng Yu <yu-cheng.yu@intel.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Borislav Petkov <bp@suse.de>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Quentin Casasnovas <quentin.casasnovas@oracle.com>
Cc: Ravi V. Shankar <ravi.v.shankar@intel.com>
Cc: Sai Praneeth Prakhya <sai.praneeth.prakhya@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: yu-cheng yu <yu-cheng.yu@intel.com>
Link: http://lkml.kernel.org/r/1452119094-7252-2-git-send-email-yu-cheng.yu@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/mm: Use PAGE_ALIGNED instead of IS_ALIGNED
Kefeng Wang [Tue, 12 Jan 2016 02:19:30 +0000 (10:19 +0800)]
x86/mm: Use PAGE_ALIGNED instead of IS_ALIGNED

Use PAGE_ALIGEND macro in <linux/mm.h> to simplify code.

Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: <guohanjun@huawei.com>
Cc: Alexander Kuleshov <kuleshovmail@gmail.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/1452565170-11083-1-git-send-email-wangkefeng.wang@huawei.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agoselftests/x86: Disable the ldt_gdt_64 test for now
Andy Lutomirski [Tue, 12 Jan 2016 01:23:32 +0000 (17:23 -0800)]
selftests/x86: Disable the ldt_gdt_64 test for now

ldt_gdt.c relies on cross-cpu invalidation of SS to do one of
its tests.  On 32-bit builds, this works fine, but on 64-bit
builds, it only works if the kernel has proper SS sigcontext
handling for 64-bit user programs.

Since the SS fixes are currently reverted, restrict the test
case to 32 bits for now.

In principle, I could change the test to use a different segment
register, but it would be messy: CS can't point to the LDT for
64-bit code, and the other registers don't result in immediate
faults because they aren't reloaded on kernel -> user
transitions.

When we fix sigcontext (in 4.6?), we can revert this.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Shuah Khan <shuahkh@osg.samsung.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/231591d9122d282402d8f53175134f8db5b3bc73.1452561752.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/mm/pat: Make split_page_count() check for empty levels to fix /proc/meminfo output
Dave Jones [Mon, 11 Jan 2016 17:04:28 +0000 (12:04 -0500)]
x86/mm/pat: Make split_page_count() check for empty levels to fix /proc/meminfo output

In CONFIG_PAGEALLOC_DEBUG=y builds, we disable 2M pages.

Unfortunatly when we split up mappings during boot,
split_page_count() doesn't take this into account, and
starts decrementing an empty direct_pages_count[] level.

This results in /proc/meminfo showing crazy things like:

  DirectMap2M:    18446744073709543424 kB

Signed-off-by: Dave Jones <davej@codemonkey.org.uk>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luis R. Rodriguez <mcgrof@suse.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Toshi Kani <toshi.kani@hp.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agoMerge commit 'linus' into x86/urgent, to pick up recent x86 changes
Ingo Molnar [Tue, 12 Jan 2016 10:08:13 +0000 (11:08 +0100)]
Merge commit 'linus' into x86/urgent, to pick up recent x86 changes

Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agoMerge branch 'x86-platform-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Tue, 12 Jan 2016 01:45:32 +0000 (17:45 -0800)]
Merge branch 'x86-platform-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 platform updates from Ingo Molnar:
 "Two changes:

   - one to quirk-save/restore certain system MSRs across
     suspend/resume, to make certain Intel systems work better
     (Chen Yu)

   - and also to constify a read only structure (Julia Lawall)"

* 'x86-platform-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/platform/calgary: Constify cal_chipset_ops structures
  x86/pm: Introduce quirk framework to save/restore extra MSR registers around suspend/resume

8 years agoMerge branch 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Tue, 12 Jan 2016 01:16:01 +0000 (17:16 -0800)]
Merge branch 'x86-mm-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 mm updates from Ingo Molnar:
 "The main changes in this cycle were:

   - make the debugfs 'kernel_page_tables' file read-only, as it only
     has read ops.  (Borislav Petkov)

   - micro-optimize clflush_cache_range() (Chris Wilson)

   - swiotlb enhancements, which fixes certain KVM emulated devices
     (Igor Mammedov)

   - fix an LDT related debug message (Jan Beulich)

   - modularize CONFIG_X86_PTDUMP (Kees Cook)

   - tone down an overly alarming warning (Laura Abbott)

   - Mark variable __initdata (Rasmus Villemoes)

   - PAT additions (Toshi Kani)"

* 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/mm: Micro-optimise clflush_cache_range()
  x86/mm/pat: Change free_memtype() to support shrinking case
  x86/mm/pat: Add untrack_pfn_moved for mremap
  x86/mm: Drop WARN from multi-BAR check
  x86/LDT: Print the real LDT base address
  x86/mm/64: Enable SWIOTLB if system has SRAT memory regions above MAX_DMA32_PFN
  x86/mm: Introduce max_possible_pfn
  x86/mm/ptdump: Make (debugfs)/kernel_page_tables read-only
  x86/mm/mtrr: Mark the 'range_new' static variable in mtrr_calc_range_state() as __initdata
  x86/mm: Turn CONFIG_X86_PTDUMP into a module

8 years agoMerge branch 'x86-fpu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Tue, 12 Jan 2016 00:56:38 +0000 (16:56 -0800)]
Merge branch 'x86-fpu-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 fpu updates from Ingo Molnar:
 "This cleans up the FPU fault handling methods to be more robust, and
  moves eligible variables to .init.data"

* 'x86-fpu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/fpu: Put a few variables in .init.data
  x86/fpu: Get rid of xstate_fault()
  x86/fpu: Add an XSTATE_OP() macro

8 years agoMerge branch 'x86-cpu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Tue, 12 Jan 2016 00:46:20 +0000 (16:46 -0800)]
Merge branch 'x86-cpu-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 cpu updates from Ingo Molnar:
 "The main changes in this cycle were:

   - Improved CPU ID handling code and related enhancements (Borislav
     Petkov)

   - RDRAND fix (Len Brown)"

* 'x86-cpu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86: Replace RDRAND forced-reseed with simple sanity check
  x86/MSR: Chop off lower 32-bit value
  x86/cpu: Fix MSR value truncation issue
  x86/cpu/amd, kvm: Satisfy guest kernel reads of IC_CFG MSR
  kvm: Add accessors for guest CPU's family, model, stepping
  x86/cpu: Unify CPU family, model, stepping calculation

8 years agoMerge branch 'x86-cleanups-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Tue, 12 Jan 2016 00:26:03 +0000 (16:26 -0800)]
Merge branch 'x86-cleanups-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 cleanups from Ingo Molnar:
 "The main changes in this cycle were:

   - code patching and cpu_has cleanups (Borislav Petkov)

   - paravirt cleanups (Juergen Gross)

   - TSC cleanup (Thomas Gleixner)

   - ptrace cleanup (Chen Gang)"

* 'x86-cleanups-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  arch/x86/kernel/ptrace.c: Remove unused arg_offs_table
  x86/mm: Align macro defines
  x86/cpu: Provide a config option to disable static_cpu_has
  x86/cpufeature: Remove unused and seldomly used cpu_has_xx macros
  x86/cpufeature: Cleanup get_cpu_cap()
  x86/cpufeature: Move some of the scattered feature bits to x86_capability
  x86/paravirt: Remove paravirt ops pmd_update[_defer] and pte_update_defer
  x86/paravirt: Remove unused pv_apic_ops structure
  x86/tsc: Remove unused tsc_pre_init() hook
  x86: Remove unused function cpu_has_ht_siblings()
  x86/paravirt: Kill some unused patching functions

8 years agoMerge branch 'x86-boot-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Tue, 12 Jan 2016 00:25:41 +0000 (16:25 -0800)]
Merge branch 'x86-boot-for-linus' of git://git./linux/kernel/git/tip/tip

Pull small x86 boot update from Ingo Molnar:
 "A single update to the MAINTAINERS file"

* 'x86-boot-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/tboot: Update maintainer list for Intel TXT

8 years agoMerge branch 'x86-asm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Mon, 11 Jan 2016 23:58:16 +0000 (15:58 -0800)]
Merge branch 'x86-asm-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 asm updates from Ingo Molnar:
 "The main changes in this cycle were:

   - vDSO and asm entry improvements (Andy Lutomirski)

   - Xen paravirt entry enhancements (Boris Ostrovsky)

   - asm entry labels enhancement (Borislav Petkov)

   - and other misc changes (Thomas Gleixner, me)"

* 'x86-asm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/vsdo: Fix build on PARAVIRT_CLOCK=y, KVM_GUEST=n
  Revert "x86/kvm: On KVM re-enable (e.g. after suspend), update clocks"
  x86/entry/64_compat: Make labels local
  x86/platform/uv: Include clocksource.h for clocksource_touch_watchdog()
  x86/vdso: Enable vdso pvclock access on all vdso variants
  x86/vdso: Remove pvclock fixmap machinery
  x86/vdso: Get pvclock data from the vvar VMA instead of the fixmap
  x86, vdso, pvclock: Simplify and speed up the vdso pvclock reader
  x86/kvm: On KVM re-enable (e.g. after suspend), update clocks
  x86/entry/64: Bypass enter_from_user_mode on non-context-tracking boots
  x86/asm: Add asm macros for static keys/jump labels
  x86/asm: Error out if asm/jump_label.h is included inappropriately
  context_tracking: Switch to new static_branch API
  x86/entry, x86/paravirt: Remove the unused usergs_sysret32 PV op
  x86/paravirt: Remove the unused irq_enable_sysexit pv op
  x86/xen: Avoid fast syscall path for Xen PV guests

8 years agoMerge branch 'x86-apic-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Mon, 11 Jan 2016 23:37:06 +0000 (15:37 -0800)]
Merge branch 'x86-apic-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 apic updates from Ingo Molnar:
 "The main changes in this cycle were:

   - introduce optimized single IPI sending methods on modern APICs
     (Linus Torvalds, Thomas Gleixner)

   - kexec/crash APIC handling fixes and enhancements (Hidehiro Kawai)

   - extend lapic vector saving/restoring to the CMCI (MCE) vector as
     well (Juergen Gross)

   - various fixes and enhancements (Jake Oshins, Len Brown)"

* 'x86-apic-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (23 commits)
  x86/irq: Export functions to allow MSI domains in modules
  Documentation: Document kernel.panic_on_io_nmi sysctl
  x86/nmi: Save regs in crash dump on external NMI
  x86/apic: Introduce apic_extnmi command line parameter
  kexec: Fix race between panic() and crash_kexec()
  panic, x86: Allow CPUs to save registers even if looping in NMI context
  panic, x86: Fix re-entrance problem due to panic on NMI
  x86/apic: Fix the saving and restoring of lapic vectors during suspend/resume
  x86/smpboot: Re-enable init_udelay=0 by default on modern CPUs
  x86/smp: Remove single IPI wrapper
  x86/apic: Use default send single IPI wrapper
  x86/apic: Provide default send single IPI wrapper
  x86/apic: Implement single IPI for apic_noop
  x86/apic: Wire up single IPI for apic_numachip
  x86/apic: Wire up single IPI for x2apic_uv
  x86/apic: Implement single IPI for x2apic_phys
  x86/apic: Wire up single IPI for bigsmp_apic
  x86/apic: Remove pointless indirections from bigsmp_apic
  x86/apic: Wire up single IPI for apic_physflat
  x86/apic: Remove pointless indirections from apic_physflat
  ...

8 years agoMerge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 11 Jan 2016 23:13:38 +0000 (15:13 -0800)]
Merge branch 'sched-core-for-linus' of git://git./linux/kernel/git/tip/tip

Pull scheduler updates from Ingo Molnar:
 "The main changes in this cycle were:

   - tickless load average calculation enhancements (Byungchul Park)

   - vtime handling enhancements (Frederic Weisbecker)

   - scalability improvement via properly aligning a key structure field
     (Jiri Olsa)

   - various stop_machine() fixes (Oleg Nesterov)

   - sched/numa enhancement (Rik van Riel)

   - various fixes and improvements (Andi Kleen, Dietmar Eggemann,
     Geliang Tang, Hiroshi Shimamoto, Joonwoo Park, Peter Zijlstra,
     Waiman Long, Wanpeng Li, Yuyang Du)"

* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (32 commits)
  sched/fair: Fix new task's load avg removed from source CPU in wake_up_new_task()
  sched/core: Move sched_entity::avg into separate cache line
  x86/fpu: Properly align size in CHECK_MEMBER_AT_END_OF() macro
  sched/deadline: Fix the earliest_dl.next logic
  sched/fair: Disable the task group load_avg update for the root_task_group
  sched/fair: Move the cache-hot 'load_avg' variable into its own cacheline
  sched/fair: Avoid redundant idle_cpu() call in update_sg_lb_stats()
  sched/core: Move the sched_to_prio[] arrays out of line
  sched/cputime: Convert vtime_seqlock to seqcount
  sched/cputime: Introduce vtime accounting check for readers
  sched/cputime: Rename vtime_accounting_enabled() to vtime_accounting_cpu_enabled()
  sched/cputime: Correctly handle task guest time on housekeepers
  sched/cputime: Clarify vtime symbols and document them
  sched/cputime: Remove extra cost in task_cputime()
  sched/fair: Make it possible to account fair load avg consistently
  sched/fair: Modify the comment about lock assumptions in migrate_task_rq_fair()
  stop_machine: Clean up the usage of the preemption counter in cpu_stopper_thread()
  stop_machine: Shift the 'done != NULL' check from cpu_stop_signal_done() to callers
  stop_machine: Kill cpu_stop_done->executed
  stop_machine: Change __stop_cpus() to rely on cpu_stop_queue_work()
  ...

8 years agoMerge branch 'ras-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Mon, 11 Jan 2016 23:07:19 +0000 (15:07 -0800)]
Merge branch 'ras-core-for-linus' of git://git./linux/kernel/git/tip/tip

Pull RAS updates from Ingo Molnar:
 "Various x86 MCE fixes and small enhancements"

* 'ras-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/mce: Make usable address checks Intel-only
  x86/mce: Add the missing memory error check on AMD
  x86/RAS: Remove mce.usable_addr
  x86/mce: Do not enter deferred errors into the generic pool twice

8 years agoMerge branch 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Mon, 11 Jan 2016 22:39:17 +0000 (14:39 -0800)]
Merge branch 'perf-core-for-linus' of git://git./linux/kernel/git/tip/tip

Pull perf updates from Ingo Molnar:
 "Kernel side changes:

   - Intel Knights Landing support.  (Harish Chegondi)

   - Intel Broadwell-EP uncore PMU support.  (Kan Liang)

   - Core code improvements.  (Peter Zijlstra.)

   - Event filter, LBR and PEBS fixes.  (Stephane Eranian)

   - Enable cycles:pp on Intel Atom.  (Stephane Eranian)

   - Add cycles:ppp support for Skylake.  (Andi Kleen)

   - Various x86 NMI overhead optimizations.  (Andi Kleen)

   - Intel PT enhancements.  (Takao Indoh)

   - AMD cache events fix.  (Vince Weaver)

  Tons of tooling changes:

   - Show random perf tool tips in the 'perf report' bottom line
     (Namhyung Kim)

   - perf report now defaults to --group if the perf.data file has
     grouped events, try it with:

      # perf record -e '{cycles,instructions}' -a sleep 1
      [ perf record: Woken up 1 times to write data ]
      [ perf record: Captured and wrote 1.093 MB perf.data (1247 samples) ]
      # perf report
      # Samples: 1K of event 'anon group { cycles, instructions }'
      # Event count (approx.): 1955219195
      #
      #       Overhead  Command     Shared Object      Symbol

         2.86%   0.22%  swapper     [kernel.kallsyms]  [k] intel_idle
         1.05%   0.33%  firefox     libxul.so          [.] js::SetObjectElement
         1.05%   0.00%  kworker/0:3 [kernel.kallsyms]  [k] gen6_ring_get_seqno
         0.88%   0.17%  chrome      chrome             [.] 0x0000000000ee27ab
         0.65%   0.86%  firefox     libxul.so          [.] js::ValueToId<(js::AllowGC)1>
         0.64%   0.23%  JS Helper   libxul.so          [.] js::SplayTree<js::jit::LiveRange*, js::jit::LiveRange>::splay
         0.62%   1.27%  firefox     libxul.so          [.] js::GetIterator
         0.61%   1.74%  firefox     libxul.so          [.] js::NativeSetProperty
         0.61%   0.31%  firefox     libxul.so          [.] js::SetPropertyByDefining

   - Introduce the 'perf stat record/report' workflow:

     Generate perf.data files from 'perf stat', to tap into the
     scripting capabilities perf has instead of defining a 'perf stat'
     specific scripting support to calculate event ratios, etc.

     Simple example:

        $ perf stat record -e cycles usleep 1

         Performance counter stats for 'usleep 1':

               1,134,996      cycles

             0.000670644 seconds time elapsed

        $ perf stat report

         Performance counter stats for '/home/acme/bin/perf stat record -e cycles usleep 1':

               1,134,996      cycles

             0.000670644 seconds time elapsed

        $

     It generates PERF_RECORD_ userspace records to store the details:

        $ perf report -D | grep PERF_RECORD
        0xf0 [0x28]: PERF_RECORD_THREAD_MAP nr: 1 thread: 27637
        0x118 [0x12]: PERF_RECORD_CPU_MAP nr: 1 cpu: 65535
        0x12a [0x40]: PERF_RECORD_STAT_CONFIG
        0x16a [0x30]: PERF_RECORD_STAT
        -1 -1 0x19a [0x40]: PERF_RECORD_MMAP -1/0: [0xffffffff81000000(0x1f000000) @ 0xffffffff81000000]: x [kernel.kallsyms]_text
        0x1da [0x18]: PERF_RECORD_STAT_ROUND
        [acme@ssdandy linux]$

     An effort was made to make perf.data files generated like this to
     not generate cryptic messages when processed by older tools.

     The 'perf script' bits need rebasing, will go up later.

   - Make command line options always available, even when they depend
     on some feature being enabled, warning the user about use of such
     options (Wang Nan)

   - Support hw breakpoint events (mem:0xAddress) in the default output
     mode in 'perf script' (Wang Nan)

   - Fixes and improvements for supporting annotating ARM binaries,
     support ARM call and jump instructions, more work needed to have
     arch specific stuff separated into tools/perf/arch/*/annotate/
     (Russell King)

   - Add initial 'perf config' command, for now just with a --list
     command to the contents of the configuration file in use and a
     basic man page describing its format, commands for doing edits and
     detailed documentation are being reviewed and proof-read.  (Taeung
     Song)

   - Allows BPF scriptlets specify arguments to be fetched using DWARF
     info, using a prologue generated at compile/build time (He Kuang,
     Wang Nan)

   - Allow attaching BPF scriptlets to module symbols (Wang Nan)

   - Allow attaching BPF scriptlets to userspace code using uprobe (Wang
     Nan)

   - BPF programs now can specify 'perf probe' tunables via its section
     name, separating key=val values using semicolons (Wang Nan)

     Testing some of these new BPF features:

        Use case: get callchains when receiving SSL packets, filter then in the
                  kernel, at arbitrary place.

        # cat ssl.bpf.c
        #define SEC(NAME) __attribute__((section(NAME), used))

        struct pt_regs;

        SEC("func=__inet_lookup_established hnum")
        int func(struct pt_regs *ctx, int err, unsigned short port)
        {
                return err == 0 && port == 443;
        }

        char _license[] SEC("license") = "GPL";
        int  _version   SEC("version") = LINUX_VERSION_CODE;
        #
        # perf record -a -g -e ssl.bpf.c
        ^C[ perf record: Woken up 1 times to write data ]
        [ perf record: Captured and wrote 0.787 MB perf.data (3 samples) ]
        # perf script | head -30
        swapper     0 [000] 58783.268118: perf_bpf_probe:func: (ffffffff816a0f60) hnum=0x1bb
           8a0f61 __inet_lookup_established (/lib/modules/4.3.0+/build/vmlinux)
           896def ip_rcv_finish (/lib/modules/4.3.0+/build/vmlinux)
           8976c2 ip_rcv (/lib/modules/4.3.0+/build/vmlinux)
           855eba __netif_receive_skb_core (/lib/modules/4.3.0+/build/vmlinux)
           8565d8 __netif_receive_skb (/lib/modules/4.3.0+/build/vmlinux)
           8572a8 process_backlog (/lib/modules/4.3.0+/build/vmlinux)
           856b11 net_rx_action (/lib/modules/4.3.0+/build/vmlinux)
           2a284b __do_softirq (/lib/modules/4.3.0+/build/vmlinux)
           2a2ba3 irq_exit (/lib/modules/4.3.0+/build/vmlinux)
           96b7a4 do_IRQ (/lib/modules/4.3.0+/build/vmlinux)
           969807 ret_from_intr (/lib/modules/4.3.0+/build/vmlinux)
           2dede5 cpu_startup_entry (/lib/modules/4.3.0+/build/vmlinux)
           95d5bc rest_init (/lib/modules/4.3.0+/build/vmlinux)
          1163ffa start_kernel ([kernel.vmlinux].init.text)
          11634d7 x86_64_start_reservations ([kernel.vmlinux].init.text)
          1163623 x86_64_start_kernel ([kernel.vmlinux].init.text)

        qemu-system-x86  9178 [003] 58785.792417: perf_bpf_probe:func: (ffffffff816a0f60) hnum=0x1bb
           8a0f61 __inet_lookup_established (/lib/modules/4.3.0+/build/vmlinux)
           896def ip_rcv_finish (/lib/modules/4.3.0+/build/vmlinux)
           8976c2 ip_rcv (/lib/modules/4.3.0+/build/vmlinux)
           855eba __netif_receive_skb_core (/lib/modules/4.3.0+/build/vmlinux)
           8565d8 __netif_receive_skb (/lib/modules/4.3.0+/build/vmlinux)
           856660 netif_receive_skb_internal (/lib/modules/4.3.0+/build/vmlinux)
           8566ec netif_receive_skb_sk (/lib/modules/4.3.0+/build/vmlinux)
             430a br_handle_frame_finish ([bridge])
             48bc br_handle_frame ([bridge])
           855f44 __netif_receive_skb_core (/lib/modules/4.3.0+/build/vmlinux)
           8565d8 __netif_receive_skb (/lib/modules/4.3.0+/build/vmlinux)
        #

   - Use 'perf probe' various options to list functions, see what
     variables can be collected at any given point, experiment first
     collecting without a filter, then filter, use it together with
     'perf trace', 'perf top', with or without callchains, if it
     explodes, please tell us!

   - Introduce a new callchain mode: "folded", that will list per line
     representations of all callchains for a give histogram entry,
     facilitating 'perf report' output processing by other tools, such
     as Brendan Gregg's flamegraph tools (Namhyung Kim)

     E.g:

        # perf report | grep -v ^# | head
           18.37%     0.00%  swapper  [kernel.kallsyms]   [k] cpu_startup_entry
                           |
                           ---cpu_startup_entry
                              |
                              |--12.07%--start_secondary
                              |
                               --6.30%--rest_init
                                         start_kernel
                                         x86_64_start_reservations
                                         x86_64_start_kernel
         #

     Becomes, in "folded" mode:

        # perf report -g folded | grep -v ^# | head -5
            18.37%     0.00%  swapper [kernel.kallsyms]   [k] cpu_startup_entry
          12.07% cpu_startup_entry;start_secondary
           6.30% cpu_startup_entry;rest_init;start_kernel;x86_64_start_reservations;x86_64_start_kernel
            16.90%     0.00%  swapper [kernel.kallsyms]   [k] call_cpuidle
          11.23% call_cpuidle;cpu_startup_entry;start_secondary
           5.67% call_cpuidle;cpu_startup_entry;rest_init;start_kernel;x86_64_start_reservations;x86_64_start_kernel
            16.90%     0.00%  swapper [kernel.kallsyms]   [k] cpuidle_enter
          11.23% cpuidle_enter;call_cpuidle;cpu_startup_entry;start_secondary
           5.67% cpuidle_enter;call_cpuidle;cpu_startup_entry;rest_init;start_kernel;x86_64_start_reservations;x86_64_start_kernel
            15.12%     0.00%  swapper [kernel.kallsyms]   [k] cpuidle_enter_state
         #

     The user can also select one of "count", "period" or "percent" as
     the first column.

  ... and lots of infrastructure enhancements, plus fixes and other
  changes, features I failed to list - see the shortlog and the git log
  for details"

* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (271 commits)
  perf evlist: Add --trace-fields option to show trace fields
  perf record: Store data mmaps for dwarf unwind
  perf libdw: Check for mmaps also in MAP__VARIABLE tree
  perf unwind: Check for mmaps also in MAP__VARIABLE tree
  perf unwind: Use find_map function in access_dso_mem
  perf evlist: Remove perf_evlist__(enable|disable)_event functions
  perf evlist: Make perf_evlist__open() open evsels with their cpus and threads (like perf record does)
  perf report: Show random usage tip on the help line
  perf hists: Export a couple of hist functions
  perf diff: Use perf_hpp__register_sort_field interface
  perf tools: Add overhead/overhead_children keys defaults via string
  perf tools: Remove list entry from struct sort_entry
  perf tools: Include all tools/lib directory for tags/cscope/TAGS targets
  perf script: Align event name properly
  perf tools: Add missing headers in perf's MANIFEST
  perf tools: Do not show trace command if it's not compiled in
  perf report: Change default to use event group view
  perf top: Decay periods in callchains
  tools lib: Move bitmap.[ch] from tools/perf/ to tools/{lib,include}/
  tools lib: Sync tools/lib/find_bit.c with the kernel
  ...

8 years agoMerge branch 'locking-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 11 Jan 2016 22:18:38 +0000 (14:18 -0800)]
Merge branch 'locking-core-for-linus' of git://git./linux/kernel/git/tip/tip

Pull locking updates from Ingo Molnar:
 "So we have a laundry list of locking subsystem changes:

   - continuing barrier API and code improvements

   - futex enhancements

   - atomics API improvements

   - pvqspinlock enhancements: in particular lock stealing and adaptive
     spinning

   - qspinlock micro-enhancements"

* 'locking-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  futex: Allow FUTEX_CLOCK_REALTIME with FUTEX_WAIT op
  futex: Cleanup the goto confusion in requeue_pi()
  futex: Remove pointless put_pi_state calls in requeue()
  futex: Document pi_state refcounting in requeue code
  futex: Rename free_pi_state() to put_pi_state()
  futex: Drop refcount if requeue_pi() acquired the rtmutex
  locking/barriers, arch: Remove ambiguous statement in the smp_store_mb() documentation
  lcoking/barriers, arch: Use smp barriers in smp_store_release()
  locking/cmpxchg, arch: Remove tas() definitions
  locking/pvqspinlock: Queue node adaptive spinning
  locking/pvqspinlock: Allow limited lock stealing
  locking/pvqspinlock: Collect slowpath lock statistics
  sched/core, locking: Document Program-Order guarantees
  locking, sched: Introduce smp_cond_acquire() and use it
  locking/pvqspinlock, x86: Optimize the PV unlock code path
  locking/qspinlock: Avoid redundant read of next pointer
  locking/qspinlock: Prefetch the next node cacheline
  locking/qspinlock: Use _acquire/_release() versions of cmpxchg() & xchg()
  atomics: Add test for atomic operations with _relaxed variants

8 years agoMerge branch 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Mon, 11 Jan 2016 21:46:11 +0000 (13:46 -0800)]
Merge branch 'core-rcu-for-linus' of git://git./linux/kernel/git/tip/tip

Pull RCU updates from Ingo Molnar:
 "The changes in this cycle were:

   - Adding transitivity uniformly to rcu_node structure ->lock
     acquisitions.  (This is implemented by the first two commits on top
     of v4.4-rc2 due to the pervasive nature of this change.)

   - Documentation updates, including RCU requirements.

   - Expedited grace-period changes.

   - Miscellaneous fixes.

   - Linked-list fixes, courtesy of KTSAN.

   - Torture-test updates.

   - Late-breaking fix to sysrq-generated crash.

  One thing I should note is that these pieces of documentation are
  fairly large files:

    .../RCU/Design/Requirements/Requirements.html      | 2897 ++++++++++++++++++++
    .../RCU/Design/Requirements/Requirements.htmlx     | 2741 ++++++++++++++++++

  and are written in HTML, not the usual .txt style.  I hope they are
  fine"

Paul McKenney explains the html docs:
 "For whatever it is worth, the reason for this unconventional choice
  was that attempts to do the diagrams in ASCII art failed miserably.

  And attempts to do ASCII art for the upcoming documentation of the
  data structures failed even more miserably"

* 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (49 commits)
  sysrq: Fix warning in sysrq generated crash.
  list: Add lockless list traversal primitives
  rcu: Make rcu_gp_init() be bool rather than int
  rcu: Move wakeup out from under rnp->lock
  rcu: Fix comment for rcu_dereference_raw_notrace
  rcu: Don't redundantly disable irqs in rcu_irq_{enter,exit}()
  rcu: Make cpu_needs_another_gp() be bool
  rcu: Eliminate unused rcu_init_one() argument
  rcu: Remove TINY_RCU bloat from pointless boot parameters
  torture: Place console.log files correctly from the get-go
  torture: Abbreviate console error dump
  rcutorture: Print symbolic name for ->gp_state
  rcutorture: Print symbolic name for rcu_torture_writer_state
  rcutorture: Remove CONFIG_RCU_USER_QS from rcutorture selftest doc
  rcutorture: Default grace period to three minutes, allow override
  rcutorture:  Dump stack when GP kthread stalls
  rcutorture: Flag nonexistent RCU GP kthread
  rcutorture: Add batch number to script printout
  Documentation/memory-barriers.txt: Fix ACCESS_ONCE thinko
  documentation: Update RCU requirements based on expedited changes
  ...

8 years agoMerge branch 'work.xattr' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Linus Torvalds [Mon, 11 Jan 2016 21:32:10 +0000 (13:32 -0800)]
Merge branch 'work.xattr' of git://git./linux/kernel/git/viro/vfs

Pull vfs xattr updates from Al Viro:
 "Andreas' xattr cleanup series.

  It's a followup to his xattr work that went in last cycle; -0.5KLoC"

* 'work.xattr' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  xattr handlers: Simplify list operation
  ocfs2: Replace list xattr handler operations
  nfs: Move call to security_inode_listsecurity into nfs_listxattr
  xfs: Change how listxattr generates synthetic attributes
  tmpfs: listxattr should include POSIX ACL xattrs
  tmpfs: Use xattr handler infrastructure
  btrfs: Use xattr handler infrastructure
  vfs: Distinguish between full xattr names and proper prefixes
  posix acls: Remove duplicate xattr name definitions
  gfs2: Remove gfs2_xattr_acl_chmod
  vfs: Remove vfs_xattr_cmp

8 years agoMerge branch 'work.symlinks' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Linus Torvalds [Mon, 11 Jan 2016 21:13:23 +0000 (13:13 -0800)]
Merge branch 'work.symlinks' of git://git./linux/kernel/git/viro/vfs

Pull vfs RCU symlink updates from Al Viro:
 "Replacement of ->follow_link/->put_link, allowing to stay in RCU mode
  even if the symlink is not an embedded one.

  No changes since the mailbomb on Jan 1"

* 'work.symlinks' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  switch ->get_link() to delayed_call, kill ->put_link()
  kill free_page_put_link()
  teach nfs_get_link() to work in RCU mode
  teach proc_self_get_link()/proc_thread_self_get_link() to work in RCU mode
  teach shmem_get_link() to work in RCU mode
  teach page_get_link() to work in RCU mode
  replace ->follow_link() with new method that could stay in RCU mode
  don't put symlink bodies in pagecache into highmem
  namei: page_getlink() and page_follow_link_light() are the same thing
  ufs: get rid of ->setattr() for symlinks
  udf: don't duplicate page_symlink_inode_operations
  logfs: don't duplicate page_symlink_inode_operations
  switch befs long symlinks to page_symlink_operations

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Linus Torvalds [Mon, 11 Jan 2016 20:54:03 +0000 (12:54 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/viro/vfs

Pull vfs compat_ioctl fixes from Al Viro:
 "This is basically Jann's patches from last week.  I have _not_
  included the stuff like switching i2c to ->compat_ioctl() into this
  one - those need more testing.

  Ideally I would like fs/compat_ioctl.c shrunk a lot, but that's a
  separate story"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  compat_ioctl: don't call do_ioctl under set_fs(KERNEL_DS)
  compat_ioctl: don't pass fd around when not needed
  compat_ioctl: don't look up the fd twice

8 years agox86/boot: Double BOOT_HEAP_SIZE to 64KB
H.J. Lu [Mon, 4 Jan 2016 18:17:09 +0000 (10:17 -0800)]
x86/boot: Double BOOT_HEAP_SIZE to 64KB

When decompressing kernel image during x86 bootup, malloc memory
for ELF program headers may run out of heap space, which leads
to system halt.  This patch doubles BOOT_HEAP_SIZE to 64KB.

Tested with 32-bit kernel which failed to boot without this patch.

Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Acked-by: H. Peter Anvin <hpa@zytor.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Cc: <stable@vger.kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agox86/mm: Add barriers and document switch_mm()-vs-flush synchronization
Andy Lutomirski [Wed, 6 Jan 2016 20:21:01 +0000 (12:21 -0800)]
x86/mm: Add barriers and document switch_mm()-vs-flush synchronization

When switch_mm() activates a new PGD, it also sets a bit that
tells other CPUs that the PGD is in use so that TLB flush IPIs
will be sent.  In order for that to work correctly, the bit
needs to be visible prior to loading the PGD and therefore
starting to fill the local TLB.

Document all the barriers that make this work correctly and add
a couple that were missing.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-mm@kvack.org
Cc: stable@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agoLinux 4.4
Linus Torvalds [Sun, 10 Jan 2016 23:01:32 +0000 (15:01 -0800)]
Linux 4.4

8 years agoMerge tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Linus Torvalds [Sat, 9 Jan 2016 22:53:48 +0000 (14:53 -0800)]
Merge tag 'scsi-fixes' of git://git./linux/kernel/git/jejb/scsi

Pull SCSI fix from James Bottomley:
 "A single fix for machines with pages > 4k (PPC mostly).

  There's a bug in our optimal transfer size code where we don't account
  for pages > 4k and can set the transfer size to be less than the page
  size causing nasty failures"

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  sd: Reject optimal transfer length smaller than page size

8 years agoMerge tag 'pci-v4.4-fixes-4' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci
Linus Torvalds [Sat, 9 Jan 2016 22:44:44 +0000 (14:44 -0800)]
Merge tag 'pci-v4.4-fixes-4' of git://git./linux/kernel/git/helgaas/pci

Pull PCI fixlet from Bjorn Helgaas:
 "This marks the TI DRA7xx host bridge driver as broken.  Apparently it
  has never worked without some additional out-of-tree code, so I'm
  going to mark it broken now and remove it completely next cycle unless
  it's fixed"

* tag 'pci-v4.4-fixes-4' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  PCI: dra7xx: Mark driver as broken

8 years agoMerge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git...
Ingo Molnar [Sat, 9 Jan 2016 16:17:33 +0000 (17:17 +0100)]
Merge tag 'perf-core-for-mingo' of git://git./linux/kernel/git/acme/linux into perf/core

Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:

New features:

- Allow using trace events fields as sort order keys, making 'perf evlist --trace_fields'
  show those, and then the user can select a subset and use like:

    perf top -e sched:sched_switch -s prev_comm,next_comm

  That works as well in 'perf report' when handling files containing
  tracepoints.

  The default when just tracepoint events are found in a perf.data file is to
  format it like ftrace, using the libtraceevent formatters, plugins, etc (Namhyung Kim)

- Add support in 'perf script' to process 'perf stat record' generated files,
  culminating in a python perf script that calculates CPI (Cycles per
  Instruction) (Jiri Olsa)

- Show random perf tool tips in the 'perf report' bottom line (Namhyung Kim)

- perf report now defaults to --group if the perf.data file has grouped events, try it with:

  # perf record -e '{cycles,instructions}' -a sleep 1
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 1.093 MB perf.data (1247 samples) ]
  # perf report
  # Samples: 1K of event 'anon group { cycles, instructions }'
  # Event count (approx.): 1955219195
  #
  #       Overhead  Command     Shared Object      Symbol

     2.86%   0.22%  swapper     [kernel.kallsyms]  [k] intel_idle
     1.05%   0.33%  firefox     libxul.so          [.] js::SetObjectElement
     1.05%   0.00%  kworker/0:3 [kernel.kallsyms]  [k] gen6_ring_get_seqno
     0.88%   0.17%  chrome      chrome             [.] 0x0000000000ee27ab
     0.65%   0.86%  firefox     libxul.so          [.] js::ValueToId<(js::AllowGC)1>
     0.64%   0.23%  JS Helper   libxul.so          [.] js::SplayTree<js::jit::LiveRange*, js::jit::LiveRange>::splay
     0.62%   1.27%  firefox     libxul.so          [.] js::GetIterator
     0.61%   1.74%  firefox     libxul.so          [.] js::NativeSetProperty
     0.61%   0.31%  firefox     libxul.so          [.] js::SetPropertyByDefining

User visible fixes:

- Coect data mmaps so that the DWARF unwinder can handle usecases needing them,
  like softice (Jiri Olsa)

- Decay callchains in fractal mode, fixing up cases where 'perf top -g' would
  show entries with more than 100% (Namhyung Kim)

Infrastructure changes:

- Sync tools/lib with the lib/ in the kernel sources for find_bit.c and
  move bitmap.[ch] from tools/perf/util/ to tools/lib/ (Arnaldo Carvalho de Melo)

- No need to set attr.sample_freq in some 'perf test' entries that only
  want to deal with PERF_RECORD_ meta-events, improve a bit error output
  for CQM test (Arnaldo Carvalho de Melo)

- Fix python binding build, adding some missing object files now required
  due to cpumap using find_bit stuff (Arnaldo Carvalho de Melo)

- tools/build improvemnts (Jiri Olsa)

- Add more files to cscope/ctags databases (Jiri Olsa)

- Do not show 'trace' in 'perf help' if it is not compiled in (Jiri Olsa)

- Make perf_evlist__open() open evsels with their cpus and threads,
  like perf record does, making them consistent (Adrian Hunter)

- Fix pmu snapshot initialization bug (Stephane Eranian)

- Add missing headers in perf's MANIFEST (Wang Nan)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agovmstat: allocate vmstat_wq before it is used
Michal Hocko [Fri, 8 Jan 2016 10:18:29 +0000 (11:18 +0100)]
vmstat: allocate vmstat_wq before it is used

kernel test robot has reported the following crash:

  BUG: unable to handle kernel NULL pointer dereference at 00000100
  IP: [<c1074df6>] __queue_work+0x26/0x390
  *pdpt = 0000000000000000 *pde = f000ff53f000ff53 *pde = f000ff53f000ff53
  Oops: 0000 [#1] PREEMPT PREEMPT SMP SMP
  CPU: 0 PID: 24 Comm: kworker/0:1 Not tainted 4.4.0-rc4-00139-g373ccbe #1
  Workqueue: events vmstat_shepherd
  task: cb684600 ti: cb7ba000 task.ti: cb7ba000
  EIP: 0060:[<c1074df6>] EFLAGS: 00010046 CPU: 0
  EIP is at __queue_work+0x26/0x390
  EAX: 00000046 EBX: cbb37800 ECX: cbb37800 EDX: 00000000
  ESI: 00000000 EDI: 00000000 EBP: cb7bbe68 ESP: cb7bbe38
   DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
  CR0: 8005003b CR2: 00000100 CR3: 01fd5000 CR4: 000006b0
  Stack:
  Call Trace:
    __queue_delayed_work+0xa1/0x160
    queue_delayed_work_on+0x36/0x60
    vmstat_shepherd+0xad/0xf0
    process_one_work+0x1aa/0x4c0
    worker_thread+0x41/0x440
    kthread+0xb0/0xd0
    ret_from_kernel_thread+0x21/0x40

The reason is that start_shepherd_timer schedules the shepherd work item
which uses vmstat_wq (vmstat_shepherd) before setup_vmstat allocates
that workqueue so if the further initialization takes more than HZ we
might end up scheduling on a NULL vmstat_wq.  This is really unlikely
but not impossible.

Fixes: 373ccbe59270 ("mm, vmstat: allow WQ concurrency to discover memory reclaim doesn't make any progress")
Reported-by: kernel test robot <ying.huang@linux.intel.com>
Signed-off-by: Michal Hocko <mhocko@suse.com>
Tested-by: Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp>
Cc: stable@vger.kernel.org
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
8 years agocompat_ioctl: don't call do_ioctl under set_fs(KERNEL_DS)
Jann Horn [Tue, 5 Jan 2016 17:27:30 +0000 (18:27 +0100)]
compat_ioctl: don't call do_ioctl under set_fs(KERNEL_DS)

This replaces all code in fs/compat_ioctl.c that translated
ioctl arguments into a in-kernel structure, then performed
do_ioctl under set_fs(KERNEL_DS), with code that allocates
data on the user stack and can call the VFS ioctl handler
under USER_DS.

This is done as a hardening measure because the caller
does not know what kind of ioctl handler will be invoked,
only that no corresponding compat_ioctl handler exists and
what the ioctl command number is. The accidental
invocation of an unlocked_ioctl handler that unexpectedly
calls copy_to_user could be a severe security issue.

Signed-off-by: Jann Horn <jann@thejh.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
8 years agocompat_ioctl: don't pass fd around when not needed
Al Viro [Thu, 7 Jan 2016 14:53:30 +0000 (09:53 -0500)]
compat_ioctl: don't pass fd around when not needed

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
8 years agocompat_ioctl: don't look up the fd twice
Jann Horn [Tue, 5 Jan 2016 17:27:29 +0000 (18:27 +0100)]
compat_ioctl: don't look up the fd twice

In code in fs/compat_ioctl.c that translates ioctl arguments
into a in-kernel structure, then performs sys_ioctl, possibly
under set_fs(KERNEL_DS), this commit changes the sys_ioctl
calls to do_ioctl calls. do_ioctl is a new function that does
the same thing as sys_ioctl, but doesn't look up the fd again.

This change is made to avoid (potential) security issues
because of ioctl handlers that accept one of the ioctl
commands I2C_FUNCS, VIDEO_GET_EVENT, MTIOCPOS, MTIOCGET,
TIOCGSERIAL, TIOCSSERIAL, RTC_IRQP_READ, RTC_EPOCH_READ.
This can happen for multiple reasons:

 - The ioctl command number could be reused.
 - The ioctl handler might not check the full ioctl
   command. This is e.g. true for drm_ioctl.
 - The ioctl handler is very special, e.g. cuse_file_ioctl

The real issue is that set_fs(KERNEL_DS) is used here,
but that's fixed in a separate commit
"compat_ioctl: don't call do_ioctl under set_fs(KERNEL_DS)".

This change mitigates potential security issues by
preventing a race that permits invocation of
unlocked_ioctl handlers under KERNEL_DS through compat
code even if a corresponding compat_ioctl handler exists.

So far, no way has been identified to use this to damage
kernel memory without having CAP_SYS_ADMIN in the init ns
(with the capability, doing reads/writes at arbitrary
kernel addresses should be easy through CUSE's ioctl
handler with FUSE_IOCTL_UNRESTRICTED set).

[AV: two missed sys_ioctl() taken care of]

Signed-off-by: Jann Horn <jann@thejh.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
8 years agoMerge tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm...
Linus Torvalds [Sat, 9 Jan 2016 00:11:05 +0000 (16:11 -0800)]
Merge tag 'fixes-for-linus' of git://git./linux/kernel/git/arm/arm-soc

Pull ARM SoC fixes from Arnd Bergmann:
 "This is the final small set of ARM SoC bug fixes for linux-4.4, almost
  all regressions:

  OMAP:
   - data corruption on the Nokia N900 flash

  Allwinner:
   - Two defconfig change to get USB working again

  ARM Versatile:
   - Interrupt numbers gone bad after an older bug fix

  Nomadik:
   - Crashes from incorrect L2 cache settings

  VIA vt8500:
   - SD/MMC support on WM8650 never worked"

* tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
  dts: vt8500: Add SDHC node to DTS file for WM8650
  ARM: Fix broken USB support in multi_v7_defconfig for sunxi devices
  ARM: versatile: fix MMC/SD interrupt assignment
  ARM: nomadik: set latencies to 8 cycles
  ARM: OMAP2+: Fix onenand rate detection to avoid filesystem corruption
  ARM: Fix broken USB support in sunxi_defconfig

8 years agoMerge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Linus Torvalds [Fri, 8 Jan 2016 23:58:14 +0000 (15:58 -0800)]
Merge tag 'for-linus' of git://git./virt/kvm/kvm

Pull KVM fix from Paolo Bonzini:
 "A simple fix.  I'm sending it before the merge window, because it
  refines a patch found in your master branch but not yet in the
  kvm/next branch that is destined for 4.5"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  kvm: x86: only channel 0 of the i8254 is linked to the HPET

8 years agoMerge tag 'pm+acpi-4.4-final' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael...
Linus Torvalds [Fri, 8 Jan 2016 23:50:59 +0000 (15:50 -0800)]
Merge tag 'pm+acpi-4.4-final' of git://git./linux/kernel/git/rafael/linux-pm

Pull ACPI fix from Rafael Wysocki:
 "Just one obvious fix that adds a missing function argument in ACPI
  code introduced recently (Kees Cook)"

* tag 'pm+acpi-4.4-final' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  ACPI / property: avoid leaking format string into kobject name

8 years agoMerge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 8 Jan 2016 23:21:48 +0000 (15:21 -0800)]
Merge branch 'x86-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 fixes from Ingo Molnar:
 "A handful of x86 fixes:

   - a syscall ABI fix, fixing an Android breakage
   - a Xen PV guest fix relating to the RTC device, causing a
     non-working console
   - a Xen guest syscall stack frame fix
   - an MCE hotplug CPU crash fix"

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/numachip: Fix NumaConnect2 MMCFG PCI access
  x86/entry: Restore traditional SYSENTER calling convention
  x86/entry: Fix some comments
  x86/paravirt: Prevent rtc_cmos platform device init on PV guests
  x86/xen: Avoid fast syscall path for Xen PV guests
  x86/mce: Ensure offline CPUs don't participate in rendezvous process

8 years agoMerge branch 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 8 Jan 2016 21:57:13 +0000 (13:57 -0800)]
Merge branch 'sched-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull scheduler fixes from Ingo Molnar:
 "Misc scheduler fixes"

* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched/core: Reset task's lockless wake-queues on fork()
  sched/core: Fix unserialized r-m-w scribbling stuff
  sched/core: Check tgid in is_global_init()
  sched/fair: Fix multiplication overflow on 32-bit systems

8 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 8 Jan 2016 21:52:59 +0000 (13:52 -0800)]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull perf fixes from Ingo Molnar:
 "Two core subsystem fixes, plus a handful of tooling fixes"

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf: Fix race in swevent hash
  perf: Fix race in perf_event_exec()
  perf list: Robustify event printing routine
  perf list: Add support for PERF_COUNT_SW_BPF_OUT
  perf hists browser: Fix segfault if use symbol filter in cmdline
  perf hists browser: Reset selection when refresh
  perf hists browser: Add NULL pointer check to prevent crash
  perf buildid-list: Fix return value of perf buildid-list -k
  perf buildid-list: Show running kernel build id fix

8 years agoMerge branch 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 8 Jan 2016 21:46:59 +0000 (13:46 -0800)]
Merge branch 'irq-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull irq fix from Ingo Molnar:
 "Fixes a core IRQ subsystem deadlock"

* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  genirq: Prevent chip buslock deadlock

8 years agoMerge branch 'for-linus' of git://git.kernel.dk/linux-block
Linus Torvalds [Fri, 8 Jan 2016 21:39:09 +0000 (13:39 -0800)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-block

Pull block revert from Jens Axboe:
 "The previous pull request had a split fix for NVMe, however there are
  corner cases where that ends up blowing up.

  So let's revert it for 4.4.  The regression isn't introduced in this
  cycle, and it's "just" a performance regression, not a
  stability/integrity issue"

* 'for-linus' of git://git.kernel.dk/linux-block:
  Revert "block: Split bios on chunk boundaries"

8 years agoMerge tag 'dmaengine-fix-4.4' of git://git.infradead.org/users/vkoul/slave-dma
Linus Torvalds [Fri, 8 Jan 2016 20:23:00 +0000 (12:23 -0800)]
Merge tag 'dmaengine-fix-4.4' of git://git.infradead.org/users/vkoul/slave-dma

Pull dmaengine fixes from Vinod Koul:
 "Late fixes for 4.4 are three fixes for drivers which include a revert
  of mic-x100 fix which is causing regression, xgene fix for double IRQ
  and async_tx fix to use GFP_NOWAIT"

* tag 'dmaengine-fix-4.4' of git://git.infradead.org/users/vkoul/slave-dma:
  dmaengine: xgene-dma: Fix double IRQ issue by setting IRQ_DISABLE_UNLAZY flag
  async_tx: use GFP_NOWAIT rather than GFP_IO
  dmaengine: Revert "dmaengine: mic_x100: add missing spin_unlock"

8 years agoMerge branch 'dmi-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvar...
Linus Torvalds [Fri, 8 Jan 2016 20:18:45 +0000 (12:18 -0800)]
Merge branch 'dmi-for-linus' of git://git./linux/kernel/git/jdelvare/staging

Pull dmi fix from Jean Delvare.

* 'dmi-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:
  firmware: dmi_scan: Fix UUID endianness for SMBIOS >= 2.6

8 years agoMerge tag 'sound-4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Linus Torvalds [Fri, 8 Jan 2016 19:52:18 +0000 (11:52 -0800)]
Merge tag 'sound-4.4' of git://git./linux/kernel/git/tiwai/sound

Pull sound fixes from Takashi Iwai:
 "A slightly higher volume than a new year's wish, but not too
  worrisome: a large LOC is only for HD-audio device-specific quirks, so
  fairly safe to apply.  The rest ASoC fixes are all trivial and small;
  a simple replacement of mutex call with nested lock version, a few
  Arizona and Realtek codec fixes, and a regression fix for Skylake
  firmware handling"

* tag 'sound-4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ASoC: Intel: Skylake: Fix the memory leak
  ASoC: Intel: Skylake: Revert previous broken fix memory leak fix
  ASoC: Use nested lock for snd_soc_dapm_mutex_lock
  ASoC: rt5645: add sys clk detection
  ALSA: hda - Add keycode map for alc input device
  ALSA: hda - Add mic mute hotkey quirk for Lenovo ThinkCentre AIO
  ASoC: arizona: Fix bclk for sample rates that are multiple of 4kHz

8 years agox86/mm: Micro-optimise clflush_cache_range()
Chris Wilson [Fri, 8 Jan 2016 09:55:33 +0000 (09:55 +0000)]
x86/mm: Micro-optimise clflush_cache_range()

Whilst inspecting the asm for clflush_cache_range() and some perf profiles
that required extensive flushing of single cachelines (from part of the
intel-gpu-tools GPU benchmarks), we noticed that gcc was reloading
boot_cpu_data.x86_clflush_size on every iteration of the loop. We can
manually hoist that read which perf regarded as taking ~25% of the
function time for a single cacheline flush.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Toshi Kani <toshi.kani@hpe.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Luis R. Rodriguez <mcgrof@suse.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Sai Praneeth <sai.praneeth.prakhya@intel.com>
Link: http://lkml.kernel.org/r/1452246933-10890-1-git-send-email-chris@chris-wilson.co.uk
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
8 years agoperf evlist: Add --trace-fields option to show trace fields
Namhyung Kim [Thu, 7 Jan 2016 00:12:29 +0000 (09:12 +0900)]
perf evlist: Add --trace-fields option to show trace fields

To use dynamic sort keys, it might be good to add an option to see the
list of field names.

  $ perf evlist -i perf.data.sched
  sched:sched_switch
  sched:sched_stat_wait
  sched:sched_stat_sleep
  sched:sched_stat_iowait
  sched:sched_stat_runtime
  sched:sched_process_fork
  sched:sched_wakeup
  sched:sched_wakeup_new
  sched:sched_migrate_task
  # Tip: use 'perf evlist --trace-fields' to show fields for tracepoint events

  $ perf evlist -i perf.data.sched --trace-fields
  sched:sched_switch: trace_fields: prev_comm,prev_pid,prev_prio,prev_state,next_comm,next_pid,next_prio
  sched:sched_stat_wait: trace_fields: comm,pid,delay
  sched:sched_stat_sleep: trace_fields: comm,pid,delay
  sched:sched_stat_iowait: trace_fields: comm,pid,delay
  sched:sched_stat_runtime: trace_fields: comm,pid,runtime,vruntime
  sched:sched_process_fork: trace_fields: parent_comm,parent_pid,child_comm,child_pid
  sched:sched_wakeup: trace_fields: comm,pid,prio,success,target_cpu
  sched:sched_wakeup_new: trace_fields: comm,pid,prio,success,target_cpu
  sched:sched_migrate_task: trace_fields: comm,pid,prio,orig_cpu,dest_cpu

Committer notes:

For another file, in verbose mode:

  # perf evlist -v --trace-fields
  sched:sched_switch: type: 2, size: 112, config: 0x10b, { sample_period, sample_freq }: 1, sample_type: IP|TID|TIME|CPU|PERIOD|RAW, disabled: 1, inherit: 1, mmap: 1, comm: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, trace_fields: prev_comm,prev_pid,prev_prio,prev_state,next_comm,next_pid,next_prio
  #

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1452125549-1511-5-git-send-email-namhyung@kernel.org
[ Replaced 'trace_fields=' with 'trace_fields: ' to make the output consistent in -v mode ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf record: Store data mmaps for dwarf unwind
Jiri Olsa [Thu, 7 Jan 2016 13:30:22 +0000 (14:30 +0100)]
perf record: Store data mmaps for dwarf unwind

Currently we don't synthesize data mmap by default. It depends on -d
option, that enables data address sampling.

But we've seen cases (softice) where DWARF unwinder went through non
executable mmaps, which we need to lookup in MAP__VARIABLE tree.

Making data mmaps to be synthesized for dwarf unwind as well.

Reported-by: Noel Grandin <noelgrandin@gmail.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20160107133022.GA32115@krava.brq.redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf libdw: Check for mmaps also in MAP__VARIABLE tree
Jiri Olsa [Thu, 7 Jan 2016 09:14:02 +0000 (10:14 +0100)]
perf libdw: Check for mmaps also in MAP__VARIABLE tree

We've seen cases (softice) where DWARF unwinder went through non
executable mmaps, which we need to lookup in MAP__VARIABLE tree.

Reported-and-Tested-by: Noel Grandin <noelgrandin@gmail.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-6-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf unwind: Check for mmaps also in MAP__VARIABLE tree
Jiri Olsa [Thu, 7 Jan 2016 09:14:01 +0000 (10:14 +0100)]
perf unwind: Check for mmaps also in MAP__VARIABLE tree

We've seen cases (softice) where DWARF unwinder went through non
executable mmaps, which we need to lookup in MAP__VARIABLE tree.

Reported-and-Tested-by: Noel Grandin <noelgrandin@gmail.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-5-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf unwind: Use find_map function in access_dso_mem
Jiri Olsa [Thu, 7 Jan 2016 09:14:00 +0000 (10:14 +0100)]
perf unwind: Use find_map function in access_dso_mem

The find_map helper is already there, so let's use it.

Also we're going to introduce wider search in following patch, so it'll
be easier to make this change on single place.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Noel Grandin <noelgrandin@gmail.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-4-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf evlist: Remove perf_evlist__(enable|disable)_event functions
Jiri Olsa [Thu, 7 Jan 2016 09:13:58 +0000 (10:13 +0100)]
perf evlist: Remove perf_evlist__(enable|disable)_event functions

Replacing them with perf_evsel__(enable|disable).

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Noel Grandin <noelgrandin@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-2-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf evlist: Make perf_evlist__open() open evsels with their cpus and threads (like...
Adrian Hunter [Thu, 7 Jan 2016 09:13:59 +0000 (10:13 +0100)]
perf evlist: Make perf_evlist__open() open evsels with their cpus and threads (like perf record does)

'perf record' uses perf_evsel__open() to open events and passes the
evsel->cpus and evsel->threads.  Many tests and some tools instead use
perf_evlist__open() which passes instead evlist->cpus and
evlist->threads.

Make perf_evlist__open() follow the 'perf record' behaviour so that a
consistent approach is taken.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Noel Grandin <noelgrandin@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-3-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoMerge tag 'omap-for-v4.4/onenand-corruption' of git://git.kernel.org/pub/scm/linux...
Arnd Bergmann [Fri, 8 Jan 2016 16:46:45 +0000 (17:46 +0100)]
Merge tag 'omap-for-v4.4/onenand-corruption' of git://git./linux/kernel/git/tmlind/linux-omap into fixes

Pull "urgent onenand file system corruption fix for n900" from Tony Lindgren:

Last minute urgent pull request to prevent file system corruption
on Nokia N900.

Looks like we have a GPMC bus timing bug that has gone unnoticed
because of bootloader configured registers until few days ago. We
are not detecting the onenand clock rate properly unless we have
CONFIG_OMAP_GPMC_DEBUG set and this causes onenand corruption
that can be easily be reproduced.

There seems to be also an additional bug still lurking around for
onenand corruption. But that is still being investigated and
it does not seem to be GPMC timings related.

Meanwhile, it would be good to get this fix into v4.4 to prevent
wrong timings from corrupting onenand.

* tag 'omap-for-v4.4/onenand-corruption' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
  ARM: OMAP2+: Fix onenand rate detection to avoid filesystem corruption

8 years agoperf report: Show random usage tip on the help line
Namhyung Kim [Thu, 7 Jan 2016 11:41:53 +0000 (20:41 +0900)]
perf report: Show random usage tip on the help line

Currently perf report only shows a help message "For a higher level
overview, try: perf report --sort comm,dso" unconditionally (even if
the sort keys were used).  Add more help tips and show randomly.

Load tips from ${prefix}/share/doc/perf-tip/tips.txt file.

  $ perf report | tail
      0.10%  swapper  [kernel.vmlinux]   [k] irq_exit
      0.09%  swapper  [kernel.vmlinux]   [k] flush_smp_call_function_queue
      0.08%  swapper  [kernel.vmlinux]   [k] native_write_msr_safe
      0.03%  swapper  [kernel.vmlinux]   [k] group_sched_in
      0.01%  perf     [kernel.vmlinux]   [k] native_write_msr_safe

  #
  # (Tip: Search options using a keyword: perf report -h <keyword>)
  #

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Ingo Molnar <mingo@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1452166913-27046-1-git-send-email-namhyung@kernel.org
[ Renamed it to perf_tip() and the parameter dirname to dirpath to fix the build on older distros ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoRevert "block: Split bios on chunk boundaries"
Jens Axboe [Fri, 8 Jan 2016 16:00:29 +0000 (09:00 -0700)]
Revert "block: Split bios on chunk boundaries"

This reverts commit d3805611130af9b911e908af9f67a3f64f4f0914.

If we end up splitting on the first segment, we don't adjust
the sector count. That results in hitting a BUG() with attempting
to split 0 sectors.

As this is just a performance issue and not a regression since
4.3 release, let's just rever this change. That gives us more
time to test a real fix for 4.5, which would be marked for
stable anyway.

8 years agoperf hists: Export a couple of hist functions
Namhyung Kim [Thu, 7 Jan 2016 09:14:10 +0000 (10:14 +0100)]
perf hists: Export a couple of hist functions

These are necessary for multi threaded sample processing:

 - hists__get__get_rotate_entries_in()
 - hists__collapse_insert_entry()
 - __hists__init()

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Noel Grandin <noelgrandin@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-14-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf diff: Use perf_hpp__register_sort_field interface
Jiri Olsa [Thu, 7 Jan 2016 09:14:09 +0000 (10:14 +0100)]
perf diff: Use perf_hpp__register_sort_field interface

Using perf_hpp__register_sort_field interface instead of directly adding
the entry.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Noel Grandin <noelgrandin@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-13-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf tools: Add overhead/overhead_children keys defaults via string
Jiri Olsa [Thu, 7 Jan 2016 09:14:08 +0000 (10:14 +0100)]
perf tools: Add overhead/overhead_children keys defaults via string

We currently set 'overhead' and 'overhead_children' as default sort keys
within perf_hpp__init function by directly adding into the sort list.

This patch adds 'overhead' and 'overhead_children' in text form into
sort_keys and let them be added by standard sort dimension interface.

We need to eliminate dirrect sort_list additions to be able to add
support for hists specific sort keys.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Noel Grandin <noelgrandin@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-12-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoPCI: dra7xx: Mark driver as broken
Richard Cochran [Fri, 8 Jan 2016 15:58:31 +0000 (09:58 -0600)]
PCI: dra7xx: Mark driver as broken

Mark the dra7xx PCI host driver as broken.  This driver was first merged in
v3.17 and has never worked.  Although the driver compiles just fine, it is
missing an essential device reset.  If the driver is included, the kernel
locks up hard shortly after booting, before any console output appears.

Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
8 years agoperf tools: Remove list entry from struct sort_entry
Jiri Olsa [Thu, 7 Jan 2016 09:14:07 +0000 (10:14 +0100)]
perf tools: Remove list entry from struct sort_entry

It's no longer needed.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Noel Grandin <noelgrandin@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-11-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf tools: Include all tools/lib directory for tags/cscope/TAGS targets
Jiri Olsa [Thu, 7 Jan 2016 09:14:06 +0000 (10:14 +0100)]
perf tools: Include all tools/lib directory for tags/cscope/TAGS targets

Besides lockdep we use all the 'tools/lib' code in perf, so include it
completely in tags.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Noel Grandin <noelgrandin@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-10-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf script: Align event name properly
Jiri Olsa [Thu, 7 Jan 2016 09:14:05 +0000 (10:14 +0100)]
perf script: Align event name properly

Adding code to align event names, so we get aligned output in case of
multiple events with different names.

Before:
  $ perf script
  :13757 13757 163918.230829: cpu/mem-snp-none/P: ffff88085f20d010
  :13757 13757 163918.230832: cpu/mem-loads,ldlat=30/P:     7f5a5f719f00
  :13757 13757 163918.230835: cpu/mem-loads,ldlat=30/P:     7f5a5f719f00
  :13758 13758 163918.230838: cpu/mem-snp-none/P: ffff88085f4ad810
  :13758 13758 163918.154093: cpu/mem-stores/P: ffff88085bb53f28
  :13757 13757 163918.155264: cpu/mem-snp-hitm/P:           601080
  ...

After:
  $ perf script
  :13757 13757 163918.228831:       cpu/mem-snp-none/P: ffffffff81a841c0
  :13757 13757 163918.228834: cpu/mem-loads,ldlat=30/P:     7f5a5f719f08
  :13757 13757 163918.228837: cpu/mem-loads,ldlat=30/P:     7f5a5f719f08
  :13758 13758 163918.228837:       cpu/mem-snp-none/P: ffff88085f4ad800
  :13758 13758 163918.154093:         cpu/mem-stores/P: ffff88085bb53f28
  :13757 13757 163918.155264:       cpu/mem-snp-hitm/P:           601080
  ...

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: David Ahern <dsahern@gmail.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Noel Grandin <noelgrandin@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-9-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf tools: Add missing headers in perf's MANIFEST
Wang Nan [Fri, 8 Jan 2016 14:23:57 +0000 (14:23 +0000)]
perf tools: Add missing headers in perf's MANIFEST

These lost headers are found in arm64 cross buildings, failing to build
perf using tarballs generated using:

 $ make perf-targz-src-pkg

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1452263041-225488-3-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf tools: Do not show trace command if it's not compiled in
Jiri Olsa [Thu, 7 Jan 2016 09:14:04 +0000 (10:14 +0100)]
perf tools: Do not show trace command if it's not compiled in

The trace command still appears in help message when you run simple
'perf' command.

It's because the generate-cmdlist.sh does not care about the
HAVE_LIBAUDIT_SUPPORT dependency of trace command and puts it into
generated common_cmds array.

Wrapping trace command under HAVE_LIBAUDIT_SUPPORT dependency, which
will exclude it from common_cmds array if HAVE_LIBAUDIT_SUPPORT is not
set.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Noel Grandin <noelgrandin@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1452158050-28061-8-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf report: Change default to use event group view
Namhyung Kim [Sun, 29 Nov 2015 14:24:17 +0000 (23:24 +0900)]
perf report: Change default to use event group view

The event group view feature is to see related events together.  To use
the group view, events should be recorded as a group with a dedicated
syntax of surrounding events by braces (-e '{ evt1, evt2, ... }').

Also 'perf report' also requires the --group option to enable it.
However it's almost always beneficial to use the group view to see the
group events as it's more expressive.  And I think it's more natural to
see events together if they are recorded as a group.

Thus this patch changes the default value to enable it.  If users don't
want to see like it and keep the original behavior, they can set the
report.group config variable to false and/or use --no-group option in
the 'perf report' command line.

Requested-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Cc: Taeung Song <treeze.taeung@gmail.com>
Link: http://lkml.kernel.org/r/1448807057-3506-1-git-send-email-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf top: Decay periods in callchains
Namhyung Kim [Tue, 5 Jan 2016 03:06:00 +0000 (12:06 +0900)]
perf top: Decay periods in callchains

It missed to decay periods in callchains when decaying hist entries.
This resulted in more than 100 percent overhead in callchains in the
fractal style output.

Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1451963160-17196-1-git-send-email-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agotools lib: Move bitmap.[ch] from tools/perf/ to tools/{lib,include}/
Arnaldo Carvalho de Melo [Fri, 8 Jan 2016 15:33:37 +0000 (12:33 -0300)]
tools lib: Move bitmap.[ch] from tools/perf/ to tools/{lib,include}/

So that lib/find_bit.c doesn't requires anything inside tools/perf/

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: David Ahern <dsahern@gmail.com>
Cc: George Spelvin <linux@horizon.com
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: Yury Norov <yury.norov@gmail.com>
Link: http://lkml.kernel.org/n/tip-7lxe7jgohaac5faodndhdmvk@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agotools lib: Sync tools/lib/find_bit.c with the kernel
Arnaldo Carvalho de Melo [Fri, 8 Jan 2016 14:26:43 +0000 (11:26 -0300)]
tools lib: Sync tools/lib/find_bit.c with the kernel

Need to move the bitmap.[ch] things from tools/perf/ to tools/lib, will
be done in the next patches.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: David Ahern <dsahern@gmail.com>
Cc: George Spelvin <linux@horizon.com
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: Yury Norov <yury.norov@gmail.com>
Link: http://lkml.kernel.org/n/tip-5fys65wkd7gu8j7a7xgukc5t@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agotools lib: Move find_next_bit.c to tools/lib/
Arnaldo Carvalho de Melo [Fri, 8 Jan 2016 13:46:52 +0000 (10:46 -0300)]
tools lib: Move find_next_bit.c to tools/lib/

The commit that introduced it should've moved it to the same place, plus
the 'tools/' prefix, but instead moved it to a bogus tools/lib/util/
directory, being the only file there.

Move it to tools/lib/find_bit.c, picking the name for the file where
these routines live since:

 8f6f19dd5143 ("lib: move find_last_bit to lib/find_next_bit.c")

Next step is to make tools/lib/find_bit.c to differ from lib/find_bit.c
just in removing what is not used by tools/.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: David Ahern <dsahern@gmail.com>
Cc: George Spelvin <linux@horizon.com
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: Yury Norov <yury.norov@gmail.com>
Link: http://lkml.kernel.org/n/tip-p391cex5mqvahp4pwrton87n@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agofirmware: dmi_scan: Fix UUID endianness for SMBIOS >= 2.6
Andrea Arcangeli [Fri, 8 Jan 2016 08:00:54 +0000 (09:00 +0100)]
firmware: dmi_scan: Fix UUID endianness for SMBIOS >= 2.6

The dmi_ver wasn't updated correctly before the dmi_decode method run
to save the uuid.

That resulted in "dmidecode -s system-uuid" and
/sys/class/dmi/id/product_uuid disagreeing. The latter was buggy and
this fixes it.

Reported-by: Federico Simoncelli <fsimonce@redhat.com>
Fixes: 9f9c9cbb6057 ("drivers/firmware/dmi_scan.c: fetch dmi version from SMBIOS if it exists")
Fixes: 79bae42d51a5 ("dmi_scan: refactor dmi_scan_machine(), {smbios,dmi}_present()")
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Jean Delvare <jdelvare@suse.de>
8 years agoACPI / property: avoid leaking format string into kobject name
Kees Cook [Thu, 7 Jan 2016 19:24:29 +0000 (11:24 -0800)]
ACPI / property: avoid leaking format string into kobject name

The dn->name is expected to be used as a literal, so add the missing
"%s".

Fixes: 263b4c1a64bc (ACPI / property: Expose data-only subnodes via sysfs)
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
8 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Thu, 7 Jan 2016 21:06:35 +0000 (13:06 -0800)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

Pull drm nouveau fix from Dave Airlie:
 "Still not back to work, but I decided to forward this fix"

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm/nouveau/gr/nv40: fix oops in interrupt handler

8 years agoMerge tag 'iommu-fixes-v4.4-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Thu, 7 Jan 2016 20:56:23 +0000 (12:56 -0800)]
Merge tag 'iommu-fixes-v4.4-rc8' of git://git./linux/kernel/git/joro/iommu

Pull IOMMU fixes from Joerg Roedel:

 - Two build issues, one in the ipmmu-vmsa driver and one for the new
   generic dma-api implemention used on arm64

 - A performance fix for said dma-api implemention

 - An issue caused by a wrong offset in map_sg in the same code as above

* tag 'iommu-fixes-v4.4-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:
  iommu/dma: Use correct offset in map_sg
  iommu/ipmmu-vmsa: Don't truncate ttbr if LPAE is not enabled
  iommu/dma: Avoid unlikely high-order allocations
  iommu/dma: Add some missing #includes

8 years agoMerge tag 'trace-v4.4-rc4-4' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt...
Linus Torvalds [Thu, 7 Jan 2016 20:42:22 +0000 (12:42 -0800)]
Merge tag 'trace-v4.4-rc4-4' of git://git./linux/kernel/git/rostedt/linux-trace

Pull ftrace fix from Steven Rostedt:
 "PeiyangX Qiu reported that if a module fails to load between calling
  ftrace_module_init() and do_init_module() that the allocations made in
  ftrace_module_init() will not be freed, resulting in a memory leak.

  The solution is to call ftrace_release_mod() on the failing module in
  the fail path befor do_init_module() is called.  This will remove any
  allocations made for that module, and nothing if ftrace_module_init()
  wasn't called yet for that module.

  Note, once do_init_module() is called, the MODULE_GOING notifiers are
  called for the failed module, which calls into the ftrace code to do
  the proper clean up (basically calling ftrace_release_mod())"

* tag 'trace-v4.4-rc4-4' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
  ftrace/module: Call clean up function when module init fails early

8 years agoperf tests: Give a bit more information on the CQM test failure path
Arnaldo Carvalho de Melo [Thu, 7 Jan 2016 19:59:27 +0000 (16:59 -0300)]
perf tests: Give a bit more information on the CQM test failure path

Before:

  $ perf test -v cqm
  48: Test intel cqm nmi context read                          :
  --- start ---
  test child forked, pid 1681
  parse_events failed
  test child finished with -2
  ---- end ----
  Test intel cqm nmi context read: Skip
  $

After:

  $ perf test -v cqm
  48: Test intel cqm nmi context read                          :
  --- start ---
  test child forked, pid 1681
  parse_events failed, is "intel_cqm/llc_occupancy/" available?
  test child finished with -2
  ---- end ----
  Test intel cqm nmi context read: Skip
  $

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Matt Fleming <matt.fleming@intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-eidpiv5x4nkbsx37xwikbnir@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf tests: No need to set attr.sample_freq for tracking !PERF_RECORD_SAMPLE
Arnaldo Carvalho de Melo [Thu, 7 Jan 2016 19:51:58 +0000 (16:51 -0300)]
perf tests: No need to set attr.sample_freq for tracking !PERF_RECORD_SAMPLE

We were asking for a 4kHz sample_freq, making the test fail needlessly
when the system reduced /proc/sys/kernel/perf_event_max_sample_rate
below that.

Before:

  # perf test -vv dummy
  23: Test using a dummy software event to keep tracking       :
  --- start ---
  test child forked, pid 32421
  ------------------------------------------------------------
  perf_event_attr:
    type                             1
    size                             112
    config                           0x9
    { sample_period, sample_freq }   4000
    sample_type                      IP|TID|ID|PERIOD
  <SNIP>
  sys_perf_event_open failed, error -22
  Unable to open dummy and cycles event
  test child finished with -2
  ---- end ----
  Test using a dummy software event to keep tracking: Skip
  #
  [root@zoo ~]# cat /proc/sys/kernel/perf_event_max_sample_rate
  1000

After:

  [root@zoo ~]# perf test dummy
  23: Test using a dummy software event to keep tracking       : Ok

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-487iquegrs2379e5n0pi0tcp@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
8 years agoperf python: Add missing files to binding link list
Arnaldo Carvalho de Melo [Thu, 7 Jan 2016 19:47:11 +0000 (16:47 -0300)]
perf python: Add missing files to binding link list

Fixing this problem, introduced recently:

  $ perf test python
  16: Try 'import perf' in python, checking link problems      : FAILED!

In verbose mode we find out what is missing:

  $ perf test -v python
  16: Try 'import perf' in python, checking link problems      :
  --- start ---
  test child forked, pid 24894
  Traceback (most recent call last):
    File "<stdin>", line 1, in <module>
  ImportError: /tmp/build/perf/python/perf.so: undefined symbol: find_next_bit
  test child finished with -1
  ---- end ----
  Try 'import perf' in python, checking link problems: FAILED!
  $

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Fixes: f77b57ad4fc4 ("perf cpu_map: Add cpu_map__new_event function")
Link: http://lkml.kernel.org/n/tip-rajx0zkz6czdrnvvwf0jp76p@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>