scripts/headers_install.sh will transform __packed to
__attribute__((packed)), so the #ifndef is not necessary.
(and, in fact, it's problematic, because we'll end up with the header
containing:
#ifndef __attribute__((packed))
#define __attribu...
and so forth.)
Hans de Goede [Thu, 18 Dec 2014 17:55:14 +0000 (09:55 -0800)]
Input: alps - v7: fix finger counting for > 2 fingers on clickpads
Protocol v7 uses the middle / right button bits on clickpads to communicate
"location" information of a 3th touch (and possible 4th) touch on
clickpads.
Specifically when 3 touches are down, if one of the 3 touches is in the
left / right button area, this will get reported in the middle / right
button bits and the touchpad will still send a TWO type packet rather then
a MULTI type packet, so when this happens we must add the finger reported
in the button area to the finger count.
Likewise we must also add fingers reported this way to the finger count
when we get MULTI packets.
Hans de Goede [Thu, 18 Dec 2014 17:53:34 +0000 (09:53 -0800)]
Input: alps - v7: sometimes a single touch is reported in mt[1]
The v7 proto differentiates between a primary touch (with high precision)
and a secondary touch (with lower precision). Normally when 2 fingers are
down and one is lifted the still present touch becomes the primary touch,
but some traces have shown that this does not happen always.
This commit deals with this by making alps_get_mt_count() not stop at the
first empty mt slot, and if a touch is present in mt[1] and not mt[0]
moving the data to mt[0] (for input_mt_assign_slots).
Hans de Goede [Thu, 18 Dec 2014 17:52:59 +0000 (09:52 -0800)]
Input: alps - v7: ignore new packets
NEW packets are send to indicate a discontinuity in the finger coordinate
reporting. Specifically a finger may have moved from slot 0 to 1 or vice
versa. INPUT_MT_TRACK takes care of this for us.
NEW packets have 3 problems:
1) They do not contain middle / right button info (on non clickpads)
this can be worked around by preserving the old button state
2) They do not contain an accurate fingercount, and they are
typically send when the number of fingers changes. We cannot use
the old finger count as that may mismatch with the amount of
touch coordinates we've available in the NEW packet
3) Their x data for the second touch is inaccurate leading to
a possible jump of the x coordinate by 16 units when the first
non NEW packet comes in
Since problems 2 & 3 cannot be worked around, just ignore them.
be2net: Fix incorrect setting of tunnel offload flag in netdev features
An earlier commit to resolve an issue with encapsulation offloads missed
setting a bit in the outer netdev features flag. This results in loss of TSO
feature on a VxLAN interface.
Fixes: 630f4b70 ("Export tunnel offloads only when a VxLAN tunnel is created") Signed-off-by: Sriharsha Basavapatna <[email protected]> Signed-off-by: David S. Miller <[email protected]>
David Vrabel [Thu, 18 Dec 2014 11:13:06 +0000 (11:13 +0000)]
xen-netback: support frontends without feature-rx-notify again
Commit bc96f648df1bbc2729abbb84513cf4f64273a1f1 (xen-netback: make
feature-rx-notify mandatory) incorrectly assumed that there were no
frontends in use that did not support this feature. But the frontend
driver in MiniOS does not and since this is used by (qemu) stubdoms,
these stopped working.
Netback sort of works as-is in this mode except:
- If there are no Rx requests and the internal Rx queue fills, only
the drain timeout will wake the thread. The default drain timeout
of 10 s would give unacceptable pauses.
- If an Rx stall was detected and the internal Rx queue is drained,
then the Rx thread would never wake.
Handle these two cases (when feature-rx-notify is disabled) by:
Andreas Larsson [Thu, 18 Dec 2014 12:23:23 +0000 (13:23 +0100)]
sparc32: destroy_context() and switch_mm() needs to disable interrupts.
Load balancing can be triggered in the critical sections protected by
srmmu_context_spinlock in destroy_context() and switch_mm() and can hang
the cpu waiting for the rq lock of another cpu that in turn has called
switch_mm hangning on srmmu_context_spinlock leading to deadlock.
So, disable interrupt while taking srmmu_context_spinlock in
destroy_context() and switch_mm() so we don't deadlock.
See also commit 77b838fa1ef0 ("[SPARC64]: destroy_context() needs to disable
interrupts.")
Merge branches 'pm-opp', 'pm-cpufreq' and 'pm-tools'
* pm-opp:
PM / OPP: do error handling at the bottom of dev_pm_opp_add_dynamic()
PM / OPP: handle allocation of device_opp in a separate routine
PM / OPP: reuse find_device_opp() instead of duplicating code
PM / OPP: Staticize __dev_pm_opp_remove()
PM / OPP: replace kfree with kfree_rcu while freeing 'struct device_opp'
* pm-cpufreq:
MAINTAINERS: add entry for intel_pstate
intel_pstate: Add a few comments
intel_pstate: add kernel parameter to force loading
* pm-tools:
Revert "tools: cpupower: fix return checks for sysfs_get_idlestate_count()"
Jesse Gross [Wed, 17 Dec 2014 02:25:32 +0000 (18:25 -0800)]
geneve: Fix races between socket add and release.
Currently, searching for a socket to add a reference to is not
synchronized with deletion of sockets. This can result in use
after free if there is another operation that is removing a
socket at the same time. Solving this requires both holding the
appropriate lock and checking the refcount to ensure that it
has not already hit zero.
Inspired by a related (but not exactly the same) issue in the
VXLAN driver.
Thomas Graf [Thu, 18 Dec 2014 10:30:26 +0000 (10:30 +0000)]
netlink: Don't reorder loads/stores before marking mmap netlink frame as available
Each mmap Netlink frame contains a status field which indicates
whether the frame is unused, reserved, contains data or needs to
be skipped. Both loads and stores may not be reordeded and must
complete before the status field is changed and another CPU might
pick up the frame for use. Use an smp_mb() to cover needs of both
types of callers to netlink_set_status(), callers which have been
reading data frame from the frame, and callers which have been
filling or releasing and thus writing to the frame.
- Example code path requiring a smp_rmb():
memcpy(skb->data, (void *)hdr + NL_MMAP_HDRLEN, hdr->nm_len);
netlink_set_status(hdr, NL_MMAP_STATUS_UNUSED);
- Example code path requiring a smp_wmb():
hdr->nm_uid = from_kuid(sk_user_ns(sk), NETLINK_CB(skb).creds.uid);
hdr->nm_gid = from_kgid(sk_user_ns(sk), NETLINK_CB(skb).creds.gid);
netlink_frame_flush_dcache(hdr);
netlink_set_status(hdr, NL_MMAP_STATUS_VALID);
Fixes: f9c228 ("netlink: implement memory mapped recvmsg()") Reported-by: Eric Dumazet <[email protected]> Signed-off-by: Thomas Graf <[email protected]> Signed-off-by: David S. Miller <[email protected]>
David Miller [Tue, 16 Dec 2014 22:58:17 +0000 (17:58 -0500)]
netlink: Always copy on mmap TX.
Checking the file f_count and the nlk->mapped count is not completely
sufficient to prevent the mmap'd area contents from changing from
under us during netlink mmap sendmsg() operations.
Be careful to sample the header's length field only once, because this
could change from under us as well.
Fixes: 5fd96123ee19 ("netlink: implement memory mapped sendmsg()") Signed-off-by: David S. Miller <[email protected]> Acked-by: Daniel Borkmann <[email protected]> Acked-by: Thomas Graf <[email protected]>
Jiri Jaburek [Thu, 18 Dec 2014 01:03:19 +0000 (02:03 +0100)]
ALSA: usb-audio: extend KEF X300A FU 10 tweak to Arcam rPAC
The Arcam rPAC seems to have the same problem - whenever anything
(alsamixer, udevd, 3.9+ kernel from 60af3d037eb8c, ..) attempts to
access mixer / control interface of the card, the firmware "locks up"
the entire device, resulting in
SNDRV_PCM_IOCTL_HW_PARAMS failed (-5): Input/output error
from alsa-lib.
Other operating systems can somehow read the mixer (there seems to be
playback volume/mute), but any manipulation is ignored by the device
(which has hardware volume controls).
Fabio Estevam [Tue, 16 Dec 2014 00:49:59 +0000 (22:49 -0200)]
watchdog: imx2_wdt: Fix the argument of watchdog_active()
Fix the following build warning by passing the expected argument type to
watchdog_active():
drivers/watchdog/imx2_wdt.c: In function 'imx2_wdt_suspend':
drivers/watchdog/imx2_wdt.c:340:2: warning: passing argument 1 of 'watchdog_active' from incompatible pointer type [enabled by default]
In file included from drivers/watchdog/imx2_wdt.c:38:0:
include/linux/watchdog.h:104:20: note: expected 'struct watchdog_device *' but argument is of type 'struct watchdog_device **'
Greg Kurz [Fri, 12 Dec 2014 11:37:40 +0000 (12:37 +0100)]
powerpc/powernv: Ignore smt-enabled on Power8 and later
Starting with POWER8, the subcore logic relies on all threads of a core
being booted so that they can participate in split mode switches. So on
those machines we ignore the smt_enabled_at_boot setting (smt-enabled on
the kernel command line).
Signed-off-by: Greg Kurz <[email protected]>
[mpe: Update comment and change log to be more precise] Signed-off-by: Michael Ellerman <[email protected]>
ACCESS_ONCE does not work reliably on non-scalar types. For
example gcc 4.6 and 4.7 might remove the volatile tag for such
accesses during the SRA (scalar replacement of aggregates) step
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145)
Commit 1365039d0cb3 ("KVM: s390: Fix ipte locking") replace
ACCESS_ONCE with barriers. Lets use READ_ONCE instead.
ACCESS_ONCE does not work reliably on non-scalar types. For
example gcc 4.6 and 4.7 might remove the volatile tag for such
accesses during the SRA (scalar replacement of aggregates) step
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145)
Change the spinlock code to replace ACCESS_ONCE with READ_ONCE.
ACCESS_ONCE does not work reliably on non-scalar types. For
example gcc 4.6 and 4.7 might remove the volatile tag for such
accesses during the SRA (scalar replacement of aggregates) step
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145)
Change the spinlock code to replace ACCESS_ONCE with READ_ONCE.
ACCESS_ONCE does not work reliably on non-scalar types. For
example gcc 4.6 and 4.7 might remove the volatile tag for such
accesses during the SRA (scalar replacement of aggregates) step
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145)
Change the gup code to replace ACCESS_ONCE with READ_ONCE.
ACCESS_ONCE does not work reliably on non-scalar types. For
example gcc 4.6 and 4.7 might remove the volatile tag for such
accesses during the SRA (scalar replacement of aggregates) step
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145)
Change the gup code to replace ACCESS_ONCE with READ_ONCE.
ACCESS_ONCE does not work reliably on non-scalar types. For
example gcc 4.6 and 4.7 might remove the volatile tag for such
accesses during the SRA (scalar replacement of aggregates) step
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145)
Change the spinlock code to replace ACCESS_ONCE with READ_ONCE.
mm: replace ACCESS_ONCE with READ_ONCE or barriers
ACCESS_ONCE does not work reliably on non-scalar types. For
example gcc 4.6 and 4.7 might remove the volatile tag for such
accesses during the SRA (scalar replacement of aggregates) step
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145)
Let's change the code to access the page table elements with
READ_ONCE that does implicit scalar accesses for the gup code.
mm_find_pmd is tricky, because m68k and sparc(32bit) define pmd_t
as array of longs. This code requires just that the pmd_present
and pmd_trans_huge check are done on the same value, so a barrier
is sufficent.
A similar case is in handle_pte_fault. On ppc44x the word size is
32 bit, but a pte is 64 bit. A barrier is ok as well.
ACCESS_ONCE does not work reliably on non-scalar types. For
example gcc 4.6 and 4.7 might remove the volatile tag for such
accesses during the SRA (scalar replacement of aggregates) step
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145)
Let's provide READ_ONCE/ASSIGN_ONCE that will do all accesses via
scalar types as suggested by Linus Torvalds. Accesses larger than
the machines word size cannot be guaranteed to be atomic. These
macros will use memcpy and emit a build warning.
Paolo Bonzini [Thu, 18 Dec 2014 08:39:55 +0000 (09:39 +0100)]
Merge tag 'signed-kvm-ppc-next' of git://github.com/agraf/linux-2.6 into HEAD
Patch queue for ppc - 2014-12-18
Highights this time around:
- Removal of HV support for 970. It became a maintenance burden and received
practically no testing. POWER8 with HV is available now, so just grab one
of those boxes if PR isn't enough for you.
- Some bug fixes and performance improvements
- Tracepoints for book3s_hv
powerpc/uaccess: Allow get_user() with bitwise types
At the moment, if p and x are both of the same bitwise type
(eg. __le32), get_user(x, p) produces a sparse warning.
This is because *p is loaded into a long then cast back to typeof(*p).
When typeof(*p) is a bitwise type (which is uncommon), such a cast needs
__force, otherwise sparse produces a warning.
For non-bitwise types __force should have no effect, and should not hide
any legitimate errors.
Note that we are casting to typeof(*p) not typeof(x). Even with the
cast, if x and *p are of different types we should get the warning, so I
think we are not loosing the ability to detect any actual errors.
virtio would like to use bitwise types with get_user() so fix these
spurious warnings by adding __force.
Signed-off-by: Michael S. Tsirkin <[email protected]>
[mpe: Fill in changelog with more details] Signed-off-by: Michael Ellerman <[email protected]>
Some products firmware is no longer being updated
e.g. dsp5000, dsp8700 but it should continue to work
with updated HPI versions.
Avoid regression by allowing this firmware to be loaded as
long as major version is the same.
Warn about mismatching versions, as matching versions are
preferred.
Ingo Molnar [Thu, 18 Dec 2014 06:23:55 +0000 (07:23 +0100)]
Merge tag 'perf-core-for-mingo-2' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/urgent
Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:
User visible changes:
- The mmap address range for the ring buffer now is calculated using the
contents of /proc/sys/kernel/perf_event_mlock_kb.
This fixes an -EPERM case where 'trace' was trying to use more than what
is configured on perf_event_mlock_kb. (Arnaldo Carvalho de Melo)
Infrastructure changes:
- Move bitops definitions so that they match the header file hierarchy
in the kernel sources where that code came from. (Arnaldo Carvalho de Melo)
- Adopt round{down,up}_pow_of_two from the kernel and use it instead of
equivalent code, so that we reuse more kernel code and make tools/ look
more like kernel source code, to encourage further contributions from
kernel hackers (Arnaldo Carvalho de Melo)
- Fix use after free in filename__read_build_id (Mitchell Krome)
Linus Torvalds [Thu, 18 Dec 2014 02:47:01 +0000 (18:47 -0800)]
Ceph: remove left-over reject file
Neither Sage nor I noticed that Zheng Yan had mistakenly committed
fs/ceph/super.h.rej as part of commit 31c542a199d7 ("ceph: add inline
data to pagecache").
Kees Cook [Fri, 12 Dec 2014 03:06:49 +0000 (13:36 +1030)]
param: do not set store func without write perm
When a module_param is defined without DAC write permissions, it can
still be changed at runtime and updated. Drivers using a 0444 permission
may be surprised that these values can still be changed.
For drivers that want to allow updates, any S_IW* flag will set the
"store" function as before. Drivers without S_IW* flags will have the
"store" function unset, unforcing a read-only value. Drivers that wish
neither "store" nor "get" can continue to use "0" for perms to stay out
of sysfs entirely.
Old behavior:
# cd /sys/module/snd/parameters
# ls -l
total 0
-r--r--r-- 1 root root 4096 Dec 11 13:55 cards_limit
-r--r--r-- 1 root root 4096 Dec 11 13:55 major
-r--r--r-- 1 root root 4096 Dec 11 13:55 slots
# cat major
116
# echo -1 > major
-bash: major: Permission denied
# chmod u+w major
# echo -1 > major
# cat major
-1
New behavior:
...
# chmod u+w major
# echo -1 > major
-bash: echo: write error: Input/output error
Dmitry Torokhov [Tue, 16 Dec 2014 23:09:39 +0000 (15:09 -0800)]
cpufreq-dt: defer probing if OPP table is not ready
cpufreq-dt driver supports mode when OPP table is provided by platform
code and not device tree. However on certain platforms code that fills
OPP table may run after cpufreq driver tries to initialize, so let's
report -EPROBE_DEFER if we do not find any entires in OPP table for the
CPU.
Dmitry Torokhov [Tue, 16 Dec 2014 23:09:38 +0000 (15:09 -0800)]
PM / OPP: take RCU lock in dev_pm_opp_get_opp_count
A lot of callers are missing the fact that dev_pm_opp_get_opp_count
needs to be called under RCU lock. Given that RCU locks can safely be
nested, instead of providing *_locked() API, let's take RCU lock inside
dev_pm_opp_get_opp_count() and leave callers as is.
Dmitry Torokhov [Tue, 16 Dec 2014 23:09:37 +0000 (15:09 -0800)]
PM / OPP: fix warning in of_free_opp_table()
Not having OPP defined for a device is not a crime, we should not splat
warning in this case. Also, it seems that we are ready to accept invalid
dev (find_device_opp will return ERR_PTR(-EINVAL) then) so let's not
crash in dev_name() in such case.
Linus Torvalds [Thu, 18 Dec 2014 00:03:12 +0000 (16:03 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client
Pull ceph updates from Sage Weil:
"The big item here is support for inline data for CephFS and for
message signatures from Zheng. There are also several bug fixes,
including interrupted flock request handling, 0-length xattrs, mksnap,
cached readdir results, and a message version compat field. Finally
there are several cleanups from Ilya, Dan, and Markus.
Note that there is another series coming soon that fixes some bugs in
the RBD 'lingering' requests, but it isn't quite ready yet"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client: (27 commits)
ceph: fix setting empty extended attribute
ceph: fix mksnap crash
ceph: do_sync is never initialized
libceph: fixup includes in pagelist.h
ceph: support inline data feature
ceph: flush inline version
ceph: convert inline data to normal data before data write
ceph: sync read inline data
ceph: fetch inline data when getting Fcr cap refs
ceph: use getattr request to fetch inline data
ceph: add inline data to pagecache
ceph: parse inline data in MClientReply and MClientCaps
libceph: specify position of extent operation
libceph: add CREATE osd operation support
libceph: add SETXATTR/CMPXATTR osd operations support
rbd: don't treat CEPH_OSD_OP_DELETE as extent op
ceph: remove unused stringification macros
libceph: require cephx message signature by default
ceph: introduce global empty snap context
ceph: message versioning fixes
...
Aniroop Mathur [Wed, 17 Dec 2014 23:33:06 +0000 (15:33 -0800)]
Input: evdev - add CLOCK_BOOTTIME support
This patch adds support for CLOCK_BOOTTIME for input event timestamp.
CLOCK_BOOTTIME includes suspend time, so it would allow aplications
to get correct time difference between two events even when system
resumes from suspend state.
haarp [Wed, 17 Dec 2014 23:22:08 +0000 (15:22 -0800)]
Input: psmouse - expose drift duration for IBM trackpoints
IBM Trackpoints have a feature to compensate for drift by recalibrating
themselves periodically. By default, if for 0.5 seconds there is no change
in position, it's used as the new zero. This duration is too low. Often,
the calibration happens when the trackpoint is in fact being used.
IBM's Trackpoint Engineering Specifications show a configuration register
that allows changing this duration, rstdft1.
Linus Torvalds [Wed, 17 Dec 2014 20:31:40 +0000 (12:31 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace
Pull user namespace related fixes from Eric Biederman:
"As these are bug fixes almost all of thes changes are marked for
backporting to stable.
The first change (implicitly adding MNT_NODEV on remount) addresses a
regression that was created when security issues with unprivileged
remount were closed. I go on to update the remount test to make it
easy to detect if this issue reoccurs.
Then there are a handful of mount and umount related fixes.
Then half of the changes deal with the a recently discovered design
bug in the permission checks of gid_map. Unix since the beginning has
allowed setting group permissions on files to less than the user and
other permissions (aka ---rwx---rwx). As the unix permission checks
stop as soon as a group matches, and setgroups allows setting groups
that can not later be dropped, results in a situtation where it is
possible to legitimately use a group to assign fewer privileges to a
process. Which means dropping a group can increase a processes
privileges.
The fix I have adopted is that gid_map is now no longer writable
without privilege unless the new file /proc/self/setgroups has been
set to permanently disable setgroups.
The bulk of user namespace using applications even the applications
using applications using user namespaces without privilege remain
unaffected by this change. Unfortunately this ix breaks a couple user
space applications, that were relying on the problematic behavior (one
of which was tools/selftests/mount/unprivileged-remount-test.c).
To hopefully prevent needing a regression fix on top of my security
fix I rounded folks who work with the container implementations mostly
like to be affected and encouraged them to test the changes.
> So far nothing broke on my libvirt-lxc test bed. :-)
> Tested with openSUSE 13.2 and libvirt 1.2.9.
> Tested-by: Richard Weinberger <[email protected]>
> Tested on Fedora20 with libvirt 1.2.11, works fine.
> Tested-by: Chen Hanxiao <[email protected]>
> Ok, thanks - yes, unprivileged lxc is working fine with your kernels.
> Just to be sure I was testing the right thing I also tested using
> my unprivileged nsexec testcases, and they failed on setgroup/setgid
> as now expected, and succeeded there without your patches.
> Tested-by: Serge Hallyn <[email protected]>
> I tested this with Sandstorm. It breaks as is and it works if I add
> the setgroups thing.
> Tested-by: Andy Lutomirski <[email protected]> # breaks things as designed :("
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace:
userns: Unbreak the unprivileged remount tests
userns; Correct the comment in map_write
userns: Allow setting gid_maps without privilege when setgroups is disabled
userns: Add a knob to disable setgroups on a per user namespace basis
userns: Rename id_map_mutex to userns_state_mutex
userns: Only allow the creator of the userns unprivileged mappings
userns: Check euid no fsuid when establishing an unprivileged uid mapping
userns: Don't allow unprivileged creation of gid mappings
userns: Don't allow setgroups until a gid mapping has been setablished
userns: Document what the invariant required for safe unprivileged mappings.
groups: Consolidate the setgroups permission checks
mnt: Clear mnt_expire during pivot_root
mnt: Carefully set CL_UNPRIVILEGED in clone_mnt
mnt: Move the clear of MNT_LOCKED from copy_tree to it's callers.
umount: Do not allow unmounting rootfs.
umount: Disallow unprivileged mount force
mnt: Update unprivileged remount test
mnt: Implicitly add MNT_NODEV on remount when it was implicitly added by mount
Marcel Holtmann [Wed, 17 Dec 2014 17:18:08 +0000 (18:18 +0100)]
Bluetooth: Fix bug with filter in service discovery optimization
The optimization for filtering out extended inquiry results, advertising
reports or scan response data based on provided UUID list has a logic
bug. In case no match is found in the advertising data, the scan
response is ignored and not checked against the filter. This will lead
to events being filtered wrongly.
Change the code to actually only drop the events when the scan response
data is not present. If it is present, it needs to be checked against
the provided filter.
The patch is a bit more complex than it needs to be. That is because
it also fixes this compiler warning that some gcc versions produce.
CC net/bluetooth/mgmt.o
net/bluetooth/mgmt.c: In function ‘mgmt_device_found’:
net/bluetooth/mgmt.c:7028:7: warning: ‘match’ may be used uninitialized in this function [-Wmaybe-uninitialized]
bool match;
^
It seems that gcc can not clearly figure out the context of the match
variable. So just change the branches for the extended inquiry response
and advertising data around so that it is clear.
Dave Hansen reports that commit fb7332a9fedf ("mmu_gather: move minimal
range calculations into generic code") caused a performance problem:
"tlb_finish_mmu() goes up about 9x in the profiles (~0.4%->3.6%) and
tlb_flush_mmu_free() takes about 3.1% of CPU time with the patch
applied, but does not show up at all on the commit before"
and the reason is that Will moved the test for whether we need to flush
from tlb_flush_mmu() into tlb_flush_mmu_tlbonly(). But that meant that
tlb_flush_mmu_free() basically lost that check.
Move it back into tlb_flush_mmu() where it belongs, so that it covers
both tlb_flush_mmu_tlbonly() _and_ tlb_flush_mmu_free().
Linus Torvalds [Wed, 17 Dec 2014 19:52:37 +0000 (11:52 -0800)]
x86: mm: fix VM_FAULT_RETRY handling
My commit 26178ec11ef3 ("x86: mm: consolidate VM_FAULT_RETRY handling")
had a really stupid typo: the FAULT_FLAG_USER bit is in the 'flags'
variable, not the 'fault' variable. Duh,
The one silver lining in this is that Dave finding this at least
confirms that trinity actually triggers this special path easily, in a
way normal use does not.
Linus Torvalds [Wed, 17 Dec 2014 18:44:22 +0000 (10:44 -0800)]
Merge tag 'vfio-v3.19-rc1' of git://github.com/awilliam/linux-vfio
Pull VFIO updates from Alex Williamson:
- s390 support (Frank Blaschka)
- Enable iommu-type1 for ARM SMMU (Will Deacon)
* tag 'vfio-v3.19-rc1' of git://github.com/awilliam/linux-vfio:
drivers/vfio: allow type-1 IOMMU instantiation on top of an ARM SMMU
vfio: make vfio run on s390
Linus Torvalds [Wed, 17 Dec 2014 18:37:56 +0000 (10:37 -0800)]
Merge tag 'virtio-next-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux
Pull virtio updates from Rusty Russell:
"A balloon enhancement, and a minor race-on-module-unload theoretical
bug which doesn't merit cc: stable.
All the exciting stuff went via MST this cycle"
* tag 'virtio-next-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux:
virtio_balloon: free some memory from balloon on OOM
virtio_balloon: return the amount of freed memory from leak_balloon()
virtio_blk: fix race at module removal
virtio: Fix comment typo 'CONFIG_S_FAILED'
i2c: sh_mobile: I2C_SH_MOBILE should depend on HAS_DMA
If NO_DMA=y:
drivers/built-in.o: In function `sh_mobile_i2c_dma_unmap':
i2c-sh_mobile.c:(.text+0x60de42): undefined reference to `dma_unmap_single'
drivers/built-in.o: In function `sh_mobile_i2c_xfer_dma':
i2c-sh_mobile.c:(.text+0x60df22): undefined reference to `dma_map_single'
i2c-sh_mobile.c:(.text+0x60df2e): undefined reference to `dma_mapping_error'
Wolfram Sang [Tue, 16 Dec 2014 12:31:26 +0000 (13:31 +0100)]
i2c: sh_mobile: rework deferred probing
DMA is opt-in for this driver. So, we can't use deferred probing for
requesting DMA channels in probe, because our driver would get endlessly
deferred if DMA support is compiled in AND the DMA driver is missing.
Because we can't know when the DMA driver might show up, we always try
again when a DMA transfer would be possible. The downside is that there
is more overhead for setting up PIO transfers under the above scenario.
But well, having DMA enabled and the proper DMA driver missing looks
like a broken or test config anyhow.
Thomas Petazzoni [Thu, 11 Dec 2014 16:33:46 +0000 (17:33 +0100)]
i2c: mv64xxx: rework offload support to fix several problems
Originally, the I2C controller supported by the i2c-mv64xxx driver
requires a lot of software support: an interrupt is generated at each
step of an I2C transaction (after the start bit, after sending the
address, etc.) and the driver is in charge of re-programming the I2C
controller to do the next step of the I2C transaction. This explains
the fairly complex state machine that the driver has.
On Marvell Armada XP and later processors (Armada 375, 38x, etc.), the
I2C controller was extended with a part called the "I2C Bridge", which
allows to offload the I2C transaction completely to the
hardware. Initial support for this mechanism was added in commit 930ab3d403a ("i2c: mv64xxx: Add I2C Transaction Generator support").
However, the implementation done in this commit has two related
issues, which this commit fixes by completely changing how the offload
implementation is done:
* SMBus read transfers, where there is one write to select the
register immediately followed in the same transaction by one read,
were making the processor hang. This was easier visible on the
Marvell Armada XP WRT1900AC platform using a driver for an I2C LED
controller, or on other Armada XP platforms by using a simple
'i2cget' command to read an I2C EEPROM.
* The implementation was based on the fact that the offload engine
was re-programmed to transfer each message of an I2C xfer: this
meant that each message sent with the offload engine was starting
with a normal I2C start sequence. However, the I2C subsystem
assumes that all messages belonging to the same xfer will use the
so-called "repeated start" so that the entire I2C xfer is seen as
one transfer by the I2C devices and cannot be interrupt by other
I2C masters on the same bus.
In fact, the "I2C Bridge" allows to offload three types of xfer:
- xfer of one write message
- xfer of one read message
- xfer of one write message followed by one read message
For all other situations, we have to fallback to not using the "I2C
Bridge" in order to get proper I2C semantics.
Therefore, this commit reworks the offload implementation to put it
not at the message level, but at the xfer level: in the
mv64xxx_i2c_xfer() function, we decide if the transaction can be
offloaded (in which case it is handled by the
mv64xxx_i2c_offload_xfer() function), or otherwise it is handled by
the slow path (implemented in the existing mv64xxx_i2c_execute_msg()).
This allows to simplify the state machine, which no longer needs to
have any state related to the offload implementation: the offload
implementation is now completely separated from the slow path (with
the exception of the interrupt handler, of course).
In summary:
- mv64xxx_i2c_can_offload() will analyze an I2C xfer and decided of
the "I2C Bridge" can be used to offload it or not.
- mv64xxx_i2c_offload_xfer() will actually program the "I2C Bridge"
to offload one xfer (of either one or two messages), and block
using mv64xxx_i2c_wait_for_completion() until the xfer completes.
- The interrupt handler mv64xxx_i2c_intr() is modified to push the
offload related code to a separate function,
mv64xxx_i2c_intr_offload(). It will take care of reading the
received data if needed.
This commit was tested on:
- Armada XP OpenBlocks AX3-4 (EEPROM on I2C and RTC on I2C)
- Armada XP WRT1900AC (LED controller on I2C)
- Armada XP GP (EEPROM on I2C)
Linus Torvalds [Wed, 17 Dec 2014 18:16:27 +0000 (10:16 -0800)]
Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux
Pull thermal management update from Zhang Rui:
"Summary:
- of-thermal extension to allow drivers to register and use its
functionality in a better way, without exploiting thermal core.
From Lukasz Majewski.
- Fix a bug in intel_soc_dts_thermal driver which calls a sleep
function in interrupt handler. From Maurice Petallo.
- add a thermal UAPI header file for exporting the thermal generic
netlink information to user-space. From Florian Fainelli.
- First round of refactoring in Exynos driver. Bartlomiej and Lukasz
are attempting to make it lean and easier to understand.
- New thermal driver for Rockchip (rk3288), with support for DT
thermal. From Caesar Wang.
- New thermal driver for Nvidia, Tegra124 SOCTHERM driver, with
support for DT thermal. From Mikko Perttunen.
- New cooling device, based on common clock framework. From Eduardo
Valentin.
- a couple of small fixes in thermal core framework. From Srinivas
Pandruvada, Javi Merino, Luis Henriques.
- Dropping Armada A375-Z1 SoC thermal support as the chip is not in
the market, armada folks decided to drop its support.
- a couple of small fixes and cleanups in int340x thermal driver"
* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux: (58 commits)
thermal: provide an UAPI header file
Thermal/int340x: Clear the error value of the last acpi_bus_get_device() call
thermal/powerclamp: add id for braswell cpu
thermal: Intel SoC DTS: Don't do thermal zone update inside spin_lock
Thermal: fix platform_no_drv_owner.cocci warnings
Thermal/int340x: avoid unnecessary pointer casting
thermal: int3403: Delete a check before thermal_zone_device_unregister()
thermal/int3400: export uuids
thermal: of: Extend current of-thermal.c code to allow setting emulated temp
thermal: of: Extend of-thermal to export table of trip points
thermal: of: Rename struct __thermal_trip to struct thermal_trip
thermal: of: Extend of-thermal.c to provide check if trip point is valid
thermal: of: Extend of-thermal.c to provide number of trip points
thermal: Fix error path in thermal_init()
thermal: lock the thermal zone when switching governors
thermal: core: ignore invalid trip temperature
thermal: armada: Remove support for A375-Z1 SoC
thermal: rockchip: add driver for thermal
dt-bindings: document Rockchip thermal
thermal: exynos: remove exynos_tmu_data.h include
...
Linus Torvalds [Wed, 17 Dec 2014 18:10:51 +0000 (10:10 -0800)]
Merge tag 'pwm/for-3.19-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/thierry.reding/linux-pwm
Pull pwm updates from Thierry Reding:
"There are two new drivers, one for the BCM2835 (Raspberry Pi) and one
used in conjunction with the LCD controller on various Atmel SoCs.
The Samsung PWM driver can now be built for 64-bit ARM (Exynos7).
A couple of fixes have been applied to the FTM PWM driver and system
sleep support was added"
* tag 'pwm/for-3.19-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/thierry.reding/linux-pwm:
pwm: atmel-hlcdc: add at91sam9x5 and sama5d3 errata handling
pwm: ftm: Add Power Management support for FTM PWM
pwm: ftm: Add regmap rbtree type cache support
pwm: ftm: Correctly track usage count
pwm: samsung: Allow Samsung PWM driver to be enabled on Exynos7
pwm: add DT bindings documentation for atmel-hlcdc-pwm driver
pwm: add support for atmel-hlcdc-pwm device
pwm: Add BCM2835 PWM driver
Linus Torvalds [Wed, 17 Dec 2014 18:06:02 +0000 (10:06 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input
Pull input subsystem updates from Dmitry Torokhov:
"Two new drivers for Elan hardware (for I2C touchpad and touchscreen
found in several Chromebooks and other devices), a driver for Goodix
touch panel, and small fixes to Cypress I2C trackpad and other input
drivers.
Also we switched to use __maybe_unused instead of gating suspend/
resume code with #ifdef guards to get better compile coverage"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input: (27 commits)
Input: gpio_keys - fix warning regarding uninitialized 'button' variable
Input: add support for Elan eKTH I2C touchscreens
Input: gpio_keys - fix warning regarding uninitialized 'irq' variable
Input: cyapa - use 'error' for error codes
Input: cyapa - fix resuming the device
Input: gpio_keys - add device tree support for interrupt only keys
Input: amikbd - allocate temporary keymap buffer on the stack
Input: amikbd - fix build if !CONFIG_HW_CONSOLE
Input: lm8323 - missing error check in lm8323_set_disable()
Input: initialize device counter variables with -1
Input: initialize input_no to -1 to avoid subtraction
Input: i8042 - do not try to load on Intel NUC D54250WYK
Input: atkbd - correct MSC_SCAN events for force_release keys
Input: cyapa - switch to using managed resources
Input: lifebook - use "static inline" instead of "inline" in lifebook.h
Input: touchscreen - use __maybe_unused instead of ifdef around suspend/resume
Input: mouse - use __maybe_unused instead of ifdef around suspend/resume
Input: misc - use __maybe_unused instead of ifdef around suspend/resume
Input: cap11xx - support for irq-active-high option
Input: cap11xx - add support for various cap11xx devices
...
Linus Torvalds [Wed, 17 Dec 2014 17:59:26 +0000 (09:59 -0800)]
Merge tag 'for-linus-20141215' of git://git.infradead.org/linux-mtd
Pull MTD updates from Brian Norris:
"Summary:
- Add device tree support for DoC3
- SPI NOR:
Refactoring, for better layering between spi-nor.c and its
driver users (e.g., m25p80.c)
New flash device support
Support 6-byte ID strings
- NAND:
New NAND driver for Allwinner SoC's (sunxi)
GPMI NAND: add support for raw (no ECC) access, for testing
purposes
Add ATO manufacturer ID
A few odd driver fixes
- MTD tests:
Allow testers to compensate for OOB bitflips in oobtest
Fix a torturetest regression
- nandsim: Support longer ID byte strings
And more"
* tag 'for-linus-20141215' of git://git.infradead.org/linux-mtd: (63 commits)
mtd: tests: abort torturetest on erase errors
mtd: physmap_of: fix potential NULL dereference
mtd: spi-nor: allow NULL as chip name and try to auto detect it
mtd: nand: gpmi: add raw oob access functions
mtd: nand: gpmi: add proper raw access support
mtd: nand: gpmi: add gpmi_copy_bits function
mtd: spi-nor: factor out write_enable() for erase commands
mtd: spi-nor: add support for s25fl128s
mtd: spi-nor: remove the jedec_id/ext_id
mtd: spi-nor: add id/id_len for flash_info{}
mtd: nand: correct the comment of function nand_block_isreserved()
jffs2: Drop bogus if in comment
mtd: atmel_nand: replace memcpy32_toio/memcpy32_fromio with memcpy
mtd: cafe_nand: drop duplicate .write_page implementation
mtd: m25p80: Add support for serial flash Spansion S25FL132K
MTD: m25p80: fix inconsistency in m25p_ids compared to spi_nor_ids
mtd: spi-nor: improve wait-till-ready timeout loop
mtd: delete unnecessary checks before two function calls
mtd: nand: omap: Fix NAND enumeration on 3430 LDP
mtd: nand: add ATO manufacturer info
...
Linus Torvalds [Wed, 17 Dec 2014 17:52:49 +0000 (09:52 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security
Pull security subsystem fixes from James Morris.
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
KEYS: remove a bogus NULL check
ima: Fix build failure on powerpc when TCG_IBMVTPM dependencies are not met
KEYS: Fix stale key registration at error path
Linus Torvalds [Wed, 17 Dec 2014 17:41:32 +0000 (09:41 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse
Pull fuse update from Miklos Szeredi:
"The first part makes sure we don't hold up umount with pending async
requests. In addition to being a cleanup, this is a small behavioral
change (for the better) and unlikely to break anything.
The second part prepares for a cleanup of the fuse device I/O code by
adding a helper for simple request submission, with some savings in
line numbers already realized"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:
fuse: use file_inode() in fuse_file_fallocate()
fuse: introduce fuse_simple_request() helper
fuse: reduce max out args
fuse: hold inode instead of path after release
fuse: flush requests on umount
fuse: don't wake up reserved req in fuse_conn_kill()
zhendong chen [Wed, 17 Dec 2014 06:37:04 +0000 (14:37 +0800)]
dm: fix missed error code if .end_io isn't implemented by target_type
In bio-based DM's clone_endio(), when target_type doesn't implement
.end_io (e.g. linear) r will be always be initialized 0. So if a
WRITE SAME bio fails WRITE SAME will not be disabled as intended.
Fix this by initializing r to error, rather than 0, in clone_endio().
Yan, Zheng [Fri, 14 Nov 2014 14:39:13 +0000 (22:39 +0800)]
ceph: flush inline version
After converting inline data to normal data, client need to flush
the new i_inline_version (CEPH_INLINE_NONE) to MDS. This commit makes
cap messages (sent to MDS) contain inline_version and inline_data.
Client always converts inline data to normal data before data write,
so the inline data length part is always zero.
Yan, Zheng [Fri, 14 Nov 2014 14:38:29 +0000 (22:38 +0800)]
ceph: convert inline data to normal data before data write
Before any data write, convert inline data to normal data and set
i_inline_version to CEPH_INLINE_NONE. The OSD request that saves
inline data to object contains 3 operations (CMPXATTR, WRITE and
SETXATTR). It compares a xattr named 'inline_version' to prevent
old data overwrites newer data.
Yan, Zheng [Fri, 14 Nov 2014 14:36:18 +0000 (22:36 +0800)]
ceph: sync read inline data
we can't use getattr to fetch inline data while holding Fr cap,
because it can cause deadlock. If we need to sync read inline data,
drop cap refs first, then use getattr to fetch inline data.
Yan, Zheng [Fri, 14 Nov 2014 14:10:07 +0000 (22:10 +0800)]
ceph: fetch inline data when getting Fcr cap refs
we can't use getattr to fetch inline data after getting Fcr caps,
because it can cause deadlock. The solution is try bringing inline
data to page cache when not holding any cap, and hope the inline
data page is still there after getting the Fcr caps. If the page
is still there, pin it in page cache for later IO.
Yan, Zheng [Thu, 13 Nov 2014 06:40:37 +0000 (14:40 +0800)]
libceph: specify position of extent operation
allow specifying position of extent operation in multi-operations
osd request. This is required for cephfs to convert inline data to
normal data (compare xattr, then write object).
Ilya Dryomov [Fri, 21 Nov 2014 19:16:43 +0000 (22:16 +0300)]
rbd: don't treat CEPH_OSD_OP_DELETE as extent op
CEPH_OSD_OP_DELETE is not an extent op, stop treating it as such. This
sneaked in with discard patches - it's one of the three osd ops (the
other two are CEPH_OSD_OP_TRUNCATE and CEPH_OSD_OP_ZERO) that discard
is implemented with.
Yan, Zheng [Thu, 6 Nov 2014 07:09:41 +0000 (15:09 +0800)]
ceph: introduce global empty snap context
Current snaphost code does not properly handle moving inode from one
empty snap realm to another empty snap realm. After changing inode's
snap realm, some dirty pages' snap context can be not equal to inode's
i_head_snap. This can trigger BUG() in ceph_put_wrbuffer_cap_refs()
The fix is introduce a global empty snap context for all empty snap
realm. This avoids triggering the BUG() for filesystem with no snapshot.
ceph, rbd: delete unnecessary checks before two function calls
The functions ceph_put_snap_context() and iput() test whether their
argument is NULL and then return immediately. Thus the test around the
call is not needed.
This issue was detected by using the Coccinelle software.
Yan, Zheng [Wed, 22 Oct 2014 01:09:56 +0000 (18:09 -0700)]
ceph: introduce a new inode flag indicating if cached dentries are ordered
After creating/deleting/renaming file, offsets of sibling dentries may
change. So we can not use cached dentries to satisfy readdir. But we can
still use the cached dentries to conclude -ENOENT for lookup.
This patch introduces a new inode flag indicating if child dentries are
ordered. The flag is set at the same time marking a directory complete.
After creating/deleting/renaming file, we clear the flag on directory
inode. This prevents ceph_readdir() from using cached dentries to satisfy
readdir syscall.
Ilya Dryomov [Thu, 23 Oct 2014 12:32:57 +0000 (16:32 +0400)]
libceph: nuke ceph_kvfree()
Use kvfree() from linux/mm.h instead, which is identical. Also fix the
ceph_buffer comment: we will allocate with kmalloc() up to 32k - the
value of PAGE_ALLOC_COSTLY_ORDER, but that really is just an
implementation detail so don't mention it at all.
projects / linux.git / log