In ad198x_power_eapd(), wrong pin NIDs are used for controlling EAPD for
HP and Front outputs of AD1988/AD1989. These are actually same with the
ones for AD1984 & co, port-A is 0x11 and port-D 0x12.
Reported-by: Raymond Yau <superquad.vortex2@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This function attempts to free one fragment beyond the number of
fragments that were actually mapped. This patch brings back the limit
to the correct spot.
Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Tested-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Because there is a possibility that skb is kfree_skb()ed and zero cleared
after ndo_start_xmit, we should not see the contents of skb like skb->len and
skb->dev->name after ndo_start_xmit. But trace_net_dev_xmit does that
and causes panic by NULL pointer dereference.
This patch fixes trace_net_dev_xmit not to see the contents of skb directly.
If you want to reproduce this panic,
1. Get tracepoint of net_dev_xmit on
2. Create 2 guests on KVM
2. Make 2 guests use virtio_net
4. Execute netperf from one to another for a long time as a network burden
5. host will panic(It takes about 30 minutes)
Signed-off-by: Koki Sanagi <sanagi.koki@jp.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
One of the error paths in
sound/usb/6fire/firmware.c::usb6fire_fw_ezusb_upload() neglects to free
the memory allocated for the firmware before returning, thus leaking the
memory.
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
By ignoring the unset values of the minconfig in deciding
what to test in the config_bisect can cause the problem
config from being tested too.
Just do not test the configs that are set in the minconfig.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The command that is called that reboots the kernel may fail
but the return code is not passed back to the ktest.pl script.
This is because a ';' is used between the two commands and
if the second command fails, only the first command's return
code is returned. Using a '&&' between the two commands fixes
this.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Because in perl the array size returned by $#arr, is the last
index and not the actually size of the array, we end the config
bisect early, thinking there is only one config left when there
are in fact two. Thus the result has a 50% chance of picking
the correct config that caused the problem.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
list_entry() and hlist_entry() are both simply aliases for
container_of(), but since io_context.cic_list.first is an hlist_node one
should at least use the correct alias.
Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
queue_fail can only be reached if cic is NULL, so its check for cic must
be bogus.
Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
In certain circumstances, we can get an oops from a torn down device.
Most notably this is from CD roms trying to call scsi_ioctl. The root
cause of the problem is the fact that after scsi_remove_device() has
been called, the queue is fully torn down. This is actually wrong
since the queue can be used until the sdev release function is called.
Therefore, we add an extra reference to the queue which is released in
sdev->release, so the queue always exists.
Reported-by: Parag Warudkar <parag.lkml@gmail.com>
Cc: stable@kernel.org
Signed-off-by: James Bottomley <jbottomley@parallels.com>
The failed_get label is used after the call to clk_get has succeeded, so it
should be moved up above the call to clk_put.
The failed_req labels doesn't do anything different than failed_get, so
delete it.
A simplified version of the semantic match that finds this problem is as
follows: (http://coccinelle.lip6.fr/)
// <smpl>
@r exists@
expression e1,e2;
statement S;
@@
e1 = clk_get@p1(...);
... when != e1 = e2
when != clk_put(e1)
when any
if (...) { ... when != clk_put(e1)
when != if (...) { ... clk_put(e1) ... }
* return@p3 ...;
} else S
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
These definitions are needed to let the runtime PM subsystem turn off
DMAC clocks, when it is suspended by the driver.
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
A recent patch has introduced a regression, where repeating a memcpy
DMA test with shdma module unloading between them skips the DMA channel
configuration. Fix this regression by always configuring the channel
during its allocation.
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Currently the DM9000 driver requests the primary interrupt before it
resets the chip and puts it into a known good state. This means that if
the chip is asserting interrupt for some reason we can end up with a
screaming IRQ that the interrupt handler is unable to deal with. Avoid
this by only requesting the interrupt after we've reset the chip so we
know what state it's in.
This started manifesting itself on one of my boards in the past month or
so, I suspect as a result of some core infrastructure changes removing
some form of mitigation against bad behaviour here, even when things boot
it seems that the new code brings the interface up more quickly.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Go to existing error handling code at the end of the function that calls
clk_put.
A simplified version of the semantic match that finds this problem is as
follows: (http://coccinelle.lip6.fr/)
// <smpl>
@r exists@
expression e1,e2;
statement S;
@@
e1 = clk_get@p1(...);
... when != e1 = e2
when != clk_put(e1)
when any
if (...) { ... when != clk_put(e1)
when != if (...) { ... clk_put(e1) ... }
* return@p3 ...;
} else S
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk>
Acked-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, user-space cannot determine if a 0 tcp_vlan_tci
means there is no VLAN tag or the VLAN ID was zero.
Add flag to make this explicit. User-space can check for
TP_STATUS_VLAN_VALID || tp_vlan_tci > 0, which will be backwards
compatible. Older could would have just checked for tp_vlan_tci,
so it will work no worse than before.
Signed-off-by: Ben Greear <greearb@candelatech.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Take the RTNL lock unconditionally when calling dev_close.
Taking the lock conditionally may cause race conditions.
Signed-off-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This avoids messages like this after suspend:
cdc_ncm 2-1.4:1.6: no reset_resume for driver cdc_ncm?
cdc_ncm 2-1.4:1.7: no reset_resume for driver cdc_ncm?
cdc_ncm 2-1.4:1.6: usb0: unregister 'cdc_ncm' usb-0000:00:1d.0-1.4, CDC NCM
This is important for the Ericsson F5521gw GSM/UMTS modem.
Otherwise modemmanager looses the fact that the cdc_ncm and cdc_acm devices
belong together.
The cdc_ether module does the same.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
commit 4af429d29b (vlan: lockless
transmit path) have a typo in vlan_dev_hard_start_xmit(), using
u64_stats_update_begin() to end the stat update, it should be
u64_stats_update_end().
Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Reviewed-by: WANG Cong <xiyou.wangcong@gmail.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Check against mistakenly passing in IPv6 addresses (which would result
in an INADDR_ANY bind) or similar incompatible sockaddrs.
Signed-off-by: Marcus Meissner <meissner@suse.de>
Cc: Reinhard Max <max@suse.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Further relax temperature range checks after reading the IA32_TEMPERATURE_TARGET
register. If the register returns a value other than 0 in bits 16..32, assume
that the returned value is correct.
This change applies to both packet and core temperature limits.
Cc: Carsten Emde <C.Emde@osadl.org>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
Acked-by: Fenghua Yu <fenghua.yu@intel.com>
Commit a321cedb12 excludes CPU models 0xe, 0xf,
0x16, and 0x1a from TjMax temperature adjustment, even though several of those
CPUs are known to have TiMax other than 100 degrees C, and even though the code
in adjust_tjmax() explicitly handles those CPUs and points to a Web document
listing several of the affected CPU IDs.
Reinstate original TjMax adjustment if TjMax can not be determined using the
IA32_TEMPERATURE_TARGET register.
https://bugzilla.kernel.org/show_bug.cgi?id=32582
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
Cc: Huaxu Wan <huaxu.wan@linux.intel.com>
Cc: Carsten Emde <C.Emde@osadl.org>
Cc: Valdis Kletnieks <valdis.kletnieks@vt.edu>
Cc: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Cc: Yong Wang <yong.y.wang@linux.intel.com>
Cc: Rudolf Marek <r.marek@assembler.cz>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Tested-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Fenghua Yu <fenghua.yu@intel.com>
Cc: <stable@kernel.org> # .35.x .36.x .37.x .38.x .39.x
This reverts commit a197b59ae6.
As rmk says:
"Commit a197b59ae6 (mm: fail GFP_DMA allocations when ZONE_DMA is not
configured) is causing regressions on ARM with various drivers which
use GFP_DMA.
The behaviour up until now has been to silently ignore that flag when
CONFIG_ZONE_DMA is not enabled, and to allocate from the normal zone.
However, as a result of the above commit, such allocations now fail
which causes drivers to fail. These are regressions compared to the
previous kernel version."
so just revert it.
Requested-by: Russell King <linux@arm.linux.org.uk>
Acked-by: Andrew Morton <akpm@linux-foundation.org>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* git://git.infradead.org/iommu-2.6:
intel-iommu: Fix off-by-one in RMRR setup
intel-iommu: Add domain check in domain_remove_one_dev_info
intel-iommu: Remove Host Bridge devices from identity mapping
intel-iommu: Use coherent DMA mask when requested
intel-iommu: Dont cache iova above 32bit
intel-iommu: Speed up processing of the identity_mapping function
intel-iommu: Check for identity mapping candidate using system dma mask
intel-iommu: Only unlink device domains from iommu
intel-iommu: Enable super page (2MiB, 1GiB, etc.) support
intel-iommu: Flush unmaps at domain_exit
intel-iommu: Remove obsolete comment from detect_intel_iommu
intel-iommu: fix VT-d PMR disable for TXT on S3 resume
Jens' back-merge commit 698567f3fa ("Merge commit 'v2.6.39' into
for-2.6.40/core") was incorrectly done, and re-introduced the
DISK_EVENT_MEDIA_CHANGE lines that had been removed earlier in commits
- 9fd097b149 ("block: unexport DISK_EVENT_MEDIA_CHANGE for
legacy/fringe drivers")
- 7eec77a181 ("ide: unexport DISK_EVENT_MEDIA_CHANGE for ide-gd
and ide-cd")
because of conflicts with the "g->flags" updates near-by by commit
d4dc210f69 ("block: don't block events on excl write for non-optical
devices")
As a result, we re-introduced the hanging behavior due to infinite disk
media change reports.
Tssk, tssk, people! Don't do back-merges at all, and *definitely* don't
do them to hide merge conflicts from me - especially as I'm likely
better at merging them than you are, since I do so many merges.
Reported-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Jens Axboe <jaxboe@fusionio.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In some cases we can read wrong temperature value. If after that
temperature value will not be updated to good one, we badly configure
tx power parameters and device is unable to send a data.
Resolves:
https://bugzilla.kernel.org/show_bug.cgi?id=35932
Cc: stable@kernel.org # 2.6.39+
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
read_lock() ... read_unlock_bh() is clearly bogus.
This was broken by
commit 23691d75cd
Author: Gustavo F. Padovan <padovan@profusion.mobi>
Date: Wed Apr 27 18:26:32 2011 -0300
Bluetooth: Remove l2cap_sk_list
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This is the same fix as
commit 841051602e
Author: Matteo Croce <technoboy85@gmail.com>
Date: Fri Dec 3 02:25:08 2010 +0100
The ath9k driver subtracts 3 dBm to the txpower as with two radios the
signal power is doubled.
The resulting value is assigned in an u16 which overflows and makes
the card work at full power.
in two more places. I grepped the ath tree and didn't find any others.
Cc: stable@kernel.org
Signed-off-by: Daniel Halperin <dhalperi@cs.washington.edu>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Commit 0a35d36 ("cfg80211: Use capability info to detect mesh beacons")
assumed that probe response with both ESS and IBSS bits cleared
means that the frame was sent by a mesh sta.
However, these capabilities are also being used in the p2p_find phase,
and the mesh-validation broke it.
Rename the WLAN_CAPABILITY_IS_MBSS macro, and verify that mesh ies
exist before assuming this frame was sent by a mesh sta.
Signed-off-by: Eliad Peller <eliad@wizery.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The current temperature range check of MSR_IA32_TEMPERATURE_TARGET
seems too strict to me, some TjMax values documented in
Documentation/hwmon/coretemp wouldn't pass. Relax the check so that
all the documented values pass.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: Carsten Emde <C.Emde@osadl.org>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
The temp_fault sysfs attribute is wrong, it should be temp2_fault instead.
Reported-by: Guenter Roeck <guenter.roeck@ericsson.com>
Signed-off-by: Per Dalen <per.dalen@appeartv.com>
Acked-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
blkbk->pending_pages can be NULL here so I added a check for it.
Signed-off-by: Dan Carpenter <error27@gmail.com>
[v1: Redid the loop a bit]
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
We were mapping an extra byte (and hence usually an extra page):
iommu_prepare_identity_map() expects to be given an 'end' argument which
is the last byte to be mapped; not the first byte *not* to be mapped.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The comment in domain_remove_one_dev_info() states "No need to compare
PCI domain; it has to be the same". But for the si_domain that isn't
going to be true, as it consists of all the PCI devices that are
identity mapped thus multiple PCI domains can be in si_domain. The
code needs to validate the PCI domain too.
Signed-off-by: Mike Habeck <habeck@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: stable@kernel.org
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
When using the 1:1 (identity) PCI DMA remapping, PCI Host Bridge devices
that do not use the IOMMU causes a kernel panic. Fix that by not
inserting those devices into the si_domain.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Mike Habeck <habeck@sgi.com>
Cc: stable@kernel.org
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The __intel_map_single function is not honoring the passed in DMA mask.
This results in not using the coherent DMA mask when called from
intel_alloc_coherent().
Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Chris Wright <chrisw@sous-sol.org>
Reviewed-by: Mike Habeck <habeck@sgi.com>
Cc: stable@kernel.org
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Mike Travis and Mike Habeck reported an issue where iova allocation
would return a range that was larger than a device's dma mask.
https://lkml.org/lkml/2011/3/29/423
The dmar initialization code will reserve all PCI MMIO regions and copy
those reservations into a domain specific iova tree. It is possible for
one of those regions to be above the dma mask of a device. It is typical
to allocate iovas with a 32bit mask (despite device's dma mask possibly
being larger) and cache the result until it exhausts the lower 32bit
address space. Freeing the iova range that is >= the last iova in the
lower 32bit range when there is still an iova above the 32bit range will
corrupt the cached iova by pointing it to a region that is above 32bit.
If that region is also larger than the device's dma mask, a subsequent
allocation will return an unusable iova and cause dma failure.
Simply don't cache an iova that is above the 32bit caching boundary.
Reported-by: Mike Travis <travis@sgi.com>
Reported-by: Mike Habeck <habeck@sgi.com>
Cc: stable@kernel.org
Acked-by: Mike Travis <travis@sgi.com>
Tested-by: Mike Habeck <habeck@sgi.com>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
When there are a large count of PCI devices, and the pass through
option for iommu is set, much time is spent in the identity_mapping
function hunting though the iommu domains to check if a specific
device is "identity mapped".
Speed up the function by checking the cached info to see if
it's mapped to the static identity domain.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Mike Habeck <habeck@sgi.com>
Cc: stable@kernel.org
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The identity mapping code appears to make the assumption that if the
devices dma_mask is greater than 32bits the device can use identity
mapping. But that is not true: take the case where we have a 40bit
device in a 44bit architecture. The device can potentially receive a
physical address that it will truncate and cause incorrect addresses
to be used.
Instead check to see if the device's dma_mask is large enough
to address the system's dma_mask.
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Mike Habeck <habeck@sgi.com>
Cc: stable@kernel.org
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Commit a97590e5 added unlinking domains from iommus to reciprocate the
iommu from domains unlinking that was already done. We actually want
to only do this for device domains and never for the static
identity map domain or VM domains. The SI domain is special and
never freed, while VM domain->id lives in their own special address
space, separate from iommu->domain_ids.
In the current code, a VM can get domain->id zero, then mark that
domain unused when unbound from pci-stub. This leads to DMAR
write faults when the device is re-bound to the host driver.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Cc: stable@kernel.org
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
There are no externally-visible changes with this. In the loop in the
internal __domain_mapping() function, we simply detect if we are mapping:
- size >= 2MiB, and
- virtual address aligned to 2MiB, and
- physical address aligned to 2MiB, and
- on hardware that supports superpages.
(and likewise for larger superpages).
We automatically use a superpage for such mappings. We never have to
worry about *breaking* superpages, since we trust that we will always
*unmap* the same range that was mapped. So all we need to do is ensure
that dma_pte_clear_range() will also cope with superpages.
Adjust pfn_to_dma_pte() to take a superpage 'level' as an argument, so
it can return a PTE at the appropriate level rather than always
extending the page tables all the way down to level 1. Again, this is
simplified by the fact that we should never encounter existing small
pages when we're creating a mapping; any old mapping that used the same
virtual range will have been entirely removed and its obsolete page
tables freed.
Provide an 'intel_iommu=sp_off' argument on the command line as a
chicken bit. Not that it should ever be required.
==
The original commit seen in the iommu-2.6.git was Youquan's
implementation (and completion) of my own half-baked code which I'd
typed into an email. Followed by half a dozen subsequent 'fixes'.
I've taken the unusual step of rewriting history and collapsing the
original commits in order to keep the main history simpler, and make
life easier for the people who are going to have to backport this to
older kernels. And also so I can give it a more coherent commit comment
which (hopefully) gives a better explanation of what's going on.
The original sequence of commits leading to identical code was:
Youquan Song (3):
intel-iommu: super page support
intel-iommu: Fix superpage alignment calculation error
intel-iommu: Fix superpage level calculation error in dma_pfn_level_pte()
David Woodhouse (4):
intel-iommu: Precalculate superpage support for dmar_domain
intel-iommu: Fix hardware_largepage_caps()
intel-iommu: Fix inappropriate use of superpages in __domain_mapping()
intel-iommu: Fix phys_pfn in __domain_mapping for sglist pages
Signed-off-by: Youquan Song <youquan.song@intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Fix build warnings in physmap.h:
include/linux/mtd/physmap.h:25: warning: 'struct platform_device' declared inside parameter list
include/linux/mtd/physmap.h:25: warning: its scope is only this definition or declaration, which is probably not what you want
include/linux/mtd/physmap.h:26: warning: 'struct platform_device' declared inside parameter list
include/linux/mtd/physmap.h:27: warning: 'struct platform_device' declared inside parameter list
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Unfortunately, the recovery fix d1606a59b6be4ea392eabd40d1250aa1eeb19efb
(UBIFS: fix extremely rare mount failure) broke recovery. This commit make
UBIFS drop the last min. I/O unit in all journal heads, but this is needed only
for the GC head. And this does not work for non-GC heads. For example, if
suppose we have min. I/O units A and B, and A contains a valid node X, which
was fsynced, and then a group of nodes Y which spans the rest of A and B. In
this case we'll drop not only Y, but also X, which is obviously incorrect.
This patch fixes the issue and additionally makes recovery to drop last min.
I/O unit only for the GC head, and leave things as they have been for ages for
the other heads - this is safer.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Instead of passing "grouped" parameter to 'ubifs_recover_leb()' which tells
whether the nodes are grouped in the LEB to recover, pass the journal head
number and let 'ubifs_recover_leb()' look at the journal head's 'grouped' flag.
This patch is a preparation to a further fix where we'll need to know the
journal head number for other purposes.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Raymond Yau
Takashi Iwai
Matt Carlson
Koki Sanagi
Chris Metcalf
Jesper Juhl
Steven Rostedt
Jens Axboe
Paul Bolle
James Bottomley
Julia Lawall
Guennadi Liakhovetski
Mark Brown
Ben Greear
sjur.brandeland@stericsson.com
Stefan Metzmacher
Wei Yongjun
Marcus Meissner
Guenter Roeck
Linus Torvalds
Stanislaw Gruszka
Johannes Berg
Daniel Halperin
Eliad Peller
Jean Delvare
Per Dalen
Dan Carpenter
Laszlo Ersek
David Woodhouse
Mike Habeck
Mike Travis
Chris Wright
Alex Williamson
Youquan Song
Randy Dunlap
Artem Bityutskiy