commit c78a10aebb upstream.
When recovering a relocation, if we run into a reloc root that has 0
refs we simply add it to the reloc_control->reloc_roots list, and then
clean it up later. The problem with this is __del_reloc_root() doesn't
do anything if the root isn't in the radix tree, which in this case it
won't be because we never call __add_reloc_root() on the reloc_root.
This exit condition simply isn't correct really. During normal
operation we can remove ourselves from the rb tree and then we're meant
to clean up later at merge_reloc_roots() time, and this happens
correctly. During recovery we're depending on free_reloc_roots() to
drop our references, but we're short-circuiting.
Fix this by continuing to check if we're on the list and dropping
ourselves from the reloc_control root list and dropping our reference
appropriately. Change the corresponding BUG_ON() to an ASSERT() that
does the correct thing if we aren't in the rb tree.
CC: stable@vger.kernel.org # 4.4+
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 867ed321f9 upstream.
While testing my error handling patches, I added a error injection site
at btrfs_inc_extent_ref, to validate the error handling I added was
doing the correct thing. However I hit a pretty ugly corruption while
doing this check, with the following error injection stack trace:
btrfs_inc_extent_ref
btrfs_copy_root
create_reloc_root
btrfs_init_reloc_root
btrfs_record_root_in_trans
btrfs_start_transaction
btrfs_update_inode
btrfs_update_time
touch_atime
file_accessed
btrfs_file_mmap
This is because we do not catch the error from btrfs_inc_extent_ref,
which in practice would be ENOMEM, which means we lose the extent
references for a root that has already been allocated and inserted,
which is the problem. Fix this by aborting the transaction if we fail
to do the reference modification.
CC: stable@vger.kernel.org # 4.4+
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 8da7520c80 upstream.
Consider the following transcript:
$ keyctl add trusted kmk "new 32 blobauth=helloworld keyhandle=80000000 migratable=1" @u
add_key: Invalid argument
The documentation has the following description:
migratable= 0|1 indicating permission to reseal to new PCR values,
default 1 (resealing allowed)
The consequence is that "migratable=1" should succeed. Fix this by
allowing this condition to pass instead of return -EINVAL.
[*] Documentation/security/keys/trusted-encrypted.rst
Cc: stable@vger.kernel.org
Cc: "James E.J. Bottomley" <jejb@linux.ibm.com>
Cc: Mimi Zohar <zohar@linux.ibm.com>
Cc: David Howells <dhowells@redhat.com>
Fixes: d00a1c72f7 ("keys: add new trusted key-type")
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e42acf104d upstream.
The current release locality code seems to be based on the
misunderstanding that the TPM interrupts when a locality is released:
it doesn't, only when the locality is acquired.
Furthermore, there seems to be no point in waiting for the locality to
be released. All it does is penalize the last TPM user. However, if
there's no next TPM user, this is a pointless wait and if there is a
next TPM user, they'll pay the penalty waiting for the new locality
(or possibly not if it's the same as the old locality).
Fix the code by making release_locality as simple write to release
with no waiting for completion.
Cc: stable@ger.kernel.org
Fixes: 33bafe9082 ("tpm_tis: verify locality released before returning from release_locality")
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 3d9ae54af1 upstream.
The TPM TIS specification says the TPM signals the acquisition of locality
when the TMP_ACCESS_REQUEST_USE bit goes to one *and* the
TPM_ACCESS_REQUEST_USE bit goes to zero. Currently we only check the
former not the latter, so check both. Adding the check on
TPM_ACCESS_REQUEST_USE should fix the case where the locality is
re-requested before the TPM has released it. In this case the locality may
get released briefly before it is reacquired, which causes all sorts of
problems. However, with the added check, TPM_ACCESS_REQUEST_USE should
remain 1 until the second request for the locality is granted.
Cc: stable@ger.kernel.org
Fixes: 27084efee0 ("[PATCH] tpm: driver for next generation TPM chips")
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com>
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ce06312918 upstream.
Currently, although set_bit() & test_bit() pairs are used as a fast-
path for initialized configurations. However, these atomic ops are
actually relaxed forms. Instead, load-acquire & store-release form is
needed to make sure uninitialized fields won't be observed in advance
here (yet no such corresponding bitops so use full barriers instead.)
Link: https://lore.kernel.org/r/20210209130618.15838-1-hsiangkao@aol.com
Fixes: 62dc45979f ("staging: erofs: fix race of initializing xattrs of a inode at the same time")
Fixes: 152a333a58 ("staging: erofs: add compacted compression indexes support")
Cc: <stable@vger.kernel.org> # 5.3+
Reported-by: Huang Jianan <huangjianan@oppo.com>
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 688f97ed3f upstream.
When run igt@kms_vrr in a device that uses DCN2.1 architecture, we
noticed multiple failures. Furthermore, when we tested a VRR demo, we
noticed a system hang where the mouse pointer still works, but the
entire system freezes; in this case, we don't see any dmesg warning or
failure messages kernel. This happens due to a lack of vupdate_no_lock
interrupt, making the userspace wait eternally to get the event back.
For fixing this issue, we need to add the vupdate_no_lock interrupt in
the interrupt list.
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Acked-by: Bindu Ramamurthy <bindu.r@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit afe78ab46f upstream.
This is potentially long running and not latency sensitive, let's get
it out of the way of other latency sensitive events.
As observed in the previous commit, the `system_wq` comes easily
congested by bcache, and this fixes a few more stalls I was observing
every once in a while.
Let's not make this `WQ_MEM_RECLAIM` as it showed to reduce performance
of boot and file system operations in my tests. Also, without
`WQ_MEM_RECLAIM`, I no longer see desktop stalls. This matches the
previous behavior as `system_wq` also does no memory reclaim:
> // workqueue.c:
> system_wq = alloc_workqueue("events", 0, 0);
Cc: Coly Li <colyli@suse.de>
Cc: stable@vger.kernel.org # 5.4+
Signed-off-by: Kai Krakow <kai@kaishome.de>
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d797bd9897 upstream.
Before killing `btree_io_wq`, the queue was allocated using
`create_singlethread_workqueue()` which has `WQ_MEM_RECLAIM`. After
killing it, it no longer had this property but `system_wq` is not
single threaded.
Let's combine both worlds and make it multi threaded but able to
reclaim memory.
Cc: Coly Li <colyli@suse.de>
Cc: stable@vger.kernel.org # 5.4+
Signed-off-by: Kai Krakow <kai@kaishome.de>
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9f233ffe02 upstream.
This reverts commit 56b30770b2.
With the btree using the `system_wq`, I seem to see a lot more desktop
latency than I should.
After some more investigation, it looks like the original assumption
of 56b3077 no longer is true, and bcache has a very high potential of
congesting the `system_wq`. In turn, this introduces laggy desktop
performance, IO stalls (at least with btrfs), and input events may be
delayed.
So let's revert this. It's important to note that the semantics of
using `system_wq` previously mean that `btree_io_wq` should be created
before and destroyed after other bcache wqs to keep the same
assumptions.
Cc: Coly Li <colyli@suse.de>
Cc: stable@vger.kernel.org # 5.4+
Signed-off-by: Kai Krakow <kai@kaishome.de>
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 0eaa1a3714 upstream.
musb_queue_resume_work() would call the provided callback if the runtime
PM status was 'active'. Otherwise, it would enqueue the request if the
hardware was still suspended (musb->is_runtime_suspended is true).
This causes a race with the runtime PM handlers, as it is possible to be
in the case where the runtime PM status is not yet 'active', but the
hardware has been awaken (PM resume function has been called).
When hitting the race, the resume work was not enqueued, which probably
triggered other bugs further down the stack. For instance, a telnet
connection on Ingenic SoCs would result in a 50/50 chance of a
segmentation fault somewhere in the musb code.
Rework the code so that either we call the callback directly if
(musb->is_runtime_suspended == 0), or enqueue the query otherwise.
Fixes: ea2f35c01d ("usb: musb: Fix sleeping function called from invalid context for hdrc glue")
Cc: stable@vger.kernel.org # v4.9+
Tested-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Link: https://lore.kernel.org/r/20210123142502.16980-1-paul@crapouillou.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit b5d6e7ab7f upstream.
After commit 77b425399f ("Input: i8042 - use chassis info to skip
selftest on Asus laptops"), all modern Asus laptops have the i8042
selftest disabled. It has done by using chassys type "10" (laptop).
The Asus Zenbook Flip suffers from similar suspend/resume issues, but
it _sometimes_ work and sometimes it doesn't. Setting noselftest makes
it work reliably. In this case, we need to add chassis type "31"
(convertible) in order to avoid selftest in this device.
Reported-by: Ludvig Norgren Guldhag <ludvigng@gmail.com>
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Link: https://lore.kernel.org/r/20210219164638.761-1-mpdesouza@suse.com
Cc: stable@vger.kernel.org
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 88f38846bf upstream.
The `wacom_feature_mapping` function is careful to only set the the
touch_max value a single time, but this care does not extend to the
`wacom_wac_finger_event` function. In particular, if a device sends
multiple HID_DG_CONTACTMAX items in a single feature report, the
driver will end up retaining the value of last item.
The HID descriptor for the Cintiq Companion 2 does exactly this. It
incorrectly sets a "Report Count" of 2, which will cause the driver
to process two HID_DG_CONTACTCOUNT items. The first item has the actual
count, while the second item should have been declared as a constant
zero. The constant zero is the value the driver ends up using, however,
since it is the last HID_DG_CONTACTCOUNT in the report.
Report ID (16),
Usage (Contact Count Maximum), ; Contact count maximum (55h, static value)
Report Count (2),
Logical Maximum (10),
Feature (Variable),
To address this, we add a check that the touch_max is not already set
within the `wacom_wac_finger_event` function that processes the
HID_DG_TOUCHMAX item. We emit a warning if the value is set and ignore
the updated value.
This could potentially cause problems if there is a tablet which has
a similar issue but requires the last item to be used. This is unlikely,
however, since it would have to have a different non-zero value for
HID_DG_CONTACTMAX earlier in the same report, which makes no sense
except in the case of a firmware bug. Note that cases where the
HID_DG_CONTACTMAX items are in different reports is already handled
(and similarly ignored) by `wacom_feature_mapping` as mentioned above.
Link: https://github.com/linuxwacom/input-wacom/issues/223
Fixes: 184eccd403 ("HID: wacom: generic: read HID_DG_CONTACTMAX from any feature report")
Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com>
CC: stable@vger.kernel.org
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ef07c116d9 upstream.
In e400071a80 I added support for the
receiver that comes with the G602 device, but unfortunately I screwed up
during testing and it seems the keyboard events were actually not being
sent to userspace.
This resulted in keyboard events being broken in userspace, please
backport the fix.
The receiver uses the normal 0x01 Logitech keyboard report descriptor,
as expected, so it is just a matter of flagging it as supported.
Reported in
https://github.com/libratbag/libratbag/issues/1124
Fixes: e400071a80 ("HID: logitech-dj: add the G602 receiver")
Cc: <stable@vger.kernel.org>
Signed-off-by: Filipe Laíns <lains@riseup.net>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 67e40054de upstream.
A list_add corruption is reported by Hulk Robot like this:
==============
list_add corruption.
Call Trace:
link_obj+0xc0/0x1c0
link_group+0x21/0x140
configfs_register_subsystem+0xdb/0x380
acpi_configfs_init+0x25/0x1000 [acpi_configfs]
do_one_initcall+0x149/0x820
do_init_module+0x1ef/0x720
load_module+0x35c8/0x4380
__do_sys_finit_module+0x10d/0x1a0
do_syscall_64+0x34/0x80
It's because of the missing check after configfs_register_default_group,
where configfs_unregister_subsystem should be called once failure.
Fixes: 612bd01fc6 ("ACPI: add support for loading SSDTs via configfs")
Reported-by: Hulk Robot <hulkci@huawei.com>
Suggested-by: Hanjun Guo <guohanjun@huawei.com>
Signed-off-by: Qinglang Miao <miaoqinglang@huawei.com>
Cc: 4.10+ <stable@vger.kernel.org> # 4.10+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e1e6bd2995 upstream.
Property matching does not work for ACPI fwnodes if the value of the
given property is not represented as a package in the _DSD package
containing it. For example, the "compatible" property in the _DSD
below
Name (_DSD, Package () {
ToUUID("daffd814-6eba-4d8c-8a91-bc9bbf4aa301"),
Package () {
Package () {"compatible", "ethernet-phy-ieee802.3-c45"}
}
})
will not be found by fwnode_property_match_string(), because the ACPI
code handling device properties does not regard the single value as a
"list" in that case.
Namely, fwnode_property_match_string() invoked to match a given
string property value first calls fwnode_property_read_string_array()
with the last two arguments equal to NULL and 0, respectively, in
order to count the items in the value of the given property, with the
assumption that this value may be an array. For ACPI fwnodes, that
operation is carried out by acpi_node_prop_read() which calls
acpi_data_prop_read() for this purpose. However, when the return
(val) pointer is NULL, that function only looks for a property whose
value is a package without checking the single-value case at all.
To fix that, make acpi_data_prop_read() check the single-value
case if its return pointer argument is NULL and modify
acpi_data_prop_read_single() handling that case to attempt to
read the value of the property if the return pointer is NULL
and return 1 if that succeeds.
Fixes: 3708184afc ("device property: Move FW type specific functionality to FW specific files")
Reported-by: Calvin Johnson <calvin.johnson@oss.nxp.com>
Cc: 4.13+ <stable@vger.kernel.org> # 4.13+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 97f433c360 upstream.
We get I/O errors when we run md-raid1 on the top of dm-integrity on the
top of ramdisk.
device-mapper: integrity: Bio not aligned on 8 sectors: 0xff00, 0xff
device-mapper: integrity: Bio not aligned on 8 sectors: 0xff00, 0xff
device-mapper: integrity: Bio not aligned on 8 sectors: 0xffff, 0x1
device-mapper: integrity: Bio not aligned on 8 sectors: 0xffff, 0x1
device-mapper: integrity: Bio not aligned on 8 sectors: 0x8048, 0xff
device-mapper: integrity: Bio not aligned on 8 sectors: 0x8147, 0xff
device-mapper: integrity: Bio not aligned on 8 sectors: 0x8246, 0xff
device-mapper: integrity: Bio not aligned on 8 sectors: 0x8345, 0xbb
The ramdisk device has logical_block_size 512 and max_sectors 255. The
dm-integrity device uses logical_block_size 4096 and it doesn't affect the
"max_sectors" value - thus, it inherits 255 from the ramdisk. So, we have
a device with max_sectors not aligned on logical_block_size.
The md-raid device sees that the underlying leg has max_sectors 255 and it
will split the bios on 255-sector boundary, making the bios unaligned on
logical_block_size.
In order to fix the bug, we round down max_sectors to logical_block_size.
Cc: stable@vger.kernel.org
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit a1858ce0cf ]
The brcmstb_send_i2c_cmd currently has a condition that is (CMD_RD ||
CMD_WR) which always evaluates to true, while the obvious fix is to test
whether the cmd variable passed as parameter holds one of these two
values.
Fixes: dd1aa2524b ("i2c: brcmstb: Add Broadcom settop SoC i2c controller driver")
Reported-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Wolfram Sang <wsa@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9d41053e8d ]
Although there has been a bit of back and forth on the subject, it
appears that invalidating TLBs requires an ISB instruction when FEAT_ETS
is not implemented by the CPU.
From the bible:
| In an implementation that does not implement FEAT_ETS, a TLB
| maintenance instruction executed by a PE, PEx, can complete at any
| time after it is issued, but is only guaranteed to be finished for a
| PE, PEx, after the execution of DSB by the PEx followed by a Context
| synchronization event
Add the missing ISB in __primary_switch, just in case.
Fixes: 3c5e9f238b ("arm64: head.S: move KASLR processing out of __enable_mmu()")
Suggested-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Link: https://lore.kernel.org/r/20210224093738.3629662-3-maz@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 15d28d0d11 ]
In the fast_find_migrateblock(), it iterates ocer the freelist to find the
proper pageblock. But there are some misbehaviors.
First, if the page we found is equal to cc->migrate_pfn, it is considered
that we didn't find a suitable pageblock. Secondly, if the loop was
terminated because order is less than PAGE_ALLOC_COSTLY_ORDER, it could be
considered that we found a suitable one. Thirdly, if the skip bit is set
on the page block and we goto continue, it doesn't check nr_scanned.
Fourthly, if the page block's skip bit is set, it checks that page block
is the last of list, which is unnecessary.
Link: https://lkml.kernel.org/r/20210128130411.6125-1-vvghjk1234@gmail.com
Fixes: 70b44595ea ("mm, compaction: use free lists to quickly locate a migration source")
Signed-off-by: Wonhyuk Yang <vvghjk1234@gmail.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 07956b6269 ]
follow_pfn() doesn't make sure that we're using the correct page
protections, get the pte with follow_pte() so that we can test
protections and get the pfn from the pte.
Fixes: 5cbf3264bc ("vfio/type1: Fix VA->PA translation for PFNMAP VMAs in vaddr_get_pfn()")
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit dc88126264 ]
When creating VFs they were sometimes not getting resources.
It was caused by not executing i40e_reset_all_vfs due to
flag __I40E_VF_DISABLE being set on PF. Because of this
IAVF was never able to finish setup sequence never
getting reset indication from PF.
Changed test_and_set_bit __I40E_VF_DISABLE in
i40e_sync_filters_subtask to test_bit and removed clear_bit.
This function should not set this bit it should only check
if it hasn't been already set.
Fixes: a7542b8760 ("i40e: check __I40E_VF_DISABLE bit in i40e_sync_filters_subtask")
Signed-off-by: Sylwester Dziedziuch <sylwesterx.dziedziuch@intel.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 28b1208e7a ]
Fix addition of VLAN filter for PF after enabling FW LLDP agent.
Changing LLDP Agent causes FW to re-initialize per NVM settings.
Remove default PF filter and move "Enable/Disable" to currently used
reset flag.
Without this patch PF would try to add MAC VLAN filter with default
switch filter present. This causes AQ error and sets promiscuous mode
on.
Fixes: c65e78f87f ("i40e: Further implementation of LLDP")
Signed-off-by: Przemyslaw Patynowski <przemyslawx.patynowski@intel.com>
Signed-off-by: Mateusz Palczewski <mateusz.palczewski@intel.com>
Reviewed-by: Sylwester Dziedziuch <sylwesterx.dziedziuch@intel.com>
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 92c6058024 ]
When a packet contains an IPv6 header with next header which is
an extension header and not a protocol one, the kernel function
skb_transport_header called with such sk_buff will return a
pointer to the extension header and not to the TCP one.
The above explained call caused a problem with packet processing
for skb with encapsulation for tunnel with I40E_TX_CTX_EXT_IP_IPV6.
The extension header was not skipped at all.
The ipv6_skip_exthdr function does check if next header of the IPV6
header is an extension header and doesn't modify the l4_proto pointer
if it points to a protocol header value so its safe to omit the
comparison of exthdr and l4.hdr pointers. The ipv6_skip_exthdr can
return value -1. This means that the skipping process failed
and there is something wrong with the packet so it will be dropped.
Fixes: a3fd9d8876 ("i40e/i40evf: Handle IPv6 extension headers in checksum offload")
Signed-off-by: Slawomir Laba <slawomirx.laba@intel.com>
Signed-off-by: Przemyslaw Patynowski <przemyslawx.patynowski@intel.com>
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
