Pull kselftest fixes from Shuah Khan:
"This update consists of important compile and run-time error fixes to
timers/freq-step, kmod, and sysctl tests"
* tag 'linux-kselftest-4.13-rc6-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest:
selftests: timers: freq-step: fix compile error
selftests: futex: fix run_tests target
test_sysctl: fix sysctl.sh by making it executable
test_kmod: fix kmod.sh by making it executable
Julian Wiedmann says:
====================
s390/net: updates for 4.14
a mixed bag of minor fixes, cleanups and refactors for net-next. Please apply.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
IP-address setting and removal are delayed when the device is not yet in
state SOFTSETUP or UP. ref_counter has been implemented only for
ip-address with type normal. In this patch ref_counter logic is also used
for ip-address with type rxip to allow appropriate handling of multiple
postponed rxip add and del calls.
Signed-off-by: Kittipon Meesompop <kmeesomp@linux.vnet.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There exist different commands to add unicast and multicast addresses on
the OSA card. rxip addresses are always set as unicast addresses and
thus just unicast addresses should be allowed.
Adding a multicast address now fails and a grace message is generated.
Signed-off-by: Kittipon Meesompop <kmeesomp@linux.vnet.ibm.com>
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Consolidation of duplicated code, no functional change.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Clean up the inline cruft in s390 net drivers. Many of the inlined
functions had only one caller anyway.
Suggested-by: Joe Perches <joe@perches.com>
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Replace some open-coded parts with their proper API calls.
Also remove two skb_[re]set_mac_header() calls in the L2
xmit paths that are clearly no longer required, since at least
commit 6d1ccff627 ("net: reset mac header in dev_start_xmit()").
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Acked-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For some xmit paths we pass down a data offset to qeth_fill_buffer(),
to indicate that the first k bytes of the skb should be skipped when
mapping it into buffer elements.
Commit acd9776b5c ("s390/qeth: no ETH header for outbound AF_IUCV")
recently switched the offset for the IUCV-over-HiperSockets path
from 0 to ETH_HLEN, and now we have
device offset
OSA = 0
IQD > 0
for all xmit paths.
OSA would previously pass down -1 from do_send_packet(), to distinguish
between 1) OSA and 2) IQD with offset 0. That's no longer needed now,
so have it pass 0, make the offset unsigned and clean up how we apply
the offset in __qeth_fill_buffer().
No change of behaviour for any of our current xmit paths.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Acked-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
1. for adjusting the buffer's next_element_to_fill in __fill_buffer(),
just pass the full qeth_qdio_out_buffer struct
2. when adding a header element, be consistent about passing
a hint ('is_first_elem') to __fill_buffer()
No functional change.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Acked-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Improve readability of the code that determines a buffer element's
fragment type, and reduce the number of cases down from 5 to 3.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Acked-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
qeth_l3_setadapter_parms() queries the device for supported
adapterparms, even though they already have been queried as part of the
device's high-level setup. Remove that extra call.
The only call chain for qeth_l3_setadapter_parms() is
__qeth_l3_set_online()
qeth_core_hardsetup_card()
qeth_query_setadapterparms()
qeth_l3_setadapter_parms()
qeth_query_setadapterparms()
, and we only reach qeth_l3_setadapter_parms() if the first
adapterparms query succeeds. Hence removing the second query results in
no loss of functionality.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Acked-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
qeth_l2_request_initial_mac() queries the device for its supported
adapterparms, even though they already have been queried as part of the
device's high-level setup. Remove that extra call.
The only call chain for qeth_l2_request_initial_mac() is
__qeth_l2_set_online()
qeth_core_hardsetup_card()
qeth_query_setadapterparms()
qeth_l2_setup_netdev()
qeth_l2_request_initial_mac()
qeth_query_setadapterparms()
, and we only reach qeth_l2_request_initial_mac() if the first
adapterparms query succeeds. Hence removing the second query results in
no loss of functionality.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Acked-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After transmitting a skb via send_packet[_fast](), the statistics
code accesses the skb once more to account for transmitted page frags.
This has a (theoretical?) race against the TX completion - if the TX
completion is processed and frees the skb before hard_start_xmit()
gets to the statistics part, we access random memory.
Fix this by caching the # of page frags, before the skb is transmitted.
Signed-off-by: Julian Wiedmann <jwi@linux.vnet.ibm.com>
Acked-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Kalle Valo says:
====================
wireless-drivers fixes for 4.13
This time quite a few fixes for iwlwifi and one major regression fix
for brcmfmac. For the iwlwifi aggregation bug a small change was
needed for mac80211, but as Johannes is still away the mac80211 patch
is taken via wireless-drivers tree.
brcmfmac
* fix firmware crash (a recent regression in bcm4343{0,1,8}
iwlwifi
* Some simple PCI HW ID fix-ups and additions for family 9000
* Remove a bogus warning message with new FWs (bug #196915)
* Don't allow illegal channel options to be used (bug #195299)
* A fix for checksum offload in family 9000
* A fix serious throughput degradation in 11ac with multiple streams
* An old bug in SMPS where the firmware was not aware of SMPS changes
* Fix a memory leak in the SAR code
* Fix a stuck queue case in AP mode;
* Convert a WARN to a simple debug in a legitimate race case (from
which we can recover)
* Fix a severe throughput aggregation on 9000-family devices due to
aggregation issues, needed a small change in mac80211
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull "DT fixes for 4.13" from Alexandre Belloni:
- Fix NAND flash support for sama5d2
* tag 'at91-ab-4.13-dt-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux:
ARM: dts: at91: sama5d2: fix EBI/NAND controllers declaration
ARM: dts: at91: sama5d2: use sama5d2 compatible string for SMC
Pull "i.MX fixes for 4.13, round 2" from Shawn Guo:
- Add missing 'ranges' property for i.MX25 device tree TSCADC node, so
that it's child nodes ADC and TSC device can be probed by kernel.
- Fix i.MX GPCv2 power domain driver to request regulator after power
domain initialization, since regulator could defer probing and
therefore cause power domain initialized twice.
* tag 'imx-fixes-4.13-2' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
ARM: dts: i.MX25: add ranges to tscadc
soc: imx: gpcv2: fix regulator deferred probe
Pull "i.MX fixes for 4.13" from Shawn Guo:
- A fix for imx7d-sdb board to move pinctrl_spi4 pins from low power
iomux controller to normal iomuxc node, as the pins belong to normal
iomuxc rather than low power one.
* tag 'imx-fixes-4.13' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
ARM: dts: imx7d-sdb: Put pinctrl_spi4 in the correct location
Since commit d05d7f4079 ("Merge branch 'for-4.8/core' of
git://git.kernel.dk/linux-block") and 3fc9d69093 ("Merge branch
'for-4.8/drivers' of git://git.kernel.dk/linux-block"), blkfront_resume()
has been using an index for iterating ring_info to check request when
iterating blk_shadow in an inner loop. This seems to have been
accidentally introduced during the massive rewrite of the block layer
macros in the commits.
This may cause crash like this:
[11798.057074] BUG: unable to handle kernel NULL pointer dereference at 0000000000000048
[11798.058832] IP: [<ffffffff814411fa>] blkfront_resume+0x10a/0x610
....
[11798.061063] Call Trace:
[11798.061063] [<ffffffff8139ce93>] xenbus_dev_resume+0x53/0x140
[11798.061063] [<ffffffff8139ce40>] ? xenbus_dev_probe+0x150/0x150
[11798.061063] [<ffffffff813f359e>] dpm_run_callback+0x3e/0x110
[11798.061063] [<ffffffff813f3a08>] device_resume+0x88/0x190
[11798.061063] [<ffffffff813f4cc0>] dpm_resume+0x100/0x2d0
[11798.061063] [<ffffffff813f5221>] dpm_resume_end+0x11/0x20
[11798.061063] [<ffffffff813950a8>] do_suspend+0xe8/0x1a0
[11798.061063] [<ffffffff813954bd>] shutdown_handler+0xfd/0x130
[11798.061063] [<ffffffff8139aba0>] ? split+0x110/0x110
[11798.061063] [<ffffffff8139ac26>] xenwatch_thread+0x86/0x120
[11798.061063] [<ffffffff810b4570>] ? prepare_to_wait_event+0x110/0x110
[11798.061063] [<ffffffff8108fe57>] kthread+0xd7/0xf0
[11798.061063] [<ffffffff811da811>] ? kfree+0x121/0x170
[11798.061063] [<ffffffff8108fd80>] ? kthread_park+0x60/0x60
[11798.061063] [<ffffffff810863b0>] ? call_usermodehelper_exec_work+0xb0/0xb0
[11798.061063] [<ffffffff810864ea>] ? call_usermodehelper_exec_async+0x13a/0x140
[11798.061063] [<ffffffff81534a45>] ret_from_fork+0x25/0x30
Use the right index in the inner loop.
Fixes: d05d7f4079 ("Merge branch 'for-4.8/core' of git://git.kernel.dk/linux-block")
Fixes: 3fc9d69093 ("Merge branch 'for-4.8/drivers' of git://git.kernel.dk/linux-block")
Signed-off-by: Munehisa Kamata <kamatam@amazon.com>
Reviewed-by: Thomas Friebel <friebelt@amazon.de>
Reviewed-by: Eduardo Valentin <eduval@amazon.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: Roger Pau Monne <roger.pau@citrix.com>
Cc: xen-devel@lists.xenproject.org
Cc: stable@vger.kernel.org
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
The current test for bio vec merging is not fully accurate and can be
tricked into merging bios when certain grant combinations are used.
The result of these malicious bio merges is a bio that extends past
the memory page used by any of the originating bios.
Take into account the following scenario, where a guest creates two
grant references that point to the same mfn, ie: grant 1 -> mfn A,
grant 2 -> mfn A.
These references are then used in a PV block request, and mapped by
the backend domain, thus obtaining two different pfns that point to
the same mfn, pfn B -> mfn A, pfn C -> mfn A.
If those grants happen to be used in two consecutive sectors of a disk
IO operation becoming two different bios in the backend domain, the
checks in xen_biovec_phys_mergeable will succeed, because bfn1 == bfn2
(they both point to the same mfn). However due to the bio merging,
the backend domain will end up with a bio that expands past mfn A into
mfn A + 1.
Fix this by making sure the check in xen_biovec_phys_mergeable takes
into account the offset and the length of the bio, this basically
replicates whats done in __BIOVEC_PHYS_MERGEABLE using mfns (bus
addresses). While there also remove the usage of
__BIOVEC_PHYS_MERGEABLE, since that's already checked by the callers
of xen_biovec_phys_mergeable.
CC: stable@vger.kernel.org
Reported-by: "Jan H. Schönherr" <jschoenh@amazon.de>
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
blk_mq_get_request() does not release the callers queue usage counter
when allocation fails. The caller still needs to account for its own
queue usage when it is unable to allocate a request.
Fixes: 1ad43c0078 ("blk-mq: don't leak preempt counter/q_usage_counter when allocating rq failed")
Reported-by: Max Gurtovoy <maxg@mellanox.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Tested-by: Max Gurtovoy <maxg@mellanox.com>
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Larry reported a CPU hotplug lock recursion in the MTRR code.
============================================
WARNING: possible recursive locking detected
systemd-udevd/153 is trying to acquire lock:
(cpu_hotplug_lock.rw_sem){.+.+.+}, at: [<c030fc26>] stop_machine+0x16/0x30
but task is already holding lock:
(cpu_hotplug_lock.rw_sem){.+.+.+}, at: [<c0234353>] mtrr_add_page+0x83/0x470
....
cpus_read_lock+0x48/0x90
stop_machine+0x16/0x30
mtrr_add_page+0x18b/0x470
mtrr_add+0x3e/0x70
mtrr_add_page() holds the hotplug rwsem already and calls stop_machine()
which acquires it again.
Call stop_machine_cpuslocked() instead.
Reported-and-tested-by: Larry Finger <Larry.Finger@lwfinger.net>
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1708140920250.1865@nanos
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Borislav Petkov <bp@suse.de>
A Senheisser headset requires the typical sample-rate quirk for
avoiding spurious errors from inquiring the current sample rate like:
usb 1-1: 2:1: cannot get freq at ep 0x4
usb 1-1: 3:1: cannot get freq at ep 0x83
The USB ID 1395:740a has to be added to the entries in
snd_usb_get_sample_rate_quirk().
Bugzilla: https://bugzilla.suse.com/show_bug.cgi?id=1052580
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
commit 4842e98f26 ("ALSA: seq: Fix race at
creating a queue") attempted to fix a race reported by syzkaller. That
fix has been described as follows:
"
When a sequencer queue is created in snd_seq_queue_alloc(),it adds the
new queue element to the public list before referencing it. Thus the
queue might be deleted before the call of snd_seq_queue_use(), and it
results in the use-after-free error, as spotted by syzkaller.
The fix is to reference the queue object at the right time.
"
Even with that fix in place, syzkaller reported a use-after-free error.
It specifically pointed to the last instruction "return q->queue" in
snd_seq_queue_alloc(). The pointer q is being used after kfree() has
been called on it.
It turned out that there is still a small window where a race can
happen. The window opens at
snd_seq_ioctl_create_queue()->snd_seq_queue_alloc()->queue_list_add()
and closes at
snd_seq_ioctl_create_queue()->queueptr()->snd_use_lock_use(). Between
these two calls, a different thread could delete the queue and possibly
re-create a different queue in the same location in queue_list.
This change prevents this situation by calling snd_use_lock_use() from
snd_seq_queue_alloc() prior to calling queue_list_add(). It is then the
caller's responsibility to call snd_use_lock_free(&q->use_lock).
Fixes: 4842e98f26 ("ALSA: seq: Fix race at creating a queue")
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Daniel Mentz <danielmentz@google.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Filtering the ACK packet was not put at the right place.
At this place, we already allocated a child and put it
into accept queue.
We absolutely need to call tcp_child_process() to release
its spinlock, or we will deadlock at accept() or close() time.
Found by syzkaller team (Thanks a lot !)
Fixes: 8fac365f63 ("tcp: Add a tcp_filter hook before handle ack packet")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: Chenbo Feng <fengc@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
copy_linear_skb() is broken; both of its callers actually
expect 'len' to be the amount we are trying to copy,
not the offset of the end.
Fix it keeping the meanings of arguments in sync with what the
callers (both of them) expect.
Also restore a saner behavior on EFAULT (i.e. preserving
the iov_iter position in case of failure):
The commit fd851ba9ca ("udp: harden copy_linear_skb()")
avoids the more destructive effect of the buggy
copy_linear_skb(), e.g. no more invalid memory access, but
said function still behaves incorrectly: when peeking with
offset it can fail with EINVAL instead of copying the
appropriate amount of memory.
Reported-by: Sasha Levin <alexander.levin@verizon.com>
Fixes: b65ac44674 ("udp: try to avoid 2 cache miss on dequeue")
Fixes: fd851ba9ca ("udp: harden copy_linear_skb()")
Signed-off-by: Al Viro <viro@ZenIV.linux.org.uk>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Tested-by: Sasha Levin <alexander.levin@verizon.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The fw_type module parameter isn't showing up in the
/sys/module/liquidio/parameters directory. Fix it by setting the read
permission bits for user, group, other in module_param_string(). Revise
the description of fw_type. Initialize the fw_type static char array with
the default value to conform to the module parameter description.
Signed-off-by: Derek Chickles <derek.chickles@cavium.com>
Signed-off-by: Felix Manlunas <felix.manlunas@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jiri Pirko says:
====================
mlxsw: Add support for nexthop group consolidation for IPv6
Arkadi says:
Due to limited ASIC resources the maximum number of routes is limited by
the nexthop resource. In order to improve the routing scale nexthop
consolidation should be performed.
In case of IPv4, the kernel does the consolidation of nexthops in the form
of the fib_info struct. In that case, the driver uses the fib_info's
address as a key for the internal nexthop group representative struct
lookup. In case of IPv6, the kernel doesn't do consolidation, thus the
driver should implement it by itself.
The hash value is calculated based on the nexthop set, by performing
bitwise xor on the ifindexs of the nexthops, in a similar way to IPV4's
kernel implementation. In case of collision a full match is performed
between the sets which include address and ifindex comparison.
In order to use the same hash table in both cases (IPv4/6), the rhashtable
is changed to operate on variable length key.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Due to limited ASIC resources the maximum number of routes is limited by
the nexthop resource. In order to improve the routing scale nexthop
consolidation should be performed.
This patch adds support for IPv6 neighbor consolidation. The hash value
is calculated based on the nexthop set, by performing bitwise xor on the
ifindexs of the nexthops, in a similar way to IPv4's kernel implementation.
In case of collision a full match is performed between the sets which
include address and ifindex comparison.
Non gateway nexthop groups are not inserted to the hash table due to
lack of nexthop device (ifindex).
Signed-off-by: Arkadi Sharshevsky <arkadis@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch does preparation before introducing IPv6 nexthop group
consolidation. Currently the nexthop group hash table is used only by
IPv4 and uses fixed key size. In order to support the IPv6's variable
length key the current table is changed.
Signed-off-by: Arkadi Sharshevsky <arkadis@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Intiyaz Basha says:
====================
liquidio: adding support for ethtool --set-ring feature
Code reorganization is required for adding ethtool --set-ring feature.
First seven patches are for code reorganization. The last patch is for
adding this feature.
Change Log:
V1 -> V2
Only patch #8 was changed: unnecessary parentheses were removed in two
if-statements in lio_ethtool_set_ringparam().
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
When a dst is created by addrconf_dst_alloc() for a host route or an
anycast route, dst->dev points to loopback dev while rt6->rt6i_idev
points to a real device.
When the real device goes down, the current cleanup code only checks for
dst->dev and assumes rt6->rt6i_idev->dev is the same. This causes the
refcount leak on the real device in the above situation.
This patch makes sure to always release the refcount taken on
rt6->rt6i_idev during dst_dev_put().
Fixes: 587fea7411 ("ipv6: mark DST_NOGC and remove the operation of
dst_free()")
Reported-by: John Stultz <john.stultz@linaro.org>
Tested-by: John Stultz <john.stultz@linaro.org>
Tested-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Wei Wang <weiwan@google.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
__tcp_ulp_find_autoload returns tcp_ulp_ops after taking a reference on
the module. Then, if ->init fails, tcp_set_ulp propagates the error but
nothing releases that reference.
Fixes: 734942cc4e ("tcp: ULP infrastructure")
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Ding Tianhong says:
====================
Add new PCI_DEV_FLAGS_NO_RELAXED_ORDERING flag
Some devices have problems with Transaction Layer Packets with the Relaxed
Ordering Attribute set. This patch set adds a new PCIe Device Flag,
PCI_DEV_FLAGS_NO_RELAXED_ORDERING, a set of PCI Quirks to catch some known
devices with Relaxed Ordering issues, and a use of this new flag by the
cxgb4 driver to avoid using Relaxed Ordering with problematic Root Complex
Ports.
It's been years since I've submitted kernel.org patches, I appolgise for the
almost certain submission errors.
v2: Alexander point out that the v1 was only a part of the whole solution,
some platform which has some issues could use the new flag to indicate
that it is not safe to enable relaxed ordering attribute, then we need
to clear the relaxed ordering enable bits in the PCI configuration when
initializing the device. So add a new second patch to modify the PCI
initialization code to clear the relaxed ordering enable bit in the
event that the root complex doesn't want relaxed ordering enabled.
The third patch was base on the v1's second patch and only be changed
to query the relaxed ordering enable bit in the PCI configuration space
to allow the Chelsio NIC to send TLPs with the relaxed ordering attributes
set.
This version didn't plan to drop the defines for Intel Drivers to use the
new checking way to enable relaxed ordering because it is not the hardest
part of the moment, we could fix it in next patchset when this patches
reach the goal.
v3: Redesigned the logic for pci_configure_relaxed_ordering when configuration,
If a PCIe device didn't enable the relaxed ordering attribute default,
we should not do anything in the PCIe configuration, otherwise we
should check if any of the devices above us do not support relaxed
ordering by the PCI_DEV_FLAGS_NO_RELAXED_ORDERING flag, then base on
the result if we get a return that indicate that the relaxed ordering
is not supported we should update our device to disable relaxed ordering
in configuration space. If the device above us doesn't exist or isn't
the PCIe device, we shouldn't do anything and skip updating relaxed ordering
because we are probably running in a guest.
v4: Rename the functions pcie_get_relaxed_ordering and pcie_disable_relaxed_ordering
according John's suggestion, and modify the description, use the true/false
as the return value.
We shouldn't enable relaxed ordering attribute by the setting in the root
complex configuration space for PCIe device, so fix it for cxgb4.
Fix some format issues.
v5: Removed the unnecessary code for some function which only return the bool
value, and add the check for VF device.
Make this patch set base on 4.12-rc5.
v6: Fix the logic error in the need to enable the relaxed ordering attribute for cxgb4.
v7: The cxgb4 drivers will enable the PCIe Capability Device Control[Relaxed
Ordering Enable] in PCI Probe() routine, this will break our current
solution for some platform which has problematic when enable the relaxed
ordering attribute. According to the latest recommendations, remove the
enable_pcie_relaxed_ordering(), although it could not cover the Peer-to-Peer
scene, but we agree to leave this problem until we really trigger it.
Make this patch set base on 4.12 release version.
v8: Change the second patch title and description to make it more reasonable,
add the acked-by from Alex and Ashok.
Add a new patch to enable the Relaxed Ordering Attribute for cxgb4vf driver.
Make this patch set base on 4.13-rc2.
v9: The document (https://software.intel.com/sites/default/files/managed/9e/
bc/64-ia-32-architectures-optimization-manual.pdf) indicate that the Xeon
processors based on Broadwell/Haswell microarchitecture has the problem
with Relaxed Ordering Attribute enabled, so add the whole list Device ID
from Intel to the patch.
v10: Significant rework based on Bjorn's feedback, reorganize the first 2 patches,
now the Intel and AMD erratum soc has been divided to the different patches,
rename the pcie_relaxed_ordering_supported() to pcie_relaxed_ordering_enabled(),
and no need to check every intervening switch except the root ports, update
some commits.
v11: We shouldn't let the Intel engineer to acked the AMD's erratum patch, fix the
funny mistake.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
cxgb4vf Ethernet driver now queries PCIe configuration space to
determine if it can send TLPs to it with the Relaxed Ordering
Attribute set, just like the pf did.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Reviewed-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
cxgb4 Ethernet driver now queries PCIe configuration space to determine
if it can send TLPs to it with the Relaxed Ordering Attribute set.
Remove the enable_pcie_relaxed_ordering() to avoid enable PCIe Capability
Device Control[Relaxed Ordering Enable] at probe routine, to make sure
the driver will not send the Relaxed Ordering TLPs to the Root Complex which
could not deal the Relaxed Ordering TLPs.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Reviewed-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Casey reported that the AMD ARM A1100 SoC has a bug in its PCIe
Root Port where Upstream Transaction Layer Packets with the Relaxed
Ordering Attribute clear are allowed to bypass earlier TLPs with
Relaxed Ordering set, it would cause Data Corruption, so we need
to disable Relaxed Ordering Attribute when Upstream TLPs to the
Root Port.
Reported-and-suggested-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Acked-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
According to the Intel spec section 3.9.1 said:
3.9.1 Optimizing PCIe Performance for Accesses Toward Coherent Memory
and Toward MMIO Regions (P2P)
In order to maximize performance for PCIe devices in the processors
listed in Table 3-6 below, the soft- ware should determine whether the
accesses are toward coherent memory (system memory) or toward MMIO
regions (P2P access to other devices). If the access is toward MMIO
region, then software can command HW to set the RO bit in the TLP
header, as this would allow hardware to achieve maximum throughput for
these types of accesses. For accesses toward coherent memory, software
can command HW to clear the RO bit in the TLP header (no RO), as this
would allow hardware to achieve maximum throughput for these types of
accesses.
Table 3-6. Intel Processor CPU RP Device IDs for Processors Optimizing
PCIe Performance
Processor CPU RP Device IDs
Intel Xeon processors based on 6F01H-6F0EH
Broadwell microarchitecture
Intel Xeon processors based on 2F01H-2F0EH
Haswell microarchitecture
It means some Intel processors has performance issue when use the Relaxed
Ordering Attribute, so disable Relaxed Ordering for these root port.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Acked-by: Alexander Duyck <alexander.h.duyck@intel.com>
Acked-by: Ashok Raj <ashok.raj@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When bit4 is set in the PCIe Device Control register, it indicates
whether the device is permitted to use relaxed ordering.
On some platforms using relaxed ordering can have performance issues or
due to erratum can cause data-corruption. In such cases devices must avoid
using relaxed ordering.
The patch adds a new flag PCI_DEV_FLAGS_NO_RELAXED_ORDERING to indicate that
Relaxed Ordering (RO) attribute should not be used for Transaction Layer
Packets (TLP) targeted towards these affected root complexes.
This patch checks if there is any node in the hierarchy that indicates that
using relaxed ordering is not safe. In such cases the patch turns off the
relaxed ordering by clearing the capability for this device.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: Ding Tianhong <dingtianhong@huawei.com>
Acked-by: Ashok Raj <ashok.raj@intel.com>
Acked-by: Alexander Duyck <alexander.h.duyck@intel.com>
Acked-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
