The nested_vmx_prepare_msr_bitmap() function doesn't directly guard the
x2APIC MSR intercepts with the "virtualize x2APIC mode" MSR. As a
result, we discovered the potential for a buggy or malicious L1 to get
access to L0's x2APIC MSRs, via an L2, as follows.
1. L1 executes WRMSR(IA32_SPEC_CTRL, 1). This causes the spec_ctrl
variable, in nested_vmx_prepare_msr_bitmap() to become true.
2. L1 disables "virtualize x2APIC mode" in VMCS12.
3. L1 enables "APIC-register virtualization" in VMCS12.
Now, KVM will set VMCS02's x2APIC MSR intercepts from VMCS12, and then
set "virtualize x2APIC mode" to 0 in VMCS02. Oops.
This patch closes the leak by explicitly guarding VMCS02's x2APIC MSR
intercepts with VMCS12's "virtualize x2APIC mode" control.
The scenario outlined above and fix prescribed here, were verified with
a related patch in kvm-unit-tests titled "Add leak scenario to
virt_x2apic_mode_test".
Note, it looks like this issue may have been introduced inadvertently
during a merge---see 15303ba5d1.
Signed-off-by: Marc Orr <marcorr@google.com>
Reviewed-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This ensures that the address and length provided to DBG_DECRYPT and
DBG_ENCRYPT do not cause an overflow.
At the same time, pass the actual number of pages pinned in memory to
sev_unpin_memory() as a cleanup.
Reported-by: Cfir Cohen <cfir@google.com>
Signed-off-by: David Rientjes <rientjes@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Pull mm/compaction fixes from Mel Gorman:
"The merge window for 5.1 introduced a number of compaction-related
patches. with intermittent reports of corruption and functional
issues. The bugs are due to sloopy checking of zone boundaries and a
corner case where invalid indexes are used to access the free lists.
Reports are not common but at least two users and 0-day have tripped
over them. There is a chance that one of the syzbot reports are
related but it has not been confirmed properly.
The normal submission path is with Andrew but there have been some
delays and I consider them urgent enough that they should be picked up
before RC4 to avoid duplicate reports.
All of these have been successfully tested on older RC windows. This
will make this branch look like a rebase but in fact, they've simply
been lifted again from Andrew's tree and placed on a fresh branch.
I've no reason to believe that this has invalidated the testing given
the lack of change in compaction and the nature of the fixes"
* tag 'mm-compaction-5.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/mel/linux:
mm/compaction.c: abort search if isolation fails
mm/compaction.c: correct zone boundary handling when resetting pageblock skip hints
The n_r3964 line discipline driver was written in a different time, when
SMP machines were rare, and users were trusted to do the right thing.
Since then, the world has moved on but not this code, it has stayed
rooted in the past with its lovely hand-crafted list structures and
loads of "interesting" race conditions all over the place.
After attempting to clean up most of the issues, I just gave up and am
now marking the driver as BROKEN so that hopefully someone who has this
hardware will show up out of the woodwork (I know you are out there!)
and will help with debugging a raft of changes that I had laying around
for the code, but was too afraid to commit as odds are they would break
things.
Many thanks to Jann and Linus for pointing out the initial problems in
this codebase, as well as many reviews of my attempts to fix the issues.
It was a case of whack-a-mole, and as you can see, the mole won.
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If a child irqchip calls irq_chip_set_wake_parent() but its parent irqchip
has the IRQCHIP_SKIP_SET_WAKE flag set an error is returned.
This is inconsistent behaviour vs. set_irq_wake_real() which returns 0 when
the irqchip has the IRQCHIP_SKIP_SET_WAKE flag set. It doesn't attempt to
walk the chain of parents and set irq wake on any chips that don't have the
flag set either. If the intent is to call the .irq_set_wake() callback of
the parent irqchip, then we expect irqchip implementations to omit the
IRQCHIP_SKIP_SET_WAKE flag and implement an .irq_set_wake() function that
calls irq_chip_set_wake_parent().
The problem has been observed on a Qualcomm sdm845 device where set wake
fails on any GPIO interrupts after applying work in progress wakeup irq
patches to the GPIO driver. The chain of chips looks like this:
QCOM GPIO -> QCOM PDC (SKIP) -> ARM GIC (SKIP)
The GPIO controllers parent is the QCOM PDC irqchip which in turn has ARM
GIC as parent. The QCOM PDC irqchip has the IRQCHIP_SKIP_SET_WAKE flag
set, and so does the grandparent ARM GIC.
The GPIO driver doesn't know if the parent needs to set wake or not, so it
unconditionally calls irq_chip_set_wake_parent() causing this function to
return a failure because the parent irqchip (PDC) doesn't have the
.irq_set_wake() callback set. Returning 0 instead makes everything work and
irqs from the GPIO controller can be configured for wakeup.
Make it consistent by returning 0 (success) from irq_chip_set_wake_parent()
when a parent chip has IRQCHIP_SKIP_SET_WAKE set.
[ tglx: Massaged changelog ]
Fixes: 08b55e2a92 ("genirq: Add irqchip_set_wake_parent")
Signed-off-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-gpio@vger.kernel.org
Cc: Lina Iyer <ilina@codeaurora.org>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20190325181026.247796-1-swboyd@chromium.org
blk_mq_try_issue_directly() can return BLK_STS*_RESOURCE for requests that
have been queued. If that happens when blk_mq_try_issue_directly() is called
by the dm-mpath driver then dm-mpath will try to resubmit a request that is
already queued and a kernel crash follows. Since it is nontrivial to fix
blk_mq_request_issue_directly(), revert the blk_mq_request_issue_directly()
changes that went into kernel v5.0.
This patch reverts the following commits:
* d6a51a97c0 ("blk-mq: replace and kill blk_mq_request_issue_directly") # v5.0.
* 5b7a6f128a ("blk-mq: issue directly with bypass 'false' in blk_mq_sched_insert_requests") # v5.0.
* 7f556a44e6 ("blk-mq: refactor the code of issue request directly") # v5.0.
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Jianchao Wang <jianchao.w.wang@oracle.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: James Smart <james.smart@broadcom.com>
Cc: Dongli Zhang <dongli.zhang@oracle.com>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: <stable@vger.kernel.org>
Reported-by: Laurence Oberman <loberman@redhat.com>
Tested-by: Laurence Oberman <loberman@redhat.com>
Fixes: 7f556a44e6 ("blk-mq: refactor the code of issue request directly") # v5.0.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
When ioctl calls are made with non-null-terminated userspace strings,
strlcpy causes an OOB-read from within strlen. Fix by changing to use
strscpy instead.
Signed-off-by: Zubin Mithra <zsm@chromium.org>
Reviewed-by: Guenter Roeck <groeck@chromium.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The "call" variable comes from the user in privcmd_ioctl_hypercall().
It's an offset into the hypercall_page[] which has (PAGE_SIZE / 32)
elements. We need to put an upper bound on it to prevent an out of
bounds access.
Cc: stable@vger.kernel.org
Fixes: 1246ae0bb9 ("xen: add variable hypercall caller")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
struct privcmd_buf_vma_private has a zero-sized array at the end
(pages), use the new struct_size() helper to determine the proper
allocation size and avoid potential type mistakes.
Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Pull drm fixes from Dave Airlie:
"Pretty quiet week, just some amdgpu and i915 fixes.
i915:
- deadlock fix
- gvt fixes
amdgpu:
- PCIE dpm feature fix
- Powerplay fixes"
* tag 'drm-fixes-2019-04-05' of git://anongit.freedesktop.org/drm/drm:
drm/i915/gvt: Fix kerneldoc typo for intel_vgpu_emulate_hotplug
drm/i915/gvt: Correct the calculation of plane size
drm/amdgpu: remove unnecessary rlc reset function on gfx9
drm/i915: Always backoff after a drm_modeset_lock() deadlock
drm/i915/gvt: do not let pin count of shadow mm go negative
drm/i915/gvt: do not deliver a workload if its creation fails
drm/amd/display: VBIOS can't be light up HDMI when restart system
drm/amd/powerplay: fix possible hang with 3+ 4K monitors
drm/amd/powerplay: correct data type to avoid overflow
drm/amd/powerplay: add ECC feature bit
drm/amd/amdgpu: fix PCIe dpm feature issue (v3)
Pull networking fixes from David Miller:
1) Several hash table refcount fixes in batman-adv, from Sven
Eckelmann.
2) Use after free in bpf_evict_inode(), from Daniel Borkmann.
3) Fix mdio bus registration in ixgbe, from Ivan Vecera.
4) Unbounded loop in __skb_try_recv_datagram(), from Paolo Abeni.
5) ila rhashtable corruption fix from Herbert Xu.
6) Don't allow upper-devices to be added to vrf devices, from Sabrina
Dubroca.
7) Add qmi_wwan device ID for Olicard 600, from Bjørn Mork.
8) Don't leave skb->next poisoned in __netif_receive_skb_list_ptype,
from Alexander Lobakin.
9) Missing IDR checks in mlx5 driver, from Aditya Pakki.
10) Fix false connection termination in ktls, from Jakub Kicinski.
11) Work around some ASPM issues with r8169 by disabling rx interrupt
coalescing on certain chips. From Heiner Kallweit.
12) Properly use per-cpu qstat values on NOLOCK qdiscs, from Paolo
Abeni.
13) Fully initialize sockaddr_in structures in SCTP, from Xin Long.
14) Various BPF flow dissector fixes from Stanislav Fomichev.
15) Divide by zero in act_sample, from Davide Caratti.
16) Fix bridging multicast regression introduced by rhashtable
conversion, from Nikolay Aleksandrov.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (106 commits)
ibmvnic: Fix completion structure initialization
ipv6: sit: reset ip header pointer in ipip6_rcv
net: bridge: always clear mcast matching struct on reports and leaves
libcxgb: fix incorrect ppmax calculation
vlan: conditional inclusion of FCoE hooks to match netdevice.h and bnx2x
sch_cake: Make sure we can write the IP header before changing DSCP bits
sch_cake: Use tc_skb_protocol() helper for getting packet protocol
tcp: Ensure DCTCP reacts to losses
net/sched: act_sample: fix divide by zero in the traffic path
net: thunderx: fix NULL pointer dereference in nicvf_open/nicvf_stop
net: hns: Fix sparse: some warnings in HNS drivers
net: hns: Fix WARNING when remove HNS driver with SMMU enabled
net: hns: fix ICMP6 neighbor solicitation messages discard problem
net: hns: Fix probabilistic memory overwrite when HNS driver initialized
net: hns: Use NAPI_POLL_WEIGHT for hns driver
net: hns: fix KASAN: use-after-free in hns_nic_net_xmit_hw()
flow_dissector: rst'ify documentation
ipv6: Fix dangling pointer when ipv6 fragment
net-gro: Fix GRO flush when receiving a GSO packet.
flow_dissector: document BPF flow dissector environment
...
Topology is not unloaded in the core during unregister_component()
anymore. So, add the remove() callback that will unload the
topology.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The chips main power supplies VA and VP are enabled during probe but
then never disabled, this will cause warnings from the regulator
framework on driver removal. Fix this by adding a remove callback and
disabling the supplies, whilst doing so follow best practice and put the
chip back into reset as well.
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Use %lu instead of %zu to fix the following warning introduced with
recent memblock refactoring:
xtensa/mm/mmu.c:36:9: warning: format '%zu' expects argument of type
'size_t', but argument 3 has type 'long unsigned int
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Pull RISC-V fixes from Palmer Dabbelt:
"I dropped the ball a bit here: these patches should all probably have
been part of rc2, but I wanted to get around to properly testing them
in the various configurations (qemu32, qeum64, unleashed) first.
Unfortunately I've been traveling and didn't have time to actually do
that, but since these fix concrete bugs and pass my old set of tests I
don't want to delay the fixes any longer.
There are four independent fixes here:
- A fix for the rv32 port that corrects the 64-bit user accesor's
fixup label address.
- A fix for a regression introduced during the merge window that
broke medlow configurations at run time. This patch also includes a
fix that disables ftrace for the same set of functions, which was
found by inspection at the same time.
- A modification of the memory map to avoid overlapping the FIXMAP
and VMALLOC regions on systems with small memory maps.
- A fix to the module handling code to use the correct syntax for
probing Kconfig entries.
These have passed my standard test flow, but I didn't have time to
expand that testing like I said I would"
* tag 'riscv-for-linus-5.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/palmer/riscv-linux:
RISC-V: Use IS_ENABLED(CONFIG_CMODEL_MEDLOW)
RISC-V: Fix FIXMAP_TOP to avoid overlap with VMALLOC area
RISC-V: Always compile mm/init.c with cmodel=medany and notrace
riscv: fix accessing 8-byte variable from RV32
We need to be careful and always zero the whole br_ip struct when it is
used for matching since the rhashtable change. This patch fixes all the
places which didn't properly clear it which in turn might've caused
mismatches.
Thanks for the great bug report with reproducing steps and bisection.
Steps to reproduce (from the bug report):
ip link add br0 type bridge mcast_querier 1
ip link set br0 up
ip link add v2 type veth peer name v3
ip link set v2 master br0
ip link set v2 up
ip link set v3 up
ip addr add 3.0.0.2/24 dev v3
ip netns add test
ip link add v1 type veth peer name v1 netns test
ip link set v1 master br0
ip link set v1 up
ip -n test link set v1 up
ip -n test addr add 3.0.0.1/24 dev v1
# Multicast receiver
ip netns exec test socat
UDP4-RECVFROM:5588,ip-add-membership=224.224.224.224:3.0.0.1,fork -
# Multicast sender
echo hello | nc -u -s 3.0.0.2 224.224.224.224 5588
Reported-by: liam.mcbirnie@boeing.com
Fixes: 19e3a9c90c ("net: bridge: convert multicast to generic rhashtable")
Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull power management fixes from Rafael Wysocki:
"These fix up the intel_pstate driver after recent changes to prevent
it from printing pointless messages and update the turbostat utility
(mostly fixes and new hardware support).
Specifics:
- Make intel_pstate only load on Intel processors and prevent it from
printing pointless failure messages (Borislav Petkov).
- Update the turbostat utility:
* Assorted fixes (Ben Hutchings, Len Brown, Prarit Bhargava).
* Support for AMD Fam 17h (Zen) RAPL and package power (Calvin
Walton).
* Support for Intel Icelake and for systems with more than one die
per package (Len Brown).
* Cleanups (Len Brown)"
* tag 'pm-5.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
cpufreq/intel_pstate: Load only on Intel hardware
tools/power turbostat: update version number
tools/power turbostat: Warn on bad ACPI LPIT data
tools/power turbostat: Add checks for failure of fgets() and fscanf()
tools/power turbostat: Also read package power on AMD F17h (Zen)
tools/power turbostat: Add support for AMD Fam 17h (Zen) RAPL
tools/power turbostat: Do not display an error on systems without a cpufreq driver
tools/power turbostat: Add Die column
tools/power turbostat: Add Icelake support
tools/power turbostat: Cleanup CNL-specific code
tools/power turbostat: Cleanup CC3-skip code
tools/power turbostat: Restore ability to execute in topology-order
Pull ACPI fix from Rafael Wysocki:
"Prevent stale GPE events from triggering spurious system wakeups from
suspend-to-idle (Furquan Shaikh)"
* tag 'acpi-5.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPICA: Clear status of GPEs before enabling them
Pull mfd fixes from Lee Jones:
- Fix failed reads due to enabled IRQs when suspended; twl-core
- Fix driver registration when using DT; sprd-sc27xx-spi
- Fix `make allyesconfig` on x86_64; SUN6I_PRCM
* tag 'mfd-fixes-5.1' of git://git.kernel.org/pub/scm/linux/kernel/git/lee/mfd:
mfd: sun6i-prcm: Allow to compile with COMPILE_TEST
mfd: sc27xx: Use SoC compatible string for PMIC devices
mfd: twl-core: Disable IRQ while suspended
BITS_TO_LONGS() uses DIV_ROUND_UP() because of
this ppmax value can be greater than available
per cpu page pods.
This patch removes BITS_TO_LONGS() to fix this
issue.
Signed-off-by: Varun Prakash <varun@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Way back in 3c9c36bced the
ndo_fcoe_get_wwn pointer was switched from depending on CONFIG_FCOE to
CONFIG_LIBFCOE in order to allow building FCoE support into the bnx2x
driver and used by bnx2fc without including the generic software fcoe
module.
But, FCoE is generally used over an 802.1q VLAN, and the implementation
of ndo_fcoe_get_wwn in the 8021q module was not similarly changed. The
result is that if CONFIG_FCOE is disabled, then bnz2fc cannot make a
call to ndo_fcoe_get_wwn through the 8021q interface to the underlying
bnx2x interface. The bnx2fc driver then falls back to a potentially
different mapping of Ethernet MAC to Fibre Channel WWN, creating an
incompatibility with the fabric and target configurations when compared
to the WWNs used by pre-boot firmware and differently-configured
kernels.
So make the conditional inclusion of FCoE code in 8021q match the
conditional inclusion in netdevice.h
Signed-off-by: Chris Leech <cleech@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
memory allocated by kmem_cache_alloc() should be freed using
kmem_cache_free(), not kfree().
Fixes: fa0ca2aee3 ("deal with get_reqs_available() in aio_get_req() itself")
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
In function do_write_buffer(), in the for loop, there is a case
chip_ready() returns 1 while chip_good() returns 0, so it never
break the loop.
To fix this, chip_good() is enough and it should timeout if it stay
bad for a while.
Fixes: dfeae1073583("mtd: cfi_cmdset_0002: Change write buffer to check correct value")
Signed-off-by: Yi Huaijie <yihuaijie@huawei.com>
Signed-off-by: Liu Jian <liujian56@huawei.com>
Reviewed-by: Tokunori Ikegami <ikegami_to@yahoo.co.jp>
Signed-off-by: Richard Weinberger <richard@nod.at>
Daniel Borkmann says:
====================
pull-request: bpf 2019-04-04
The following pull-request contains BPF updates for your *net* tree.
The main changes are:
1) Batch of fixes to the existing BPF flow dissector API to support
calling BPF programs from the eth_get_headlen context (support for
latter is planned to be added in bpf-next), from Stanislav.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
* pm-tools:
tools/power turbostat: update version number
tools/power turbostat: Warn on bad ACPI LPIT data
tools/power turbostat: Add checks for failure of fgets() and fscanf()
tools/power turbostat: Also read package power on AMD F17h (Zen)
tools/power turbostat: Add support for AMD Fam 17h (Zen) RAPL
tools/power turbostat: Do not display an error on systems without a cpufreq driver
tools/power turbostat: Add Die column
tools/power turbostat: Add Icelake support
tools/power turbostat: Cleanup CNL-specific code
tools/power turbostat: Cleanup CC3-skip code
tools/power turbostat: Restore ability to execute in topology-order
Storage devices which report supporting discard commands like
WRITE_SAME_16 with unmap, but reject discard commands sent to the
storage device. This is a clear storage firmware bug but it doesn't
change the fact that should a program cause discards to be sent to a
multipath device layered on this buggy storage, all paths can end up
failed at the same time from the discards, causing possible I/O loss.
The first discard to a path will fail with Illegal Request, Invalid
field in cdb, e.g.:
kernel: sd 8:0:8:19: [sdfn] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
kernel: sd 8:0:8:19: [sdfn] tag#0 Sense Key : Illegal Request [current]
kernel: sd 8:0:8:19: [sdfn] tag#0 Add. Sense: Invalid field in cdb
kernel: sd 8:0:8:19: [sdfn] tag#0 CDB: Write same(16) 93 08 00 00 00 00 00 a0 08 00 00 00 80 00 00 00
kernel: blk_update_request: critical target error, dev sdfn, sector 10487808
The SCSI layer converts this to the BLK_STS_TARGET error number, the sd
device disables its support for discard on this path, and because of the
BLK_STS_TARGET error multipath fails the discard without failing any
path or retrying down a different path. But subsequent discards can
cause path failures. Any discards sent to the path which already failed
a discard ends up failing with EIO from blk_cloned_rq_check_limits with
an "over max size limit" error since the discard limit was set to 0 by
the sd driver for the path. As the error is EIO, this now fails the
path and multipath tries to send the discard down the next path. This
cycle continues as discards are sent until all paths fail.
Fix this by training DM core to disable DISCARD if the underlying
storage already did so.
Also, fix branching in dm_done() and clone_endio() to reflect the
mutually exclussive nature of the IO operations in question.
Cc: stable@vger.kernel.org
Reported-by: David Jeffery <djeffery@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
return_address returns the address that is one level higher in the call
stack than requested in its argument, because level 0 corresponds to its
caller's return address. Use requested level as the number of stack
frames to skip.
This fixes the address reported by might_sleep and friends.
Cc: stable@vger.kernel.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Toke Høiland-Jørgensen says:
====================
sched: A few small fixes for sch_cake
Kevin noticed a few issues with the way CAKE reads the skb protocol and the IP
diffserv fields. This series fixes those two issues, and should probably go to
in 4.19 as well. However, the previous refactoring patch means they don't apply
as-is; I can send a follow-up directly to stable if that's OK with you?
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
There is not actually any guarantee that the IP headers are valid before we
access the DSCP bits of the packets. Fix this using the same approach taken
in sch_dsmark.
Reported-by: Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We shouldn't be using skb->protocol directly as that will miss cases with
hardware-accelerated VLAN tags. Use the helper instead to get the right
protocol number.
Reported-by: Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
RFC8257 §3.5 explicitly states that "A DCTCP sender MUST react to
loss episodes in the same way as conventional TCP".
Currently, Linux DCTCP performs no cwnd reduction when losses
are encountered. Optionally, the dctcp_clamp_alpha_on_loss resets
alpha to its maximal value if a RTO happens. This behavior
is sub-optimal for at least two reasons: i) it ignores losses
triggering fast retransmissions; and ii) it causes unnecessary large
cwnd reduction in the future if the loss was isolated as it resets
the historical term of DCTCP's alpha EWMA to its maximal value (i.e.,
denoting a total congestion). The second reason has an especially
noticeable effect when using DCTCP in high BDP environments, where
alpha normally stays at low values.
This patch replace the clamping of alpha by setting ssthresh to
half of cwnd for both fast retransmissions and RTOs, at most once
per RTT. Consequently, the dctcp_clamp_alpha_on_loss module parameter
has been removed.
The table below shows experimental results where we measured the
drop probability of a PIE AQM (not applying ECN marks) at a
bottleneck in the presence of a single TCP flow with either the
alpha-clamping option enabled or the cwnd halving proposed by this
patch. Results using reno or cubic are given for comparison.
| Link | RTT | Drop
TCP CC | speed | base+AQM | probability
==================|=========|==========|============
CUBIC | 40Mbps | 7+20ms | 0.21%
RENO | | | 0.19%
DCTCP-CLAMP-ALPHA | | | 25.80%
DCTCP-HALVE-CWND | | | 0.22%
------------------|---------|----------|------------
CUBIC | 100Mbps | 7+20ms | 0.03%
RENO | | | 0.02%
DCTCP-CLAMP-ALPHA | | | 23.30%
DCTCP-HALVE-CWND | | | 0.04%
------------------|---------|----------|------------
CUBIC | 800Mbps | 1+1ms | 0.04%
RENO | | | 0.05%
DCTCP-CLAMP-ALPHA | | | 18.70%
DCTCP-HALVE-CWND | | | 0.06%
We see that, without halving its cwnd for all source of losses,
DCTCP drives the AQM to large drop probabilities in order to keep
the queue length under control (i.e., it repeatedly faces RTOs).
Instead, if DCTCP reacts to all source of losses, it can then be
controlled by the AQM using similar drop levels than cubic or reno.
Signed-off-by: Koen De Schepper <koen.de_schepper@nokia-bell-labs.com>
Signed-off-by: Olivier Tilmans <olivier.tilmans@nokia-bell-labs.com>
Cc: Bob Briscoe <research@bobbriscoe.net>
Cc: Lawrence Brakmo <brakmo@fb.com>
Cc: Florian Westphal <fw@strlen.de>
Cc: Daniel Borkmann <borkmann@iogearbox.net>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Andrew Shewmaker <agshew@gmail.com>
Cc: Glenn Judd <glenn.judd@morganstanley.com>
Acked-by: Florian Westphal <fw@strlen.de>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Yonglong Liu says:
====================
net: hns: bugfixes for HNS Driver
This patchset fix some bugs that were found in the test of
various scenarios, or identify by KASAN/sparse.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
There are some sparse warnings in the HNS drivers:
warning: incorrect type in assignment (different address spaces)
expected void [noderef] <asn:2> *io_base
got void *vaddr
warning: cast removes address space '<asn:2>' of expression
[...]
Add __iomem and change all the u8 __iomem to void __iomem to
fix these kind of warnings.
warning: incorrect type in argument 1 (different address spaces)
expected void [noderef] <asn:2> *base
got unsigned char [usertype] *base_addr
warning: cast to restricted __le16
warning: incorrect type in assignment (different base types)
expected unsigned int [usertype] tbl_tcam_data_high
got restricted __le32 [usertype]
warning: cast to restricted __le32
[...]
These variables used u32/u16 as their type, and finally as a
parameter of writel(), writel() will do the cpu_to_le32 coversion
so remove the little endian covert code to fix these kind of warnings.
Signed-off-by: Yonglong Liu <liuyonglong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
