We need to be better at making sure we don't have a link check
watchdog go off while we're shutting things down, so let's stop
the timer as soon as we start the remove.
Meanwhile, since that was the only thing in
ionic_dev_teardown(), simplify and remove that function.
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
Geetha sowjanya says:
====================
Fix bugs in Octeontx2 netdev driver
In existing Octeontx2 network drivers code has issues
like stale entries in broadcast replication list, missing
L3TYPE for IPv6 frames, running tx queues on error and
race condition in mbox reset.
This patch set fixes the above issues.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Mbox implementation in octeontx2 driver has three states
alloc, send and reset in mbox response. VF allocate and
sends message to PF for processing, PF ACKs them back and
reset the mbox memory. In some case we see synchronization
issue where after msgs_acked is incremented and before
mbox_reset API is called, if current execution is scheduled
out and a different thread is scheduled in which checks for
msgs_acked. Since the new thread sees msgs_acked == msgs_sent
it will try to allocate a new message and to send a new mbox
message to PF.Now if mbox_reset is scheduled in, PF will see
'0' in msgs_send.
This patch fixes the issue by calling mbox_reset before
incrementing msgs_acked flag for last processing message and
checks for valid message size.
Fixes: d424b6c02 ("octeontx2-pf: Enable SRIOV and added VF mbox handling")
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently in otx2_open on failure of nix_lf_start
transmit queues are not stopped which are already
started in link_event. Since the tx queues are not
stopped network stack still try's to send the packets
leading to driver crash while access the device resources.
Fixes: 50fe6c02e ("octeontx2-pf: Register and handle link notifications")
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For TCP/UDP checksum offload feature in Octeontx2
expects L3TYPE to be set irrespective of IP header
checksum is being offloaded or not. Currently for
IPv6 frames L3TYPE is not being set resulting in
packet drop with checksum error. This patch fixes
this issue.
Fixes: 3ca6c4c88 ("octeontx2-pf: Add packet transmission support")
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Packet replication feature present in Octeontx2
is a hardware linked list of PF and its VF
interfaces so that broadcast packets are sent
to all interfaces present in the list. It is
driver job to add and delete a PF/VF interface
to/from the list when the interface is brought
up and down. This patch fixes the
npc_enadis_default_entries function to handle
broadcast replication properly if packet replication
feature is present.
Fixes: 40df309e41 ("octeontx2-af: Support to enable/disable default MCAM entries")
Signed-off-by: Subbaraya Sundeep <sbhatta@marvell.com>
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This reverts commit 7e24bc347e.
7e24bc347e was based on PCIe r5.0, sec 5.9, which claims we need a 200 ms
delay when transitioning to or from D2. However, sec 5.3.1.3 states the
delay as 200 μs (microseconds), as does the table in PCIe r4.0, sec 5.9.1.
This looks like a typo in the r5.0 spec, so revert back to a 200 μs delay
instead of a 200 ms delay.
Fixes: 7e24bc347e ("PCI/PM: Apply D2 delay as milliseconds, not microseconds")
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2020-09-30
This series contains updates to ice driver only.
Jake increases the wait time for firmware response as it can take longer
than the current wait time. Preserves the NVM capabilities of the device in
safe mode so the device reports its NVM update capabilities properly
when in this state.
v2: Added cover letter
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
476e7faefc ("PCI PM: Do not wait for buses in B2 or B3 during resume")
removed the last use of PCI_PM_BUS_WAIT. Remove the definition as well.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
sg_init_table() zeroes its first argument, so the allocation of that
argument doesn't have to.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
// <smpl>
@@
expression x;
@@
x =
- kzalloc
+ kmalloc
(...)
...
sg_init_table(x,...)
// </smpl>
Link: https://lore.kernel.org/r/1600601186-7420-15-git-send-email-Julia.Lawall@inria.fr
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Take care about Coccinelle warnings:
drivers/pci/pci.c:6008:6-12: WARNING: Comparison to bool
drivers/pci/pci.c:6024:7-13: WARNING: Comparison to bool
No change to functionality intended.
Link: https://lore.kernel.org/r/20200925224555.1752460-1-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Jonathan reports that the strict policy for memory mapped by the
ACPI core breaks the use case of passing ACPI table overrides via
initramfs. This is due to the fact that the memory type used for
loading the initramfs in memory is not recognized as a memory type
that is typically used by firmware to pass firmware tables.
Since the purpose of the strict policy is to ensure that no AML or
other ACPI code can manipulate any memory that is used by the kernel
to keep its internal state or the state of user tasks, we can relax
the permission check, and allow mappings of memory that is reserved
and marked as NOMAP via memblock, and therefore not covered by the
linear mapping to begin with.
Fixes: 1583052d11 ("arm64/acpi: disallow AML memory opregions to access kernel memory")
Fixes: 325f5585ec ("arm64/acpi: disallow writeable AML opregion mapping for EFI code regions")
Reported-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Tested-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Sudeep Holla <sudeep.holla@arm.com>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Link: https://lore.kernel.org/r/20200929132522.18067-1-ardb@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Eric Dumazet says:
====================
tcp: exponential backoff in tcp_send_ack()
We had outages caused by repeated skb allocation failures in tcp_send_ack()
It is time to add exponential backoff to reduce number of attempts.
Before doing so, first patch removes icsk_ack.blocked to make
room for a new field (icsk_ack.retry)
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Whenever host is under very high memory pressure,
__tcp_send_ack() skb allocation fails, and we setup
a 200 ms (TCP_DELACK_MAX) timer before retrying.
On hosts with high number of TCP sockets, we can spend
considerable amount of cpu cycles in these attempts,
add high pressure on various spinlocks in mm-layer,
ultimately blocking threads attempting to free space
from making any progress.
This patch adds standard exponential backoff to avoid
adding fuel to the fire.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
TCP has been using it to work around the possibility of tcp_delack_timer()
finding the socket owned by user.
After commit 6f458dfb40 ("tcp: improve latencies of timer triggered events")
we added TCP_DELACK_TIMER_DEFERRED atomic bit for more immediate recovery,
so we can get rid of icsk_ack.blocked
This frees space that following patch will reuse.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
- Make sure DRAM and ChipID region doesn't get disabled on Exynos
- Fix a SATA failure on Tegra
- Fix the emac_ptp clk divider on stratix10
-----BEGIN PGP SIGNATURE-----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=CJSx
-----END PGP SIGNATURE-----
Merge tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux
Pull clk fixes from Stephen Boyd:
"Another batch of clk driver fixes:
- Make sure DRAM and ChipID region doesn't get disabled on Exynos
- Fix a SATA failure on Tegra
- Fix the emac_ptp clk divider on stratix10"
* tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux:
clk: socfpga: stratix10: fix the divider for the emac_ptp_free_clk
clk: samsung: exynos4: mark 'chipid' clock as CLK_IGNORE_UNUSED
clk: tegra: Fix missing prototype for tegra210_clk_register_emc()
clk: tegra: Always program PLL_E when enabled
clk: tegra: Capitalization fixes
clk: samsung: Keep top BPLL mux on Exynos542x enabled
struct macb_platform_data is only used by macb_pci to register the platform
device, move its definition to cadence/macb.h and remove platform_data/macb.h
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Petr Machata says:
====================
mlxsw: PFC and headroom selftests
Recent changes in the headroom management code made it clear that an
automated way of testing this functionality is needed. This patchset brings
two tests: a synthetic headroom behavior test, which verifies mechanics of
headroom management. And a PFC test, which verifies whether this behavior
actually translates into a working lossless configuration.
Both of these tests rely on mlnx_qos[1], a tool that interfaces with Linux
DCB API. The tool was originally written to work with Mellanox NICs, but
does not actually rely on anything Mellanox-specific, and can be used for
mlxsw as well as for any other NIC-like driver. Unlike Open LLDP it does
support buffer commands and permits a fire-and-forget approach to
configuration, which makes it very handy for writing of selftests.
Patches #1-#3 extend the selftest devlink_lib.sh in various ways. Patch #4
then adds a helper wrapper for mlnx_qos to mlxsw's qos_lib.sh.
Patch #5 adds a test for management of port headroom.
Patch #6 adds a PFC test.
[1] https://github.com/Mellanox/mlnx-tools/
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a test for PFC. Runs 10MB of traffic through a bottleneck and checks
that none of it gets lost.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a test for headroom configuration. This covers projection of ETS
configuration to ingress, PFC, adjustments for MTU, the qdisc / TC
mode and the effect of egress SPAN session on buffer configuration.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
mlnx_qos is a script for configuration of DCB. Despite the name it is not
actually Mellanox-specific in any way. It is currently the only ad-hoc tool
available (in contrast to a daemon that manages an interface on an ongoing
basis). However, it is very verbose and parsing out error messages is not
really possible. Add a wrapper that makes it easier to use the tool.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some selftests may not need any actual ports. Technically those are not
forwarding selftests, but devlink_lib can still be handy. Fall back on
NETIF_NO_CABLE in those cases.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper that answers the cell size of the devlink device.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Changing pool type from static to dynamic causes reinterpretation of
threshold values. They therefore need to be saved before pool type is
changed, then the pool type can be changed, and then the new values need
to be set up.
For that reason, set cannot subsume save, because it would be saving the
wrong thing, with possibly a nonsensical value, and restore would then fail
to restore the nonsensical value.
Thus extract a _save() from each of the relevant _set()'s. This way it is
possible to save everything up front, then to tweak it, and then restore in
the required order.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add vivaldi HID driver. This driver allows us to read and report the top
row layout of keyboards which provide a vendor-defined (Google) HID
usage.
Signed-off-by: Sean O'Brien <seobrien@chromium.org>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
CNL+ can report DIMM sizes in .5 GB units. In order to not trauncate
away the .5 GB let's switch to storing the DIMM size in Gb units.
Cc: Swati Sharma <swati2.sharma@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200929131312.12999-1-ville.syrjala@linux.intel.com
One issue was reported at a baremetal environment, which is used for
FPGA verification. "The first transfer will fail for extended ID
format(for both 2.0B and FD format), following frames can be transmitted
and received successfully for extended format, and standard format don't
have this issue. This issue occurred randomly with high possiblity, when
it occurs, the transmitter will detect a BIT1 error, the receiver a CRC
error. According to the spec, a non-correctable error may cause this
transfer failure."
With FLEXCAN_QUIRK_DISABLE_MECR quirk, it supports correctable errors,
disable non-correctable errors interrupt and freeze mode. Platform has
ECC hardware support, but select this quirk, this issue may not come to
light. Initialize all FlexCAN memory before accessing them, at least it
can avoid non-correctable errors detected due to memory uninitialized.
The internal region can't be initialized when the hardware doesn't support
ECC.
According to IMX8MPRM, Rev.C, 04/2020. There is a NOTE at the section
11.8.3.13 Detection and correction of memory errors:
"All FlexCAN memory must be initialized before starting its operation in
order to have the parity bits in memory properly updated. CTRL2[WRMFRZ]
grants write access to all memory positions that require initialization,
ranging from 0x080 to 0xADF and from 0xF28 to 0xFFF when the CAN FD feature
is enabled. The RXMGMASK, RX14MASK, RX15MASK, and RXFGMASK registers need to
be initialized as well. MCR[RFEN] must not be set during memory initialization."
Memory range from 0x080 to 0xADF, there are reserved memory (unimplemented
by hardware, e.g. only configure 64 MBs), these memory can be initialized or not.
In this patch, initialize all flexcan memory which includes reserved memory.
In this patch, create FLEXCAN_QUIRK_SUPPORT_ECC for platforms which has ECC
feature. If you have a ECC platform in your hand, please select this
qurik to initialize all flexcan memory firstly, then you can select
FLEXCAN_QUIRK_DISABLE_MECR to only enable correctable errors.
Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Link: https://lore.kernel.org/r/20200929211557.14153-2-qiangqing.zhang@nxp.com
[mkl: wrap long lines]
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
In [1] Geert noted that the autodetect compatible for the mcp25xxfd driver,
which is "microchip,mcp25xxfd" might be too generic and overlap with upcoming,
but incompatible chips.
In the previous patch the autodetect DT compatbile has been renamed to
"microchip,mcp251xfd", this patch changes all user facing strings from
"mcp25xxfd" to "mcp251xfd" and "MCP25XXFD" to "MCP251XFD", including:
- kconfig symbols
- name of kernel module
- DT and SPI compatible
[1] http://lore.kernel.org/r/CAMuHMdVkwGjr6dJuMyhQNqFoJqbh6Ec5V2b5LenCshwpM2SDsQ@mail.gmail.com
Link: https://lore.kernel.org/r/20200930091424.792165-9-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
arch/mips/kernel/process.c:696:15: error: no previous prototype for 'arch_align_stack' [-Werror=missing-prototypes]
Signed-off-by: Pujin Shi <shipujin.t@gmail.com>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
This commit adds a prototype to fix warning at W=1:
arch/mips/kernel/process.c:95:5: error: no previous prototype for 'arch_dup_task_struct' [-Werror=missing-prototypes]
Signed-off-by: Pujin Shi <shipujin.t@gmail.com>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
If a DM device was suspended when bios were issued to it, those bios
would be deferred using queue_io(). Once the DM device was resumed
dm_process_bio() could be called by dm_wq_work() for original bio that
still needs splitting. dm_process_bio()'s check for current->bio_list
(meaning call chain is within ->submit_bio) as a prerequisite for
calling blk_queue_split() for "abnormal IO" would result in
dm_process_bio() never imposing corresponding queue_limits
(e.g. discard_granularity, discard_max_bytes, etc).
Fix this by always having dm_wq_work() resubmit deferred bios using
submit_bio_noacct().
Side-effect is blk_queue_split() is always called for "abnormal IO" from
->submit_bio, be it from application thread or dm_wq_work() workqueue,
so proper bio splitting and depth-first bio submission is performed.
For sake of clarity, remove current->bio_list check before call to
blk_queue_split().
Also, remove dm_wq_work()'s use of dm_{get,put}_live_table() -- no
longer needed since IO will be reissued in terms of ->submit_bio.
And rename bio variable from 'c' to 'bio'.
Fixes: cf9c378655 ("dm: fix comment in dm_process_bio()")
Reported-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Add test validating that btf_dump works fine with BTFs that are modified and
incrementally generated.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200929232843.1249318-5-andriin@fb.com
Ensure that btf_dump can accommodate new BTF types being appended to BTF
instance after struct btf_dump was created. This came up during attemp to
use btf_dump for raw type dumping in selftests, but given changes are not
excessive, it's good to not have any gotchas in API usage, so I decided to
support such use case in general.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200929232843.1249318-2-andriin@fb.com
compute units that are in use.
[Why]
Allow user to know how many compute units (CU) are in use at any given
moment.
[How]
Surface files in Sysfs that allow user to determine the number of compute
units that are in use for a given process. One Sysfs file is used per
device.
Signed-off-by: Ramesh Errabolu <Ramesh.Errabolu@amd.com>
Reviewed-By: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
waves that are in flight.
[Why]
Allow user to know how many compute units (CU) are in use at any given
moment.
[How]
Read registers of SQ that give number of waves that are in flight
of various queues. Use this information to determine number of CU's
in use.
Signed-off-by: Ramesh Errabolu <Ramesh.Errabolu@amd.com>
Reviewed-By: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Daniel Borkmann says:
====================
This series adds two BPF helpers, that is, one for retrieving the classid
of an skb and another one to redirect via the neigh subsystem, and improves
also the cookie helpers by removing the atomic counter. I've also added
the bpf_tail_call_static() helper to the libbpf API that we've been using
in Cilium for a while now, and last but not least the series adds a few
selftests. For details, please check individual patches, thanks!
v3 -> v4:
- Removed out_rec error path (Martin)
- Integrate BPF_F_NEIGH flag into rejecting invalid flags (Martin)
- I think this way it's better to avoid bit overlaps given it's
right in the place that would need to be extended on new flags
v2 -> v3:
- Removed double skb->dev = dev assignment (David)
- Added headroom check for v6 path (David)
- Set set flowi4_proto for ip_route_output_flow (David)
- Rebased onto latest bpf-next
v1 -> v2:
- Rework cookie generator to support nested contexts (Eric)
- Use ip_neigh_gw6() and container_of() (David)
- Rename __throw_build_bug() and improve comments (Andrii)
- Use bpf_tail_call_static() also in BPF samples (Maciej)
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Port of tail_call_static() helper function from Cilium's BPF code base [0]
to libbpf, so others can easily consume it as well. We've been using this
in production code for some time now. The main idea is that we guarantee
that the kernel's BPF infrastructure and JIT (here: x86_64) can patch the
JITed BPF insns with direct jumps instead of having to fall back to using
expensive retpolines. By using inline asm, we guarantee that the compiler
won't merge the call from different paths with potentially different
content of r2/r3.
We're also using Cilium's __throw_build_bug() macro (here as: __bpf_unreachable())
in different places as a neat trick to trigger compilation errors when
compiler does not remove code at compilation time. This works for the BPF
back end as it does not implement the __builtin_trap().
[0] f5537c2602
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/1656a082e077552eb46642d513b4a6bde9a7dd01.1601477936.git.daniel@iogearbox.net
Add a redirect_neigh() helper as redirect() drop-in replacement
for the xmit side. Main idea for the helper is to be very similar
in semantics to the latter just that the skb gets injected into
the neighboring subsystem in order to let the stack do the work
it knows best anyway to populate the L2 addresses of the packet
and then hand over to dev_queue_xmit() as redirect() does.
This solves two bigger items: i) skbs don't need to go up to the
stack on the host facing veth ingress side for traffic egressing
the container to achieve the same for populating L2 which also
has the huge advantage that ii) the skb->sk won't get orphaned in
ip_rcv_core() when entering the IP routing layer on the host stack.
Given that skb->sk neither gets orphaned when crossing the netns
as per 9c4c325252 ("skbuff: preserve sock reference when scrubbing
the skb.") the helper can then push the skbs directly to the phys
device where FQ scheduler can do its work and TCP stack gets proper
backpressure given we hold on to skb->sk as long as skb is still
residing in queues.
With the helper used in BPF data path to then push the skb to the
phys device, I observed a stable/consistent TCP_STREAM improvement
on veth devices for traffic going container -> host -> host ->
container from ~10Gbps to ~15Gbps for a single stream in my test
environment.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Reviewed-by: David Ahern <dsahern@gmail.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Cc: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/bpf/f207de81629e1724899b73b8112e0013be782d35.1601477936.git.daniel@iogearbox.net
With its use in BPF, the cookie generator can be called very frequently
in particular when used out of cgroup v2 hooks (e.g. connect / sendmsg)
and attached to the root cgroup, for example, when used in v1/v2 mixed
environments. In particular, when there's a high churn on sockets in the
system there can be many parallel requests to the bpf_get_socket_cookie()
and bpf_get_netns_cookie() helpers which then cause contention on the
atomic counter.
As similarly done in f991bd2e14 ("fs: introduce a per-cpu last_ino
allocator"), add a small helper library that both can use for the 64 bit
counters. Given this can be called from different contexts, we also need
to deal with potential nested calls even though in practice they are
considered extremely rare. One idea as suggested by Eric Dumazet was
to use a reverse counter for this situation since we don't expect 64 bit
overflows anyways; that way, we can avoid bigger gaps in the 64 bit
counter space compared to just batch-wise increase. Even on machines
with small number of cores (e.g. 4) the cookie generation shrinks from
min/max/med/avg (ns) of 22/50/40/38.9 down to 10/35/14/17.3 when run
in parallel from multiple CPUs.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Link: https://lore.kernel.org/bpf/8a80b8d27d3c49f9a14e1d5213c19d8be87d1dc8.1601477936.git.daniel@iogearbox.net
Similarly to 5a52ae4e32 ("bpf: Allow to retrieve cgroup v1 classid
from v2 hooks"), add a helper to retrieve cgroup v1 classid solely
based on the skb->sk, so it can be used as key as part of BPF map
lookups out of tc from host ns, in particular given the skb->sk is
retained these days when crossing net ns thanks to 9c4c325252
("skbuff: preserve sock reference when scrubbing the skb."). This
is similar to bpf_skb_cgroup_id() which implements the same for v2.
Kubernetes ecosystem is still operating on v1 however, hence net_cls
needs to be used there until this can be dropped in with the v2
helper of bpf_skb_cgroup_id().
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/ed633cf27a1c620e901c5aa99ebdefb028dce600.1601477936.git.daniel@iogearbox.net
This three thread race can result in the work being run once the callback
becomes NULL:
CPU1 CPU2 CPU3
netevent_callback()
process_one_req() rdma_addr_cancel()
[..]
spin_lock_bh()
set_timeout()
spin_unlock_bh()
spin_lock_bh()
list_del_init(&req->list);
spin_unlock_bh()
req->callback = NULL
spin_lock_bh()
if (!list_empty(&req->list))
// Skipped!
// cancel_delayed_work(&req->work);
spin_unlock_bh()
process_one_req() // again
req->callback() // BOOM
cancel_delayed_work_sync()
The solution is to always cancel the work once it is completed so any
in between set_timeout() does not result in it running again.
Cc: stable@vger.kernel.org
Fixes: 44e75052bc ("RDMA/rdma_cm: Make rdma_addr_cancel into a fence")
Link: https://lore.kernel.org/r/20200930072007.1009692-1-leon@kernel.org
Reported-by: Dan Aloni <dan@kernelim.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Nothing reads this any more, and the reason for its existence has passed
due to the deferred fput() scheme.
Fixes: 8ea1f989aa ("drivers/IB,usnic: reduce scope of mmap_sem")
Link: https://lore.kernel.org/r/0-v1-df64ff042436+42-uctx_closing_jgg@nvidia.com
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Be consistent and use unsigned long throughout the chunk copies to
avoid the inherent clumsiness of mixing integer types of different
widths and signs. Failing to take acount of a wider unsigned type when
using min_t can lead to treating it as a negative, only for it flip back
to a large unsigned value after passing a boundary check.
Fixes: ed13033f02 ("drm/i915/cmdparser: Only cache the dst vmap")
Testcase: igt/gen9_exec_parse/bb-large
Reported-by: "Candelaria, Jared" <jared.candelaria@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: "Candelaria, Jared" <jared.candelaria@intel.com>
Cc: "Bloomfield, Jon" <jon.bloomfield@intel.com>
Cc: <stable@vger.kernel.org> # v4.9+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200928215942.31917-1-chris@chris-wilson.co.uk
(cherry picked from commit b7eeb2b413)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
