| Age | Commit message (Collapse) | Author |
|---|
|
gcc 4.8 reports a warning when initializing with = {0}.
Dropping the "0" from the braces fixes the issue.
This fix is not ANSI compatible but is allowed by gcc.
Signed-off-by: Sameeh Jubran <sameehj@amazon.com>
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add a reserved PCI device ID to the driver's table
Used for internal testing purposes.
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
For an overview of the race created by this patch goto synchronization
label.
In napi busy-poll mode, the kernel invokes the napi handler of the
device repeatedly to poll the NIC's receive queues. This process
repeats until a timeout, specific for each connection, is up.
By polling packets in busy-poll mode the user may gain lower latency
and higher throughput (since the kernel no longer waits for interrupts
to poll the queues) in expense of CPU usage.
Upon completing a napi routine, the driver checks whether
the routine was called by an interrupt handler. If so, the driver
re-enables interrupts for the device. This is needed since an
interrupt routine invocation disables future invocations until
explicitly re-enabled.
The driver avoids re-enabling the interrupts if they were not disabled
in the first place (e.g. if driver in busy mode).
Originally, the driver checked whether interrupt re-enabling is needed
by reading the 'ena_napi->unmask_interrupt' variable. This atomic
variable was set upon interrupt and cleared after re-enabling it.
In the 4.10 Linux version, the 'napi_complete_done' call was changed
so that it returns 'false' when device should not re-enable
interrupts, and 'true' otherwise. The change includes reading the
"NAPIF_STATE_IN_BUSY_POLL" flag to check if the napi call is in
busy-poll mode, and if so, return 'false'.
The driver was changed to re-enable interrupts according to this
routine's return value.
The Linux community rejected the use of the
'ena_napi->unmaunmask_interrupt' variable to determine whether
unmasking is needed, and urged to use napi_napi_complete_done()
return value solely.
See https://lore.kernel.org/patchwork/patch/741149/ for more details
As explained, a busy-poll session exists for a specified timeout
value, after which it exits the busy-poll mode and re-enters it later.
This leads to many invocations of the napi handler where
napi_complete_done() false indicates that interrupts should be
re-enabled.
This creates a bug in which the interrupts are re-enabled
unnecessarily.
To reproduce this bug:
1) echo 50 | sudo tee /proc/sys/net/core/busy_poll
2) echo 50 | sudo tee /proc/sys/net/core/busy_read
3) Add counters that check whether
'ena_unmask_interrupt(tx_ring, rx_ring);'
is called without disabling the interrupts in the first
place (i.e. with calling the interrupt routine
ena_intr_msix_io())
Steps 1+2 enable busy-poll as the default mode for new connections.
The busy poll routine rearms the interrupts after every session by
design, and so we need to add an extra check that the interrupts were
masked in the first place.
synchronization:
This patch introduces a race between the interrupt handler
ena_intr_msix_io() and the napi routine ena_io_poll().
Some macros and instruction were added to prevent this race from leaving
the interrupts masked. The following specifies the different race
scenarios in this patch:
1) interrupt handler and napi routine run sequentially
i) interrupt handler is called, sets 'interrupts_masked' flag and
successfully schedules the napi handler via softirq.
In this scenario the napi routine might not see the flag change
for several reasons:
a) The flag is stored in a register by the compiler. For this
case the WRITE_ONCE macro which prevents this.
b) The compiler might reorder the instruction. For this the
smp_wmb() instruction was used which implies a compiler memory
barrier.
c) On archs with weak consistency model (like ARM64) the napi
routine might be scheduled and start running before the flag
STORE instruction is committed to cache/memory. To ensure this
doesn't happen, the smp_wmb() instruction was added. It ensures
that the flag set instruction is committed before scheduling
napi.
ii) compiler reorders the flag's value check in the 'if' with
the flag set in the napi routine.
This scenario is prevented by smp_rmb() call after the flag check.
2) interrupt handler and napi routine run in parallel (can happen when
busy poll routine invokes the napi handler)
i) interrupt handler sets the flag in one core, while the napi
routine reads it in another core.
This scenario also is divided into two cases:
a) napi_complete_done() doesn't finish running, in which case
napi_sched() would just set NAPIF_STATE_MISSED and the napi
routine would reschedule itself without changing the flag's value.
b) napi_complete_done() finishes running. In this case the
napi routine might override the flag's value.
This doesn't present any rise since it later unmasks the
interrupt vector.
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
- Free ILT lines used for XRC-SRQ's contexts.
- Free XRCD bitmap
Fixes: b8204ad878ce7 ("qed: changes to ILT to support XRC")
Fixes: 7bfb399eca460 ("qed: Add XRC to RoCE")
Signed-off-by: Michal Kalderon <mkalderon@marvell.com>
Signed-off-by: Yuval Basson <ybason@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Russell King says:
====================
Phylink PCS updates
This series updates the rudimentary phylink PCS support with the
results of the last four months of development of that. Phylink
PCS support was initially added back at the end of March, when it
became clear that the current approach of treating everything at
the MAC end as being part of the MAC was inadequate.
However, this rudimentary implementation was fine initially for
mvneta and similar, but in practice had a fair number of issues,
particularly when ethtool interfaces were used to change various
link properties.
It became apparent that relying on the phylink_config structure for
the PCS was also bad when it became clear that the same PCS was used
in DSA drivers as well as in NXPs other offerings, and there was a
desire to re-use that code.
It also became apparent that splitting the "configuration" step on
an interface mode configuration between the MAC and PCS using just
mac_config() and pcs_config() methods was not sufficient for some
setups, as the MAC needed to be "taken down" prior to making changes,
and once all settings were complete, the MAC could only then be
resumed.
This series addresses these points, progressing PCS support, and
has been developed with mvneta and DPAA2 setups, with work on both
those drivers to prove this approach. It has been rigorously tested
with mvneta, as that provides the most flexibility for testing the
various code paths.
To solve the phylink_config reuse problem, we introduce a struct
phylink_pcs, which contains the minimal information necessary, and it
is intended that this is embedded in the PCS private data structure.
To solve the interface mode configuration problem, we introduce two
new MAC methods, mac_prepare() and mac_finish() which wrap the entire
interface mode configuration only. This has the additional benefit of
relieving MAC drivers from working out whether an interface change has
occurred, and whether they need to do some major work.
I have not yet updated all the interface documentation for these
changes yet, that work remains, but this patch set is provided in the
hope that those working on PCS support in NXP will find this useful.
Since there is a lot of change here, this is the reason why I strongly
advise that everyone has converted to the mac_link_up() way of
configuring the link parameters when the link comes up, rather than
the old way of using mac_config() - especially as splitting the PCS
changes how and when phylink calls mac_config(). Although no change
for existing users is intended, that is something I no longer am able
to test.
Changes since RFC:
- fix bisect build failure
- add patch to use config.an_enabled
- rename phylink_config_interface to phylink_major_reconfig
- add expanded documentation for phylink_set_pcs()
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add an interface to configure the advertisement for a clause 22 PCS
PHY, and set the AN enable flag in the BMCR appropriately.
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add a way for MAC PCS to have private data while keeping independence
from struct phylink_config, which is used for the MAC itself. We need
this independence as we will have stand-alone code for PCS that is
independent of the MAC. Introduce struct phylink_pcs, which is
designed to be embedded in a driver private data structure.
This structure does not include a mdio_device as there are PCS
implementations such as the Marvell DSA and network drivers where this
is not necessary.
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
With PCS support, how we implement interface reconfiguration (or other
major reconfiguration) is not up to the job; we end up reconfiguring
the PCS for an interface change while the link could potentially be up.
In order to solve this, add two additional MAC methods for major
configuration, one to prepare for the change, and one to finish the
change.
This allows mvneta and mvpp2 to shutdown what they require prior to the
MAC and PCS configuration calls, and then restart as appropriate.
This impacts ksettings_set(), which now needs to identify whether the
change is a minor tweak to the advertisement masks or whether the
interface mode has changed, and call the appropriate function for that
update.
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Re-code the pause in-band advertisement update in light of the addition
of PCS support, so that we perform the minimum required; only the PCS
configuration function needs to be called in this case, followed by the
request to trigger a restart of negotiation if the programmed
advertisement changed.
We need to change the pcs_config() signature to pass whether resolved
pause should be passed to the MAC for setups such as mvneta and mvpp2
where doing so overrides the MAC manual flow controls.
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
For fixed links, we only allow the current settings, so this should be
a matter of merely rejecting an attempt to change the settings. If the
settings agree, then there is nothing more we need to do.
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Rather than recomputing whether AN is enabled, use config.an_enabled.
Suggested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When we have a PHY attached, an ethtool ksettings_set() call only
really needs to call through to the phylib equivalent; phylib will
call back to us when the link changes so we can update our state.
Therefore, we can bypass most of our ksettings_set() call for this
case.
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Simplify the ksettings_set() implementation to look more like phylib's
implementation; use a switch() for validating the autoneg setting, and
use the linkmode_modify() helper to set the autoneg bit in the
advertisement mask.
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Avoid calling mac_config() when using split PCS, and the interface
remains the same.
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The only PHYs that are used with phylink which change their interface
are the BCM84881 and MV88X3310 family, both of which only change their
interface modes on link-up events. This will break when drivers are
converted to split-PCS. Fix this.
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The only PHYs that are used with phylink which change their interface
are the BCM84881 and MV88X3310 family, both of which only change their
interface modes on link-up events. However, rather than relying upon
this behaviour by the PHY, we should give a stronger guarantee when
resolving that the link will be down whenever we change the interface
mode. This patch implements that stronger guarantee for resolve.
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use a boolean to indicate whether mac_config() should be called during
a resolution. This allows resolution to have a single location where
mac_config() will be called, which will allow us to make decisions
about how and what we do.
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Rejig the link state tracking, so that we can use the current state
in a future patch.
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Comparing the ethtool output from phylink and non-phylink fixed-link
setups shows that we have some differences:
- The "auto-negotiation" fields are different; phylink reports these
as "No", non-phylink reports these as "Yes" for the supported and
advertising masks.
- The link partner advertisement is set to the link speed with non-
phylink, but phylink leaves this unset, causing all link partner
fields to be omitted.
The phylink ethtool output also disagrees with the software emulated
PHY dump via the MII registers.
Update the phylink fixed-link parsing code so that we better reflect
the behaviour of the non-phylink code that this facility replaces, and
bring the ethtool interface more into line with the report from via the
MII interface.
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Claudiu Manoil says:
====================
enetc: Add adaptive interrupt coalescing
Apart from some related cleanup patches, this set
introduces in a straightforward way the support needed
to enable and configure interrupt coalescing for ENETC.
Patch 5 introduces the support needed for configuring the
interrupt coalescing parameters and for switching between
moderated (int. coalescing) and per-packet interrupt modes.
When interrupt coalescing is enabled the Rx/Tx time
thresholds are configurable, packet thresholds are fixed.
To make this work reliably, patch 5 uses the traffic
pause procedure introduced in patch 2.
Patch 6 adds DIM (Dynamic Interrupt Moderation) to implement
adaptive coalescing based on time thresholds, for the Rx 'channel'.
On the Tx side a default optimal value is used instead, optimized for
TCP traffic over 1G and 2.5G links. This default 'optimal' value can
be overridden anytime via 'ethtool -C tx-usecs'.
netperf -t TCP_MAERTS measurements show a significant CPU load
reduction correlated w/ reduced interrupt rates. For the
measurement results refer to the comments in patch 6.
v2: Replaced Tx DIM with predefined optimal value, giving
better results. This was also suggested by Jakub (cc).
Switched order of patches 4 and 5, for better grouping.
v3: minor cleanup/improvements
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use the generic dynamic interrupt moderation (dim)
framework to implement adaptive interrupt coalescing
on Rx. With the per-packet interrupt scheme, a high
interrupt rate has been noted for moderate traffic flows
leading to high CPU utilization. The 'dim' scheme
implemented by the current patch addresses this issue
improving CPU utilization while using minimal coalescing
time thresholds in order to preserve a good latency.
On the Tx side use an optimal time threshold value by
default. This value has been optimized for Tx TCP
streams at a rate of around 85kpps on a 1G link,
at which rate half of the Tx ring size (128) gets filled
in 1500 usecs. Scaling this down to 2.5G links yields
the current value of 600 usecs, which is conservative
and gives good enough results for 1G links too (see
next).
Below are some measurement results for before and after
this patch (and related dependencies) basically, for a
2 ARM Cortex-A72 @1.3Ghz CPUs system (32 KB L1 data cache),
using 60secs log netperf TCP stream tests @ 1Gbit link
(maximum throughput):
1) 1 Rx TCP flow, both Rx and Tx processed by the same NAPI
thread on the same CPU:
CPU utilization int rate (ints/sec)
Before: 50%-60% (over 50%) 92k
After: 13%-22% 3.5k-12k
Comment: Major CPU utilization improvement for a single flow
Rx TCP flow (i.e. netperf -t TCP_MAERTS) on a single
CPU. Usually settles under 16% for longer tests.
2) 4 Rx TCP flows + 4 Tx TCP flows (+ pings to check the latency):
Total CPU utilization Total int rate (ints/sec)
Before: ~80% (spikes to 90%) ~100k
After: 60% (more steady) ~4k
Comment: Important improvement for this load test, while the
ping test outcome does not show any notable
difference compared to before.
Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Enable programming of the interrupt coalescing registers
and allow manual configuration of the coalescing time
thresholds via ethtool. Packet thresholds have been fixed
to predetermined values as there's no point in making them
run-time configurable, also anticipating the dynamic interrupt
moderation (DIM) algorithm which uses fixed packet thresholds
as well. If the interface is up when the operation mode of
traffic interrupt events is changed by the user (i.e. switching
from default per-packet interrupts to coalesced interrupts),
the traffic needs to be paused in the process.
This patch also prepares the ground for introducing DIM on Rx.
Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
'struct enetc_bdr' is already '____cacheline_aligned_in_smp'.
Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Interrupt coalescing registers naming in the current revision
of the Ref Man (RM) is ICR, deprecating the ICIR name used
in earlier (draft) versions of the RM.
Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
A reliable traffic pause (and reconfiguration) procedure
is needed to be able to safely make h/w configuration
changes during run-time, like changing the mode in which the
interrupts are operating (i.e. with or without coalescing),
as opposed to making on-the-fly register updates that
may be subject to h/w or s/w concurrency issues.
To this end, the code responsible of the run-time device
configurations that basically starts resp. stops the traffic
flow through the device has been extracted from the
the enetc_open/_close procedures, to the separate standalone
enetc_start/_stop procedures. Traffic stop should be as
graceful as possible, it lets the executing napi threads to
to finish while the interrupts stay disabled. But since
the napi thread will try to re-enable interrupts by clearing
the device's unmask register, the enable_irq/ disable_irq
API has been used to avoid this potential concurrency issue
and make the traffic pause procedure more reliable.
Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
It's time to differentiate between Rx and Tx ring sizes.
Not only Tx rings are processed differently than Rx rings,
but their default number also differs - i.e. up to 8 Tx rings
per device (8 traffic classes) vs. 2 Rx rings (one per CPU).
So let's set Tx rings sizes to half the size of the Rx rings
for now, to be conservative.
The default ring sizes were decreased as well (to the next
lower power of 2), to reduce the memory footprint, buffering
etc., since the measurements I've made so far show that the
rings are very unlikely to get full.
This change also anticipates the introduction of the
dynamic interrupt moderation (dim) algorithm which operates
on maximum packet thresholds of 256 packets for Rx and 128
packets for Tx.
Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use devm_gpiod_get_array() to simplify the error handling and exit
code path.
Signed-off-by: Jisheng Zhang <Jisheng.Zhang@synaptics.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The bpftool sources include code to walk file trees, but use multiple
frameworks to do so: nftw and fts. While nftw conforms to POSIX/SUSv3 and
is widely available, fts is not conformant and less common, especially on
non-glibc systems. The inconsistent framework usage hampers maintenance
and portability of bpftool, in particular for embedded systems.
Standardize code usage by rewriting one fts-based function to use nftw and
clean up some related function warnings by extending use of "const char *"
arguments. This change helps in building bpftool against musl for OpenWrt.
Also fix an unsafe call to dirname() by duplicating the string to pass,
since some implementations may directly alter it. The same approach is
used in libbpf.c.
Signed-off-by: Tony Ambardar <Tony.Ambardar@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Quentin Monnet <quentin@isovalent.com>
Link: https://lore.kernel.org/bpf/20200721024817.13701-1-Tony.Ambardar@gmail.com
|
|
Yonghong Song says:
====================
Commit 5a2798ab32ba
("bpf: Add BTF_ID_LIST/BTF_ID/BTF_ID_UNUSED macros")
implemented a mechanism to compute btf_ids at kernel build
time which can simplify kernel implementation and reduce
runtime overhead by removing in-kernel btf_id calculation.
This patch set tried to use this mechanism to compute
btf_ids for bpf_skc_to_*() helpers and for btf_id_or_null ctx
arguments specified during bpf iterator registration.
Please see individual patch for details.
Changelogs:
v1 -> v2:
- v1 ([1]) is only for bpf_skc_to_*() helpers. This version
expanded it to cover ctx btf_id_or_null arguments
- abandoned the change of "extern u32 name[]" to
"static u32 name[]" for BPF_ID_LIST local "name" definition.
gcc 9 incurred a compilation error.
[1]: https://lore.kernel.org/bpf/20200717184706.3476992-1-yhs@fb.com/T
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
A handful of samples and selftests fail to build on s390, because
after commit 0ebeea8ca8a4 ("bpf: Restrict bpf_probe_read{, str}()
only to archs where they work") bpf_probe_read is not available
anymore.
Fix by using bpf_probe_read_kernel.
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200720114806.88823-1-iii@linux.ibm.com
|
|
One additional field btf_id is added to struct
bpf_ctx_arg_aux to store the precomputed btf_ids.
The btf_id is computed at build time with
BTF_ID_LIST or BTF_ID_LIST_GLOBAL macro definitions.
All existing bpf iterators are changed to used
pre-compute btf_ids.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200720163403.1393551-1-yhs@fb.com
|
|
OpenBSD netcat (Debian patchlevel 1.195-2) does not seem to react to
SIGINT for whatever reason, causing prefix.pl to hang after
test_lwt_seg6local.sh exits due to netcat inheriting
test_lwt_seg6local.sh's file descriptors.
Fix by using SIGTERM instead.
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200720101810.84299-1-iii@linux.ibm.com
|
|
tcp and udp bpf_iter can reuse some socket ids in
btf_sock_ids, so make it global.
I put the extern definition in btf_ids.h as a central
place so it can be easily discovered by developers.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200720163402.1393427-1-yhs@fb.com
|
|
Luke Nelson says:
====================
his patch series enables using compressed riscv (RVC) instructions
in the rv64 BPF JIT.
RVC is a standard riscv extension that adds a set of compressed,
2-byte instructions that can replace some regular 4-byte instructions
for improved code density.
This series first modifies the JIT to support using 2-byte instructions
(e.g., in jump offset computations), then adds RVC encoding and
helper functions, and finally uses the helper functions to optimize
the rv64 JIT.
I used our formal verification framework, Serval, to verify the
correctness of the RVC encodings and their uses in the rv64 JIT.
The JIT continues to pass all tests in lib/test_bpf.c, and introduces
no new failures to test_verifier; both with and without RVC being enabled.
The following are examples of the JITed code for the verifier selftest
"direct packet read test#3 for CGROUP_SKB OK", without and with RVC
enabled, respectively. The former uses 178 bytes, and the latter uses 112,
for a ~37% reduction in code size for this example.
Without RVC:
0: 02000813 addi a6,zero,32
4: fd010113 addi sp,sp,-48
8: 02813423 sd s0,40(sp)
c: 02913023 sd s1,32(sp)
10: 01213c23 sd s2,24(sp)
14: 01313823 sd s3,16(sp)
18: 01413423 sd s4,8(sp)
1c: 03010413 addi s0,sp,48
20: 03056683 lwu a3,48(a0)
24: 02069693 slli a3,a3,0x20
28: 0206d693 srli a3,a3,0x20
2c: 03456703 lwu a4,52(a0)
30: 02071713 slli a4,a4,0x20
34: 02075713 srli a4,a4,0x20
38: 03856483 lwu s1,56(a0)
3c: 02049493 slli s1,s1,0x20
40: 0204d493 srli s1,s1,0x20
44: 03c56903 lwu s2,60(a0)
48: 02091913 slli s2,s2,0x20
4c: 02095913 srli s2,s2,0x20
50: 04056983 lwu s3,64(a0)
54: 02099993 slli s3,s3,0x20
58: 0209d993 srli s3,s3,0x20
5c: 09056a03 lwu s4,144(a0)
60: 020a1a13 slli s4,s4,0x20
64: 020a5a13 srli s4,s4,0x20
68: 00900313 addi t1,zero,9
6c: 006a7463 bgeu s4,t1,0x74
70: 00000a13 addi s4,zero,0
74: 02d52823 sw a3,48(a0)
78: 02e52a23 sw a4,52(a0)
7c: 02952c23 sw s1,56(a0)
80: 03252e23 sw s2,60(a0)
84: 05352023 sw s3,64(a0)
88: 00000793 addi a5,zero,0
8c: 02813403 ld s0,40(sp)
90: 02013483 ld s1,32(sp)
94: 01813903 ld s2,24(sp)
98: 01013983 ld s3,16(sp)
9c: 00813a03 ld s4,8(sp)
a0: 03010113 addi sp,sp,48
a4: 00078513 addi a0,a5,0
a8: 00008067 jalr zero,0(ra)
With RVC:
0: 02000813 addi a6,zero,32
4: 7179 c.addi16sp sp,-48
6: f422 c.sdsp s0,40(sp)
8: f026 c.sdsp s1,32(sp)
a: ec4a c.sdsp s2,24(sp)
c: e84e c.sdsp s3,16(sp)
e: e452 c.sdsp s4,8(sp)
10: 1800 c.addi4spn s0,sp,48
12: 03056683 lwu a3,48(a0)
16: 1682 c.slli a3,0x20
18: 9281 c.srli a3,0x20
1a: 03456703 lwu a4,52(a0)
1e: 1702 c.slli a4,0x20
20: 9301 c.srli a4,0x20
22: 03856483 lwu s1,56(a0)
26: 1482 c.slli s1,0x20
28: 9081 c.srli s1,0x20
2a: 03c56903 lwu s2,60(a0)
2e: 1902 c.slli s2,0x20
30: 02095913 srli s2,s2,0x20
34: 04056983 lwu s3,64(a0)
38: 1982 c.slli s3,0x20
3a: 0209d993 srli s3,s3,0x20
3e: 09056a03 lwu s4,144(a0)
42: 1a02 c.slli s4,0x20
44: 020a5a13 srli s4,s4,0x20
48: 4325 c.li t1,9
4a: 006a7363 bgeu s4,t1,0x50
4e: 4a01 c.li s4,0
50: d914 c.sw a3,48(a0)
52: d958 c.sw a4,52(a0)
54: dd04 c.sw s1,56(a0)
56: 03252e23 sw s2,60(a0)
5a: 05352023 sw s3,64(a0)
5e: 4781 c.li a5,0
60: 7422 c.ldsp s0,40(sp)
62: 7482 c.ldsp s1,32(sp)
64: 6962 c.ldsp s2,24(sp)
66: 69c2 c.ldsp s3,16(sp)
68: 6a22 c.ldsp s4,8(sp)
6a: 6145 c.addi16sp sp,48
6c: 853e c.mv a0,a5
6e: 8082 c.jr ra
RFC -> v1:
- From Björn Töpel:
* Changed RVOFF macro to static inline "ninsns_rvoff".
* Changed return type of rvc_ functions from u32 to u16.
* Changed sizeof(u16) to sizeof(*ctx->insns).
* Factored unsigned immediate checks into helper functions
(is_8b_uint, etc.)
* Changed to use IS_ENABLED instead of #ifdef to check if RVC is
enabled.
* Changed type of immediate arguments to rvc_* encoding to u32
to avoid issues from promotion of u16 to signed int.
* Cleaned up RVC checks in emit_{addi,slli,srli,srai}.
+ Wrapped lines at 100 instead of 80 columns for increased clarity.
+ Move !imm checks into each branch instead of checking
separately.
+ Strengthed checks for c.{slli,srli,srai} to check that
imm < XLEN. Otherwise, imm could be non-zero but the lower
XLEN bits could all be zero, leading to invalid RVC encoding.
* Changed emit_imm to sign-extend the 12-bit value in "lower"
+ The immediate checks for emit_{addiw,li,addi} use signed
comparisons, so this enables the RVC variants to be used
more often (e.g., if val == -1, then lower should be -1
as opposed to 4095).
====================
Reviewed-by: Björn Töpel <bjorn.topel@gmail.com>
Acked-by: Björn Töpel <bjorn.topel@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Existing BTF_ID_LIST used a local static variable
to store btf_ids. This patch provided a new macro
BTF_ID_LIST_GLOBAL to store btf_ids in a global
variable which can be shared among multiple files.
The existing BTF_ID_LIST is still retained.
Two reasons. First, BTF_ID_LIST is also used to build
btf_ids for helper arguments which typically
is an array of 5. Since typically different
helpers have different signature, it makes
little sense to share them. Second, some
current computed btf_ids are indeed local.
If later those btf_ids are shared between
different files, they can use BTF_ID_LIST_GLOBAL then.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Link: https://lore.kernel.org/bpf/20200720163401.1393159-1-yhs@fb.com
|
|
Sync kernel header btf_ids.h to tools directory.
Also define macro CONFIG_DEBUG_INFO_BTF before
including btf_ids.h in prog_tests/resolve_btfids.c
since non-stub definitions for BTF_ID_LIST etc. macros
are defined under CONFIG_DEBUG_INFO_BTF. This
prevented test_progs from failing.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200720163359.1393079-1-yhs@fb.com
|
|
Currently, socket types (struct tcp_sock, udp_sock, etc.)
used by bpf_skc_to_*() helpers are computed when vmlinux_btf
is first built in the kernel.
Commit 5a2798ab32ba
("bpf: Add BTF_ID_LIST/BTF_ID/BTF_ID_UNUSED macros")
implemented a mechanism to compute btf_ids at kernel build
time which can simplify kernel implementation and reduce
runtime overhead by removing in-kernel btf_id calculation.
This patch did exactly this, removing in-kernel btf_id
computation and utilizing build-time btf_id computation.
If CONFIG_DEBUG_INFO_BTF is not defined, BTF_ID_LIST will
define an array with size of 5, which is not enough for
btf_sock_ids. So define its own static array if
CONFIG_DEBUG_INFO_BTF is not defined.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200720163358.1393023-1-yhs@fb.com
|
|
Fix pass 0 to PTR_ERR, also dump more err info using
libbpf_strerror.
Fixes: 5dc7a8b21144 ("bpftool, selftests/bpf: Embed object file inside skeleton")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Reviewed-by: Quentin Monnet <quentin@isovalent.com>
Link: https://lore.kernel.org/bpf/20200717123059.29624-1-yuehaibing@huawei.com
|
|
This patch uses the RVC support and encodings from bpf_jit.h to optimize
the rv64 jit.
The optimizations work by replacing emit(rv_X(...)) with a call to a
helper function emit_X, which will emit a compressed version of the
instruction when possible, and when RVC is enabled.
The JIT continues to pass all tests in lib/test_bpf.c, and introduces
no new failures to test_verifier; both with and without RVC being enabled.
Most changes are straightforward replacements of emit(rv_X(...), ctx)
with emit_X(..., ctx), with the following exceptions bearing mention;
* Change emit_imm to sign-extend the value in "lower", since the
checks for RVC (and the instructions themselves) treat the value as
signed. Otherwise, small negative immediates will not be recognized as
encodable using an RVC instruction. For example, without this change,
emit_imm(rd, -1, ctx) would cause lower to become 4095, which is not a
6b int even though a "c.li rd, -1" instruction suffices.
* For {BPF_MOV,BPF_ADD} BPF_X, drop using addiw,addw in the 32-bit
cases since the values are zero-extended into the upper 32 bits in
the following instructions anyways, and the addition commutes with
zero-extension. (BPF_SUB BPF_X must still use subw since subtraction
does not commute with zero-extension.)
This patch avoids optimizing branches and jumps to use RVC instructions
since surrounding code often makes assumptions about the sizes of
emitted instructions. Optimizing these will require changing these
functions (e.g., emit_branch) to dynamically compute jump offsets.
The following are examples of the JITed code for the verifier selftest
"direct packet read test#3 for CGROUP_SKB OK", without and with RVC
enabled, respectively. The former uses 178 bytes, and the latter uses 112,
for a ~37% reduction in code size for this example.
Without RVC:
0: 02000813 addi a6,zero,32
4: fd010113 addi sp,sp,-48
8: 02813423 sd s0,40(sp)
c: 02913023 sd s1,32(sp)
10: 01213c23 sd s2,24(sp)
14: 01313823 sd s3,16(sp)
18: 01413423 sd s4,8(sp)
1c: 03010413 addi s0,sp,48
20: 03056683 lwu a3,48(a0)
24: 02069693 slli a3,a3,0x20
28: 0206d693 srli a3,a3,0x20
2c: 03456703 lwu a4,52(a0)
30: 02071713 slli a4,a4,0x20
34: 02075713 srli a4,a4,0x20
38: 03856483 lwu s1,56(a0)
3c: 02049493 slli s1,s1,0x20
40: 0204d493 srli s1,s1,0x20
44: 03c56903 lwu s2,60(a0)
48: 02091913 slli s2,s2,0x20
4c: 02095913 srli s2,s2,0x20
50: 04056983 lwu s3,64(a0)
54: 02099993 slli s3,s3,0x20
58: 0209d993 srli s3,s3,0x20
5c: 09056a03 lwu s4,144(a0)
60: 020a1a13 slli s4,s4,0x20
64: 020a5a13 srli s4,s4,0x20
68: 00900313 addi t1,zero,9
6c: 006a7463 bgeu s4,t1,0x74
70: 00000a13 addi s4,zero,0
74: 02d52823 sw a3,48(a0)
78: 02e52a23 sw a4,52(a0)
7c: 02952c23 sw s1,56(a0)
80: 03252e23 sw s2,60(a0)
84: 05352023 sw s3,64(a0)
88: 00000793 addi a5,zero,0
8c: 02813403 ld s0,40(sp)
90: 02013483 ld s1,32(sp)
94: 01813903 ld s2,24(sp)
98: 01013983 ld s3,16(sp)
9c: 00813a03 ld s4,8(sp)
a0: 03010113 addi sp,sp,48
a4: 00078513 addi a0,a5,0
a8: 00008067 jalr zero,0(ra)
With RVC:
0: 02000813 addi a6,zero,32
4: 7179 c.addi16sp sp,-48
6: f422 c.sdsp s0,40(sp)
8: f026 c.sdsp s1,32(sp)
a: ec4a c.sdsp s2,24(sp)
c: e84e c.sdsp s3,16(sp)
e: e452 c.sdsp s4,8(sp)
10: 1800 c.addi4spn s0,sp,48
12: 03056683 lwu a3,48(a0)
16: 1682 c.slli a3,0x20
18: 9281 c.srli a3,0x20
1a: 03456703 lwu a4,52(a0)
1e: 1702 c.slli a4,0x20
20: 9301 c.srli a4,0x20
22: 03856483 lwu s1,56(a0)
26: 1482 c.slli s1,0x20
28: 9081 c.srli s1,0x20
2a: 03c56903 lwu s2,60(a0)
2e: 1902 c.slli s2,0x20
30: 02095913 srli s2,s2,0x20
34: 04056983 lwu s3,64(a0)
38: 1982 c.slli s3,0x20
3a: 0209d993 srli s3,s3,0x20
3e: 09056a03 lwu s4,144(a0)
42: 1a02 c.slli s4,0x20
44: 020a5a13 srli s4,s4,0x20
48: 4325 c.li t1,9
4a: 006a7363 bgeu s4,t1,0x50
4e: 4a01 c.li s4,0
50: d914 c.sw a3,48(a0)
52: d958 c.sw a4,52(a0)
54: dd04 c.sw s1,56(a0)
56: 03252e23 sw s2,60(a0)
5a: 05352023 sw s3,64(a0)
5e: 4781 c.li a5,0
60: 7422 c.ldsp s0,40(sp)
62: 7482 c.ldsp s1,32(sp)
64: 6962 c.ldsp s2,24(sp)
66: 69c2 c.ldsp s3,16(sp)
68: 6a22 c.ldsp s4,8(sp)
6a: 6145 c.addi16sp sp,48
6c: 853e c.mv a0,a5
6e: 8082 c.jr ra
Signed-off-by: Luke Nelson <luke.r.nels@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Cc: Björn Töpel <bjorn.topel@gmail.com>
Link: https://lore.kernel.org/bpf/20200721025241.8077-4-luke.r.nels@gmail.com
|
|
The non-builtin route for offsetof has a dependency on size_t from
stdlib.h/stdint.h that is undeclared and may break targets.
The offsetof macro in bpf_helpers may disable the same macro in other
headers that have a #ifdef offsetof guard. Rather than add additional
dependencies improve the offsetof macro declared here to use the
builtin that is available since llvm 3.7 (the first with a BPF backend).
Signed-off-by: Ian Rogers <irogers@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20200720061741.1514673-1-irogers@google.com
|
|
This patch adds functions for encoding and emitting compressed riscv
(RVC) instructions to the BPF JIT.
Some regular riscv instructions can be compressed into an RVC instruction
if the instruction fields meet some requirements. For example, "add rd,
rs1, rs2" can be compressed into "c.add rd, rs2" when rd == rs1.
To make using RVC encodings simpler, this patch also adds helper
functions that selectively emit either a regular instruction or a
compressed instruction if possible.
For example, emit_add will produce a "c.add" if possible and regular
"add" otherwise.
Signed-off-by: Luke Nelson <luke.r.nels@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200721025241.8077-3-luke.r.nels@gmail.com
|
|
Now that we have bpf_skip() for emitting nops, use it in
bpf_jit_prologue() in order to reduce code duplication.
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200717165326.6786-6-iii@linux.ibm.com
|
|
This patch makes the necessary changes to struct rv_jit_context and to
bpf_int_jit_compile to support compressed riscv (RVC) instructions in
the BPF JIT.
It changes the JIT image to be u16 instead of u32, since RVC instructions
are 2 bytes as opposed to 4.
It also changes ctx->offset and ctx->ninsns to refer to 2-byte
instructions rather than 4-byte ones. The riscv PC is required to be
16-bit aligned with or without RVC, so this is sufficient to refer to
any valid riscv offset.
The code for computing jump offsets in bytes is updated accordingly,
and factored into a new "ninsns_rvoff" function to simplify the code.
Signed-off-by: Luke Nelson <luke.r.nels@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200721025241.8077-2-luke.r.nels@gmail.com
|
|
"BPF_MAXINSNS: Maximum possible literals" unnecessarily falls back to
the interpreter because of failing sanity check in bpf_set_addr. The
problem is that there are a lot of branches that can be shrunk, and
doing so opens up the possibility to shrink even more. This process
does not converge after 3 passes, causing code offsets to change during
the codegen pass, which must never happen.
Fix by inserting nops during codegen pass in order to preserve code
offets.
Fixes: 4e9b4a6883dd ("s390/bpf: Use relative long branches")
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200717165326.6786-5-iii@linux.ibm.com
|
|
"BPF_MAXINSNS: Maximum possible literals" test causes panic with
bpf_jit_harden = 2. The reason is that BPF_JMP | BPF_EXIT is always
emitted as brc, however, after removal of JITed image size
limitations, brcl might be required.
Fix by using brcl when necessary.
Fixes: 4e9b4a6883dd ("s390/bpf: Use relative long branches")
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200717165326.6786-4-iii@linux.ibm.com
|
|
Both signed and unsigned variants of BPF_JMP | BPF_K require
sign-extending the immediate. JIT emits cgfi for the signed case,
which is correct, and clgfi for the unsigned case, which is not
correct: clgfi zero-extends the immediate.
s390 does not provide an instruction that does sign-extension and
unsigned comparison at the same time. Therefore, fix by first loading
the sign-extended immediate into work register REG_1 and proceeding
as if it's BPF_X.
Fixes: 4e9b4a6883dd ("s390/bpf: Use relative long branches")
Reported-by: Seth Forshee <seth.forshee@canonical.com>
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Tested-by: Seth Forshee <seth.forshee@canonical.com>
Link: https://lore.kernel.org/bpf/20200717165326.6786-3-iii@linux.ibm.com
|
|
When running out of srctree, relative path to lib/test_bpf.ko is
different than when running in srctree. Check $building_out_of_srctree
environment variable and use a different relative path if needed.
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200717165326.6786-2-iii@linux.ibm.com
|
|
Fix the following cpumap kthread hung. The issue is currently occurring
when __cpu_map_load_bpf_program fails (e.g if the bpf prog has not
BPF_XDP_CPUMAP as expected_attach_type)
$./test_progs -n 101
101/1 cpumap_with_progs:OK
101 xdp_cpumap_attach:OK
Summary: 1/1 PASSED, 0 SKIPPED, 0 FAILED
[ 369.996478] INFO: task cpumap/0/map:7:205 blocked for more than 122 seconds.
[ 369.998463] Not tainted 5.8.0-rc4-01472-ge57892f50a07 #212
[ 370.000102] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 370.001918] cpumap/0/map:7 D 0 205 2 0x00004000
[ 370.003228] Call Trace:
[ 370.003930] __schedule+0x5c7/0xf50
[ 370.004901] ? io_schedule_timeout+0xb0/0xb0
[ 370.005934] ? static_obj+0x31/0x80
[ 370.006788] ? mark_held_locks+0x24/0x90
[ 370.007752] ? cpu_map_bpf_prog_run_xdp+0x6c0/0x6c0
[ 370.008930] schedule+0x6f/0x160
[ 370.009728] schedule_preempt_disabled+0x14/0x20
[ 370.010829] kthread+0x17b/0x240
[ 370.011433] ? kthread_create_worker_on_cpu+0xd0/0xd0
[ 370.011944] ret_from_fork+0x1f/0x30
[ 370.012348]
Showing all locks held in the system:
[ 370.013025] 1 lock held by khungtaskd/33:
[ 370.013432] #0: ffffffff82b24720 (rcu_read_lock){....}-{1:2}, at: debug_show_all_locks+0x28/0x1c3
[ 370.014461] =============================================
Fixes: 9216477449f3 ("bpf: cpumap: Add the possibility to attach an eBPF program to cpumap")
Reported-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Tested-by: Jakub Sitnicki <jakub@cloudflare.com>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/e54f2aabf959f298939e5507b09c48f8c2e380be.1595170625.git.lorenzo@kernel.org
|
|
When CONFIG_NET is set but CONFIG_INET isn't, build fails with:
ld: kernel/bpf/net_namespace.o: in function `netns_bpf_attach_type_unneed':
kernel/bpf/net_namespace.c:32: undefined reference to `bpf_sk_lookup_enabled'
ld: kernel/bpf/net_namespace.o: in function `netns_bpf_attach_type_need':
kernel/bpf/net_namespace.c:43: undefined reference to `bpf_sk_lookup_enabled'
This is because without CONFIG_INET bpf_sk_lookup_enabled symbol is not
available. Wrap references to bpf_sk_lookup_enabled with preprocessor
conditionals.
Fixes: 1559b4aa1db4 ("inet: Run SK_LOOKUP BPF program on socket lookup")
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
Link: https://lore.kernel.org/bpf/20200721100716.720477-1-jakub@cloudflare.com
|
|
Now that DSA supports MTU configuration, undo the effects of commit
8b1efc0f83f1 ("net: remove MTU limits on a few ether_setup callers") and
let DSA interfaces use the default min_mtu and max_mtu specified by
ether_setup(). This is more important for min_mtu: since DSA is
Ethernet, the minimum MTU is the same as of any other Ethernet
interface, and definitely not zero. For the max_mtu, we have a callback
through which drivers can override that, if they want to.
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
