* Re: [PATCH net-next 04/12] net: sched: add tcf_block_setup()
From: Jiri Pirko @ 2019-06-26 12:12 UTC (permalink / raw)
To: Pablo Neira Ayuso
Cc: netdev, netfilter-devel, davem, thomas.lendacky, f.fainelli,
ariel.elior, michael.chan, santosh, madalin.bucur, yisen.zhuang,
salil.mehta, jeffrey.t.kirsher, tariqt, saeedm, jiri, idosch,
jakub.kicinski, peppe.cavallaro, grygorii.strashko, andrew,
vivien.didelot, alexandre.torgue, joabreu, linux-net-drivers,
ganeshgr, ogerlitz, Manish.Chopra, marcelo.leitner, mkubecek,
venkatkumar.duvvuru, cphealy
In-Reply-To: <20190625083154.jfzhh22zsl3fu2ik@salvia>
Tue, Jun 25, 2019 at 10:31:54AM CEST, pablo@netfilter.org wrote:
>On Fri, Jun 21, 2019 at 07:16:03PM +0200, Jiri Pirko wrote:
>> Thu, Jun 20, 2019 at 09:49:09PM CEST, pablo@netfilter.org wrote:
>>
>> [...]
>>
>> >
>> >+static LIST_HEAD(tcf_block_cb_list);
>>
>> I still don't like the global list. Have to go throught the code more
>> carefully, but why you can't pass the priv/ctx from tc/netfilter. From
>> tc it would be tcf_block as it is now, from netfilter something else.
>
>This tcf_block_cb_list should go away at some point, once drivers know
>how to deal with multiple subsystems using the setup block
>infrastructure. As I said in my previous email, only one can set up
>the block at this stage, the ones coming later will hit busy.
The driver should know if it can bind or is busy. Also, the bind cmd
should contain type of binder (tc/nft/whatever) or perhaps rather binder
priority (according to the hook order in rx/tx).
^ permalink raw reply
* [PATCH net-next v3 0/8] net: aquantia: implement vlan offloads
From: Igor Russkikh @ 2019-06-26 12:35 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev@vger.kernel.org, Igor Russkikh
This patchset introduces hardware VLAN offload support and also does some
maintenance: we replace driver version with uts version string, add
documentation file for atlantic driver, and update maintainers
adding Igor as a maintainer.
v3: shuffle doc sections, per Andrew's comments
v2: updates in doc, gpl spdx tag cleanup
Igor Russkikh (8):
net: aquantia: replace internal driver version code with uts
net: aquantia: add documentation for the atlantic driver
maintainers: declare aquantia atlantic driver maintenance
net: aquantia: make all files GPL-2.0-only
net: aquantia: added vlan offload related macros and functions
net: aquantia: adding fields and device features for vlan offload
net: aquantia: vlan offloads logic in datapath
net: aquantia: implement vlan offload configuration
.../device_drivers/aquantia/atlantic.txt | 439 ++++++++++++++++++
MAINTAINERS | 9 +
.../net/ethernet/aquantia/atlantic/aq_cfg.h | 7 +-
.../ethernet/aquantia/atlantic/aq_drvinfo.c | 2 +-
.../ethernet/aquantia/atlantic/aq_drvinfo.h | 2 +-
.../ethernet/aquantia/atlantic/aq_filters.c | 2 +-
.../ethernet/aquantia/atlantic/aq_filters.h | 2 +-
.../net/ethernet/aquantia/atlantic/aq_main.c | 34 +-
.../net/ethernet/aquantia/atlantic/aq_nic.c | 28 +-
.../net/ethernet/aquantia/atlantic/aq_nic.h | 2 +
.../net/ethernet/aquantia/atlantic/aq_ring.c | 4 +
.../net/ethernet/aquantia/atlantic/aq_ring.h | 9 +-
.../aquantia/atlantic/hw_atl/hw_atl_a0.c | 2 +-
.../aquantia/atlantic/hw_atl/hw_atl_b0.c | 62 ++-
.../atlantic/hw_atl/hw_atl_b0_internal.h | 7 +
.../aquantia/atlantic/hw_atl/hw_atl_llh.c | 16 +
.../aquantia/atlantic/hw_atl/hw_atl_llh.h | 5 +
.../atlantic/hw_atl/hw_atl_llh_internal.h | 18 +
drivers/net/ethernet/aquantia/atlantic/ver.h | 5 -
19 files changed, 608 insertions(+), 47 deletions(-)
create mode 100644 Documentation/networking/device_drivers/aquantia/atlantic.txt
--
2.17.1
^ permalink raw reply
* [PATCH net-next v3 1/8] net: aquantia: replace internal driver version code with uts
From: Igor Russkikh @ 2019-06-26 12:35 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev@vger.kernel.org, Igor Russkikh
In-Reply-To: <cover.1561552290.git.igor.russkikh@aquantia.com>
As it was discussed some time previously, driver is better to
report kernel version string, as it in a best way identifies
the codebase.
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
---
drivers/net/ethernet/aquantia/atlantic/aq_cfg.h | 7 +++----
drivers/net/ethernet/aquantia/atlantic/ver.h | 5 -----
2 files changed, 3 insertions(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_cfg.h b/drivers/net/ethernet/aquantia/atlantic/aq_cfg.h
index 173be45463ee..02f1b70c4e25 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_cfg.h
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_cfg.h
@@ -9,6 +9,8 @@
#ifndef AQ_CFG_H
#define AQ_CFG_H
+#include <generated/utsrelease.h>
+
#define AQ_CFG_VECS_DEF 8U
#define AQ_CFG_TCS_DEF 1U
@@ -86,10 +88,7 @@
#define AQ_CFG_DRV_AUTHOR "aQuantia"
#define AQ_CFG_DRV_DESC "aQuantia Corporation(R) Network Driver"
#define AQ_CFG_DRV_NAME "atlantic"
-#define AQ_CFG_DRV_VERSION __stringify(NIC_MAJOR_DRIVER_VERSION)"."\
- __stringify(NIC_MINOR_DRIVER_VERSION)"."\
- __stringify(NIC_BUILD_DRIVER_VERSION)"."\
- __stringify(NIC_REVISION_DRIVER_VERSION) \
+#define AQ_CFG_DRV_VERSION UTS_RELEASE \
AQ_CFG_DRV_VERSION_SUFFIX
#endif /* AQ_CFG_H */
diff --git a/drivers/net/ethernet/aquantia/atlantic/ver.h b/drivers/net/ethernet/aquantia/atlantic/ver.h
index 23374bffa92b..597654b51e01 100644
--- a/drivers/net/ethernet/aquantia/atlantic/ver.h
+++ b/drivers/net/ethernet/aquantia/atlantic/ver.h
@@ -7,11 +7,6 @@
#ifndef VER_H
#define VER_H
-#define NIC_MAJOR_DRIVER_VERSION 2
-#define NIC_MINOR_DRIVER_VERSION 0
-#define NIC_BUILD_DRIVER_VERSION 4
-#define NIC_REVISION_DRIVER_VERSION 0
-
#define AQ_CFG_DRV_VERSION_SUFFIX "-kern"
#endif /* VER_H */
--
2.17.1
^ permalink raw reply related
* [PATCH net-next v3 3/8] maintainers: declare aquantia atlantic driver maintenance
From: Igor Russkikh @ 2019-06-26 12:35 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev@vger.kernel.org, Igor Russkikh
In-Reply-To: <cover.1561552290.git.igor.russkikh@aquantia.com>
Aquantia is resposible now for all new features and bugfixes.
Reflect that in MAINTAINERS.
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
---
MAINTAINERS | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 606d1f80bc49..82f762ddbe7a 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -1140,6 +1140,15 @@ L: linux-media@vger.kernel.org
S: Maintained
F: drivers/media/i2c/aptina-pll.*
+AQUANTIA ETHERNET DRIVER (atlantic)
+M: Igor Russkikh <igor.russkikh@aquantia.com>
+L: netdev@vger.kernel.org
+S: Supported
+W: http://www.aquantia.com
+Q: http://patchwork.ozlabs.org/project/netdev/list/
+F: drivers/net/ethernet/aquantia/atlantic/
+F: Documentation/networking/device_drivers/aquantia/atlantic.txt
+
ARC FRAMEBUFFER DRIVER
M: Jaya Kumar <jayalk@intworks.biz>
S: Maintained
--
2.17.1
^ permalink raw reply related
* [PATCH net-next v3 2/8] net: aquantia: add documentation for the atlantic driver
From: Igor Russkikh @ 2019-06-26 12:35 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev@vger.kernel.org, Igor Russkikh
In-Reply-To: <cover.1561552290.git.igor.russkikh@aquantia.com>
Document contains configuration options description,
details and examples of driver various settings.
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
---
.../device_drivers/aquantia/atlantic.txt | 439 ++++++++++++++++++
1 file changed, 439 insertions(+)
create mode 100644 Documentation/networking/device_drivers/aquantia/atlantic.txt
diff --git a/Documentation/networking/device_drivers/aquantia/atlantic.txt b/Documentation/networking/device_drivers/aquantia/atlantic.txt
new file mode 100644
index 000000000000..427443f74a41
--- /dev/null
+++ b/Documentation/networking/device_drivers/aquantia/atlantic.txt
@@ -0,0 +1,439 @@
+aQuantia AQtion Driver for the aQuantia Multi-Gigabit PCI Express Family of
+Ethernet Adapters
+=============================================================================
+
+Contents
+========
+
+- Identifying Your Adapter
+- Configuration
+- Supported ethtool options
+- Command Line Parameters
+- Config file parameters
+- Support
+- License
+
+Identifying Your Adapter
+========================
+
+The driver in this release is compatible with AQC-100, AQC-107, AQC-108 based ethernet adapters.
+
+
+SFP+ Devices (for AQC-100 based adapters)
+----------------------------------
+
+This release tested with passive Direct Attach Cables (DAC) and SFP+/LC Optical Transceiver.
+
+Configuration
+=========================
+ Viewing Link Messages
+ ---------------------
+ Link messages will not be displayed to the console if the distribution is
+ restricting system messages. In order to see network driver link messages on
+ your console, set dmesg to eight by entering the following:
+
+ dmesg -n 8
+
+ NOTE: This setting is not saved across reboots.
+
+ Jumbo Frames
+ ------------
+ The driver supports Jumbo Frames for all adapters. Jumbo Frames support is
+ enabled by changing the MTU to a value larger than the default of 1500.
+ The maximum value for the MTU is 16000. Use the `ip` command to
+ increase the MTU size. For example:
+
+ ip link set mtu 16000 dev enp1s0
+
+ ethtool
+ -------
+ The driver utilizes the ethtool interface for driver configuration and
+ diagnostics, as well as displaying statistical information. The latest
+ ethtool version is required for this functionality.
+
+ NAPI
+ ----
+ NAPI (Rx polling mode) is supported in the atlantic driver.
+
+Supported ethtool options
+============================
+ Viewing adapter settings
+ ---------------------
+ ethtool <ethX>
+
+ Output example:
+
+ Settings for enp1s0:
+ Supported ports: [ TP ]
+ Supported link modes: 100baseT/Full
+ 1000baseT/Full
+ 10000baseT/Full
+ 2500baseT/Full
+ 5000baseT/Full
+ Supported pause frame use: Symmetric
+ Supports auto-negotiation: Yes
+ Supported FEC modes: Not reported
+ Advertised link modes: 100baseT/Full
+ 1000baseT/Full
+ 10000baseT/Full
+ 2500baseT/Full
+ 5000baseT/Full
+ Advertised pause frame use: Symmetric
+ Advertised auto-negotiation: Yes
+ Advertised FEC modes: Not reported
+ Speed: 10000Mb/s
+ Duplex: Full
+ Port: Twisted Pair
+ PHYAD: 0
+ Transceiver: internal
+ Auto-negotiation: on
+ MDI-X: Unknown
+ Supports Wake-on: g
+ Wake-on: d
+ Link detected: yes
+
+ ---
+ Note: AQrate speeds (2.5/5 Gb/s) will be displayed only with linux kernels > 4.10.
+ But you can still use these speeds:
+ ethtool -s eth0 autoneg off speed 2500
+
+ Viewing adapter information
+ ---------------------
+ ethtool -i <ethX>
+
+ Output example:
+
+ driver: atlantic
+ version: 5.2.0-050200rc5-generic-kern
+ firmware-version: 3.1.78
+ expansion-rom-version:
+ bus-info: 0000:01:00.0
+ supports-statistics: yes
+ supports-test: no
+ supports-eeprom-access: no
+ supports-register-dump: yes
+ supports-priv-flags: no
+
+
+ Viewing Ethernet adapter statistics:
+ ---------------------
+ ethtool -S <ethX>
+
+ Output example:
+ NIC statistics:
+ InPackets: 13238607
+ InUCast: 13293852
+ InMCast: 52
+ InBCast: 3
+ InErrors: 0
+ OutPackets: 23703019
+ OutUCast: 23704941
+ OutMCast: 67
+ OutBCast: 11
+ InUCastOctects: 213182760
+ OutUCastOctects: 22698443
+ InMCastOctects: 6600
+ OutMCastOctects: 8776
+ InBCastOctects: 192
+ OutBCastOctects: 704
+ InOctects: 2131839552
+ OutOctects: 226938073
+ InPacketsDma: 95532300
+ OutPacketsDma: 59503397
+ InOctetsDma: 1137102462
+ OutOctetsDma: 2394339518
+ InDroppedDma: 0
+ Queue[0] InPackets: 23567131
+ Queue[0] OutPackets: 20070028
+ Queue[0] InJumboPackets: 0
+ Queue[0] InLroPackets: 0
+ Queue[0] InErrors: 0
+ Queue[1] InPackets: 45428967
+ Queue[1] OutPackets: 11306178
+ Queue[1] InJumboPackets: 0
+ Queue[1] InLroPackets: 0
+ Queue[1] InErrors: 0
+ Queue[2] InPackets: 3187011
+ Queue[2] OutPackets: 13080381
+ Queue[2] InJumboPackets: 0
+ Queue[2] InLroPackets: 0
+ Queue[2] InErrors: 0
+ Queue[3] InPackets: 23349136
+ Queue[3] OutPackets: 15046810
+ Queue[3] InJumboPackets: 0
+ Queue[3] InLroPackets: 0
+ Queue[3] InErrors: 0
+
+ Interrupt coalescing support
+ ---------------------------------
+ ITR mode, TX/RX coalescing timings could be viewed with:
+
+ ethtool -c <ethX>
+
+ and changed with:
+
+ ethtool -C <ethX> tx-usecs <usecs> rx-usecs <usecs>
+
+ To disable coalescing:
+
+ ethtool -C <ethX> tx-usecs 0 rx-usecs 0 tx-max-frames 1 tx-max-frames 1
+
+ Wake on LAN support
+ ---------------------------------
+
+ WOL support by magic packet:
+
+ ethtool -s <ethX> wol g
+
+ To disable WOL:
+
+ ethtool -s <ethX> wol d
+
+ Set and check the driver message level
+ ---------------------------------
+
+ Set message level
+
+ ethtool -s <ethX> msglvl <level>
+
+ Level values:
+
+ 0x0001 - general driver status.
+ 0x0002 - hardware probing.
+ 0x0004 - link state.
+ 0x0008 - periodic status check.
+ 0x0010 - interface being brought down.
+ 0x0020 - interface being brought up.
+ 0x0040 - receive error.
+ 0x0080 - transmit error.
+ 0x0200 - interrupt handling.
+ 0x0400 - transmit completion.
+ 0x0800 - receive completion.
+ 0x1000 - packet contents.
+ 0x2000 - hardware status.
+ 0x4000 - Wake-on-LAN status.
+
+ By default, the level of debugging messages is set 0x0001(general driver status).
+
+ Check message level
+
+ ethtool <ethX> | grep "Current message level"
+
+ If you want to disable the output of messages
+
+ ethtool -s <ethX> msglvl 0
+
+ RX flow rules (ntuple filters)
+ ---------------------------------
+ There are separate rules supported, that applies in that order:
+ 1. 16 VLAN ID rules
+ 2. 16 L2 EtherType rules
+ 3. 8 L3/L4 5-Tuple rules
+
+
+ The driver utilizes the ethtool interface for configuring ntuple filters,
+ via "ethtool -N <device> <filter>".
+
+ To enable or disable the RX flow rules:
+
+ ethtool -K ethX ntuple <on|off>
+
+ When disabling ntuple filters, all the user programed filters are
+ flushed from the driver cache and hardware. All needed filters must
+ be re-added when ntuple is re-enabled.
+
+ Because of the fixed order of the rules, the location of filters is also fixed:
+ - Locations 0 - 15 for VLAN ID filters
+ - Locations 16 - 31 for L2 EtherType filters
+ - Locations 32 - 39 for L3/L4 5-tuple filters (locations 32, 36 for IPv6)
+
+ The L3/L4 5-tuple (protocol, source and destination IP address, source and
+ destination TCP/UDP/SCTP port) is compared against 8 filters. For IPv4, up to
+ 8 source and destination addresses can be matched. For IPv6, up to 2 pairs of
+ addresses can be supported. Source and destination ports are only compared for
+ TCP/UDP/SCTP packets.
+
+ To add a filter that directs packet to queue 5, use <-N|-U|--config-nfc|--config-ntuple> switch:
+
+ ethtool -N <ethX> flow-type udp4 src-ip 10.0.0.1 dst-ip 10.0.0.2 src-port 2000 dst-port 2001 action 5 <loc 32>
+
+ - action is the queue number.
+ - loc is the rule number.
+
+ For "flow-type ip4|udp4|tcp4|sctp4|ip6|udp6|tcp6|sctp6" you must set the loc
+ number within 32 - 39.
+ For "flow-type ip4|udp4|tcp4|sctp4|ip6|udp6|tcp6|sctp6" you can set 8 rules
+ for traffic IPv4 or you can set 2 rules for traffic IPv6. Loc number traffic
+ IPv6 is 32 and 36.
+ At the moment you can not use IPv4 and IPv6 filters at the same time.
+
+ Example filter for IPv6 filter traffic:
+
+ sudo ethtool -N <ethX> flow-type tcp6 src-ip 2001:db8:0:f101::1 dst-ip 2001:db8:0:f101::2 action 1 loc 32
+ sudo ethtool -N <ethX> flow-type ip6 src-ip 2001:db8:0:f101::2 dst-ip 2001:db8:0:f101::5 action -1 loc 36
+
+ Example filter for IPv4 filter traffic:
+
+ sudo ethtool -N <ethX> flow-type udp4 src-ip 10.0.0.4 dst-ip 10.0.0.7 src-port 2000 dst-port 2001 loc 32
+ sudo ethtool -N <ethX> flow-type tcp4 src-ip 10.0.0.3 dst-ip 10.0.0.9 src-port 2000 dst-port 2001 loc 33
+ sudo ethtool -N <ethX> flow-type ip4 src-ip 10.0.0.6 dst-ip 10.0.0.4 loc 34
+
+ If you set action -1, then all traffic corresponding to the filter will be discarded.
+ The maximum value action is 31.
+
+
+ The VLAN filter (VLAN id) is compared against 16 filters.
+ VLAN id must be accompanied by mask 0xF000. That is to distinguish VLAN filter
+ from L2 Ethertype filter with UserPriority since both User Priority and VLAN ID
+ are passed in the same 'vlan' parameter.
+
+ To add a filter that directs packets from VLAN 2001 to queue 5:
+ ethtool -N <ethX> flow-type ip4 vlan 2001 m 0xF000 action 1 loc 0
+
+
+ L2 EtherType filters allows filter packet by EtherType field or both EtherType
+ and User Priority (PCP) field of 802.1Q.
+ UserPriority (vlan) parameter must be accompanied by mask 0x1FFF. That is to
+ distinguish VLAN filter from L2 Ethertype filter with UserPriority since both
+ User Priority and VLAN ID are passed in the same 'vlan' parameter.
+
+ To add a filter that directs IP4 packess of priority 3 to queue 3:
+ ethtool -N <ethX> flow-type ether proto 0x800 vlan 0x600 m 0x1FFF action 3 loc 16
+
+
+ To see the list of filters currently present:
+
+ ethtool <-u|-n|--show-nfc|--show-ntuple> <ethX>
+
+ Rules may be deleted from the table itself. This is done using:
+
+ sudo ethtool <-N|-U|--config-nfc|--config-ntuple> <ethX> delete <loc>
+
+ - loc is the rule number to be deleted.
+
+ Rx filters is an interface to load the filter table that funnels all flow
+ into queue 0 unless an alternative queue is specified using "action". In that
+ case, any flow that matches the filter criteria will be directed to the
+ appropriate queue. RX filters is supported on all kernels 2.6.30 and later.
+
+ RSS for UDP
+ ---------------------------------
+ Currently, NIC does not support RSS for fragmented IP packets, which leads to
+ incorrect working of RSS for fragmented UDP traffic. To disable RSS for UDP the
+ RX Flow L3/L4 rule may be used.
+
+ Example:
+ ethtool -N eth0 flow-type udp4 action 0 loc 32
+
+Command Line Parameters
+=======================
+The following command line parameters are available on atlantic driver:
+
+aq_itr -Interrupt throttling mode
+----------------------------------------
+Accepted values: 0, 1, 0xFFFF
+Default value: 0xFFFF
+0 - Disable interrupt throttling.
+1 - Enable interrupt throttling and use specified tx and rx rates.
+0xFFFF - Auto throttling mode. Driver will choose the best RX and TX
+ interrupt throtting settings based on link speed.
+
+aq_itr_tx - TX interrupt throttle rate
+----------------------------------------
+Accepted values: 0 - 0x1FF
+Default value: 0
+TX side throttling in microseconds. Adapter will setup maximum interrupt delay
+to this value. Minimum interrupt delay will be a half of this value
+
+aq_itr_rx - RX interrupt throttle rate
+----------------------------------------
+Accepted values: 0 - 0x1FF
+Default value: 0
+RX side throttling in microseconds. Adapter will setup maximum interrupt delay
+to this value. Minimum interrupt delay will be a half of this value
+
+Note: ITR settings could be changed in runtime by ethtool -c means (see below)
+
+Config file parameters
+=======================
+For some fine tuning and performance optimizations,
+some parameters can be changed in the {source_dir}/aq_cfg.h file.
+
+AQ_CFG_RX_PAGEORDER
+----------------------------------------
+Default value: 0
+RX page order override. Thats a power of 2 number of RX pages allocated for
+each descriptor. Received descriptor size is still limited by AQ_CFG_RX_FRAME_MAX.
+Increasing pageorder makes page reuse better (actual on iommu enabled systems).
+
+AQ_CFG_RX_REFILL_THRES
+----------------------------------------
+Default value: 32
+RX refill threshold. RX path will not refill freed descriptors until the
+specified number of free descriptors is observed. Larger values may help
+better page reuse but may lead to packet drops as well.
+
+AQ_CFG_VECS_DEF
+------------------------------------------------------------
+Number of queues
+Valid Range: 0 - 8 (up to AQ_CFG_VECS_MAX)
+Default value: 8
+Notice this value will be capped by the number of cores available on the system.
+
+AQ_CFG_IS_RSS_DEF
+------------------------------------------------------------
+Enable/disable Receive Side Scaling
+
+This feature allows the adapter to distribute receive processing
+across multiple CPU-cores and to prevent from overloading a single CPU core.
+
+Valid values
+0 - disabled
+1 - enabled
+
+Default value: 1
+
+AQ_CFG_NUM_RSS_QUEUES_DEF
+------------------------------------------------------------
+Number of queues for Receive Side Scaling
+Valid Range: 0 - 8 (up to AQ_CFG_VECS_DEF)
+
+Default value: AQ_CFG_VECS_DEF
+
+AQ_CFG_IS_LRO_DEF
+------------------------------------------------------------
+Enable/disable Large Receive Offload
+
+This offload enables the adapter to coalesce multiple TCP segments and indicate
+them as a single coalesced unit to the OS networking subsystem.
+The system consumes less energy but it also introduces more latency in packets processing.
+
+Valid values
+0 - disabled
+1 - enabled
+
+Default value: 1
+
+AQ_CFG_TX_CLEAN_BUDGET
+----------------------------------------
+Maximum descriptors to cleanup on TX at once.
+Default value: 256
+
+After the aq_cfg.h file changed the driver must be rebuilt to take effect.
+
+Support
+=======
+
+If an issue is identified with the released source code on the supported
+kernel with a supported adapter, email the specific information related
+to the issue to support@aquantia.com
+
+License
+=======
+
+aQuantia Corporation Network Driver
+Copyright(c) 2014 - 2019 aQuantia Corporation.
+
+This program is free software; you can redistribute it and/or modify it
+under the terms and conditions of the GNU General Public License,
+version 2, as published by the Free Software Foundation.
--
2.17.1
^ permalink raw reply related
* [PATCH net-next v3 4/8] net: aquantia: make all files GPL-2.0-only
From: Igor Russkikh @ 2019-06-26 12:35 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev@vger.kernel.org, Igor Russkikh
In-Reply-To: <cover.1561552290.git.igor.russkikh@aquantia.com>
It was noticed some files had -or-later, however overall driver has
-only license. Clean this up.
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
---
drivers/net/ethernet/aquantia/atlantic/aq_drvinfo.c | 2 +-
drivers/net/ethernet/aquantia/atlantic/aq_drvinfo.h | 2 +-
drivers/net/ethernet/aquantia/atlantic/aq_filters.c | 2 +-
drivers/net/ethernet/aquantia/atlantic/aq_filters.h | 2 +-
4 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_drvinfo.c b/drivers/net/ethernet/aquantia/atlantic/aq_drvinfo.c
index adad6a7acabe..6da65099047d 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_drvinfo.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_drvinfo.c
@@ -1,4 +1,4 @@
-// SPDX-License-Identifier: GPL-2.0-or-later
+// SPDX-License-Identifier: GPL-2.0-only
/* Copyright (C) 2014-2019 aQuantia Corporation. */
/* File aq_drvinfo.c: Definition of common code for firmware info in sys.*/
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_drvinfo.h b/drivers/net/ethernet/aquantia/atlantic/aq_drvinfo.h
index 41fbb1358068..23a0487893a7 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_drvinfo.h
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_drvinfo.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* SPDX-License-Identifier: GPL-2.0-only */
/* Copyright (C) 2014-2017 aQuantia Corporation. */
/* File aq_drvinfo.h: Declaration of common code for firmware info in sys.*/
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_filters.c b/drivers/net/ethernet/aquantia/atlantic/aq_filters.c
index 18bc035da850..04a4cb7cfcc5 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_filters.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_filters.c
@@ -1,4 +1,4 @@
-// SPDX-License-Identifier: GPL-2.0-or-later
+// SPDX-License-Identifier: GPL-2.0-only
/* Copyright (C) 2014-2017 aQuantia Corporation. */
/* File aq_filters.c: RX filters related functions. */
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_filters.h b/drivers/net/ethernet/aquantia/atlantic/aq_filters.h
index c6a08c6585d5..122e06c88a33 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_filters.h
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_filters.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* SPDX-License-Identifier: GPL-2.0-only */
/* Copyright (C) 2014-2017 aQuantia Corporation. */
/* File aq_filters.h: RX filters related functions. */
--
2.17.1
^ permalink raw reply related
* [PATCH net-next v3 5/8] net: aquantia: added vlan offload related macros and functions
From: Igor Russkikh @ 2019-06-26 12:35 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev@vger.kernel.org, Igor Russkikh
In-Reply-To: <cover.1561552290.git.igor.russkikh@aquantia.com>
Register declaration macros required to work with vlan offload mode.
Tested-by: Nikita Danilov <ndanilov@aquantia.com>
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
---
.../atlantic/hw_atl/hw_atl_b0_internal.h | 7 +++++++
.../aquantia/atlantic/hw_atl/hw_atl_llh.c | 16 ++++++++++++++++
.../aquantia/atlantic/hw_atl/hw_atl_llh.h | 5 +++++
.../atlantic/hw_atl/hw_atl_llh_internal.h | 18 ++++++++++++++++++
4 files changed, 46 insertions(+)
diff --git a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0_internal.h b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0_internal.h
index e4ba2ccf9830..808d8cd4252a 100644
--- a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0_internal.h
+++ b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0_internal.h
@@ -107,10 +107,17 @@
#define HW_ATL_B0_RXD_NCEA0 (0x1)
#define HW_ATL_B0_RXD_WB_STAT_RSSTYPE (0x0000000F)
+#define HW_ATL_B0_RXD_WB_STAT_RSSTYPE_SHIFT (0x0)
#define HW_ATL_B0_RXD_WB_STAT_PKTTYPE (0x00000FF0)
+#define HW_ATL_B0_RXD_WB_STAT_PKTTYPE_SHIFT (0x4)
#define HW_ATL_B0_RXD_WB_STAT_RXCTRL (0x00180000)
+#define HW_ATL_B0_RXD_WB_STAT_RXCTRL_SHIFT (0x13)
#define HW_ATL_B0_RXD_WB_STAT_SPLHDR (0x00200000)
#define HW_ATL_B0_RXD_WB_STAT_HDRLEN (0xFFC00000)
+#define HW_ATL_B0_RXD_WB_STAT_HDRLEN_SHIFT (0x16)
+
+#define HW_ATL_B0_RXD_WB_PKTTYPE_VLAN BIT(5)
+#define HW_ATL_B0_RXD_WB_PKTTYPE_VLAN_DOUBLE BIT(6)
#define HW_ATL_B0_RXD_WB_STAT2_DD (0x0001)
#define HW_ATL_B0_RXD_WB_STAT2_EOP (0x0002)
diff --git a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh.c b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh.c
index 451529069f28..1149812ae463 100644
--- a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh.c
+++ b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh.c
@@ -1004,6 +1004,22 @@ void hw_atl_rpo_rx_desc_vlan_stripping_set(struct aq_hw_s *aq_hw,
rx_desc_vlan_stripping);
}
+void hw_atl_rpo_outer_vlan_tag_mode_set(void *context,
+ u32 outervlantagmode)
+{
+ aq_hw_write_reg_bit(context, HW_ATL_RPO_OUTER_VL_INS_MODE_ADR,
+ HW_ATL_RPO_OUTER_VL_INS_MODE_MSK,
+ HW_ATL_RPO_OUTER_VL_INS_MODE_SHIFT,
+ outervlantagmode);
+}
+
+u32 hw_atl_rpo_outer_vlan_tag_mode_get(void *context)
+{
+ return aq_hw_read_reg_bit(context, HW_ATL_RPO_OUTER_VL_INS_MODE_ADR,
+ HW_ATL_RPO_OUTER_VL_INS_MODE_MSK,
+ HW_ATL_RPO_OUTER_VL_INS_MODE_SHIFT);
+}
+
void hw_atl_rpo_tcp_udp_crc_offload_en_set(struct aq_hw_s *aq_hw,
u32 tcp_udp_crc_offload_en)
{
diff --git a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh.h b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh.h
index 34b42ce43512..0c37abbabca5 100644
--- a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh.h
+++ b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh.h
@@ -488,6 +488,11 @@ void hw_atl_rpo_rx_desc_vlan_stripping_set(struct aq_hw_s *aq_hw,
u32 rx_desc_vlan_stripping,
u32 descriptor);
+void hw_atl_rpo_outer_vlan_tag_mode_set(void *context,
+ u32 outervlantagmode);
+
+u32 hw_atl_rpo_outer_vlan_tag_mode_get(void *context);
+
/* set tcp/udp checksum offload enable */
void hw_atl_rpo_tcp_udp_crc_offload_en_set(struct aq_hw_s *aq_hw,
u32 tcp_udp_crc_offload_en);
diff --git a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh_internal.h b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh_internal.h
index fc1446f737bb..c3febcdfa92e 100644
--- a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh_internal.h
+++ b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_llh_internal.h
@@ -1383,6 +1383,24 @@
/* default value of bitfield l4_chk_en */
#define HW_ATL_RPOL4CHK_EN_DEFAULT 0x0
+/* RX outer_vl_ins_mode Bitfield Definitions
+ * Preprocessor definitions for the bitfield "outer_vl_ins_mode".
+ * PORT="pif_rpo_outer_vl_mode_i"
+ */
+
+/* Register address for bitfield outer_vl_ins_mode */
+#define HW_ATL_RPO_OUTER_VL_INS_MODE_ADR 0x00005580
+/* Bitmask for bitfield outer_vl_ins_mode */
+#define HW_ATL_RPO_OUTER_VL_INS_MODE_MSK 0x00000004
+/* Inverted bitmask for bitfield outer_vl_ins_mode */
+#define HW_ATL_RPO_OUTER_VL_INS_MODE_MSKN 0xFFFFFFFB
+/* Lower bit position of bitfield outer_vl_ins_mode */
+#define HW_ATL_RPO_OUTER_VL_INS_MODE_SHIFT 2
+/* Width of bitfield outer_vl_ins_mode */
+#define HW_ATL_RPO_OUTER_VL_INS_MODE_WIDTH 1
+/* Default value of bitfield outer_vl_ins_mode */
+#define HW_ATL_RPO_OUTER_VL_INS_MODE_DEFAULT 0x0
+
/* rx reg_res_dsbl bitfield definitions
* preprocessor definitions for the bitfield "reg_res_dsbl".
* port="pif_rx_reg_res_dsbl_i"
--
2.17.1
^ permalink raw reply related
* [PATCH net-next v3 6/8] net: aquantia: adding fields and device features for vlan offload
From: Igor Russkikh @ 2019-06-26 12:35 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev@vger.kernel.org, Igor Russkikh
In-Reply-To: <cover.1561552290.git.igor.russkikh@aquantia.com>
Updating features and vlan_features with vlan HW offload.
Added vlan_tag fields to rx/tx ring_buff to track vlan related data.
Tested-by: Nikita Danilov <ndanilov@aquantia.com>
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
---
drivers/net/ethernet/aquantia/atlantic/aq_nic.c | 9 ++++++---
drivers/net/ethernet/aquantia/atlantic/aq_nic.h | 2 ++
drivers/net/ethernet/aquantia/atlantic/aq_ring.h | 9 ++++++---
.../net/ethernet/aquantia/atlantic/hw_atl/hw_atl_a0.c | 2 +-
.../net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c | 10 +++++++---
5 files changed, 22 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
index 0da5e161ec5d..aa0521e8e71a 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
@@ -126,6 +126,8 @@ void aq_nic_cfg_start(struct aq_nic_s *self)
cfg->link_speed_msk &= cfg->aq_hw_caps->link_speed_msk;
cfg->features = cfg->aq_hw_caps->hw_features;
+ cfg->is_vlan_rx_strip = !!(cfg->features & NETIF_F_HW_VLAN_CTAG_RX);
+ cfg->is_vlan_tx_insert = !!(cfg->features & NETIF_F_HW_VLAN_CTAG_TX);
}
static int aq_nic_update_link_status(struct aq_nic_s *self)
@@ -285,7 +287,8 @@ void aq_nic_ndev_init(struct aq_nic_s *self)
self->ndev->hw_features |= aq_hw_caps->hw_features;
self->ndev->features = aq_hw_caps->hw_features;
self->ndev->vlan_features |= NETIF_F_HW_CSUM | NETIF_F_RXCSUM |
- NETIF_F_RXHASH | NETIF_F_SG | NETIF_F_LRO;
+ NETIF_F_RXHASH | NETIF_F_SG |
+ NETIF_F_LRO | NETIF_F_TSO;
self->ndev->priv_flags = aq_hw_caps->hw_priv_flags;
self->ndev->priv_flags |= IFF_LIVE_ADDR_CHANGE;
@@ -434,7 +437,7 @@ static unsigned int aq_nic_map_skb(struct aq_nic_s *self,
dx_buff->len_l3 = ip_hdrlen(skb);
dx_buff->len_l4 = tcp_hdrlen(skb);
dx_buff->mss = skb_shinfo(skb)->gso_size;
- dx_buff->is_txc = 1U;
+ dx_buff->is_gso = 1U;
dx_buff->eop_index = 0xffffU;
dx_buff->is_ipv6 =
@@ -534,7 +537,7 @@ static unsigned int aq_nic_map_skb(struct aq_nic_s *self,
--ret, dx = aq_ring_next_dx(ring, dx)) {
dx_buff = &ring->buff_ring[dx];
- if (!dx_buff->is_txc && dx_buff->pa) {
+ if (!dx_buff->is_gso && dx_buff->pa) {
if (unlikely(dx_buff->is_sop)) {
dma_unmap_single(aq_nic_get_dev(self),
dx_buff->pa,
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.h b/drivers/net/ethernet/aquantia/atlantic/aq_nic.h
index eb2e3c7c36f9..26c72f298684 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.h
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.h
@@ -35,6 +35,8 @@ struct aq_nic_cfg_s {
u32 flow_control;
u32 link_speed_msk;
u32 wol;
+ u8 is_vlan_rx_strip;
+ u8 is_vlan_tx_insert;
u16 is_mc_list_enabled;
u16 mc_list_count;
bool is_autoneg;
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_ring.h b/drivers/net/ethernet/aquantia/atlantic/aq_ring.h
index 6bd67210d0b7..47abd09d06c2 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_ring.h
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_ring.h
@@ -27,7 +27,7 @@ struct aq_rxpage {
* +----------+----------+----------+-----------
* 4/8bytes|len pkt |len pkt | | skb
* +----------+----------+----------+-----------
- * 4/8bytes|is_txc |len,flags |len |len,is_eop
+ * 4/8bytes|is_gso |len,flags |len |len,is_eop
* +----------+----------+----------+-----------
*
* This aq_ring_buff_s doesn't have endianness dependency.
@@ -44,6 +44,7 @@ struct __packed aq_ring_buff_s {
u8 is_hash_l4;
u8 rsvd1;
struct aq_rxpage rxdata;
+ u16 vlan_rx_tag;
};
/* EOP */
struct {
@@ -59,6 +60,7 @@ struct __packed aq_ring_buff_s {
u8 is_ipv6:1;
u8 rsvd2:7;
u32 len_pkt;
+ u16 vlan_tx_tag;
};
};
union {
@@ -70,11 +72,12 @@ struct __packed aq_ring_buff_s {
u32 is_cso_err:1;
u32 is_sop:1;
u32 is_eop:1;
- u32 is_txc:1;
+ u32 is_gso:1;
u32 is_mapped:1;
u32 is_cleaned:1;
u32 is_error:1;
- u32 rsvd3:6;
+ u32 is_vlan:1;
+ u32 rsvd3:5;
u16 eop_index;
u16 rsvd4;
};
diff --git a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_a0.c b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_a0.c
index 0f140a9fe404..359a4d387185 100644
--- a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_a0.c
+++ b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_a0.c
@@ -451,7 +451,7 @@ static int hw_atl_a0_hw_ring_tx_xmit(struct aq_hw_s *self,
buff = &ring->buff_ring[ring->sw_tail];
- if (buff->is_txc) {
+ if (buff->is_gso) {
txd->ctl |= (buff->len_l3 << 31) |
(buff->len_l2 << 24) |
HW_ATL_A0_TXD_CTL_CMD_TCP |
diff --git a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c
index 1c7593d54035..773345371bcf 100644
--- a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c
+++ b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c
@@ -40,7 +40,9 @@
NETIF_F_TSO | \
NETIF_F_LRO | \
NETIF_F_NTUPLE | \
- NETIF_F_HW_VLAN_CTAG_FILTER, \
+ NETIF_F_HW_VLAN_CTAG_FILTER | \
+ NETIF_F_HW_VLAN_CTAG_RX | \
+ NETIF_F_HW_VLAN_CTAG_TX, \
.hw_priv_flags = IFF_UNICAST_FLT, \
.flow_control = true, \
.mtu = HW_ATL_B0_MTU_JUMBO, \
@@ -501,7 +503,7 @@ static int hw_atl_b0_hw_ring_tx_xmit(struct aq_hw_s *self,
buff = &ring->buff_ring[ring->sw_tail];
- if (buff->is_txc) {
+ if (buff->is_gso) {
txd->ctl |= (buff->len_l3 << 31) |
(buff->len_l2 << 24) |
HW_ATL_B0_TXD_CTL_CMD_TCP |
@@ -559,6 +561,7 @@ static int hw_atl_b0_hw_ring_rx_init(struct aq_hw_s *self,
{
u32 dma_desc_addr_lsw = (u32)aq_ring->dx_ring_pa;
u32 dma_desc_addr_msw = (u32)(((u64)aq_ring->dx_ring_pa) >> 32);
+ u32 vlan_rx_stripping = self->aq_nic_cfg->is_vlan_rx_strip;
hw_atl_rdm_rx_desc_en_set(self, false, aq_ring->idx);
@@ -578,7 +581,8 @@ static int hw_atl_b0_hw_ring_rx_init(struct aq_hw_s *self,
hw_atl_rdm_rx_desc_head_buff_size_set(self, 0U, aq_ring->idx);
hw_atl_rdm_rx_desc_head_splitting_set(self, 0U, aq_ring->idx);
- hw_atl_rpo_rx_desc_vlan_stripping_set(self, 0U, aq_ring->idx);
+ hw_atl_rpo_rx_desc_vlan_stripping_set(self, !!vlan_rx_stripping,
+ aq_ring->idx);
/* Rx ring set mode */
--
2.17.1
^ permalink raw reply related
* [PATCH net-next v3 7/8] net: aquantia: vlan offloads logic in datapath
From: Igor Russkikh @ 2019-06-26 12:35 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev@vger.kernel.org, Igor Russkikh
In-Reply-To: <cover.1561552290.git.igor.russkikh@aquantia.com>
Update datapath by adding logic related to hardware assisted
vlan strip/insert behaviour.
Tested-by: Nikita Danilov <ndanilov@aquantia.com>
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
---
.../net/ethernet/aquantia/atlantic/aq_nic.c | 23 +++++---
.../net/ethernet/aquantia/atlantic/aq_ring.c | 4 ++
.../aquantia/atlantic/hw_atl/hw_atl_b0.c | 52 ++++++++++++++-----
3 files changed, 60 insertions(+), 19 deletions(-)
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
index aa0521e8e71a..746f85e6de13 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
@@ -429,26 +429,37 @@ static unsigned int aq_nic_map_skb(struct aq_nic_s *self,
unsigned int dx = ring->sw_tail;
struct aq_ring_buff_s *first = NULL;
struct aq_ring_buff_s *dx_buff = &ring->buff_ring[dx];
+ bool need_context_tag = false;
+
+ dx_buff->flags = 0U;
if (unlikely(skb_is_gso(skb))) {
- dx_buff->flags = 0U;
+ dx_buff->mss = skb_shinfo(skb)->gso_size;
+ dx_buff->is_gso = 1U;
dx_buff->len_pkt = skb->len;
dx_buff->len_l2 = ETH_HLEN;
dx_buff->len_l3 = ip_hdrlen(skb);
dx_buff->len_l4 = tcp_hdrlen(skb);
- dx_buff->mss = skb_shinfo(skb)->gso_size;
- dx_buff->is_gso = 1U;
dx_buff->eop_index = 0xffffU;
-
dx_buff->is_ipv6 =
(ip_hdr(skb)->version == 6) ? 1U : 0U;
+ need_context_tag = true;
+ }
+ if (self->aq_nic_cfg.is_vlan_tx_insert && skb_vlan_tag_present(skb)) {
+ dx_buff->vlan_tx_tag = skb_vlan_tag_get(skb);
+ dx_buff->len_pkt = skb->len;
+ dx_buff->is_vlan = 1U;
+ need_context_tag = true;
+ }
+
+ if (need_context_tag) {
dx = aq_ring_next_dx(ring, dx);
dx_buff = &ring->buff_ring[dx];
+ dx_buff->flags = 0U;
++ret;
}
- dx_buff->flags = 0U;
dx_buff->len = skb_headlen(skb);
dx_buff->pa = dma_map_single(aq_nic_get_dev(self),
skb->data,
@@ -537,7 +548,7 @@ static unsigned int aq_nic_map_skb(struct aq_nic_s *self,
--ret, dx = aq_ring_next_dx(ring, dx)) {
dx_buff = &ring->buff_ring[dx];
- if (!dx_buff->is_gso && dx_buff->pa) {
+ if (!dx_buff->is_gso && !dx_buff->is_vlan && dx_buff->pa) {
if (unlikely(dx_buff->is_sop)) {
dma_unmap_single(aq_nic_get_dev(self),
dx_buff->pa,
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_ring.c b/drivers/net/ethernet/aquantia/atlantic/aq_ring.c
index 2a7b91ed17c5..3901d7994ca1 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_ring.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_ring.c
@@ -409,6 +409,10 @@ int aq_ring_rx_clean(struct aq_ring_s *self,
}
}
+ if (buff->is_vlan)
+ __vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q),
+ buff->vlan_rx_tag);
+
skb->protocol = eth_type_trans(skb, ndev);
aq_rx_checksum(self, buff, skb);
diff --git a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c
index 773345371bcf..1d216757bb9d 100644
--- a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c
+++ b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c
@@ -247,6 +247,9 @@ static int hw_atl_b0_hw_offload_set(struct aq_hw_s *self,
/* LSO offloads*/
hw_atl_tdm_large_send_offload_en_set(self, 0xFFFFFFFFU);
+ /* Outer VLAN tag offload */
+ hw_atl_rpo_outer_vlan_tag_mode_set(self, 1U);
+
/* LRO offloads */
{
unsigned int val = (8U < HW_ATL_B0_LRO_RXD_MAX) ? 0x3U :
@@ -489,6 +492,7 @@ static int hw_atl_b0_hw_ring_tx_xmit(struct aq_hw_s *self,
unsigned int buff_pa_len = 0U;
unsigned int pkt_len = 0U;
unsigned int frag_count = 0U;
+ bool is_vlan = false;
bool is_gso = false;
buff = &ring->buff_ring[ring->sw_tail];
@@ -504,35 +508,43 @@ static int hw_atl_b0_hw_ring_tx_xmit(struct aq_hw_s *self,
buff = &ring->buff_ring[ring->sw_tail];
if (buff->is_gso) {
+ txd->ctl |= HW_ATL_B0_TXD_CTL_CMD_TCP;
+ txd->ctl |= HW_ATL_B0_TXD_CTL_DESC_TYPE_TXC;
txd->ctl |= (buff->len_l3 << 31) |
- (buff->len_l2 << 24) |
- HW_ATL_B0_TXD_CTL_CMD_TCP |
- HW_ATL_B0_TXD_CTL_DESC_TYPE_TXC;
- txd->ctl2 |= (buff->mss << 16) |
- (buff->len_l4 << 8) |
- (buff->len_l3 >> 1);
+ (buff->len_l2 << 24);
+ txd->ctl2 |= (buff->mss << 16);
+ is_gso = true;
pkt_len -= (buff->len_l4 +
buff->len_l3 +
buff->len_l2);
- is_gso = true;
-
if (buff->is_ipv6)
txd->ctl |= HW_ATL_B0_TXD_CTL_CMD_IPV6;
- } else {
+ txd->ctl2 |= (buff->len_l4 << 8) |
+ (buff->len_l3 >> 1);
+ }
+ if (buff->is_vlan) {
+ txd->ctl |= HW_ATL_B0_TXD_CTL_DESC_TYPE_TXC;
+ txd->ctl |= buff->vlan_tx_tag << 4;
+ is_vlan = true;
+ }
+ if (!buff->is_gso && !buff->is_vlan) {
buff_pa_len = buff->len;
txd->buf_addr = buff->pa;
txd->ctl |= (HW_ATL_B0_TXD_CTL_BLEN &
((u32)buff_pa_len << 4));
txd->ctl |= HW_ATL_B0_TXD_CTL_DESC_TYPE_TXD;
+
/* PAY_LEN */
txd->ctl2 |= HW_ATL_B0_TXD_CTL2_LEN & (pkt_len << 14);
- if (is_gso) {
- txd->ctl |= HW_ATL_B0_TXD_CTL_CMD_LSO;
+ if (is_gso || is_vlan) {
+ /* enable tx context */
txd->ctl2 |= HW_ATL_B0_TXD_CTL2_CTX_EN;
}
+ if (is_gso)
+ txd->ctl |= HW_ATL_B0_TXD_CTL_CMD_LSO;
/* Tx checksum offloads */
if (buff->is_ip_cso)
@@ -541,13 +553,16 @@ static int hw_atl_b0_hw_ring_tx_xmit(struct aq_hw_s *self,
if (buff->is_udp_cso || buff->is_tcp_cso)
txd->ctl |= HW_ATL_B0_TXD_CTL_CMD_TUCSO;
+ if (is_vlan)
+ txd->ctl |= HW_ATL_B0_TXD_CTL_CMD_VLAN;
+
if (unlikely(buff->is_eop)) {
txd->ctl |= HW_ATL_B0_TXD_CTL_EOP;
txd->ctl |= HW_ATL_B0_TXD_CTL_CMD_WB;
is_gso = false;
+ is_vlan = false;
}
}
-
ring->sw_tail = aq_ring_next_dx(ring, ring->sw_tail);
}
@@ -685,11 +700,15 @@ static int hw_atl_b0_hw_ring_rx_receive(struct aq_hw_s *self,
buff = &ring->buff_ring[ring->hw_head];
+ buff->flags = 0U;
+ buff->is_hash_l4 = 0U;
+
rx_stat = (0x0000003CU & rxd_wb->status) >> 2;
is_rx_check_sum_enabled = (rxd_wb->type >> 19) & 0x3U;
- pkt_type = 0xFFU & (rxd_wb->type >> 4);
+ pkt_type = (rxd_wb->type & HW_ATL_B0_RXD_WB_STAT_PKTTYPE) >>
+ HW_ATL_B0_RXD_WB_STAT_PKTTYPE_SHIFT;
if (is_rx_check_sum_enabled & BIT(0) &&
(0x0U == (pkt_type & 0x3U)))
@@ -710,6 +729,13 @@ static int hw_atl_b0_hw_ring_rx_receive(struct aq_hw_s *self,
buff->is_cso_err = 0U;
}
+ if (self->aq_nic_cfg->is_vlan_rx_strip &&
+ ((pkt_type & HW_ATL_B0_RXD_WB_PKTTYPE_VLAN) ||
+ (pkt_type & HW_ATL_B0_RXD_WB_PKTTYPE_VLAN_DOUBLE))) {
+ buff->is_vlan = 1;
+ buff->vlan_rx_tag = le16_to_cpu(rxd_wb->vlan);
+ }
+
if ((rx_stat & BIT(0)) || rxd_wb->type & 0x1000U) {
/* MAC error or DMA error */
buff->is_error = 1U;
--
2.17.1
^ permalink raw reply related
* [PATCH net-next v3 8/8] net: aquantia: implement vlan offload configuration
From: Igor Russkikh @ 2019-06-26 12:35 UTC (permalink / raw)
To: David S . Miller; +Cc: netdev@vger.kernel.org, Igor Russkikh
In-Reply-To: <cover.1561552290.git.igor.russkikh@aquantia.com>
set_features should update flags and reinit hardware if
vlan offload settings were changed.
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Tested-by: Nikita Danilov <ndanilov@aquantia.com>
---
.../net/ethernet/aquantia/atlantic/aq_main.c | 34 +++++++++++++++----
1 file changed, 27 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_main.c b/drivers/net/ethernet/aquantia/atlantic/aq_main.c
index 5315df5ff6f8..100722ad5c2d 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_main.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_main.c
@@ -108,11 +108,16 @@ static int aq_ndev_change_mtu(struct net_device *ndev, int new_mtu)
static int aq_ndev_set_features(struct net_device *ndev,
netdev_features_t features)
{
+ bool is_vlan_rx_strip = !!(features & NETIF_F_HW_VLAN_CTAG_RX);
+ bool is_vlan_tx_insert = !!(features & NETIF_F_HW_VLAN_CTAG_TX);
struct aq_nic_s *aq_nic = netdev_priv(ndev);
- struct aq_nic_cfg_s *aq_cfg = aq_nic_get_cfg(aq_nic);
+ bool need_ndev_restart = false;
+ struct aq_nic_cfg_s *aq_cfg;
bool is_lro = false;
int err = 0;
+ aq_cfg = aq_nic_get_cfg(aq_nic);
+
if (!(features & NETIF_F_NTUPLE)) {
if (aq_nic->ndev->features & NETIF_F_NTUPLE) {
err = aq_clear_rxnfc_all_rules(aq_nic);
@@ -135,17 +140,32 @@ static int aq_ndev_set_features(struct net_device *ndev,
if (aq_cfg->is_lro != is_lro) {
aq_cfg->is_lro = is_lro;
-
- if (netif_running(ndev)) {
- aq_ndev_close(ndev);
- aq_ndev_open(ndev);
- }
+ need_ndev_restart = true;
}
}
- if ((aq_nic->ndev->features ^ features) & NETIF_F_RXCSUM)
+
+ if ((aq_nic->ndev->features ^ features) & NETIF_F_RXCSUM) {
err = aq_nic->aq_hw_ops->hw_set_offload(aq_nic->aq_hw,
aq_cfg);
+ if (unlikely(err))
+ goto err_exit;
+ }
+
+ if (aq_cfg->is_vlan_rx_strip != is_vlan_rx_strip) {
+ aq_cfg->is_vlan_rx_strip = is_vlan_rx_strip;
+ need_ndev_restart = true;
+ }
+ if (aq_cfg->is_vlan_tx_insert != is_vlan_tx_insert) {
+ aq_cfg->is_vlan_tx_insert = is_vlan_tx_insert;
+ need_ndev_restart = true;
+ }
+
+ if (need_ndev_restart && netif_running(ndev)) {
+ aq_ndev_close(ndev);
+ aq_ndev_open(ndev);
+ }
+
err_exit:
return err;
}
--
2.17.1
^ permalink raw reply related
* Re: [PATCH v4 net-next 1/4] net: core: page_pool: add user cnt preventing pool deletion
From: Ivan Khoronzhuk @ 2019-06-26 12:39 UTC (permalink / raw)
To: Jesper Dangaard Brouer
Cc: davem, grygorii.strashko, saeedm, leon, ast, linux-kernel,
linux-omap, ilias.apalodimas, netdev, daniel, jakub.kicinski,
john.fastabend
In-Reply-To: <20190626135128.5724f40e@carbon>
On Wed, Jun 26, 2019 at 01:51:28PM +0200, Jesper Dangaard Brouer wrote:
>On Wed, 26 Jun 2019 13:49:49 +0300
>Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org> wrote:
>
>> On Wed, Jun 26, 2019 at 12:42:16PM +0200, Jesper Dangaard Brouer wrote:
>> >On Tue, 25 Jun 2019 20:59:45 +0300
>> >Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org> wrote:
>> >
>> >> Add user counter allowing to delete pool only when no users.
>> >> It doesn't prevent pool from flush, only prevents freeing the
>> >> pool instance. Helps when no need to delete the pool and now
>> >> it's user responsibility to free it by calling page_pool_free()
>> >> while destroying procedure. It also makes to use page_pool_free()
>> >> explicitly, not fully hidden in xdp unreg, which looks more
>> >> correct after page pool "create" routine.
>> >
>> >No, this is wrong.
>> below.
>>
>> >
>> >> Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
>> >> ---
>> >> drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 8 +++++---
>> >> include/net/page_pool.h | 7 +++++++
>> >> net/core/page_pool.c | 7 +++++++
>> >> net/core/xdp.c | 3 +++
>> >> 4 files changed, 22 insertions(+), 3 deletions(-)
>> >>
>> >> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
>> >> index 5e40db8f92e6..cb028de64a1d 100644
>> >> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
>> >> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
>> >> @@ -545,10 +545,8 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c,
>> >> }
>> >> err = xdp_rxq_info_reg_mem_model(&rq->xdp_rxq,
>> >> MEM_TYPE_PAGE_POOL, rq->page_pool);
>> >> - if (err) {
>> >> - page_pool_free(rq->page_pool);
>> >> + if (err)
>> >> goto err_free;
>> >> - }
>> >>
>> >> for (i = 0; i < wq_sz; i++) {
>> >> if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) {
>> >> @@ -613,6 +611,8 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c,
>> >> if (rq->xdp_prog)
>> >> bpf_prog_put(rq->xdp_prog);
>> >> xdp_rxq_info_unreg(&rq->xdp_rxq);
>> >> + if (rq->page_pool)
>> >> + page_pool_free(rq->page_pool);
>> >> mlx5_wq_destroy(&rq->wq_ctrl);
>> >>
>> >> return err;
>> >> @@ -643,6 +643,8 @@ static void mlx5e_free_rq(struct mlx5e_rq *rq)
>> >> }
>> >>
>> >> xdp_rxq_info_unreg(&rq->xdp_rxq);
>> >> + if (rq->page_pool)
>> >> + page_pool_free(rq->page_pool);
>> >
>> >No, this is wrong. The hole point with the merged page_pool fixes
>> >patchset was that page_pool_free() needs to be delayed until no-more
>> >in-flight packets exist.
>>
>> Probably it's not so obvious, but it's still delayed and deleted only
>> after no-more in-flight packets exist. Here question is only who is able
>> to do this first based on refcnt.
>
>Hmm... then I find this API is rather misleading, even the function
>name page_pool_free is misleading ("free"). (Now, I do see, below, that
>page_pool_create() take an extra reference).
In feneral "free" looks not bad after "create".
It's called after "create" if some error with registering it rxq.
and it looks logical, if it's called after no need in pool.
obj = create()
/* a lot of different stuff */
free(obj);
>
>But it is still wrong / problematic. As you allow
>__page_pool_request_shutdown() to be called with elevated refcnt. Your
>use-case is to have more than 1 xdp_rxq_info struct using the same
>page_pool. Then you have to call xdp_rxq_info_unreg_mem_model() for
>each, which will call __page_pool_request_shutdown().
>
>For this to be safe, your driver have to stop RX for all the
>xdp_rxq_info structs that share the page_pool. The page_pool already
>have this requirement, but it comes as natural step when shutting down
>an RXQ. With your change, you have to take care of stopping the RXQs
>first, and then call xdp_rxq_info_unreg_mem_model() for each
>xdp_rxq_info afterwards. I assume you do this, but it is just a driver
>bug waiting to happen.
All rxq queues are stopped before this, and only after this the pools are freed,
exactly as it required for one xdp_rxq_info_unreg_mem_model(), w/o exclusions,
as it requires the API.
>
>> >> diff --git a/net/core/page_pool.c b/net/core/page_pool.c
>> >> index b366f59885c1..169b0e3c870e 100644
>> >> --- a/net/core/page_pool.c
>> >> +++ b/net/core/page_pool.c
>[...]
>> >> @@ -70,6 +71,8 @@ struct page_pool *page_pool_create(const struct page_pool_params *params)
>> >> kfree(pool);
>> >> return ERR_PTR(err);
>> >> }
>> >> +
>> >> + page_pool_get(pool);
>> >> return pool;
>> >> }
>> >> EXPORT_SYMBOL(page_pool_create);
>
>The thing (perhaps) like about your API change, is that you also allow
>the driver to explicitly keep the page_pool object across/after a
>xdp_rxq_info_unreg_mem_model(). And this way possibly reuse it for
>another RXQ.
>The problem is of-cause that on driver shutdown, this
>will force drivers to implement the same shutdown logic with
>schedule_delayed_work as the core xdp.c code already does.
I see.
The cpsw dosn't re-use it, so here all is fine, but if a driver needs
to re-use it again, lets suppose, as it can happen, the pool needs to
be registered with xdp_rxq_info_reg_mem_model() again, and for that
potentially can be added verification on in-flight packets
or some register state...but better mention in some place
to not do this, frankly, I don't know where it should be at this moment.
--
Regards,
Ivan Khoronzhuk
^ permalink raw reply
* · can you assist me?
From: Laura @ 2019-06-26 11:32 UTC (permalink / raw)
To: netdev
Hi,
Did you see my email below from last week?
We manufacture ALL custom LOGO and branded products – over 300,000 to
choose from.
The most asked about product that we make, are the custom printed USB flash
drives!
We can print your logo on them and load your digital images, videos and
files!
http://inaflashweb.com
Here is what we include:
-Any size memory you need: 64MB up to 128GB
-We will print your logo on both sides, just ask!
-Very Low Order Minimums
-Need them quickly? Not a problem, we offer Rush Service
http://inaflashweb.com
NEW: We can make a custom shaped USB drive to look like your Logo or
product!
Email over a copy of your logo and we will create a design mock up for you
at no cost!
Our higher memory sizes are a really good option right now!
Ask about the “Double Your Memory” upgrade promotion going on right
now!
Pricing is low right now, so let us know what you need and we will get you
a quick quote.
We will beat any competitors pricing, send us your last invoice and we will
beat it!
We always offer great rates for schools and nonprofits as well.
Let us know what you would like quoted?
http://inaflashweb.com
Regards,
Sabrina Millons
+1-888-480-8218
Custom USB Account Manager
http://inaflashweb.com
^ permalink raw reply
* · can you assist me?
From: Laura @ 2019-06-26 11:16 UTC (permalink / raw)
To: netdev
Hi,
Did you see my email below from last week?
We manufacture ALL custom LOGO and branded products – over 300,000 to
choose from.
The most asked about product that we make, are the custom printed USB flash
drives!
We can print your logo on them and load your digital images, videos and
files!
http://inaflashweb.com
Here is what we include:
-Any size memory you need: 64MB up to 128GB
-We will print your logo on both sides, just ask!
-Very Low Order Minimums
-Need them quickly? Not a problem, we offer Rush Service
http://inaflashweb.com
NEW: We can make a custom shaped USB drive to look like your Logo or
product!
Email over a copy of your logo and we will create a design mock up for you
at no cost!
Our higher memory sizes are a really good option right now!
Ask about the “Double Your Memory” upgrade promotion going on right
now!
Pricing is low right now, so let us know what you need and we will get you
a quick quote.
We will beat any competitors pricing, send us your last invoice and we will
beat it!
We always offer great rates for schools and nonprofits as well.
Let us know what you would like quoted?
http://inaflashweb.com
Regards,
Sabrina Millons
+1-888-480-8218
Custom USB Account Manager
http://inaflashweb.com
^ permalink raw reply
* Re: [PATCH nf-next v2 2/2] netfilter: nft_meta: Add NFT_META_BRI_VLAN support
From: wenxu @ 2019-06-26 12:42 UTC (permalink / raw)
To: Pablo Neira Ayuso; +Cc: fw, netfilter-devel, netdev
In-Reply-To: <20190626102935.ztxcfb3kysvohzi3@salvia>
I agree with you, It's a more generic way to set the vlan tag not base on
any bridge. I will resubmit NFT_META_BRI_VLAN_PROTO and
NFT_META_VLAN patches
在 2019/6/26 18:29, Pablo Neira Ayuso 写道:
> Could you add a NFT_META_BRI_VLAN_PROTO? Similar to patch 1/2, to
> retrieve p->br->vlan_proto.
>
> Then, add a generic way to set the vlan metadata. I'm attaching an
> incomplete patch, so there is something like:
>
> meta vlan set 0x88a8:20
>
> to set q-in-q.
>
> we could also add a shortcut for simple vlan case (no q-in-q), ie.
> assuming protocol is 0x8100:
>
> meta vlan set 20
>
> Does this make sense to you?
>
> And we have a way to set the meta vlan information from ingress to
> then, which is something I also need here.
>
> Thanks.
^ permalink raw reply
* RE: [PATCH net-next 0/2] net: ipv4: remove erroneous advancement of list pointer
From: Ran Rozenstein @ 2019-06-26 12:48 UTC (permalink / raw)
To: Florian Westphal
Cc: Tariq Toukan, netdev@vger.kernel.org, Maor Gottlieb,
edumazet@google.com
In-Reply-To: <20190625091903.gepfjgpiksslnyqy@breakpoint.cc>
> -----Original Message-----
> From: Florian Westphal [mailto:fw@strlen.de]
> Sent: Tuesday, June 25, 2019 12:19
> To: Ran Rozenstein <ranro@mellanox.com>
> Cc: Tariq Toukan <tariqt@mellanox.com>; Florian Westphal <fw@strlen.de>;
> netdev@vger.kernel.org; Maor Gottlieb <maorg@mellanox.com>;
> edumazet@google.com
> Subject: Re: [PATCH net-next 0/2] net: ipv4: remove erroneous
> advancement of list pointer
>
> Ran Rozenstein <ranro@mellanox.com> wrote:
> > > On 6/17/2019 5:02 PM, Florian Westphal wrote:
> > > > Tariq reported a soft lockup on net-next that Mellanox was able to
> > > > bisect to 2638eb8b50cf ("net: ipv4: provide __rcu annotation for
> ifa_list").
> > > >
> > > > While reviewing above patch I found a regression when addresses
> > > > have a lifetime specified.
> > > >
> > > > Second patch extends rtnetlink.sh to trigger crash (without first
> > > > patch applied).
> > > >
> > >
> > > Thanks Florian.
> > >
> > > Ran, can you please test?
> >
> > Tested, still reproduce.
>
> Can you be a little more specific? Is there any reproducer?
The test dose stress on the interface by running this 2 commands in loop:
command is: /sbin/ip -f inet addr add $IP/16 brd + dev ens8f1
command is: ifconfig ens8f1 $IP netmask 255.255.0.0
when $IP change every iteration.
It execute every second when we see the reproduce somewhere between 40 to 200 seconds of execution.
Thanks,
Ran
^ permalink raw reply
* Re: net: never suspend the ethernet PHY on certain boards?
From: Andrew Lunn @ 2019-06-26 12:55 UTC (permalink / raw)
To: Alexander Dahl; +Cc: netdev, linux-kernel, Thomas Pfahl
In-Reply-To: <4693980.Yko7hG0E1C@ada>
> What I could do:
>
> 1) Revert that change on my tree, which would mean reverting a generic bugfix
> 2) Patch smsc phy driver to not suspend anymore
> 3) Invent some new way to prevent suspend on a configuration basis (dt?)
> 4) Anything I did not think of yet
5) Enable WoL?
There are other boards which have PHY clock issues. Let me check how
they work around this.
Andrew
^ permalink raw reply
* Re: [PATCH nf-next v2 2/2] netfilter: nft_meta: Add NFT_META_BRI_VLAN support
From: Pablo Neira Ayuso @ 2019-06-26 12:57 UTC (permalink / raw)
To: wenxu; +Cc: fw, netfilter-devel, netdev
In-Reply-To: <b037a0a9-4729-41ff-81bb-ca76c0e3fba9@ucloud.cn>
On Wed, Jun 26, 2019 at 08:42:27PM +0800, wenxu wrote:
> I agree with you, It's a more generic way to set the vlan tag not base on
>
> any bridge. I will resubmit NFT_META_BRI_VLAN_PROTO and
>
> NFT_META_VLAN patches
Thank you very much.
^ permalink raw reply
* [PATCH resend] can: rcar_canfd: fix possible IRQ storm on high load
From: Nikita Yushchenko @ 2019-06-26 13:08 UTC (permalink / raw)
To: Wolfgang Grandegger, Marc Kleine-Budde, David S. Miller,
Simon Horman, Wolfram Sang
Cc: linux-can, netdev, linux-kernel, Artemi Ivanov, Sergei Shtylyov,
Nikita Yushchenko
We have observed rcar_canfd driver entering IRQ storm under high load,
with following scenario:
- rcar_canfd_global_interrupt() in entered due to Rx available,
- napi_schedule_prep() is called, and sets NAPIF_STATE_SCHED in state
- Rx fifo interrupts are masked,
- rcar_canfd_global_interrupt() is entered again, this time due to
error interrupt (e.g. due to overflow),
- since scheduled napi poller has not yet executed, condition for calling
napi_schedule_prep() from rcar_canfd_global_interrupt() remains true,
thus napi_schedule_prep() gets called and sets NAPIF_STATE_MISSED flag
in state,
- later, napi poller function rcar_canfd_rx_poll() gets executed, and
calls napi_complete_done(),
- due to NAPIF_STATE_MISSED flag in state, this call does not clear
NAPIF_STATE_SCHED flag from state,
- on return from napi_complete_done(), rcar_canfd_rx_poll() unmasks Rx
interrutps,
- Rx interrupt happens, rcar_canfd_global_interrupt() gets called
and calls napi_schedule_prep(),
- since NAPIF_STATE_SCHED is set in state at this time, this call
returns false,
- due to that false return, rcar_canfd_global_interrupt() returns
without masking Rx interrupt
- and this results into IRQ storm: unmasked Rx interrupt happens again
and again is misprocessed in the same way.
This patch fixes that scenario by unmasking Rx interrupts only when
napi_complete_done() returns true, which means it has cleared
NAPIF_STATE_SCHED in state.
Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
---
drivers/net/can/rcar/rcar_canfd.c | 9 +++++----
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/drivers/net/can/rcar/rcar_canfd.c b/drivers/net/can/rcar/rcar_canfd.c
index 05410008aa6b..de34a4b82d4a 100644
--- a/drivers/net/can/rcar/rcar_canfd.c
+++ b/drivers/net/can/rcar/rcar_canfd.c
@@ -1508,10 +1508,11 @@ static int rcar_canfd_rx_poll(struct napi_struct *napi, int quota)
/* All packets processed */
if (num_pkts < quota) {
- napi_complete_done(napi, num_pkts);
- /* Enable Rx FIFO interrupts */
- rcar_canfd_set_bit(priv->base, RCANFD_RFCC(ridx),
- RCANFD_RFCC_RFIE);
+ if (napi_complete_done(napi, num_pkts)) {
+ /* Enable Rx FIFO interrupts */
+ rcar_canfd_set_bit(priv->base, RCANFD_RFCC(ridx),
+ RCANFD_RFCC_RFIE);
+ }
}
return num_pkts;
}
--
2.11.0
^ permalink raw reply related
* Re: [PATCH resend] can: rcar_canfd: fix possible IRQ storm on high load
From: Wolfram Sang @ 2019-06-26 13:12 UTC (permalink / raw)
To: Nikita Yushchenko, Ramesh Shanmugasundaram
Cc: Wolfgang Grandegger, Marc Kleine-Budde, David S. Miller,
Simon Horman, Wolfram Sang, linux-can, netdev, linux-kernel,
Artemi Ivanov, Sergei Shtylyov
In-Reply-To: <20190626130848.6671-1-nikita.yoush@cogentembedded.com>
[-- Attachment #1: Type: text/plain, Size: 2605 bytes --]
On Wed, Jun 26, 2019 at 04:08:48PM +0300, Nikita Yushchenko wrote:
> We have observed rcar_canfd driver entering IRQ storm under high load,
> with following scenario:
> - rcar_canfd_global_interrupt() in entered due to Rx available,
> - napi_schedule_prep() is called, and sets NAPIF_STATE_SCHED in state
> - Rx fifo interrupts are masked,
> - rcar_canfd_global_interrupt() is entered again, this time due to
> error interrupt (e.g. due to overflow),
> - since scheduled napi poller has not yet executed, condition for calling
> napi_schedule_prep() from rcar_canfd_global_interrupt() remains true,
> thus napi_schedule_prep() gets called and sets NAPIF_STATE_MISSED flag
> in state,
> - later, napi poller function rcar_canfd_rx_poll() gets executed, and
> calls napi_complete_done(),
> - due to NAPIF_STATE_MISSED flag in state, this call does not clear
> NAPIF_STATE_SCHED flag from state,
> - on return from napi_complete_done(), rcar_canfd_rx_poll() unmasks Rx
> interrutps,
> - Rx interrupt happens, rcar_canfd_global_interrupt() gets called
> and calls napi_schedule_prep(),
> - since NAPIF_STATE_SCHED is set in state at this time, this call
> returns false,
> - due to that false return, rcar_canfd_global_interrupt() returns
> without masking Rx interrupt
> - and this results into IRQ storm: unmasked Rx interrupt happens again
> and again is misprocessed in the same way.
>
> This patch fixes that scenario by unmasking Rx interrupts only when
> napi_complete_done() returns true, which means it has cleared
> NAPIF_STATE_SCHED in state.
>
> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
CCing the driver author...
> ---
> drivers/net/can/rcar/rcar_canfd.c | 9 +++++----
> 1 file changed, 5 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/net/can/rcar/rcar_canfd.c b/drivers/net/can/rcar/rcar_canfd.c
> index 05410008aa6b..de34a4b82d4a 100644
> --- a/drivers/net/can/rcar/rcar_canfd.c
> +++ b/drivers/net/can/rcar/rcar_canfd.c
> @@ -1508,10 +1508,11 @@ static int rcar_canfd_rx_poll(struct napi_struct *napi, int quota)
>
> /* All packets processed */
> if (num_pkts < quota) {
> - napi_complete_done(napi, num_pkts);
> - /* Enable Rx FIFO interrupts */
> - rcar_canfd_set_bit(priv->base, RCANFD_RFCC(ridx),
> - RCANFD_RFCC_RFIE);
> + if (napi_complete_done(napi, num_pkts)) {
> + /* Enable Rx FIFO interrupts */
> + rcar_canfd_set_bit(priv->base, RCANFD_RFCC(ridx),
> + RCANFD_RFCC_RFIE);
> + }
> }
> return num_pkts;
> }
> --
> 2.11.0
>
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
^ permalink raw reply
* Re: [PATCH net-next 04/12] net: sched: add tcf_block_setup()
From: Pablo Neira Ayuso @ 2019-06-26 13:16 UTC (permalink / raw)
To: Jiri Pirko
Cc: netdev, netfilter-devel, davem, thomas.lendacky, f.fainelli,
ariel.elior, michael.chan, santosh, madalin.bucur, yisen.zhuang,
salil.mehta, jeffrey.t.kirsher, tariqt, saeedm, jiri, idosch,
jakub.kicinski, peppe.cavallaro, grygorii.strashko, andrew,
vivien.didelot, alexandre.torgue, joabreu, linux-net-drivers,
ganeshgr, ogerlitz, Manish.Chopra, marcelo.leitner, mkubecek,
venkatkumar.duvvuru, cphealy
In-Reply-To: <20190626121256.GA2424@nanopsycho>
On Wed, Jun 26, 2019 at 02:12:56PM +0200, Jiri Pirko wrote:
> Tue, Jun 25, 2019 at 10:31:54AM CEST, pablo@netfilter.org wrote:
> >On Fri, Jun 21, 2019 at 07:16:03PM +0200, Jiri Pirko wrote:
> >> Thu, Jun 20, 2019 at 09:49:09PM CEST, pablo@netfilter.org wrote:
> >>
> >> [...]
> >>
> >> >
> >> >+static LIST_HEAD(tcf_block_cb_list);
> >>
> >> I still don't like the global list. Have to go throught the code more
> >> carefully, but why you can't pass the priv/ctx from tc/netfilter. From
> >> tc it would be tcf_block as it is now, from netfilter something else.
> >
> >This tcf_block_cb_list should go away at some point, once drivers know
> >how to deal with multiple subsystems using the setup block
> >infrastructure. As I said in my previous email, only one can set up
> >the block at this stage, the ones coming later will hit busy.
>
> The driver should know if it can bind or is busy. Also, the bind cmd
> should contain type of binder (tc/nft/whatever) or perhaps rather binder
> priority (according to the hook order in rx/tx).
OK, so I see two possible paths then:
#1 Add global list and allow one single subsystem to bind by now. Then
later, in a follow up patchset. Add binder type and priority once
there is a driver that can handle the three subsystems, remove
this global list and each driver deals/knows what to do from the
binder path.
#2 Remove the global list now, each driver maintains a list of flow blocks
internally, allow one single flow block by now. This will need a bit more
code, since there will be code in the driver to maintain the list of
existing flow blocks, per driver, instead of global. So it will be
a per-driver global local to check if there is a flow block with
this [ cb, cb_ident ] already in place.
#1 is almost ready - it's this batch :-) - then #2 may need more code -
this batch is slightly large.
I understand though that path #2 may make it easier for the first
driver client allowing for the three subsystems to bind.
Let me know what path your prefer.
Thanks for reviewing.
^ permalink raw reply
* RE: [PATCH net-next 01/16] qlge: Remove irq_cnt
From: Manish Chopra @ 2019-06-26 13:21 UTC (permalink / raw)
To: Benjamin Poirier; +Cc: GR-Linux-NIC-Dev, netdev@vger.kernel.org
In-Reply-To: <20190626113619.GA27420@f1>
> In msix mode there's no need to explicitly disable completion interrupts, they
> are reliably auto-masked, according to my observations.
> I tested this on two QLE8142 adapters.
>
> Do you have reason to believe this might not always be the case?
How did you check auto-masking of MSI-X interrupts ?
I was just wondering about the below comment in ql_disable_completion_interrupt(), where for MSI-X it does disable completion intr for zeroth intr.
Seems special case for zeroth intr in MSI-X particular to this device.
/* HW disables for us if we're MSIX multi interrupts and
* it's not the default (zeroeth) interrupt.
*/
if (likely(test_bit(QL_MSIX_ENABLED, &qdev->flags) && intr))
return 0;
^ permalink raw reply
* Re: [PATCH] rt2x00: fix rx queue hang
From: Soeren Moch @ 2019-06-26 13:28 UTC (permalink / raw)
To: Stanislaw Gruszka
Cc: Helmut Schaa, Kalle Valo, David S. Miller, linux-wireless, netdev,
linux-kernel, stable
In-Reply-To: <20190625095734.GA2886@redhat.com>
On 25.06.19 11:57, Stanislaw Gruszka wrote:
> Hello
>
> On Fri, Jun 21, 2019 at 01:30:01PM +0200, Soeren Moch wrote:
>> On 18.06.19 11:34, Stanislaw Gruszka wrote:
>>> Hi
>>>
>>> On Mon, Jun 17, 2019 at 11:46:56AM +0200, Soeren Moch wrote:
>>>> Since commit ed194d136769 ("usb: core: remove local_irq_save() around
>>>> ->complete() handler") the handlers rt2x00usb_interrupt_rxdone() and
>>>> rt2x00usb_interrupt_txdone() are not running with interrupts disabled
>>>> anymore. So these handlers are not guaranteed to run completely before
>>>> workqueue processing starts. So only mark entries ready for workqueue
>>>> processing after proper accounting in the dma done queue.
>>> It was always the case on SMP machines that rt2x00usb_interrupt_{tx/rx}done
>>> can run concurrently with rt2x00_work_{rx,tx}done, so I do not
>>> understand how removing local_irq_save() around complete handler broke
>>> things.
>> I think because completion handlers can be interrupted now and scheduled
>> away
>> in the middle of processing.
>>> Have you reverted commit ed194d136769 and the revert does solve the problem ?
>> Yes, I already sent a patch for this, see [1]. But this was not considered
>> an acceptablesolution. Especially RT folks do not like code running with
>> interrupts disabled,particularly when trying to acquire spinlocks then.
>>
>> [1] https://lkml.org/lkml/2019/5/31/863
>>> Between 4.19 and 4.20 we have some quite big changes in rt2x00 driver:
>>>
>>> 0240564430c0 rt2800: flush and txstatus rework for rt2800mmio
>>> adf26a356f13 rt2x00: use different txstatus timeouts when flushing
>>> 5022efb50f62 rt2x00: do not check for txstatus timeout every time on tasklet
>>> 0b0d556e0ebb rt2800mmio: use txdone/txstatus routines from lib
>>> 5c656c71b1bf rt2800: move usb specific txdone/txstatus routines to rt2800lib
>>>
>>> so I'm a bit afraid that one of those changes is real cause of
>>> the issue not ed194d136769 .
>> I tested 4.20 and 5.1 and see the exact same behavior. Reverting this
>> usb core patchsolves the problem.
>> 4.19.x (before this usb core patch) is running fine.
>>>> Note that rt2x00usb_work_rxdone() processes all available entries, not
>>>> only such for which queue_work() was called.
>>>>
>>>> This fixes a regression on a RT5370 based wifi stick in AP mode, which
>>>> suddenly stopped data transmission after some period of heavy load. Also
>>>> stopping the hanging hostapd resulted in the error message "ieee80211
>>>> phy0: rt2x00queue_flush_queue: Warning - Queue 14 failed to flush".
>>>> Other operation modes are probably affected as well, this just was
>>>> the used testcase.
>>> Do you know what actually make the traffic stop,
>>> TX queue hung or RX queue hung?
>> I think RX queue hang, as stated in the patch title. "Queue 14" means QID_RX
>> (rt2x00queue.h, enum data_queue_qid).
>> I also tried to re-add local_irq_save() in only one of the handlers. Adding
>> this tort2x00usb_interrupt_rxdone() alone solved the issue, while doing so
>> for tx alonedid not.
>>
>> Note that this doesn't mean there is no problem for tx, that's maybe
>> just more
>> difficult to trigger.
>>>> diff --git a/drivers/net/wireless/ralink/rt2x00/rt2x00dev.c b/drivers/net/wireless/ralink/rt2x00/rt2x00dev.c
>>>> index 1b08b01db27b..9c102a501ee6 100644
>>>> --- a/drivers/net/wireless/ralink/rt2x00/rt2x00dev.c
>>>> +++ b/drivers/net/wireless/ralink/rt2x00/rt2x00dev.c
>>>> @@ -263,9 +263,9 @@ EXPORT_SYMBOL_GPL(rt2x00lib_dmastart);
>>>>
>>>> void rt2x00lib_dmadone(struct queue_entry *entry)
>>>> {
>>>> - set_bit(ENTRY_DATA_STATUS_PENDING, &entry->flags);
>>>> clear_bit(ENTRY_OWNER_DEVICE_DATA, &entry->flags);
>>>> rt2x00queue_index_inc(entry, Q_INDEX_DMA_DONE);
>>>> + set_bit(ENTRY_DATA_STATUS_PENDING, &entry->flags);
>>> Unfortunately I do not understand how this suppose to fix the problem,
>>> could you elaborate more about this change?
>>>
>> Re-adding local_irq_save() around thisrt2x00lib_dmadone()solved
>> the issue. So I also tried to reverse the order of these calls.
>> It seems totally plausible to me, that the correct sequence is to
>> first clear the device assignment, then to set the status to dma_done,
>> then to trigger the workqueue processing for this entry. When the handler
>> is scheduled away in the middle of this sequence, now there is no
>> strange state where the entry can be processed by the workqueue while
>> not declared dma_done for it.
>> With this changed sequence there is no need anymore to disable interrupts
>> for solving the hang issue.
> Thanks very much for explanations. However I still do not fully
> understand the issue. Q_INDEX_DMA_DONE index is only checked on TX
> processing (on RX we use only Q_INDEX_DONE and Q_INDEX) and
> ENTRY_OWNER_DEVICE_DATA is already cleared before rt2x00lib_dmadone()
> in rt2x00usb_interrupt_rxdone() .
>
> So I'm not sure how changing the order solve the problem. Looks
> for me that the issue is triggered by some rt2x00lib_dmadone()
> call done on error path (not in rt2x00usb_interrupt_rxdone())
> and it race with this check:
>
> if (test_and_set_bit(ENTRY_OWNER_DEVICE_DATA, &entry->flags) ||
> test_bit(ENTRY_DATA_STATUS_PENDING, &entry->flags))
> return false;
>
> in rt2x00usb_kick_rx_entry() - we return instead of submit urb.
Hi Stanislaw,
the good news is: your patch below also solves the issue for me. But
removing the ENTRY_DATA_STATUS_PENDING check in
rt2x00usb_kick_rx_entry() alone does not help, while removing this check
in rt2x00usb_work_rxdone() alone does the trick.
So the real race seems to be that the flags set in the completion
handler are not yet visible on the cpu core running the workqueue. And
because the worker is not rescheduled when aborted, the entry can just
wait forever.
Do you think this could make sense?
> I'm somewhat reluctant to change the order, because TX processing
> might relay on it (we first mark we wait for TX status and
> then mark entry is no longer owned by hardware).
OK, maybe it's just good luck that changing the order solves the rx
problem. Or can memory barriers associated with the spinlock in
rt2x00lib_dmadone() be responsible for that?
(I'm testing on a armv7 system, Cortex-A9 quadcore.)
While looking at it, why we double-clear ENTRY_OWNER_DEVICE_DATA in
rt2x00usb_interrupt_rxdone() directly and in rt2x00lib_dmadone() again,
while not doing the same for tx? Would it make more sense to possibly
set ENTRY_DATA_IO_FAILED before clearing ENTRY_OWNER_DEVICE_DATA in
rt2x00usb_interrupt_rxdone() as for tx?
> However on RX
> side ENTRY_DATA_STATUS_PENDING bit make no sense as we do not
> wait for status. We should remove ENTRY_DATA_STATUS_PENDING on
> RX side and perhaps this also will solve issue you observe.
I agree that removing the unnecessary checks is a good idea in any case.
> Could you please check below patch, if it fixes the problem as well?
At least I could not trigger the problem within transferring 10GB of
data. But maybe the probability for triggering this bug is just lower
because ENTRY_OWNER_DEVICE_DATA is cleared some time before
ENTRY_DATA_STATUS_PENDING is set?
Soeren
>
> Stanislaw
>
> diff --git a/drivers/net/wireless/ralink/rt2x00/rt2x00usb.c b/drivers/net/wireless/ralink/rt2x00/rt2x00usb.c
> index b6c1344..731e633 100644
> --- a/drivers/net/wireless/ralink/rt2x00/rt2x00usb.c
> +++ b/drivers/net/wireless/ralink/rt2x00/rt2x00usb.c
> @@ -360,8 +360,7 @@ static void rt2x00usb_work_rxdone(struct work_struct *work)
> while (!rt2x00queue_empty(rt2x00dev->rx)) {
> entry = rt2x00queue_get_entry(rt2x00dev->rx, Q_INDEX_DONE);
>
> - if (test_bit(ENTRY_OWNER_DEVICE_DATA, &entry->flags) ||
> - !test_bit(ENTRY_DATA_STATUS_PENDING, &entry->flags))
> + if (test_bit(ENTRY_OWNER_DEVICE_DATA, &entry->flags))
> break;
>
> /*
> @@ -413,8 +412,7 @@ static bool rt2x00usb_kick_rx_entry(struct queue_entry *entry, void *data)
> struct queue_entry_priv_usb *entry_priv = entry->priv_data;
> int status;
>
> - if (test_and_set_bit(ENTRY_OWNER_DEVICE_DATA, &entry->flags) ||
> - test_bit(ENTRY_DATA_STATUS_PENDING, &entry->flags))
> + if (test_and_set_bit(ENTRY_OWNER_DEVICE_DATA, &entry->flags))
> return false;
>
> rt2x00lib_dmastart(entry);
^ permalink raw reply
* Re: [PATCH bpf-next 1/4] bpf: unprivileged BPF access via /dev/bpf
From: Daniel Borkmann @ 2019-06-26 13:32 UTC (permalink / raw)
To: Song Liu, netdev, bpf; +Cc: ast, kernel-team, jannh
In-Reply-To: <20190625182303.874270-2-songliubraving@fb.com>
On 06/25/2019 08:23 PM, Song Liu wrote:
> This patch introduce unprivileged BPF access. The access control is
> achieved via device /dev/bpf. Users with access to /dev/bpf are able
> to access BPF syscall.
>
> Two ioctl command are added to /dev/bpf:
>
> The first two commands get/put permission to access sys_bpf. This
> permission is noted by setting bit TASK_BPF_FLAG_PERMITTED of
> current->bpf_flags. This permission cannot be inherited via fork().
>
> Helper function bpf_capable() is added to check whether the task has got
> permission via /dev/bpf.
>
> Signed-off-by: Song Liu <songliubraving@fb.com>
[ Lets Cc Jann so he has a chance to review as he was the one who suggested
the idea. ]
> ---
> Documentation/ioctl/ioctl-number.txt | 1 +
> include/linux/bpf.h | 12 +++++
> include/linux/sched.h | 8 ++++
> include/uapi/linux/bpf.h | 5 ++
> kernel/bpf/arraymap.c | 2 +-
> kernel/bpf/cgroup.c | 2 +-
> kernel/bpf/core.c | 4 +-
> kernel/bpf/cpumap.c | 2 +-
> kernel/bpf/devmap.c | 2 +-
> kernel/bpf/hashtab.c | 4 +-
> kernel/bpf/lpm_trie.c | 2 +-
> kernel/bpf/offload.c | 2 +-
> kernel/bpf/queue_stack_maps.c | 2 +-
> kernel/bpf/reuseport_array.c | 2 +-
> kernel/bpf/stackmap.c | 2 +-
> kernel/bpf/syscall.c | 72 +++++++++++++++++++++-------
> kernel/bpf/verifier.c | 2 +-
> kernel/bpf/xskmap.c | 2 +-
> kernel/fork.c | 4 ++
> net/core/filter.c | 6 +--
> 20 files changed, 104 insertions(+), 34 deletions(-)
>
> diff --git a/Documentation/ioctl/ioctl-number.txt b/Documentation/ioctl/ioctl-number.txt
> index c9558146ac58..19998b99d603 100644
> --- a/Documentation/ioctl/ioctl-number.txt
> +++ b/Documentation/ioctl/ioctl-number.txt
> @@ -327,6 +327,7 @@ Code Seq#(hex) Include File Comments
> 0xB4 00-0F linux/gpio.h <mailto:linux-gpio@vger.kernel.org>
> 0xB5 00-0F uapi/linux/rpmsg.h <mailto:linux-remoteproc@vger.kernel.org>
> 0xB6 all linux/fpga-dfl.h
> +0xBP 01-02 uapi/linux/bpf.h <mailto:bpf@vger.kernel.org>
> 0xC0 00-0F linux/usb/iowarrior.h
> 0xCA 00-0F uapi/misc/cxl.h
> 0xCA 10-2F uapi/misc/ocxl.h
> diff --git a/include/linux/bpf.h b/include/linux/bpf.h
> index a62e7889b0b6..dbba7870f6df 100644
> --- a/include/linux/bpf.h
> +++ b/include/linux/bpf.h
> @@ -14,6 +14,10 @@
> #include <linux/numa.h>
> #include <linux/wait.h>
> #include <linux/u64_stats_sync.h>
> +#include <linux/sched.h>
> +#include <linux/capability.h>
> +
> +#include <asm/current.h>
>
> struct bpf_verifier_env;
> struct perf_event;
> @@ -742,6 +746,12 @@ int bpf_prog_test_run_skb(struct bpf_prog *prog, const union bpf_attr *kattr,
> int bpf_prog_test_run_flow_dissector(struct bpf_prog *prog,
> const union bpf_attr *kattr,
> union bpf_attr __user *uattr);
> +
> +static inline bool bpf_capable(int cap)
> +{
> + return test_bit(TASK_BPF_FLAG_PERMITTED, ¤t->bpf_flags) ||
> + capable(cap);
> +}
> #else /* !CONFIG_BPF_SYSCALL */
> static inline struct bpf_prog *bpf_prog_get(u32 ufd)
> {
> @@ -874,6 +884,8 @@ static inline int bpf_prog_test_run_flow_dissector(struct bpf_prog *prog,
> {
> return -ENOTSUPP;
> }
> +
> +#define bpf_capable(cap) capable((cap))
> #endif /* CONFIG_BPF_SYSCALL */
>
> static inline struct bpf_prog *bpf_prog_get_type(u32 ufd,
> diff --git a/include/linux/sched.h b/include/linux/sched.h
> index 11837410690f..ddd33d4476c5 100644
> --- a/include/linux/sched.h
> +++ b/include/linux/sched.h
> @@ -1200,6 +1200,10 @@ struct task_struct {
> unsigned long prev_lowest_stack;
> #endif
>
> +#ifdef CONFIG_BPF_SYSCALL
> + unsigned long bpf_flags;
> +#endif
There are plenty of bits available here:
/* --- cacheline 14 boundary (896 bytes) --- */
unsigned int in_execve:1; /* 896:31 4 */
unsigned int in_iowait:1; /* 896:30 4 */
unsigned int restore_sigmask:1; /* 896:29 4 */
unsigned int in_user_fault:1; /* 896:28 4 */
unsigned int no_cgroup_migration:1; /* 896:27 4 */
unsigned int frozen:1; /* 896:26 4 */
unsigned int use_memdelay:1; /* 896:25 4 */
/* XXX 25 bits hole, try to pack */
/* XXX 4 bytes hole, try to pack */
Given that bpf is pretty much enabled by default everywhere, I don't think we
should waste so much space in task_struct just for this flag (pretty sure that
task_struct is the equivalent of sk_buff that rather needs a diet). Other options
could be to add to atomic_flags which also still has space.
> /*
> * New fields for task_struct should be added above here, so that
> * they are included in the randomized portion of task_struct.
> @@ -1772,6 +1776,10 @@ static inline void set_task_cpu(struct task_struct *p, unsigned int cpu)
>
> #endif /* CONFIG_SMP */
>
> +enum task_struct_bpf_flags {
> + TASK_BPF_FLAG_PERMITTED,
> +};
> +
> /*
> * In order to reduce various lock holder preemption latencies provide an
> * interface to see if a vCPU is currently running or not.
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index b077507efa3f..ec3ae452cfd7 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -3541,4 +3541,9 @@ struct bpf_sysctl {
> */
> };
>
> +#define BPF_IOCTL 0xBF
> +
> +#define BPF_DEV_IOCTL_GET_PERM _IO(BPF_IOCTL, 0x01)
> +#define BPF_DEV_IOCTL_PUT_PERM _IO(BPF_IOCTL, 0x02)
> +
> #endif /* _UAPI__LINUX_BPF_H__ */
> diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c
> index 1c65ce0098a9..9ae668fa9185 100644
> --- a/kernel/bpf/arraymap.c
> +++ b/kernel/bpf/arraymap.c
> @@ -73,7 +73,7 @@ static struct bpf_map *array_map_alloc(union bpf_attr *attr)
> bool percpu = attr->map_type == BPF_MAP_TYPE_PERCPU_ARRAY;
> int ret, numa_node = bpf_map_attr_numa_node(attr);
> u32 elem_size, index_mask, max_entries;
> - bool unpriv = !capable(CAP_SYS_ADMIN);
> + bool unpriv = !bpf_capable(CAP_SYS_ADMIN);
> u64 cost, array_size, mask64;
> struct bpf_map_memory mem;
> struct bpf_array *array;
> diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
> index c225c42e114a..fd9bea70f8f3 100644
> --- a/kernel/bpf/cgroup.c
> +++ b/kernel/bpf/cgroup.c
> @@ -777,7 +777,7 @@ cgroup_base_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog)
> case BPF_FUNC_get_current_cgroup_id:
> return &bpf_get_current_cgroup_id_proto;
> case BPF_FUNC_trace_printk:
> - if (capable(CAP_SYS_ADMIN))
> + if (bpf_capable(CAP_SYS_ADMIN))
> return bpf_get_trace_printk_proto();
> /* fall through */
> default:
> diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
> index ad3be85f1411..25c1e3c59699 100644
> --- a/kernel/bpf/core.c
> +++ b/kernel/bpf/core.c
> @@ -646,7 +646,7 @@ static bool bpf_prog_kallsyms_verify_off(const struct bpf_prog *fp)
> void bpf_prog_kallsyms_add(struct bpf_prog *fp)
> {
> if (!bpf_prog_kallsyms_candidate(fp) ||
> - !capable(CAP_SYS_ADMIN))
> + !bpf_capable(CAP_SYS_ADMIN))
> return;
>
> spin_lock_bh(&bpf_lock);
> @@ -768,7 +768,7 @@ static int bpf_jit_charge_modmem(u32 pages)
> {
> if (atomic_long_add_return(pages, &bpf_jit_current) >
> (bpf_jit_limit >> PAGE_SHIFT)) {
> - if (!capable(CAP_SYS_ADMIN)) {
> + if (!bpf_capable(CAP_SYS_ADMIN)) {
> atomic_long_sub(pages, &bpf_jit_current);
> return -EPERM;
> }
> diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c
> index 8dff08768087..4c6054626b4f 100644
> --- a/kernel/bpf/cpumap.c
> +++ b/kernel/bpf/cpumap.c
> @@ -83,7 +83,7 @@ static struct bpf_map *cpu_map_alloc(union bpf_attr *attr)
> u64 cost;
> int ret;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return ERR_PTR(-EPERM);
>
> /* check sanity of attributes */
> diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
> index 40e86a7e0ef0..b7c3785be289 100644
> --- a/kernel/bpf/devmap.c
> +++ b/kernel/bpf/devmap.c
> @@ -83,7 +83,7 @@ static struct bpf_map *dev_map_alloc(union bpf_attr *attr)
> u64 cost;
> int err;
>
> - if (!capable(CAP_NET_ADMIN))
> + if (!bpf_capable(CAP_NET_ADMIN))
> return ERR_PTR(-EPERM);
>
> /* check sanity of attributes */
> diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
> index 22066a62c8c9..461a75c311a4 100644
> --- a/kernel/bpf/hashtab.c
> +++ b/kernel/bpf/hashtab.c
> @@ -244,13 +244,13 @@ static int htab_map_alloc_check(union bpf_attr *attr)
> BUILD_BUG_ON(offsetof(struct htab_elem, fnode.next) !=
> offsetof(struct htab_elem, hash_node.pprev));
>
> - if (lru && !capable(CAP_SYS_ADMIN))
> + if (lru && !bpf_capable(CAP_SYS_ADMIN))
> /* LRU implementation is much complicated than other
> * maps. Hence, limit to CAP_SYS_ADMIN for now.
> */
> return -EPERM;
>
> - if (zero_seed && !capable(CAP_SYS_ADMIN))
> + if (zero_seed && !bpf_capable(CAP_SYS_ADMIN))
> /* Guard against local DoS, and discourage production use. */
> return -EPERM;
>
> diff --git a/kernel/bpf/lpm_trie.c b/kernel/bpf/lpm_trie.c
> index 56e6c75d354d..571962022fdf 100644
> --- a/kernel/bpf/lpm_trie.c
> +++ b/kernel/bpf/lpm_trie.c
> @@ -543,7 +543,7 @@ static struct bpf_map *trie_alloc(union bpf_attr *attr)
> u64 cost = sizeof(*trie), cost_per_node;
> int ret;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return ERR_PTR(-EPERM);
>
> /* check sanity of attributes */
> diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c
> index ba635209ae9a..d3e5378c5a15 100644
> --- a/kernel/bpf/offload.c
> +++ b/kernel/bpf/offload.c
> @@ -366,7 +366,7 @@ struct bpf_map *bpf_map_offload_map_alloc(union bpf_attr *attr)
> struct bpf_offloaded_map *offmap;
> int err;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return ERR_PTR(-EPERM);
> if (attr->map_type != BPF_MAP_TYPE_ARRAY &&
> attr->map_type != BPF_MAP_TYPE_HASH)
> diff --git a/kernel/bpf/queue_stack_maps.c b/kernel/bpf/queue_stack_maps.c
> index f697647ceb54..01d848f1a783 100644
> --- a/kernel/bpf/queue_stack_maps.c
> +++ b/kernel/bpf/queue_stack_maps.c
> @@ -45,7 +45,7 @@ static bool queue_stack_map_is_full(struct bpf_queue_stack *qs)
> /* Called from syscall */
> static int queue_stack_map_alloc_check(union bpf_attr *attr)
> {
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
>
> /* check sanity of attributes */
> diff --git a/kernel/bpf/reuseport_array.c b/kernel/bpf/reuseport_array.c
> index 50c083ba978c..840f38a58c7d 100644
> --- a/kernel/bpf/reuseport_array.c
> +++ b/kernel/bpf/reuseport_array.c
> @@ -154,7 +154,7 @@ static struct bpf_map *reuseport_array_alloc(union bpf_attr *attr)
> struct bpf_map_memory mem;
> u64 array_size;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return ERR_PTR(-EPERM);
>
> array_size = sizeof(*array);
> diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c
> index 052580c33d26..1eab27b0bc17 100644
> --- a/kernel/bpf/stackmap.c
> +++ b/kernel/bpf/stackmap.c
> @@ -90,7 +90,7 @@ static struct bpf_map *stack_map_alloc(union bpf_attr *attr)
> u64 cost, n_buckets;
> int err;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return ERR_PTR(-EPERM);
>
> if (attr->map_flags & ~STACK_CREATE_FLAG_MASK)
> diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
> index 7713cf39795a..d80b04b6a5fa 100644
> --- a/kernel/bpf/syscall.c
> +++ b/kernel/bpf/syscall.c
> @@ -23,6 +23,8 @@
> #include <linux/timekeeping.h>
> #include <linux/ctype.h>
> #include <linux/nospec.h>
> +#include <linux/miscdevice.h>
> +#include <linux/resource.h>
>
> #define IS_FD_ARRAY(map) ((map)->map_type == BPF_MAP_TYPE_PROG_ARRAY || \
> (map)->map_type == BPF_MAP_TYPE_PERF_EVENT_ARRAY || \
> @@ -1166,7 +1168,7 @@ static int map_freeze(const union bpf_attr *attr)
> err = -EBUSY;
> goto err_put;
> }
> - if (!capable(CAP_SYS_ADMIN)) {
> + if (!bpf_capable(CAP_SYS_ADMIN)) {
> err = -EPERM;
> goto err_put;
> }
> @@ -1616,7 +1618,7 @@ static int bpf_prog_load(union bpf_attr *attr, union bpf_attr __user *uattr)
>
> if (!IS_ENABLED(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) &&
> (attr->prog_flags & BPF_F_ANY_ALIGNMENT) &&
> - !capable(CAP_SYS_ADMIN))
> + !bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
>
> /* copy eBPF program license from user space */
> @@ -1629,11 +1631,12 @@ static int bpf_prog_load(union bpf_attr *attr, union bpf_attr __user *uattr)
> is_gpl = license_is_gpl_compatible(license);
>
> if (attr->insn_cnt == 0 ||
> - attr->insn_cnt > (capable(CAP_SYS_ADMIN) ? BPF_COMPLEXITY_LIMIT_INSNS : BPF_MAXINSNS))
> + attr->insn_cnt > (bpf_capable(CAP_SYS_ADMIN) ?
> + BPF_COMPLEXITY_LIMIT_INSNS : BPF_MAXINSNS))
> return -E2BIG;
> if (type != BPF_PROG_TYPE_SOCKET_FILTER &&
> type != BPF_PROG_TYPE_CGROUP_SKB &&
> - !capable(CAP_SYS_ADMIN))
> + !bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
>
> bpf_prog_load_fixup_attach_type(attr);
> @@ -1861,7 +1864,7 @@ static int bpf_prog_attach(const union bpf_attr *attr)
> struct bpf_prog *prog;
> int ret;
>
> - if (!capable(CAP_NET_ADMIN))
> + if (!bpf_capable(CAP_NET_ADMIN))
> return -EPERM;
>
> if (CHECK_ATTR(BPF_PROG_ATTACH))
> @@ -1951,7 +1954,7 @@ static int bpf_prog_detach(const union bpf_attr *attr)
> {
> enum bpf_prog_type ptype;
>
> - if (!capable(CAP_NET_ADMIN))
> + if (!bpf_capable(CAP_NET_ADMIN))
> return -EPERM;
>
> if (CHECK_ATTR(BPF_PROG_DETACH))
> @@ -2007,7 +2010,7 @@ static int bpf_prog_detach(const union bpf_attr *attr)
> static int bpf_prog_query(const union bpf_attr *attr,
> union bpf_attr __user *uattr)
> {
> - if (!capable(CAP_NET_ADMIN))
> + if (!bpf_capable(CAP_NET_ADMIN))
> return -EPERM;
> if (CHECK_ATTR(BPF_PROG_QUERY))
> return -EINVAL;
> @@ -2051,7 +2054,7 @@ static int bpf_prog_test_run(const union bpf_attr *attr,
> struct bpf_prog *prog;
> int ret = -ENOTSUPP;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
> if (CHECK_ATTR(BPF_PROG_TEST_RUN))
> return -EINVAL;
> @@ -2088,7 +2091,7 @@ static int bpf_obj_get_next_id(const union bpf_attr *attr,
> if (CHECK_ATTR(BPF_OBJ_GET_NEXT_ID) || next_id >= INT_MAX)
> return -EINVAL;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
>
> next_id++;
> @@ -2114,7 +2117,7 @@ static int bpf_prog_get_fd_by_id(const union bpf_attr *attr)
> if (CHECK_ATTR(BPF_PROG_GET_FD_BY_ID))
> return -EINVAL;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
>
> spin_lock_bh(&prog_idr_lock);
> @@ -2148,7 +2151,7 @@ static int bpf_map_get_fd_by_id(const union bpf_attr *attr)
> attr->open_flags & ~BPF_OBJ_FLAG_MASK)
> return -EINVAL;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
>
> f_flags = bpf_get_file_flag(attr->open_flags);
> @@ -2323,7 +2326,7 @@ static int bpf_prog_get_info_by_fd(struct bpf_prog *prog,
> info.run_time_ns = stats.nsecs;
> info.run_cnt = stats.cnt;
>
> - if (!capable(CAP_SYS_ADMIN)) {
> + if (!bpf_capable(CAP_SYS_ADMIN)) {
> info.jited_prog_len = 0;
> info.xlated_prog_len = 0;
> info.nr_jited_ksyms = 0;
> @@ -2641,7 +2644,7 @@ static int bpf_btf_load(const union bpf_attr *attr)
> if (CHECK_ATTR(BPF_BTF_LOAD))
> return -EINVAL;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
>
> return btf_new_fd(attr);
> @@ -2654,7 +2657,7 @@ static int bpf_btf_get_fd_by_id(const union bpf_attr *attr)
> if (CHECK_ATTR(BPF_BTF_GET_FD_BY_ID))
> return -EINVAL;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
>
> return btf_get_fd_by_id(attr->btf_id);
> @@ -2723,7 +2726,7 @@ static int bpf_task_fd_query(const union bpf_attr *attr,
> if (CHECK_ATTR(BPF_TASK_FD_QUERY))
> return -EINVAL;
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
>
> if (attr->task_fd_query.flags != 0)
> @@ -2791,7 +2794,7 @@ SYSCALL_DEFINE3(bpf, int, cmd, union bpf_attr __user *, uattr, unsigned int, siz
> union bpf_attr attr = {};
> int err;
>
> - if (sysctl_unprivileged_bpf_disabled && !capable(CAP_SYS_ADMIN))
> + if (sysctl_unprivileged_bpf_disabled && !bpf_capable(CAP_SYS_ADMIN))
> return -EPERM;
>
> err = bpf_check_uarg_tail_zero(uattr, sizeof(attr), size);
> @@ -2886,3 +2889,40 @@ SYSCALL_DEFINE3(bpf, int, cmd, union bpf_attr __user *, uattr, unsigned int, siz
>
> return err;
> }
> +
> +static long bpf_dev_ioctl(struct file *filp,
> + unsigned int ioctl, unsigned long arg)
> +{
> + switch (ioctl) {
> + case BPF_DEV_IOCTL_GET_PERM:
> + set_bit(TASK_BPF_FLAG_PERMITTED, ¤t->bpf_flags);
> + break;
> + case BPF_DEV_IOCTL_PUT_PERM:
> + clear_bit(TASK_BPF_FLAG_PERMITTED, ¤t->bpf_flags);
I think the get/put for uapi is a bit misleading, first thought at least for
me is on get/put_user() when I read the name.
> + break;
> + default:
> + return -EINVAL;
> + }
> + return 0;
> +}
> +
> +static const struct file_operations bpf_chardev_ops = {
> + .unlocked_ioctl = bpf_dev_ioctl,
> +};
> +
> +static struct miscdevice bpf_dev = {
> + .minor = MISC_DYNAMIC_MINOR,
> + .name = "bpf",
> + .fops = &bpf_chardev_ops,
> + .mode = 0440,
> + .nodename = "bpf",
Here's what kvm does:
static struct miscdevice kvm_dev = {
KVM_MINOR,
"kvm",
&kvm_chardev_ops,
};
Is there an actual reason that mode is not 0 by default in bpf case? Why
we need to define nodename?
> +};
> +
> +static int __init bpf_dev_init(void)
> +{
> + if (misc_register(&bpf_dev))
> + pr_warn("BPF: Failed to create /dev/bpf. Continue without it...\n");
> +
> + return 0;
> +}
> +device_initcall(bpf_dev_init);
> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> index 0e079b2298f8..79dc4d641cf3 100644
> --- a/kernel/bpf/verifier.c
> +++ b/kernel/bpf/verifier.c
> @@ -9134,7 +9134,7 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr,
> env->insn_aux_data[i].orig_idx = i;
> env->prog = *prog;
> env->ops = bpf_verifier_ops[env->prog->type];
> - is_priv = capable(CAP_SYS_ADMIN);
> + is_priv = bpf_capable(CAP_SYS_ADMIN);
>
> /* grab the mutex to protect few globals used by verifier */
> if (!is_priv)
> diff --git a/kernel/bpf/xskmap.c b/kernel/bpf/xskmap.c
> index ef7338cebd18..06063679c27a 100644
> --- a/kernel/bpf/xskmap.c
> +++ b/kernel/bpf/xskmap.c
> @@ -21,7 +21,7 @@ static struct bpf_map *xsk_map_alloc(union bpf_attr *attr)
> int cpu, err;
> u64 cost;
>
> - if (!capable(CAP_NET_ADMIN))
> + if (!bpf_capable(CAP_NET_ADMIN))
> return ERR_PTR(-EPERM);
>
> if (attr->max_entries == 0 || attr->key_size != 4 ||
> diff --git a/kernel/fork.c b/kernel/fork.c
> index 75675b9bf6df..18f914d54d92 100644
> --- a/kernel/fork.c
> +++ b/kernel/fork.c
> @@ -923,6 +923,10 @@ static struct task_struct *dup_task_struct(struct task_struct *orig, int node)
> #ifdef CONFIG_MEMCG
> tsk->active_memcg = NULL;
> #endif
> +
> +#ifdef CONFIG_BPF_SYSCALL
> + tsk->bpf_flags = 0;
> +#endif
> return tsk;
>
> free_stack:
> diff --git a/net/core/filter.c b/net/core/filter.c
> index 2014d76e0d2a..01ccf031849c 100644
> --- a/net/core/filter.c
> +++ b/net/core/filter.c
> @@ -5875,7 +5875,7 @@ bpf_base_func_proto(enum bpf_func_id func_id)
> break;
> }
>
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return NULL;
>
> switch (func_id) {
> @@ -6438,7 +6438,7 @@ static bool cg_skb_is_valid_access(int off, int size,
> return false;
> case bpf_ctx_range(struct __sk_buff, data):
> case bpf_ctx_range(struct __sk_buff, data_end):
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return false;
> break;
> }
> @@ -6450,7 +6450,7 @@ static bool cg_skb_is_valid_access(int off, int size,
> case bpf_ctx_range_till(struct __sk_buff, cb[0], cb[4]):
> break;
> case bpf_ctx_range(struct __sk_buff, tstamp):
> - if (!capable(CAP_SYS_ADMIN))
> + if (!bpf_capable(CAP_SYS_ADMIN))
> return false;
> break;
> default:
>
^ permalink raw reply
* Re: [PATCH resend] can: rcar_canfd: fix possible IRQ storm on high load
From: Wolfram Sang @ 2019-06-26 13:32 UTC (permalink / raw)
To: Nikita Yushchenko
Cc: Wolfgang Grandegger, Marc Kleine-Budde, David S. Miller,
Simon Horman, Wolfram Sang, linux-can, netdev, linux-kernel,
Artemi Ivanov, Sergei Shtylyov
In-Reply-To: <20190626131251.GB801@ninjato>
[-- Attachment #1: Type: text/plain, Size: 1854 bytes --]
On Wed, Jun 26, 2019 at 03:12:51PM +0200, Wolfram Sang wrote:
> On Wed, Jun 26, 2019 at 04:08:48PM +0300, Nikita Yushchenko wrote:
> > We have observed rcar_canfd driver entering IRQ storm under high load,
> > with following scenario:
> > - rcar_canfd_global_interrupt() in entered due to Rx available,
> > - napi_schedule_prep() is called, and sets NAPIF_STATE_SCHED in state
> > - Rx fifo interrupts are masked,
> > - rcar_canfd_global_interrupt() is entered again, this time due to
> > error interrupt (e.g. due to overflow),
> > - since scheduled napi poller has not yet executed, condition for calling
> > napi_schedule_prep() from rcar_canfd_global_interrupt() remains true,
> > thus napi_schedule_prep() gets called and sets NAPIF_STATE_MISSED flag
> > in state,
> > - later, napi poller function rcar_canfd_rx_poll() gets executed, and
> > calls napi_complete_done(),
> > - due to NAPIF_STATE_MISSED flag in state, this call does not clear
> > NAPIF_STATE_SCHED flag from state,
> > - on return from napi_complete_done(), rcar_canfd_rx_poll() unmasks Rx
> > interrutps,
> > - Rx interrupt happens, rcar_canfd_global_interrupt() gets called
> > and calls napi_schedule_prep(),
> > - since NAPIF_STATE_SCHED is set in state at this time, this call
> > returns false,
> > - due to that false return, rcar_canfd_global_interrupt() returns
> > without masking Rx interrupt
> > - and this results into IRQ storm: unmasked Rx interrupt happens again
> > and again is misprocessed in the same way.
> >
> > This patch fixes that scenario by unmasking Rx interrupts only when
> > napi_complete_done() returns true, which means it has cleared
> > NAPIF_STATE_SCHED in state.
> >
> > Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
>
> CCing the driver author...
Bounced :(
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
^ permalink raw reply
* Re: [PATCH net-next 2/5] net: sched: em_ipt: set the family based on the protocol when matching
From: Eyal Birger @ 2019-06-26 13:33 UTC (permalink / raw)
To: Nikolay Aleksandrov
Cc: netdev, roopa, pablo, xiyou.wangcong, davem, jiri, jhs
In-Reply-To: <20190626115855.13241-3-nikolay@cumulusnetworks.com>
Hi Nikolay,
On Wed, 26 Jun 2019 14:58:52 +0300
Nikolay Aleksandrov <nikolay@cumulusnetworks.com> wrote:
> Set the family based on the protocol otherwise protocol-neutral
> matches will have wrong information (e.g. NFPROTO_UNSPEC). In
> preparation for using NFPROTO_UNSPEC xt matches.
>
> Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
> ---
> net/sched/em_ipt.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/net/sched/em_ipt.c b/net/sched/em_ipt.c
> index 64dbafe4e94c..23965a071177 100644
> --- a/net/sched/em_ipt.c
> +++ b/net/sched/em_ipt.c
> @@ -189,10 +189,12 @@ static int em_ipt_match(struct sk_buff *skb,
> struct tcf_ematch *em, case htons(ETH_P_IP):
> if (!pskb_network_may_pull(skb, sizeof(struct
> iphdr))) return 0;
> + state.pf = NFPROTO_IPV4;
> break;
> case htons(ETH_P_IPV6):
> if (!pskb_network_may_pull(skb, sizeof(struct
> ipv6hdr))) return 0;
> + state.pf = NFPROTO_IPV6;
> break;
> default:
> return 0;
> @@ -203,7 +205,7 @@ static int em_ipt_match(struct sk_buff *skb,
> struct tcf_ematch *em, if (skb->skb_iif)
> indev = dev_get_by_index_rcu(em->net, skb->skb_iif);
>
> - nf_hook_state_init(&state, im->hook, im->match->family,
> + nf_hook_state_init(&state, im->hook, state.pf,
> indev ?: skb->dev, skb->dev, NULL,
> em->net, NULL);
> acpar.match = im->match;
I think this change is incompatible with current behavior.
Consider the 'policy' match which matches the packet's xfrm state (sec_path)
with the provided user space parameters. The sec_path includes information
about the encapsulating packet's parameters whereas the current skb points to
the encapsulated packet, and the match is done on the encapsulating
packet's info.
So if you have an IPv6 packet encapsulated within an IPv4 packet, the match
parameters should be done using IPv4 parameters, not IPv6.
Maybe use the packet's family only if the match family is UNSPEC?
Eyal.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox