| Age | Commit message (Collapse) | Author |
|---|
|
Function QoS related fields are already defined in qos related struct.
min and max rate are left out to mlx5_vport_info struct.
Move them to existing qos struct.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Structure mlx5_vport_info consumes 40 bytes of space due to a hole
in it. After packing it reduces to 32 bytes.
Currently:
pahole -C mlx5_vport_info drivers/net/ethernet/mellanox/mlx5/core/eswitch.o
struct mlx5_vport_info {
u8 mac[6]; /* 0 6 */
u16 vlan; /* 6 2 */
u8 qos; /* 8 1 */
/* XXX 7 bytes hole, try to pack */
u64 node_guid; /* 16 8 */
int link_state; /* 24 4 */
u32 min_rate; /* 28 4 */
u32 max_rate; /* 32 4 */
bool spoofchk; /* 36 1 */
bool trusted; /* 37 1 */
/* size: 40, cachelines: 1, members: 9 */
/* sum members: 31, holes: 1, sum holes: 7 */
/* padding: 2 */
/* last cacheline: 40 bytes */
};
After packing:
$ pahole -C mlx5_vport_info drivers/net/ethernet/mellanox/mlx5/core/eswitch.o
struct mlx5_vport_info {
u8 mac[6]; /* 0 6 */
u16 vlan; /* 6 2 */
u64 node_guid; /* 8 8 */
int link_state; /* 16 4 */
u32 min_rate; /* 20 4 */
u32 max_rate; /* 24 4 */
u8 qos; /* 28 1 */
u8 spoofchk:1; /* 29: 0 1 */
u8 trusted:1; /* 29: 1 1 */
/* size: 32, cachelines: 1, members: 9 */
/* padding: 2 */
/* bit_padding: 6 bits */
/* last cacheline: 32 bytes */
};
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Add missing mutex_destroy() to pair with mutex_init().
This should be done only when table is initialized, hence perform
mutex_init() only when table is initialized.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
A device supports 128 rate limiters. A static table allocation consumes
8KB of memory even when rate is not configured.
Instead, allocate the table when at least one rate is configured.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Rate limit entry refcount can be incremented uniformly when it is newly
allocated or reused.
So simplify the code to increment refcount at one place.
Use decrement refcount helper in two routines.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
User helper routines to allocate and free rate limit table entries.
Subsequent patch extends use of these helpers to do allocation
during rate entry allocation callback.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Table max_size, min and max rate are constants initialized while table
is created. Reading it doesn't need to hold a table mutex. Hence, read
them without holding table mutex.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Fix the warning due to missing int.
WARNING: Prefer 'unsigned int' to bare use of 'unsigned'
+ unsigned free_count;
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Function QoS related fields are already defined in qos related struct.
min and max rate are left out to mlx5_vport_info struct.
Move them to existing qos struct.
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Structure mlx5_vport_info consumes 40 bytes of space due to a hole
in it. After packing it reduces to 32 bytes.
Currently:
pahole -C mlx5_vport_info drivers/net/ethernet/mellanox/mlx5/core/eswitch.o
struct mlx5_vport_info {
u8 mac[6]; /* 0 6 */
u16 vlan; /* 6 2 */
u8 qos; /* 8 1 */
/* XXX 7 bytes hole, try to pack */
u64 node_guid; /* 16 8 */
int link_state; /* 24 4 */
u32 min_rate; /* 28 4 */
u32 max_rate; /* 32 4 */
bool spoofchk; /* 36 1 */
bool trusted; /* 37 1 */
/* size: 40, cachelines: 1, members: 9 */
/* sum members: 31, holes: 1, sum holes: 7 */
/* padding: 2 */
/* last cacheline: 40 bytes */
};
After packing:
$ pahole -C mlx5_vport_info drivers/net/ethernet/mellanox/mlx5/core/eswitch.o
struct mlx5_vport_info {
u8 mac[6]; /* 0 6 */
u16 vlan; /* 6 2 */
u64 node_guid; /* 8 8 */
int link_state; /* 16 4 */
u32 min_rate; /* 20 4 */
u32 max_rate; /* 24 4 */
u8 qos; /* 28 1 */
u8 spoofchk:1; /* 29: 0 1 */
u8 trusted:1; /* 29: 1 1 */
/* size: 32, cachelines: 1, members: 9 */
/* padding: 2 */
/* bit_padding: 6 bits */
/* last cacheline: 32 bytes */
};
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Add support for matching on ct_state inv and rel flags.
Currently the support is only for match on -inv and -rel.
Matching on +inv and +rel will be rejected.
Example:
$ tc filter add dev ens1f0_0 ingress prio 1 chain 1 proto ip flower \
ct_state -est-rel+trk \
action mirred egress redirect dev ens1f0_1
$ tc filter add dev ens1f0_1 ingress prio 1 chain 1 proto ip flower \
ct_state +trk+est-inv \
action mirred egress redirect dev ens1f0_0
Signed-off-by: Ariel Levkovich <lariel@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Restore CMP screener registers on resume path.
Fixes: c1e85c6ce57ef ("net: macb: save/restore the remaining registers and features")
Signed-off-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Acked-by: Nicolas Ferre <nicolas.ferre@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
It's useful, especially for debugging purposes, to have the Rx copybreak
value changeable at runtime. Export it as an ethtool tunable.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
DMA unmapping, allocating a new buffer and DMA mapping it back on the
refill path is really not that efficient. Proper buffer recycling (page
pool, flipping the page and using the other half) cannot be done for
DPAA2 since it's not a ring based controller but it rather deals with
multiple queues which all get their buffers from the same buffer pool on
Rx.
To circumvent these limitations, add support for Rx copybreak. For small
sized packets instead of creating a skb around the buffer in which the
frame was received, allocate a new sk buffer altogether, copy the
contents of the frame and release the initial page back into the buffer
pool.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Rename the dpaa2_eth_xdp_release_buf function into dpaa2_eth_recycle_buf
since in the next patches we'll be using the same recycle mechanism for
the normal stack path beside for XDP_DROP.
Also, rename the array which holds the buffers to be recycled so that it
does not have any reference to XDP.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Let the network stack know we've got support for timestamping
the packets.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add the new hwstamp stats to our ethtool stats output.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add the get_ts_info() callback for ethtool support of
timestamping information.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The Tx and Rx timestamped packets are handled through separate
queues. Here we set them up, service them, and tear them down
along with the normal Tx and Rx queues.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We do hardware timestamping through a separate Tx queue,
and optionally through a separate Rx queue. These queues
are allocated, freed, and tracked separately from the basic
queue arrays.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add handling of the new Rx packet classification filter type.
This simple bit of classification allows for steering packets
to a separate Rx queue for processing.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
These are changes to compile and link the new code, but no
new feature support is available or advertised yet.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This adds the file of code for supporting Tx and Rx hardware
timestamps and the raw clock interface, but does not yet link
it in for compiling or use.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Split the wait part out of adminq_post_wait() into a separate
function so that a caller can have finer grain control over
the sequencing of operations and locking.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The interface for hardware timestamping includes a new FW
request, device identity fields, Tx and Rx queue feature bits, a
new Rx filter type, the beginnings of Rx packet classifications,
and hardware timestamp registers.
If the IONIC_ETH_HW_TIMESTAMP bit is shown in the
ionic_lif_config features bit string, then we have support
for the hw clock registers. If the IONIC_RXQ_F_HWSTAMP and
IONIC_TXQ_F_HWSTAMP features are shown in the ionic_q_identity
features, then the queues can support HW timestamps on packets.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
In preparating for hardware timestamping, we need to support
large Tx and Rx completion descriptors. Here we add the new
queue feature ids and handling for the completion descriptor
sizes.
We only are adding support for the Rx 2x sized completion
descriptors in the general Rx queues for now as we will be
using it for PTP Rx support, and we don't have an immediate
use for the large descriptors in the general Tx queues yet;
it will be used in a special Tx queues added in one of the
next few patches.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add queue feature extensions to prepare for features that
can be queue specific, in addition to the general queue
features already defined. While we're here, change the
existing feature ids from #defines to enum.
Signed-off-by: Allen Hubbe <allenbh@pensando.io>
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Alexei Starovoitov says:
====================
pull-request: bpf-next 2021-04-01
The following pull-request contains BPF updates for your *net-next* tree.
We've added 68 non-merge commits during the last 7 day(s) which contain
a total of 70 files changed, 2944 insertions(+), 1139 deletions(-).
The main changes are:
1) UDP support for sockmap, from Cong.
2) Verifier merge conflict resolution fix, from Daniel.
3) xsk selftests enhancements, from Maciej.
4) Unstable helpers aka kernel func calling, from Martin.
5) Batches ops for LPM map, from Pedro.
6) Fix race in bpf_get_local_storage, from Yonghong.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use memset to initialize local array in drivers/net/usb/ax88179_178a.c, and
also set a local u16 and u32 variable to 0. Fixes a KMSAN found uninit-value bug
reported by syzbot at:
https://syzkaller.appspot.com/bug?id=00371c73c72f72487c1d0bfe0cc9d00de339d5aa
Reported-by: syzbot+4993e4a0e237f1b53747@syzkaller.appspotmail.com
Signed-off-by: Phillip Potter <phil@philpotter.co.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2021-04-01
This series contains updates to i40e driver only.
Arkadiusz fixes warnings for inconsistent indentation.
Magnus fixes an issue on xsk receive where single packets over time
are batched rather than received immediately.
Eryk corrects warnings and reporting of veb-stats.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
All Gigabit PHYs use the same register layout as far as fetching
statistics goes. Fast Ethernet PHYs do not all support statistics, and
the BCM54616S would require some switching between the coper and fiber
modes to fetch the appropriate statistics which is not supported yet.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
syzbot reported memory leak in peak_usb.
The problem was in case of failure after calling
->dev_init()[2] in peak_usb_create_dev()[1]. The data
allocated int dev_init() wasn't freed, so simple
->dev_free() call fix this problem.
backtrace:
[<0000000079d6542a>] kmalloc include/linux/slab.h:552 [inline]
[<0000000079d6542a>] kzalloc include/linux/slab.h:682 [inline]
[<0000000079d6542a>] pcan_usb_fd_init+0x156/0x210 drivers/net/can/usb/peak_usb/pcan_usb_fd.c:868 [2]
[<00000000c09f9057>] peak_usb_create_dev drivers/net/can/usb/peak_usb/pcan_usb_core.c:851 [inline] [1]
[<00000000c09f9057>] peak_usb_probe+0x389/0x490 drivers/net/can/usb/peak_usb/pcan_usb_core.c:949
Reported-by: syzbot+91adee8d9ebb9193d22d@syzkaller.appspotmail.com
Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Fix some typos.
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Lu Wei <luwei32@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The commit d2a029bde37b ("stmmac: pci: add MSI support for Intel Quark
X1000") introduced a pci_enable_msi() call in stmmac_pci.c.
With the commit 58da0cfa6cf1 ("net: stmmac: create dwmac-intel.c to
contain all Intel platform"), Intel Quark platform related codes
have been moved to the newly created driver.
Removing this unnecessary pci_enable_msi() call as there are no other
devices that uses stmmac-pci and need MSI to be enabled.
Signed-off-by: Wong Vee Khee <vee.khee.wong@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Update dwmac-intel to use managed function, i.e. pcim_enable_device().
This will allow devres framework to call resource free function for us.
Signed-off-by: Wong Vee Khee <vee.khee.wong@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
syzbot reported memory leak in atusb_probe()[1].
The problem was in atusb_alloc_urbs().
Since urb is anchored, we need to release the reference
to correctly free the urb
backtrace:
[<ffffffff82ba0466>] kmalloc include/linux/slab.h:559 [inline]
[<ffffffff82ba0466>] usb_alloc_urb+0x66/0xe0 drivers/usb/core/urb.c:74
[<ffffffff82ad3888>] atusb_alloc_urbs drivers/net/ieee802154/atusb.c:362 [inline][2]
[<ffffffff82ad3888>] atusb_probe+0x158/0x820 drivers/net/ieee802154/atusb.c:1038 [1]
Reported-by: syzbot+28a246747e0a465127f3@syzkaller.appspotmail.com
Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue
Tony Nguyen says:
====================
100GbE Intel Wired LAN Driver Updates 2021-03-31
This series contains updates to ice driver only.
Benita adds support for XPS.
Ani moves netdev registration to the end of probe to prevent use before
the interface is ready and moves up an error check to possibly avoid
an unneeded call. He also consolidates the VSI state and flag fields to
a single field.
Dan changes the segment where package information is pulled.
Paul S ensures correct ITR values are set when increasing ring size.
Paul G rewords a link misconfiguration message as this could be
expected.
Bruce removes setting an unnecessary AQ flag and corrects a memory
allocation call. Also fixes checkpatch issues for 'COMPLEX_MACRO'.
Qi aligns PTYPE bitmap naming by adding 'ptype' prefix to the bitmaps
missing it.
Brett removes limiting Rx queue mapping to RSS size as there is not a
dependency on this. He also refactors RSS configuration by introducing
individual functions for LUT and key configuration and by passing a
structure containing pertinent information instead of individual
arguments.
Tony corrects a comment block to follow netdev style.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux into arm/drivers
SoCFPGA updates for v5.13
- Patches from Krzysztof Kozlowski the cleans up and consolidate support for
SoCFPGA platforms
- Rename ARCH_SOCFPGA into ARCH_INTEL_SOCFPGA
- Consolidate ARCH_STRATIX10 into ARCH_INTEL_SOCFPGA
- Consolidate ARCH_AGILEX into ARCH_INTEL_SOCFPGA
- Consolidate ARCH_N5X into ARCH_INTEL_SOCFPGA
* tag 'socfpga_update_for_v5.13' of git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux:
ARM: socfpga: drop ARCH_SOCFPGA
reset: socfpga: use ARCH_INTEL_SOCFPGA also for 32-bit ARM SoCs
i2c: altera: use ARCH_INTEL_SOCFPGA also for 32-bit ARM SoCs
fpga: altera: use ARCH_INTEL_SOCFPGA also for 32-bit ARM SoCs
dmaengine: socfpga: use ARCH_INTEL_SOCFPGA also for 32-bit ARM SoCs
clk: socfpga: use ARCH_INTEL_SOCFPGA also for 32-bit ARM SoCs (and compile test)
clk: socfpga: allow compile testing of Stratix 10 / Agilex clocks
arm64: socfpga: merge Agilex and N5X into ARCH_INTEL_SOCFPGA
EDAC: altera: merge ARCH_SOCFPGA and ARCH_STRATIX10
clk: socfpga: merge ARCH_SOCFPGA and ARCH_STRATIX10
clk: socfpga: build together Stratix 10, Agilex and N5X clock drivers
net: stmmac: merge ARCH_SOCFPGA and ARCH_STRATIX10
mfd: altera: merge ARCH_SOCFPGA and ARCH_STRATIX10
ARM: socfpga: introduce common ARCH_INTEL_SOCFPGA
clk: socfpga: allow building N5X clocks with ARCH_N5X
Link: https://lore.kernel.org/r/20210330110430.558182-1-dinguyen@kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
|
|
If veb-stats was enabled, the ethtool stats triggered a warning
due to invalid size: 'unexpected stat size for veb.tc_%u_tx_packets'.
This was due to an incorrect structure definition for the statistics.
Structures and functions have been improved in line with requirements
for the presentation of statistics, in particular for the functions:
'i40e_add_ethtool_stats' and 'i40e_add_stat_strings'.
Fixes: 1510ae0be2a4 ("i40e: convert VEB TC stats to use an i40e_stats array")
Signed-off-by: Eryk Rybak <eryk.roch.rybak@intel.com>
Signed-off-by: Grzegorz Szczurek <grzegorzx.szczurek@intel.com>
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Tested-by: Dave Switzer <david.switzer@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
Fix so that single packets are received immediately instead of in
batches of 8. If you sent 1 pps to a system, you received 8 packets
every 8 seconds instead of 1 packet every second. The problem behind
this was that the work_done reporting from the Tx part of the driver
was broken. The work_done reporting in i40e controls not only the
reporting back to the napi logic but also the setting of the interrupt
throttling logic. When Tx or Rx reports that it has more to do,
interrupts are throttled or coalesced and when they both report that
they are done, interrupts are armed right away. If the wrong work_done
value is returned, the logic will start to throttle interrupts in a
situation where it should have just enabled them. This leads to the
undesired batching behavior seen in user-space.
Fix this by returning the correct boolean value from the Tx xsk
zero-copy path. Return true if there is nothing to do or if we got
fewer packets to process than we asked for. Return false if we got as
many packets as the budget since there might be more packets we can
process.
Fixes: 3106c580fb7c ("i40e: Use batched xsk Tx interfaces to increase performance")
Reported-by: Sreedevi Joshi <sreedevi.joshi@intel.com>
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Kiran Bhandare <kiranx.bhandare@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
Fixed new static analysis findings:
"warn: inconsistent indenting" - introduced lately,
reported with lkp and smatch build.
Fixes: 4b208eaa8078 ("i40e: Add init and default config of software based DCB")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Signed-off-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Tested-by: Dave Switzer <david.switzer@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
|
|
mt7623 uses offload version 2 too
tested on Bananapi-R2
Signed-off-by: Frank Wunderlich <frank-w@public-files.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Turn on the MEEAO field of MTL_ECC_Control_Register by default.
As the MTL ECC Error Address Status Over-ride(MEEAO) is set by default,
the following error address fields will hold the last valid address
where the error is detected.
Signed-off-by: Voon Weifeng <weifeng.voon@intel.com>
Signed-off-by: Tan Tee Min <tee.min.tan@intel.com>
Co-developed-by: Wong Vee Khee <vee.khee.wong@linux.intel.com>
Signed-off-by: Wong Vee Khee <vee.khee.wong@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The driver implementation of the XDP_REDIRECT action reuses parts from
XDP_TX, most notably the enetc_xdp_tx function which transmits an array
of TX software BDs. Only this time, the buffers don't have DMA mappings,
we need to create them.
When a BPF program reaches the XDP_REDIRECT verdict for a frame, we can
employ the same buffer reuse strategy as for the normal processing path
and for XDP_PASS: we can flip to the other page half and seed that to
the RX ring.
Note that scatter/gather support is there, but disabled due to lack of
multi-buffer support in XDP (which is added by this series):
https://patchwork.kernel.org/project/netdevbpf/cover/cover.1616179034.git.lorenzo@kernel.org/
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
As explained in the XDP_TX patch, when receiving a burst of frames with
the XDP_TX verdict, there is a momentary dip in the number of available
RX buffers. The system will eventually recover as TX completions will
start kicking in and refilling our RX BD ring again. But until that
happens, we need to survive with as few out-of-buffer discards as
possible.
This increases the memory footprint of the driver in order to avoid
discards at 2.5Gbps line rate 64B packet sizes, the maximum speed
available for testing on 1 port on NXP LS1028A.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
For reflecting packets back into the interface they came from, we create
an array of TX software BDs derived from the RX software BDs. Therefore,
we need to extend the TX software BD structure to contain most of the
stuff that's already present in the RX software BD structure, for
reasons that will become evident in a moment.
For a frame with the XDP_TX verdict, we don't reuse any buffer right
away as we do for XDP_DROP (the same page half) or XDP_PASS (the other
page half, same as the skb code path).
Because the buffer transfers ownership from the RX ring to the TX ring,
reusing any page half right away is very dangerous. So what we can do is
we can recycle the same page half as soon as TX is complete.
The code path is:
enetc_poll
-> enetc_clean_rx_ring_xdp
-> enetc_xdp_tx
-> enetc_refill_rx_ring
(time passes, another MSI interrupt is raised)
enetc_poll
-> enetc_clean_tx_ring
-> enetc_recycle_xdp_tx_buff
But that creates a problem, because there is a potentially large time
window between enetc_xdp_tx and enetc_recycle_xdp_tx_buff, period in
which we'll have less and less RX buffers.
Basically, when the ship starts sinking, the knee-jerk reaction is to
let enetc_refill_rx_ring do what it does for the standard skb code path
(refill every 16 consumed buffers), but that turns out to be very
inefficient. The problem is that we have no rx_swbd->page at our
disposal from the enetc_reuse_page path, so enetc_refill_rx_ring would
have to call enetc_new_page for every buffer that we refill (if we
choose to refill at this early stage). Very inefficient, it only makes
the problem worse, because page allocation is an expensive process, and
CPU time is exactly what we're lacking.
Additionally, there is an even bigger problem: if we let
enetc_refill_rx_ring top up the ring's buffers again from the RX path,
remember that the buffers sent to transmission haven't disappeared
anywhere. They will be eventually sent, and processed in
enetc_clean_tx_ring, and an attempt will be made to recycle them.
But surprise, the RX ring is already full of new buffers, because we
were premature in deciding that we should refill. So not only we took
the expensive decision of allocating new pages, but now we must throw
away perfectly good and reusable buffers.
So what we do is we implement an elastic refill mechanism, which keeps
track of the number of in-flight XDP_TX buffer descriptors. We top up
the RX ring only up to the total ring capacity minus the number of BDs
that are in flight (because we know that those BDs will return to us
eventually).
The enetc driver manages 1 RX ring per CPU, and the default TX ring
management is the same. So we do XDP_TX towards the TX ring of the same
index, because it is affined to the same CPU. This will probably not
produce great results when we have a tc-taprio/tc-mqprio qdisc on the
interface, because in that case, the number of TX rings might be
greater, but I didn't add any checks for that yet (mostly because I
didn't know what checks to add).
It should also be noted that we need to change the DMA mapping direction
for RX buffers, since they may now be reflected into the TX ring of the
same device. We choose to use DMA_BIDIRECTIONAL instead of unmapping and
remapping as DMA_TO_DEVICE, because performance is better this way.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
For the RX ring, enetc uses an allocation scheme based on pages split
into two buffers, which is already very efficient in terms of preventing
reallocations / maximizing reuse, so I see no reason why I would change
that.
+--------+--------+--------+--------+--------+--------+--------+
| | | | | | | |
| half B | half B | half B | half B | half B | half B | half B |
| | | | | | | |
+--------+--------+--------+--------+--------+--------+--------+
| | | | | | | |
| half A | half A | half A | half A | half A | half A | half A | RX ring
| | | | | | | |
+--------+--------+--------+--------+--------+--------+--------+
^ ^
| |
next_to_clean next_to_alloc
next_to_use
+--------+--------+--------+--------+--------+
| | | | | |
| half B | half B | half B | half B | half B |
| | | | | |
+--------+--------+--------+--------+--------+--------+--------+
| | | | | | | |
| half B | half B | half A | half A | half A | half A | half A | RX ring
| | | | | | | |
+--------+--------+--------+--------+--------+--------+--------+
| | | ^ ^
| half A | half A | | |
| | | next_to_clean next_to_use
+--------+--------+
^
|
next_to_alloc
then when enetc_refill_rx_ring is called, whose purpose is to advance
next_to_use, it sees that it can take buffers up to next_to_alloc, and
it says "oh, hey, rx_swbd->page isn't NULL, I don't need to allocate
one!".
The only problem is that for default PAGE_SIZE values of 4096, buffer
sizes are 2048 bytes. While this is enough for normal skb allocations at
an MTU of 1500 bytes, for XDP it isn't, because the XDP headroom is 256
bytes, and including skb_shared_info and alignment, we end up being able
to make use of only 1472 bytes, which is insufficient for the default
MTU.
To solve that problem, we implement scatter/gather processing in the
driver, because we would really like to keep the existing allocation
scheme. A packet of 1500 bytes is received in a buffer of 1472 bytes and
another one of 28 bytes.
Because the headroom required by XDP is different (and much larger) than
the one required by the network stack, whenever a BPF program is added
or deleted on the port, we drain the existing RX buffers and seed new
ones with the required headroom. We also keep the required headroom in
rx_ring->buffer_offset.
The simplest way to implement XDP_PASS, where an skb must be created, is
to create an xdp_buff based on the next_to_clean RX BDs, but not clear
those BDs from the RX ring yet, just keep the original index at which
the BDs for this frame started. Then, if the verdict is XDP_PASS,
instead of converting the xdb_buff to an skb, we replay a call to
enetc_build_skb (just as in the normal enetc_clean_rx_ring case),
starting from the original BD index.
We would also like to be minimally invasive to the regular RX data path,
and not check whether there is a BPF program attached to the ring on
every packet. So we create a separate RX ring processing function for
XDP.
Because we only install/remove the BPF program while the interface is
down, we forgo the rcu_read_lock() in enetc_clean_rx_ring, since there
shouldn't be any circumstance in which we are processing packets and
there is a potentially freed BPF program attached to the RX ring.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
For XDP_TX, we need to call enetc_reuse_page from enetc_clean_tx_ring,
so we need to avoid a forward declaration.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
With the future introduction of some new fields into enetc_tx_swbd such
as is_xdp_tx, is_xdp_redirect etc, we need not only to set these bits
to true from the XDP_TX/XDP_REDIRECT code path, but also to false from
the old code paths.
This is because TX software buffer descriptors are kept in a ring that
is shadow of the hardware TX ring, so these structures keep getting
reused, and there is always the possibility that when a software BD is
reused (after we ran a full circle through the TX ring), the old user of
the tx_swbd had set is_xdp_tx = true, and now we are sending a regular
skb, which would need to set is_xdp_tx = false.
To be minimally invasive to the old code paths, let's just scrub the
software TX BD in the TX confirmation path (enetc_clean_tx_ring), once
we know that nobody uses this software TX BD (tx_ring->next_to_clean
hasn't yet been updated, and the TX paths check enetc_bd_unused which
tells them if there's any more space in the TX ring for a new enqueue).
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
