From: Jakub Kicinski
Date: Fri, 1 Aug 2025 15:33:43 -0700
> On Wed, 30 Jul 2025 18:07:15 +0200 Alexander Lobakin wrote:
>> Use __LIBETH_WORD_ACCESS to parse descriptors more efficiently when
>> applicable. It really gives some good boosts and code size reduction
>> on x
From: Jakub Kicinski
Date: Fri, 1 Aug 2025 15:30:23 -0700
> On Wed, 30 Jul 2025 18:07:12 +0200 Alexander Lobakin wrote:
>> Note that "other count" in Ethtool will now also include XDP Tx queues.
>
> Erm, what? You mean other channels? Channels are interrupts,
>
From: Joshua Hay
Date: Fri, 25 Jul 2025 11:42:21 -0700
> Replace the TxQ buffer ring with one large pool/array of buffers (only
> for flow scheduling). This eliminates the tag generation and makes it
> impossible for a tag to be associated with more than one packet.
[...]
> -static bool idpf_tx
From: Alexander Lobakin
Date: Fri, 1 Aug 2025 15:12:43 +0200
> From: Kees Cook
> Date: Thu, 31 Jul 2025 10:05:47 -0700
>
>> On Thu, Jul 31, 2025 at 01:37:34PM +0100, Simon Horman wrote:
>>> While I appreciate the desire for improved performance and nicer code
>>
From: Kees Cook
Date: Thu, 31 Jul 2025 10:05:47 -0700
> On Thu, Jul 31, 2025 at 01:37:34PM +0100, Simon Horman wrote:
>> While I appreciate the desire for improved performance and nicer code
>> generation. I think the idea of writing 64 bits of data to the
>> address of a 32 bit member of a struc
From: Simon Horman
Date: Thu, 31 Jul 2025 14:35:57 +0100
> On Wed, Jul 30, 2025 at 06:07:15PM +0200, Alexander Lobakin wrote:
>> Use libeth XDP infra to support running XDP program on Rx polling.
>> This includes all of the possible verdicts/actions.
>> XDP Tx queues are
Add &xdp_metadata_ops with a callback to get RSS hash hint from the
descriptor. Declare the splitq 32-byte descriptor as 4 u64s to parse
them more efficiently when possible.
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/xdp.h | 64 +++
drivers
ki
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/xdp.h | 2 ++
drivers/net/ethernet/intel/idpf/idpf_lib.c | 1 +
drivers/net/ethernet/intel/idpf/xdp.c | 20
3 files changed, 23 insertions(+)
diff --git a/drivers/net/ethernet/intel/idp
tions make sure the compiler could uninline
them when needed.
Use __LIBETH_WORD_ACCESS to parse descriptors more efficiently when
applicable. It really gives some good boosts and code size reduction
on x86_64.
Co-developed-by: Michal Kubiak
Signed-off-by: Michal Kubiak
Signed-off-by: Alexand
d to the frame.
Store &libeth_xdp_stash instead of the skb pointer on the Rx queue.
It's only 8 bytes wider, but contains everything we may need.
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 17 +-
.../ethernet/intel/idpf/idpf_singleq_t
From: Michal Kubiak
Implement loading/removing XDP program using .ndo_bpf callback
in the split queue mode. Reconfigure and restart the queues if needed
(!!old_prog != !!new_prog), otherwise, just update the pointers.
Signed-off-by: Michal Kubiak
Signed-off-by: Alexander Lobakin
---
drivers
h regular Tx traffic.
Also add functions dedicated to support XDP initialization for Rx and
Tx queues and call those functions from the existing algorithms of
queues configuration.
Note that "other count" in Ethtool will now also include XDP Tx queues.
Signed-off-by: Michal Kubiak
Co-dev
creating vport, so that it can be used later by
queues when needed (not only XDP ones).
Co-developed-by: Michal Kubiak
Signed-off-by: Michal Kubiak
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf.h| 8 +++
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 4 ++
dr
t for the replies. If replies for all of the
queues are received, this means the synchronization is done correctly
and we can go forth with stopping the link.
Signed-off-by: Michal Kubiak
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf.h| 7 +-
drivers/net/eth
completions during Tx polling, as
for now, the only user of QB will be XDP, which has its own routines.
Signed-off-by: Michal Kubiak
Signed-off-by: Alexander Lobakin
---
.../net/ethernet/intel/idpf/idpf_lan_txrx.h | 6 +++-
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 11 --
drivers/net
/enable/disable large amount of queues within the minimum
timeout (2 seconds). For now, fall back to the default timeout for
every operation until this is resolved.
Signed-off-by: Alexander Lobakin
---
.../net/ethernet/intel/idpf/idpf_virtchnl.h | 1 -
drivers/net/ethernet/intel/idpf/idp
ken, needs it.
Expand common idpf_vport_{start,stop}() to take an additional bool
telling whether we need to manually take the RTNL lock.
Suggested-by: Maciej Fijalkowski # helper
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf_lib.c | 38 +++--
driv
need to stop polling. Debatable on coherent DMA as the Rx descriptor
size is <= cacheline size, but anyway, the current barrier position
only makes the codegen worse.
Fixes: 3a8845af66ed ("idpf: add RX splitq napi poll support")
Reviewed-by: Maciej Fijalkowski
Signed-off-by: Alex
From: Joshua Hay
With the new Tx buffer management scheme, there is no need for all of
the stashing mechanisms, the hash table, the reserve buffer stack, etc.
Remove all of that.
Signed-off-by: Joshua Hay
Reviewed-by: Madhu Chittim
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet
hua Hay
Reviewed-by: Madhu Chittim
Reviewed-by: Aleksandr Loktionov
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 10 +-
drivers/net/ethernet/intel/idpf/idpf_txrx.c | 203 +---
2 files changed, 102 insertions(+), 111 deletions(-)
diff
resources
to send the packet, and stop the queue if not.
Fixes: 7292af042bcf ("idpf: fix a race in txq wakeup")
Signed-off-by: Joshua Hay
Reviewed-by: Madhu Chittim
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 15 +-
.../ethernet/
r advance. The previous implementation was dependent on the start
of a packet falling on a 64th index in the descriptor ring, which is not
guaranteed with large packets.
Signed-off-by: Luigi Rizzo
Signed-off-by: Brian Vazquez
Signed-off-by: Joshua Hay
Reviewed-by: Madhu Chittim
Signed-off-by: Al
the PTP context descriptor.
Fixes: 1a49cf814fe1 ("idpf: add Tx timestamp flows")
Signed-off-by: Joshua Hay
Reviewed-by: Madhu Chittim
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 5 +-
.../ethernet/intel/idpf/idpf_singleq_txr
arios.
idpf doesn't support VLAN Rx offload, so only the hash hint is
available for now.
Patches 1-6 are prereqs, without which XDP would either not work at all or
work slower/worse/...
Alexander Lobakin (8):
idpf: fix Rx descriptor ready check barrier in splitq
idpf: use a saner limit f
: Madhu Chittim
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 8 +-
drivers/net/ethernet/intel/idpf/idpf_txrx.c | 93 +++--
2 files changed, 91 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.h
b/drivers/net
From: Tony Nguyen
Date: Thu, 17 Jul 2025 13:25:05 -0700
>
>
> On 7/16/2025 2:12 PM, Samiullah Khawaja wrote:
>> Use netif_queue_set_napi to associate TX/RX queues to the relevant napi.
>> This allows fetching napi for a TX or RX queue using netlink queue-get
>> op.
>>
>> Tested:
>> python3 tool
From: Przemek Kitszel
Date: Fri, 11 Jul 2025 12:25:09 +0200
> On 7/11/25 08:47, Malin Jonsson wrote:
>> The idpf driver depends on both AUXILIARY_BUS and NET_DEVLINK, but
>> they're not selected by the Kconfig. Let's fix this so idpf builds even
>> with a stripped config.
>>
>> Reported-by: Yong
From: Jacob Keller
Date: Thu, 10 Jul 2025 15:43:20 -0700
>
>
> On 7/7/2025 4:36 PM, Jacob Keller wrote:
[...]
> I got this to work with the following diff:
>
> diff --git i/drivers/net/ethernet/intel/ice/ice_txrx.h
> w/drivers/net/ethernet/intel/ice/ice_txrx.h
> index 42e74925b9df..6b72608a2
all(), where rxq->q_vector is always initialized
already.
Signed-off-by: Alexander Lobakin
---
To Tony: this is a hotfix for
"idpf: implement XDP_SETUP_PROG in ndo_bpf for splitq"
where this call was introduced, please squash.
---
drivers/net/ethernet/intel/idpf/idpf_txrx.c | 4 ++--
1
From: Joshua Hay
Date: Wed, 25 Jun 2025 09:11:55 -0700
> The Tx refillq logic will cause packets to be silently dropped if there
> are not enough buffer resources available to send a packet in flow
> scheduling mode. Instead, determine how many buffers are needed along
> with number of descriptor
From: Joshua Hay
Date: Wed, 25 Jun 2025 09:11:52 -0700
> This is the start of a 5 patch series intended to fix a stability issue
> in the flow scheduling Tx send/clean path that results in a Tx timeout.
No need to mention "series", "start", "patch" in commit messages.
>
> In certain production
From: Joshua Hay
Date: Wed, 25 Jun 2025 09:11:54 -0700
> Replace the TxQ buffer ring with one large pool/array of buffers (only
> for flow scheduling). The completion tag passed to HW through the
[...]
> diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.c
> b/drivers/net/ethernet/intel/id
from the start
> of the structure. Move this member before the CPU mask instead, in all three
> drivers.
>
> Signed-off-by: Arnd Bergmann
Reviewed-by: Alexander Lobakin
> ---
> v2: move rcu to just after the napi_struct [Alexander Lobakin]
Thanks!
Olek
arios.
idpf doesn't support VLAN Rx offload, so only the hash hint is
available for now.
Patches 1-6 are prereqs, without which XDP would either not work at all or
work slower/worse/...
Alexander Lobakin (8):
idpf: fix Rx descriptor ready check barrier in splitq
idpf: use a saner limit f
ki
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/xdp.h | 2 ++
drivers/net/ethernet/intel/idpf/idpf_lib.c | 1 +
drivers/net/ethernet/intel/idpf/xdp.c | 20
3 files changed, 23 insertions(+)
diff --git a/drivers/net/ethernet/intel/idp
mod, which happens outside of RTNL context,
but calls idpf_vport_stop() which assumes RTNL protection.
Suggested-by: Maciej Fijalkowski # helper
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf_lib.c | 2 ++
drivers/net/ethernet/intel/idpf/idpf_txrx.c |
Add &xdp_metadata_ops with a callback to get RSS hash hint from the
descriptor. Declare the splitq 32-byte descriptor as 4 u64s to parse
them more efficiently when possible.
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/xdp.h | 64 +++
drivers
d to the frame.
Store &libeth_xdp_stash instead of the skb pointer on the Rx queue.
It's only 8 bytes wider, but contains everything we may need.
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 17 +-
.../ethernet/intel/idpf/idpf_singleq_t
need to stop polling. Debatable on coherent DMA as the Rx descriptor
size is <= cacheline size, but anyway, the current barrier position
only makes the codegen worse.
Fixes: 3a8845af66ed ("idpf: add RX splitq napi poll support")
Reviewed-by: Maciej Fijalkowski
Signed-off-by: Alex
tions make sure the compiler could uninline
them when needed.
Use __LIBETH_WORD_ACCESS to parse descriptors more efficiently when
applicable. It really gives some good boosts and code size reduction
on x86_64.
Co-developed-by: Michal Kubiak
Signed-off-by: Michal Kubiak
Signed-off-by: Alexand
From: Michal Kubiak
Implement loading/removing XDP program using .ndo_bpf callback
in the split queue mode. Reconfigure and restart the queues if needed
(!!old_prog != !!new_prog), otherwise, just update the pointers.
Signed-off-by: Michal Kubiak
Signed-off-by: Alexander Lobakin
---
drivers
h regular Tx traffic.
Also add functions dedicated to support XDP initialization for Rx and
Tx queues and call those functions from the existing algorithms of
queues configuration.
Note that "other count" in Ethtool will now also include XDP Tx queues.
Signed-off-by: Michal Kubiak
Co-dev
creating vport, so that it can be used later by
queues when needed (not only XDP ones).
Co-developed-by: Michal Kubiak
Signed-off-by: Michal Kubiak
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf.h| 8 +++
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 4 ++
dr
t for the replies. If replies for all of the
queues are received, this means the synchronization is done correctly
and we can go forth with stopping the link.
Signed-off-by: Michal Kubiak
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf.h| 7 +-
drivers/net/eth
that anyhow.
On a 128-thread Xeon, this now gives me 32 regular Tx queues and leaves
224 free for XDP (128 of which will handle XDP_TX, .ndo_xdp_xmit(), and
XSk xmit when enabled).
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/idpf/idpf_txrx.c | 8 +---
drivers/net/eth
completions during Tx polling, as
for now, the only user of QB will be XDP, which has its own routines.
Signed-off-by: Michal Kubiak
Signed-off-by: Alexander Lobakin
---
.../net/ethernet/intel/idpf/idpf_lan_txrx.h | 6 +++-
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 11 --
drivers/net
From: Arnd Bergmann
Date: Fri, 20 Jun 2025 13:11:17 +0200
> From: Arnd Bergmann
>
> With large values of CONFIG_NR_CPUS, three Intel ethernet drivers fail to
> compile like:
>
> In function ‘i40e_free_q_vector’,
> inlined from ‘i40e_vsi_alloc_q_vectors’ at
> drivers/net/ethernet/intel/i40
From: Colin Ian King
Date: Wed, 18 Jun 2025 14:54:08 +0100
> Don't populate the const read-only array supported_sizes on the
> stack at run time, instead make it static.
>
> Signed-off-by: Colin Ian King
Reviewed-by: Alexander Lobakin
> ---
> drivers/net/ethernet/
s on.
There's a few places in libeth_xdp winning up to several percent from
combined access (both performance and object code size, especially
when unrolling). Add __LIBETH_WORD_ACCESS and use it there on LE.
Drivers are free to optimize HW-specific callbacks under the same
definition.
Sign
o provides better control for userspace on which CPU
will handle the xmit, as SW interrupts honor IRQ affinity no matter
which core produces XSk xmit descs (while XDPSQs are associated 1:1
with cores having the same ID).
Signed-off-by: Alexander Lobakin
---
include/net/libeth/xsk.h
he same file where dev->xsk_tx_metadata_ops gets set to
avoid duplication of slowpath; at the same time; XSk xmit functions
use local "fast" copy to inline XMO callbacks.
Tx descriptor filling loop is unrolled by 8.
Suggested-by: Maciej Fijalkowski # optimizations
Signed-off-by: Alexander
.
LIBETH_XDP_DEFINE_{BEGIN,END}() is needed for GCC 8+ unfortunately to
let the drivers control which functions will be static and which global
without hitting `-Wold-style-declaration`.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/xdp.h | 195 +++
1 file changed
ader split gives no benefits for XSk Rx, at
least for now.
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/libeth/Kconfig | 2 +-
drivers/net/ethernet/intel/libeth/Makefile | 1 +
drivers/net/ethernet/intel/libeth/priv.h | 6 +
include/net/libeth/tx.h|
of
the last send.
The timer is triggered when finalizing the queue. As long as there is
regular active traffic, the timer doesn't fire.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/types.h | 21 -
include/net/libeth/xdp.h| 57
the corresponding Rx queue
structure via container_of(buff::rxq). If not, a driver can extend
the buff with more fields directly on the stack without touching
libeth_xdp definitions.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/types.h | 22 ++
include/net/libeth/x
gram presence as it's always true for XSk.
Suggested-by: Maciej Fijalkowski # optimizations
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/libeth/priv.h | 3 +
include/net/libeth/xdp.h | 17 +-
include/net/libeth/xsk.h
r
the time of cleaning and filling the descriptors.
As long as this particular queue is not used by more than 1 CPU,
the impact is minimal (runtime check for boolean twice per 16+
descriptors).
Suggested-by: Maciej Fijalkowski # static key
Signed-off-by: Alexander Lobakin
---
include/net/l
End the XDP section by adding helpers to setup XDP features, flipping
.ndo_xdp_xmit() support at runtime (in case when it's not always on),
and calculating the queue clean/refill threshold.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/xdp.h
n its headroom, assuming it
has enough of it for an 8 (or 4) byte value.
In addition to @prep and @xmit driver callbacks in XDP_TX, xmit also
needs @finalize to kick the XDPSQ after filling.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/tx.h | 6 +
include/net/libeth/
here. Define a
minimal structure, which would store only the fields essential to
restore it.
I was able to pack it into 16 bytes, which is only 8 bytes bigger
than `struct sk_buff *skb` on x64.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/types.h | 23
in
we don't need inlining but
convenience.
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/libeth/Makefile | 1 +
include/net/libeth/types.h | 21 ++-
drivers/net/ethernet/intel/libeth/priv.h | 26 +
include/net/libeth/tx.h
Suggested-by: Maciej Fijalkowski # xmit logic
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/libeth/Kconfig | 10 +-
drivers/net/ethernet/intel/libeth/Makefile | 6 +-
include/net/libeth/tx.h| 11 +-
include/net/libeth/xdp.h
faster
"unsafe" underscored netmem <--> virt accessors as netmem_is_net_iov()
is always false in that case, but consumes some cycles (bit test +
true branch).
Reviewed-by: Mina Almasry
Signed-off-by: Alexander Lobakin
---
include/net/libeth/rx.h | 22 +++
ed-off-by: Alexander Lobakin
---
include/net/libeth/rx.h| 6 +-
drivers/net/ethernet/intel/libeth/rx.c | 20 +++-
2 files changed, 20 insertions(+), 6 deletions(-)
diff --git a/include/net/libeth/rx.h b/include/net/libeth/rx.h
index 7d5dc58984b1..5d99140484
sts here ("XDP for idpf"). Macros for building
driver XDP functions lead to that some implementations (XDP_TX,
ndo_xdp_xmit etc.) consist of really only a few lines.
Alexander Lobakin (17):
libeth, libie: clean symbol exports up a little
libeth: convert to netmem
libeth: support na
Change EXPORT_SYMBOL_NS_GPL(x, "LIBETH") to EXPORT_SYMBOL_GPL(x) +
DEFAULT_SYMBOL_NAMESPACE "LIBETH" to make the code more compact.
Also, explicitly include to satisfy new
requirements from scripts/misc-check.
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/in
From: Maciej Fijalkowski
Date: Thu, 24 Apr 2025 14:48:39 +0200
> On Tue, Apr 15, 2025 at 07:28:12PM +0200, Alexander Lobakin wrote:
>> Start adding XDP-specific code to libeth, namely handling XDP_TX buffers
>> (only sending).
[...]
>> +static __always_inline u32
>&g
From: Loktionov, Aleksandr
Date: Wed, 16 Apr 2025 08:54:39 +0200
>
>
>> -Original Message-
>> From: Intel-wired-lan On Behalf Of
>> Alexander Lobakin
>> Sent: Tuesday, April 15, 2025 7:28 PM
>> To: intel-wired-...@lists.osuosl.org
>>
From: Larysa Zaremba
Date: Tue, 8 Apr 2025 14:47:46 +0200
> This patch series adds the iXD driver, which supports the Intel(R)
> Control Plane PCI Function on Intel E2100 and later IPUs and FNICs.
> It facilitates a centralized control over multiple IDPF PFs/VFs/SFs
> exposed by the same card. Th
s on.
There's a few places in libeth_xdp winning up to several percent from
combined access (both performance and object code size, especially
when unrolling). Add __LIBETH_WORD_ACCESS and use it there on LE.
Drivers are free to optimize HW-specific callbacks under the same
definition.
Sign
o provides better control for userspace on which CPU
will handle the xmit, as SW interrupts honor IRQ affinity no matter
which core produces XSk xmit descs (while XDPSQs are associated 1:1
with cores having the same ID).
Signed-off-by: Alexander Lobakin
---
include/net/libeth/xsk.h
End the XDP section by adding helpers to setup XDP features, flipping
.ndo_xdp_xmit() support at runtime (in case when it's not always on),
and calculating the queue clean/refill threshold.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/xdp.h
gram presence as it's always true for XSk.
Suggested-by: Maciej Fijalkowski # optimizations
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/libeth/priv.h | 3 +
include/net/libeth/xdp.h | 17 +-
include/net/libeth/xsk.h
ader split gives no benefits for XSk Rx, at
least for now.
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/libeth/Kconfig | 2 +-
drivers/net/ethernet/intel/libeth/Makefile | 1 +
drivers/net/ethernet/intel/libeth/priv.h | 6 +
include/net/libeth/tx.h|
he same file where dev->xsk_tx_metadata_ops gets set to
avoid duplication of slowpath; at the same time; XSk xmit functions
use local "fast" copy to inline XMO callbacks.
Tx descriptor filling loop is unrolled by 8.
Suggested-by: Maciej Fijalkowski # optimizations
Signed-off-by: Alexander
.
LIBETH_XDP_DEFINE_{BEGIN,END}() is needed for GCC 8+ unfortunately to
let the drivers control which functions will be static and which global
without hitting `-Wold-style-declaration`.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/xdp.h | 195 +++
1 file changed
here. Define a
minimal structure, which would store only the fields essential to
restore it.
I was able to pack it into 16 bytes, which is only 8 bytes bigger
than `struct sk_buff *skb` on x64.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/types.h | 23
in
the corresponding Rx queue
structure via container_of(buff::rxq). If not, a driver can extend
the buff with more fields directly on the stack without touching
libeth_xdp definitions.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/types.h | 22 ++
include/net/libeth/x
r
the time of cleaning and filling the descriptors.
As long as this particular queue is not used by more than 1 CPU,
the impact is minimal (runtime check for boolean twice per 16+
descriptors).
Suggested-by: Maciej Fijalkowski # static key
Signed-off-by: Alexander Lobakin
---
include/net/l
of
the last send.
The timer is triggered when finalizing the queue. As long as there is
regular active traffic, the timer doesn't fire.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/types.h | 21 -
include/net/libeth/xdp.h| 57
n its headroom, assuming it
has enough of it for an 8 (or 4) byte value.
In addition to @prep and @xmit driver callbacks in XDP_TX, xmit also
needs @finalize to kick the XDPSQ after filling.
Signed-off-by: Alexander Lobakin
---
include/net/libeth/tx.h | 6 +
include/net/libeth/
we don't need inlining but
convenience.
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/libeth/Makefile | 1 +
include/net/libeth/types.h | 21 ++-
drivers/net/ethernet/intel/libeth/priv.h | 26 +
include/net/libeth/tx.h
Suggested-by: Maciej Fijalkowski # xmit logic
Signed-off-by: Alexander Lobakin
---
drivers/net/ethernet/intel/libeth/Kconfig | 10 +-
drivers/net/ethernet/intel/libeth/Makefile | 6 +-
include/net/libeth/tx.h| 11 +-
include/net/libeth/xdp.h
ed-off-by: Alexander Lobakin
---
include/net/libeth/rx.h| 6 +-
drivers/net/ethernet/intel/libeth/rx.c | 20 +++-
2 files changed, 20 insertions(+), 6 deletions(-)
diff --git a/include/net/libeth/rx.h b/include/net/libeth/rx.h
index 7d5dc58984b1..5d99140484
faster
"unsafe" underscored netmem <--> virt accessors as netmem_is_net_iov()
is always false in that case, but consumes some cycles (bit test +
true branch).
Misc: replace explicit EXPORT_SYMBOL_NS_GPL("NS") with
DEFAULT_SYMBOL_NAMESPACE.
Reviewed-by: Mina Almasry
Signe
sts here ("XDP for idpf"). Macros for building
driver XDP functions lead to that some implementations (XDP_TX,
ndo_xdp_xmit etc.) consist of really only a few lines.
Alexander Lobakin (16):
libeth: convert to netmem
libeth: support native XDP and register memory model
libeth: xdp
From: Maciej Fijalkowski
Date: Tue, 11 Mar 2025 16:50:07 +0100
> On Wed, Mar 05, 2025 at 05:21:30PM +0100, Alexander Lobakin wrote:
>> Use libeth XDP infra to support running XDP program on Rx polling.
>> This includes all of the possible verdicts/actions.
>> XDP Tx queue
From: Leon Romanovsky
Date: Thu, 10 Apr 2025 14:23:49 +0300
> On Thu, Apr 10, 2025 at 12:44:33PM +0200, Larysa Zaremba wrote:
>> On Thu, Apr 10, 2025 at 11:21:37AM +0300, Leon Romanovsky wrote:
>>> On Tue, Apr 08, 2025 at 02:47:51PM +0200, Larysa Zaremba wrote:
From: Phani R Burra
From: Leon Romanovsky
Date: Thu, 10 Apr 2025 16:27:06 +0300
> On Thu, Apr 10, 2025 at 02:58:28PM +0200, Larysa Zaremba wrote:
>> On Thu, Apr 10, 2025 at 02:23:49PM +0300, Leon Romanovsky wrote:
>>> On Thu, Apr 10, 2025 at 12:44:33PM +0200, Larysa Zaremba wrote:
On Thu, Apr 10, 2025 at 11:21:
From: Leon Romanovsky
Date: Thu, 10 Apr 2025 16:44:43 +0300
> On Thu, Apr 10, 2025 at 03:05:19PM +0200, Alexander Lobakin wrote:
>> From: Leon Romanovsky
>> Date: Thu, 10 Apr 2025 14:23:49 +0300
>>
>>> On Thu, Apr 10, 2025 at 12:44:33PM +0200, Larysa Zaremba wrote
From: Leon Romanovsky
Date: Thu, 10 Apr 2025 16:58:43 +0300
> On Thu, Apr 10, 2025 at 03:33:40PM +0200, Alexander Lobakin wrote:
>> From: Leon Romanovsky
>> Date: Thu, 10 Apr 2025 16:27:06 +0300
>>
>>> On Thu, Apr 10, 2025 at 02:58:28PM +0200, Larysa Zaremba wrote
From: Maciej Fijalkowski
Date: Wed, 19 Mar 2025 17:29:37 +0100
> On Mon, Mar 17, 2025 at 03:50:11PM +0100, Alexander Lobakin wrote:
>> From: Maciej Fijalkowski
>> Date: Fri, 7 Mar 2025 14:27:13 +0100
>>
>>> On Wed, Mar 05, 2025 at 05:21:27PM +0100, Alexander Lo
From: Alexander Lobakin
Date: Tue, 8 Apr 2025 15:22:48 +0200
> From: Maciej Fijalkowski
> Date: Tue, 11 Mar 2025 15:05:38 +0100
>
>> On Wed, Mar 05, 2025 at 05:21:19PM +0100, Alexander Lobakin wrote:
>>> "Couple" is a bit humbly... Add the
From: Alexander Lobakin
Date: Tue, 1 Apr 2025 15:11:50 +0200
> From: Maciej Fijalkowski
> Date: Wed, 19 Mar 2025 17:19:44 +0100
[...]
>> Not sure what to say here. Your time dedicated for making this work easier
>> to swallow means less time dedicated for going through th
From: Maciej Fijalkowski
Date: Tue, 11 Mar 2025 17:08:22 +0100
> On Wed, Mar 05, 2025 at 05:21:31PM +0100, Alexander Lobakin wrote:
>> Use libeth XDP infra to implement .ndo_xdp_xmit() in idpf.
>> The Tx callbacks are reused from XDP_TX code. XDP redirect target
>> f
From: Maciej Fijalkowski
Date: Tue, 11 Mar 2025 15:05:38 +0100
> On Wed, Mar 05, 2025 at 05:21:19PM +0100, Alexander Lobakin wrote:
>> "Couple" is a bit humbly... Add the following functionality to libeth:
[...]
>> +struct libeth_rq_napi_stats {
>> +
From: Maciej Fijalkowski
Date: Wed, 19 Mar 2025 17:19:44 +0100
> On Mon, Mar 17, 2025 at 04:26:04PM +0100, Alexander Lobakin wrote:
>> From: Maciej Fijalkowski
>> Date: Tue, 11 Mar 2025 15:05:38 +0100
>>
>>> On Wed, Mar 05, 2025 at 05:21:19PM +0100, Alexander Lob
From: Ahmed Zaki
Date: Mon, 24 Mar 2025 07:49:38 -0600
> Use the new virtchnl2 OP codes to communicate with the Control Plane to
> add flow steering filters. We add the basic functionality for ADD/Delete
> with TCP/UDP IPv4 only. Support for other OP codes and protocols will be
> added later.
>
From: Ahmed Zaki
Date: Mon, 24 Mar 2025 07:49:36 -0600
> Add basic flow steering. For now, we support IPv4 and TCP/UDP only.
1. Very poor cover letter. I'd suggest describing a bit more here.
2. net-next is closed, so this is RFC at max.
3. I haven't seen this reviewed properly on our internal M
From: Alexander Lobakin
Date: Wed, 12 Mar 2025 18:16:11 +0100
> From: Eric Dumazet
> Date: Fri, 7 Mar 2025 11:28:36 +0100
>
>> On Wed, Mar 5, 2025 at 5:22 PM Alexander Lobakin
>> wrote:
>>>
>>> Add the missing linking of NAPIs to netdev queues when ena
1 - 100 of 540 matches
Mail list logo
