Hi Sriram,
On Fri Aug 04 2023, Sriram Yagnaraman wrote:
> The first couple of patches adds helper funcctions to prepare for AF_XDP
> zero-copy support which comes in the last couple of patches, one each
> for Rx and TX paths.
>
> As mentioned in v1 patchset [0], I don't have access to an actual IG
VSI for subfunction is only allocated when port representor is created.
The VSI id isn't correct at this state. Update the representor with
correct VSI id after the VSI is configured.
Signed-off-by: Michal Swiatkowski
---
It should be squashed with:
8a469e1a1cf9 ("ice: create port representor for
On Mon, Jun 24, 2024 at 10:05:10AM +0200, Michal Swiatkowski wrote:
> Since the port representors are added one by one there is no need to do
> eswitch rebuild. Each port representor is detached and attached in VF
> reset path.
>
> Reviewed-by: Wojciech Drewek
> Signed-off-by: Michal Swiatkowski
On Thu, Jun 06, 2024 at 01:24:52PM +0200, Michal Swiatkowski wrote:
> When subfunction VSI is open the same code as for PF VSI should be
> executed. Also when up is complete. Reflect that in code by adding
> subfunction VSI to consideration.
>
> In case of stopping, PF doesn't have additional task
On Thu, Jun 06, 2024 at 01:24:55PM +0200, Michal Swiatkowski wrote:
> From: Piotr Raczynski
>
> Configure netdevice for subfunction usecase. Mostly it is reusing ops
> from the PF netdevice.
>
> SF netdev is linked to devlink port registered after SF activation.
>
> Reviewed-by: Simon Horman
>
On Thu, Jun 27, 2024 at 01:07:22PM +0200, Maciej Fijalkowski wrote:
> On Thu, Jun 06, 2024 at 01:24:52PM +0200, Michal Swiatkowski wrote:
> > When subfunction VSI is open the same code as for PF VSI should be
> > executed. Also when up is complete. Reflect that in code by adding
> > subfunction VSI
On Thu, Jun 27, 2024 at 01:13:47PM +0200, Maciej Fijalkowski wrote:
> On Thu, Jun 06, 2024 at 01:24:55PM +0200, Michal Swiatkowski wrote:
> > From: Piotr Raczynski
> >
> > Configure netdevice for subfunction usecase. Mostly it is reusing ops
> > from the PF netdevice.
> >
> > SF netdev is linked
Hi,
changes included in this patchset address an issue that customer has
been facing when AF_XDP ZC Tx sockets were used in combination with flow
control and regular Tx traffic.
After executing:
ethtool --set-priv-flags $dev link-down-on-close on
ethtool -A $dev rx on tx on
launching multiple ZC
From: Michal Kubiak
Address a scenario in which XSK ZC Tx produces descriptors to XDP Tx
ring when link is either not yet fully initialized or process of
stopping the netdev has already started. To avoid this, add checks
against carrier readiness in ice_xsk_wakeup() and in ice_xmit_zc().
One coul
When ice driver is spammed with multiple xdpsock instances and flow
control is enabled, there are cases when Rx queue gets stuck and unable
to reflect the disable state in QRX_CTRL register. Similar issue has
previously been addressed in commit 13a6233b033f ("ice: Add support to
enable/disable all
Given that ice_qp_dis() is called under rtnl_lock, synchronize_net() can
be called instead of synchronize_rcu() so that XDP rings can finish its
job in a faster way. Also let us do this as earlier in XSK queue disable
flow.
Additionally, turn off regular Tx queue before disabling irqs and NAPI.
F
Don't bail out right when spotting an error within ice_qp_{dis,ena}()
but rather track error and go through whole flow of disabling and
enabling queue pair.
Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Reviewed-by: Shannon Nelson
Tested-by: Chandan Kumar Rout (A Contingent Worker at
Inte
This so we prevent Tx timeout issues. One of conditions checked on
running in the background dev_watchdog() is netif_carrier_ok(), so let
us turn it off when we disable the queues that belong to a q_vector
where XSK pool is being configured. Turn carrier on in ice_qp_ena()
only when ice_get_link_st
xsk_buff_pool pointers that ice ring structs hold are updated via
ndo_bpf that is executed in process context while it can be read by
remote CPU at the same time within NAPI poll. Use synchronize_net()
after pointer update and {READ,WRITE}_ONCE() when working with mentioned
pointer.
Fixes: 2d4238f
It is read by data path and modified from process context on remote cpu
so it is needed to use WRITE_ONCE to clear the pointer.
Fixes: efc2214b6047 ("ice: Add support for XDP")
Reviewed-by: Shannon Nelson
Tested-by: Chandan Kumar Rout (A Contingent Worker at
Intel)
Signed-off-by: Maciej Fijalko
ice_cfg_txq_interrupt() internally handles XDP Tx ring. Do not use
ice_for_each_tx_ring() in ice_qvec_cfg_msix() as this causing us to
treat XDP ring that belongs to queue vector as Tx ring and therefore
misconfiguring the interrupts.
Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Reviewed-by
Optimize the process of creating a recipe in the switch block by removing
duplicate switch ID words and changing how result indexes are fitted into
recipes. In many cases this can decrease the number of recipes required to
add a certain set of rules, potentially allowing a more varied set of rules
From: Michal Swiatkowski
The content of the first read recipe is used as a template when adding a
recipe. It isn't needed - only prune index is directly set from there. Set
it in the code instead. Also, now there's no need to set rid and lookup
indexes to 0, as the whole recipe buffer is initiali
From: Michal Swiatkowski
Remove root_buf from recipe struct. Its only usage was in ice_find_recp(),
where if recipe had an inverse action, it was skipped, but actually the
driver never adds inverse actions, so effectively it was pointless.
Without root_buf, the recipe data element in ice_add_sw_
Currently when creating switch recipes, switch ID is always added as the
first word in every recipe. There are only 5 words in a recipe, so one
word is always wasted. This is also true for the last recipe, which stores
result indexes (in case of chain recipes). Therefore the maximum usable
length o
Remove field_off as it's never used.
Remove done bitmap, as its value is only checked and never assigned.
Reusing sub-recipes while creating new root recipes is currently not
supported in the driver.
Reviewed-by: Przemek Kitszel
Signed-off-by: Marcin Szycik
---
.../ethernet/intel/ice/ice_proto
Track the number of rules and recipes added to switch. Add a tracepoint to
ice_aq_sw_rules(), which shows both rule and recipe count. This information
can be helpful when designing a set of rules to program to the hardware, as
it shows where the practical limit is. Actual limits are known (64 recip
From: Michal Swiatkowski
Remove unnecessary size checks when copying bitmaps in ice_add_sw_recipe()
and replace them with compile time assert. Check if the bitmaps are equal
size, as they are copied both ways.
Signed-off-by: Michal Swiatkowski
Reviewed-by: Przemek Kitszel
Co-developed-by: Marc
Remove several members of struct ice_sw_recipe and struct
ice_prot_lkup_ext. Remove struct ice_recp_grp_entry and struct
ice_pref_recipe_group, since they are now unused as well.
All of the deleted members were only written to and never read, so it's
pointless to keep them.
Reviewed-by: Przemek K
This series cleans up current PTP GPIO pin handling, fixes minor bugs,
refactors implementation for all products, introduces SDP (Software
Definable Pins) for E825C and implements reading SDP section from NVM
for E810 products.
Karol Kolacinski (5):
ice: Implement ice_ptp_pin_desc
ice: Add SDP
Add a new internal structure describing PTP pins.
Use the new structure for all non-E810T products.
Reviewed-by: Arkadiusz Kubalewski
Signed-off-by: Karol Kolacinski
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 271 ---
drivers/net/ethernet/intel/ice/ice_ptp.h | 65 --
Add support of PTP SDPs (Software Definable Pins) for E825C products.
Reviewed-by: Arkadiusz Kubalewski
Signed-off-by: Karol Kolacinski
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 22 +++---
1 file changed, 19 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/i
Instead of having separate PTP GPIO implementation for E810T, use
existing one from all other products.
Reviewed-by: Arkadiusz Kubalewski
Signed-off-by: Karol Kolacinski
---
drivers/net/ethernet/intel/ice/ice_gnss.c | 4 +-
drivers/net/ethernet/intel/ice/ice_ptp.c| 460 +
Cache original PTP GPIO requests instead of saving each parameter in
internal structures for periodic output or external timestamp request.
Factor out all periodic output register writes from ice_ptp_cfg_clkout
to a separate function to improve readability.
Reviewed-by: Arkadiusz Kubalewski
Sign
When setting a new supported function for a pin on E810, disable other
enabled pin that shares the same GPIO.
Reviewed-by: Arkadiusz Kubalewski
Signed-off-by: Karol Kolacinski
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 65
1 file changed, 65 insertions(+)
diff --gi
From: Sergey Temerkhanov
Implement 1PPS signal enabling/disabling in CGU. The amplitude is
always the maximum in this implementation
Reviewed-by: Arkadiusz Kubalewski
Signed-off-by: Sergey Temerkhanov
Co-developed-by: Karol Kolacinski
Signed-off-by: Karol Kolacinski
---
drivers/net/ethernet
From: Yochai Hagvi
PTP pins assignment and their related SDPs (Software Definable Pins) are
currently hardcoded.
Fix that by reading NVM section instead on products supporting this,
which are E810 products.
If SDP section is not defined in NVM, the driver continues to use the
hardcoded table.
Re
Dear Karol, dear Sergey,
Thank you for the patch.
Am 27.06.24 um 17:09 schrieb Karol Kolacinski:
From: Sergey Temerkhanov
Implement 1PPS signal enabling/disabling in CGU. The amplitude is
always the maximum in this implementation
(Please add a dot/period at the end of sentences.)
Could y
On Thursday, 27 June 2024, 09:07:55 CEST, Kurt Kanzenbach wrote:
> Hi Sriram,
>
> On Fri Aug 04 2023, Sriram Yagnaraman wrote:
> > The first couple of patches adds helper funcctions to prepare for AF_XDP
> > zero-copy support which comes in the last couple of patches, one each
> > for Rx and TX pa
Hi Benjamin,
On Thu Jun 27 2024, Benjamin Steinke wrote:
> On Thursday, 27 June 2024, 09:07:55 CEST, Kurt Kanzenbach wrote:
>> Hi Sriram,
>>
>> On Fri Aug 04 2023, Sriram Yagnaraman wrote:
>> > The first couple of patches adds helper funcctions to prepare for AF_XDP
>> > zero-copy support which c
On Tue, Jun 25, 2024 at 08:49:53PM +0200, Aleksandr Loktionov wrote:
> Remove wrong EIO to EGAIN conversion and pass all errors as is.
>
> After commit 230f3d53a547 ("i40e: remove i40e_status"), which should only
> replace F/W specific error codes with Linux kernel generic, all EIO errors
> sudden
On Wed, Jun 26, 2024 at 09:19:12AM +0800, Abdul Rahim, Faizal wrote:
> Added Vladimir and Husaini in CC.
>
> On 25/6/2024 4:26 pm, Faizal Rahim wrote:
> > Following the implementation of "igc: Add TransmissionOverrun counter"
> > patch, when a taprio command is triggered by user, igc processes two
37 matches
Mail list logo
