> -Original Message-
> From: Intel-wired-lan On Behalf Of
> Jacob Keller
> Sent: Thursday, October 17, 2024 1:28 AM
> To: Kolacinski, Karol ; intel-wired-
> l...@lists.osuosl.org
> Cc: net...@vger.kernel.org; Nguyen, Anthony L
> ; Kitszel, Przemyslaw
>
> Subject: Re: [Intel-wired-lan] [
Link queues to NAPI instances via netdev-genl API so that users can
query this information with netlink. Handle a few cases in the driver:
1. Link/unlink the NAPIs when XDP is enabled/disabled
2. Handle IGC_FLAG_QUEUE_PAIRS enabled and disabled
Example output when IGC_FLAG_QUEUE_PAIRS is enabl
Greetings:
Welcome to v5.
See changelog below and in each patch for changes from v4 [1].
This revision was created due to a report from Vitaly [2], that my v4
was re-introducing a potential deadlock in runtime_resume which was
fixed in commit: 6f31d6b: "igc: Refactor runtime power management flo
Link IRQs to NAPI instances via netdev-genl API so that users can query
this information with netlink.
Compare the output of /proc/interrupts (noting that IRQ 128 is the
"other" IRQ which does not appear to have a NAPI instance):
$ cat /proc/interrupts | grep enp86s0 | cut --delimiter=":" -f1
12
From: Karol Kolacinski
Fix ETH56G FC-FEC incorrect Rx offset value by changing it from -255.96
to -469.26 ns.
Those values are derived from HW spec and reflect internal delays.
Hex value is a fixed point representation in Q23.9 format.
Fixes: 7cab44f1c35f ("ice: Introduce ETH56G PHY model for E
From: Karol Kolacinski
Current implementation checks revision of all PHYs on all PFs, which is
incorrect and may result in initialization failure. Check only the
revision of the current PHY.
Fixes: 7cab44f1c35f ("ice: Introduce ETH56G PHY model for E825C products")
Reviewed-by: Arkadiusz Kubalew
E825 products have incorrect initialization procedure, which may lead to
initialization failures and register values.
Fix E825 products initialization by adding correct sync delay, checking
the PHY revision only for current PHY and adding proper destination
device when reading port/quad.
In addit
From: Karol Kolacinski
Quad registers are read/written incorrectly. E825 devices always use
quad 0 address and differentiate between the PHYs by changing SBQ
destination device (phy_0 or phy_0_peer).
Add helpers for reading/writing PTP registers shared per quad and use
correct quad address and S
From: Karol Kolacinski
Driver always naively assumes, that for PTP purposes, PHY lane to
configure is corresponding to PF ID.
This is not true for some port configurations, e.g.:
- 2x50G per quad, where lanes used are 0 and 2 on each quad, but PF IDs
are 0 and 1
- 100G per quad on 2 quads, whe
Currently HW support of PTP/timesync solutions in network PHY chips can be
implemented with two different approaches, the timestamp maybe latched
either at the beginning or after the Start of Frame Delimiter (SFD) [1].
Allow ptp device drivers to provide user with control over the HW
timestamp lat
Allow user to control the latch point of ptp HW timestamps in E825
devices.
Usage, examples:
** Obtain current state:
$ cat /sys/class/net/eth/device/ptp/ts_point
Command returns enum/integer:
* 0 - timestamp latched by PHY at the beginning of SFD,
* 1 - timestamp latched by PHY after the SFD,
*
HW support of PTP/timesync solutions in network PHY chips can be
achieved with two different approaches, the timestamp maybe latched
either in the beginning or after the Start of Frame Delimiter (SFD) [1].
Allow ptp device drivers to provide user with control over the timestamp
latch point.
[1] h
> -Original Message-
> From: Intel-wired-lan On Behalf Of
> Karol Kolacinski
> Sent: Thursday, October 10, 2024 4:21 PM
> To: intel-wired-...@lists.osuosl.org
> Cc: net...@vger.kernel.org; Kolacinski, Karol ;
> Kubalewski, Arkadiusz ; Nguyen, Anthony
> L ; Kitszel, Przemyslaw
>
> Subje
On 10/25/24 22:17, Rosen Penev wrote:
The latter is the preferred way to copy ethtool strings.
Avoids manually incrementing the pointer. Cleans up the code quite well.
Signed-off-by: Rosen Penev
---
v2: add iwl-next tag. use inline int in for loops.
.../net/ethernet/intel/e1000/e1000_ethto
On 10/27/24 15:19, R Sundar wrote:
Use string choice helpers for better readability.
Reported-by: kernel test robot
Reported-by: Julia Lawall
Closes: https://lore.kernel.org/r/202410121553.srnfzc2m-...@intel.com/
Signed-off-by: R Sundar
---
thanks, this indeed covers all "enabled/disabled"
Use generic devlink PF MSI-X parameter to allow user to change MSI-X
range.
Add notes about this parameters into ice devlink documentation.
Reviewed-by: Wojciech Drewek
Signed-off-by: Michal Swiatkowski
---
Documentation/networking/devlink/ice.rst | 11 +++
.../net/ethernet/intel/ice/devl
With dynamic approach to alloc MSI-X there is no sense to statically
split MSI-X between PF features.
Splitting was also calculating needed MSI-X. Move this part to separate
function and use as max value.
Remove ICE_ESWITCH_MSIX, as there is no need for additional MSI-X for
switchdev.
Reviewed-b
Move responsibility of MSI-X requesting for RDMA feature from ice driver
to irdma driver. It is done to allow simple fallback when there is not
enough MSI-X available.
Change amount of MSI-X used for control from 4 to 1, as it isn't needed
to have more than one MSI-X for this purpose.
Signed-off-
Remove the field to allow having more queues than MSI-X on VSI. As
default the number will be the same, but if there won't be more MSI-X
available VSI can run with at least one MSI-X.
Reviewed-by: Wojciech Drewek
Signed-off-by: Michal Swiatkowski
---
drivers/net/ethernet/intel/ice/ice.h
Previous implementation assumes that there is 1:1 matching between
vectors and queues. It isn't always true.
Get minimum value from Rx/Tx queues to determine combined queues number.
Signed-off-by: Michal Swiatkowski
---
drivers/net/ethernet/intel/ice/ice_ethtool.c | 3 +--
1 file changed, 1 ins
After implementing pf->msix.max field, base vector for other use cases
(like VFs) can be fixed. This simplify code when changing MSI-X amount
on particular VF, because there is no need to move a base vector.
A fixed base vector allows to reserve vectors from the beginning
instead of from the end,
Implement enable_rdma devlink parameter to allow user to turn RDMA
feature on and off.
It is useful when there is no enough interrupts and user doesn't need
RDMA feature.
Reviewed-by: Jan Sokolowski
Reviewed-by: Przemek Kitszel
Signed-off-by: Michal Swiatkowski
---
.../net/ethernet/intel/ice/
Flow director needs only one MSI-X. Load it before RDMA to save MSI-X
for it.
Signed-off-by: Michal Swiatkowski
---
drivers/net/ethernet/intel/ice/ice_main.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c
b/drivers/net/etherne
Hi,
It is another try to allow user to manage amount of MSI-X used for each
feature in ice. First was via devlink resources API, it wasn't accepted
in upstream. Also static MSI-X allocation using devlink resources isn't
really user friendly.
This try is using more dynamic way. "Dynamic" across wh
It can be needed to have some MSI-X allocated as static and rest as
dynamic. For example on PF VSI. We want to always have minimum one MSI-X
on it, because of that it is allocated as a static one, rest can be
dynamic if it is supported.
Change the ice_get_irq_res() to allow using static entries if
From: Intel-wired-lan on behalf of
Diomidis Spinellis
Sent: Thursday, October 17, 2024 10:58 AM
To: Nguyen, Anthony L ; Kitszel, Przemyslaw
Cc: intel-wired-...@lists.osuosl.org ;
Diomidis Spinellis
Subject: [Intel-wired-lan] [PATCH] ixgbe: Break include dependency cycle
Header ixgbe_type.h
On Mon, Oct 28, 2024 at 09:00:06AM -0700, Joe Damato wrote:
> On Mon, Oct 28, 2024 at 08:50:38AM -0700, Joe Damato wrote:
> > On Sun, Oct 27, 2024 at 11:49:33AM +0200, Lifshits, Vitaly wrote:
> > >
> > > On 10/23/2024 12:52 AM, Joe Damato wrote:
> > > > Link queues to NAPI instances via netdev-gen
On 10/28/2024 9:00 AM, Joe Damato wrote:
>
> I see, so it looks like there is:
>- resume
>- runtime_resume
>
> The bug I am reintroducing is runtime_resume already holding RTNL
> before my added call to rtnl_lock.
>
> OK.
>
> Does resume also hold rtnl before the driver's igc_resume
On Mon, Oct 28, 2024 at 11:53:55AM -0700, Jacob Keller wrote:
>
>
> On 10/28/2024 9:00 AM, Joe Damato wrote:
> >
> > I see, so it looks like there is:
> >- resume
> >- runtime_resume
> >
> > The bug I am reintroducing is runtime_resume already holding RTNL
> > before my added call to rt
On Sun, Oct 27, 2024 at 11:49:33AM +0200, Lifshits, Vitaly wrote:
>
> On 10/23/2024 12:52 AM, Joe Damato wrote:
> > Link queues to NAPI instances via netdev-genl API so that users can
> > query this information with netlink. Handle a few cases in the driver:
> >1. Link/unlink the NAPIs when XD
Cleans things up nicely :)
Reviewed-by: Lee Trager
On 10/25/24 1:37 PM, Rosen Penev wrote:
The latter is the preferred way to copy ethtool strings.
Avoids manually incrementing the pointer. Cleans up the code quite well.
Signed-off-by: Rosen Penev
---
v2: fix wrong variable in for loop
Use string choice helpers for better readability.
Reported-by: kernel test robot
Reported-by: Julia Lawall
Closes: https://lore.kernel.org/r/202410121553.srnfzc2m-...@intel.com/
Signed-off-by: R Sundar
---
Reported in linux repository.
tree: https://git.kernel.org/pub/scm/linux/kernel/git/t
Fix Flow Director not allowing to re-map traffic to 0th queue when action
is configured to drop (and vice versa).
The current implementation of ethtool callback in the ice driver forbids
change Flow Director action from 0 to -1 and from -1 to 0 with an error,
e.g:
# ethtool -U eth2 flow-type tcp
On Mon, Oct 28, 2024 at 08:50:38AM -0700, Joe Damato wrote:
> On Sun, Oct 27, 2024 at 11:49:33AM +0200, Lifshits, Vitaly wrote:
> >
> > On 10/23/2024 12:52 AM, Joe Damato wrote:
> > > Link queues to NAPI instances via netdev-genl API so that users can
> > > query this information with netlink. Han
34 matches
Mail list logo
