From: Jason Xing
Since we never support this feature for I40E driver, we don't have to
display the value when using 'ethtool -c eth0'.
Before this patch applied, the rx-frames-irq is 256 which is consistent
with tx-frames-irq. Apparently it could mislead users.
Signed-off
From: Jason Xing
Since we never support this feature for I40E driver, we don't have to
display the value when using 'ethtool -c eth0'.
Before this patch applied, the rx-frames-irq is 256 which is consistent
with tx-frames-irq. Apparently it could mislead users.
Signed-off
On Tue, May 13, 2025 at 12:08 PM Haifeng Xu wrote:
>
> Hi all,
>
> If the packets arrive at the rx and then raise soft irq to handle it,
> but in i40e_clean_rx_irq, status_error_len is 0 and return.
Directly "return"? What version of I40E are you looking at?
> The data isn't fet
Hi all,
I'm currently faced with one tough issue caused by zero copy mode in
xsk with ixgbe driver loaded. The case is that if we use xdpsock to
send descs, nearly at the same time normal packets from other tx
queues cannot be transmitted/completed at all.
Here is how I try:
1. run iperf or ping
On Fri, Jul 18, 2025 at 6:27 PM Larysa Zaremba wrote:
>
> On Wed, Jul 16, 2025 at 11:41:42AM +0800, Jason Xing wrote:
> > Hi all,
> >
> > I'm currently faced with one tough issue caused by zero copy mode in
> > xsk with ixgbe driver loaded. The case is that
From: Jason Xing
Resolve the budget underflow which leads to returning true in ixgbe_xmit_zc
even when the budget of descs are thoroughly consumed.
Before this patch, when the budget is decreased to zero and finishes
sending the last allowed desc in ixgbe_xmit_zc, it will always turn back
and
From: Jason Xing
Since 'budget' parameter in ixgbe_clean_xdp_tx_irq() takes no effect,
the patch removes it. No functional change here.
Signed-off-by: Jason Xing
---
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c| 2 +-
drivers/net/ethernet/intel/ixgbe/ixgbe_txrx_co
From: Jason Xing
- Adjust ixgbe_desc_unused as the budget value.
- Avoid checking desc_unused over and over again in the loop.
The patch makes ixgbe follow i40e driver that was done in commit
1fd972ebe523 ("i40e: move check of full Tx ring to outside of send loop").
[ Note that the
From: Jason Xing
Use the common interface to see if the desc is the end of packets. If
so, set IXGBE_TXD_CMD_EOP bit instead of setting for all preceding
descriptors. This is also how i40e driver did in commit a92b96c4ae10
("i40e: xsk: add TX multi-buffer support").
Signed-off-by:
From: Jason Xing
The series mostly follows the development of i40e/ice to improve the
performance for zerocopy mode in the tx path.
Jason Xing (5):
ixgbe: xsk: remove budget from ixgbe_clean_xdp_tx_irq
ixgbe: xsk: resolve the underflow of budget in ixgbe_xmit_zc
ixgbe: xsk: use
From: Jason Xing
Like what i40e driver initially did in commit 3106c580fb7cf
("i40e: Use batched xsk Tx interfaces to increase performance"), use
the batched xsk feature to transmit packets.
Signed-off-by: Jason Xing
---
drivers/net/ethernet/intel/ixgbe/ixgbe_x
On Tue, Jul 22, 2025 at 7:05 AM Jason Xing wrote:
>
> On Mon, Jul 21, 2025 at 11:37 PM Stanislav Fomichev
> wrote:
> >
> > On 07/21, Jason Xing wrote:
> > > From: Jason Xing
> > >
> > > The issue can happen when the budget number of descs are co
On Mon, Jul 21, 2025 at 11:37 PM Stanislav Fomichev
wrote:
>
> On 07/21, Jason Xing wrote:
> > From: Jason Xing
> >
> > The issue can happen when the budget number of descs are consumed. As
> > long as the budget is decreased to zero, it will again go into
> >
From: Jason Xing
There is no break time in the while() loop, so every time at the end of
igb_xmit_zc(), underflow of nb_pkts will occur, which renders the return
value always false. But theoretically, the result should be set after
calling xsk_tx_peek_release_desc_batch(). We can take
Hi Paul,
On Mon, Jul 21, 2025 at 4:56 PM Paul Menzel wrote:
>
> Dear Jason,
>
>
> Thank you for your patch.
Thanks for your quick response and review :)
>
> Am 21.07.25 um 10:33 schrieb Jason Xing:
> > From: Jason Xing
> >
> > The issue can happen when
From: Jason Xing
Fix two underflow issues around {stmmac_xdp|igb}_xmit_zc().
Jason Xing (2):
stmmac: xsk: fix underflow of budget in zerocopy mode
igb: xsk: solve underflow of nb_pkts in zerocopy mode
drivers/net/ethernet/intel/igb/igb_xsk.c | 2 +-
drivers/net/ethernet/stmicro
From: Jason Xing
The issue can happen when the budget number of descs are consumed. As
long as the budget is decreased to zero, it will again go into
while (budget-- > 0) statement and get decreased by one, so the
underflow issue can happen. It will lead to returning true whereas the
expec
On Mon, Jul 21, 2025 at 3:25 PM Larysa Zaremba wrote:
>
> On Sat, Jul 19, 2025 at 01:26:18PM +0800, Jason Xing wrote:
> > On Fri, Jul 18, 2025 at 6:27 PM Larysa Zaremba
> > wrote:
> > >
> > > On Wed, Jul 16, 2025 at 11:41:42AM +0800, Jason Xing wrote:
Hi Simon,
On Mon, Jul 21, 2025 at 6:12 PM Simon Horman wrote:
>
> On Mon, Jul 21, 2025 at 04:33:43PM +0800, Jason Xing wrote:
> > From: Jason Xing
> >
> > There is no break time in the while() loop, so every time at the end of
> > igb_xmit_zc(), underflow of nb_
On Mon, Jul 21, 2025 at 5:25 PM Larysa Zaremba wrote:
>
> On Mon, Jul 21, 2025 at 05:03:07PM +0800, Jason Xing wrote:
> > On Mon, Jul 21, 2025 at 3:25 PM Larysa Zaremba
> > wrote:
> > >
> > > On Sat, Jul 19, 2025 at 01:26:18PM +0800, Jason Xing wrote:
>
From: Jason Xing
A negative overflow can happen when the budget number of descs are
consumed. as long as the budget is decreased to zero, it will again go
into while (budget-- > 0) statement and get decreased by one, so the
overflow issue can happen. It will lead to returning true whereas
From: Jason Xing
There is no break time in the while() loop, so every time at the end of
igb_xmit_zc(), negative overflow of nb_pkts will occur, which renders
the return value always false. But theoretically, the result should be
set after calling xsk_tx_peek_release_desc_batch(). We can take
From: Jason Xing
Fix two negative overflow issues around {stmmac_xdp|igb}_xmit_zc().
Jason Xing (2):
stmmac: xsk: fix negative overflow of budget in zerocopy mode
igb: xsk: solve negative overflow of nb_pkts in zerocopy mode
drivers/net/ethernet/intel/igb/igb_xsk.c | 3
On Wed, Jul 23, 2025 at 12:10 PM kernel test robot wrote:
>
> Hi Jason,
>
> kernel test robot noticed the following build errors:
>
> [auto build test ERROR on net/main]
>
> url:
> https://github.com/intel-lab-lkp/linux/commits/Jason-Xing/stmmac-xsk-fix-underflow
From: Jason Xing
Fix two underflow issues around {stmmac_xdp|igb}_xmit_zc().
Jason Xing (2):
stmmac: xsk: fix underflow of budget in zerocopy mode
igb: xsk: solve underflow of nb_pkts in zerocopy mode
drivers/net/ethernet/intel/igb/igb_xsk.c | 3 +--
drivers/net/ethernet/stmicro
From: Jason Xing
There is no break time in the while() loop, so every time at the end of
igb_xmit_zc(), underflow of nb_pkts will occur, which renders the return
value always false. But theoretically, the result should be set after
calling xsk_tx_peek_release_desc_batch(). We can take
From: Jason Xing
An underflow can happen when the budget number of descs are consumed.
as long as the budget is decreased to zero, it will again go into
while (budget-- > 0) statement and get decreased by one, so the
underflow issue can happen. It will lead to returning true whereas the
expec
On Tue, Jul 22, 2025 at 10:16 PM Willem de Bruijn
wrote:
>
> Jason Xing wrote:
> > Hi Paul,
> >
> > On Mon, Jul 21, 2025 at 4:56 PM Paul Menzel wrote:
> > >
> > > Dear Jason,
> > >
> > >
> > > Thank you for your patch.
&g
On Wed, Jul 23, 2025 at 1:29 AM Willem de Bruijn
wrote:
>
> Jason Xing wrote:
> > On Tue, Jul 22, 2025 at 10:16 PM Willem de Bruijn
> > wrote:
> > >
> > > Jason Xing wrote:
> > > > Hi Paul,
> > > >
> > > > On Mon, Jul 2
On Tue, Jul 22, 2025 at 9:51 PM Jason Xing wrote:
>
> From: Jason Xing
>
> Fix two underflow issues around {stmmac_xdp|igb}_xmit_zc().
>
> Jason Xing (2):
> stmmac: xsk: fix underflow of budget in zerocopy mode
> igb: xsk: solve underflow of nb_pkts in zerocopy mode
From: Jason Xing
Resolve the budget negative overflow which leads to returning true in
ixgbe_xmit_zc even when the budget of descs are thoroughly consumed.
Before this patch, when the budget is decreased to zero and finishes
sending the last allowed desc in ixgbe_xmit_zc, it will always turn
Hi Maciej,
On Fri, Jul 25, 2025 at 6:00 PM Maciej Fijalkowski
wrote:
>
> On Sun, Jul 20, 2025 at 05:11:23PM +0800, Jason Xing wrote:
> > From: Jason Xing
> >
> > Use the common interface to see if the desc is the end of packets. If
> > so, set IXGBE_TXD_CMD_EOP
Hi Tony,
On Fri, Jul 25, 2025 at 4:21 AM Tony Nguyen wrote:
>
>
>
> On 7/20/2025 2:11 AM, Jason Xing wrote:
> > From: Jason Xing
> >
> > Resolve the budget underflow which leads to returning true in ixgbe_xmit_zc
> > even when the budget of descs are thor
On Sun, Jul 27, 2025 at 10:47 PM Simon Horman wrote:
>
> On Sun, Jul 27, 2025 at 10:16:10PM +0800, Jason Xing wrote:
> > Hi Simon,
> >
> > On Sun, Jul 27, 2025 at 9:55 PM Simon Horman wrote:
> > >
> > > On Sun, Jul 27, 2025 at 06:06:55P
Hi Simon,
On Sun, Jul 27, 2025 at 9:55 PM Simon Horman wrote:
>
> On Sun, Jul 27, 2025 at 06:06:55PM +0800, Jason Xing wrote:
> > Hi Paul,
> >
> > On Sun, Jul 27, 2025 at 4:36 PM Paul Menzel wrote:
> > >
> > > Dear Jason,
> > >
> > &
Hi Paul,
On Sun, Jul 27, 2025 at 4:36 PM Paul Menzel wrote:
>
> Dear Jason,
>
>
> Thank you for the improved version.
>
> Am 26.07.25 um 09:03 schrieb Jason Xing:
> > From: Jason Xing
> >
> > Resolve the budget negative overflow which leads to returning
On Fri, Jul 25, 2025 at 6:58 PM Larysa Zaremba wrote:
>
> On Fri, Jul 25, 2025 at 07:18:11AM +0800, Jason Xing wrote:
> > Hi Tony,
> >
> > On Fri, Jul 25, 2025 at 4:21 AM Tony Nguyen
> > wrote:
> > >
> > >
> > >
> > >
On Fri, Jul 25, 2025 at 6:52 PM Larysa Zaremba wrote:
>
> On Sun, Jul 20, 2025 at 05:11:22PM +0800, Jason Xing wrote:
> > From: Jason Xing
> >
> > Like what i40e driver initially did in commit 3106c580fb7cf
> > ("i40e: Use batched xsk Tx interfaces to increase
On Fri, Jul 25, 2025 at 5:45 PM Larysa Zaremba wrote:
>
> On Sun, Jul 20, 2025 at 05:11:21PM +0800, Jason Xing wrote:
> > From: Jason Xing
> >
> > - Adjust ixgbe_desc_unused as the budget value.
> > - Avoid checking desc_unused over and over again in the loop.
On Sat, Jul 26, 2025 at 12:54 AM Tony Nguyen wrote:
>
>
>
> On 7/25/2025 3:57 AM, Larysa Zaremba wrote:
> > On Fri, Jul 25, 2025 at 07:18:11AM +0800, Jason Xing wrote:
> >> Hi Tony,
> >>
> >> On Fri, Jul 25, 2025 at 4:21 AM Tony Nguyen
> >&g
40 matches
Mail list logo
