[PATCH net 0/2][pull request] Intel Wired LAN Driver Updates 2022-03-02

[PATCH net 0/2][pull request] Intel Wired LAN Driver Updates 2022-03-02

[Tony Nguyen]

This series contains updates to ixgbe and ice drivers.

Maciej fixes an issue that occurs when carrier is not ok and xsk_pool
is present that will cause ksoftirqd to consume 100% CPU by changing
the value returned when the carrier state is not ok for ixgbe. He also
removes checks against ice_ring_is_xdp() that can't occur for ice.

The following are changes since commit 4761df52f1549cc8c5ffcad0b2095fffe2c5435d:
  Merge git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
and are available in the git repository at:
  git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue 10GbE

Maciej Fijalkowski (2):
  ixgbe: xsk: change !netif_carrier_ok() handling in ixgbe_xmit_zc()
  ice: avoid XDP checks in ice_clean_tx_irq()

 drivers/net/ethernet/intel/ice/ice_txrx.c    | 7 +------
 drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c | 6 ++++--
 2 files changed, 5 insertions(+), 8 deletions(-)

-- 
2.31.1

[PATCH net 1/2] ixgbe: xsk: change !netif_carrier_ok() handling in ixgbe_xmit_zc()

[Tony Nguyen]

From: Maciej Fijalkowski <maciej.fijalkowski@intel.com>

Commit c685c69fba71 ("ixgbe: don't do any AF_XDP zero-copy transmit if
netif is not OK") addressed the ring transient state when
MEM_TYPE_XSK_BUFF_POOL was being configured which in turn caused the
interface to through down/up. Maurice reported that when carrier is not
ok and xsk_pool is present on ring pair, ksoftirqd will consume 100% CPU
cycles due to the constant NAPI rescheduling as ixgbe_poll() states that
there is still some work to be done.

To fix this, do not set work_done to false for a !netif_carrier_ok().

Fixes: c685c69fba71 ("ixgbe: don't do any AF_XDP zero-copy transmit if netif is not OK")
Reported-by: Maurice Baijens <maurice.baijens@ellips.com>
Tested-by: Maurice Baijens <maurice.baijens@ellips.com>
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Sandeep Penigalapati <sandeep.penigalapati@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
 drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c
index b3fd8e5cd85b..6a5e9cf6b5da 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c
@@ -390,12 +390,14 @@ static bool ixgbe_xmit_zc(struct ixgbe_ring *xdp_ring, unsigned int budget)
 	u32 cmd_type;
 
 	while (budget-- > 0) {
-		if (unlikely(!ixgbe_desc_unused(xdp_ring)) ||
-		    !netif_carrier_ok(xdp_ring->netdev)) {
+		if (unlikely(!ixgbe_desc_unused(xdp_ring))) {
 			work_done = false;
 			break;
 		}
 
+		if (!netif_carrier_ok(xdp_ring->netdev))
+			break;
+
 		if (!xsk_tx_peek_desc(pool, &desc))
 			break;
 
-- 
2.31.1

[PATCH net 2/2] ice: avoid XDP checks in ice_clean_tx_irq()

[Tony Nguyen]

From: Maciej Fijalkowski <maciej.fijalkowski@intel.com>

Commit 9610bd988df9 ("ice: optimize XDP_TX workloads") introduced Tx IRQ
cleaning routine dedicated for XDP rings. Currently it is impossible to
call ice_clean_tx_irq() against XDP ring, so it is safe to drop
ice_ring_is_xdp() calls in there.

Fixes: 1c96c16858ba ("ice: update to newer kernel API")
Fixes: cc14db11c8a4 ("ice: use prefetch methods")
Reviewed-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Kiran Bhandare <kiranx.bhandare@intel.com>  (A Contingent Worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_txrx.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c
index 3e38695f1c9d..2a1a12299fbb 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.c
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
@@ -221,8 +221,7 @@ static bool ice_clean_tx_irq(struct ice_tx_ring *tx_ring, int napi_budget)
 	struct ice_tx_buf *tx_buf;
 
 	/* get the bql data ready */
-	if (!ice_ring_is_xdp(tx_ring))
-		netdev_txq_bql_complete_prefetchw(txring_txq(tx_ring));
+	netdev_txq_bql_complete_prefetchw(txring_txq(tx_ring));
 
 	tx_buf = &tx_ring->tx_buf[i];
 	tx_desc = ICE_TX_DESC(tx_ring, i);
@@ -311,10 +310,6 @@ static bool ice_clean_tx_irq(struct ice_tx_ring *tx_ring, int napi_budget)
 	tx_ring->next_to_clean = i;
 
 	ice_update_tx_ring_stats(tx_ring, total_pkts, total_bytes);
-
-	if (ice_ring_is_xdp(tx_ring))
-		return !!budget;
-
 	netdev_tx_completed_queue(txring_txq(tx_ring), total_pkts, total_bytes);
 
 #define TX_WAKE_THRESHOLD ((s16)(DESC_NEEDED * 2))
-- 
2.31.1

Re: [PATCH net 2/2] ice: avoid XDP checks in ice_clean_tx_irq()

[Jakub Kicinski]

On Wed,  2 Mar 2022 09:59:28 -0800 Tony Nguyen wrote:
> From: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
> 
> Commit 9610bd988df9 ("ice: optimize XDP_TX workloads") introduced Tx IRQ
> cleaning routine dedicated for XDP rings. Currently it is impossible to
> call ice_clean_tx_irq() against XDP ring, so it is safe to drop
> ice_ring_is_xdp() calls in there.
> 
> Fixes: 1c96c16858ba ("ice: update to newer kernel API")
> Fixes: cc14db11c8a4 ("ice: use prefetch methods")
> Reviewed-by: Alexander Lobakin <alexandr.lobakin@intel.com>
> Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
> Tested-by: Kiran Bhandare <kiranx.bhandare@intel.com>  (A Contingent Worker at Intel)
> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>

Is this really a fix?

Re: [PATCH net 2/2] ice: avoid XDP checks in ice_clean_tx_irq()

[Jakub Kicinski]

On Thu, 3 Mar 2022 08:19:01 -0800 Jakub Kicinski wrote:
> On Wed,  2 Mar 2022 09:59:28 -0800 Tony Nguyen wrote:
> > From: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
> > 
> > Commit 9610bd988df9 ("ice: optimize XDP_TX workloads") introduced Tx IRQ
> > cleaning routine dedicated for XDP rings. Currently it is impossible to
> > call ice_clean_tx_irq() against XDP ring, so it is safe to drop
> > ice_ring_is_xdp() calls in there.
> > 
> > Fixes: 1c96c16858ba ("ice: update to newer kernel API")
> > Fixes: cc14db11c8a4 ("ice: use prefetch methods")
> > Reviewed-by: Alexander Lobakin <alexandr.lobakin@intel.com>
> > Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
> > Tested-by: Kiran Bhandare <kiranx.bhandare@intel.com>  (A Contingent Worker at Intel)
> > Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>  
> 
> Is this really a fix?

I'll apply patch 1 from the list, please reroute this one to net-next.

Re: [PATCH net 2/2] ice: avoid XDP checks in ice_clean_tx_irq()

[Tony Nguyen]

On 3/3/2022 8:28 AM, Jakub Kicinski wrote:
> On Thu, 3 Mar 2022 08:19:01 -0800 Jakub Kicinski wrote:
>> On Wed,  2 Mar 2022 09:59:28 -0800 Tony Nguyen wrote:
>>> From: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
>>>
>>> Commit 9610bd988df9 ("ice: optimize XDP_TX workloads") introduced Tx IRQ
>>> cleaning routine dedicated for XDP rings. Currently it is impossible to
>>> call ice_clean_tx_irq() against XDP ring, so it is safe to drop
>>> ice_ring_is_xdp() calls in there.
>>>
>>> Fixes: 1c96c16858ba ("ice: update to newer kernel API")
>>> Fixes: cc14db11c8a4 ("ice: use prefetch methods")
>>> Reviewed-by: Alexander Lobakin <alexandr.lobakin@intel.com>
>>> Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
>>> Tested-by: Kiran Bhandare <kiranx.bhandare@intel.com>  (A Contingent Worker at Intel)
>>> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
>> Is this really a fix?
> I'll apply patch 1 from the list, please reroute this one to net-next.

Thanks Jakub, will do.

-Tony

Re: [PATCH net 0/2][pull request] Intel Wired LAN Driver Updates 2022-03-02

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net.git (master)
by Jakub Kicinski <kuba@kernel.org>:

On Wed,  2 Mar 2022 09:59:26 -0800 you wrote:
> This series contains updates to ixgbe and ice drivers.
> 
> Maciej fixes an issue that occurs when carrier is not ok and xsk_pool
> is present that will cause ksoftirqd to consume 100% CPU by changing
> the value returned when the carrier state is not ok for ixgbe. He also
> removes checks against ice_ring_is_xdp() that can't occur for ice.
> 
> [...]

Here is the summary with links:
  - [net,1/2] ixgbe: xsk: change !netif_carrier_ok() handling in ixgbe_xmit_zc()
    https://git.kernel.org/netdev/net/c/6c7273a26675
  - [net,2/2] ice: avoid XDP checks in ice_clean_tx_irq()
    (no matching commit)

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html