patches for DPDK stable branches
 help / color / mirror / Atom feed
From: Kevin Traynor <ktraynor@redhat.com>
To: Jie Hai <haijie1@huawei.com>
Cc: Ferruh Yigit <ferruh.yigit@amd.com>, dpdk stable <stable@dpdk.org>
Subject: patch 'app/testpmd: fix primary process not polling all queues' has been queued to stable release 21.11.6
Date: Thu,  2 Nov 2023 13:12:47 +0000	[thread overview]
Message-ID: <20231102131254.315947-101-ktraynor@redhat.com> (raw)
In-Reply-To: <20231102131254.315947-1-ktraynor@redhat.com>

Hi,

FYI, your patch has been queued to stable release 21.11.6

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 11/08/23. So please
shout if anyone has objections.

Also note that after the patch there's a diff of the upstream commit vs the
patch applied to the branch. This will indicate if there was any rebasing
needed to apply to the stable branch. If there were code changes for rebasing
(ie: not only metadata diffs), please double check that the rebase was
correctly done.

Queued patches are on a temporary branch at:
https://github.com/kevintraynor/dpdk-stable

This queued commit can be viewed at:
https://github.com/kevintraynor/dpdk-stable/commit/097657f9bc334a0fc69215e33bf25e5277d41e22

Thanks.

Kevin

---
From 097657f9bc334a0fc69215e33bf25e5277d41e22 Mon Sep 17 00:00:00 2001
From: Jie Hai <haijie1@huawei.com>
Date: Fri, 8 Sep 2023 19:29:01 +0800
Subject: [PATCH] app/testpmd: fix primary process not polling all queues

[ upstream commit 37b68fa1445cb6fe287178607bac6fa3d90b835f ]

Here's how the problem arises.
step1: Start the app.
    dpdk-testpmd -a 0000:35:00.0 -l 0-3 -- -i --rxq=10 --txq=10

step2: Perform the following steps and send traffic. As expected,
queue 7 does not send or receive packets, and other queues do.
    port 0 rxq 7 stop
    port 0 txq 7 stop
    set fwd mac
    start

step3: Perform the following steps and send traffic. All queues
are expected to send and receive packets normally, but that's not
the case for queue 7.
    stop
    port stop all
    port start all
    start
    show port xstats all

In fact, only the value of rx_q7_packets for queue 7 is not zero,
which means queue 7 is enabled for the driver but is not involved
in packet receiving and forwarding by software. If we check queue
state by command 'show rxq info 0 7' and 'show txq info 0 7',
we see queue 7 is started as other queues are.
    Rx queue state: started
    Tx queue state: started
The queue 7 is started but cannot forward. That's the problem.

We know that each stream has a read-only "disabled" field that
control if this stream should be used to forward. This field
depends on testpmd local queue state, please see
commit 3c4426db54fc ("app/testpmd: do not poll stopped queues").
DPDK framework maintains ethdev queue state that drivers reported,
which indicates the real state of queues.

There are commands that update these two kind queue state such as
'port X rxq|txq start|stop'. But these operations take effect only
in one stop-start round. In the following stop-start round, the
preceding operations do not take effect anymore. However, only
the ethdev queue state is updated, causing the testpmd and ethdev
state information to diverge and causing unexpected side effects
as above problem.

There was a similar problem for the secondary process, please see
commit 5028f207a4fa ("app/testpmd: fix secondary process packet
forwarding").

This patch applies its workaround with some difference to the
primary process. Not all PMDs implement rte_eth_rx_queue_info_get and
rte_eth_tx_queue_info_get, however they may support deferred_start
with primary process. To not break their behavior, retain the original
testpmd local queue state for those PMDs.

Fixes: 3c4426db54fc ("app/testpmd: do not poll stopped queues")

Signed-off-by: Jie Hai <haijie1@huawei.com>
Acked-by: Ferruh Yigit <ferruh.yigit@amd.com>
---
 app/test-pmd/testpmd.c | 25 ++++++++++++++++++++-----
 1 file changed, 20 insertions(+), 5 deletions(-)

diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
index 9f4a0e0a78..5f34641e90 100644
--- a/app/test-pmd/testpmd.c
+++ b/app/test-pmd/testpmd.c
@@ -2376,4 +2376,11 @@ update_rx_queue_state(uint16_t port_id, uint16_t queue_id)
 			rx_qinfo.queue_state;
 	} else if (rc == -ENOTSUP) {
+		/*
+		 * Do not change the rxq state for primary process
+		 * to ensure that the PMDs do not implement
+		 * rte_eth_rx_queue_info_get can forward as before.
+		 */
+		if (rte_eal_process_type() == RTE_PROC_PRIMARY)
+			return;
 		/*
 		 * Set the rxq state to RTE_ETH_QUEUE_STATE_STARTED
@@ -2401,4 +2408,11 @@ update_tx_queue_state(uint16_t port_id, uint16_t queue_id)
 			tx_qinfo.queue_state;
 	} else if (rc == -ENOTSUP) {
+		/*
+		 * Do not change the txq state for primary process
+		 * to ensure that the PMDs do not implement
+		 * rte_eth_tx_queue_info_get can forward as before.
+		 */
+		if (rte_eal_process_type() == RTE_PROC_PRIMARY)
+			return;
 		/*
 		 * Set the txq state to RTE_ETH_QUEUE_STATE_STARTED
@@ -2415,5 +2429,5 @@ update_tx_queue_state(uint16_t port_id, uint16_t queue_id)
 
 static void
-update_queue_state(void)
+update_queue_state(portid_t pid)
 {
 	portid_t pi;
@@ -2421,4 +2435,7 @@ update_queue_state(void)
 
 	RTE_ETH_FOREACH_DEV(pi) {
+		if (pid != pi && pid != (portid_t)RTE_PORT_ALL)
+			continue;
+
 		for (qi = 0; qi < nb_rxq; qi++)
 			update_rx_queue_state(pi, qi);
@@ -2468,6 +2485,5 @@ start_packet_forwarding(int with_tx_first)
 
 	if (stream_init != NULL) {
-		if (rte_eal_process_type() == RTE_PROC_SECONDARY)
-			update_queue_state();
+		update_queue_state(RTE_PORT_ALL);
 		for (i = 0; i < cur_fwd_config.nb_fwd_streams; i++)
 			stream_init(fwd_streams[i]);
@@ -3127,6 +3143,5 @@ start_port(portid_t pid)
 	}
 
-	if (rte_eal_process_type() == RTE_PROC_SECONDARY)
-		update_queue_state();
+	update_queue_state(pi);
 
 	if (at_least_one_port_successfully_started && !no_link_check)
-- 
2.41.0

---
  Diff of the applied patch vs upstream commit (please double-check if non-empty:
---
--- -	2023-11-02 13:09:43.209083868 +0000
+++ 0101-app-testpmd-fix-primary-process-not-polling-all-queu.patch	2023-11-02 13:09:40.923163688 +0000
@@ -1 +1 @@
-From 37b68fa1445cb6fe287178607bac6fa3d90b835f Mon Sep 17 00:00:00 2001
+From 097657f9bc334a0fc69215e33bf25e5277d41e22 Mon Sep 17 00:00:00 2001
@@ -5,0 +6,2 @@
+[ upstream commit 37b68fa1445cb6fe287178607bac6fa3d90b835f ]
+
@@ -61 +62,0 @@
-Cc: stable@dpdk.org
@@ -70 +71 @@
-index b3d6ce7127..595b77748c 100644
+index 9f4a0e0a78..5f34641e90 100644
@@ -73 +74 @@
-@@ -2426,4 +2426,11 @@ update_rx_queue_state(uint16_t port_id, uint16_t queue_id)
+@@ -2376,4 +2376,11 @@ update_rx_queue_state(uint16_t port_id, uint16_t queue_id)
@@ -85 +86 @@
-@@ -2451,4 +2458,11 @@ update_tx_queue_state(uint16_t port_id, uint16_t queue_id)
+@@ -2401,4 +2408,11 @@ update_tx_queue_state(uint16_t port_id, uint16_t queue_id)
@@ -97 +98 @@
-@@ -2465,5 +2479,5 @@ update_tx_queue_state(uint16_t port_id, uint16_t queue_id)
+@@ -2415,5 +2429,5 @@ update_tx_queue_state(uint16_t port_id, uint16_t queue_id)
@@ -104 +105 @@
-@@ -2471,4 +2485,7 @@ update_queue_state(void)
+@@ -2421,4 +2435,7 @@ update_queue_state(void)
@@ -112 +113 @@
-@@ -2518,6 +2535,5 @@ start_packet_forwarding(int with_tx_first)
+@@ -2468,6 +2485,5 @@ start_packet_forwarding(int with_tx_first)
@@ -120 +121 @@
-@@ -3282,6 +3298,5 @@ start_port(portid_t pid)
+@@ -3127,6 +3143,5 @@ start_port(portid_t pid)


  parent reply	other threads:[~2023-11-02 13:15 UTC|newest]

Thread overview: 104+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-11-02 13:11 patch 'mempool: fix default ops for an empty mempool' " Kevin Traynor
2023-11-02 13:11 ` patch 'eal/unix: fix firmware reading with external xz helper' " Kevin Traynor
2023-11-02 13:11 ` patch 'rawdev: fix device class in log message' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/bonding: fix header for C++' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/sfc: set max Rx packet length for representors' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/sfc: account for data offset on Tx' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/sfc: add missing error code indication to MAE init path' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/hns3: fix VF default MAC modified when set failed' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/hns3: fix error code for multicast resource' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/hns3: fix flushing multicast MAC address' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/hns3: fix order in NEON Rx' " Kevin Traynor
2023-11-02 13:11 ` patch 'random: initialize state for unregistered non-EAL threads' " Kevin Traynor
2023-11-02 13:11 ` patch 'bus/dpaa: fix build with asserts for GCC 13' " Kevin Traynor
2023-11-02 13:11 ` patch 'hash: align SSE lookup to scalar implementation' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/netvsc: increase VSP response timeout to 60 seconds' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/txgbe: add Tx queue maximum limit' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/txgbe: fix GRE tunnel packet checksum' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/ngbe: fix flow control' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/txgbe: reconfigure MAC Rx when link update' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/ngbe: " Kevin Traynor
2023-11-02 13:11 ` patch 'net/txgbe: keep link down after device close' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/ngbe: " Kevin Traynor
2023-11-02 13:11 ` patch 'net/txgbe: check process type in close operation' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/ngbe: " Kevin Traynor
2023-11-02 13:11 ` patch 'net/tap: use MAC address parse API instead of local parser' " Kevin Traynor
2023-11-02 13:11 ` patch 'app/testpmd: fix help string' " Kevin Traynor
2023-11-02 13:11 ` patch 'cryptodev: add missing doc for security context' " Kevin Traynor
2023-11-02 13:11 ` patch 'doc: replace code blocks with includes in security guide' " Kevin Traynor
2023-11-02 13:11 ` patch 'test/crypto: fix IV in some vectors' " Kevin Traynor
2023-11-02 13:11 ` patch 'test/crypto: skip some synchronous tests with CPU crypto' " Kevin Traynor
2023-11-02 13:11 ` patch 'test/crypto: fix typo in asym tests' " Kevin Traynor
2023-11-02 13:11 ` patch 'crypto/ipsec_mb: add dependency check for cross build' " Kevin Traynor
2023-11-02 13:11 ` patch 'eventdev: fix symbol export for port maintenance' " Kevin Traynor
2023-11-02 13:11 ` patch 'bus/pci: fix device ID log' " Kevin Traynor
2023-11-02 13:11 ` patch 'event/cnxk: fix getwork mode devargs parsing' " Kevin Traynor
2023-11-02 13:11 ` patch 'event/cnxk: fix return values for capability API' " Kevin Traynor
2023-11-02 13:11 ` patch 'test/event: fix crypto null device creation' " Kevin Traynor
2023-11-02 13:11 ` patch 'event/sw: remove obsolete comment' " Kevin Traynor
2023-11-02 13:11 ` patch 'eventdev/eth_rx: fix timestamp field register in mbuf' " Kevin Traynor
2023-11-02 13:11 ` patch 'event/sw: fix ordering corruption with op release' " Kevin Traynor
2023-11-02 13:11 ` patch 'common/cnxk: fix default flow action setting' " Kevin Traynor
2023-11-02 13:11 ` patch 'common/cnxk: fix xstats for different packet sizes' " Kevin Traynor
2023-11-02 13:11 ` patch 'common/cnxk: fix different size bit operations' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/cnxk: fix uninitialized variable' " Kevin Traynor
2023-11-02 13:11 ` Kevin Traynor
2023-11-02 13:11 ` patch 'common/cnxk: fix DPI memzone name' " Kevin Traynor
2023-11-02 13:11 ` patch 'dma/cnxk: fix device state' " Kevin Traynor
2023-11-02 13:11 ` patch 'mempool/cnxk: fix free from non-EAL threads' " Kevin Traynor
2023-11-02 13:11 ` patch 'common/cnxk: fix aura disable handling' " Kevin Traynor
2023-11-02 13:11 ` patch 'common/cnxk: remove dead Meson code' " Kevin Traynor
2023-11-02 13:11 ` patch 'baseband/acc: fix ACC100 HARQ input alignment' " Kevin Traynor
2023-11-02 13:11 ` patch 'app/bbdev: fix link with NXP LA12XX' " Kevin Traynor
2023-11-02 13:11 ` patch 'net/i40e: fix FDIR queue receives broadcast packets' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/ice: write timestamp to first segment in scattered Rx' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/iavf: fix VLAN offload strip flag' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/iavf: fix checksum offloading' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/i40e: fix buffer leak on Rx reconfiguration' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/ice: fix TM configuration clearing' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/iavf: fix port stats " Kevin Traynor
2023-11-02 13:12 ` patch 'net/iavf: unregister interrupt handler before FD close' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/iavf: fix ESN session update' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/ice: fix initial link status' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/iavf: fix Tx debug' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/iavf: remove log from Tx prepare function' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/iavf: fix TSO with big segments' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/ice: remove log from Tx prepare function' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/ice: fix TSO with big segments' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/mlx5: fix leak in sysfs port name translation' " Kevin Traynor
2023-11-02 13:12 ` patch 'fib: fix adding default route overwriting entire table' " Kevin Traynor
2023-11-02 13:12 ` patch 'fib6: fix adding default route as first route' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/af_packet: fix Rx and Tx queue state' " Kevin Traynor
2023-11-02 13:12 ` patch 'net/af_xdp: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/avp: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/bnx2x: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/bnxt: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/bonding: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/cxgbe: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/dpaa: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/dpaa2: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/e1000: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/ena: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/enetc: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/enic: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/hinic: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/ipn3ke: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/memif: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/mlx4: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/mvneta: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/mvpp2: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/ngbe: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/null: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/octeon_ep: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/pfe: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/ring: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/sfc: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/softnic: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/txgbe: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/vhost: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/virtio: " Kevin Traynor
2023-11-02 13:12 ` patch 'net/vmxnet3: " Kevin Traynor
2023-11-02 13:12 ` Kevin Traynor [this message]
2023-11-02 13:12 ` patch 'net/bonding: fix link status callback stop' " Kevin Traynor
2023-11-02 13:12 ` patch 'app/procinfo: remove unnecessary rte_malloc' " Kevin Traynor
2023-11-02 13:12 ` patch 'malloc: remove return from void functions' " Kevin Traynor

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20231102131254.315947-101-ktraynor@redhat.com \
    --to=ktraynor@redhat.com \
    --cc=ferruh.yigit@amd.com \
    --cc=haijie1@huawei.com \
    --cc=stable@dpdk.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).