From: Kevin Traynor <ktraynor@redhat.com>
To: Chengwen Feng <fengchengwen@huawei.com>
Cc: dpdk stable <stable@dpdk.org>
Subject: patch 'dma/hisilicon: fix stop with pending transfers' has been queued to stable release 24.11.4
Date: Fri, 31 Oct 2025 14:33:09 +0000 [thread overview]
Message-ID: <20251031143421.324432-67-ktraynor@redhat.com> (raw)
In-Reply-To: <20251031143421.324432-1-ktraynor@redhat.com>
Hi,
FYI, your patch has been queued to stable release 24.11.4
Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 11/05/25. 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/caf6bf8424e73315be792d4759a593509f1d50e6
Thanks.
Kevin
---
From caf6bf8424e73315be792d4759a593509f1d50e6 Mon Sep 17 00:00:00 2001
From: Chengwen Feng <fengchengwen@huawei.com>
Date: Mon, 13 Oct 2025 17:22:11 +0800
Subject: [PATCH] dma/hisilicon: fix stop with pending transfers
[ upstream commit 8aa458f1c44b9f6e73f75047aca77191dc79746f ]
Stop dmadev may fail if there are pending DMA transfers, we need make
sure there are no pending DMA transfers when stop.
This commit uses following scheme:
1. flag stop proc so that new request will not process.
2. setting drop flag for all descriptor to quick complete.
3. waiting dmadev to complete.
Fixes: 3c5f5f03a047 ("dma/hisilicon: add control path")
Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
---
drivers/dma/hisilicon/hisi_dmadev.c | 45 ++++++++++++++++++++++++-----
drivers/dma/hisilicon/hisi_dmadev.h | 2 ++
2 files changed, 39 insertions(+), 8 deletions(-)
diff --git a/drivers/dma/hisilicon/hisi_dmadev.c b/drivers/dma/hisilicon/hisi_dmadev.c
index fdd2fa225f..feac145e29 100644
--- a/drivers/dma/hisilicon/hisi_dmadev.c
+++ b/drivers/dma/hisilicon/hisi_dmadev.c
@@ -377,4 +377,5 @@ hisi_dma_start(struct rte_dma_dev *dev)
hw->cqs_completed = 0;
hw->cqe_vld = 1;
+ hw->stop_proc = 0;
hw->submitted = 0;
hw->completed = 0;
@@ -388,10 +389,4 @@ hisi_dma_start(struct rte_dma_dev *dev)
}
-static int
-hisi_dma_stop(struct rte_dma_dev *dev)
-{
- return hisi_dma_reset_hw(dev->data->dev_private);
-}
-
static int
hisi_dma_close(struct rte_dma_dev *dev)
@@ -455,4 +450,35 @@ hisi_dma_vchan_status(const struct rte_dma_dev *dev, uint16_t vchan,
}
+static int
+hisi_dma_stop(struct rte_dma_dev *dev)
+{
+#define MAX_WAIT_MSEC 10
+ struct hisi_dma_dev *hw = dev->data->dev_private;
+ enum rte_dma_vchan_status status;
+ uint32_t i;
+
+ /* Flag stop processing new requests. */
+ hw->stop_proc = 1;
+ rte_delay_ms(1);
+
+ /* Force set drop flag so that the hardware can quickly complete. */
+ for (i = 0; i <= hw->sq_depth_mask; i++)
+ hw->sqe[i].dw0 |= SQE_DROP_FLAG;
+
+ i = 0;
+ do {
+ hisi_dma_vchan_status(dev, 0, &status);
+ if (status != RTE_DMA_VCHAN_ACTIVE)
+ break;
+ rte_delay_ms(1);
+ } while (i++ < MAX_WAIT_MSEC);
+ if (status == RTE_DMA_VCHAN_ACTIVE) {
+ HISI_DMA_ERR(hw, "dev is still active!");
+ return -EBUSY;
+ }
+
+ return hisi_dma_reset_hw(dev->data->dev_private);
+}
+
static void
hisi_dma_dump_range(struct hisi_dma_dev *hw, FILE *f, uint32_t start,
@@ -549,5 +575,5 @@ hisi_dma_dump(const struct rte_dma_dev *dev, FILE *f)
" ridx: %u cridx: %u\n"
" sq_head: %u sq_tail: %u cq_sq_head: %u\n"
- " cq_head: %u cqs_completed: %u cqe_vld: %u\n"
+ " cq_head: %u cqs_completed: %u cqe_vld: %u stop_proc: %u\n"
" submitted: %" PRIu64 " completed: %" PRIu64 " errors: %"
PRIu64 " qfulls: %" PRIu64 "\n",
@@ -556,5 +582,5 @@ hisi_dma_dump(const struct rte_dma_dev *dev, FILE *f)
hw->ridx, hw->cridx,
hw->sq_head, hw->sq_tail, hw->cq_sq_head,
- hw->cq_head, hw->cqs_completed, hw->cqe_vld,
+ hw->cq_head, hw->cqs_completed, hw->cqe_vld, hw->stop_proc,
hw->submitted, hw->completed, hw->errors, hw->qfulls);
hisi_dma_dump_queue(hw, f);
@@ -574,4 +600,7 @@ hisi_dma_copy(void *dev_private, uint16_t vchan,
RTE_SET_USED(vchan);
+ if (unlikely(hw->stop_proc > 0))
+ return -EPERM;
+
if (((hw->sq_tail + 1) & hw->sq_depth_mask) == hw->sq_head) {
hw->qfulls++;
diff --git a/drivers/dma/hisilicon/hisi_dmadev.h b/drivers/dma/hisilicon/hisi_dmadev.h
index 786fe3cc0e..5e3c8debcc 100644
--- a/drivers/dma/hisilicon/hisi_dmadev.h
+++ b/drivers/dma/hisilicon/hisi_dmadev.h
@@ -143,4 +143,5 @@ enum {
struct hisi_dma_sqe {
uint32_t dw0;
+#define SQE_DROP_FLAG BIT(4)
#define SQE_FENCE_FLAG BIT(10)
#define SQE_OPCODE_M2M 0x4
@@ -213,4 +214,5 @@ struct hisi_dma_dev {
uint16_t cqs_completed;
uint8_t cqe_vld; /**< valid bit for CQE, will change for every round. */
+ volatile uint8_t stop_proc; /**< whether stop processing new requests. */
uint64_t submitted;
--
2.51.0
---
Diff of the applied patch vs upstream commit (please double-check if non-empty:
---
--- - 2025-10-31 13:53:54.301364499 +0000
+++ 0067-dma-hisilicon-fix-stop-with-pending-transfers.patch 2025-10-31 13:53:52.172444764 +0000
@@ -1 +1 @@
-From 8aa458f1c44b9f6e73f75047aca77191dc79746f Mon Sep 17 00:00:00 2001
+From caf6bf8424e73315be792d4759a593509f1d50e6 Mon Sep 17 00:00:00 2001
@@ -5,0 +6,2 @@
+[ upstream commit 8aa458f1c44b9f6e73f75047aca77191dc79746f ]
+
@@ -15 +16,0 @@
-Cc: stable@dpdk.org
@@ -24 +25 @@
-index 019c4a8189..7575fb12d9 100644
+index fdd2fa225f..feac145e29 100644
@@ -103 +104 @@
-index 90301e6b00..aab87c40be 100644
+index 786fe3cc0e..5e3c8debcc 100644
@@ -106 +107 @@
-@@ -142,4 +142,5 @@ enum {
+@@ -143,4 +143,5 @@ enum {
@@ -112 +113 @@
-@@ -212,4 +213,5 @@ struct hisi_dma_dev {
+@@ -213,4 +214,5 @@ struct hisi_dma_dev {
next prev parent reply other threads:[~2025-10-31 14:37 UTC|newest]
Thread overview: 135+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-10-31 14:32 patch 'test/telemetry: fix test calling all commands' " Kevin Traynor
2025-10-31 14:32 ` patch 'eal: fix plugin dir walk' " Kevin Traynor
2025-10-31 14:32 ` patch 'cmdline: fix port list parsing' " Kevin Traynor
2025-10-31 14:32 ` patch 'tailq: fix lookup macro' " Kevin Traynor
2025-10-31 14:32 ` patch 'hash: fix unaligned access in predictable RSS' " Kevin Traynor
2025-10-31 14:32 ` patch 'graph: fix stats query with no node xstats' " Kevin Traynor
2025-10-31 14:32 ` patch 'graph: fix unaligned access in stats' " Kevin Traynor
2025-10-31 14:32 ` patch 'eventdev: fix listing timer adapters with telemetry' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/gve: send whole packet when mbuf is large' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/gve: clean when insufficient Tx descriptors' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/gve: do not write zero-length " Kevin Traynor
2025-10-31 14:32 ` patch 'net/gve: validate Tx packet before sending' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/gve: add DQO Tx descriptor limit' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/gve: fix DQO TSO " Kevin Traynor
2025-10-31 14:32 ` patch 'net/gve: clear DQO Tx descriptors before writing' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/vmxnet3: fix mapping of mempools to queues' " Kevin Traynor
2025-10-31 14:32 ` patch 'app/testpmd: increase size of set cores list command' " Kevin Traynor
2025-10-31 14:32 ` patch 'bus/dpaa: improve cleanup' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/gve: free device resources on close' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/dpaa2: fix extract buffer preparation' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/dpaa2: fix shaper rate' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/tap: fix BPF with cross-compilation' " Kevin Traynor
2025-10-31 14:32 ` patch 'app/testpmd: fix mask in flow random item' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/zxdh: fix Arm build' " Kevin Traynor
2025-10-31 14:32 ` patch 'app/testpmd: monitor state of primary process' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/gve: fix disabling interrupts on DQ' " Kevin Traynor
2025-10-31 14:32 ` patch 'app/testpmd: fix conntrack action query' " Kevin Traynor
2025-10-31 14:32 ` patch 'doc: add conntrack state inspect command to testpmd guide' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/gve: free Rx mbufs if allocation fails on ring setup' " Kevin Traynor
2025-10-31 14:32 ` patch 'app/testpmd: validate DSCP and VLAN for meter creation' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix index-based flow rules' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix default flow rules start' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix storage of shared Rx queues' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix use after scope of RSS configuration' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix unsupported flow rule port action' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix double free in non-template flow destroy' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix non-template age rules flush' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix non-template RSS expansion' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix connection tracking state item validation' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5/hws: fix TIR action support in FDB' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix indirect flow age action handling' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix Direct Verbs counter offset detection' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/mlx5: fix interface name parameter definition' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/iavf: fix Tx vector path selection logic' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/ixgbe: fix SCTP port filtering on E610' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/intel: fix assumption about tag placement order' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/ice: fix VLAN tag reporting on Rx' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/ice/base: fix adding special words' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/ice/base: fix memory leak in HW profile handling' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/ice/base: fix memory leak in recipe " Kevin Traynor
2025-10-31 14:32 ` patch 'gro: fix payload corruption in coalescing packets' " Kevin Traynor
2025-10-31 14:32 ` patch 'eal: fix DMA mask validation with IOVA mode option' " Kevin Traynor
2025-10-31 14:32 ` patch 'eal: fix MP socket cleanup' " Kevin Traynor
2025-10-31 14:32 ` patch 'crypto/ipsec_mb: fix QP release in secondary' " Kevin Traynor
2025-10-31 14:32 ` patch 'efd: fix AVX2 support' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/octeon_ep: handle interrupt enable failure' " Kevin Traynor
2025-10-31 14:32 ` patch 'net/octeon_ep: fix mbuf data offset update' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/octeon_ep: fix device start' " Kevin Traynor
2025-10-31 14:33 ` patch 'common/cnxk: fix async event handling' " Kevin Traynor
2025-10-31 14:33 ` patch 'doc: fix feature list of ice driver' " Kevin Traynor
2025-10-31 14:33 ` patch 'doc: fix feature list of iavf " Kevin Traynor
2025-10-31 14:33 ` patch 'eal/arm: fix C++ build for 32-bit memcpy' " Kevin Traynor
2025-10-31 14:33 ` patch 'test/debug: fix crash with mlx5 devices' " Kevin Traynor
2025-10-31 14:33 ` patch 'bus/pci: fix build with MinGW 13' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/mlx5: " Kevin Traynor
2025-10-31 14:33 ` patch 'bbdev: " Kevin Traynor
2025-10-31 14:33 ` Kevin Traynor [this message]
2025-10-31 14:33 ` patch 'test/dma: fix failure condition' " Kevin Traynor
2025-10-31 14:33 ` patch 'eal/x86: enable timeout in AMD power monitor' " Kevin Traynor
2025-11-01 15:31 ` Tummala, Sivaprasad
2025-10-31 14:33 ` patch 'test/func_reentrancy: fix args to EAL init call' " Kevin Traynor
2025-10-31 14:33 ` patch 'fib6: fix memory leak on delete operation' " Kevin Traynor
2025-10-31 14:33 ` patch 'fib6: fix tbl8 allocation check logic' " Kevin Traynor
2025-10-31 14:33 ` patch 'vhost: add VDUSE virtqueue ready state polling workaround' " Kevin Traynor
2025-10-31 14:33 ` patch 'vhost: fix virtqueue info init in VDUSE vring setup' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/virtio-user: fix used ring address calculation' " Kevin Traynor
2025-10-31 14:33 ` patch 'vhost: fix double fetch when dequeue offloading' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/bnxt: fix free of not allocated object' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/ice/base: fix integer overflow on NVM init' " Kevin Traynor
2025-10-31 14:33 ` patch 'doc: fix display of commands in cpfl guide' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/ice: fix initialization with 8 ports' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/ice: remove indirection for FDIR filters' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/ice: fix memory leak in raw pattern parse' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/i40e: fix symmetric Toeplitz hashing for SCTP' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/mlx5: fix ESP item validation to match on seqnum' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/mlx5: fix ESP header match after UDP for group 0' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/mlx5: fix multicast' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/mlx5: fix flow encapsulation hash' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/mlx5: fix indirect flow action memory leak' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/mlx5: fix MTU initialization' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/mlx5: fix leak of flow indexed pools' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/mlx5: fix flow aging race condition' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/hns3: fix inconsistent lock' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/hns3: fix VLAN resources freeing' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/hns3: fix overwrite mbuf in vector path' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/af_packet: fix crash in secondary process' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/ark: remove double mbuf free' " Kevin Traynor
2025-10-31 14:33 ` patch 'app/testpmd: stop forwarding in secondary process' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/tap: fix build with LTO' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/hns3: fix VLAN tag loss for short tunnel frame' " Kevin Traynor
2025-10-31 14:33 ` patch 'ethdev: fix VLAN filter parameter description' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/enetfec: fix file descriptor leak on read error' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/enetfec: fix out-of-bounds access in UIO mapping' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/enetfec: fix buffer descriptor size configuration' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/enetfec: fix Tx queue free' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/enetfec: fix checksum flag handling and error return' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/enetfec: reject multi-queue configuration' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/enetfec: fix memory leak in Rx buffer cleanup' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/enetfec: reject Tx deferred queue' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/tap: fix interrupt callback crash after failed start' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/ntnic: fix potential format overflow' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/ena: fix PCI BAR mapping on 64K page size' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/ena/base: fix unsafe memcpy on invalid memory' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/dpaa2: fix uninitialized variable' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/dpaa2: free buffers from error queue' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/dpaa2: fix L3/L4 checksum results' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/dpaa2: receive packets with additional parse errors' " Kevin Traynor
2025-10-31 14:33 ` patch 'net/dpaa2: fix error frame dump' " Kevin Traynor
2025-10-31 14:34 ` patch 'net/dpaa2: fix flow rule resizing' " Kevin Traynor
2025-10-31 14:34 ` patch 'crypto/qat: fix ECDH' " Kevin Traynor
2025-10-31 14:34 ` patch 'crypto/qat: fix source buffer alignment' " Kevin Traynor
2025-10-31 14:34 ` patch 'crypto/cnxk: refactor RSA verification' " Kevin Traynor
2025-10-31 14:34 ` patch 'test/crypto: fix mbuf handling' " Kevin Traynor
2025-10-31 14:34 ` patch 'app/crypto-perf: fix plaintext size exceeds buffer size' " Kevin Traynor
2025-10-31 14:34 ` patch 'test/crypto: fix vector initialization' " Kevin Traynor
2025-10-31 14:34 ` patch 'crypto/virtio: fix cookies leak' " Kevin Traynor
2025-10-31 14:34 ` patch 'bitops: improve power of 2 alignment documentation' " Kevin Traynor
2025-10-31 14:34 ` patch 'sched: fix WRR parameter data type' " Kevin Traynor
2025-10-31 14:34 ` patch 'test/argparse: change initialization to workaround LTO' " Kevin Traynor
2025-10-31 14:34 ` patch 'config/arm: enable NUMA for Neoverse N2' " Kevin Traynor
2025-10-31 14:34 ` patch 'dts: fix docstring in checksum suite' " Kevin Traynor
2025-10-31 14:34 ` patch 'bus/pci: fix resource leak in secondary process' " Kevin Traynor
2025-10-31 14:34 ` patch 'net/ice: fix vector Rx VLAN offload flags' " Kevin Traynor
2025-10-31 14:34 ` patch 'net/ice: remove unsupported SCTP Rx offload' " Kevin Traynor
2025-10-31 14:34 ` patch 'cmdline: fix highest bit port list parsing' " 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=20251031143421.324432-67-ktraynor@redhat.com \
--to=ktraynor@redhat.com \
--cc=fengchengwen@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).