From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
patches@lists.linux.dev, Michal Pecio <michal.pecio@gmail.com>,
Mathias Nyman <mathias.nyman@linux.intel.com>,
Sasha Levin <sashal@kernel.org>
Subject: [PATCH 6.1 65/81] usb: xhci: Avoid queuing redundant Stop Endpoint commands
Date: Mon, 6 Jan 2025 16:16:37 +0100 [thread overview]
Message-ID: <20250106151131.887998986@linuxfoundation.org> (raw)
In-Reply-To: <20250106151129.433047073@linuxfoundation.org>
6.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Michal Pecio <michal.pecio@gmail.com>
[ Upstream commit 474538b8dd1cd9c666e56cfe8ef60fbb0fb513f4 ]
Stop Endpoint command on an already stopped endpoint fails and may be
misinterpreted as a known hardware bug by the completion handler. This
results in an unnecessary delay with repeated retries of the command.
Avoid queuing this command when endpoint state flags indicate that it's
stopped or halted and the command will fail. If commands are pending on
the endpoint, their completion handlers will process cancelled TDs so
it's done. In case of waiting for external operations like clearing TT
buffer, the endpoint is stopped and cancelled TDs can be processed now.
This eliminates practically all unnecessary retries because an endpoint
with pending URBs is maintained in Running state by the driver, unless
aforementioned commands or other operations are pending on it. This is
guaranteed by xhci_ring_ep_doorbell() and by the fact that it is called
every time any of those operations completes.
The only known exceptions are hardware bugs (the endpoint never starts
at all) and Stream Protocol errors not associated with any TRB, which
cause an endpoint reset not followed by restart. Sounds like a bug.
Generally, these retries are only expected to happen when the endpoint
fails to start for unknown/no reason, which is a worse problem itself,
and fixing the bug eliminates the retries too.
All cases were tested and found to work as expected. SET_DEQ_PENDING
was produced by patching uvcvideo to unlink URBs in 100us intervals,
which then runs into this case very often. EP_HALTED was produced by
restarting 'cat /dev/ttyUSB0' on a serial dongle with broken cable.
EP_CLEARING_TT by the same, with the dongle on an external hub.
Fixes: fd9d55d190c0 ("xhci: retry Stop Endpoint on buggy NEC controllers")
CC: stable@vger.kernel.org
Signed-off-by: Michal Pecio <michal.pecio@gmail.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20241106101459.775897-34-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/usb/host/xhci-ring.c | 13 +++++++++++++
drivers/usb/host/xhci.c | 19 +++++++++++++++----
drivers/usb/host/xhci.h | 1 +
3 files changed, 29 insertions(+), 4 deletions(-)
diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c
index e5b2a3b551e3..2503022a3123 100644
--- a/drivers/usb/host/xhci-ring.c
+++ b/drivers/usb/host/xhci-ring.c
@@ -1052,6 +1052,19 @@ static int xhci_invalidate_cancelled_tds(struct xhci_virt_ep *ep)
return 0;
}
+/*
+ * Erase queued TDs from transfer ring(s) and give back those the xHC didn't
+ * stop on. If necessary, queue commands to move the xHC off cancelled TDs it
+ * stopped on. Those will be given back later when the commands complete.
+ *
+ * Call under xhci->lock on a stopped endpoint.
+ */
+void xhci_process_cancelled_tds(struct xhci_virt_ep *ep)
+{
+ xhci_invalidate_cancelled_tds(ep);
+ xhci_giveback_invalidated_tds(ep);
+}
+
/*
* Returns the TD the endpoint ring halted on.
* Only call for non-running rings without streams.
diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c
index ae14c7ade9bc..e726c5edee03 100644
--- a/drivers/usb/host/xhci.c
+++ b/drivers/usb/host/xhci.c
@@ -1903,10 +1903,21 @@ static int xhci_urb_dequeue(struct usb_hcd *hcd, struct urb *urb, int status)
}
}
- /* Queue a stop endpoint command, but only if this is
- * the first cancellation to be handled.
- */
- if (!(ep->ep_state & EP_STOP_CMD_PENDING)) {
+ /* These completion handlers will sort out cancelled TDs for us */
+ if (ep->ep_state & (EP_STOP_CMD_PENDING | EP_HALTED | SET_DEQ_PENDING)) {
+ xhci_dbg(xhci, "Not queuing Stop Endpoint on slot %d ep %d in state 0x%x\n",
+ urb->dev->slot_id, ep_index, ep->ep_state);
+ goto done;
+ }
+
+ /* In this case no commands are pending but the endpoint is stopped */
+ if (ep->ep_state & EP_CLEARING_TT) {
+ /* and cancelled TDs can be given back right away */
+ xhci_dbg(xhci, "Invalidating TDs instantly on slot %d ep %d in state 0x%x\n",
+ urb->dev->slot_id, ep_index, ep->ep_state);
+ xhci_process_cancelled_tds(ep);
+ } else {
+ /* Otherwise, queue a new Stop Endpoint command */
command = xhci_alloc_command(xhci, false, GFP_ATOMIC);
if (!command) {
ret = -ENOMEM;
diff --git a/drivers/usb/host/xhci.h b/drivers/usb/host/xhci.h
index a75b8122538d..1a641f281c00 100644
--- a/drivers/usb/host/xhci.h
+++ b/drivers/usb/host/xhci.h
@@ -1952,6 +1952,7 @@ void xhci_ring_doorbell_for_active_rings(struct xhci_hcd *xhci,
void xhci_cleanup_command_queue(struct xhci_hcd *xhci);
void inc_deq(struct xhci_hcd *xhci, struct xhci_ring *ring);
unsigned int count_trbs(u64 addr, u64 len);
+void xhci_process_cancelled_tds(struct xhci_virt_ep *ep);
/* xHCI roothub code */
void xhci_set_link_state(struct xhci_hcd *xhci, struct xhci_port *port,
--
2.39.5
next prev parent reply other threads:[~2025-01-06 15:21 UTC|newest]
Thread overview: 93+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-01-06 15:15 [PATCH 6.1 00/81] 6.1.124-rc1 review Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 01/81] x86/hyperv: Fix hv tsc page based sched_clock for hibernation Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 02/81] selinux: ignore unknown extended permissions Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 03/81] btrfs: fix use-after-free in btrfs_encoded_read_endio() Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 04/81] tracing: Have process_string() also allow arrays Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 05/81] thunderbolt: Add support for Intel Lunar Lake Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 06/81] thunderbolt: Add support for Intel Panther Lake-M/P Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 07/81] thunderbolt: Dont display nvm_version unless upgrade supported Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 08/81] xhci: retry Stop Endpoint on buggy NEC controllers Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 09/81] usb: xhci: Limit Stop Endpoint retries Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 10/81] xhci: Turn NEC specific quirk for handling Stop Endpoint errors generic Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 11/81] net: mctp: handle skb cleanup on sock_queue failures Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 12/81] RDMA/mlx5: Enforce same type port association for multiport RoCE Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 13/81] RDMA/bnxt_re: Add check for path mtu in modify_qp Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 14/81] RDMA/bnxt_re: Fix reporting hw_ver in query_device Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 15/81] RDMA/bnxt_re: Fix max_qp_wrs reported Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 16/81] RDMA/bnxt_re: Fix the locking while accessing the QP table Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 17/81] drm/bridge: adv7511_audio: Update Audio InfoFrame properly Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 18/81] net: dsa: microchip: Fix KSZ9477 set_ageing_time function Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 19/81] net: dsa: microchip: add ksz_rmw8() function Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 20/81] net: dsa: microchip: Fix LAN937X set_ageing_time function Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 21/81] RDMA/hns: Refactor mtr find Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 22/81] RDMA/hns: Remove unused parameters and variables Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 23/81] RDMA/hns: Fix mapping error of zero-hop WQE buffer Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 24/81] RDMA/hns: Fix warning storm caused by invalid input in IO path Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 25/81] RDMA/hns: Fix missing flush CQE for DWQE Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 26/81] net: stmmac: platform: provide devm_stmmac_probe_config_dt() Greg Kroah-Hartman
2025-01-06 15:15 ` [PATCH 6.1 27/81] net: stmmac: dont create a MDIO bus if unnecessary Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 28/81] net: stmmac: restructure the error path of stmmac_probe_config_dt() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 29/81] net: fix memory leak in tcp_conn_request() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 30/81] ipip,ip_tunnel,sit: Add FOU support for externally controlled ipip devices Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 31/81] ip_tunnel: annotate data-races around t->parms.link Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 32/81] ipv4: ip_tunnel: Unmask upper DSCP bits in ip_tunnel_bind_dev() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 33/81] ipv4: ip_tunnel: Unmask upper DSCP bits in ip_md_tunnel_xmit() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 34/81] ipv4: ip_tunnel: Unmask upper DSCP bits in ip_tunnel_xmit() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 35/81] net: Fix netns for ip_tunnel_init_flow() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 36/81] netrom: check buffer length before accessing it Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 37/81] drm/i915/dg1: Fix power gate sequence Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 38/81] netfilter: nft_set_hash: unaligned atomic read on struct nft_set_ext Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 39/81] net: llc: reset skb->transport_header Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 40/81] ALSA: usb-audio: US16x08: Initialize array before use Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 41/81] eth: bcmsysport: fix call balance of priv->clk handling routines Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 42/81] net: mv643xx_eth: fix an OF node reference leak Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 43/81] net: wwan: t7xx: Fix FSM command timeout issue Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 44/81] RDMA/rtrs: Ensure ib_sge list is accessible Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 45/81] net: reenable NETIF_F_IPV6_CSUM offload for BIG TCP packets Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 46/81] net: restrict SO_REUSEPORT to inet sockets Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 47/81] net: wwan: iosm: Properly check for valid exec stage in ipc_mmio_init() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 48/81] af_packet: fix vlan_get_tci() vs MSG_PEEK Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 49/81] af_packet: fix vlan_get_protocol_dgram() " Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 50/81] ila: serialize calls to nf_register_net_hooks() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 51/81] btrfs: rename and export __btrfs_cow_block() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 52/81] btrfs: fix use-after-free when COWing tree bock and tracing is enabled Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 53/81] wifi: mac80211: wake the queues in case of failure in resume Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 54/81] drm/amdkfd: Correct the migration DMA map direction Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 55/81] btrfs: flush delalloc workers queue before stopping cleaner kthread during unmount Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 56/81] ALSA: hda/realtek: Add new alc2xx-fixup-headset-mic model Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 57/81] sound: usb: enable DSD output for ddHiFi TC44C Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 58/81] sound: usb: format: dont warn that raw DSD is unsupported Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 59/81] bpf: fix potential error return Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 60/81] ksmbd: retry iterate_dir in smb2_query_dir Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 61/81] net: usb: qmi_wwan: add Telit FE910C04 compositions Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 62/81] Bluetooth: hci_core: Fix sleeping function called from invalid context Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 63/81] irqchip/gic: Correct declaration of *percpu_base pointer in union gic_base Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 64/81] ARC: build: Try to guess GCC variant of cross compiler Greg Kroah-Hartman
2025-01-06 15:16 ` Greg Kroah-Hartman [this message]
2025-01-06 15:16 ` [PATCH 6.1 66/81] modpost: fix input MODULE_DEVICE_TABLE() built for 64-bit on 32-bit host Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 67/81] modpost: fix the missed iteration for the max bit in do_input() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 68/81] ALSA hda/realtek: Add quirk for Framework F111:000C Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 69/81] ALSA: seq: oss: Fix races at processing SysEx messages Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 70/81] kcov: mark in_softirq_really() as __always_inline Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 71/81] RDMA/uverbs: Prevent integer overflow issue Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 72/81] pinctrl: mcp23s08: Fix sleeping in atomic context due to regmap locking Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 73/81] sky2: Add device ID 11ab:4373 for Marvell 88E8075 Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 74/81] net/sctp: Prevent autoclose integer overflow in sctp_association_init() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 75/81] drm: adv7511: Drop dsi single lane support Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 76/81] dt-bindings: display: adi,adv7533: Drop " Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 77/81] mm/readahead: fix large folio support in async readahead Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 78/81] mm: vmscan: account for free pages to prevent infinite Loop in throttle_direct_reclaim() Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 79/81] mptcp: fix TCP options overflow Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 80/81] mptcp: fix recvbuffer adjust on sleeping rcvmsg Greg Kroah-Hartman
2025-01-06 15:16 ` [PATCH 6.1 81/81] mptcp: dont always assume copied data in mptcp_cleanup_rbuf() Greg Kroah-Hartman
2025-01-06 18:22 ` [PATCH 6.1 00/81] 6.1.124-rc1 review Pavel Machek
2025-01-06 19:29 ` Florian Fainelli
2025-01-06 22:26 ` Peter Schneider
2025-01-07 0:22 ` SeongJae Park
2025-01-07 7:10 ` Ron Economos
2025-01-07 12:33 ` Mark Brown
2025-01-07 12:36 ` Naresh Kamboju
2025-01-07 12:44 ` Jon Hunter
2025-01-07 20:59 ` [PATCH 6.1] " Hardik Garg
2025-01-07 23:16 ` [PATCH 6.1 00/81] " Shuah Khan
2025-01-08 12:54 ` Muhammad Usama Anjum
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=20250106151131.887998986@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=mathias.nyman@linux.intel.com \
--cc=michal.pecio@gmail.com \
--cc=patches@lists.linux.dev \
--cc=sashal@kernel.org \
--cc=stable@vger.kernel.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