From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
patches@lists.linux.dev, Dexuan Cui <decui@microsoft.com>,
Michael Kelley <mikelley@microsoft.com>,
Lorenzo Pieralisi <lpieralisi@kernel.org>,
Wei Liu <wei.liu@kernel.org>
Subject: [PATCH 5.15 23/96] PCI: hv: Add a per-bus mutex state_lock
Date: Mon, 26 Jun 2023 20:11:38 +0200 [thread overview]
Message-ID: <20230626180747.895428436@linuxfoundation.org> (raw)
In-Reply-To: <20230626180746.943455203@linuxfoundation.org>
From: Dexuan Cui <decui@microsoft.com>
commit 067d6ec7ed5b49380688e06c1e5f883a71bef4fe upstream.
In the case of fast device addition/removal, it's possible that
hv_eject_device_work() can start to run before create_root_hv_pci_bus()
starts to run; as a result, the pci_get_domain_bus_and_slot() in
hv_eject_device_work() can return a 'pdev' of NULL, and
hv_eject_device_work() can remove the 'hpdev', and immediately send a
message PCI_EJECTION_COMPLETE to the host, and the host immediately
unassigns the PCI device from the guest; meanwhile,
create_root_hv_pci_bus() and the PCI device driver can be probing the
dead PCI device and reporting timeout errors.
Fix the issue by adding a per-bus mutex 'state_lock' and grabbing the
mutex before powering on the PCI bus in hv_pci_enter_d0(): when
hv_eject_device_work() starts to run, it's able to find the 'pdev' and call
pci_stop_and_remove_bus_device(pdev): if the PCI device driver has
loaded, the PCI device driver's probe() function is already called in
create_root_hv_pci_bus() -> pci_bus_add_devices(), and now
hv_eject_device_work() -> pci_stop_and_remove_bus_device() is able
to call the PCI device driver's remove() function and remove the device
reliably; if the PCI device driver hasn't loaded yet, the function call
hv_eject_device_work() -> pci_stop_and_remove_bus_device() is able to
remove the PCI device reliably and the PCI device driver's probe()
function won't be called; if the PCI device driver's probe() is already
running (e.g., systemd-udev is loading the PCI device driver), it must
be holding the per-device lock, and after the probe() finishes and releases
the lock, hv_eject_device_work() -> pci_stop_and_remove_bus_device() is
able to proceed to remove the device reliably.
Fixes: 4daace0d8ce8 ("PCI: hv: Add paravirtual PCI front-end for Microsoft Hyper-V VMs")
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Reviewed-by: Michael Kelley <mikelley@microsoft.com>
Acked-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20230615044451.5580-6-decui@microsoft.com
Signed-off-by: Wei Liu <wei.liu@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
drivers/pci/controller/pci-hyperv.c | 29 ++++++++++++++++++++++++++---
1 file changed, 26 insertions(+), 3 deletions(-)
--- a/drivers/pci/controller/pci-hyperv.c
+++ b/drivers/pci/controller/pci-hyperv.c
@@ -483,7 +483,10 @@ struct hv_pcibus_device {
struct fwnode_handle *fwnode;
/* Protocol version negotiated with the host */
enum pci_protocol_version_t protocol_version;
+
+ struct mutex state_lock;
enum hv_pcibus_state state;
+
struct hv_device *hdev;
resource_size_t low_mmio_space;
resource_size_t high_mmio_space;
@@ -2191,6 +2194,8 @@ static void pci_devices_present_work(str
if (!dr)
return;
+ mutex_lock(&hbus->state_lock);
+
/* First, mark all existing children as reported missing. */
spin_lock_irqsave(&hbus->device_list_lock, flags);
list_for_each_entry(hpdev, &hbus->children, list_entry) {
@@ -2272,6 +2277,8 @@ static void pci_devices_present_work(str
break;
}
+ mutex_unlock(&hbus->state_lock);
+
kfree(dr);
}
@@ -2420,6 +2427,8 @@ static void hv_eject_device_work(struct
hpdev = container_of(work, struct hv_pci_dev, wrk);
hbus = hpdev->hbus;
+ mutex_lock(&hbus->state_lock);
+
/*
* Ejection can come before or after the PCI bus has been set up, so
* attempt to find it and tear down the bus state, if it exists. This
@@ -2456,6 +2465,8 @@ static void hv_eject_device_work(struct
put_pcichild(hpdev);
put_pcichild(hpdev);
/* hpdev has been freed. Do not use it any more. */
+
+ mutex_unlock(&hbus->state_lock);
}
/**
@@ -3218,6 +3229,7 @@ static int hv_pci_probe(struct hv_device
return -ENOMEM;
hbus->bridge = bridge;
+ mutex_init(&hbus->state_lock);
hbus->state = hv_pcibus_init;
hbus->wslot_res_allocated = -1;
@@ -3322,9 +3334,11 @@ static int hv_pci_probe(struct hv_device
if (ret)
goto free_irq_domain;
+ mutex_lock(&hbus->state_lock);
+
ret = hv_pci_enter_d0(hdev);
if (ret)
- goto free_irq_domain;
+ goto release_state_lock;
ret = hv_pci_allocate_bridge_windows(hbus);
if (ret)
@@ -3342,12 +3356,15 @@ static int hv_pci_probe(struct hv_device
if (ret)
goto free_windows;
+ mutex_unlock(&hbus->state_lock);
return 0;
free_windows:
hv_pci_free_bridge_windows(hbus);
exit_d0:
(void) hv_pci_bus_exit(hdev, true);
+release_state_lock:
+ mutex_unlock(&hbus->state_lock);
free_irq_domain:
irq_domain_remove(hbus->irq_domain);
free_fwnode:
@@ -3580,20 +3597,26 @@ static int hv_pci_resume(struct hv_devic
if (ret)
goto out;
+ mutex_lock(&hbus->state_lock);
+
ret = hv_pci_enter_d0(hdev);
if (ret)
- goto out;
+ goto release_state_lock;
ret = hv_send_resources_allocated(hdev);
if (ret)
- goto out;
+ goto release_state_lock;
prepopulate_bars(hbus);
hv_pci_restore_msi_state(hbus);
hbus->state = hv_pcibus_installed;
+ mutex_unlock(&hbus->state_lock);
return 0;
+
+release_state_lock:
+ mutex_unlock(&hbus->state_lock);
out:
vmbus_close(hdev->channel);
return ret;
next prev parent reply other threads:[~2023-06-26 18:39 UTC|newest]
Thread overview: 104+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-06-26 18:11 [PATCH 5.15 00/96] 5.15.119-rc1 review Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 01/96] drm/amd/display: fix the system hang while disable PSR Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 02/96] tracing: Add tracing_reset_all_online_cpus_unlocked() function Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 03/96] tpm, tpm_tis: Claim locality in interrupt handler Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 04/96] drm/amd/display: Add minimal pipe split transition state Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 05/96] drm/amd/display: Use dc_update_planes_and_stream Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 06/96] drm/amd/display: Add wrapper to call planes and stream update Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 07/96] tick/common: Align tick period during sched_timer setup Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 08/96] selftests: mptcp: lib: skip if missing symbol Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 09/96] selftests: mptcp: lib: skip if not below kernel version Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 10/96] selftests/mount_setattr: fix redefine struct mount_attr build error Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 11/96] selftests: mptcp: pm nl: remove hardcoded default limits Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 12/96] selftests: mptcp: join: use iptables-legacy if available Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 13/96] selftests: mptcp: join: skip check if MIB counter not supported Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 14/96] nilfs2: fix buffer corruption due to concurrent device reads Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 15/96] ACPI: sleep: Avoid breaking S3 wakeup due to might_sleep() Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 16/96] KVM: Avoid illegal stage2 mapping on invalid memory slot Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 17/96] Drivers: hv: vmbus: Call hv_synic_free() if hv_synic_alloc() fails Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 18/96] Drivers: hv: vmbus: Fix vmbus_wait_for_unload() to scan present CPUs Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 19/96] PCI: hv: Fix a race condition bug in hv_pci_query_relations() Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 20/96] Revert "PCI: hv: Fix a timing issue which causes kdump to fail occasionally" Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 21/96] PCI: hv: Remove the useless hv_pcichild_state from struct hv_pci_dev Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 22/96] PCI: hv: Fix a race condition in hv_irq_unmask() that can cause panic Greg Kroah-Hartman
2023-06-26 18:11 ` Greg Kroah-Hartman [this message]
2023-06-26 18:11 ` [PATCH 5.15 24/96] cgroup: Do not corrupt task iteration when rebinding subsystem Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 25/96] mmc: sdhci-msm: Disable broken 64-bit DMA on MSM8916 Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 26/96] mmc: meson-gx: remove redundant mmc_request_done() call from irq context Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 27/96] mmc: mmci: stm32: fix max busy timeout calculation Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 28/96] ip_tunnels: allow VXLAN/GENEVE to inherit TOS/TTL from VLAN Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 29/96] regulator: pca9450: Fix LDO3OUT and LDO4OUT MASK Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 30/96] regmap: spi-avmm: Fix regmap_bus max_raw_write Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 31/96] writeback: fix dereferencing NULL mapping->host on writeback_page_template Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 32/96] io_uring/net: save msghdr->msg_control for retries Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 33/96] io_uring/net: clear msg_controllen on partial sendmsg retry Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 34/96] io_uring/net: disable partial retries for recvmsg with cmsg Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 35/96] nilfs2: prevent general protection fault in nilfs_clear_dirty_page() Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 36/96] x86/mm: Avoid using set_pgd() outside of real PGD pages Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 37/96] memfd: check for non-NULL file_seals in memfd_create() syscall Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 38/96] mmc: meson-gx: fix deferred probing Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 39/96] ieee802154: hwsim: Fix possible memory leaks Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 40/96] xfrm: Treat already-verified secpath entries as optional Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 41/96] xfrm: interface: rename xfrm_interface.c to xfrm_interface_core.c Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 42/96] xfrm: Ensure policies always checked on XFRM-I input path Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 43/96] bpf: track immediate values written to stack by BPF_ST instruction Greg Kroah-Hartman
2023-06-26 18:11 ` [PATCH 5.15 44/96] bpf: Fix verifier id tracking of scalars on spill Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 45/96] xfrm: fix inbound ipv4/udp/esp packets to UDPv6 dualstack sockets Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 46/96] selftests: net: fcnal-test: check if FIPS mode is enabled Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 47/96] xfrm: Linearize the skb after offloading if needed Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 48/96] net: qca_spi: Avoid high load if QCA7000 is not available Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 49/96] mmc: mtk-sd: fix deferred probing Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 50/96] mmc: mvsdio: " Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 51/96] mmc: omap: " Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 52/96] mmc: omap_hsmmc: " Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 53/96] mmc: owl: " Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 54/96] mmc: sdhci-acpi: " Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 55/96] mmc: sh_mmcif: " Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 56/96] mmc: usdhi60rol0: " Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 57/96] ipvs: align inner_mac_header for encapsulation Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 58/96] net: dsa: mt7530: fix trapping frames on non-MT7621 SoC MT7530 switch Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 59/96] net: dsa: mt7530: fix handling of BPDUs on " Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 60/96] be2net: Extend xmit workaround to BE3 chip Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 61/96] netfilter: nf_tables: fix chain binding transaction logic Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 62/96] netfilter: nf_tables: add NFT_TRANS_PREPARE_ERROR to deal with bound set/chain Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 63/96] netfilter: nft_set_pipapo: .walk does not deal with generations Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 64/96] netfilter: nf_tables: disallow element updates of bound anonymous sets Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 65/96] netfilter: nf_tables: reject unbound anonymous set before commit phase Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 66/96] netfilter: nf_tables: reject unbound chain " Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 67/96] netfilter: nf_tables: disallow updates of anonymous sets Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 68/96] netfilter: nfnetlink_osf: fix module autoload Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 69/96] Revert "net: phy: dp83867: perform soft reset and retain established link" Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 70/96] bpf/btf: Accept function names that contain dots Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 71/96] selftests: forwarding: Fix race condition in mirror installation Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 72/96] sch_netem: acquire qdisc lock in netem_change() Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 73/96] gpio: Allow per-parent interrupt data Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 74/96] gpiolib: Fix GPIO chip IRQ initialization restriction Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 75/96] gpio: sifive: add missing check for platform_get_irq Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 76/96] scsi: target: iscsi: Prevent login threads from racing between each other Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 77/96] HID: wacom: Add error check to wacom_parse_and_register() Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 78/96] arm64: Add missing Set/Way CMO encodings Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 79/96] media: cec: core: dont set last_initiator if tx in progress Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 80/96] nfcsim.c: Fix error checking for debugfs_create_dir Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 81/96] usb: gadget: udc: fix NULL dereference in remove() Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 82/96] nvme: double KA polling frequency to avoid KATO with TBKAS on Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 83/96] Input: soc_button_array - add invalid acpi_index DMI quirk handling Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 84/96] s390/cio: unregister device when the only path is gone Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 85/96] spi: lpspi: disable lpspi module irq in DMA mode Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 86/96] ASoC: simple-card: Add missing of_node_put() in case of error Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 87/96] soundwire: dmi-quirks: add new mapping for HP Spectre x360 Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 88/96] ASoC: nau8824: Add quirk to active-high jack-detect Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 89/96] s390/purgatory: disable branch profiling Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 90/96] ARM: dts: Fix erroneous ADS touchscreen polarities Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 91/96] drm/exynos: vidi: fix a wrong error return Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 92/96] drm/exynos: fix race condition UAF in exynos_g2d_exec_ioctl Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 93/96] drm/radeon: fix race condition UAF in radeon_gem_set_domain_ioctl Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 94/96] vhost_net: revert upend_idx only on retriable error Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 95/96] x86/apic: Fix kernel panic when booting with intremap=off and x2apic_phys Greg Kroah-Hartman
2023-06-26 18:12 ` [PATCH 5.15 96/96] i2c: imx-lpi2c: fix type char overflow issue when calculating the clock cycle Greg Kroah-Hartman
2023-06-27 9:04 ` [PATCH 5.15 00/96] 5.15.119-rc1 review Jon Hunter
2023-06-27 20:09 ` Chris Paterson
2023-06-27 21:34 ` Guenter Roeck
2023-06-28 6:42 ` Naresh Kamboju
2023-06-28 7:27 ` Ron Economos
2023-06-28 17:38 ` Allen Pais
2023-07-21 23:29 ` Guenter Roeck
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=20230626180747.895428436@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=decui@microsoft.com \
--cc=lpieralisi@kernel.org \
--cc=mikelley@microsoft.com \
--cc=patches@lists.linux.dev \
--cc=stable@vger.kernel.org \
--cc=wei.liu@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;
as well as URLs for NNTP newsgroup(s).