Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Add support for multiple disks #2

Closed
rbradford opened this issue May 14, 2019 · 0 comments · Fixed by #29
Closed

Add support for multiple disks #2

rbradford opened this issue May 14, 2019 · 0 comments · Fixed by #29

Comments

@rbradford
Copy link
Member

In order to be able to configure with cloud-init we need to be able to support specifying multiple disks at start time.
@rbradford rbradford mentioned this issue May 22, 2019
Merged
@sameo sameo mentioned this issue Aug 30, 2019
Merged
sameo referenced this issue in sameo/cloud-hypervisor-1 Sep 19, 2019
main: Group cli options logically
4fe331f 
With the API server socket option, we want to support new UX patterns:

1) The CLI model: The user entirely configures a guest from the
command line and does not expect to be able to control it dynamically
through the API socket. This is the current model.

2) The API control model: The user does not configure anything from
the command line, and calls the cloud-hypervisor binary with the API
socket path command line option only. Everything will be done from
the API socket, including the guest initial configuration and creation.

3) The hybrid model: The user wants to quickly create and start a guest
and then asynchronously control it from the API socket. The initial
configuration setup is done through the command line and further control
go through the API socket. In this model, we need to make sure that the
guest configuration is valid. We don't want to support invalid
configurations that could implicitly be overwritten through the API
socket.

In order to support the new models (#2 and #3) we need to be able to
tell if an empty guest configuration was passed through the command line
or not. If it is empty, then we go with model #2. If it is not empty, we
either go with #1 or #3. One way to check if a guest configuration is
empty or not is by logically grouping the command line options into a VM
config group and a VMM config one. If the VM config group is empty, then
the user is not trying to configure a guest.

Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
@sameo sameo mentioned this issue Sep 19, 2019
Closed
sameo referenced this issue in sameo/cloud-hypervisor-1 Sep 19, 2019
main: Group cli options logically
48da5d9 
With the API server socket option, we want to support new UX patterns:

1) The CLI model: The user entirely configures a guest from the
command line and does not expect to be able to control it dynamically
through the API socket. This is the current model.

2) The API control model: The user does not configure anything from
the command line, and calls the cloud-hypervisor binary with the API
socket path command line option only. Everything will be done from
the API socket, including the guest initial configuration and creation.

3) The hybrid model: The user wants to quickly create and start a guest
and then asynchronously control it from the API socket. The initial
configuration setup is done through the command line and further control
go through the API socket. In this model, we need to make sure that the
guest configuration is valid. We don't want to support invalid
configurations that could implicitly be overwritten through the API
socket.

In order to support the new models (#2 and #3) we need to be able to
tell if an empty guest configuration was passed through the command line
or not. If it is empty, then we go with model #2. If it is not empty, we
either go with #1 or #3. One way to check if a guest configuration is
empty or not is by logically grouping the command line options into a VM
config group and a VMM config one. If the VM config group is empty, then
the user is not trying to configure a guest.

Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
@russell-islam russell-islam mentioned this issue Jun 18, 2020
Merged
@magnate3 magnate3 mentioned this issue Dec 29, 2020
Closed
@rbradford rbradford mentioned this issue Feb 10, 2021
Merged
arafatms pushed a commit to arafatms/cloud-hypervisor that referenced this issue Aug 3, 2021
@arafatms
virtio-devices: vhost-user: Send set_vring_base before setup inflight…
1ab5d5a 
… I/O tracking

backend like SPDK required to know how many virt queues to be handled
before gets VHOST_USER_SET_INFLIGHT_FD message.

fix dpdk core dump while processing vhost_user_set_inflight_fd:

#0  0x00007fffef47c347 in vhost_user_set_inflight_fd (pdev=0x7fffe2895998, msg=0x7fffe28956f0, main_fd=545) at ../lib/librte_vhost/vhost_user.c:1570
cloud-hypervisor#1  0x00007fffef47e7b9 in vhost_user_msg_handler (vid=0, fd=545) at ../lib/librte_vhost/vhost_user.c:2735
cloud-hypervisor#2  0x00007fffef46bac0 in vhost_user_read_cb (connfd=545, dat=0x7fffdc0008c0, remove=0x7fffe2895a64) at ../lib/librte_vhost/socket.c:309
cloud-hypervisor#3  0x00007fffef45b3f6 in fdset_event_dispatch (arg=0x7fffef6dc2e0 <vhost_user+8192>) at ../lib/librte_vhost/fd_man.c:286
cloud-hypervisor#4  0x00007ffff09926f3 in rte_thread_init (arg=0x15ee180) at ../lib/librte_eal/common/eal_common_thread.c:175

Signed-off-by: Arafatms <aierpatijiang1@kingsoft.com>
@arafatms arafatms mentioned this issue Aug 3, 2021
Closed
arafatms added a commit to arafatms/cloud-hypervisor that referenced this issue Aug 3, 2021
@arafatms
virtio-devices: vhost-user: Send set_vring_base before setup inflight…
125d7ab 
… I/O tracking

backend like SPDK required to know how many virt queues to be handled
before gets VHOST_USER_SET_INFLIGHT_FD message.

fix dpdk core dump while processing vhost_user_set_inflight_fd:
#0 0x00007fffef47c347 in vhost_user_set_inflight_fd (pdev=0x7fffe2895998, msg=0x7fffe28956f0, main_fd=545) at ../lib/librte_vhost/vhost_user.c:1570
cloud-hypervisor#1 0x00007fffef47e7b9 in vhost_user_msg_handler (vid=0, fd=545) at ../lib/librte_vhost/vhost_user.c:2735
cloud-hypervisor#2 0x00007fffef46bac0 in vhost_user_read_cb (connfd=545, dat=0x7fffdc0008c0, remove=0x7fffe2895a64) at ../lib/librte_vhost/socket.c:309
cloud-hypervisor#3 0x00007fffef45b3f6 in fdset_event_dispatch (arg=0x7fffef6dc2e0 <vhost_user+8192>) at ../lib/librte_vhost/fd_man.c:286
cloud-hypervisor#4 0x00007ffff09926f3 in rte_thread_init (arg=0x15ee180) at ../lib/librte_eal/common/eal_common_thread.c:175

Signed-off-by: Arafatms <arafatms@outlook.com>
@arafatms arafatms mentioned this issue Aug 3, 2021
Merged
arafatms added a commit to arafatms/cloud-hypervisor that referenced this issue Aug 4, 2021
@arafatms
virtio-devices: vhost-user: Send set_vring_num before setup inflight …
3791f18 
…I/O tracking

backend like SPDK required to know how many virt queues to be handled
before gets VHOST_USER_SET_INFLIGHT_FD message.

fix dpdk core dump while processing vhost_user_set_inflight_fd:
    #0 0x00007fffef47c347 in vhost_user_set_inflight_fd (pdev=0x7fffe2895998, msg=0x7fffe28956f0, main_fd=545) at ../lib/librte_vhost/vhost_user.c:1570
    cloud-hypervisor#1 0x00007fffef47e7b9 in vhost_user_msg_handler (vid=0, fd=545) at ../lib/librte_vhost/vhost_user.c:2735
    cloud-hypervisor#2 0x00007fffef46bac0 in vhost_user_read_cb (connfd=545, dat=0x7fffdc0008c0, remove=0x7fffe2895a64) at ../lib/librte_vhost/socket.c:309
    cloud-hypervisor#3 0x00007fffef45b3f6 in fdset_event_dispatch (arg=0x7fffef6dc2e0 <vhost_user+8192>) at ../lib/librte_vhost/fd_man.c:286
    cloud-hypervisor#4 0x00007ffff09926f3 in rte_thread_init (arg=0x15ee180) at ../lib/librte_eal/common/eal_common_thread.c:175

Signed-off-by: Arafatms <arafatms@outlook.com>
sboeuf pushed a commit that referenced this issue Aug 4, 2021
@arafatms
virtio-devices: vhost-user: Send set_vring_num before setup inflight …
8fb53eb 
…I/O tracking

backend like SPDK required to know how many virt queues to be handled
before gets VHOST_USER_SET_INFLIGHT_FD message.

fix dpdk core dump while processing vhost_user_set_inflight_fd:
    #0 0x00007fffef47c347 in vhost_user_set_inflight_fd (pdev=0x7fffe2895998, msg=0x7fffe28956f0, main_fd=545) at ../lib/librte_vhost/vhost_user.c:1570
    #1 0x00007fffef47e7b9 in vhost_user_msg_handler (vid=0, fd=545) at ../lib/librte_vhost/vhost_user.c:2735
    #2 0x00007fffef46bac0 in vhost_user_read_cb (connfd=545, dat=0x7fffdc0008c0, remove=0x7fffe2895a64) at ../lib/librte_vhost/socket.c:309
    #3 0x00007fffef45b3f6 in fdset_event_dispatch (arg=0x7fffef6dc2e0 <vhost_user+8192>) at ../lib/librte_vhost/fd_man.c:286
    #4 0x00007ffff09926f3 in rte_thread_init (arg=0x15ee180) at ../lib/librte_eal/common/eal_common_thread.c:175

Signed-off-by: Arafatms <arafatms@outlook.com>
sboeuf pushed a commit that referenced this issue Jan 6, 2022
@rbradford
vmm: Fix potential deadlock in CpuManager
e7db354 
The lock on the config should not be held whilst calling into
CpuManager::new().

WARNING: ThreadSanitizer: lock-order-inversion (potential deadlock) (pid=24176)
  Cycle in lock order graph: M13 (0x7b0c000001e0) => M43 (0x7b0c00001a70) => M13

  Mutex M43 acquired here while holding mutex M13 in thread T1:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x663984)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x491bde)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::h8a843a6e74b34c4a /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36cd8b)
    #4 vmm::cpu::CpuManager::new::h1cc88224a2a50d87 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:575:30 (cloud-hypervisor+0x372e16)
    #5 vmm::vm::Vm::new_from_memory_manager::ha2a4467be260e93c /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:592:27 (cloud-hypervisor+0x330b86)
    #6 vmm::vm::Vm::new::h43efe7c6cd97ede5 /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:784:22 (cloud-hypervisor+0x3343df)
    #7 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:379:26 (cloud-hypervisor+0x2e5ba8)
    #8 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f44e0)
    #9 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1dda20)
    #10 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x659bfa)
    #11 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x44103e)
    #12 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dda8e)
    #13 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d3701)
    #14 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d4748)
    #15 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d2489)
    #16 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57bdfe)
    #17 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f943)
    #18 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cf425)
    #19 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d4c2)
    #20 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d4c2)
    #21 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d4c2)

  Mutex M13 previously acquired by the same thread here:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x663984)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x491bde)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::ha29f58bbf496356a /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36d03b)
    #4 vmm::vm::Vm::new_from_memory_manager::ha2a4467be260e93c /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:593:14 (cloud-hypervisor+0x3307ee)
    #5 vmm::vm::Vm::new::h43efe7c6cd97ede5 /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:784:22 (cloud-hypervisor+0x3343df)
    #6 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:379:26 (cloud-hypervisor+0x2e5ba8)
    #7 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f44e0)
    #8 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1dda20)
    #9 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x659bfa)
    #10 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x44103e)
    #11 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dda8e)
    #12 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d3701)
    #13 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d4748)
    #14 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d2489)
    #15 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57bdfe)
    #16 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f943)
    #17 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cf425)
    #18 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d4c2)
    #19 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d4c2)
    #20 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d4c2)

  Mutex M13 acquired here while holding mutex M43 in thread T1:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x663984)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x491bde)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::ha29f58bbf496356a /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36d03b)
    #4 vmm::device_manager::DeviceManager::add_console_device::h1d2b419feef80564 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:1839:29 (cloud-hypervisor+0x3972f4)
    #5 vmm::device_manager::DeviceManager::create_devices::h29fc5b8a20e1aea5 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:1143:24 (cloud-hypervisor+0x38f068)
    #6 vmm::vm::Vm::new::h43efe7c6cd97ede5 /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:798:9 (cloud-hypervisor+0x334671)
    #7 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:379:26 (cloud-hypervisor+0x2e5ba8)
    #8 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f44e0)
    #9 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1dda20)
    #10 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x659bfa)
    #11 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x44103e)
    #12 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dda8e)
    #13 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d3701)
    #14 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d4748)
    #15 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d2489)
    #16 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57bdfe)
    #17 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f943)
    #18 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cf425)
    #19 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d4c2)
    #20 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d4c2)
    #21 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d4c2)

  Mutex M43 previously acquired by the same thread here:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x663984)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x491bde)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::h8a843a6e74b34c4a /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36cd8b)
    #4 vmm::vm::Vm::new::h43efe7c6cd97ede5 /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:798:9 (cloud-hypervisor+0x334537)
    #5 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:379:26 (cloud-hypervisor+0x2e5ba8)
    #6 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f44e0)
    #7 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1dda20)
    #8 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x659bfa)
    #9 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x44103e)
    #10 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dda8e)
    #11 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d3701)
    #12 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d4748)
    #13 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d2489)
    #14 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57bdfe)
    #15 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f943)
    #16 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cf425)
    #17 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d4c2)
    #18 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d4c2)
    #19 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d4c2)

Signed-off-by: Rob Bradford <robert.bradford@intel.com>
sboeuf pushed a commit that referenced this issue Jan 6, 2022
@rbradford
vmm, pci: Fix potential deadlock in PCI BAR allocation
9ef1187 
The allocator is locked by both the BAR allocation code and the
interrupt allocation code. Resulting in a potential lock inversion
error.

WARNING: ThreadSanitizer: lock-order-inversion (potential deadlock) (pid=26318)
  Cycle in lock order graph: M87 (0x7b0c00001e30) => M28 (0x7b0c00001830) => M87

  Mutex M28 acquired here while holding mutex M87 in thread T1:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x663954)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x491bae)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::hc61622e5536f5b72 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36d07b)
    #4 _$LT$vmm..interrupt..MsiInterruptManager$LT$kvm_bindings..x86..bindings..kvm_irq_routing_entry$GT$$u20$as$u20$vm_device..interrupt..InterruptManager$GT$::create_group::hd412b5e1e8eeacc2 /home/rob/src/cloud-hypervisor/vmm/src/interrupt.rs:310:29 (cloud-hypervisor+0x6d1403)
    #5 virtio_devices::transport::pci_device::VirtioPciDevice::new::h3af603c3f00f4b3d /home/rob/src/cloud-hypervisor/virtio-devices/src/transport/pci_device.rs:376:38 (cloud-hypervisor+0x8e6137)
    #6 vmm::device_manager::DeviceManager::add_virtio_pci_device::h23608151d7668a1c /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:3333:37 (cloud-hypervisor+0x3b6339)
    #7 vmm::device_manager::DeviceManager::add_pci_devices::h136cc20cbeb6b977 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:1236:30 (cloud-hypervisor+0x390aad)
    #8 vmm::device_manager::DeviceManager::create_devices::h29fc5b8a20e1aea5 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:1155:9 (cloud-hypervisor+0x38f48c)
    #9 vmm::vm::Vm::new::h43efe7c6cd97ede5 /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:799:9 (cloud-hypervisor+0x334641)
    #10 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:379:26 (cloud-hypervisor+0x2e5ba8)
    #11 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f44e0)
    #12 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1dda20)
    #13 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x659bca)
    #14 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x44100e)
    #15 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dda5e)
    #16 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d36d1)
    #17 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d4718)
    #18 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d2459)
    #19 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57bdce)
    #20 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f913)
    #21 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cf3f5)
    #22 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d492)
    #23 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d492)
    #24 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d492)

  Mutex M87 previously acquired by the same thread here:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::h9a2d3e97e05c6430 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x9ea344)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::h8abb3b5cf55c0264 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x96face)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::hecec128d40c6dd44 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x97120a)
    #4 virtio_devices::transport::pci_device::VirtioPciDevice::new::h3af603c3f00f4b3d /home/rob/src/cloud-hypervisor/virtio-devices/src/transport/pci_device.rs:356:29 (cloud-hypervisor+0x8e5c0e)
    #5 vmm::device_manager::DeviceManager::add_virtio_pci_device::h23608151d7668a1c /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:3333:37 (cloud-hypervisor+0x3b6339)
    #6 vmm::device_manager::DeviceManager::add_pci_devices::h136cc20cbeb6b977 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:1236:30 (cloud-hypervisor+0x390aad)
    #7 vmm::device_manager::DeviceManager::create_devices::h29fc5b8a20e1aea5 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:1155:9 (cloud-hypervisor+0x38f48c)
    #8 vmm::vm::Vm::new::h43efe7c6cd97ede5 /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:799:9 (cloud-hypervisor+0x334641)
    #9 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:379:26 (cloud-hypervisor+0x2e5ba8)
    #10 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f44e0)
    #11 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1dda20)
    #12 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x659bca)
    #13 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x44100e)
    #14 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dda5e)
    #15 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d36d1)
    #16 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d4718)
    #17 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d2459)
    #18 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57bdce)
    #19 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f913)
    #20 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cf3f5)
    #21 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d492)
    #22 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d492)
    #23 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d492)

  Mutex M87 acquired here while holding mutex M28 in thread T1:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::h9a2d3e97e05c6430 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x9ea344)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::h8abb3b5cf55c0264 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x96face)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::hecec128d40c6dd44 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x97120a)
    #4 _$LT$virtio_devices..transport..pci_device..VirtioPciDevice$u20$as$u20$pci..device..PciDevice$GT$::allocate_bars::h39dc42b48fc8264c /home/rob/src/cloud-hypervisor/virtio-devices/src/transport/pci_device.rs:850:22 (cloud-hypervisor+0x8eb1a4)
    #5 vmm::device_manager::DeviceManager::add_pci_device::h561f6c8ed61db117 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:3087:20 (cloud-hypervisor+0x3b0c62)
    #6 vmm::device_manager::DeviceManager::add_virtio_pci_device::h23608151d7668a1c /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:3359:20 (cloud-hypervisor+0x3b6707)
    #7 vmm::device_manager::DeviceManager::add_pci_devices::h136cc20cbeb6b977 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:1236:30 (cloud-hypervisor+0x390aad)
    #8 vmm::device_manager::DeviceManager::create_devices::h29fc5b8a20e1aea5 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:1155:9 (cloud-hypervisor+0x38f48c)
    #9 vmm::vm::Vm::new::h43efe7c6cd97ede5 /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:799:9 (cloud-hypervisor+0x334641)
    #10 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:379:26 (cloud-hypervisor+0x2e5ba8)
    #11 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f44e0)
    #12 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1dda20)
    #13 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x659bca)
    #14 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x44100e)
    #15 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dda5e)
    #16 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d36d1)
    #17 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d4718)
    #18 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d2459)
    #19 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57bdce)
    #20 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f913)
    #21 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cf3f5)
    #22 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d492)
    #23 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d492)
    #24 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d492)

  Mutex M28 previously acquired by the same thread here:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x663954)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x491bae)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::hc61622e5536f5b72 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36d07b)
    #4 vmm::device_manager::DeviceManager::add_pci_device::h561f6c8ed61db117 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:3091:22 (cloud-hypervisor+0x3b0a95)
    #5 vmm::device_manager::DeviceManager::add_virtio_pci_device::h23608151d7668a1c /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:3359:20 (cloud-hypervisor+0x3b6707)
    #6 vmm::device_manager::DeviceManager::add_pci_devices::h136cc20cbeb6b977 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:1236:30 (cloud-hypervisor+0x390aad)
    #7 vmm::device_manager::DeviceManager::create_devices::h29fc5b8a20e1aea5 /home/rob/src/cloud-hypervisor/vmm/src/device_manager.rs:1155:9 (cloud-hypervisor+0x38f48c)
    #8 vmm::vm::Vm::new::h43efe7c6cd97ede5 /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:799:9 (cloud-hypervisor+0x334641)
    #9 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:379:26 (cloud-hypervisor+0x2e5ba8)
    #10 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f44e0)
    #11 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1dda20)
    #12 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x659bca)
    #13 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x44100e)
    #14 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dda5e)
    #15 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d36d1)
    #16 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d4718)
    #17 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d2459)
    #18 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57bdce)
    #19 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f913)
    #20 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cf3f5)
    #21 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d492)
    #22 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d492)
    #23 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d492)

Signed-off-by: Rob Bradford <robert.bradford@intel.com>
sboeuf pushed a commit that referenced this issue Jan 6, 2022
@rbradford
vmm: Fix potential deadlock in CpuManager
2f46647 
Delay creating the mutex on the Vcpu until later preventing a potential
lock inversion between CpuManager and the Vcpu.

WARNING: ThreadSanitizer: lock-order-inversion (potential deadlock) (pid=28799)
  Cycle in lock order graph: M48 (0x7b0c00001aa0) => M117 (0x7b0c00002280) => M48

  Mutex M117 acquired here while holding mutex M48 in thread T1:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x662fc4)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x4915ae)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::h14cfa3c8f5ba878a /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36c82b)
    #4 vmm::cpu::CpuManager::create_vcpu::hd5878da6efae8d68 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:665:13 (cloud-hypervisor+0x3743de)
    #5 vmm::cpu::CpuManager::create_vcpus::h3c747553a1d5bc4e /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:704:13 (cloud-hypervisor+0x374d87)
    #6 vmm::cpu::CpuManager::create_boot_vcpus::he8eeca10785067c1 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:938:9 (cloud-hypervisor+0x377305)
    #7 vmm::vm::Vm::boot::hc2ca6b16f996267b /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:1986:9 (cloud-hypervisor+0x3432d3)
    #8 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:397:13 (cloud-hypervisor+0x2e5d35)
    #9 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f42d0)
    #10 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1dd8d0)
    #11 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x65923a)
    #12 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x440a0e)
    #13 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dd0ce)
    #14 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2d41)
    #15 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3d88)
    #16 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d1ac9)
    #17 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57b72e)
    #18 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f313)
    #19 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cef15)
    #20 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d662)
    #21 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d662)
    #22 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d662)

  Mutex M48 previously acquired by the same thread here:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x662fc4)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x4915ae)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::hecf671add5fe1762 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36cfbb)
    #4 vmm::vm::Vm::boot::hc2ca6b16f996267b /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:1986:9 (cloud-hypervisor+0x343237)
    #5 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:397:13 (cloud-hypervisor+0x2e5d35)
    #6 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f42d0)
    #7 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1dd8d0)
    #8 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x65923a)
    #9 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x440a0e)
    #10 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dd0ce)
    #11 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2d41)
    #12 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3d88)
    #13 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d1ac9)
    #14 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57b72e)
    #15 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f313)
    #16 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cef15)
    #17 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d662)
    #18 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d662)
    #19 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d662)

  Mutex M48 acquired here while holding mutex M117 in thread T3:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::h967991d72ceb6eb0 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0xd94dc4)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::h8779639163126a21 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0xd90c9e)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::hd85239d207beb12f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0xd6e98a)
    #4 vm_device::bus::Bus::write::hf20f991e71af3199 /home/rob/src/cloud-hypervisor/vm-device/src/bus.rs:235:16 (cloud-hypervisor+0xd8dcfd)
    #5 _$LT$vmm..vm..VmOps$u20$as$u20$hypervisor..vm..VmmOps$GT$::mmio_write::hc759194aaebc7399 /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:424:15 (cloud-hypervisor+0x32d94f)
    #6 _$LT$hypervisor..kvm..KvmVcpu$u20$as$u20$hypervisor..cpu..Vcpu$GT$::run::h94762dfba6642fb2 /home/rob/src/cloud-hypervisor/hypervisor/src/kvm/mod.rs:1003:32 (cloud-hypervisor+0xcc3ea8)
    #7 vmm::cpu::Vcpu::run::hd5cf042157f95bea /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:327:9 (cloud-hypervisor+0x3700c4)
    #8 vmm::cpu::CpuManager::start_vcpu::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h37e4dd8619b3a5e5 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:819:35 (cloud-hypervisor+0x47782b)
    #9 std::panicking::try::do_call::h093e4d1434150d77 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2aba)
    #10 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3d88)
    #11 std::panicking::try::hee9535cb997282b4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d265f)
    #12 std::panic::catch_unwind::he3908c4d08a8a028 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57bac9)
    #13 vmm::cpu::CpuManager::start_vcpu::_$u7b$$u7b$closure$u7d$$u7d$::h29472aaa3a600231 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:788:21 (cloud-hypervisor+0x477126)
    #14 std::sys_common::backtrace::__rust_begin_short_backtrace::hcfc2f02361c98808 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x6592fb)
    #15 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h08b82db41d7af2f2 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x4408bf)
    #16 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h7ebad9d94e64fa5f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dd1af)
    #17 std::panicking::try::do_call::h121fafbdf5cf84af /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2b51)
    #18 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3d88)
    #19 std::panicking::try::h79e25f019cd90522 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d1def)
    #20 std::panic::catch_unwind::h5a0619a53bbd611d /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57b8cf)
    #21 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h1cfd689c9d362e48 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43facb)
    #22 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::h6642b1b3a2289640 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3ced25)
    #23 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d662)
    #24 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d662)
    #25 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d662)

  Mutex M117 previously acquired by the same thread here:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x662fc4)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x4915ae)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::h14cfa3c8f5ba878a /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36c82b)
    #4 vmm::cpu::CpuManager::start_vcpu::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h37e4dd8619b3a5e5 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:819:35 (cloud-hypervisor+0x477799)
    #5 std::panicking::try::do_call::h093e4d1434150d77 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2aba)
    #6 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3d88)
    #7 std::panicking::try::hee9535cb997282b4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d265f)
    #8 std::panic::catch_unwind::he3908c4d08a8a028 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57bac9)
    #9 vmm::cpu::CpuManager::start_vcpu::_$u7b$$u7b$closure$u7d$$u7d$::h29472aaa3a600231 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:788:21 (cloud-hypervisor+0x477126)
    #10 std::sys_common::backtrace::__rust_begin_short_backtrace::hcfc2f02361c98808 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x6592fb)
    #11 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h08b82db41d7af2f2 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x4408bf)
    #12 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h7ebad9d94e64fa5f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dd1af)
    #13 std::panicking::try::do_call::h121fafbdf5cf84af /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2b51)
    #14 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3d88)
    #15 std::panicking::try::h79e25f019cd90522 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d1def)
    #16 std::panic::catch_unwind::h5a0619a53bbd611d /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57b8cf)
    #17 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h1cfd689c9d362e48 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43facb)
    #18 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::h6642b1b3a2289640 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3ced25)
    #19 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d662)
    #20 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d662)
    #21 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d662)

Signed-off-by: Rob Bradford <robert.bradford@intel.com>
sboeuf pushed a commit that referenced this issue Jan 6, 2022
@rbradford
vmm: Fix potential deadlock in CpuManager
d57c496 
Remove requirement for CpuManager to lock the Vcpu when starting the
vCPU as the numerical id corresponds to the index in the the vector.
This avoids a potential lock inversion between the Vcpu and CpuManager.

WARNING: ThreadSanitizer: lock-order-inversion (potential deadlock) (pid=30497)
  Cycle in lock order graph: M48 (0x7b0c00001aa0) => M121 (0x7b0c000022b0) => M48

  Mutex M121 acquired here while holding mutex M48 in thread T1:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x662ff4)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x4915de)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::h14cfa3c8f5ba878a /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36ca3b)
    #4 vmm::cpu::CpuManager::start_vcpu::h290fdbb4b7124ec5 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:710:22 (cloud-hypervisor+0x375023)
    #5 vmm::cpu::CpuManager::activate_vcpus::h2eab380826588391 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:902:13 (cloud-hypervisor+0x376b7b)
    #6 vmm::cpu::CpuManager::start_boot_vcpus::hd80cafe6aa4e8279 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:937:9 (cloud-hypervisor+0x3773af)
    #7 vmm::vm::Vm::boot::hc2ca6b16f996267b /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:2063:9 (cloud-hypervisor+0x343d57)
    #8 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:397:13 (cloud-hypervisor+0x2e5f45)
    #9 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f44e0)
    #10 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1ddae0)
    #11 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x65926a)
    #12 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x440a3e)
    #13 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dd0fe)
    #14 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2d71)
    #15 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3db8)
    #16 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d1af9)
    #17 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57b75e)
    #18 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f343)
    #19 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cef45)
    #20 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d692)
    #21 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d692)
    #22 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d692)

  Mutex M48 previously acquired by the same thread here:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x662ff4)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x4915de)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::hecf671add5fe1762 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36d1cb)
    #4 vmm::vm::Vm::boot::hc2ca6b16f996267b /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:2063:9 (cloud-hypervisor+0x343cd1)
    #5 vmm::Vmm::vm_boot::h06bdf54b95d5e14f /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:397:13 (cloud-hypervisor+0x2e5f45)
    #6 vmm::Vmm::control_loop::h40c9b48c7b800bed /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:1299:48 (cloud-hypervisor+0x2f44e0)
    #7 vmm::start_vmm_thread::_$u7b$$u7b$closure$u7d$$u7d$::h016d2f7cff698175 /home/rob/src/cloud-hypervisor/vmm/src/lib.rs:263:17 (cloud-hypervisor+0x1ddae0)
    #8 std::sys_common::backtrace::__rust_begin_short_backtrace::h7fd2df3e7cfba503 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x65926a)
    #9 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h89880b05fe892d7e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x440a3e)
    #10 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h487382524d80571f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dd0fe)
    #11 std::panicking::try::do_call::h1d9c2ccdc39f3322 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2d71)
    #12 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3db8)
    #13 std::panicking::try::h251306df23d21913 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d1af9)
    #14 std::panic::catch_unwind::h2a9ac2fb12c3c64e /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57b75e)
    #15 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h10f4c340611b55e4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43f343)
    #16 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::hdd9b37241caf97b3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3cef45)
    #17 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d692)
    #18 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d692)
    #19 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d692)

  Mutex M48 acquired here while holding mutex M121 in thread T4:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::h967991d72ceb6eb0 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0xd94df4)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::h8779639163126a21 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0xd90cce)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::hd85239d207beb12f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0xd6e9ba)
    #4 vm_device::bus::Bus::write::hf20f991e71af3199 /home/rob/src/cloud-hypervisor/vm-device/src/bus.rs:235:16 (cloud-hypervisor+0xd8dd2d)
    #5 _$LT$vmm..vm..VmOps$u20$as$u20$hypervisor..vm..VmmOps$GT$::mmio_write::hc759194aaebc7399 /home/rob/src/cloud-hypervisor/vmm/src/vm.rs:424:15 (cloud-hypervisor+0x32db5f)
    #6 _$LT$hypervisor..kvm..KvmVcpu$u20$as$u20$hypervisor..cpu..Vcpu$GT$::run::h94762dfba6642fb2 /home/rob/src/cloud-hypervisor/hypervisor/src/kvm/mod.rs:1003:32 (cloud-hypervisor+0xcc3ed8)
    #7 vmm::cpu::Vcpu::run::hd5cf042157f95bea /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:327:9 (cloud-hypervisor+0x370234)
    #8 vmm::cpu::CpuManager::start_vcpu::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h37e4dd8619b3a5e5 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:813:35 (cloud-hypervisor+0x47785b)
    #9 std::panicking::try::do_call::h093e4d1434150d77 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2aea)
    #10 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3db8)
    #11 std::panicking::try::hee9535cb997282b4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d268f)
    #12 std::panic::catch_unwind::he3908c4d08a8a028 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57baf9)
    #13 vmm::cpu::CpuManager::start_vcpu::_$u7b$$u7b$closure$u7d$$u7d$::h29472aaa3a600231 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:782:21 (cloud-hypervisor+0x477156)
    #14 std::sys_common::backtrace::__rust_begin_short_backtrace::hcfc2f02361c98808 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x65932b)
    #15 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h08b82db41d7af2f2 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x4408ef)
    #16 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h7ebad9d94e64fa5f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dd1df)
    #17 std::panicking::try::do_call::h121fafbdf5cf84af /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2b81)
    #18 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3db8)
    #19 std::panicking::try::h79e25f019cd90522 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d1e1f)
    #20 std::panic::catch_unwind::h5a0619a53bbd611d /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57b8ff)
    #21 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h1cfd689c9d362e48 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43fafb)
    #22 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::h6642b1b3a2289640 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3ced55)
    #23 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d692)
    #24 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d692)
    #25 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d692)

  Mutex M121 previously acquired by the same thread here:
    #0 pthread_mutex_lock /rustc/llvm/src/llvm-project/compiler-rt/lib/tsan/../sanitizer_common/sanitizer_common_interceptors.inc:4249:3 (cloud-hypervisor+0x9c368)
    #1 std::sys::unix::mutex::Mutex::lock::hcd1b9aa06ff775d3 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/mutex.rs:63:17 (cloud-hypervisor+0x662ff4)
    #2 std::sys_common::mutex::MovableMutex::raw_lock::hff98d0b036469bca /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/mutex.rs:76:18 (cloud-hypervisor+0x4915de)
    #3 std::sync::mutex::Mutex$LT$T$GT$::lock::h14cfa3c8f5ba878a /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sync/mutex.rs:267:13 (cloud-hypervisor+0x36ca3b)
    #4 vmm::cpu::CpuManager::start_vcpu::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h37e4dd8619b3a5e5 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:813:35 (cloud-hypervisor+0x4777c9)
    #5 std::panicking::try::do_call::h093e4d1434150d77 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2aea)
    #6 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3db8)
    #7 std::panicking::try::hee9535cb997282b4 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d268f)
    #8 std::panic::catch_unwind::he3908c4d08a8a028 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57baf9)
    #9 vmm::cpu::CpuManager::start_vcpu::_$u7b$$u7b$closure$u7d$$u7d$::h29472aaa3a600231 /home/rob/src/cloud-hypervisor/vmm/src/cpu.rs:782:21 (cloud-hypervisor+0x477156)
    #10 std::sys_common::backtrace::__rust_begin_short_backtrace::hcfc2f02361c98808 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys_common/backtrace.rs:123:18 (cloud-hypervisor+0x65932b)
    #11 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::_$u7b$$u7b$closure$u7d$$u7d$::h08b82db41d7af2f2 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:477:17 (cloud-hypervisor+0x4408ef)
    #12 _$LT$core..panic..unwind_safe..AssertUnwindSafe$LT$F$GT$$u20$as$u20$core..ops..function..FnOnce$LT$$LP$$RP$$GT$$GT$::call_once::h7ebad9d94e64fa5f /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/panic/unwind_safe.rs:271:9 (cloud-hypervisor+0x6dd1df)
    #13 std::panicking::try::do_call::h121fafbdf5cf84af /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:406:40 (cloud-hypervisor+0x6d2b81)
    #14 __rust_try 3hkmq3dzyyv5ejsx (cloud-hypervisor+0x6d3db8)
    #15 std::panicking::try::h79e25f019cd90522 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panicking.rs:370:19 (cloud-hypervisor+0x6d1e1f)
    #16 std::panic::catch_unwind::h5a0619a53bbd611d /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/panic.rs:133:14 (cloud-hypervisor+0x57b8ff)
    #17 std::thread::Builder::spawn_unchecked::_$u7b$$u7b$closure$u7d$$u7d$::h1cfd689c9d362e48 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/thread/mod.rs:476:30 (cloud-hypervisor+0x43fafb)
    #18 core::ops::function::FnOnce::call_once$u7b$$u7b$vtable.shim$u7d$$u7d$::h6642b1b3a2289640 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/core/src/ops/function.rs:227:5 (cloud-hypervisor+0x3ced55)
    #19 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::ha5022a6bb7833f62 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d692)
    #20 _$LT$alloc..boxed..Box$LT$F$C$A$GT$$u20$as$u20$core..ops..function..FnOnce$LT$Args$GT$$GT$::call_once::h481697829cbc6746 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/alloc/src/boxed.rs:1854:9 (cloud-hypervisor+0x119d692)
    #21 std::sys::unix::thread::Thread::new::thread_start::h6fad62c4c393bbe7 /rustc/7d6f948173ccb18822bab13d548c65632db5f0aa/library/std/src/sys/unix/thread.rs:108:17 (cloud-hypervisor+0x119d692)

Signed-off-by: Rob Bradford <robert.bradford@intel.com>
up2wing pushed a commit to up2wing/cloud-hypervisor that referenced this issue Mar 9, 2022
vmm: interrupt: fix msi mask irq causing kernel panic on AMD
f474c3f 
When mask a msi irq, we set the entry.masked to be true, so kvm
hypervisor will not pass the gsi to kernel through KVM_SET_GSI_ROUTING
ioctl which update kvm->irq_routing. This will trigger kernel
panic on AMD platform when the gsi is the largest one in kernel
kvm->irqfds.items:

crash> bt
PID: 22218  TASK: ffff951a6ad74980  CPU: 73  COMMAND: "vcpu8"
 #0 [ffffb1ba6707fa40] machine_kexec at ffffffff8565b397
 cloud-hypervisor#1 [ffffb1ba6707fa90] __crash_kexec at ffffffff85788a6d
 cloud-hypervisor#2 [ffffb1ba6707fb58] crash_kexec at ffffffff8578995d
 cloud-hypervisor#3 [ffffb1ba6707fb70] oops_end at ffffffff85623c0d
 cloud-hypervisor#4 [ffffb1ba6707fb90] no_context at ffffffff856692c9
 cloud-hypervisor#5 [ffffb1ba6707fbf8] exc_page_fault at ffffffff85f95b51
 cloud-hypervisor#6 [ffffb1ba6707fc50] asm_exc_page_fault at ffffffff86000ace
    [exception RIP: svm_update_pi_irte+227]
    RIP: ffffffffc0761b53  RSP: ffffb1ba6707fd08  RFLAGS: 00010086
    RAX: ffffb1ba6707fd78  RBX: ffffb1ba66d91000  RCX: 0000000000000001
    RDX: 00003c803f63f1c0  RSI: 000000000000019a  RDI: ffffb1ba66db2ab8
    RBP: 000000000000019a   R8: 0000000000000040   R9: ffff94ca41b82200
    R10: ffffffffffffffcf  R11: 0000000000000001  R12: 0000000000000001
    R13: 0000000000000001  R14: ffffffffffffffcf  R15: 000000000000005f
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 cloud-hypervisor#7 [ffffb1ba6707fdb8] kvm_irq_routing_update at ffffffffc09f19a1 [kvm]
 cloud-hypervisor#8 [ffffb1ba6707fde0] kvm_set_irq_routing at ffffffffc09f2133 [kvm]
 cloud-hypervisor#9 [ffffb1ba6707fe18] kvm_vm_ioctl at ffffffffc09ef544 [kvm]
    RIP: 00007f143c36488b  RSP: 00007f143a4e04b8  RFLAGS: 00000246
    RAX: ffffffffffffffda  RBX: 00007f05780041d0  RCX: 00007f143c36488b
    RDX: 00007f05780041d0  RSI: 000000004008ae6a  RDI: 0000000000000020
    RBP: 00000000000004e8   R8: 0000000000000008   R9: 00007f05780041e0
    R10: 00007f0578004560  R11: 0000000000000246  R12: 00000000000004e0
    R13: 000000000000001a  R14: 00007f1424001c60  R15: 00007f0578003bc0
    ORIG_RAX: 0000000000000010  CS: 0033  SS: 002b

To solve this problem, move route.disable() before set_gsi_routes() to
remove the gsi from irqfds.items first.

This problem only exists on AMD platform, 'cause on Intel platform
kernel just return when update irte while it only prints a warning on
AMD.

Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
@up2wing up2wing mentioned this issue Mar 9, 2022
Merged
up2wing pushed a commit to up2wing/cloud-hypervisor that referenced this issue Mar 10, 2022
vmm: interrupt: fix msi mask irq causing kernel panic on AMD
38c8743 
When mask a msi irq, we set the entry.masked to be true, so kvm
hypervisor will not pass the gsi to kernel through KVM_SET_GSI_ROUTING
ioctl which update kvm->irq_routing. This will trigger kernel
panic on AMD platform when the gsi is the largest one in kernel
kvm->irqfds.items:

crash> bt
PID: 22218  TASK: ffff951a6ad74980  CPU: 73  COMMAND: "vcpu8"
 #0 [ffffb1ba6707fa40] machine_kexec at ffffffff8565b397
 cloud-hypervisor#1 [ffffb1ba6707fa90] __crash_kexec at ffffffff85788a6d
 cloud-hypervisor#2 [ffffb1ba6707fb58] crash_kexec at ffffffff8578995d
 cloud-hypervisor#3 [ffffb1ba6707fb70] oops_end at ffffffff85623c0d
 cloud-hypervisor#4 [ffffb1ba6707fb90] no_context at ffffffff856692c9
 cloud-hypervisor#5 [ffffb1ba6707fbf8] exc_page_fault at ffffffff85f95b51
 cloud-hypervisor#6 [ffffb1ba6707fc50] asm_exc_page_fault at ffffffff86000ace
    [exception RIP: svm_update_pi_irte+227]
    RIP: ffffffffc0761b53  RSP: ffffb1ba6707fd08  RFLAGS: 00010086
    RAX: ffffb1ba6707fd78  RBX: ffffb1ba66d91000  RCX: 0000000000000001
    RDX: 00003c803f63f1c0  RSI: 000000000000019a  RDI: ffffb1ba66db2ab8
    RBP: 000000000000019a   R8: 0000000000000040   R9: ffff94ca41b82200
    R10: ffffffffffffffcf  R11: 0000000000000001  R12: 0000000000000001
    R13: 0000000000000001  R14: ffffffffffffffcf  R15: 000000000000005f
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 cloud-hypervisor#7 [ffffb1ba6707fdb8] kvm_irq_routing_update at ffffffffc09f19a1 [kvm]
 cloud-hypervisor#8 [ffffb1ba6707fde0] kvm_set_irq_routing at ffffffffc09f2133 [kvm]
 cloud-hypervisor#9 [ffffb1ba6707fe18] kvm_vm_ioctl at ffffffffc09ef544 [kvm]
    RIP: 00007f143c36488b  RSP: 00007f143a4e04b8  RFLAGS: 00000246
    RAX: ffffffffffffffda  RBX: 00007f05780041d0  RCX: 00007f143c36488b
    RDX: 00007f05780041d0  RSI: 000000004008ae6a  RDI: 0000000000000020
    RBP: 00000000000004e8   R8: 0000000000000008   R9: 00007f05780041e0
    R10: 00007f0578004560  R11: 0000000000000246  R12: 00000000000004e0
    R13: 000000000000001a  R14: 00007f1424001c60  R15: 00007f0578003bc0
    ORIG_RAX: 0000000000000010  CS: 0033  SS: 002b

To solve this problem, move route.disable() before set_gsi_routes() to
remove the gsi from irqfds.items first.

This problem only exists on AMD platform, 'cause on Intel platform
kernel just return when update irte while it only prints a warning on
AMD.

Also, this patch adjusts the order of enable() and set_gsi_routes() in
unmask(), which should do no harm.

Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
sboeuf pushed a commit that referenced this issue Mar 10, 2022
vmm: interrupt: fix msi mask irq causing kernel panic on AMD
db9e5e5 
When mask a msi irq, we set the entry.masked to be true, so kvm
hypervisor will not pass the gsi to kernel through KVM_SET_GSI_ROUTING
ioctl which update kvm->irq_routing. This will trigger kernel
panic on AMD platform when the gsi is the largest one in kernel
kvm->irqfds.items:

crash> bt
PID: 22218  TASK: ffff951a6ad74980  CPU: 73  COMMAND: "vcpu8"
 #0 [ffffb1ba6707fa40] machine_kexec at ffffffff8565b397
 #1 [ffffb1ba6707fa90] __crash_kexec at ffffffff85788a6d
 #2 [ffffb1ba6707fb58] crash_kexec at ffffffff8578995d
 #3 [ffffb1ba6707fb70] oops_end at ffffffff85623c0d
 #4 [ffffb1ba6707fb90] no_context at ffffffff856692c9
 #5 [ffffb1ba6707fbf8] exc_page_fault at ffffffff85f95b51
 #6 [ffffb1ba6707fc50] asm_exc_page_fault at ffffffff86000ace
    [exception RIP: svm_update_pi_irte+227]
    RIP: ffffffffc0761b53  RSP: ffffb1ba6707fd08  RFLAGS: 00010086
    RAX: ffffb1ba6707fd78  RBX: ffffb1ba66d91000  RCX: 0000000000000001
    RDX: 00003c803f63f1c0  RSI: 000000000000019a  RDI: ffffb1ba66db2ab8
    RBP: 000000000000019a   R8: 0000000000000040   R9: ffff94ca41b82200
    R10: ffffffffffffffcf  R11: 0000000000000001  R12: 0000000000000001
    R13: 0000000000000001  R14: ffffffffffffffcf  R15: 000000000000005f
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 #7 [ffffb1ba6707fdb8] kvm_irq_routing_update at ffffffffc09f19a1 [kvm]
 #8 [ffffb1ba6707fde0] kvm_set_irq_routing at ffffffffc09f2133 [kvm]
 #9 [ffffb1ba6707fe18] kvm_vm_ioctl at ffffffffc09ef544 [kvm]
    RIP: 00007f143c36488b  RSP: 00007f143a4e04b8  RFLAGS: 00000246
    RAX: ffffffffffffffda  RBX: 00007f05780041d0  RCX: 00007f143c36488b
    RDX: 00007f05780041d0  RSI: 000000004008ae6a  RDI: 0000000000000020
    RBP: 00000000000004e8   R8: 0000000000000008   R9: 00007f05780041e0
    R10: 00007f0578004560  R11: 0000000000000246  R12: 00000000000004e0
    R13: 000000000000001a  R14: 00007f1424001c60  R15: 00007f0578003bc0
    ORIG_RAX: 0000000000000010  CS: 0033  SS: 002b

To solve this problem, move route.disable() before set_gsi_routes() to
remove the gsi from irqfds.items first.

This problem only exists on AMD platform, 'cause on Intel platform
kernel just return when update irte while it only prints a warning on
AMD.

Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
sboeuf pushed a commit that referenced this issue Mar 10, 2022
vmm: interrupt: fix msi mask irq causing kernel panic on AMD
5375b84 
When mask a msi irq, we set the entry.masked to be true, so kvm
hypervisor will not pass the gsi to kernel through KVM_SET_GSI_ROUTING
ioctl which update kvm->irq_routing. This will trigger kernel
panic on AMD platform when the gsi is the largest one in kernel
kvm->irqfds.items:

crash> bt
PID: 22218  TASK: ffff951a6ad74980  CPU: 73  COMMAND: "vcpu8"
 #0 [ffffb1ba6707fa40] machine_kexec at ffffffff8565b397
 #1 [ffffb1ba6707fa90] __crash_kexec at ffffffff85788a6d
 #2 [ffffb1ba6707fb58] crash_kexec at ffffffff8578995d
 #3 [ffffb1ba6707fb70] oops_end at ffffffff85623c0d
 #4 [ffffb1ba6707fb90] no_context at ffffffff856692c9
 #5 [ffffb1ba6707fbf8] exc_page_fault at ffffffff85f95b51
 #6 [ffffb1ba6707fc50] asm_exc_page_fault at ffffffff86000ace
    [exception RIP: svm_update_pi_irte+227]
    RIP: ffffffffc0761b53  RSP: ffffb1ba6707fd08  RFLAGS: 00010086
    RAX: ffffb1ba6707fd78  RBX: ffffb1ba66d91000  RCX: 0000000000000001
    RDX: 00003c803f63f1c0  RSI: 000000000000019a  RDI: ffffb1ba66db2ab8
    RBP: 000000000000019a   R8: 0000000000000040   R9: ffff94ca41b82200
    R10: ffffffffffffffcf  R11: 0000000000000001  R12: 0000000000000001
    R13: 0000000000000001  R14: ffffffffffffffcf  R15: 000000000000005f
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 #7 [ffffb1ba6707fdb8] kvm_irq_routing_update at ffffffffc09f19a1 [kvm]
 #8 [ffffb1ba6707fde0] kvm_set_irq_routing at ffffffffc09f2133 [kvm]
 #9 [ffffb1ba6707fe18] kvm_vm_ioctl at ffffffffc09ef544 [kvm]
    RIP: 00007f143c36488b  RSP: 00007f143a4e04b8  RFLAGS: 00000246
    RAX: ffffffffffffffda  RBX: 00007f05780041d0  RCX: 00007f143c36488b
    RDX: 00007f05780041d0  RSI: 000000004008ae6a  RDI: 0000000000000020
    RBP: 00000000000004e8   R8: 0000000000000008   R9: 00007f05780041e0
    R10: 00007f0578004560  R11: 0000000000000246  R12: 00000000000004e0
    R13: 000000000000001a  R14: 00007f1424001c60  R15: 00007f0578003bc0
    ORIG_RAX: 0000000000000010  CS: 0033  SS: 002b

To solve this problem, move route.disable() before set_gsi_routes() to
remove the gsi from irqfds.items first.

This problem only exists on AMD platform, 'cause on Intel platform
kernel just return when update irte while it only prints a warning on
AMD.

Also, this patch adjusts the order of enable() and set_gsi_routes() in
unmask(), which should do no harm.

Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
rbradford pushed a commit that referenced this issue Mar 11, 2022
@rbradford
vmm: interrupt: fix msi mask irq causing kernel panic on AMD
4dfe1ff 
When mask a msi irq, we set the entry.masked to be true, so kvm
hypervisor will not pass the gsi to kernel through KVM_SET_GSI_ROUTING
ioctl which update kvm->irq_routing. This will trigger kernel
panic on AMD platform when the gsi is the largest one in kernel
kvm->irqfds.items:

crash> bt
PID: 22218  TASK: ffff951a6ad74980  CPU: 73  COMMAND: "vcpu8"
 #0 [ffffb1ba6707fa40] machine_kexec at ffffffff8565b397
 #1 [ffffb1ba6707fa90] __crash_kexec at ffffffff85788a6d
 #2 [ffffb1ba6707fb58] crash_kexec at ffffffff8578995d
 #3 [ffffb1ba6707fb70] oops_end at ffffffff85623c0d
 #4 [ffffb1ba6707fb90] no_context at ffffffff856692c9
 #5 [ffffb1ba6707fbf8] exc_page_fault at ffffffff85f95b51
 #6 [ffffb1ba6707fc50] asm_exc_page_fault at ffffffff86000ace
    [exception RIP: svm_update_pi_irte+227]
    RIP: ffffffffc0761b53  RSP: ffffb1ba6707fd08  RFLAGS: 00010086
    RAX: ffffb1ba6707fd78  RBX: ffffb1ba66d91000  RCX: 0000000000000001
    RDX: 00003c803f63f1c0  RSI: 000000000000019a  RDI: ffffb1ba66db2ab8
    RBP: 000000000000019a   R8: 0000000000000040   R9: ffff94ca41b82200
    R10: ffffffffffffffcf  R11: 0000000000000001  R12: 0000000000000001
    R13: 0000000000000001  R14: ffffffffffffffcf  R15: 000000000000005f
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 #7 [ffffb1ba6707fdb8] kvm_irq_routing_update at ffffffffc09f19a1 [kvm]
 #8 [ffffb1ba6707fde0] kvm_set_irq_routing at ffffffffc09f2133 [kvm]
 #9 [ffffb1ba6707fe18] kvm_vm_ioctl at ffffffffc09ef544 [kvm]
    RIP: 00007f143c36488b  RSP: 00007f143a4e04b8  RFLAGS: 00000246
    RAX: ffffffffffffffda  RBX: 00007f05780041d0  RCX: 00007f143c36488b
    RDX: 00007f05780041d0  RSI: 000000004008ae6a  RDI: 0000000000000020
    RBP: 00000000000004e8   R8: 0000000000000008   R9: 00007f05780041e0
    R10: 00007f0578004560  R11: 0000000000000246  R12: 00000000000004e0
    R13: 000000000000001a  R14: 00007f1424001c60  R15: 00007f0578003bc0
    ORIG_RAX: 0000000000000010  CS: 0033  SS: 002b

To solve this problem, move route.disable() before set_gsi_routes() to
remove the gsi from irqfds.items first.

This problem only exists on AMD platform, 'cause on Intel platform
kernel just return when update irte while it only prints a warning on
AMD.

Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
(cherry picked from commit db9e5e5)
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
rbradford pushed a commit that referenced this issue Mar 11, 2022
@rbradford
vmm: interrupt: fix msi mask irq causing kernel panic on AMD
6275d4f 
When mask a msi irq, we set the entry.masked to be true, so kvm
hypervisor will not pass the gsi to kernel through KVM_SET_GSI_ROUTING
ioctl which update kvm->irq_routing. This will trigger kernel
panic on AMD platform when the gsi is the largest one in kernel
kvm->irqfds.items:

crash> bt
PID: 22218  TASK: ffff951a6ad74980  CPU: 73  COMMAND: "vcpu8"
 #0 [ffffb1ba6707fa40] machine_kexec at ffffffff8565b397
 #1 [ffffb1ba6707fa90] __crash_kexec at ffffffff85788a6d
 #2 [ffffb1ba6707fb58] crash_kexec at ffffffff8578995d
 #3 [ffffb1ba6707fb70] oops_end at ffffffff85623c0d
 #4 [ffffb1ba6707fb90] no_context at ffffffff856692c9
 #5 [ffffb1ba6707fbf8] exc_page_fault at ffffffff85f95b51
 #6 [ffffb1ba6707fc50] asm_exc_page_fault at ffffffff86000ace
    [exception RIP: svm_update_pi_irte+227]
    RIP: ffffffffc0761b53  RSP: ffffb1ba6707fd08  RFLAGS: 00010086
    RAX: ffffb1ba6707fd78  RBX: ffffb1ba66d91000  RCX: 0000000000000001
    RDX: 00003c803f63f1c0  RSI: 000000000000019a  RDI: ffffb1ba66db2ab8
    RBP: 000000000000019a   R8: 0000000000000040   R9: ffff94ca41b82200
    R10: ffffffffffffffcf  R11: 0000000000000001  R12: 0000000000000001
    R13: 0000000000000001  R14: ffffffffffffffcf  R15: 000000000000005f
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 #7 [ffffb1ba6707fdb8] kvm_irq_routing_update at ffffffffc09f19a1 [kvm]
 #8 [ffffb1ba6707fde0] kvm_set_irq_routing at ffffffffc09f2133 [kvm]
 #9 [ffffb1ba6707fe18] kvm_vm_ioctl at ffffffffc09ef544 [kvm]
    RIP: 00007f143c36488b  RSP: 00007f143a4e04b8  RFLAGS: 00000246
    RAX: ffffffffffffffda  RBX: 00007f05780041d0  RCX: 00007f143c36488b
    RDX: 00007f05780041d0  RSI: 000000004008ae6a  RDI: 0000000000000020
    RBP: 00000000000004e8   R8: 0000000000000008   R9: 00007f05780041e0
    R10: 00007f0578004560  R11: 0000000000000246  R12: 00000000000004e0
    R13: 000000000000001a  R14: 00007f1424001c60  R15: 00007f0578003bc0
    ORIG_RAX: 0000000000000010  CS: 0033  SS: 002b

To solve this problem, move route.disable() before set_gsi_routes() to
remove the gsi from irqfds.items first.

This problem only exists on AMD platform, 'cause on Intel platform
kernel just return when update irte while it only prints a warning on
AMD.

Also, this patch adjusts the order of enable() and set_gsi_routes() in
unmask(), which should do no harm.

Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
(cherry picked from commit 5375b84)
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
rbradford pushed a commit that referenced this issue Mar 11, 2022
@rbradford
vmm: interrupt: fix msi mask irq causing kernel panic on AMD
9e2d051 
When mask a msi irq, we set the entry.masked to be true, so kvm
hypervisor will not pass the gsi to kernel through KVM_SET_GSI_ROUTING
ioctl which update kvm->irq_routing. This will trigger kernel
panic on AMD platform when the gsi is the largest one in kernel
kvm->irqfds.items:

crash> bt
PID: 22218  TASK: ffff951a6ad74980  CPU: 73  COMMAND: "vcpu8"
 #0 [ffffb1ba6707fa40] machine_kexec at ffffffff8565b397
 #1 [ffffb1ba6707fa90] __crash_kexec at ffffffff85788a6d
 #2 [ffffb1ba6707fb58] crash_kexec at ffffffff8578995d
 #3 [ffffb1ba6707fb70] oops_end at ffffffff85623c0d
 #4 [ffffb1ba6707fb90] no_context at ffffffff856692c9
 #5 [ffffb1ba6707fbf8] exc_page_fault at ffffffff85f95b51
 #6 [ffffb1ba6707fc50] asm_exc_page_fault at ffffffff86000ace
    [exception RIP: svm_update_pi_irte+227]
    RIP: ffffffffc0761b53  RSP: ffffb1ba6707fd08  RFLAGS: 00010086
    RAX: ffffb1ba6707fd78  RBX: ffffb1ba66d91000  RCX: 0000000000000001
    RDX: 00003c803f63f1c0  RSI: 000000000000019a  RDI: ffffb1ba66db2ab8
    RBP: 000000000000019a   R8: 0000000000000040   R9: ffff94ca41b82200
    R10: ffffffffffffffcf  R11: 0000000000000001  R12: 0000000000000001
    R13: 0000000000000001  R14: ffffffffffffffcf  R15: 000000000000005f
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 #7 [ffffb1ba6707fdb8] kvm_irq_routing_update at ffffffffc09f19a1 [kvm]
 #8 [ffffb1ba6707fde0] kvm_set_irq_routing at ffffffffc09f2133 [kvm]
 #9 [ffffb1ba6707fe18] kvm_vm_ioctl at ffffffffc09ef544 [kvm]
    RIP: 00007f143c36488b  RSP: 00007f143a4e04b8  RFLAGS: 00000246
    RAX: ffffffffffffffda  RBX: 00007f05780041d0  RCX: 00007f143c36488b
    RDX: 00007f05780041d0  RSI: 000000004008ae6a  RDI: 0000000000000020
    RBP: 00000000000004e8   R8: 0000000000000008   R9: 00007f05780041e0
    R10: 00007f0578004560  R11: 0000000000000246  R12: 00000000000004e0
    R13: 000000000000001a  R14: 00007f1424001c60  R15: 00007f0578003bc0
    ORIG_RAX: 0000000000000010  CS: 0033  SS: 002b

To solve this problem, move route.disable() before set_gsi_routes() to
remove the gsi from irqfds.items first.

This problem only exists on AMD platform, 'cause on Intel platform
kernel just return when update irte while it only prints a warning on
AMD.

Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
(cherry picked from commit db9e5e5)
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
rbradford pushed a commit that referenced this issue Mar 11, 2022
@rbradford
vmm: interrupt: fix msi mask irq causing kernel panic on AMD
e0fda7b 
When mask a msi irq, we set the entry.masked to be true, so kvm
hypervisor will not pass the gsi to kernel through KVM_SET_GSI_ROUTING
ioctl which update kvm->irq_routing. This will trigger kernel
panic on AMD platform when the gsi is the largest one in kernel
kvm->irqfds.items:

crash> bt
PID: 22218  TASK: ffff951a6ad74980  CPU: 73  COMMAND: "vcpu8"
 #0 [ffffb1ba6707fa40] machine_kexec at ffffffff8565b397
 #1 [ffffb1ba6707fa90] __crash_kexec at ffffffff85788a6d
 #2 [ffffb1ba6707fb58] crash_kexec at ffffffff8578995d
 #3 [ffffb1ba6707fb70] oops_end at ffffffff85623c0d
 #4 [ffffb1ba6707fb90] no_context at ffffffff856692c9
 #5 [ffffb1ba6707fbf8] exc_page_fault at ffffffff85f95b51
 #6 [ffffb1ba6707fc50] asm_exc_page_fault at ffffffff86000ace
    [exception RIP: svm_update_pi_irte+227]
    RIP: ffffffffc0761b53  RSP: ffffb1ba6707fd08  RFLAGS: 00010086
    RAX: ffffb1ba6707fd78  RBX: ffffb1ba66d91000  RCX: 0000000000000001
    RDX: 00003c803f63f1c0  RSI: 000000000000019a  RDI: ffffb1ba66db2ab8
    RBP: 000000000000019a   R8: 0000000000000040   R9: ffff94ca41b82200
    R10: ffffffffffffffcf  R11: 0000000000000001  R12: 0000000000000001
    R13: 0000000000000001  R14: ffffffffffffffcf  R15: 000000000000005f
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 #7 [ffffb1ba6707fdb8] kvm_irq_routing_update at ffffffffc09f19a1 [kvm]
 #8 [ffffb1ba6707fde0] kvm_set_irq_routing at ffffffffc09f2133 [kvm]
 #9 [ffffb1ba6707fe18] kvm_vm_ioctl at ffffffffc09ef544 [kvm]
    RIP: 00007f143c36488b  RSP: 00007f143a4e04b8  RFLAGS: 00000246
    RAX: ffffffffffffffda  RBX: 00007f05780041d0  RCX: 00007f143c36488b
    RDX: 00007f05780041d0  RSI: 000000004008ae6a  RDI: 0000000000000020
    RBP: 00000000000004e8   R8: 0000000000000008   R9: 00007f05780041e0
    R10: 00007f0578004560  R11: 0000000000000246  R12: 00000000000004e0
    R13: 000000000000001a  R14: 00007f1424001c60  R15: 00007f0578003bc0
    ORIG_RAX: 0000000000000010  CS: 0033  SS: 002b

To solve this problem, move route.disable() before set_gsi_routes() to
remove the gsi from irqfds.items first.

This problem only exists on AMD platform, 'cause on Intel platform
kernel just return when update irte while it only prints a warning on
AMD.

Also, this patch adjusts the order of enable() and set_gsi_routes() in
unmask(), which should do no harm.

Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
(cherry picked from commit 5375b84)
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
@MarkGass MarkGass mentioned this issue Jan 8, 2023
Closed
@up2wing up2wing mentioned this issue Apr 3, 2024
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Add support for multiple disks rbradford/cloud-hypervisor
1 participant
@rbradford