* [PATCH RESEND rdma-rc] IB/core: Fix a nested dead lock as part of ODP flow
@ 2022-08-24 6:10 Leon Romanovsky
2022-08-29 8:15 ` Leon Romanovsky
0 siblings, 1 reply; 3+ messages in thread
From: Leon Romanovsky @ 2022-08-24 6:10 UTC (permalink / raw)
To: Jason Gunthorpe, Andrew Morton
Cc: Yishai Hadas, linux-rdma, Maor Gottlieb, linux-mm, linux-kernel,
Thomas Gleixner
From: Yishai Hadas <yishaih@nvidia.com>
Fix a nested dead lock as part of ODP flow by using mmput_async().
From the below call trace [1] can see that calling mmput() once we have
the umem_odp->umem_mutex locked as required by
ib_umem_odp_map_dma_and_lock() might trigger in the same task the
exit_mmap()->__mmu_notifier_release()->mlx5_ib_invalidate_range() which
may dead lock when trying to lock the same mutex.
Moving to use mmput_async() will solve the problem as the above
exit_mmap() flow will be called in other task and will be executed once
the lock will be available.
[1]
[64843.077665] task:kworker/u133:2 state:D stack: 0 pid:80906 ppid:
2 flags:0x00004000
[64843.077672] Workqueue: mlx5_ib_page_fault mlx5_ib_eqe_pf_action [mlx5_ib]
[64843.077719] Call Trace:
[64843.077722] <TASK>
[64843.077724] __schedule+0x23d/0x590
[64843.077729] schedule+0x4e/0xb0
[64843.077735] schedule_preempt_disabled+0xe/0x10
[64843.077740] __mutex_lock.constprop.0+0x263/0x490
[64843.077747] __mutex_lock_slowpath+0x13/0x20
[64843.077752] mutex_lock+0x34/0x40
[64843.077758] mlx5_ib_invalidate_range+0x48/0x270 [mlx5_ib]
[64843.077808] __mmu_notifier_release+0x1a4/0x200
[64843.077816] exit_mmap+0x1bc/0x200
[64843.077822] ? walk_page_range+0x9c/0x120
[64843.077828] ? __cond_resched+0x1a/0x50
[64843.077833] ? mutex_lock+0x13/0x40
[64843.077839] ? uprobe_clear_state+0xac/0x120
[64843.077860] mmput+0x5f/0x140
[64843.077867] ib_umem_odp_map_dma_and_lock+0x21b/0x580 [ib_core]
[64843.077931] pagefault_real_mr+0x9a/0x140 [mlx5_ib]
[64843.077962] pagefault_mr+0xb4/0x550 [mlx5_ib]
[64843.077992] pagefault_single_data_segment.constprop.0+0x2ac/0x560
[mlx5_ib]
[64843.078022] mlx5_ib_eqe_pf_action+0x528/0x780 [mlx5_ib]
[64843.078051] process_one_work+0x22b/0x3d0
[64843.078059] worker_thread+0x53/0x410
[64843.078065] ? process_one_work+0x3d0/0x3d0
[64843.078073] kthread+0x12a/0x150
[64843.078079] ? set_kthread_struct+0x50/0x50
[64843.078085] ret_from_fork+0x22/0x30
[64843.078093] </TASK>
Fixes: 36f30e486dce ("IB/core: Improve ODP to use hmm_range_fault()")
Reviewed-by: Maor Gottlieb <maorg@nvidia.com>
Signed-off-by: Yishai Hadas <yishaih@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
---
Resend to larger forum.
https://lore.kernel.org/all/74d93541ea533ef7daec6f126deb1072500aeb16.1661251841.git.leonro@nvidia.com
---
drivers/infiniband/core/umem_odp.c | 2 +-
kernel/fork.c | 1 +
2 files changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/infiniband/core/umem_odp.c b/drivers/infiniband/core/umem_odp.c
index 186ed8859920..d39e16c211e8 100644
--- a/drivers/infiniband/core/umem_odp.c
+++ b/drivers/infiniband/core/umem_odp.c
@@ -462,7 +462,7 @@ int ib_umem_odp_map_dma_and_lock(struct ib_umem_odp *umem_odp, u64 user_virt,
mutex_unlock(&umem_odp->umem_mutex);
out_put_mm:
- mmput(owning_mm);
+ mmput_async(owning_mm);
out_put_task:
if (owning_process)
put_task_struct(owning_process);
diff --git a/kernel/fork.c b/kernel/fork.c
index 90c85b17bf69..8a9e92068b15 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -1225,6 +1225,7 @@ void mmput_async(struct mm_struct *mm)
schedule_work(&mm->async_put_work);
}
}
+EXPORT_SYMBOL_GPL(mmput_async);
#endif
/**
--
2.37.2
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH RESEND rdma-rc] IB/core: Fix a nested dead lock as part of ODP flow
2022-08-24 6:10 [PATCH RESEND rdma-rc] IB/core: Fix a nested dead lock as part of ODP flow Leon Romanovsky
@ 2022-08-29 8:15 ` Leon Romanovsky
2022-09-05 11:48 ` Leon Romanovsky
0 siblings, 1 reply; 3+ messages in thread
From: Leon Romanovsky @ 2022-08-29 8:15 UTC (permalink / raw)
To: Jason Gunthorpe, Andrew Morton, linux-mm
Cc: Yishai Hadas, linux-rdma, Maor Gottlieb, linux-kernel, Thomas Gleixner
On Wed, Aug 24, 2022 at 09:10:36AM +0300, Leon Romanovsky wrote:
> From: Yishai Hadas <yishaih@nvidia.com>
>
> Fix a nested dead lock as part of ODP flow by using mmput_async().
>
> From the below call trace [1] can see that calling mmput() once we have
> the umem_odp->umem_mutex locked as required by
> ib_umem_odp_map_dma_and_lock() might trigger in the same task the
> exit_mmap()->__mmu_notifier_release()->mlx5_ib_invalidate_range() which
> may dead lock when trying to lock the same mutex.
>
> Moving to use mmput_async() will solve the problem as the above
> exit_mmap() flow will be called in other task and will be executed once
> the lock will be available.
>
> [1]
> [64843.077665] task:kworker/u133:2 state:D stack: 0 pid:80906 ppid:
> 2 flags:0x00004000
> [64843.077672] Workqueue: mlx5_ib_page_fault mlx5_ib_eqe_pf_action [mlx5_ib]
> [64843.077719] Call Trace:
> [64843.077722] <TASK>
> [64843.077724] __schedule+0x23d/0x590
> [64843.077729] schedule+0x4e/0xb0
> [64843.077735] schedule_preempt_disabled+0xe/0x10
> [64843.077740] __mutex_lock.constprop.0+0x263/0x490
> [64843.077747] __mutex_lock_slowpath+0x13/0x20
> [64843.077752] mutex_lock+0x34/0x40
> [64843.077758] mlx5_ib_invalidate_range+0x48/0x270 [mlx5_ib]
> [64843.077808] __mmu_notifier_release+0x1a4/0x200
> [64843.077816] exit_mmap+0x1bc/0x200
> [64843.077822] ? walk_page_range+0x9c/0x120
> [64843.077828] ? __cond_resched+0x1a/0x50
> [64843.077833] ? mutex_lock+0x13/0x40
> [64843.077839] ? uprobe_clear_state+0xac/0x120
> [64843.077860] mmput+0x5f/0x140
> [64843.077867] ib_umem_odp_map_dma_and_lock+0x21b/0x580 [ib_core]
> [64843.077931] pagefault_real_mr+0x9a/0x140 [mlx5_ib]
> [64843.077962] pagefault_mr+0xb4/0x550 [mlx5_ib]
> [64843.077992] pagefault_single_data_segment.constprop.0+0x2ac/0x560
> [mlx5_ib]
> [64843.078022] mlx5_ib_eqe_pf_action+0x528/0x780 [mlx5_ib]
> [64843.078051] process_one_work+0x22b/0x3d0
> [64843.078059] worker_thread+0x53/0x410
> [64843.078065] ? process_one_work+0x3d0/0x3d0
> [64843.078073] kthread+0x12a/0x150
> [64843.078079] ? set_kthread_struct+0x50/0x50
> [64843.078085] ret_from_fork+0x22/0x30
> [64843.078093] </TASK>
>
> Fixes: 36f30e486dce ("IB/core: Improve ODP to use hmm_range_fault()")
> Reviewed-by: Maor Gottlieb <maorg@nvidia.com>
> Signed-off-by: Yishai Hadas <yishaih@nvidia.com>
> Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
> ---
> Resend to larger forum.
> https://lore.kernel.org/all/74d93541ea533ef7daec6f126deb1072500aeb16.1661251841.git.leonro@nvidia.com
> ---
> drivers/infiniband/core/umem_odp.c | 2 +-
> kernel/fork.c | 1 +
> 2 files changed, 2 insertions(+), 1 deletion(-)
Any objections?
Thanks
>
> diff --git a/drivers/infiniband/core/umem_odp.c b/drivers/infiniband/core/umem_odp.c
> index 186ed8859920..d39e16c211e8 100644
> --- a/drivers/infiniband/core/umem_odp.c
> +++ b/drivers/infiniband/core/umem_odp.c
> @@ -462,7 +462,7 @@ int ib_umem_odp_map_dma_and_lock(struct ib_umem_odp *umem_odp, u64 user_virt,
> mutex_unlock(&umem_odp->umem_mutex);
>
> out_put_mm:
> - mmput(owning_mm);
> + mmput_async(owning_mm);
> out_put_task:
> if (owning_process)
> put_task_struct(owning_process);
> diff --git a/kernel/fork.c b/kernel/fork.c
> index 90c85b17bf69..8a9e92068b15 100644
> --- a/kernel/fork.c
> +++ b/kernel/fork.c
> @@ -1225,6 +1225,7 @@ void mmput_async(struct mm_struct *mm)
> schedule_work(&mm->async_put_work);
> }
> }
> +EXPORT_SYMBOL_GPL(mmput_async);
> #endif
>
> /**
> --
> 2.37.2
>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH RESEND rdma-rc] IB/core: Fix a nested dead lock as part of ODP flow
2022-08-29 8:15 ` Leon Romanovsky
@ 2022-09-05 11:48 ` Leon Romanovsky
0 siblings, 0 replies; 3+ messages in thread
From: Leon Romanovsky @ 2022-09-05 11:48 UTC (permalink / raw)
To: Jason Gunthorpe, Andrew Morton, linux-mm
Cc: Yishai Hadas, linux-rdma, Maor Gottlieb, linux-kernel, Thomas Gleixner
On Mon, Aug 29, 2022 at 11:15:56AM +0300, Leon Romanovsky wrote:
> On Wed, Aug 24, 2022 at 09:10:36AM +0300, Leon Romanovsky wrote:
> > From: Yishai Hadas <yishaih@nvidia.com>
> >
> > Fix a nested dead lock as part of ODP flow by using mmput_async().
> >
> > From the below call trace [1] can see that calling mmput() once we have
> > the umem_odp->umem_mutex locked as required by
> > ib_umem_odp_map_dma_and_lock() might trigger in the same task the
> > exit_mmap()->__mmu_notifier_release()->mlx5_ib_invalidate_range() which
> > may dead lock when trying to lock the same mutex.
> >
> > Moving to use mmput_async() will solve the problem as the above
> > exit_mmap() flow will be called in other task and will be executed once
> > the lock will be available.
> >
> > [1]
> > [64843.077665] task:kworker/u133:2 state:D stack: 0 pid:80906 ppid:
> > 2 flags:0x00004000
> > [64843.077672] Workqueue: mlx5_ib_page_fault mlx5_ib_eqe_pf_action [mlx5_ib]
> > [64843.077719] Call Trace:
> > [64843.077722] <TASK>
> > [64843.077724] __schedule+0x23d/0x590
> > [64843.077729] schedule+0x4e/0xb0
> > [64843.077735] schedule_preempt_disabled+0xe/0x10
> > [64843.077740] __mutex_lock.constprop.0+0x263/0x490
> > [64843.077747] __mutex_lock_slowpath+0x13/0x20
> > [64843.077752] mutex_lock+0x34/0x40
> > [64843.077758] mlx5_ib_invalidate_range+0x48/0x270 [mlx5_ib]
> > [64843.077808] __mmu_notifier_release+0x1a4/0x200
> > [64843.077816] exit_mmap+0x1bc/0x200
> > [64843.077822] ? walk_page_range+0x9c/0x120
> > [64843.077828] ? __cond_resched+0x1a/0x50
> > [64843.077833] ? mutex_lock+0x13/0x40
> > [64843.077839] ? uprobe_clear_state+0xac/0x120
> > [64843.077860] mmput+0x5f/0x140
> > [64843.077867] ib_umem_odp_map_dma_and_lock+0x21b/0x580 [ib_core]
> > [64843.077931] pagefault_real_mr+0x9a/0x140 [mlx5_ib]
> > [64843.077962] pagefault_mr+0xb4/0x550 [mlx5_ib]
> > [64843.077992] pagefault_single_data_segment.constprop.0+0x2ac/0x560
> > [mlx5_ib]
> > [64843.078022] mlx5_ib_eqe_pf_action+0x528/0x780 [mlx5_ib]
> > [64843.078051] process_one_work+0x22b/0x3d0
> > [64843.078059] worker_thread+0x53/0x410
> > [64843.078065] ? process_one_work+0x3d0/0x3d0
> > [64843.078073] kthread+0x12a/0x150
> > [64843.078079] ? set_kthread_struct+0x50/0x50
> > [64843.078085] ret_from_fork+0x22/0x30
> > [64843.078093] </TASK>
> >
> > Fixes: 36f30e486dce ("IB/core: Improve ODP to use hmm_range_fault()")
> > Reviewed-by: Maor Gottlieb <maorg@nvidia.com>
> > Signed-off-by: Yishai Hadas <yishaih@nvidia.com>
> > Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
> > ---
> > Resend to larger forum.
> > https://lore.kernel.org/all/74d93541ea533ef7daec6f126deb1072500aeb16.1661251841.git.leonro@nvidia.com
> > ---
> > drivers/infiniband/core/umem_odp.c | 2 +-
> > kernel/fork.c | 1 +
> > 2 files changed, 2 insertions(+), 1 deletion(-)
>
> Any objections?
I didn't hear any.
Applied to rdma-rc.
Thanks
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2022-09-05 11:48 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-08-24 6:10 [PATCH RESEND rdma-rc] IB/core: Fix a nested dead lock as part of ODP flow Leon Romanovsky
2022-08-29 8:15 ` Leon Romanovsky
2022-09-05 11:48 ` Leon Romanovsky
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).