Re: waitqueue lockdep annotation V3

2018-07-17 Thread Christoph Hellwig
On Tue, Jul 17, 2018 at 05:04:39PM +0200, Peter Zijlstra wrote: > On Tue, Jul 17, 2018 at 04:24:37PM +0200, Christoph Hellwig wrote: > > On Thu, Jul 12, 2018 at 12:17:53PM -0700, Davidlohr Bueso wrote: > > > On Thu, 14 Dec 2017, Christoph Hellwig wrote: > > > > > >> Hi all, > > >> > > >> this

Re: waitqueue lockdep annotation V3

2018-07-17 Thread Christoph Hellwig
On Tue, Jul 17, 2018 at 05:04:39PM +0200, Peter Zijlstra wrote: > On Tue, Jul 17, 2018 at 04:24:37PM +0200, Christoph Hellwig wrote: > > On Thu, Jul 12, 2018 at 12:17:53PM -0700, Davidlohr Bueso wrote: > > > On Thu, 14 Dec 2017, Christoph Hellwig wrote: > > > > > >> Hi all, > > >> > > >> this

Re: waitqueue lockdep annotation V3

2018-07-17 Thread Peter Zijlstra
On Tue, Jul 17, 2018 at 04:24:37PM +0200, Christoph Hellwig wrote: > On Thu, Jul 12, 2018 at 12:17:53PM -0700, Davidlohr Bueso wrote: > > On Thu, 14 Dec 2017, Christoph Hellwig wrote: > > > >> Hi all, > >> > >> this series adds a strategic lockdep_assert_held to __wake_up_common > >> to ensure

Re: waitqueue lockdep annotation V3

2018-07-17 Thread Peter Zijlstra
On Tue, Jul 17, 2018 at 04:24:37PM +0200, Christoph Hellwig wrote: > On Thu, Jul 12, 2018 at 12:17:53PM -0700, Davidlohr Bueso wrote: > > On Thu, 14 Dec 2017, Christoph Hellwig wrote: > > > >> Hi all, > >> > >> this series adds a strategic lockdep_assert_held to __wake_up_common > >> to ensure

Re: waitqueue lockdep annotation V3

2018-07-17 Thread Christoph Hellwig
On Thu, Jul 12, 2018 at 12:17:53PM -0700, Davidlohr Bueso wrote: > On Thu, 14 Dec 2017, Christoph Hellwig wrote: > >> Hi all, >> >> this series adds a strategic lockdep_assert_held to __wake_up_common >> to ensure callers really do hold the wait_queue_head lock when calling >> the unlocked wake_up

Re: waitqueue lockdep annotation V3

2018-07-17 Thread Christoph Hellwig
On Thu, Jul 12, 2018 at 12:17:53PM -0700, Davidlohr Bueso wrote: > On Thu, 14 Dec 2017, Christoph Hellwig wrote: > >> Hi all, >> >> this series adds a strategic lockdep_assert_held to __wake_up_common >> to ensure callers really do hold the wait_queue_head lock when calling >> the unlocked wake_up

Re: waitqueue lockdep annotation V3

2018-07-12 Thread Davidlohr Bueso
On Thu, 14 Dec 2017, Christoph Hellwig wrote: Hi all, this series adds a strategic lockdep_assert_held to __wake_up_common to ensure callers really do hold the wait_queue_head lock when calling the unlocked wake_up variants. It turns out epoll did not do this for a fairly common path (hit all

Re: waitqueue lockdep annotation V3

2018-07-12 Thread Davidlohr Bueso
On Thu, 14 Dec 2017, Christoph Hellwig wrote: Hi all, this series adds a strategic lockdep_assert_held to __wake_up_common to ensure callers really do hold the wait_queue_head lock when calling the unlocked wake_up variants. It turns out epoll did not do this for a fairly common path (hit all

waitqueue lockdep annotation V3

2017-12-14 Thread Christoph Hellwig
Hi all, this series adds a strategic lockdep_assert_held to __wake_up_common to ensure callers really do hold the wait_queue_head lock when calling the unlocked wake_up variants. It turns out epoll did not do this for a fairly common path (hit all the time by systemd during bootup), so the

waitqueue lockdep annotation V3

2017-12-14 Thread Christoph Hellwig
Hi all, this series adds a strategic lockdep_assert_held to __wake_up_common to ensure callers really do hold the wait_queue_head lock when calling the unlocked wake_up variants. It turns out epoll did not do this for a fairly common path (hit all the time by systemd during bootup), so the