Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-16 Thread James Simmons
> > > Anyway, I understand that Intel has been ignoring kernel.org instead of > > > sending forwarding their patches properly so you're doing a difficult > > > and thankless job... Thanks for that. I'm sure it's frustrating to > > > look at these patches for you as well. > > > > Thank you for

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-16 Thread Joe Perches
On Wed, 2018-05-16 at 09:12 +, Dilger, Andreas wrote: > On May 16, 2018, at 02:00, Dan Carpenter wrote: > > > > On Tue, May 15, 2018 at 04:02:55PM +0100, James Simmons wrote: > > > > > > > > /* > > > > >* Allocate new object. This may result in rather

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-16 Thread Greg Kroah-Hartman
On Tue, May 15, 2018 at 04:02:55PM +0100, James Simmons wrote: > > > /* > > >* Allocate new object. This may result in rather complicated > > >* operations, including fld queries, inode loading, etc. > > >*/ > > > o = lu_object_alloc(env, dev, f, conf); > > > - if (IS_ERR(o)) > > >

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-16 Thread Dilger, Andreas
On May 16, 2018, at 02:00, Dan Carpenter wrote: > > On Tue, May 15, 2018 at 04:02:55PM +0100, James Simmons wrote: >> /* * Allocate new object. This may result in rather complicated * operations, including fld queries, inode loading, etc.

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-16 Thread Dan Carpenter
On Tue, May 15, 2018 at 04:02:55PM +0100, James Simmons wrote: > > > > /* > > >* Allocate new object. This may result in rather complicated > > >* operations, including fld queries, inode loading, etc. > > >*/ > > > o = lu_object_alloc(env, dev, f, conf); > > > - if (IS_ERR(o)) >

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-15 Thread James Simmons
> > /* > > * Allocate new object. This may result in rather complicated > > * operations, including fld queries, inode loading, etc. > > */ > > o = lu_object_alloc(env, dev, f, conf); > > - if (IS_ERR(o)) > > + if (unlikely(IS_ERR(o))) > > return o; > > >

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-14 Thread James Simmons
> >> On Wed, May 02 2018, James Simmons wrote: > >> > >> > From: Lai Siyao > >> > > >> > Currently we set LU_OBJECT_HEARD_BANSHEE on object when we want > >> > to remove object from cache, but this may lead to deadlock, because > >> > when other process lookup such object,

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-14 Thread NeilBrown
On Tue, May 15 2018, James Simmons wrote: >> On Wed, May 02 2018, James Simmons wrote: >> >> > From: Lai Siyao >> > >> > Currently we set LU_OBJECT_HEARD_BANSHEE on object when we want >> > to remove object from cache, but this may lead to deadlock, because >> > when other

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-14 Thread James Simmons
> On Wed, May 02 2018, James Simmons wrote: > > > From: Lai Siyao > > > > Currently we set LU_OBJECT_HEARD_BANSHEE on object when we want > > to remove object from cache, but this may lead to deadlock, because > > when other process lookup such object, it needs to wait for

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-08 Thread Dan Carpenter
> /* >* Allocate new object. This may result in rather complicated >* operations, including fld queries, inode loading, etc. >*/ > o = lu_object_alloc(env, dev, f, conf); > - if (IS_ERR(o)) > + if (unlikely(IS_ERR(o))) > return o; > This

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-07 Thread Greg Kroah-Hartman
On Wed, May 02, 2018 at 02:21:48PM -0400, James Simmons wrote: > From: Lai Siyao > > Currently we set LU_OBJECT_HEARD_BANSHEE on object when we want > to remove object from cache, but this may lead to deadlock, because > when other process lookup such object, it needs to

Re: [PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-03 Thread NeilBrown
On Wed, May 02 2018, James Simmons wrote: > From: Lai Siyao > > Currently we set LU_OBJECT_HEARD_BANSHEE on object when we want > to remove object from cache, but this may lead to deadlock, because > when other process lookup such object, it needs to wait for this > object

[PATCH 4/4] staging: lustre: obdclass: change object lookup to no wait mode

2018-05-02 Thread James Simmons
From: Lai Siyao Currently we set LU_OBJECT_HEARD_BANSHEE on object when we want to remove object from cache, but this may lead to deadlock, because when other process lookup such object, it needs to wait for this object until release (done at last refcount put), while that