[PATCH 4.4 030/131] Hang/soft lockup in d_invalidate with simultaneous calls

4.4-stable review patch. If anyone has any objections, please let me know. -- From: Al Viro commit 81be24d263dbeddaba35827036d6f6787a59c2c3 upstream. It's not hard to trigger a bunch of d_invalidate() on the same dentry in parallel. They end up fighting each other - any

[PATCH 4.9 21/30] Hang/soft lockup in d_invalidate with simultaneous calls

4.9-stable review patch. If anyone has any objections, please let me know. -- From: Al Viro commit 81be24d263dbeddaba35827036d6f6787a59c2c3 upstream. It's not hard to trigger a bunch of d_invalidate() on the same dentry in parallel. They end up fighting each other - any

Re: [BACKPORT 4.4.y 20/25] Hang/soft lockup in d_invalidate with simultaneous calls

On Fri, Mar 22, 2019 at 04:44:11PM +0100, Arnd Bergmann wrote: > From: Al Viro > > It's not hard to trigger a bunch of d_invalidate() on the same > dentry in parallel. They end up fighting each other - any > dentry picked for removal by one will be skipped by the rest > and we'll go for the

[BACKPORT 4.4.y 20/25] Hang/soft lockup in d_invalidate with simultaneous calls

From: Al Viro It's not hard to trigger a bunch of d_invalidate() on the same dentry in parallel. They end up fighting each other - any dentry picked for removal by one will be skipped by the rest and we'll go for the next iteration through the entire subtree, even if everything is being

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Mon, Jun 12, 2017 at 04:00:45PM -0700, Khazhismel Kumykov wrote: > On Fri, Jun 2, 2017 at 11:47 PM, Khazhismel Kumykov wrote: > > On Fri, Jun 2, 2017 at 11:20 PM, Al Viro wrote: > >> The thing is, unlike shrink_dcache_parent() we *can* bugger off as

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Mon, Jun 12, 2017 at 04:00:45PM -0700, Khazhismel Kumykov wrote: > On Fri, Jun 2, 2017 at 11:47 PM, Khazhismel Kumykov wrote: > > On Fri, Jun 2, 2017 at 11:20 PM, Al Viro wrote: > >> The thing is, unlike shrink_dcache_parent() we *can* bugger off as > >> soon as we'd found no victims, nothing

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Fri, Jun 2, 2017 at 11:47 PM, Khazhismel Kumykov wrote: > On Fri, Jun 2, 2017 at 11:20 PM, Al Viro wrote: >> The thing is, unlike shrink_dcache_parent() we *can* bugger off as >> soon as we'd found no victims, nothing mounted and dentry itself >> is

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Fri, Jun 2, 2017 at 11:47 PM, Khazhismel Kumykov wrote: > On Fri, Jun 2, 2017 at 11:20 PM, Al Viro wrote: >> The thing is, unlike shrink_dcache_parent() we *can* bugger off as >> soon as we'd found no victims, nothing mounted and dentry itself >> is unhashed. We can't do anything in

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Fri, Jun 2, 2017 at 11:20 PM, Al Viro wrote: > The thing is, unlike shrink_dcache_parent() we *can* bugger off as > soon as we'd found no victims, nothing mounted and dentry itself > is unhashed. We can't do anything in select_collect() (we would've > broken

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Fri, Jun 2, 2017 at 11:20 PM, Al Viro wrote: > The thing is, unlike shrink_dcache_parent() we *can* bugger off as > soon as we'd found no victims, nothing mounted and dentry itself > is unhashed. We can't do anything in select_collect() (we would've > broken shrink_dcache_parent() that way),

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Fri, Jun 02, 2017 at 10:22:39PM -0700, Khazhismel Kumykov wrote: > On Fri, Jun 2, 2017 at 6:12 PM, Al Viro wrote: > > Part of that could be relieved if we turned check_and_drop() into > > static void check_and_drop(void *_data) > > { > > struct detach_data

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Fri, Jun 02, 2017 at 10:22:39PM -0700, Khazhismel Kumykov wrote: > On Fri, Jun 2, 2017 at 6:12 PM, Al Viro wrote: > > Part of that could be relieved if we turned check_and_drop() into > > static void check_and_drop(void *_data) > > { > > struct detach_data *data = _data; > > > >

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Fri, Jun 2, 2017 at 6:12 PM, Al Viro wrote: > Part of that could be relieved if we turned check_and_drop() into > static void check_and_drop(void *_data) > { > struct detach_data *data = _data; > > if (!data->mountpoint && list_empty(>select.dispose)) >

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Fri, Jun 2, 2017 at 6:12 PM, Al Viro wrote: > Part of that could be relieved if we turned check_and_drop() into > static void check_and_drop(void *_data) > { > struct detach_data *data = _data; > > if (!data->mountpoint && list_empty(>select.dispose)) >

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Wed, May 17, 2017 at 02:58:11PM -0700, Khazhismel Kumykov wrote: > Once the dentry is on a shrink list would > it be unreachable anyways, Why would it be? Suppose e.g. shrink_dcache_parent() finds a dentry with zero refcount; to the shrink list it goes, right? Then, before we actually get

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Wed, May 17, 2017 at 02:58:11PM -0700, Khazhismel Kumykov wrote: > Once the dentry is on a shrink list would > it be unreachable anyways, Why would it be? Suppose e.g. shrink_dcache_parent() finds a dentry with zero refcount; to the shrink list it goes, right? Then, before we actually get

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Mon, May 22, 2017 at 11:18 AM, Khazhismel Kumykov wrote: > On Wed, May 17, 2017 at 2:58 PM Khazhismel Kumykov wrote: >> >> On Mon, May 15, 2017 at 5:05 PM, Khazhismel Kumykov >> wrote: >> > Hi, >> > >> > I'm seeing behavior in

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Mon, May 22, 2017 at 11:18 AM, Khazhismel Kumykov wrote: > On Wed, May 17, 2017 at 2:58 PM Khazhismel Kumykov wrote: >> >> On Mon, May 15, 2017 at 5:05 PM, Khazhismel Kumykov >> wrote: >> > Hi, >> > >> > I'm seeing behavior in d_invalidate, if multiple threads call d_invalidate >> > on >>

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Wed, May 17, 2017 at 2:58 PM Khazhismel Kumykov wrote: > > On Mon, May 15, 2017 at 5:05 PM, Khazhismel Kumykov wrote: > > Hi, > > > > I'm seeing behavior in d_invalidate, if multiple threads call d_invalidate > > on > > the same tree at the same,

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Wed, May 17, 2017 at 2:58 PM Khazhismel Kumykov wrote: > > On Mon, May 15, 2017 at 5:05 PM, Khazhismel Kumykov wrote: > > Hi, > > > > I'm seeing behavior in d_invalidate, if multiple threads call d_invalidate > > on > > the same tree at the same, behavior time blows up and all the calls hang

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Mon, May 15, 2017 at 5:05 PM, Khazhismel Kumykov wrote: > Hi, > > I'm seeing behavior in d_invalidate, if multiple threads call d_invalidate on > the same tree at the same, behavior time blows up and all the calls hang with > large enough trees/enough simultaneous callers.

Re: Hang/soft lockup in d_invalidate with simultaneous calls

On Mon, May 15, 2017 at 5:05 PM, Khazhismel Kumykov wrote: > Hi, > > I'm seeing behavior in d_invalidate, if multiple threads call d_invalidate on > the same tree at the same, behavior time blows up and all the calls hang with > large enough trees/enough simultaneous callers. (e.g. a directory w/

Hang/soft lockup in d_invalidate with simultaneous calls

Hi, I'm seeing behavior in d_invalidate, if multiple threads call d_invalidate on the same tree at the same, behavior time blows up and all the calls hang with large enough trees/enough simultaneous callers. (e.g. a directory w/ 100k entries in d_subdir, and 5 or so threads calling d_invalidate

Hang/soft lockup in d_invalidate with simultaneous calls

Hi, I'm seeing behavior in d_invalidate, if multiple threads call d_invalidate on the same tree at the same, behavior time blows up and all the calls hang with large enough trees/enough simultaneous callers. (e.g. a directory w/ 100k entries in d_subdir, and 5 or so threads calling d_invalidate