CC Dave Chinner, who I recall had strong opinions on the mempool model
The context is commit f9054c7 ("mm, mempool: only set __GFP_NOMEMALLOC
if there are free elements"), which gives MEMALLOC/TIF_MEMDIE mempool
allocations access to the system emergency reserves when there is no
reserved object c
On Mon 18-07-16 19:00:57, David Rientjes wrote:
> On Mon, 18 Jul 2016, Michal Hocko wrote:
>
> > David Rientjes was objecting that such an approach wouldn't help if the
> > oom victim was blocked on a lock held by process doing mempool_alloc. This
> > is very similar to other oom deadlock situatio
On Tue, Jul 12, 2016 at 10:32:33PM -0400, Mike Snitzer wrote:
> On Tue, Jul 12 2016 at 10:18pm -0400,
> Eric Wheeler wrote:
>
> > On Tue, 12 Jul 2016, NeilBrown wrote:
> >
> > > On Tue, Jul 12 2016, Lars Ellenberg wrote:
> > >
> > > >
> > > > Instead, I suggest to distinguish between recurs
From: Heinz Mauelshagen
raid_io_hints() was retrieving the number of data stripes
used for the calculation of io_opt from struct r5conf,
which is not defined for raid0 mappings.
Base the calculation on the in-core raid_set structure instead.
Adjust to use to_bytes() for the sector -> bytes conv
From: Heinz Mauelshagen
Use 'unsigned int' where appropriate.
Return negative errors.
Correct an indentation.
Signed-off-by: Heinz Mauelshagen
---
drivers/md/dm-raid.c | 40
1 file changed, 20 insertions(+), 20 deletions(-)
diff --git a/drivers/md/dm-r
Mike Christie wrote:
> Add rbd to hwtable. These defaults are for the HA type of setup
> supported by the checker. We do no support features like multibus
> at the dm-multipath level yet.
Please use this patch as guide:
https://patchwork.kernel.org/patch/9225521/
> Signed-off-by: Mike Christie
On Tue 19-07-16 09:54:26, Johannes Weiner wrote:
> On Mon, Jul 18, 2016 at 10:41:24AM +0200, Michal Hocko wrote:
> > The original intention of f9054c70d28b was to help with the OOM
> > situations where the oom victim depends on mempool allocation to make a
> > forward progress. We can handle that c
If it is removed, the compiler warns: "undefined reference to `FREE'"
Cc: Christophe Varoqui
Cc: device-mapper development
Signed-off-by: Xose Vazquez Perez
---
multipathd/uxclnt.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/multipathd/uxclnt.c b/multipathd/uxclnt.c
ind
Added in 57a6c5c7.
But it could be removed because it's already included in "checkers.h"
Cc: Christophe Varoqui
Cc: device-mapper development
Signed-off-by: Xose Vazquez Perez
---
libmultipath/checkers/emc_clariion.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/libmultip
On Tue, 19 Jul 2016, Johannes Weiner wrote:
> Mempool guarantees forward progress by having all necessary memory
> objects for the guaranteed operation in reserve. Think about it this
> way: you should be able to delete the pool->alloc() call entirely and
> still make reliable forward progress. It
Cc: Christophe Varoqui
Cc: device-mapper development
Signed-off-by: Xose Vazquez Perez
---
libmultipath/blacklist.h | 2 +-
libmultipath/dict.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/libmultipath/blacklist.h b/libmultipath/blacklist.h
index 24a5fa5..443025d
On Mon, 18 Jul 2016, Michal Hocko wrote:
> From: Michal Hocko
>
> There has been a report about OOM killer invoked when swapping out to
> a dm-crypt device. The primary reason seems to be that the swapout
> out IO managed to completely deplete memory reserves. Mikulas was
> able to bisect and
On Mon, 18 Jul 2016, Michal Hocko wrote:
> From: Michal Hocko
>
> Mikulas has reported that a swap backed by dm-crypt doesn't work
> properly because the swapout cannot make a sufficient forward progress
> as the writeout path depends on dm_crypt worker which has to allocate
> memory to perfor
On Tue, 19 Jul 2016, Michal Hocko wrote:
> On Tue 19-07-16 09:54:26, Johannes Weiner wrote:
> > On Mon, Jul 18, 2016 at 10:41:24AM +0200, Michal Hocko wrote:
> > > The original intention of f9054c70d28b was to help with the OOM
> > > situations where the oom victim depends on mempool allocation
Hello Mike,
If I run a fio data integrity test against kernel v4.7-rc7 then I see
often that fio reports I/O errors if a path is removed despite
queue_if_no_path having been set in /etc/multipath.conf. Further
analysis showed that this happens because during SCSI device removal a
SCSI device
On Tue 19-07-16 17:50:29, Mikulas Patocka wrote:
>
>
> On Mon, 18 Jul 2016, Michal Hocko wrote:
>
> > From: Michal Hocko
> >
> > There has been a report about OOM killer invoked when swapping out to
> > a dm-crypt device. The primary reason seems to be that the swapout
> > out IO managed to co
16 matches
Mail list logo