On Tue, Nov 27, 2012 at 10:16:28PM -0500, Steven Rostedt wrote:
> On Fri, 2012-11-23 at 18:15 +0100, Lars Ellenberg wrote:
> > When toying around with debugfs, intentionally trying to break things,
> > I managed to get it into a reproducible endless loop w
On Wed, Nov 28, 2012 at 10:37:54AM +0100, Lars Ellenberg wrote:
> Note that all but rmmod may be done as normal user,
Ok, that is not exactly correct,
I forgot about debugfs being mounted 0700 ;-)
Still...
Lars
--
To unsubscribe from this list: send the line "unsubscribe linu
d to a BUG()") from the block tree and commit
> "wait: add wait_event_lock_irq() interface" from the akpm tree.
Thanks.
We can just drop our copy-n-paste-from-md.h then.
Lars
---
commit 2712ab592def9f0171b6d47349879e975db57657
Author: Lars Ellen
On Thu, Apr 04, 2013 at 02:17:09PM +0800, Wei Yongjun wrote:
> From: Wei Yongjun
>
> Fix to return a negative error code from the error handling
> case instead of 0, as returned elsewhere in this function.
We used to have an err = -ENOMEM; "just in case"...
How about this instead:
diff --git a
On Wed, Sep 05, 2012 at 01:49:15AM -0700, Tejun Heo wrote:
> On Wed, Sep 05, 2012 at 10:44:55AM +0200, Philipp Reisner wrote:
> > > Currently, FLUSH/FUA doesn't enforce any ordering requirement. File
> > > systems are responsible for draining all writes which have to happen
> > > before and not is
On Tue, Sep 11, 2012 at 03:58:20PM +1000, NeilBrown wrote:
> On Mon, 10 Sep 2012 16:31:59 -0700 Kent Overstreet
> wrote:
>
> > cc'ing Neil
> >
> > On Mon, Sep 10, 2012 at 04:06:54PM -0700, Tejun Heo wrote:
> > > Hello, again.
> > >
> > > cc'ing Kent and Vivek. The original thread is at
> > >
On Mon, Sep 24, 2012 at 03:34:46PM -0700, Kent Overstreet wrote:
> Just a little convenience macro - main reason to add it now is preparing
> for immutable bio vecs, it'll reduce the size of the patch that puts
> bi_sector/bi_size/bi_idx into a struct bvec_iter.
For the DRBD part:
> diff --git a
On Tue, Sep 25, 2012 at 03:06:24PM -0700, Kent Overstreet wrote:
> On Tue, Sep 25, 2012 at 01:54:52PM +0200, Lars Ellenberg wrote:
> > On Mon, Sep 24, 2012 at 03:34:46PM -0700, Kent Overstreet wrote:
> > > Just a little convenience macro - main reason to add it now is preparing
&
On Thu, Sep 06, 2012 at 02:29:52PM -0700, Tejun Heo wrote:
> Hello,
>
> On Wed, Sep 05, 2012 at 12:07:24PM +0200, Lars Ellenberg wrote:
> > So reiterating the situation:
> >
> > If I'd submit a non-empty bio with FLUSH/FUA set,
> > on a qu
e the hard limit for the number of
minor devices (allocation of an array of corresponding size), that has
long since changed, and now it is really only used as scaling factor for
these mempools.
Lars
--
: Lars Ellenberg
: LINBIT | Your Way to High Availability
: DRBD/HA support and co
On Fri, Mar 22, 2013 at 11:17:06AM -0600, Jens Axboe wrote:
> On Tue, Mar 19 2013, Philipp Reisner wrote:
> > This patch series improves DRBD's behavior. A submitter is no longer blocked
> > in the case of a AL-miss. For this a dedicated submitter worker is
> > introduced
> > (patch 13).
> Patch
From: Lars Ellenberg
Signed-off-by: Philipp Reisner
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_int.h | 13 +
drivers/block/drbd/drbd_main.c | 26 +-
drivers/block/drbd/drbd_nl.c |1 +
drivers/block/drbd/drbd_req.c | 29
On Mon, Feb 25, 2013 at 11:28:03AM -0800, Randy Dunlap wrote:
> On 02/24/13 21:00, Stephen Rothwell wrote:
> > Hi all,
> >
> > Please do not add any work destined for v3.10 to your -next included
> > branches until after Linus has release v3.9-rc1.
> >
> > Changes since 20130222:
> >
>
>
> on
> >
> > --
> >
> > From: "Philipp Reisner "
> >
> > commit ef57f9e6bb9278720c8a5278728f252ab85d7ac6 upstream
> >
> > Signed-off-by: Philipp Reisner
> > Signed-off-by: Lars Ellenberg
> > Signed-off-by: Jens
09-06-10 "Fixed a deadlock when using automatic split brain recovery when
> > both nodes are"
> > replacing drbd_set_role() with drbd_change_state() in that code-path,
> > but the first hunk of that patch forgets to remove the drbd_set_role().
> >
> > We app
On Wed, Apr 03, 2013 at 04:00:31PM +0800, Chen Gang wrote:
> Hello maintainers:
>
> when you have time, please help to check this patch whether is OK.
>
> thanks.
Thanks.
It has already been queued in:
http://git.kernel.dk/?p=linux-block.git;a=commit;h=0b6ef4164f50698eee536903d69d086add1a78
On Tue, Apr 09, 2013 at 09:22:40AM +0800, Chen Gang wrote:
> On 2013年04月09日 04:45, Lars Ellenberg wrote:
> > On Wed, Apr 03, 2013 at 04:00:31PM +0800, Chen Gang wrote:
> >> > Hello maintainers:
> >> >
> >> > when you have time,
When toying around with debugfs, intentionally trying to break things,
I managed to get it into a reproducible endless loop when cleaning up.
debugfs_remove_recursive() completely ignores that entries found
on ->d_subdirs may already be d_delete()d, but not yet d_kill()ed.
In this case, the firs
altime. Generally, you combine with a program called heartbeat to
> > implement high-availabilty failover. It's very neat ;-)
>
> They should get it merged then. Anyway, I'm glad it's working...
will do, once we have it (or, the next generation of it) in shape.
may
k device driver as developed mainly
by Lars Ellenberg and Philipp Reisner
both employed at Linbit.
We implement shared-disk semantics in a shared-nothing cluster.
yes, we even support cluster file systems (OCFS2, GFS2)
[support for this is somewhat clumsy, still,
mainly userland/setup/convenience i
On Sat, Jul 21, 2007 at 11:17:43PM +0200, Jan Engelhardt wrote:
>
> On Jul 21 2007 22:38, Lars Ellenberg wrote:
> >
> >We implement shared-disk semantics in a shared-nothing cluster.
>
> If nothing is shared, the disk is not shared, but got shared-disk
> semantics? A
On Sun, Jul 22, 2007 at 01:50:09AM +0200, Andi Kleen wrote:
> Lars Ellenberg <[EMAIL PROTECTED]> writes:
> >
> > Jens, Andrew, anyone: please review,
> > and give me advice how to proceed from here.
>
> The standard procedure would be to post all the source cod
On Sun, Jul 22, 2007 at 07:52:36AM +0200, Jens Axboe wrote:
> On Sun, Jul 22 2007, Andi Kleen wrote:
> > Lars Ellenberg <[EMAIL PROTECTED]> writes:
> > >
> > > Jens, Andrew, anyone: please review,
> > > and give me advice how to proceed from here.
> &
On Sun, Jul 22, 2007 at 11:06:59AM +0200, Jan Engelhardt wrote:
>
> On Jul 22 2007 00:43, Lars Ellenberg wrote:
> >On Sat, Jul 21, 2007 at 11:17:43PM +0200, Jan Engelhardt wrote:
> >> On Jul 21 2007 22:38, Lars Ellenberg wrote:
> >> >
> >> >We implem
On Mon, Jul 23, 2007 at 11:19:40AM +0200, Sam Ravnborg wrote:
> The message is quite clear.
> For a driver like this do something like:
> - Describe the functionality / the problem is solved
> - Describe the overall design (maybe as a file in Documentation/?)
I'm currently writing the paper for li
On Sun, Jul 22, 2007 at 09:32:02PM -0400, Kyle Moffett wrote:
> Ok, I didn't have a chance to get through anywhere near all of it, but
> here's my comments so far. I didn't really go through things in any
> particular order but most of these comments are about your drbd_int.h
> header file. Hopef
On Mon, Jul 23, 2007 at 03:37:04PM +0200, Jens Axboe wrote:
> On Mon, Jul 23 2007, Lars Ellenberg wrote:
> > > +/* THINK maybe we actually want to use the default "event/%s" worker
> > > threads
> > > + * or similar in linux 2.6, which uses per cpu dat
On Mon, Jul 23, 2007 at 07:10:58PM +0530, Satyam Sharma wrote:
> On 7/23/07, Lars Ellenberg <[EMAIL PROTECTED]> wrote:
> >On Sun, Jul 22, 2007 at 09:32:02PM -0400, Kyle Moffett wrote:
> >[...]
> >> Don't use signals between kernel threads, use proper primitives
On Wed, Jul 25, 2007 at 04:41:53AM +0530, Satyam Sharma wrote:
> Hi Lars,
>
>
> On 7/24/07, Lars Ellenberg <[EMAIL PROTECTED]> wrote:
> >On Mon, Jul 23, 2007 at 07:10:58PM +0530, Satyam Sharma wrote:
> >> On 7/23/07, Lars Ellenberg <[EMAIL PROTECTED]> wro
/* netfilter subsystem */
>
> That's not only buggy but also not used by web100.
>
>
> cu
> Adrian
thank you very much for looking at this,
--
: Lars EllenbergTel +43-1-8178292-0 :
: LINBIT Information Technologies GmbH Fax +43-1-817829
On Fri, Jul 27, 2007 at 06:46:17PM +, Pavel Machek wrote:
> Hi!
>
> > > >We implement shared-disk semantics in a shared-nothing cluster.
> > >
> > > If nothing is shared, the disk is not shared, but got shared-disk
> > > semantics? A little confusing.
> >
> > Think of it as RAID1 over TCP.
>
sorry, I did not have the time to "beautify" our code base further,
so still no patch submission for the in-depth review. looking at my
schedule, this probably has to wait for september at the very least
:(
meanwhile, please, anyone interessted,
the drbd paper for LinuxConf Eu 2007 is finalized.
those "drbd_requests"
does more than just the mempoll_free().
Fixes: bdfafc4ffdd2 ("locking/atomic, kref: Kill kref_sub()")
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_req.c | 27 +++
1 file changed, 15 insertions(+), 12 deletions(-)
diff -
On Mon, Jan 15, 2018 at 03:52:15PM -0800, Eric Wheeler wrote:
> From: Eric Wheeler
>
> For DRBD resources with long names that start with the same prefix,
> it was difficult to find all process pids for that minor since names
> are truncated to the task_struct's comm field (16 bytes).
>
> This p
On Mon, Jan 15, 2018 at 11:26:15PM -0800, Christoph Hellwig wrote:
> NAK. Calling a discard and expecting zeroing is simply buggy.
What he said.
The bug/misunderstanding was that we now use zeroout even for discards,
with the assumption that it would try to do discards where it can.
Which is ev
On Wed, May 16, 2018 at 11:43:32AM +0200, Christoph Hellwig wrote:
> And stop messing with try_module_get on THIS_MODULE, which doesn't make
> any sense here.
The idea was to increase module count on /proc/drbd access.
If someone holds /proc/drbd open, previously rmmod would
"succeed" in starting
On Mon, Dec 17, 2018 at 10:29:38AM -0700, Jens Axboe wrote:
> > Hi Lars and Philipp,
> >
> > Could you please make sure that this patch and the other one I sent make
> > it into 4.21/5.0? I am not sure when you were planning on sending the
> > pull request to Jens that you mentioned in the other t
quot;.
All others are sufficiently simple and supposedly "obvious".
Lars Ellenberg (13):
drbd: must not use connection after kref_put(&connection->kref)
drbd: centralize printk reporting of new size into drbd_set_my_capacity()
drbd: ignore "all zero" peer volume
-> may sleep
While using GFP_ATOMIC whould have been possible in the first two cases,
the real fix is to narrow the rcu_read_lock.
Reported-by: Jia-Ju Bai
Reviewed-by: Lars Ellenberg
Signed-off-by: Roland Kammerer
---
drivers/block/drbd/drbd_receiver.c | 11 ++-
1 file changed, 6 i
Previously, some implicit resizes that happend during handshake
have not been reported as prominently as explicit resize.
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_int.h | 8 +---
drivers/block/drbd/drbd_main.c | 17 -
drivers/block/drbd/drbd_nl.c | 3
nt size, am Primary, and he has the only disk,
which is larger than my current size
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_receiver.c | 33 +++---
1 file changed, 30 insertions(+), 3 deletions(-)
diff --git a/drivers/block/drbd/drbd_receiver.c
b
If we would reject a new handshake, if the peer had attached first,
and then connected, we should force disconnect if the peer first connects,
and only then attaches.
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_receiver.c | 11 ++-
1 file changed, 6 insertions(+), 5
cess).
Fix this by skipping the activity log adjustment if the "al-extents"
setting did not change. Also, in case it did change, fail early without
blocking if it looks like we would block forever.
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_nl.c | 37
ix is to forcefully disconnect as soon as we notice that the peer
attached to the "wrong" dataset.
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_receiver.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/block/drbd/drbd_receiver.c
b/drivers/block
asily abort the attach on the peer without more
refactoring of the handshake.
We now reject any attach to "unsuitable" uuids,
as long as we can see a Primary role,
unless we already have access to "good" data.
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_nl.c
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_state.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/block/drbd/drbd_state.c b/drivers/block/drbd/drbd_state.c
index 0813c654c893..18d53fe60d1d 100644
--- a/drivers/block/drbd/drbd_state.c
+++ b/drivers
ing to attach.
It also allows to attach with an explicit size.
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_nl.c | 49
1 file changed, 44 insertions(+), 5 deletions(-)
diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_receiver.c | 2 +-
drivers/block/drbd/drbd_state.c| 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/block/drbd/drbd_receiver.c
b/drivers/block/drbd/drbd_receiver.c
index 76d74b2122d6..3a0fe357b68b 100644
an AuthChallenge,
but the peer is not even configured to expect or use a shared secret.
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_receiver.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/block/drbd/drbd_receiver.c
b/drivers/block/drbd/drbd
delaying the error return becomes annoying.
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_nl.c | 15 ---
1 file changed, 8 insertions(+), 7 deletions(-)
diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c
index 82915880c5e9..bfe1b0062d62 100644
--- a/dri
l,
we want to do that, even if our backend does not announce
max_write_zeroes_sectors itself.
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_debugfs.c | 2 +
drivers/block/drbd/drbd_int.h | 11 +-
drivers/block/drbd/drbd_main.c | 11 +-
drivers/block/drbd/drbd_nl.c
From: Luc Van Oostenryck
print_st_err() is defined with its 4th argument taking an
'enum drbd_state_rv' but its prototype use an int for it.
Fix this by using 'enum drbd_state_rv' in the prototype too.
Signed-off-by: Luc Van Oostenryck
Signed-off-by: Roland Kammerer
udate yourself".
Try to catch this by checking for current connection status,
and possibly retry as local-only state change instead.
Signed-off-by: Lars Ellenberg
---
drivers/block/drbd/drbd_nl.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/drivers/block/drbd/drbd_n
From: Nathan Chancellor
Clang warns when an implicit conversion is done between enumerated
types:
drivers/block/drbd/drbd_state.c:708:8: warning: implicit conversion from
enumeration type 'enum drbd_ret_code' to different enumeration type
'enum drbd_state_rv' [-Wenum-conversion]
y sent by Arnd Bergmann with a default case
statement: https://lore.kernel.org/patchwork/patch/756723/
Link: https://github.com/ClangBuiltLinux/linux/issues/43
Suggested-by: Lars Ellenberg
Signed-off-by: Nathan Chancellor
---
include/linux/genl_magic_struct.h | 5 -
1 file changed, 4 insertions
t;, I disagree.
I think it is more cpu cache friendly to hweight_long then set to ~0UL
each word in turn, than to first bitmap_weight() all words, then
bitmap_fill() all words.
Thanks,
Lars Ellenberg
BTW, you swapped names and email addresses of phil and me. Are those
listed incorrectly s
On Tue, May 26, 2015 at 01:45:09AM +0530, Shailendra Verma wrote:
> In function overlapping_resync_write() variable rv is bool type.
> Hence assigning the values as true/false instead of 1/0.
I like 0 and 1 better. So unless the powers that be rule this to be the
recommended or mandatory coding s
without taking up any real network bandwidth.
In short: I don't see the point.
Lars Ellenberg
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
11:43 +0800
| From: Fengguang Wu
| To: Philipp Reisner , drbd-u...@lists.linbit.com,
linux-kernel@vger.kernel.org
| Subject: Re: [drbd?] Kernel panic - not syncing: Out of memory and no
killable processes...
| Message-ID: <20130612101143.GA13837@localhost>
|
| On Tue, Jun 11, 2013 at 05:33:27
blk_add_request_payload,
which then tries to use bio->bi_io_vec->page.
Fix: try to be even smarter in bch_generic_make_request_hack(),
and always pretend to have at least bi_max_vecs of 1,
unless the incoming bio was already created without a single bvec.
Signed-off-by: Lars Ellenberg
---
drivers/md/bcach
On Wed, Feb 19, 2014 at 01:07:14PM +, Lars Ellenberg wrote:
> bch_generic_make_request_hack() tries to be smart,
> and fake a bi_max_bvecs = bi_vcnt.
I meant to include a cover letter with the Oops backtrace,
and this:
Note that this patch is relevant for 3.13 and before.
With 3.14
On Tue, Jun 17, 2014 at 08:54:07AM +0200, Martin Kepplinger wrote:
> The one-bit bitfields are assigned true (1) or false (0) and checked
> for them respectively. While it should work either way and -1 is true
> as well it is more clear to see what's going on when using an unsigned int
> because 1
to properly initialize rs_last_events instead.
I'll think about that...
Lars Ellenberg
> Reported-by: Mikhail Sugakov
> Cc: Matt Wilson
> Signed-off-by: Imre Palik
> ---
> drivers/block/drbd/drbd_receiver.c |3 ++-
> 1 file changed, 2 insertions(+), 1 deletio
l need to review if the ->ldev->backing_bdev-> ... deref
is always safe to do. But I think it is.
commit c90cac67d6c12dcf3945f67a28718c40b644cc32
Author: Lars Ellenberg
Date: Thu Sep 4 13:15:45 2014 +0200
drbd: fix resync throttling initialization
If for some reason DRBD resync was the only activity
,
# drbdadm resize ... --assume-clean ...
No need to touch either the DRBD module, or the DRBD utils at all.
All there already.
Lars Ellenberg
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
e interesting so that new module won't use it. What is your
> opinion?
This was supposed to not be DRBD specific. But it might even still
need some massaging before it was truly generic. And obviously,
it does not meet the taste of genetlink folks, to say the least :(
I don't care either way.
Lars Ellenberg
t always use 0 to pad.
Patch is much shorter as well, see below.
Attribute type "0" is not used,
and will never be of semantic value,
but always be ignored in the DRBD netlink family.
Whereas using some arbitrary value will be wrong,
and will needlessly break userland.
Thanks,
On Tue, May 03, 2016 at 12:05:56PM -0400, David Miller wrote:
> From: Lars Ellenberg
> Date: Tue, 3 May 2016 12:06:44 +0200
>
> > Please just NOT use an additional "field",
> > but always use 0 to pad.
>
> You can't, it doesn't work.
I did, and it
On Wed, May 04, 2016 at 02:49:00PM +0200, Nicolas Dichtel wrote:
> Le 04/05/2016 11:05, Lars Ellenberg a écrit :
> [snip]
> > We don't have an "alignment problem" there, btw.
> > Last time I checked, we did work fine without this alignment magic,
> > we
On Tue, May 10, 2016 at 11:09:53AM +0200, Nicolas Dichtel wrote:
> Le 09/05/2016 15:15, Lars Ellenberg a écrit :
> > On Mon, May 09, 2016 at 11:40:20AM +0200, Nicolas Dichtel wrote:
> [snip]
> >> Maybe prefixing genl_magic_func.h and genl_magic_struct.h by 'drbd_'
&g
On Tue, May 10, 2016 at 11:39:49AM -0400, David Miller wrote:
> From: Lars Ellenberg
> Date: Tue, 10 May 2016 11:40:23 +0200
excuse me for reordering the original:
> Anyways, back to the topic, can you please just relent and come to
> some kind of agreement about the fix for this a
On Mon, Feb 15, 2016 at 09:26:25PM -0500, Insu Yun wrote:
> crypto_alloc_hash returns an error code, not NULL.
You are correct, of course. Was broken since its introduction
five years ago. Strange though, we have a helper function further
down in that file, and other, even much older, call sites a
On Thu, Nov 20, 2014 at 05:37:19PM -0500, nick wrote:
> Greetings Lars,
> I am wondering about removing the spin locks around the linked list
> operators in worker.c or do we need to test it first it first on
> actual hardware.
> Cheers Nick
Unless you have deep insight in the inner workings of DR
ze it to 0.
We used to have the explicit memset there,
but then changed that, because
I was under the impression that since
2007-07-17 d07dbea, Slab allocators: support __GFP_ZERO in all allocators
it was supported?
Lars Ellenberg
--
To unsubscribe from this list: send the line &qu
On Tue, Apr 05, 2016 at 07:56:48PM +0800, Ming Lei wrote:
> For a non-cloned bio, bio_add_page() only returns failure when
> the io vec table is full, but in that case, bio->bi_vcnt can't
> be zero at all.
>
> So remove the impossible failure handling.
Before the immutable bvecs,
we did in fact s
ber "struct blk_plug_cb cb" at an arbitrary offset?
As it is, raid1_check_plugged() below is actually just a cast.
Signed-off-by: Lars Ellenberg
---
drivers/md/raid1.c | 19 +--
drivers/md/raid10.c | 21 +
drivers/md/raid5.c | 5 +
fs/btrfs/raid56
On Wed, Apr 06, 2016 at 01:10:57PM +1000, NeilBrown wrote:
> On Wed, Apr 06 2016, Shaohua Li wrote:
>
> > On Tue, Apr 05, 2016 at 03:36:57PM +0200, Lars Ellenberg wrote:
> >> blk_check_plugged() will return a pointer
> >> to an object linked on current->plug->c
On Mon, Apr 25, 2016 at 09:37:28AM -0700, Bart Van Assche wrote:
> On 04/25/2016 05:13 AM, Philipp Reisner wrote:
> >+while (nr_sectors >= granularity) {
> >+nr = min_t(sector_t, nr_sectors, max_discard_sectors);
> >+err |= blkdev_issue_discard(bdev, start, nr, GFP_NOIO,
On Mon, Apr 25, 2016 at 02:49:11PM -0500, Philipp Reisner wrote:
> Am Montag, 25. April 2016, 11:48:30 schrieb Bart Van Assche:
> > On 04/25/2016 09:42 AM, Philipp Reisner wrote:
> > > Am Montag, 25. April 2016, 08:35:26 schrieb Bart Van Assche:
> > >> On 04/25/2016 05:10 AM, Philipp Reisner wrote:
On Mon, Apr 25, 2016 at 02:07:38PM +0200, Philipp Reisner wrote:
> + max_discard_sectors = min(q->limits.max_discard_sectors, (1U << 22));
> + max_discard_sectors -= max_discard_sectors % granularity;
there
On Mon, Apr 25, 2016 at 10:32:17PM +0200, Lars Ellenberg
On Tue, Apr 26, 2016 at 10:06:10AM +0200, Nicolas Dichtel wrote:
>
> This is the continuation (series #3) of the work done to align netlink
> attributes when these attributes contain some 64-bit fields.
>
> It's the last patchset from what I've seen.
>
> The last user of nla_put_u64() is block/d
On Tue, Apr 26, 2016 at 01:54:27PM +0200, Lars Ellenberg wrote:
> On Tue, Apr 26, 2016 at 10:06:10AM +0200, Nicolas Dichtel wrote:
> >
> > This is the continuation (series #3) of the work done to align netlink
> > attributes when these attributes contain some 64-bit fields.
&
On Mon, Nov 09, 2015 at 08:56:10PM +0900, Tetsuo Handa wrote:
> There are many locations that do
>
> if (memory_was_allocated_by_vmalloc)
> vfree(ptr);
> else
> kfree(ptr);
>
> but kvfree() can handle both kmalloc()ed memory and vmalloc()ed memory
> using is_vmalloc_addr(). Unless cal
gt; +
> r = __clone_and_map_data_bio(ci, ti, ci->sector, &len);
> if (r < 0)
> return r;
>
> though I haven't tested, and the change (if it works) should probably be
> more fully integrated into surrounding code.
>
> You probably don&
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:
> > >
> > > >
ack bio-parts resulting from blk_queue_split() will be processed
in LIFO order, one-by-one, whenever the recursion list becomes empty.
Signed-off-by: Lars Ellenberg
Signed-off-by: Roland Kammerer
---
This is not a theoretical problem.
At least int DRBD, and an unfortunately high IO concurrency wr
On Fri, Jun 24, 2016 at 07:36:57PM +0800, Ming Lei wrote:
> >
> > This is not a theoretical problem.
> > At least int DRBD, and an unfortunately high IO concurrency wrt. the
> > "max-buffers" setting, without this patch we have a reproducible deadlock.
>
> Is there any log about the deadlock? And
On Mon, Jul 04, 2016 at 06:47:29PM +0800, Ming Lei wrote:
> >> One clean solution may be to convert the loop of generic_make_request()
> >> into the following way:
> >>
> >> do {
> >> struct bio *splitted, *remainder = NULL;
> >> struct request_queue *q = bdev_get_queue(bio->bi_bdev);
> >>
On Thu, Jul 07, 2016 at 03:35:48PM +1000, NeilBrown wrote:
> On Wed, Jun 22 2016, Lars Ellenberg wrote:
>
> > For a long time, generic_make_request() converts recursion into
> > iteration by queuing recursive arguments on current->bio_list.
> >
> > This is conveni
On Wed, Jul 06, 2016 at 11:57:51PM +0800, Ming Lei wrote:
> > my suggestion
> >
> > generic_make_request(bio_orig)
> > NULLin-flight=0
> > bio_origempty in-flight=0
> > qA->make_request_fn(bio_orig)
> > blk_queue_split()
> > res
On Thu, Jul 07, 2016 at 10:16:16AM +0200, Lars Ellenberg wrote:
> > > Instead, I suggest to distinguish between recursive calls to
> > > generic_make_request(), and pushing back the remainder part in
> > > blk_queue_split(), by pointing current->
incoming = [ bio(l=2,aa), bio(l=2,ab), ...,
bio(l=1,a,remainder), bio(l=1,b), bio(l=1,c), ...,
bio(l=0,remainder_1) ]
.recursion = []
...
process away ... until back at l=0
.incoming = [ bio(l=0,remainder_1) ]
.recursion = []
potentially split further
.incoming = [ bio(l=0,now_2), bio(l=0,remainder_2) ]
.recursion = []
rinse, repeat.
Thanks,
Lars Ellenberg
On Fri, Jul 08, 2016 at 07:08:32PM +0800, Ming Lei wrote:
> > So after processing a particular bio, we should then process all the
> > 'child' bios - bios send to underlying devices. Then the 'sibling'
> > bios, that were split off, and then any remaining parents and ancestors.
>
> IMHO, that is
r .incoming, though.
>
> :-) neither "remainder" or "recursion" seem like brilliant names to me,
> but I don't have anything better to suggest. Naming is hard!
> As long as a comment explains the name clearly I could cope with X and Y.
...
> I think we just might be in violent agreement.
I thought so, too :-)
Should I merge both patches,
rename to ".queue" and ".tmp",
and submit for inclusion?
Lars Ellenberg
".recursion", or maybe use something like
> ".downstream".
"Naming is hard" :-)
.recursion and .queue, then,
because .remainder does no longer fit.
Also, what about "Cc: Stable"?
The patch does not strictly fall into the "stable" rules,
(too many lines...).
But it should go there anyways, I think.
I'll just add those tags, and wait for the heat?
Lars Ellenberg
Result of RFC previously discussed here:
https://lkml.org/lkml/2016/6/22/172
[RFC] block: fix blk_queue_split() resource exhaustion
Rebased to linux-block/for-4.8/core as of today.
Would also need to go to Stable 4.3 and later.
Lars Ellenberg (1):
block: fix blk_queue_split() resource
el bios first,
and "sibling" bios of the same level in "natural" order.
Signed-off-by: Lars Ellenberg
Signed-off-by: Roland Kammerer
---
block/bio.c | 27 +
block/blk-core.c | 50 +-
On Sat, Jul 02, 2016 at 06:28:29PM +0800, Ming Lei wrote:
> The idea looks good, but not sure it can cover all cases of
> dm over brbd or brbd over dm and maintaining two lists
> becomes too complicated.
>
> One clean solution may be to convert the loop of generic_make_request()
> into the followi
ndmsg().
> > >
> > > Cc: sta...@vger.kernel.org
> > > Cc: v...@zeniv.linux.org.uk
> > > Cc: christoph.lechleit...@iteg.at
> > > Cc: wolfgang.g...@iteg.at
> > > Reported-by: Christoph Lechleitner
> > > Tested-by: Christoph Lechleitne
On Wed, Nov 09, 2016 at 04:47:15PM +0100, Richard Weinberger wrote:
> > Should I sent two patches, one that applies to 4.5 and later,
> > and one that applies to 2.6.33 ... 4.4, or are you or stable
> > willing to resolve the trivial "missing comment block" conflict yourself?
>
> BTW: Why did you
1 - 100 of 117 matches
Mail list logo
