* Re: [git pull] device-mapper patches for 2.6.29
2009-01-06 2:51 ` Linus Torvalds
@ 2009-01-06 3:04 ` Stephen Rothwell
2009-01-06 3:14 ` [dm-devel] " Alasdair G Kergon
2009-01-06 3:15 ` Cheng Renquan (程任全)
2 siblings, 0 replies; 10+ messages in thread
From: Stephen Rothwell @ 2009-01-06 3:04 UTC (permalink / raw)
To: Alasdair G Kergon
Cc: Randy Dunlap, Linus Torvalds, dm-devel, linux-kernel,
Jonathan Brassow, Mikulas Patocka, Milan Broz, Andi Kleen,
Cheng Renquan, Jun'ichi Nomura, Kiyoshi Ueda, Takahiro Yasui
[-- Attachment #1: Type: text/plain, Size: 613 bytes --]
Hi Alasdair.
On Mon, 5 Jan 2009 18:51:00 -0800 (PST) Linus Torvalds <torvalds@linux-foundation.org> wrote:
>
> Oh no, this is totally broken, and doesn't compile even with my fairly
> minimal install:
>
> drivers/md/dm-target.c: In function ‘dm_unregister_target’:
> drivers/md/dm-target.c:106: error: implicit declaration of function ‘module_refcount’
This is from a patch added to linux-next only yesterday and was
reported by Randy Dunlap from his overnight build testing.
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [dm-devel] Re: [git pull] device-mapper patches for 2.6.29
2009-01-06 2:51 ` Linus Torvalds
2009-01-06 3:04 ` Stephen Rothwell
@ 2009-01-06 3:14 ` Alasdair G Kergon
2009-01-06 3:15 ` Cheng Renquan (程任全)
2 siblings, 0 replies; 10+ messages in thread
From: Alasdair G Kergon @ 2009-01-06 3:14 UTC (permalink / raw)
To: Linus Torvalds
Cc: Kiyoshi Ueda, Andi Kleen, linux-kernel, dm-devel, Cheng Renquan,
Mikulas Patocka, Jun'ichi Nomura, Milan Broz
On Mon, Jan 05, 2009 at 06:51:00PM -0800, Linus Torvalds wrote:
> On Mon, 5 Jan 2009, Alasdair G Kergon wrote:
> > Please pull from:
> > master.kernel.org:/pub/scm/linux/kernel/git/agk/linux-2.6-dm.git
> > to get the following device-mapper patches for 2.6.29:
> Oh no, this is totally broken, and doesn't compile even with my fairly
> minimal install:
> drivers/md/dm-target.c: In function ‘dm_unregister_target’:
> drivers/md/dm-target.c:106: error: implicit declaration of function ‘module_refcount’
Missing an ifdef:-( Sorry about that. Better drop those two patches for now
then.
Please pull from:
master.kernel.org:/pub/scm/linux/kernel/git/agk/linux-2.6-dm.git
to get the following device-mapper patches for 2.6.29:
Alasdair G Kergon (2):
dm mpath: move trigger_event to system workqueue
dm snapshot: split out exception store implementations
Andi Kleen (1):
dm: support barriers on simple devices
Jonathan Brassow (4):
dm raid1: fix error count
dm snapshot: separate out exception store interface
dm snapshot: rename struct exception_store
dm snapshot: extend exception store functions
Kiyoshi Ueda (2):
dm request: add caches
dm request: extend target interface
Mikulas Patocka (4):
dm table: drop reference at unbind
dm snapshot: change yield to msleep
dm: consolidate target deregistration error handling
dm table: rework reference counting
Milan Broz (4):
dm log: move region_size validation
dm log: ensure log bitmap fits on log device
dm ioctl: allow dm_copy_name_and_uuid to return only one field
dm: add name and uuid to sysfs
Takahiro Yasui (2):
dm log: fix dm_io_client leak on error paths
dm log: avoid reinitialising io_req on every operation
drivers/md/Makefile | 5 +-
drivers/md/dm-crypt.c | 6 +-
drivers/md/dm-delay.c | 6 +-
drivers/md/dm-exception-store.c | 749 +--------------------------------------
drivers/md/dm-exception-store.h | 148 ++++++++
drivers/md/dm-ioctl.c | 16 +-
drivers/md/dm-linear.c | 6 +-
drivers/md/dm-log.c | 40 ++-
drivers/md/dm-mpath.c | 14 +-
drivers/md/dm-raid1.c | 24 +-
drivers/md/dm-snap-persistent.c | 704 ++++++++++++++++++++++++++++++++++++
drivers/md/dm-snap-transient.c | 98 +++++
drivers/md/dm-snap.c | 48 ++-
drivers/md/dm-snap.h | 129 +-------
drivers/md/dm-stripe.c | 4 +-
drivers/md/dm-sysfs.c | 99 +++++
drivers/md/dm-table.c | 47 +++-
drivers/md/dm-target.c | 15 +-
drivers/md/dm-zero.c | 5 +-
drivers/md/dm.c | 101 ++++--
drivers/md/dm.h | 10 +
include/linux/device-mapper.h | 28 ++-
22 files changed, 1319 insertions(+), 983 deletions(-)
create mode 100644 drivers/md/dm-exception-store.h
create mode 100644 drivers/md/dm-snap-persistent.c
create mode 100644 drivers/md/dm-snap-transient.c
create mode 100644 drivers/md/dm-sysfs.c
Alasdair
--
agk@redhat.com
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [git pull] device-mapper patches for 2.6.29
2009-01-06 2:51 ` Linus Torvalds
2009-01-06 3:04 ` Stephen Rothwell
2009-01-06 3:14 ` [dm-devel] " Alasdair G Kergon
@ 2009-01-06 3:15 ` Cheng Renquan (程任全)
2009-01-06 3:21 ` Alasdair G Kergon
2 siblings, 1 reply; 10+ messages in thread
From: Cheng Renquan (程任全) @ 2009-01-06 3:15 UTC (permalink / raw)
To: Linus Torvalds
Cc: Alasdair G Kergon, dm-devel, linux-kernel, Jonathan Brassow,
Mikulas Patocka, Milan Broz, Andi Kleen, Jun'ichi Nomura,
Kiyoshi Ueda, Takahiro Yasui
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=UTF-8, Size: 1038 bytes --]
2009/1/6 Linus Torvalds <torvalds@linux-foundation.org>:>>> On Mon, 5 Jan 2009, Alasdair G Kergon wrote:>>>> Please pull from:>>>> master.kernel.org:/pub/scm/linux/kernel/git/agk/linux-2.6-dm.git>>>> to get the following device-mapper patches for 2.6.29:>> Oh no, this is totally broken, and doesn't compile even with my fairly> minimal install:>> drivers/md/dm-target.c: In function 'dm_unregister_target':> drivers/md/dm-target.c:106: error: implicit declaration of function 'module_refcount'>> Not pulled.>> Linus>
That's my fault, module_refcount can only be used ifCONFIG_MODULE_UNLOAD defined according to <linux/module.h>,
So now maybe we can add a null macro to module_refcount ifCONFIG_MODULE_UNLOAD undefined?
-- Cheng Renquan (ç¨ä»»å
¨), Shenzhen, ChinaPaula Poundstone - "I don't have a bank account because I don't knowmy mother's maiden name."ÿôèº{.nÇ+·®+%Ëÿ±éݶ\x17¥wÿº{.nÇ+·¥{±þG«éÿ{ayº\x1dÊÚë,j\a¢f£¢·hïêÿêçz_è®\x03(éÝ¢j"ú\x1a¶^[m§ÿÿ¾\a«þG«éÿ¢¸?¨èÚ&£ø§~á¶iOæ¬z·vØ^\x14\x04\x1a¶^[m§ÿÿÃ\fÿ¶ìÿ¢¸?I¥
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [git pull] device-mapper patches for 2.6.29
2009-01-06 3:15 ` Cheng Renquan (程任全)
@ 2009-01-06 3:21 ` Alasdair G Kergon
2009-01-06 3:32 ` Cheng Renquan (程任全)
0 siblings, 1 reply; 10+ messages in thread
From: Alasdair G Kergon @ 2009-01-06 3:21 UTC (permalink / raw)
To: Cheng Renquan (程任全)
Cc: Linus Torvalds, dm-devel, linux-kernel, Jonathan Brassow,
Mikulas Patocka, Milan Broz, Andi Kleen, Jun'ichi Nomura,
Kiyoshi Ueda, Takahiro Yasui
On Tue, Jan 06, 2009 at 11:15:58AM +0800, Cheng Renquan (程任全) wrote:
> So now maybe we can add a null macro to module_refcount if
> CONFIG_MODULE_UNLOAD undefined?
Do we even need the test at all?
Alasdair
--
agk@redhat.com
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [git pull] device-mapper patches for 2.6.29
2009-01-06 3:21 ` Alasdair G Kergon
@ 2009-01-06 3:32 ` Cheng Renquan (程任全)
2009-01-06 3:48 ` [dm-devel] " Alasdair G Kergon
2009-01-06 11:59 ` Stefan Richter
0 siblings, 2 replies; 10+ messages in thread
From: Cheng Renquan (程任全) @ 2009-01-06 3:32 UTC (permalink / raw)
To: Alasdair G Kergon
Cc: Linus Torvalds, dm-devel, linux-kernel, Jonathan Brassow,
Mikulas Patocka, Milan Broz, Andi Kleen, Jun'ichi Nomura,
Kiyoshi Ueda, Takahiro Yasui
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=UTF-8, Size: 726 bytes --]
On Tue, Jan 6, 2009 at 11:21 AM, Alasdair G Kergon <agk@redhat.com> wrote:> On Tue, Jan 06, 2009 at 11:15:58AM +0800, Cheng Renquan (ç¨ä»»å
¨) wrote:>> So now maybe we can add a null macro to module_refcount if>> CONFIG_MODULE_UNLOAD undefined?>> Do we even need the test at all?
So you mean add "#ifdef CONFIG_MODULE_UNLOAD ..." to dm-target.c? Ican do that, but I really don't know a better solution withoutconditional compiling?
>> Alasdair> --> agk@redhat.com
-- Cheng Renquan, Shenzhen, ChinaEddie Izzard - "I grew up in Europe, where the history comes from."ÿôèº{.nÇ+·®+%Ëÿ±éݶ\x17¥wÿº{.nÇ+·¥{±þG«éÿ{ayº\x1dÊÚë,j\a¢f£¢·hïêÿêçz_è®\x03(éÝ¢j"ú\x1a¶^[m§ÿÿ¾\a«þG«éÿ¢¸?¨èÚ&£ø§~á¶iOæ¬z·vØ^\x14\x04\x1a¶^[m§ÿÿÃ\fÿ¶ìÿ¢¸?I¥
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [dm-devel] Re: [git pull] device-mapper patches for 2.6.29
2009-01-06 3:32 ` Cheng Renquan (程任全)
@ 2009-01-06 3:48 ` Alasdair G Kergon
2009-01-06 11:59 ` Stefan Richter
1 sibling, 0 replies; 10+ messages in thread
From: Alasdair G Kergon @ 2009-01-06 3:48 UTC (permalink / raw)
To: crquan
Cc: Kiyoshi Ueda, linux-kernel, Andi Kleen, Mikulas Patocka,
device-mapper development, Jun'ichi Nomura, Linus Torvalds,
Milan Broz
On Tue, Jan 06, 2009 at 11:32:51AM +0800, Cheng Renquan (程任全) wrote:
> On Tue, Jan 6, 2009 at 11:21 AM, Alasdair G Kergon <agk@redhat.com> wrote:
> > On Tue, Jan 06, 2009 at 11:15:58AM +0800, Cheng Renquan (程任全) wrote:
> >> So now maybe we can add a null macro to module_refcount if
> >> CONFIG_MODULE_UNLOAD undefined?
> > Do we even need the test at all?
> So you mean add "#ifdef CONFIG_MODULE_UNLOAD ..." to dm-target.c? I
> can do that, but I really don't know a better solution without
> conditional compiling?
I mean under what set of circumstances could the test of the module refcount
here fail, and if the test was simply removed are there any cases where that
would matter?
Alasdair
--
agk@redhat.com
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [git pull] device-mapper patches for 2.6.29
2009-01-06 3:32 ` Cheng Renquan (程任全)
2009-01-06 3:48 ` [dm-devel] " Alasdair G Kergon
@ 2009-01-06 11:59 ` Stefan Richter
1 sibling, 0 replies; 10+ messages in thread
From: Stefan Richter @ 2009-01-06 11:59 UTC (permalink / raw)
To: "Cheng Renquan (程任全)"
Cc: Alasdair G Kergon, Linus Torvalds, dm-devel, linux-kernel,
Jonathan Brassow, Mikulas Patocka, Milan Broz, Andi Kleen,
Jun'ichi Nomura, Kiyoshi Ueda, Takahiro Yasui
Cheng Renquan (程任全) wrote:
> On Tue, Jan 6, 2009 at 11:21 AM, Alasdair G Kergon <agk@redhat.com> wrote:
>> On Tue, Jan 06, 2009 at 11:15:58AM +0800, Cheng Renquan (程任全) wrote:
>>> So now maybe we can add a null macro to module_refcount if
>>> CONFIG_MODULE_UNLOAD undefined?
>>
>> Do we even need the test at all?
>
> So you mean add "#ifdef CONFIG_MODULE_UNLOAD ..." to dm-target.c? I
> can do that, but I really don't know a better solution without
> conditional compiling?
The fact that module_refcount may be undefined is an indicator that this
is not an API for general use.
Why are you wanting to peek into the reference count of a module?
Shouldn't you have everything covered by properly recounting all your
data objects? (Preferably via the kref API.)
--
Stefan Richter
-=====-==--= ---= --==-
http://arcgraph.de/sr/
^ permalink raw reply [flat|nested] 10+ messages in thread