From: Vladislav Bolkhovitin <vst@vlnb.net>
To: Greg KH <greg@kroah.com>
Cc: linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org,
scst-devel <scst-devel@lists.sourceforge.net>,
James Bottomley <James.Bottomley@HansenPartnership.com>,
Andrew Morton <akpm@linux-foundation.org>,
FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>,
Mike Christie <michaelc@cs.wisc.edu>,
Vu Pham <vuhuong@mellanox.com>,
Bart Van Assche <bart.vanassche@gmail.com>,
James Smart <James.Smart@Emulex.Com>,
Joe Eykholt <jeykholt@cisco.com>, Andy Yan <ayan@marvell.com>,
Chetan Loke <generationgnu@yahoo.com>,
Dmitry Torokhov <dmitry.torokhov@gmail.com>,
Hannes Reinecke <hare@suse.de>,
Richard Sharpe <realrichardsharpe@gmail.com>,
Daniel Henrique Debonzi <debonzi@linux.vnet.ibm.com>
Subject: Re: [PATCH 8/19]: SCST SYSFS interface implementation
Date: Thu, 14 Oct 2010 23:48:17 +0400 [thread overview]
Message-ID: <4CB75E81.7000208@vlnb.net> (raw)
In-Reply-To: <20101012190345.GA25737@kroah.com>
Greg KH, on 10/12/2010 11:03 PM wrote:
> On Tue, Oct 12, 2010 at 10:53:45PM +0400, Vladislav Bolkhovitin
> wrote:
>>> Seriously, you CAN NOT DO THIS! If you embed a kobject in a
>>> different structure, then you have to rely on the kobject to
>>> handle the reference counting for that larger structure. To do
>>> ANYTHING else is a bug and wrong.
>>>
>>> Please read the kobject documentation and fix this code up
>>> before submitting it again.
>>
>> Sure, I have read it and we rely on the kobject to handle the
>> reference counting for the larger structure. It's only done not in
>> a straightforward way, because the way it is implemented is
>> simpler for us + for some other reasons.
>
> Sorry, but I don't buy it.
>
>> For instance, for structure scst_tgt it is done using
>> tgt_kobj_release_cmpl completion. When a target driver calls
>> scst_unregister_target(), scst_unregister_target() in the end
>> calls scst_tgt_sysfs_del(), which calls kobject_put(&tgt->tgt_kobj)
>> and wait for tgt_kobj_release_cmpl to complete.
>
> Wait, why shouldn't the release then free the memory?
>
>> At this point tgt_kobj can be taken only by the SYSFS.
>> Scst_tgt_sysfs_del() can wait as much as needed until the SYSFS
>> code released it. As far as I can see, it can't be forever, so it's
>> OK.
>
> I don't understand, why can't you just free the memory, what are you
> having to wait for?
>
> You are only having one kobject for your structure, right? If so,
> then free the memory in the release, to wait for something else to
> free the memory is wrong.
>
>> Then, after scst_tgt_sysfs_del() returned,
>> scst_unregister_target() will free scst_tgt together with embedded
>> tgt_kobj.
>
> As no other kernel code is like this, I don't think it's valid to be
> doing so, sorry.
>
> Please fix this.
I'm sorry, but after I started implementing it I'm confused..
Originally I thought you are asking to make tgt_kobj be not embedded in
struct scst_tgt, but a pointer in it, so scst_tgtt_release() will
kfree() tgt_kobj. Hence, all the above I wrote about why we have
tgt_kobj embedded.
But now I feel like you are asking that scst_tgtt_release() should
kfree() tgt, not tgt_kobj.
Is it correct?
If yes, we did it this way to make errors processing uniform and
straightforward. In the code (simplified) our current errors recovery
looks like:
void scst_tgt_sysfs_del(struct scst_tgt *tgt)
{
...
kobject_put(&tgt->tgt_kobj);
wait_for_completion(&tgt->tgt_kobj_release_cmpl);
}
int scst_tgt_sysfs_create(struct scst_tgt *tgt)
{
init_completion(&tgt->tgt_kobj_release_cmpl);
res = kobject_init_and_add(&tgt->tgt_kobj, &tgt_ktype,
&tgt->tgtt->tgtt_kobj, tgt->tgt_name);
if (res != 0)
goto out;
res = sysfs_create_file(&tgt->tgt_kobj,
&tgt_enable_attr.attr);
if (res != 0)
goto out_err;
...
out:
return res;
out_err:
scst_tgt_sysfs_del(tgt);
goto out;
}
struct scst_tgt *scst_register_target()
{
struct scst_tgt *tgt;
int rc = 0;
TRACE_ENTRY();
rc = scst_alloc_tgt();
if (rc != 0)
goto out;
...
tgt->tgt_name = kmalloc(strlen(target_name) + 1, GFP_KERNEL);
if (tgt->tgt_name == NULL)
goto out_free_tgt;
...
mutex_lock();
rc = scst_tgt_sysfs_create(tgt);
if (rc < 0)
goto out_unlock;
tgt->default_acg = scst_alloc_add_acg();
if (tgt->default_acg == NULL)
goto out_sysfs_del;
...
out:
return tgt;
out_sysfs_del:
mutex_unlock();
scst_tgt_sysfs_del(tgt)
goto out_free_tgt;
out_unlock:
mutex_unlock();
out_free_tgt:
scst_free_tgt(tgt);
}
We have a simple and straightforward errors recovery semantic: if
scst_tgt_sysfs_create() failed, then tgt_kobj returned deinited as if
scst_tgt_sysfs_create() has never been called. This is a regular
practice in the kernel: don't return half-initialized objects.
If we implement freeing tgt in scst_tgtt_release() as you requesting, we
will need to add in the error recovery path additional recovery code to
track and delete half-initialized tgt_kobj. Particularly, we will need
to add additional flag to track that tgt_kobj initialized, hence needs
deleting. Similar flag and code will be added in all similar to scst_tgt
SCST objects.
This code will be quite errors prone as you can see on the example of
device_register() which on failure requires device_put() to be called
(http://lkml.org/lkml/2010/9/19/93). (I'm not questioning
device_register() implementation, there might be very good reasons to
implement it this way (I don't know), I mean, it is too easy to forget
to do the needed recovery of the half-created objects as this case
demonstrating.)
Could you confirm if I understand you correctly and need to implement
freeing tgt in the kobject release() function, please?
Thanks you very much,
Vlad
next prev parent reply other threads:[~2010-10-14 19:49 UTC|newest]
Thread overview: 93+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-10-01 21:34 [PATCHv4 0/19]: New SCSI target framework (SCST) with dev handlers and 2 target drivers Vladislav Bolkhovitin
2010-10-01 21:36 ` [PATCH 1/19]: Integration of SCST into the Linux kernel tree Vladislav Bolkhovitin
2010-10-01 21:36 ` [PATCH 2/19]: SCST core's Makefile and Kconfig Vladislav Bolkhovitin
2010-10-01 21:38 ` [PATCH 3/19]: SCST public headers Vladislav Bolkhovitin
2010-10-01 21:39 ` [PATCH 4/19]: SCST main management files and private headers Vladislav Bolkhovitin
2010-10-01 21:42 ` [PATCH 5/19]: SCST implementation of the SCSI target state machine Vladislav Bolkhovitin
2010-10-01 21:43 ` [PATCH 6/19]: SCST internal library functions Vladislav Bolkhovitin
2010-10-01 21:44 ` [PATCH 7/19]: SCST Persistent Reservations implementation Vladislav Bolkhovitin
2010-10-01 21:46 ` [PATCH 8/19]: SCST SYSFS interface implementation Vladislav Bolkhovitin
2010-10-09 21:20 ` Greg KH
2010-10-11 19:29 ` Vladislav Bolkhovitin
2010-10-11 21:32 ` Greg KH
2010-10-12 18:53 ` Vladislav Bolkhovitin
2010-10-12 19:03 ` Greg KH
2010-10-14 19:48 ` Vladislav Bolkhovitin [this message]
2010-10-14 20:04 ` Greg KH
2010-10-22 17:30 ` Vladislav Bolkhovitin
2010-10-22 17:56 ` Greg KH
2010-10-22 18:40 ` Vladislav Bolkhovitin
2010-10-22 18:54 ` Greg KH
2010-11-08 19:58 ` Vladislav Bolkhovitin
2010-11-09 0:28 ` Greg KH
2010-11-09 20:06 ` Vladislav Bolkhovitin
2010-11-10 9:58 ` Boaz Harrosh
2010-11-10 20:19 ` Vladislav Bolkhovitin
2010-11-10 20:29 ` Joe Eykholt
2010-11-10 20:38 ` Vladislav Bolkhovitin
2010-11-10 20:42 ` Joe Eykholt
2010-11-11 9:59 ` Boaz Harrosh
2010-11-11 12:04 ` Greg KH
2010-11-11 14:05 ` Boaz Harrosh
2010-11-11 14:16 ` Greg KH
2010-11-11 14:19 ` Boaz Harrosh
2010-11-11 20:50 ` Vladislav Bolkhovitin
2010-11-12 1:23 ` Dmitry Torokhov
2010-11-12 12:09 ` Bart Van Assche
2010-11-12 18:44 ` Dmitry Torokhov
2010-11-13 10:52 ` Bart Van Assche
2010-11-13 17:20 ` Vladislav Bolkhovitin
2010-11-13 23:59 ` Greg KH
2010-11-15 6:59 ` Dmitry Torokhov
2010-11-15 17:53 ` Bart Van Assche
2010-11-15 20:36 ` Vladislav Bolkhovitin
2010-11-15 9:46 ` Boaz Harrosh
2010-11-15 16:16 ` Greg KH
2010-11-15 17:19 ` Boaz Harrosh
2010-11-15 17:49 ` Bart Van Assche
2010-11-15 20:19 ` Nicholas A. Bellinger
2010-11-16 13:12 ` Vladislav Bolkhovitin
2010-11-16 11:59 ` [Scst-devel] " Richard Williams
2010-11-16 13:17 ` Vladislav Bolkhovitin
2010-11-18 21:02 ` Vladislav Bolkhovitin
2010-11-18 21:46 ` Greg KH
2010-11-19 18:00 ` Vladislav Bolkhovitin
2010-11-19 20:22 ` Dmitry Torokhov
2010-11-19 20:50 ` Vladislav Bolkhovitin
2010-11-19 21:16 ` Greg KH
2010-11-24 20:35 ` Vladislav Bolkhovitin
2010-11-19 21:19 ` Greg KH
2010-12-10 12:06 ` Bart Van Assche
2010-12-10 19:36 ` Greg KH
2010-12-14 14:10 ` Bart Van Assche
2010-11-19 18:01 ` Bart Van Assche
2010-11-15 20:39 ` Vladislav Bolkhovitin
2010-11-15 20:39 ` Vladislav Bolkhovitin
2010-11-15 17:45 ` Bart Van Assche
2010-11-15 18:44 ` Greg KH
2010-11-15 20:39 ` Vladislav Bolkhovitin
2010-11-15 22:13 ` Greg KH
2010-11-16 5:04 ` Joe Eykholt
2010-11-16 6:03 ` Nicholas A. Bellinger
2010-11-16 8:49 ` Florian Mickler
2010-11-16 13:18 ` Vladislav Bolkhovitin
2010-11-16 7:15 ` Bart Van Assche
2010-11-16 13:19 ` Vladislav Bolkhovitin
2010-11-15 20:36 ` Vladislav Bolkhovitin
2010-11-15 7:04 ` Dmitry Torokhov
2010-11-15 20:37 ` Vladislav Bolkhovitin
2010-11-15 21:14 ` Dmitry Torokhov
2010-11-16 13:13 ` Vladislav Bolkhovitin
2010-10-01 21:46 ` [PATCH 9/19]: SCST debugging support routines Vladislav Bolkhovitin
2010-10-01 21:48 ` [PATCH 10/19]: SCST SGV cache Vladislav Bolkhovitin
2010-10-01 21:48 ` [PATCH 11/19]: SCST core's docs Vladislav Bolkhovitin
2010-10-01 21:49 ` [PATCH 12/19]: SCST dev handlers' Makefile Vladislav Bolkhovitin
2010-10-01 21:50 ` [PATCH 13/19]: SCST vdisk dev handler Vladislav Bolkhovitin
2010-10-01 21:51 ` [PATCH 14/19]: SCST pass-through dev handlers Vladislav Bolkhovitin
2010-10-01 21:53 ` [PATCH 15/19]: Implementation of blk_rq_map_kern_sg() Vladislav Bolkhovitin
2010-10-01 21:57 ` [PATCH 16/19]: scst_local target driver Vladislav Bolkhovitin
2010-10-01 21:58 ` [PATCH 17/19]: SCST InfiniBand SRP " Vladislav Bolkhovitin
2010-10-01 22:04 ` [PATCH 18/19]: ibmvstgt: Port from tgt to SCST Vladislav Bolkhovitin
2010-10-01 22:05 ` [PATCH 19/19]: tgt: Removal Vladislav Bolkhovitin
2010-10-02 7:40 ` [PATCHv4 0/19]: New SCSI target framework (SCST) with dev handlers and 2 target drivers Bart Van Assche
2010-10-06 20:21 ` [Scst-devel] " Steve Modica
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4CB75E81.7000208@vlnb.net \
--to=vst@vlnb.net \
--cc=James.Bottomley@HansenPartnership.com \
--cc=James.Smart@Emulex.Com \
--cc=akpm@linux-foundation.org \
--cc=ayan@marvell.com \
--cc=bart.vanassche@gmail.com \
--cc=debonzi@linux.vnet.ibm.com \
--cc=dmitry.torokhov@gmail.com \
--cc=fujita.tomonori@lab.ntt.co.jp \
--cc=generationgnu@yahoo.com \
--cc=greg@kroah.com \
--cc=hare@suse.de \
--cc=jeykholt@cisco.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=michaelc@cs.wisc.edu \
--cc=realrichardsharpe@gmail.com \
--cc=scst-devel@lists.sourceforge.net \
--cc=vuhuong@mellanox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome