mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Veaceslav Falico <vfalico@redhat.com>
To: linux-kernel@vger.kernel.org
Cc: Bjorn Helgaas <bhelgaas@google.com>,
	linux-pci@vger.kernel.org, Veaceslav Falico <vfalico@redhat.com>
Subject: [PATCH 0/3] msi: fix kobject/sysfs removal from msi_list
Date: Tue, 17 Sep 2013 03:47:41 +0200	[thread overview]
Message-ID: <1379382464-7920-1-git-send-email-vfalico@redhat.com> (raw)

Currently, while removing msi_list's ->kobj, we just do kobject_put() on it
and after that free the entry itself. However, kobject_put() doesn't
guarantee that the kobject itself is freed - it can be used by someone else
and thus, when we'll free the entry, we'll free the kobject too - leading
to bugs in the other users (or when we'll finally release it).

Also, in some cases we might fail to register the kobjects, but we forget
to remove pdev->msi_kset, and this can lead to errors if we try to
re-register it - cause we already have that kset initialized.

Fix both issues by moving msi_kset/kobject deinitialization code completely
to free_msi_irqs(), which is called every time we fail and need to roll
back (and on the proper device irqs deinit). Also, move kfree-ing of the
msi_list entry to kobject->release (msi_kobj_release()), so that the entry
containing kobject will only be delisted in free_msi_irqs(), and free only
when there are no other users of its kobject.

CC: Bjorn Helgaas <bhelgaas@google.com>
CC: linux-pci@vger.kernel.org
CC: linux-kernel@vger.kernel.org
Signed-off-by: Veaceslav Falico <vfalico@redhat.com>

---
 drivers/pci/msi.c | 38 +++++++++++++++-----------------------
 1 file changed, 15 insertions(+), 23 deletions(-)

             reply	other threads:[~2013-09-17  1:47 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-09-17  1:47 Veaceslav Falico [this message]
2013-09-17  1:47 ` [PATCH 1/3] msi: add forgotten pci_dev_put(pdev) to populate_msi_sysfs() Veaceslav Falico
2013-09-25 21:08   ` Bjorn Helgaas
2013-09-25 21:30     ` Bjorn Helgaas
2013-09-25 22:09       ` Veaceslav Falico
2013-09-25 23:23     ` Neil Horman
2013-09-25 23:35       ` Bjorn Helgaas
2013-09-26  9:27         ` Veaceslav Falico
2013-09-26 12:25         ` Veaceslav Falico
2013-09-26 14:07           ` Veaceslav Falico
2013-09-26 22:16             ` Bjorn Helgaas
2013-09-26 23:05               ` Veaceslav Falico
2013-09-26 14:40           ` Neil Horman
2013-09-17  1:47 ` [PATCH 2/3] msi: always unregister ->msi_kset within free_msi_irqs() Veaceslav Falico
2013-09-17  1:47 ` [PATCH 3/3] msi: free msi_desc entry only after we've released the kobject Veaceslav Falico
2013-09-25 21:34   ` Bjorn Helgaas
2013-09-25 22:12     ` Veaceslav Falico

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=1379382464-7920-1-git-send-email-vfalico@redhat.com \
    --to=vfalico@redhat.com \
    --cc=bhelgaas@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    /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

all inboxes | Powered by JetHome®