From: Heiner Kallweit <hkallweit1@gmail.com>
To: Lukas Wunner <lukas@wunner.de>
Cc: Kurt Kanzenbach <kurt@linutronix.de>,
Roman Lozko <lozko.roma@gmail.com>,
Dave Hansen <dave.hansen@linux.intel.com>,
linux-pci@vger.kernel.org,
Christian Marangi <ansuelsmth@gmail.com>,
Eric Dumazet <edumazet@google.com>,
netdev@vger.kernel.org, Tony Nguyen <anthony.l.nguyen@intel.com>,
Sean Christopherson <seanjc@google.com>,
Bjorn Helgaas <bhelgaas@google.com>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
"David S. Miller" <davem@davemloft.net>,
intel-wired-lan@lists.osuosl.org
Subject: Re: [Intel-wired-lan] Deadlock in pciehp on dock disconnect
Date: Fri, 5 Apr 2024 22:16:56 +0200 [thread overview]
Message-ID: <4c11ca11-8585-4cb5-bd0e-cd67eb1c208f@gmail.com> (raw)
In-Reply-To: <ZhBN9p1yOyciXkzw@wunner.de>
On 05.04.2024 21:16, Lukas Wunner wrote:
> On Fri, Apr 05, 2024 at 07:48:08PM +0200, Lukas Wunner wrote:
>> Roman, does the below patch fix the issue?
>
> Actually the patch in my previous e-mail was crap as the unregistering
> of the LEDs happened after unbind of the pdev, i.e. after
> igc_release_hw_control() and pci_disable_device().
>
For r8169 the first version is sufficient because everything is
device-managed.
> The driver otherwise doesn't seem to be using devm_*() and with
> devm_*() it's always all or nothing. A mix of devm_*() and manual
> teardown is prone to ordering issues.
>
> Here's another attempt:
>
> -- >8 --
>
> diff --git a/drivers/net/ethernet/intel/igc/igc.h b/drivers/net/ethernet/intel/igc/igc.h
> index 90316dc58630..f9ffe9df9a96 100644
> --- a/drivers/net/ethernet/intel/igc/igc.h
> +++ b/drivers/net/ethernet/intel/igc/igc.h
> @@ -298,6 +298,7 @@ struct igc_adapter {
>
> /* LEDs */
> struct mutex led_mutex;
> + struct igc_led_classdev *leds;
> };
>
> void igc_up(struct igc_adapter *adapter);
> @@ -723,6 +724,7 @@ void igc_ptp_read(struct igc_adapter *adapter, struct timespec64 *ts);
> void igc_ptp_tx_tstamp_event(struct igc_adapter *adapter);
>
> int igc_led_setup(struct igc_adapter *adapter);
> +void igc_led_teardown(struct igc_adapter *adapter);
>
> #define igc_rx_pg_size(_ring) (PAGE_SIZE << igc_rx_pg_order(_ring))
>
> diff --git a/drivers/net/ethernet/intel/igc/igc_leds.c b/drivers/net/ethernet/intel/igc/igc_leds.c
> index bf240c5daf86..4c2806c0878a 100644
> --- a/drivers/net/ethernet/intel/igc/igc_leds.c
> +++ b/drivers/net/ethernet/intel/igc/igc_leds.c
> @@ -236,8 +236,8 @@ static void igc_led_get_name(struct igc_adapter *adapter, int index, char *buf,
> pci_dev_id(adapter->pdev), index);
> }
>
> -static void igc_setup_ldev(struct igc_led_classdev *ldev,
> - struct net_device *netdev, int index)
> +static int igc_setup_ldev(struct igc_led_classdev *ldev,
> + struct net_device *netdev, int index)
> {
> struct igc_adapter *adapter = netdev_priv(netdev);
> struct led_classdev *led_cdev = &ldev->led;
> @@ -257,15 +257,15 @@ static void igc_setup_ldev(struct igc_led_classdev *ldev,
> led_cdev->hw_control_get = igc_led_hw_control_get;
> led_cdev->hw_control_get_device = igc_led_hw_control_get_device;
>
> - devm_led_classdev_register(&netdev->dev, led_cdev);
> + return led_classdev_register(&netdev->dev, led_cdev);
> }
>
> int igc_led_setup(struct igc_adapter *adapter)
> {
> struct net_device *netdev = adapter->netdev;
> - struct device *dev = &netdev->dev;
> + struct device *dev = &adapter->pdev->dev;
> struct igc_led_classdev *leds;
> - int i;
> + int i, ret;
>
> mutex_init(&adapter->led_mutex);
>
> @@ -273,8 +273,27 @@ int igc_led_setup(struct igc_adapter *adapter)
> if (!leds)
> return -ENOMEM;
>
> - for (i = 0; i < IGC_NUM_LEDS; i++)
> - igc_setup_ldev(leds + i, netdev, i);
> + for (i = 0; i < IGC_NUM_LEDS; i++) {
> + ret = igc_setup_ldev(leds + i, netdev, i);
> + if (ret)
> + goto err;
> + }
> +
> + adapter->leds = leds;
>
> return 0;
> +
> +err:
> + for (i--; i >= 0; i--)
> + led_classdev_unregister(&((leds + i)->led));
> + return ret;
> +}
> +
> +void igc_led_teardown(struct igc_adapter *adapter)
> +{
> + struct igc_led_classdev *leds = adapter->leds;
> + int i;
> +
> + for (i = 0; i < IGC_NUM_LEDS; i++)
> + led_classdev_unregister(&((leds + i)->led));
> }
> diff --git a/drivers/net/ethernet/intel/igc/igc_main.c b/drivers/net/ethernet/intel/igc/igc_main.c
> index 2e1cfbd82f4f..cd164442ab35 100644
> --- a/drivers/net/ethernet/intel/igc/igc_main.c
> +++ b/drivers/net/ethernet/intel/igc/igc_main.c
> @@ -7025,6 +7025,9 @@ static void igc_remove(struct pci_dev *pdev)
> cancel_work_sync(&adapter->watchdog_task);
> hrtimer_cancel(&adapter->hrtimer);
>
> + if (IS_ENABLED(CONFIG_IGC_LEDS))
> + igc_led_teardown(adapter);
> +
> /* Release control of h/w to f/w. If f/w is AMT enabled, this
> * would have already happened in close and is redundant.
> */
>
prev parent reply other threads:[~2024-04-05 20:17 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <CAEhC_B=ksywxCG_+aQqXUrGEgKq+4mqnSV8EBHOKbC3-Obj9+Q@mail.gmail.com>
2024-04-05 10:02 ` [Intel-wired-lan] Deadlock in pciehp on dock disconnect Lukas Wunner
2024-04-05 13:31 ` Heiner Kallweit
2024-04-05 17:48 ` Lukas Wunner
2024-04-05 19:01 ` Heiner Kallweit
2024-04-05 19:16 ` Lukas Wunner
2024-04-05 20:16 ` Heiner Kallweit [this message]
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=4c11ca11-8585-4cb5-bd0e-cd67eb1c208f@gmail.com \
--to=hkallweit1@gmail.com \
--cc=ansuelsmth@gmail.com \
--cc=anthony.l.nguyen@intel.com \
--cc=bhelgaas@google.com \
--cc=dave.hansen@linux.intel.com \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=intel-wired-lan@lists.osuosl.org \
--cc=kuba@kernel.org \
--cc=kurt@linutronix.de \
--cc=linux-pci@vger.kernel.org \
--cc=lozko.roma@gmail.com \
--cc=lukas@wunner.de \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=seanjc@google.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).