Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Alexander Stein <alexander.stein@ew.tq-group.com>
To:     Bjorn Helgaas <bhelgaas@google.com>,
        Korneliusz Osmenda <korneliuszo@gmail.com>
Cc:     linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] Guard pci_create_sysfs_dev_files with atomic value
Date:   Thu, 06 Oct 2022 12:08:27 +0200
Message-ID: <5610142.DvuYhMxLoT@steina-w>
Organization: TQ-Systems GmbH
In-Reply-To: <4469eba2-188b-aab7-07d1-5c77313fc42f@gmail.com>
References: <4469eba2-188b-aab7-07d1-5c77313fc42f@gmail.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 7Bit
Content-Type: text/plain; charset="us-ascii"
Precedence: bulk

Hi,

Am Montag, 29. August 2022, 21:14:43 CEST schrieb Korneliusz Osmenda:
> On Gateworks Ventana there is a number of PCI devices and:
>   - imx6_pcie_probe takes longer than start of late init
>   - pci_sysfs_init sets up flag sysfs_initialized
>   - pci_sysfs_init initializes already found devices
>   - imx6_pcie_probe tries to reinitialize device
> 
> Bug: https://bugzilla.kernel.org/show_bug.cgi?id=215515
> 
> Signed-off-by: Korneliusz Osmenda <korneliuszo@gmail.com>
> ---
>   drivers/pci/pci-sysfs.c | 6 ++++++
>   include/linux/pci.h     | 2 ++
>   2 files changed, 8 insertions(+)
> 
> diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c
> index fc804e08e3cb..a6648239e235 100644
> --- a/drivers/pci/pci-sysfs.c
> +++ b/drivers/pci/pci-sysfs.c
> @@ -1378,6 +1378,9 @@ int __must_check pci_create_sysfs_dev_files(struct
> pci_dev *pdev)
>   	if (!sysfs_initialized)
>   		return -EACCES;
> 
> +	if (atomic_cmpxchg(&pdev->sysfs_init_cnt,0,1) == 1)
> +		return 0;		/* already added */
> +
>   	return pci_create_resource_files(pdev);
>   }
> 
> @@ -1392,6 +1395,9 @@ void pci_remove_sysfs_dev_files(struct pci_dev *pdev)
>   	if (!sysfs_initialized)
>   		return;
> 
> +	if (atomic_cmpxchg(&pdev->sysfs_init_cnt,1,0) == 0)
> +		return;		/* already removed */
> +
>   	pci_remove_resource_files(pdev);
>   }
> 
> diff --git a/include/linux/pci.h b/include/linux/pci.h
> index 060af91bafcd..5477de2ef057 100644
> --- a/include/linux/pci.h
> +++ b/include/linux/pci.h
> @@ -465,6 +465,8 @@ struct pci_dev {
>   	pci_dev_flags_t dev_flags;
>   	atomic_t	enable_cnt;	/* pci_enable_device has been 
called */
> 
> +	atomic_t	sysfs_init_cnt;	/* pci_create_sysfs_dev_files has 
been called */
> +
>   	u32		saved_config_space[16]; /* Config space saved 
at suspend time */
>   	struct hlist_head saved_cap_space;
>   	int		rom_attr_enabled;	/* Display of ROM 
attribute enabled? */

Is there any feedback here? I'm hitting the same issue on TQMa6x+MBa6x.
Having this patch applied my PCIe device using several bridges is detected 
fine.

root@tqma6-common:~# lspci
00:00.0 PCI bridge: Synopsys, Inc. DWC_usb3 / PCIe bridge (rev 01)
01:00.0 PCI bridge: Pericom Semiconductor Device a303 (rev 03)
02:01.0 PCI bridge: Pericom Semiconductor Device a303 (rev 03)
02:02.0 PCI bridge: Pericom Semiconductor Device a303 (rev 03)
03:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 
PCI Express Gigabit Ethernet Controller (rev 0c)
04:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 
PCI Express Gigabit Ethernet Controller (rev 0c)

root@tqma6-common:~# lspci -t
-[0000:00]---00.0-[01-ff]----00.0-[02-04]--+-01.0-[03]----00.0
                                           \-02.0-[04]----00.0

Best regards,
Alexander