linux-kernel - Re: [PATCH v2 7/7] dmaengine: idxd: Refactor remove call with idxd

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <4b5b45b3-76a1-4850-aeba-ff4d6777e97c@nvidia.com>
Date: Tue, 18 Feb 2025 13:01:56 -0800
From: Fenghua Yu <fenghuay@...dia.com>
To: Shuai Xue <xueshuai@...ux.alibaba.com>, vinicius.gomes@...el.com,
 dave.jiang@...el.com, vkoul@...nel.org
Cc: nikhil.rao@...el.com, dmaengine@...r.kernel.org,
 linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 7/7] dmaengine: idxd: Refactor remove call with
 idxd_cleanup() helper

Hi, Shuai,

On 2/14/25 21:44, Shuai Xue wrote:
> The idxd_cleanup() helper clean up perfmon, interrupts, internals and so

s/clean/cleans/


> on. Refactor remove call with idxd_cleanup() helper to avoid code
s/idxd_cleanup()/the idxd_cleanup()/
> duplication. Note, this also fixes the missing put_device() for idxd
> groups, enginces and wqs.
>
> Fixes: bfe1d56091c1 ("dmaengine: idxd: Init and probe for Intel data accelerators")
> Suggested-by: Vinicius Costa Gomes <vinicius.gomes@...el.com>
> Signed-off-by: Shuai Xue <xueshuai@...ux.alibaba.com>
> ---
>   drivers/dma/idxd/init.c | 13 ++-----------
>   1 file changed, 2 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/dma/idxd/init.c b/drivers/dma/idxd/init.c
> index f40f1c44a302..0fbfbe024c29 100644
> --- a/drivers/dma/idxd/init.c
> +++ b/drivers/dma/idxd/init.c
> @@ -1282,20 +1282,11 @@ static void idxd_remove(struct pci_dev *pdev)
>   	get_device(idxd_confdev(idxd));
get_device() is called here.
>   	device_unregister(idxd_confdev(idxd));
>   	idxd_shutdown(pdev);
> -	if (device_pasid_enabled(idxd))
> -		idxd_disable_system_pasid(idxd);
>   	idxd_device_remove_debugfs(idxd);
> -
> -	irq_entry = idxd_get_ie(idxd, 0);
> -	free_irq(irq_entry->vector, irq_entry);
> -	pci_free_irq_vectors(pdev);
> +	idxd_cleanup(idxd);
>   	pci_iounmap(pdev, idxd->reg_base);
> -	if (device_user_pasid_enabled(idxd))
> -		idxd_disable_sva(pdev);
> -	pci_disable_device(pdev);
> -	destroy_workqueue(idxd->wq);
> -	perfmon_pmu_remove(idxd);
>   	idxd_free(idxd);

put_device() is called inside idxd_free(). Seems not easy to read code 
to match the pair.

Plus idxd_free() is called only in non FLR case.

Maybe it's better to change this code to:

1. call put_device() outside idxd_free() so that it's easy to match the 
get_device() and put_deivce in the same level of function.

2. idxd_free() called here is OK because this is not in FLR handler. But 
only call it in non FLR path in idxd_pci_probe_alloc()

?

> +	pci_disable_device(pdev);
>   }
>   
>   static struct pci_driver idxd_pci_driver = {

Thanks.


-Fenghua