netdev - Re: [PATCH net] dpaa2-ptp: fix double free of the ptp

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: <20191214213005.701756d0@cakuba.netronome.com>
Date:   Sat, 14 Dec 2019 21:30:05 -0800
From:   Jakub Kicinski <jakub.kicinski@...ronome.com>
To:     Ioana Ciornei <ioana.ciornei@....com>
Cc:     davem@...emloft.net, netdev@...r.kernel.org,
        Yangbo Lu <yangbo.lu@....com>
Subject: Re: [PATCH net] dpaa2-ptp: fix double free of the ptp_qoriq IRQ

On Fri, 13 Dec 2019 12:04:22 +0200, Ioana Ciornei wrote:
> Upon reusing the ptp_qoriq driver, the ptp_qoriq_free() function was
> used on the remove path to free any allocated resources.
> The ptp_qoriq IRQ is among these resources that are freed in
> ptp_qoriq_free() even though it is also a managed one (allocated using
> devm_request_threaded_irq).
> 
> Drop the resource managed version of requesting the IRQ in order to not
> trigger a double free of the interrupt as below:
> 
> [  226.731005] Trying to free already-free IRQ 126
> [  226.735533] WARNING: CPU: 6 PID: 749 at kernel/irq/manage.c:1707
> __free_irq+0x9c/0x2b8
> [  226.743435] Modules linked in:
> [  226.746480] CPU: 6 PID: 749 Comm: bash Tainted: G        W
> 5.4.0-03629-gfd7102c32b2c-dirty #912
> [  226.755857] Hardware name: NXP Layerscape LX2160ARDB (DT)
> [  226.761244] pstate: 40000085 (nZcv daIf -PAN -UAO)
> [  226.766022] pc : __free_irq+0x9c/0x2b8
> [  226.769758] lr : __free_irq+0x9c/0x2b8
> [  226.773493] sp : ffff8000125039f0
> (...)
> [  226.856275] Call trace:
> [  226.858710]  __free_irq+0x9c/0x2b8
> [  226.862098]  free_irq+0x30/0x70
> [  226.865229]  devm_irq_release+0x14/0x20
> [  226.869054]  release_nodes+0x1b0/0x220
> [  226.872790]  devres_release_all+0x34/0x50
> [  226.876790]  device_release_driver_internal+0x100/0x1c0
> 
> Fixes: d346c9e86d86 ("dpaa2-ptp: reuse ptp_qoriq driver")
> Cc: Yangbo Lu <yangbo.lu@....com>
> Signed-off-by: Ioana Ciornei <ioana.ciornei@....com>
> ---
>  drivers/net/ethernet/freescale/dpaa2/dpaa2-ptp.c | 12 +++++++-----
>  1 file changed, 7 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-ptp.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-ptp.c
> index a9503aea527f..04a4b316f1dc 100644
> --- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-ptp.c
> +++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-ptp.c
> @@ -160,10 +160,10 @@ static int dpaa2_ptp_probe(struct fsl_mc_device *mc_dev)
>  	irq = mc_dev->irqs[0];
>  	ptp_qoriq->irq = irq->msi_desc->irq;
>  
> -	err = devm_request_threaded_irq(dev, ptp_qoriq->irq, NULL,
> -					dpaa2_ptp_irq_handler_thread,
> -					IRQF_NO_SUSPEND | IRQF_ONESHOT,
> -					dev_name(dev), ptp_qoriq);
> +	err = request_threaded_irq(ptp_qoriq->irq, NULL,
> +				   dpaa2_ptp_irq_handler_thread,
> +				   IRQF_NO_SUSPEND | IRQF_ONESHOT,
> +				   dev_name(dev), ptp_qoriq);
>  	if (err < 0) {
>  		dev_err(dev, "devm_request_threaded_irq(): %d\n", err);
>  		goto err_free_mc_irq;
> @@ -173,7 +173,7 @@ static int dpaa2_ptp_probe(struct fsl_mc_device *mc_dev)
>  				   DPRTC_IRQ_INDEX, 1);
>  	if (err < 0) {
>  		dev_err(dev, "dprtc_set_irq_enable(): %d\n", err);
> -		goto err_free_mc_irq;
> +		goto err_free_threaded_irq;
>  	}
>  
>  	err = ptp_qoriq_init(ptp_qoriq, base, &dpaa2_ptp_caps);

There is another goto right here which still jumps to err_free_mc_irq
rather than err_free_threaded_irq. Is that intentional?

> @@ -185,6 +185,8 @@ static int dpaa2_ptp_probe(struct fsl_mc_device *mc_dev)
>  
>  	return 0;
>  
> +err_free_threaded_irq:
> +	free_irq(ptp_qoriq->irq, ptp_qoriq);
>  err_free_mc_irq:
>  	fsl_mc_free_irqs(mc_dev);
>  err_unmap: