linux-kernel - RE: [PATCH] drm/hyperv: Fix an error handling path in hyperv_vmbus

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]

Date:   Fri, 5 Aug 2022 18:35:01 +0000
From:   "Michael Kelley (LINUX)" <mikelley@...rosoft.com>
To:     Christophe JAILLET <christophe.jaillet@...adoo.fr>,
        Deepak Rawat <drawat.floss@...il.com>,
        David Airlie <airlied@...ux.ie>,
        Daniel Vetter <daniel@...ll.ch>,
        Thomas Zimmermann <tzimmermann@...e.de>
CC:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "kernel-janitors@...r.kernel.org" <kernel-janitors@...r.kernel.org>,
        "linux-hyperv@...r.kernel.org" <linux-hyperv@...r.kernel.org>,
        "dri-devel@...ts.freedesktop.org" <dri-devel@...ts.freedesktop.org>
Subject: RE: [PATCH] drm/hyperv: Fix an error handling path in
 hyperv_vmbus_probe()

From: Christophe JAILLET <christophe.jaillet@...adoo.fr> Sent: Sunday, July 31, 2022 1:02 PM
> 
> hyperv_setup_vram() calls vmbus_allocate_mmio().
> This must be undone in the error handling path of the probe, as already
> done in the remove function.
> 
> This patch depends on commit a0ab5abced55 ("drm/hyperv : Removing the
> restruction of VRAM allocation with PCI bar size").
> Without it, something like what is done in commit e048834c209a
> ("drm/hyperv: Fix device removal on Gen1 VMs") should be done.

Should the above paragraph be below the '---' as a comment, rather than
part of the commit message?  It's more about staging instructions than a
long-term record of the actual functional/code change.

> 
> Fixes: 76c56a5affeb ("drm/hyperv: Add DRM driver for hyperv synthetic video device")

I wonder if the Fixes: dependency should be on a0ab5abced55.  As you noted,
this patch won't apply cleanly on stable kernel versions that lack that commit,
so we'll need a separate patch for stable if we want to make the fix there.

> Signed-off-by: Christophe JAILLET <christophe.jaillet@...adoo.fr>

All that said, the fix looks good, so

Reviewed-by: Michael Kelley <mikelley@...rosoft.com>

> ---
>  drivers/gpu/drm/hyperv/hyperv_drm_drv.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/gpu/drm/hyperv/hyperv_drm_drv.c
> b/drivers/gpu/drm/hyperv/hyperv_drm_drv.c
> index 6d11e7938c83..fc8b4e045f5d 100644
> --- a/drivers/gpu/drm/hyperv/hyperv_drm_drv.c
> +++ b/drivers/gpu/drm/hyperv/hyperv_drm_drv.c
> @@ -133,7 +133,6 @@ static int hyperv_vmbus_probe(struct hv_device *hdev,
>  	}
> 
>  	ret = hyperv_setup_vram(hv, hdev);
> -
>  	if (ret)
>  		goto err_vmbus_close;
> 
> @@ -150,18 +149,20 @@ static int hyperv_vmbus_probe(struct hv_device *hdev,
> 
>  	ret = hyperv_mode_config_init(hv);
>  	if (ret)
> -		goto err_vmbus_close;
> +		goto err_free_mmio;
> 
>  	ret = drm_dev_register(dev, 0);
>  	if (ret) {
>  		drm_err(dev, "Failed to register drm driver.\n");
> -		goto err_vmbus_close;
> +		goto err_free_mmio;
>  	}
> 
>  	drm_fbdev_generic_setup(dev, 0);
> 
>  	return 0;
> 
> +err_free_mmio:
> +	vmbus_free_mmio(hv->mem->start, hv->fb_size);
>  err_vmbus_close:
>  	vmbus_close(hdev->channel);
>  err_hv_set_drv_data:
> --
> 2.34.1