FAILED: patch "[PATCH] PCI: hv: Fix a memory leak in hv_eject_device_work()" failed to apply to 4.9-stable tree

FAILED: patch "[PATCH] PCI: hv: Fix a memory leak in hv_eject_device_work()" failed to apply to 4.9-stable tree

[gregkh]

The patch below does not apply to the 4.9-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable@vger.kernel.org>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From 05f151a73ec2b23ffbff706e5203e729a995cdc2 Mon Sep 17 00:00:00 2001
From: Dexuan Cui <decui@microsoft.com>
Date: Mon, 4 Mar 2019 21:34:48 +0000
Subject: [PATCH] PCI: hv: Fix a memory leak in hv_eject_device_work()

When a device is created in new_pcichild_device(), hpdev->refs is set
to 2 (i.e. the initial value of 1 plus the get_pcichild()).

When we hot remove the device from the host, in a Linux VM we first call
hv_pci_eject_device(), which increases hpdev->refs by get_pcichild() and
then schedules a work of hv_eject_device_work(), so hpdev->refs becomes
3 (let's ignore the paired get/put_pcichild() in other places). But in
hv_eject_device_work(), currently we only call put_pcichild() twice,
meaning the 'hpdev' struct can't be freed in put_pcichild().

Add one put_pcichild() to fix the memory leak.

The device can also be removed when we run "rmmod pci-hyperv". On this
path (hv_pci_remove() -> hv_pci_bus_exit() -> hv_pci_devices_present()),
hpdev->refs is 2, and we do correctly call put_pcichild() twice in
pci_devices_present_work().

Fixes: 4daace0d8ce8 ("PCI: hv: Add paravirtual PCI front-end for Microsoft Hyper-V VMs")
Signed-off-by: Dexuan Cui <decui@microsoft.com>
[lorenzo.pieralisi@arm.com: commit log rework]
Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
Reviewed-by:  Michael Kelley <mikelley@microsoft.com>
Cc: stable@vger.kernel.org

diff --git a/drivers/pci/controller/pci-hyperv.c b/drivers/pci/controller/pci-hyperv.c
index 95441a35eceb..30f16b882746 100644
--- a/drivers/pci/controller/pci-hyperv.c
+++ b/drivers/pci/controller/pci-hyperv.c
@@ -1900,6 +1900,9 @@ static void hv_eject_device_work(struct work_struct *work)
 			 sizeof(*ejct_pkt), (unsigned long)&ctxt.pkt,
 			 VM_PKT_DATA_INBAND, 0);
 
+	/* For the get_pcichild() in hv_pci_eject_device() */
+	put_pcichild(hpdev);
+	/* For the two refs got in new_pcichild_device() */
 	put_pcichild(hpdev);
 	put_pcichild(hpdev);
 	put_hvpcibus(hpdev->hbus);

RE: FAILED: patch "[PATCH] PCI: hv: Fix a memory leak in hv_eject_device_work()" failed to apply to 4.9-stable tree

[Dexuan Cui]

[-- Attachment #1: Type: text/plain, Size: 3032 bytes --]

> From: gregkh@linuxfoundation.org <gregkh@linuxfoundation.org>
> Sent: Wednesday, May 15, 2019 1:35 AM
> To: Dexuan Cui <decui@microsoft.com>; lorenzo.pieralisi@arm.com; Michael
> Kelley <mikelley@microsoft.com>; stephen@networkplumber.org
> Cc: stable@vger.kernel.org
> Subject: FAILED: patch "[PATCH] PCI: hv: Fix a memory leak in
> hv_eject_device_work()" failed to apply to 4.9-stable tree
> 
> 
> The patch below does not apply to the 4.9-stable tree.
> If someone wants it applied there, or to any other stable or longterm
> tree, then please email the backport, including the original git commit
> id to <stable@vger.kernel.org>.
> 
> thanks,
> 
> greg k-h
> 
> ------------------ original commit in Linus's tree ------------------
> 
> From 05f151a73ec2b23ffbff706e5203e729a995cdc2 Mon Sep 17 00:00:00
> 2001
> From: Dexuan Cui <decui@microsoft.com>
> Date: Mon, 4 Mar 2019 21:34:48 +0000
> Subject: [PATCH] PCI: hv: Fix a memory leak in hv_eject_device_work()
> 
> When a device is created in new_pcichild_device(), hpdev->refs is set
> to 2 (i.e. the initial value of 1 plus the get_pcichild()).
> 
> When we hot remove the device from the host, in a Linux VM we first call
> hv_pci_eject_device(), which increases hpdev->refs by get_pcichild() and
> then schedules a work of hv_eject_device_work(), so hpdev->refs becomes
> 3 (let's ignore the paired get/put_pcichild() in other places). But in
> hv_eject_device_work(), currently we only call put_pcichild() twice,
> meaning the 'hpdev' struct can't be freed in put_pcichild().
> 
> Add one put_pcichild() to fix the memory leak.
> 
> The device can also be removed when we run "rmmod pci-hyperv". On this
> path (hv_pci_remove() -> hv_pci_bus_exit() -> hv_pci_devices_present()),
> hpdev->refs is 2, and we do correctly call put_pcichild() twice in
> pci_devices_present_work().
> 
> Fixes: 4daace0d8ce8 ("PCI: hv: Add paravirtual PCI front-end for Microsoft
> Hyper-V VMs")
> Signed-off-by: Dexuan Cui <decui@microsoft.com>
> [lorenzo.pieralisi@arm.com: commit log rework]
> Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
> Reviewed-by:  Michael Kelley <mikelley@microsoft.com>
> Cc: stable@vger.kernel.org
> 
> diff --git a/drivers/pci/controller/pci-hyperv.c
> b/drivers/pci/controller/pci-hyperv.c
> index 95441a35eceb..30f16b882746 100644
> --- a/drivers/pci/controller/pci-hyperv.c
> +++ b/drivers/pci/controller/pci-hyperv.c
> @@ -1900,6 +1900,9 @@ static void hv_eject_device_work(struct work_struct
> *work)
>  			 sizeof(*ejct_pkt), (unsigned long)&ctxt.pkt,
>  			 VM_PKT_DATA_INBAND, 0);
> 
> +	/* For the get_pcichild() in hv_pci_eject_device() */
> +	put_pcichild(hpdev);
> +	/* For the two refs got in new_pcichild_device() */
>  	put_pcichild(hpdev);
>  	put_pcichild(hpdev);
>  	put_hvpcibus(hpdev->hbus);

Hi,
I backported the patch for linux-4.9.y. Please use the attached patch.

Thanks,
-- Dexuan

[-- Attachment #2: For-linux-4.9.y_v4.9.176_PCI-hv-Fix-a-memory-leak-in-hv_eject_device_work.patch --]
[-- Type: application/octet-stream, Size: 1004 bytes --]

From 151f2bc2afcc71ddba7b1978cb736cf13c4eeff0 Mon Sep 17 00:00:00 2001
From: Dexuan Cui <decui@microsoft.com>
Date: Wed, 15 May 2019 15:42:07 -0700
Subject: [PATCH][For v4.9.176] PCI: hv: Fix a memory leak in hv_eject_device_work()

This patch is backported for v4.9.176 from the mainline
commit 05f151a73ec2 ("PCI: hv: Fix a memory leak in hv_eject_device_work()")

Signed-off-by: Dexuan Cui <decui@microsoft.com>
---
 drivers/pci/host/pci-hyperv.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/pci/host/pci-hyperv.c b/drivers/pci/host/pci-hyperv.c
index b4d8ccfd9f7c..200b41576526 100644
--- a/drivers/pci/host/pci-hyperv.c
+++ b/drivers/pci/host/pci-hyperv.c
@@ -1620,6 +1620,7 @@ static void hv_eject_device_work(struct work_struct *work)
 	spin_unlock_irqrestore(&hpdev->hbus->device_list_lock, flags);
 
 	put_pcichild(hpdev, hv_pcidev_ref_childlist);
+	put_pcichild(hpdev, hv_pcidev_ref_initial);
 	put_pcichild(hpdev, hv_pcidev_ref_pnp);
 	put_hvpcibus(hpdev->hbus);
 }
-- 
2.17.1

Re: FAILED: patch "[PATCH] PCI: hv: Fix a memory leak in hv_eject_device_work()" failed to apply to 4.9-stable tree

[Sasha Levin]

On Wed, May 15, 2019 at 11:14:31PM +0000, Dexuan Cui wrote:
>> From: gregkh@linuxfoundation.org <gregkh@linuxfoundation.org>
>> Sent: Wednesday, May 15, 2019 1:35 AM
>> To: Dexuan Cui <decui@microsoft.com>; lorenzo.pieralisi@arm.com; Michael
>> Kelley <mikelley@microsoft.com>; stephen@networkplumber.org
>> Cc: stable@vger.kernel.org
>> Subject: FAILED: patch "[PATCH] PCI: hv: Fix a memory leak in
>> hv_eject_device_work()" failed to apply to 4.9-stable tree
>>
>>
>> The patch below does not apply to the 4.9-stable tree.
>> If someone wants it applied there, or to any other stable or longterm
>> tree, then please email the backport, including the original git commit
>> id to <stable@vger.kernel.org>.
>>
>> thanks,
>>
>> greg k-h
>>
>> ------------------ original commit in Linus's tree ------------------
>>
>> From 05f151a73ec2b23ffbff706e5203e729a995cdc2 Mon Sep 17 00:00:00
>> 2001
>> From: Dexuan Cui <decui@microsoft.com>
>> Date: Mon, 4 Mar 2019 21:34:48 +0000
>> Subject: [PATCH] PCI: hv: Fix a memory leak in hv_eject_device_work()
>>
>> When a device is created in new_pcichild_device(), hpdev->refs is set
>> to 2 (i.e. the initial value of 1 plus the get_pcichild()).
>>
>> When we hot remove the device from the host, in a Linux VM we first call
>> hv_pci_eject_device(), which increases hpdev->refs by get_pcichild() and
>> then schedules a work of hv_eject_device_work(), so hpdev->refs becomes
>> 3 (let's ignore the paired get/put_pcichild() in other places). But in
>> hv_eject_device_work(), currently we only call put_pcichild() twice,
>> meaning the 'hpdev' struct can't be freed in put_pcichild().
>>
>> Add one put_pcichild() to fix the memory leak.
>>
>> The device can also be removed when we run "rmmod pci-hyperv". On this
>> path (hv_pci_remove() -> hv_pci_bus_exit() -> hv_pci_devices_present()),
>> hpdev->refs is 2, and we do correctly call put_pcichild() twice in
>> pci_devices_present_work().
>>
>> Fixes: 4daace0d8ce8 ("PCI: hv: Add paravirtual PCI front-end for Microsoft
>> Hyper-V VMs")
>> Signed-off-by: Dexuan Cui <decui@microsoft.com>
>> [lorenzo.pieralisi@arm.com: commit log rework]
>> Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
>> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
>> Reviewed-by:  Michael Kelley <mikelley@microsoft.com>
>> Cc: stable@vger.kernel.org
>>
>> diff --git a/drivers/pci/controller/pci-hyperv.c
>> b/drivers/pci/controller/pci-hyperv.c
>> index 95441a35eceb..30f16b882746 100644
>> --- a/drivers/pci/controller/pci-hyperv.c
>> +++ b/drivers/pci/controller/pci-hyperv.c
>> @@ -1900,6 +1900,9 @@ static void hv_eject_device_work(struct work_struct
>> *work)
>>  			 sizeof(*ejct_pkt), (unsigned long)&ctxt.pkt,
>>  			 VM_PKT_DATA_INBAND, 0);
>>
>> +	/* For the get_pcichild() in hv_pci_eject_device() */
>> +	put_pcichild(hpdev);
>> +	/* For the two refs got in new_pcichild_device() */
>>  	put_pcichild(hpdev);
>>  	put_pcichild(hpdev);
>>  	put_hvpcibus(hpdev->hbus);
>
>Hi,
>I backported the patch for linux-4.9.y. Please use the attached patch.
>
>Thanks,
>-- Dexuan

Queued for 4.9, thanks.

--
Thanks,
Sasha