[PATCH 0/4] Do not DMA on the stack

[PATCH 0/4] Do not DMA on the stack

[Ksenija Stanojevic]

Fix error "doing DMA on the stack" by using kzalloc for buffer
allocation in rtl8192u/r8192U_core.c in the following functions:
	write_nic_byte_E()
	write_nic_byte()
	write_nic_word()
	write_nic_dword()
Issue found by smatch.

Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>

Ksenija Stanojevic (4):
  Staging: rtl8192u: Do not DMA on the stack in write_nic_byte_E()
  Staging: rtl8192u: Do not DMA on the stack in write_nic_byte()
  Staging: rtl8192u: Do not DMA on the stack in write_nic_word()
  Staging: rtl8192u: Do not DMA on the stack in write_nic_dword()

 drivers/staging/rtl8192u/r8192U_core.c | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)

-- 
1.9.1

[PATCH 1/4] Staging: rtl8192u: Do not DMA on the stack in write_nic_byte_E()

[Ksenija Stanojevic]

Fix error "doing DMA on the stack" by using kzalloc for buffer
allocation.
Issue found by smatch.

Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
---
 drivers/staging/rtl8192u/r8192U_core.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/staging/rtl8192u/r8192U_core.c b/drivers/staging/rtl8192u/r8192U_core.c
index 28b54ba..287e2ce 100644
--- a/drivers/staging/rtl8192u/r8192U_core.c
+++ b/drivers/staging/rtl8192u/r8192U_core.c
@@ -259,10 +259,15 @@ void write_nic_byte_E(struct net_device *dev, int indx, u8 data)
 	int status;
 	struct r8192_priv *priv = (struct r8192_priv *)ieee80211_priv(dev);
 	struct usb_device *udev = priv->udev;
+	u8 *pdata = kzalloc(sizeof(data), GFP_KERNEL);
+
+	if (!pdata)
+		return;
 
 	status = usb_control_msg(udev, usb_sndctrlpipe(udev, 0),
 				 RTL8187_REQ_SET_REGS, RTL8187_REQT_WRITE,
 				 indx | 0xfe00, 0, &data, 1, HZ / 2);
+	kfree(pdata);
 
 	if (status < 0)
 		netdev_err(dev, "write_nic_byte_E TimeOut! status: %d\n",
-- 
1.9.1

[PATCH 2/4] Staging: rtl8192u: Do not DMA on the stack in write_nic_byte()

[Ksenija Stanojevic]

Fix error "doing DMA on the stack" by using kzalloc for buffer
allocation.
Issue found by smatch.

Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
---
 drivers/staging/rtl8192u/r8192U_core.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/staging/rtl8192u/r8192U_core.c b/drivers/staging/rtl8192u/r8192U_core.c
index 287e2ce..94e3610 100644
--- a/drivers/staging/rtl8192u/r8192U_core.c
+++ b/drivers/staging/rtl8192u/r8192U_core.c
@@ -298,11 +298,16 @@ void write_nic_byte(struct net_device *dev, int indx, u8 data)
 
 	struct r8192_priv *priv = (struct r8192_priv *)ieee80211_priv(dev);
 	struct usb_device *udev = priv->udev;
+	u8 *pdata = kzalloc(sizeof(data), GFP_KERNEL);
+
+	if (!pdata)
+		return;
 
 	status = usb_control_msg(udev, usb_sndctrlpipe(udev, 0),
 				 RTL8187_REQ_SET_REGS, RTL8187_REQT_WRITE,
 				 (indx & 0xff) | 0xff00, (indx >> 8) & 0x0f,
 				 &data, 1, HZ / 2);
+	kfree(pdata);
 
 	if (status < 0)
 		netdev_err(dev, "write_nic_byte TimeOut! status: %d\n", status);
-- 
1.9.1

[PATCH 3/4] Staging: rtl8192u: Do not DMA on the stack in write_nic_word()

[Ksenija Stanojevic]

Fix error "doing DMA on the stack" by using kzalloc for buffer
allocation.
Issue found by smatch.

Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
---
 drivers/staging/rtl8192u/r8192U_core.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/staging/rtl8192u/r8192U_core.c b/drivers/staging/rtl8192u/r8192U_core.c
index 94e3610..6d0127f 100644
--- a/drivers/staging/rtl8192u/r8192U_core.c
+++ b/drivers/staging/rtl8192u/r8192U_core.c
@@ -323,11 +323,16 @@ void write_nic_word(struct net_device *dev, int indx, u16 data)
 
 	struct r8192_priv *priv = (struct r8192_priv *)ieee80211_priv(dev);
 	struct usb_device *udev = priv->udev;
+	u16 *pdata = kzalloc(sizeof(data), GFP_KERNEL);
+
+	if (!pdata)
+		return;
 
 	status = usb_control_msg(udev, usb_sndctrlpipe(udev, 0),
 				 RTL8187_REQ_SET_REGS, RTL8187_REQT_WRITE,
 				 (indx & 0xff) | 0xff00, (indx >> 8) & 0x0f,
 				 &data, 2, HZ / 2);
+	kfree(pdata);
 
 	if (status < 0)
 		netdev_err(dev, "write_nic_word TimeOut! status: %d\n", status);
-- 
1.9.1

[PATCH 4/4] Staging: rtl8192u: Do not DMA on the stack in write_nic_dword()

[Ksenija Stanojevic]

Fix error "doing DMA on the stack" by using kzalloc for buffer
allocation.
Issue found by smatch.

Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
---
 drivers/staging/rtl8192u/r8192U_core.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/staging/rtl8192u/r8192U_core.c b/drivers/staging/rtl8192u/r8192U_core.c
index 6d0127f..a5bf8b4 100644
--- a/drivers/staging/rtl8192u/r8192U_core.c
+++ b/drivers/staging/rtl8192u/r8192U_core.c
@@ -347,11 +347,16 @@ void write_nic_dword(struct net_device *dev, int indx, u32 data)
 
 	struct r8192_priv *priv = (struct r8192_priv *)ieee80211_priv(dev);
 	struct usb_device *udev = priv->udev;
+	u32 *pdata = kzalloc(sizeof(data), GFP_KERNEL);
+
+	if (!pdata)
+		return;
 
 	status = usb_control_msg(udev, usb_sndctrlpipe(udev, 0),
 				 RTL8187_REQ_SET_REGS, RTL8187_REQT_WRITE,
 				 (indx & 0xff) | 0xff00, (indx >> 8) & 0x0f,
 				 &data, 4, HZ / 2);
+	kfree(pdata);
 
 
 	if (status < 0)
-- 
1.9.1

Re: [Outreachy kernel] [PATCH 0/4] Do not DMA on the stack

[Arnd Bergmann]

On Sunday 04 October 2015 22:13:02 Ksenija Stanojevic wrote:
> Fix error "doing DMA on the stack" by using kzalloc for buffer
> allocation in rtl8192u/r8192U_core.c in the following functions:
>         write_nic_byte_E()
>         write_nic_byte()
>         write_nic_word()
>         write_nic_dword()
> Issue found by smatch.
> 
> Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
> 
> Ksenija Stanojevic (4):
>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte_E()
>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte()
>   Staging: rtl8192u: Do not DMA on the stack in write_nic_word()
>   Staging: rtl8192u: Do not DMA on the stack in write_nic_dword()
> 
 
The changes all look wrong to me, can you guess what the problem is
when you read them again?

Also, these are four instances of the same code, in the same file,
so it would be reasonable to do a single patch, and then you can
do the other five instances in the same file as well.

	Arnd

Re: [Outreachy kernel] [PATCH 1/4] Staging: rtl8192u: Do not DMA on the stack in write_nic_byte_E()

[Greg KH]

On Sun, Oct 04, 2015 at 10:14:00PM +0200, Ksenija Stanojevic wrote:
> Fix error "doing DMA on the stack" by using kzalloc for buffer
> allocation.
> Issue found by smatch.

What does smatch say after you made your changes and you run it on the
code again?

thanks,

greg k-h

Re: [PATCH 4/4] Staging: rtl8192u: Do not DMA on the stack in write_nic_dword()

[Ksenija Stanojević]

On Sun, Oct 4, 2015 at 10:15 PM, Ksenija Stanojevic
<ksenija.stanojevic@gmail.com> wrote:
> Fix error "doing DMA on the stack" by using kzalloc for buffer
> allocation.
> Issue found by smatch.
>
> Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
> ---
>  drivers/staging/rtl8192u/r8192U_core.c | 5 +++++
>  1 file changed, 5 insertions(+)
>
> diff --git a/drivers/staging/rtl8192u/r8192U_core.c b/drivers/staging/rtl8192u/r8192U_core.c
> index 6d0127f..a5bf8b4 100644
> --- a/drivers/staging/rtl8192u/r8192U_core.c
> +++ b/drivers/staging/rtl8192u/r8192U_core.c
> @@ -347,11 +347,16 @@ void write_nic_dword(struct net_device *dev, int indx, u32 data)
>
>         struct r8192_priv *priv = (struct r8192_priv *)ieee80211_priv(dev);
>         struct usb_device *udev = priv->udev;
> +       u32 *pdata = kzalloc(sizeof(data), GFP_KERNEL);
> +
> +       if (!pdata)
> +               return;
>
>         status = usb_control_msg(udev, usb_sndctrlpipe(udev, 0),
>                                  RTL8187_REQ_SET_REGS, RTL8187_REQT_WRITE,
>                                  (indx & 0xff) | 0xff00, (indx >> 8) & 0x0f,
>                                  &data, 4, HZ / 2);

I made a mistake i should have replace &data with pdata.

> +       kfree(pdata);
>
>
>         if (status < 0)
> --
> 1.9.1
>

Re: [Outreachy kernel] [PATCH 1/4] Staging: rtl8192u: Do not DMA on the stack in write_nic_byte_E()

[Ksenija Stanojević]

On Mon, Oct 5, 2015 at 5:17 AM, Greg KH <gregkh@linuxfoundation.org> wrote:
> On Sun, Oct 04, 2015 at 10:14:00PM +0200, Ksenija Stanojevic wrote:
>> Fix error "doing DMA on the stack" by using kzalloc for buffer
>> allocation.
>> Issue found by smatch.
>
> What does smatch say after you made your changes and you run it on the
> code again?


I made a mistake i should have replace &data with pdata.
> thanks,
>
> greg k-h
>
> --
> You received this message because you are subscribed to the Google Groups "outreachy-kernel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to outreachy-kernel+unsubscribe@googlegroups.com.
> To post to this group, send email to outreachy-kernel@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/outreachy-kernel/20151005031746.GD27303%40kroah.com.
> For more options, visit https://groups.google.com/d/optout.

Re: [Outreachy kernel] [PATCH 0/4] Do not DMA on the stack

[Ksenija Stanojević]

On Sun, Oct 4, 2015 at 10:39 PM, Arnd Bergmann <arnd@arndb.de> wrote:
> On Sunday 04 October 2015 22:13:02 Ksenija Stanojevic wrote:
>> Fix error "doing DMA on the stack" by using kzalloc for buffer
>> allocation in rtl8192u/r8192U_core.c in the following functions:
>>         write_nic_byte_E()
>>         write_nic_byte()
>>         write_nic_word()
>>         write_nic_dword()
>> Issue found by smatch.
>>
>> Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
>>
>> Ksenija Stanojevic (4):
>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte_E()
>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte()
>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_word()
>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_dword()
>>
>
> The changes all look wrong to me, can you guess what the problem is
> when you read them again?

I forgot to replace &data with pdata in usb_control_msg()
>
> Also, these are four instances of the same code, in the same file,
> so it would be reasonable to do a single patch, and then you can
> do the other five instances in the same file as well.
>
>         Arnd

Re: [Outreachy kernel] [PATCH 0/4] Do not DMA on the stack

[Ksenija Stanojević]

On Mon, Oct 5, 2015 at 7:28 AM, Ksenija Stanojević
<ksenija.stanojevic@gmail.com> wrote:
> On Sun, Oct 4, 2015 at 10:39 PM, Arnd Bergmann <arnd@arndb.de> wrote:
>> On Sunday 04 October 2015 22:13:02 Ksenija Stanojevic wrote:
>>> Fix error "doing DMA on the stack" by using kzalloc for buffer
>>> allocation in rtl8192u/r8192U_core.c in the following functions:
>>>         write_nic_byte_E()
>>>         write_nic_byte()
>>>         write_nic_word()
>>>         write_nic_dword()
>>> Issue found by smatch.
>>>
>>> Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
>>>
>>> Ksenija Stanojevic (4):
>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte_E()
>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte()
>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_word()
>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_dword()
>>>
>>
>> The changes all look wrong to me, can you guess what the problem is
>> when you read them again?
>
> I forgot to replace &data with pdata in usb_control_msg()
And also assign pdata to point to data
 pdata = &data;

>>
>> Also, these are four instances of the same code, in the same file,
>> so it would be reasonable to do a single patch, and then you can
>> do the other five instances in the same file as well.
>>
>>         Arnd

Re: [Outreachy kernel] [PATCH 0/4] Do not DMA on the stack

[Ksenija Stanojević]

On Mon, Oct 5, 2015 at 8:40 AM, Ksenija Stanojević
<ksenija.stanojevic@gmail.com> wrote:
> On Mon, Oct 5, 2015 at 7:28 AM, Ksenija Stanojević
> <ksenija.stanojevic@gmail.com> wrote:
>> On Sun, Oct 4, 2015 at 10:39 PM, Arnd Bergmann <arnd@arndb.de> wrote:
>>> On Sunday 04 October 2015 22:13:02 Ksenija Stanojevic wrote:
>>>> Fix error "doing DMA on the stack" by using kzalloc for buffer
>>>> allocation in rtl8192u/r8192U_core.c in the following functions:
>>>>         write_nic_byte_E()
>>>>         write_nic_byte()
>>>>         write_nic_word()
>>>>         write_nic_dword()
>>>> Issue found by smatch.
>>>>
>>>> Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
>>>>
>>>> Ksenija Stanojevic (4):
>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte_E()
>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte()
>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_word()
>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_dword()
>>>>
>>>
>>> The changes all look wrong to me, can you guess what the problem is
>>> when you read them again?
>>
Please ignore this.
> And also assign pdata to point to data
>  pdata = &data;

thanks,
Ksenija
>>>
>>> Also, these are four instances of the same code, in the same file,
>>> so it would be reasonable to do a single patch, and then you can
>>> do the other five instances in the same file as well.
>>>
>>>         Arnd

Re: [Outreachy kernel] [PATCH 0/4] Do not DMA on the stack

[Ksenija Stanojević]

On Mon, Oct 5, 2015 at 8:44 AM, Ksenija Stanojević
<ksenija.stanojevic@gmail.com> wrote:
> On Mon, Oct 5, 2015 at 8:40 AM, Ksenija Stanojević
> <ksenija.stanojevic@gmail.com> wrote:
>> On Mon, Oct 5, 2015 at 7:28 AM, Ksenija Stanojević
>> <ksenija.stanojevic@gmail.com> wrote:
>>> On Sun, Oct 4, 2015 at 10:39 PM, Arnd Bergmann <arnd@arndb.de> wrote:
>>>> On Sunday 04 October 2015 22:13:02 Ksenija Stanojevic wrote:
>>>>> Fix error "doing DMA on the stack" by using kzalloc for buffer
>>>>> allocation in rtl8192u/r8192U_core.c in the following functions:
>>>>>         write_nic_byte_E()
>>>>>         write_nic_byte()
>>>>>         write_nic_word()
>>>>>         write_nic_dword()
>>>>> Issue found by smatch.
>>>>>
>>>>> Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
>>>>>
>>>>> Ksenija Stanojevic (4):
>>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte_E()
>>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte()
>>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_word()
>>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_dword()
>>>>>
>>>>
>>>> The changes all look wrong to me, can you guess what the problem is
>>>> when you read them again?
>>>
> Please ignore this.
>> And also assign pdata to point to data
>>  pdata = &data;

I meant *pdata = data.

> thanks,
> Ksenija
>>>>
>>>> Also, these are four instances of the same code, in the same file,
>>>> so it would be reasonable to do a single patch, and then you can
>>>> do the other five instances in the same file as well.
>>>>
>>>>         Arnd

Re: [Outreachy kernel] [PATCH 0/4] Do not DMA on the stack

[Arnd Bergmann]

On Monday 05 October 2015 08:50:09 Ksenija Stanojević wrote:
> On Mon, Oct 5, 2015 at 8:44 AM, Ksenija Stanojević
> <ksenija.stanojevic@gmail.com> wrote:
> > On Mon, Oct 5, 2015 at 8:40 AM, Ksenija Stanojević
> > <ksenija.stanojevic@gmail.com> wrote:
> >> On Mon, Oct 5, 2015 at 7:28 AM, Ksenija Stanojević
> >> <ksenija.stanojevic@gmail.com> wrote:
> >>> On Sun, Oct 4, 2015 at 10:39 PM, Arnd Bergmann <arnd@arndb.de> wrote:
> >>>> On Sunday 04 October 2015 22:13:02 Ksenija Stanojevic wrote:
> >>>>> Fix error "doing DMA on the stack" by using kzalloc for buffer
> >>>>> allocation in rtl8192u/r8192U_core.c in the following functions:
> >>>>>         write_nic_byte_E()
> >>>>>         write_nic_byte()
> >>>>>         write_nic_word()
> >>>>>         write_nic_dword()
> >>>>> Issue found by smatch.
> >>>>>
> >>>>> Signed-off-by: Ksenija Stanojevic <ksenija.stanojevic@gmail.com>
> >>>>>
> >>>>> Ksenija Stanojevic (4):
> >>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte_E()
> >>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_byte()
> >>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_word()
> >>>>>   Staging: rtl8192u: Do not DMA on the stack in write_nic_dword()
> >>>>>
> >>>>
> >>>> The changes all look wrong to me, can you guess what the problem is
> >>>> when you read them again?
> >>>
> > Please ignore this.
> >> And also assign pdata to point to data
> >>  pdata = &data;
> 
> I meant *pdata = data.

Right. Once you change that, it should be fine.

	Arnd