2013-09-10 23:33:06

by Ryan Mallon

Subject: [PATCH 1/2] uinput: Support injecting multiple events in one write() call

Rework the code in uinput_inject_event so that it matches the code in
evdev_write and allows injecting more than one event, or zero events.

Signed-off-by: Ryan Mallon <[email protected]>
---
drivers/input/misc/uinput.c | 14 +++++++++-----
1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/drivers/input/misc/uinput.c b/drivers/input/misc/uinput.c
index a0a4bba..6aea346 100644
--- a/drivers/input/misc/uinput.c
+++ b/drivers/input/misc/uinput.c
@@ -434,16 +434,20 @@ static ssize_t uinput_inject_event(struct uinput_device *udev,
const char __user *buffer, size_t count)
{
struct input_event ev;
+ size_t bytes = 0;

- if (count < input_event_size())
+ if (count != 0 && count < input_event_size())
return -EINVAL;

- if (input_event_from_user(buffer, &ev))
- return -EFAULT;
+ while (bytes + input_event_size() <= count) {
+ if (input_event_from_user(buffer + bytes, &ev))
+ return -EFAULT;

- input_event(udev->dev, ev.type, ev.code, ev.value);
+ input_event(udev->dev, ev.type, ev.code, ev.value);
+ bytes += input_event_size();
+ }

- return input_event_size();
+ return bytes;
}

static ssize_t uinput_write(struct file *file, const char __user *buffer,
--
1.7.9.7

2013-09-10 23:33:10

by Ryan Mallon

Subject: [PATCH 2/2] uinput: Use input_inject_event in uinput_inject_event

Call input_inject_event rather than input_event in uinput_inject_event. This
mirrors the behaviour of evdev_write. input_inject_event will ignore the
injected event if the uinput input device has been grabbed for exclusive
access by a handler other than uinput.

Signed-off-by: Ryan Mallon <[email protected]>
---
drivers/input/misc/uinput.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/input/misc/uinput.c b/drivers/input/misc/uinput.c
index 6aea346..0962ecb 100644
--- a/drivers/input/misc/uinput.c
+++ b/drivers/input/misc/uinput.c
@@ -443,7 +443,7 @@ static ssize_t uinput_inject_event(struct uinput_device *udev,
if (input_event_from_user(buffer + bytes, &ev))
return -EFAULT;

- input_event(udev->dev, ev.type, ev.code, ev.value);
+ input_inject_event(udev->dev, ev.type, ev.code, ev.value);
bytes += input_event_size();
}

--
1.7.9.7

2013-09-11 00:12:42

by Dmitry Torokhov

Subject: Re: [PATCH 2/2] uinput: Use input_inject_event in uinput_inject_event

Hi Ryan,

On Wed, Sep 11, 2013 at 09:32:53AM +1000, Ryan Mallon wrote:
> Call input_inject_event rather than input_event in uinput_inject_event. This
> mirrors the behaviour of evdev_write. input_inject_event will ignore the
> injected event if the uinput input device has been grabbed for exclusive
> access by a handler other than uinput.

uinput is not an input handler but input driver and therefore should
continue using input_event().

Thanks.

--
Dmitry

2013-09-11 00:14:33

by Dmitry Torokhov

Subject: Re: [PATCH 1/2] uinput: Support injecting multiple events in one write() call

Hi Ryan,

On Wed, Sep 11, 2013 at 09:32:52AM +1000, Ryan Mallon wrote:
> Rework the code in uinput_inject_event so that it matches the code in
> evdev_write and allows injecting more than one event, or zero events.
>
> Signed-off-by: Ryan Mallon <[email protected]>
> ---
> drivers/input/misc/uinput.c | 14 +++++++++-----
> 1 file changed, 9 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/input/misc/uinput.c b/drivers/input/misc/uinput.c
> index a0a4bba..6aea346 100644
> --- a/drivers/input/misc/uinput.c
> +++ b/drivers/input/misc/uinput.c
> @@ -434,16 +434,20 @@ static ssize_t uinput_inject_event(struct uinput_device *udev,
> const char __user *buffer, size_t count)
> {
> struct input_event ev;
> + size_t bytes = 0;
>
> - if (count < input_event_size())
> + if (count != 0 && count < input_event_size())
> return -EINVAL;
>
> - if (input_event_from_user(buffer, &ev))
> - return -EFAULT;
> + while (bytes + input_event_size() <= count) {
> + if (input_event_from_user(buffer + bytes, &ev))
> + return -EFAULT;

Unless the failure is in the very first packet we should tell the user
how many bytes we have transferred successfully instead of returning
-EFAULT.

Thanks.

--
Dmitry

2013-09-11 00:26:16

by Ryan Mallon

Subject: Re: [PATCH 2/2] uinput: Use input_inject_event in uinput_inject_event

On 11/09/13 10:12, Dmitry Torokhov wrote:
> Hi Ryan,
>
> On Wed, Sep 11, 2013 at 09:32:53AM +1000, Ryan Mallon wrote:
>> Call input_inject_event rather than input_event in uinput_inject_event. This
>> mirrors the behaviour of evdev_write. input_inject_event will ignore the
>> injected event if the uinput input device has been grabbed for exclusive
>> access by a handler other than uinput.
>
> uinput is not an input handler but input driver and therefore should
> continue using input_event().

Ah, I wasn't sure about this, which is why I broke it out into a
separate patch. Sorry for the noise.

~Ryan

2013-09-11 00:29:46

by Ryan Mallon

Subject: Re: [PATCH 1/2] uinput: Support injecting multiple events in one write() call

On 11/09/13 10:14, Dmitry Torokhov wrote:
> Hi Ryan,
>
> On Wed, Sep 11, 2013 at 09:32:52AM +1000, Ryan Mallon wrote:
>> Rework the code in uinput_inject_event so that it matches the code in
>> evdev_write and allows injecting more than one event, or zero events.
>>
>> Signed-off-by: Ryan Mallon <[email protected]>
>> ---
>> drivers/input/misc/uinput.c | 14 +++++++++-----
>> 1 file changed, 9 insertions(+), 5 deletions(-)
>>
>> diff --git a/drivers/input/misc/uinput.c b/drivers/input/misc/uinput.c
>> index a0a4bba..6aea346 100644
>> --- a/drivers/input/misc/uinput.c
>> +++ b/drivers/input/misc/uinput.c
>> @@ -434,16 +434,20 @@ static ssize_t uinput_inject_event(struct uinput_device *udev,
>> const char __user *buffer, size_t count)
>> {
>> struct input_event ev;
>> + size_t bytes = 0;
>>
>> - if (count < input_event_size())
>> + if (count != 0 && count < input_event_size())
>> return -EINVAL;
>>
>> - if (input_event_from_user(buffer, &ev))
>> - return -EFAULT;
>> + while (bytes + input_event_size() <= count) {
>> + if (input_event_from_user(buffer + bytes, &ev))
>> + return -EFAULT;
> Unless the failure is in the very first packet we should tell the user
> how many bytes we have transferred successfully instead of returning
> -EFAULT.

Good point. evdev is also broken in this regard, I'll send a follow-up
patch for that once you are happy with this one. Updated patch is below:

~Ryan

---
diff --git a/drivers/input/misc/uinput.c b/drivers/input/misc/uinput.c
index a0a4bba..92ca2c5 100644
--- a/drivers/input/misc/uinput.c
+++ b/drivers/input/misc/uinput.c
@@ -434,16 +434,20 @@ static ssize_t uinput_inject_event(struct uinput_device *u
const char __user *buffer, size_t count)
{
struct input_event ev;
+ size_t bytes = 0;

- if (count < input_event_size())
+ if (count != 0 && count < input_event_size())
return -EINVAL;

- if (input_event_from_user(buffer, &ev))
- return -EFAULT;
+ while (bytes + input_event_size() <= count) {
+ if (input_event_from_user(buffer + bytes, &ev))
+ return bytes ?: -EFAULT;

- input_event(udev->dev, ev.type, ev.code, ev.value);
+ input_event(udev->dev, ev.type, ev.code, ev.value);
+ bytes += input_event_size();
+ }

- return input_event_size();
+ return bytes;
}

static ssize_t uinput_write(struct file *file, const char __user *buffer,