Skip to contentSkip to conversation

Input: feed more data into entropy pool

PATCH In review Dmitry Torokhov 4 participants 4 messages May 18, 2016 – May 19, 2016

Subsystems (2)
HID Input drivers
lore
Share & tools
lore

Dmitry Torokhov

author maintainer
#0 lore raw

PATCH 1/1 Input: feed more data into entropy pool

Commit 4369c64c79a22 ("Input: Send events one packet at a time") significantly reduced amount of entropy input core was feeding to the rest of the system, because only the very first event in the event block would be used as source of entropy.

With this change we will be calling add_input_randomness() for every event that is not filtered by the input core as a duplicate. In addition, all EV_SYN events are ignored.

Signed-off-by Dmitry Torokhov <[email protected]>
---
 drivers/input/input.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
drivers/input/input.c +3 -4
index b87ffbd..d95c34e 100644
@@ -153,8 +153,6 @@ static void input_pass_values(struct input_dev *dev,
 
 	rcu_read_unlock();
 
-	add_input_randomness(vals->type, vals->code, vals->value);
-
 	/* trigger auto repeat for key events */
 	if (test_bit(EV_REP, dev->evbit) && test_bit(EV_KEY, dev->evbit)) {
 		for (v = vals; v != vals + count; v++) {

@@ -371,9 +369,10 @@ static int input_get_disposition(struct input_dev *dev,
 static void input_handle_event(struct input_dev *dev,
 			       unsigned int type, unsigned int code, int value)
 {
-	int disposition;
+	int disposition = input_get_disposition(dev, type, code, &value);
 
-	disposition = input_get_disposition(dev, type, code, &value);
+	if (disposition != INPUT_IGNORE_EVENT && type != EV_SYN)
+		add_input_randomness(type, code, value);
 
 	if ((disposition & INPUT_PASS_TO_DEVICE) && dev->event)
 		dev->event(dev, type, code, value);
Signature · 5 lines
2.8.0.rc3.226.g39d4020


-- 
Dmitry
Acked-by on v1 1/1
#1 lore raw

Hi Dmitry,

Quoted 10 lines · Dmitry Torokhov

Commit 4369c64c79a22 ("Input: Send events one packet at a time") significantly reduced amount of entropy input core was feeding to the rest of the system, because only the very first event in the event block would be used as source of entropy.

With this change we will be calling add_input_randomness() for every event that is not filtered by the input core as a duplicate. In addition, all EV_SYN events are ignored.

Signed-off-by Dmitry Torokhov <[email protected]>
Acked-by Henrik Rydberg <[email protected]>

Thanks for this, it has been nagging me as well.

drivers/input/input.c static int input_get_disposition(struct input_dev *dev, · line 375

Quoted 3 lines
drivers/input/input.c +2 -1
-	disposition = input_get_disposition(dev, type, code, &value);
+	if (disposition != INPUT_IGNORE_EVENT && type != EV_SYN)
+		add_input_randomness(type, code, value);

As a comment, it is not immediately clear to me how this will affect latency, although it appears rather harmless.

Thanks,
Henrik

Acked-by on v1 1/1
#2 lore raw
In reply to Dmitry Torokhov
Quoted 10 lines · Dmitry Torokhov

Commit 4369c64c79a22 ("Input: Send events one packet at a time") significantly reduced amount of entropy input core was feeding to the rest of the system, because only the very first event in the event block would be used as source of entropy.

With this change we will be calling add_input_randomness() for every event that is not filtered by the input core as a duplicate. In addition, all EV_SYN events are ignored.

Signed-off-by Dmitry Torokhov <[email protected]>
Acked-by Benjamin Tissoires <[email protected]>

Thanks!

maintainer Acked-by on v1 1/1
#3 lore raw
In reply to Dmitry Torokhov
Quoted 10 lines · Dmitry Torokhov

Commit 4369c64c79a22 ("Input: Send events one packet at a time") significantly reduced amount of entropy input core was feeding to the rest of the system, because only the very first event in the event block would be used as source of entropy.

With this change we will be calling add_input_randomness() for every event that is not filtered by the input core as a duplicate. In addition, all EV_SYN events are ignored.

Signed-off-by Dmitry Torokhov <[email protected]>

IMO absolutely reasonable thing to do. FWIW

Acked-by Jiri Kosina <[email protected]>

Thanks,

Signature · 2 lines
Jiri Kosina
SUSE Labs