fix NULL pointer deference in ALPS
Subsystems (2)
PATCH 1/1 fix NULL pointer deference in ALPS
I get a NULL pointer deference in with alps while suspending.
The following patch fixes it: alps_get_model returns a pointer or NULL in case of errors, so we need to check for the results being NULL, not negative.
Since it is trivial, it is maybe a candidate for 2.6.11.2.
It does not apply to -mm since the last occurence of alps_get_model was corrected (but not the others), if needed i can send a patch for -mm as well.
regards,
Benoit
drivers/input/mouse/alps.c +3 -3
@@ -325,7 +325,7 @@ static int alps_reconnect(struct psmouseint model;unsigned char param[4 ]; -if ((model =alps_get_model (psmouse)) <0 ) +if (!(model =alps_get_model (psmouse)))return -1 ;if (model ==ALPS_MODEL_DUALPOINT &&alps_passthrough_mode (psmouse,1 ))
@@ -358,7 +358,7 @@ int alps_init(struct psmouse *psmouse)unsigned char param[4 ];int model; -if ((model =alps_get_model (psmouse)) <0 ) +if (!(model =alps_get_model (psmouse)))return -1 ; printk(KERN_INFO "ALPS Touchpad (%s) detected\n",
@@ -412,7 +412,7 @@ int alps_init(struct psmouse *psmouse)int alps_detect (struct psmouse *psmouse,int set_properties) { -if (alps_get_model (psmouse) <0 ) +if (!alps_get_model (psmouse))return -1 ;if (set_properties) {
Quoted 2 lines · Benoit Boissinot
alps_get_modelreturns a pointer or NULL in case of errors, so we need to check for the results being NULL, not negative.
2.6.11-bk2: int alps_get_model(struct psmouse *psmouse)
takes 1 argument, returns -1 on error
2.6.11-mm1: static struct alps_model_info *alps_get_model(struct psmouse *psmouse, int *version)
takes 2 arguments, returns NULL on error
drivers/input/mouse/alps.c
Quoted 11 lines
drivers/input/mouse/alps.c +3
-3-if ((model =alps_get_model (psmouse)) <0 ) +if (!(model =alps_get_model (psmouse))) -if ((model =alps_get_model (psmouse)) <0 ) +if (!(model =alps_get_model (psmouse))) -if (alps_get_model (psmouse) <0 ) +if (!alps_get_model (psmouse))
To what version of kernel this patch should be applied?
Alexey
Quoted 11 lines · Alexey Dobriyan
Quoted 2 lines · Benoit Boissinot
alps_get_modelreturns a pointer or NULL in case of errors, so we need to check for the results being NULL, not negative.2.6.11-bk2: int
alps_get_model(struct psmouse *psmouse)
takes 1 argument, returns -1 on error2.6.11-mm1: static struct
alps_model_info*alps_get_model(struct psmouse *psmouse, int *version)
takes 2 arguments, returns NULL on error
Sorry, i misreaded the vanilla code, it only applies to -mm.
Since it seems to be fixed in bk-input, please forget the patch.
Sorry.
Benoit
Signature · 1 line
powered by bash/screen/(urxvt/fvwm|linux-console)/gentoo/gnu/linux OS
Quoted 11 lines · Benoit Boissinot
I get a NULL pointer deference in with alps while suspending.
The following patch fixes it:
alps_get_modelreturns a pointer or NULL in case of errors, so we need to check for the results being NULL, not negative.Since it is trivial, it is maybe a candidate for 2.6.11.2.
It does not apply to -mm since the last occurence of
alps_get_modelwas corrected (but not the others), if needed i can send a patch for -mm as well.
I already fixed it in my tree, but feel free to push it for the sucker tree.
Quoted 38 lines
regards,
Benoit
Signed-off-by Benoit Boissinot <[email protected]>drivers/input/mouse/alps.c +3
-3@@ -325,7 +325,7 @@ static int alps_reconnect(struct psmouseint model;unsigned char param[4 ]; -if ((model =alps_get_model (psmouse)) <0 ) +if (!(model =alps_get_model (psmouse)))return -1 ;if (model ==ALPS_MODEL_DUALPOINT &&alps_passthrough_mode (psmouse,1 ))@@ -358,7 +358,7 @@ int alps_init(struct psmouse *psmouse)unsigned char param[4 ];int model; -if ((model =alps_get_model (psmouse)) <0 ) +if (!(model =alps_get_model (psmouse)))return -1 ; printk(KERN_INFO "ALPS Touchpad (%s) detected\n",@@ -412,7 +412,7 @@ int alps_init(struct psmouse *psmouse)int alps_detect (struct psmouse *psmouse,int set_properties) { -if (alps_get_model (psmouse) <0 ) +if (!alps_get_model (psmouse))return -1 ;if (set_properties) {
Signature · 2 lines
Vojtech Pavlik SuSE Labs, SuSE CR
Quoted 7 lines · Vojtech Pavlik
Quoted 6 lines · Benoit Boissinot
I get a NULL pointer deference in with alps while suspending.
The following patch fixes it:
alps_get_modelreturns a pointer or NULL in case of errors, so we need to check for the results being NULL, not negative.
Quoted 8 lines
Quoted 5 lines
Since it is trivial, it is maybe a candidate for 2.6.11.2.
It does not apply to -mm since the last occurence of
alps_get_modelwas corrected (but not the others), if needed i can send a patch for -mm as well.I already fixed it in my tree, but feel free to push it for the sucker tree.
Oops. No, 2.6.11 doesn't need that fix. Only -mm does, and it's already queued.
Signature · 2 lines
Vojtech Pavlik SuSE Labs, SuSE CR