integer overflow issue in 'appletouch' driver
in [Kernel]
|
Prev: [PATCH] workqueue: add missing __percpu markup in kernel/workqueue.c
Next: [PATCH] percpu: add missing markup in mm/memcontrol.c
From: Serge Belyshev on 7 Aug 2010 10:30 Hi! Since 2.6.34 the touchpad on my feb 2005 powerbook (mod. A1106) has stopped working. I've identified the following guilty patch: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=04b4b88cca0ebe3813b4b6f014fb6a0db380b137 > ... Testing on one of such PowerBooks I have > noticed that touchpad always generates positive values, but some of > them are greater that 127, and thus, when cast to 'signed char' being > interpreted as a negative. My device is 05ac:020e "fountain", actually generates *signed* values, thus the patch completely breaks it. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: Johannes Berg on 7 Aug 2010 11:50 On Sat, 2010-08-07 at 14:00 +0000, Serge Belyshev wrote: > Hi! > > Since 2.6.34 the touchpad on my feb 2005 powerbook (mod. A1106) has > stopped working. I've identified the following guilty patch: > > http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=04b4b88cca0ebe3813b4b6f014fb6a0db380b137 > > > ... Testing on one of such PowerBooks I have > > noticed that touchpad always generates positive values, but some of > > them are greater that 127, and thus, when cast to 'signed char' being > > interpreted as a negative. > > My device is 05ac:020e "fountain", actually generates *signed* values, > thus the patch completely breaks it. I think Ben might have the same issue. I guess we need per-touchpad functions to read the data. johannes -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: Benjamin Herrenschmidt on 8 Aug 2010 20:30 > > My device is 05ac:020e "fountain", actually generates *signed* values, > > thus the patch completely breaks it. > > I think Ben might have the same issue. I guess we need per-touchpad > functions to read the data. Yup, I confirm, same deal. This patch completely breaks it on my powerbook (same USB ID) Should we revert the commit for now ? Cheers, Ben. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: Vadim Zaliva on 8 Aug 2010 22:40 I have tested the patch on device I had I will have an access to it to retest later thus week. We can wait for my test to try to identify how my PowerBook is different from yours. Or you can roll the patch back. -- Sent from my iPhone On Aug 8, 2010, at 17:20, Benjamin Herrenschmidt <benh(a)kernel.crashing.org> wrote: > >>> My device is 05ac:020e "fountain", actually generates *signed* values, >>> thus the patch completely breaks it. >> >> I think Ben might have the same issue. I guess we need per-touchpad >> functions to read the data. > > Yup, I confirm, same deal. This patch completely breaks it on my > powerbook (same USB ID) > > Should we revert the commit for now ? > > Cheers, > Ben. > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: Benjamin Herrenschmidt on 8 Aug 2010 23:50
On Sun, 2010-08-08 at 18:24 -0700, Vadim Zaliva wrote: > I have tested the patch on device I had > I will have an access to it to retest later thus week. We can wait for my > test to try to identify how my PowerBook is different from yours. Or you > can roll the patch back. I'd rather roll the patch back for now. I'll send the revert as part of the next powerpc update. We'll sort things out then. Cheers, Ben. > > -- > Sent from my iPhone > > On Aug 8, 2010, at 17:20, Benjamin Herrenschmidt <benh(a)kernel.crashing.org> wrote: > > > > >>> My device is 05ac:020e "fountain", actually generates *signed* values, > >>> thus the patch completely breaks it. > >> > >> I think Ben might have the same issue. I guess we need per-touchpad > >> functions to read the data. > > > > Yup, I confirm, same deal. This patch completely breaks it on my > > powerbook (same USB ID) > > > > Should we revert the commit for now ? > > > > Cheers, > > Ben. > > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ |