On Mon, Aug 5, 2019 at 4:34 PM Oliver Neukum <oneukum@xxxxxxxx> wrote: > > Am Montag, den 05.08.2019, 05:38 -0700 schrieb syzbot: > > Hello, > > > > syzbot found the following crash on: > > > > HEAD commit: e96407b4 usb-fuzzer: main usb gadget fuzzer driver > > git tree: https://github.com/google/kasan.git usb-fuzzer > > console output: https://syzkaller.appspot.com/x/log.txt?x=144c21dc600000 > > kernel config: https://syzkaller.appspot.com/x/.config?x=cfa2c18fb6a8068e > > dashboard link: https://syzkaller.appspot.com/bug?extid=94e2b9e9c7d1dd332345 > > compiler: gcc (GCC) 9.0.0 20181231 (experimental) > > syz repro: https://syzkaller.appspot.com/x/repro.syz?x=169e8542600000 > > C reproducer: https://syzkaller.appspot.com/x/repro.c?x=10ec8262600000 > > > > IMPORTANT: if you fix the bug, please add the following tag to the commit: > > Reported-by: syzbot+94e2b9e9c7d1dd332345@xxxxxxxxxxxxxxxxxxxxxxxxx > > > > usb 1-1: config 0 interface 0 altsetting 0 has 1 endpoint descriptor, > > different from the interface descriptor's value: 9 > > usb 1-1: New USB device found, idVendor=046d, idProduct=c298, bcdDevice= > > 0.00 > > usb 1-1: New USB device strings: Mfr=0, Product=0, SerialNumber=0 > > usb 1-1: config 0 descriptor?? > > logitech 0003:046D:C298.0001: unknown main item tag 0x0 > > logitech 0003:046D:C298.0001: unknown main item tag 0x0 > > logitech 0003:046D:C298.0001: hidraw0: USB HID v0.00 Device [HID 046d:c298] > > on usb-dummy_hcd.0-1/input0 > > BUG: KASAN: slab-out-of-bounds in set_bit > > include/asm-generic/bitops-instrumented.h:28 [inline] > > #syz test: https://github.com/google/kasan.git e96407b4 > > From 7e7f8ce9108b69613f8bb4ff2f95c258e22c3228 Mon Sep 17 00:00:00 2001 > From: Oliver Neukum <oneukum@xxxxxxxx> > Date: Mon, 5 Aug 2019 16:14:47 +0200 > Subject: [PATCH] hid-lg4ff: sanity check for offsets of FF effects > > Malicious devices could provide huge offsets which would lead > to setting bits in random kernel memory. Adding a sanity check. > > Reported-by: syzbot+94e2b9e9c7d1dd332345@xxxxxxxxxxxxxxxxxxxxxxxxx > Signed-off-by: Oliver Neukum <oneukum@xxxxxxxx> > --- > drivers/hid/hid-lg4ff.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/drivers/hid/hid-lg4ff.c b/drivers/hid/hid-lg4ff.c > index cefba038520c..f9572750d889 100644 > --- a/drivers/hid/hid-lg4ff.c > +++ b/drivers/hid/hid-lg4ff.c > @@ -1327,8 +1327,12 @@ int lg4ff_init(struct hid_device *hid) > } > > /* Set supported force feedback capabilities */ > + error = -ENODEV; > for (j = 0; lg4ff_devices[i].ff_effects[j] >= 0; j++) > - set_bit(lg4ff_devices[i].ff_effects[j], dev->ffbit); > + if (lg4ff_devices[i].ff_effects[j] <= 15) Can't ff_effects have one of the FF_CONSTANT, FF_PERIODIC, etc. values? Those are 0x50, 0x51, ... Or maybe I'm just misunderstanding something. Are those ff_effects provided by the device? > + set_bit(lg4ff_devices[i].ff_effects[j], dev->ffbit); > + else > + goto err_init; > > error = input_ff_create_memless(dev, NULL, lg4ff_play); > > -- > 2.16.4 >
