Re: [PATCH] usbhid: fix array-index-out-of-bounds in usbhid_parse UBSAN warning

From: Aleksandr Nogikh
Date: Wed Jan 03 2024 - 09:30:05 EST

Next message: Maxime Chevallier: "[PATCH ethtool-next 0/3] ethtool: Introduce PHY listing and targetting"
Previous message: Alex Deucher: "Re: [PATCH v2 00/14] Unified cross-architecture kernel-mode FPU API"
In reply to: Dan Carpenter: "Re: [PATCH] usbhid: fix array-index-out-of-bounds in usbhid_parse UBSAN warning"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Hi Dan,

In this particular case syzbot just forwarded a user's patch testing
request to the LKML. I think there's not much value in kernel test
robot analyzing such emails.

--
Aleksandr

On Wed, Jan 3, 2024 at 3:12 PM Dan Carpenter <dan.carpenter@xxxxxxxxxx> wrote:
>
> Hi syzbot,
>
> kernel test robot noticed the following build warnings:
>
> https://git-scm.com/docs/git-format-patch#_base_tree_information]
>
> url: https://github.com/intel-lab-lkp/linux/commits/syzbot/usbhid-fix-array-index-out-of-bounds-in-usbhid_parse-UBSAN-warning/20231225-153341
> base: https://git.kernel.org/pub/scm/linux/kernel/git/hid/hid.git for-next
> patch link: https://lore.kernel.org/r/0000000000009ae37b060d32c643%40google.com
> patch subject: [PATCH] usbhid: fix array-index-out-of-bounds in usbhid_parse UBSAN warning
> config: x86_64-randconfig-161-20231225 (https://download.01.org/0day-ci/archive/20231226/202312260900.gRDPofL9-lkp@xxxxxxxxx/config)
> compiler: clang version 16.0.4 (https://github.com/llvm/llvm-project.git ae42196bc493ffe877a7e3dff8be32035dea4d07)
>
> If you fix the issue in a separate patch/commit (i.e. not just a new version of
> the same patch/commit), kindly add following tags
> | Reported-by: kernel test robot <lkp@xxxxxxxxx>
> | Reported-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
> | Closes: https://lore.kernel.org/r/202312260900.gRDPofL9-lkp@xxxxxxxxx/
>
> smatch warnings:
> drivers/hid/usbhid/hid-core.c:1026 usbhid_parse() warn: curly braces intended?
> drivers/hid/usbhid/hid-core.c:1029 usbhid_parse() warn: inconsistent indenting
>
> vim +1026 drivers/hid/usbhid/hid-core.c
>
> c500c9714011ed drivers/hid/usbhid/hid-core.c Jiri Slaby 2008-05-16 979 static int usbhid_parse(struct hid_device *hid)
> c500c9714011ed drivers/hid/usbhid/hid-core.c Jiri Slaby 2008-05-16 980 {
> c500c9714011ed drivers/hid/usbhid/hid-core.c Jiri Slaby 2008-05-16 981 struct usb_interface *intf = to_usb_interface(hid->dev.parent);
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 982 struct usb_host_interface *interface = intf->cur_altsetting;
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 983 struct usb_device *dev = interface_to_usbdev (intf);
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 984 struct hid_descriptor *hdesc;
> 2eb5dc30eb87aa drivers/hid/usbhid/hid-core.c Paul Walmsley 2007-04-19 985 u32 quirks = 0;
> c500c9714011ed drivers/hid/usbhid/hid-core.c Jiri Slaby 2008-05-16 986 unsigned int rsize = 0;
> c5b7c7c395a34f drivers/usb/input/hid-core.c Dmitry Torokhov 2005-09-15 987 char *rdesc;
> c500c9714011ed drivers/hid/usbhid/hid-core.c Jiri Slaby 2008-05-16 988 int ret, n;
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 989 int num_descriptors;
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 990 size_t offset = offsetof(struct hid_descriptor, desc);
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 991
> d5d3e202753cc0 drivers/hid/usbhid/hid-core.c Benjamin Tissoires 2017-11-20 992 quirks = hid_lookup_quirk(hid);
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 993
> 6f4303fb2ec680 drivers/hid/usbhid/hid-core.c Jiri Kosina 2009-01-29 994 if (quirks & HID_QUIRK_IGNORE)
> 6f4303fb2ec680 drivers/hid/usbhid/hid-core.c Jiri Kosina 2009-01-29 995 return -ENODEV;
> 6f4303fb2ec680 drivers/hid/usbhid/hid-core.c Jiri Kosina 2009-01-29 996
> 0f28b55db54300 drivers/usb/input/hid-core.c Alan Stern 2006-05-15 997 /* Many keyboards and mice don't like to be polled for reports,
> 0f28b55db54300 drivers/usb/input/hid-core.c Alan Stern 2006-05-15 998 * so we will always set the HID_QUIRK_NOGET flag for them. */
> 0f28b55db54300 drivers/usb/input/hid-core.c Alan Stern 2006-05-15 999 if (interface->desc.bInterfaceSubClass == USB_INTERFACE_SUBCLASS_BOOT) {
> 0f28b55db54300 drivers/usb/input/hid-core.c Alan Stern 2006-05-15 1000 if (interface->desc.bInterfaceProtocol == USB_INTERFACE_PROTOCOL_KEYBOARD ||
> 0f28b55db54300 drivers/usb/input/hid-core.c Alan Stern 2006-05-15 1001 interface->desc.bInterfaceProtocol == USB_INTERFACE_PROTOCOL_MOUSE)
> 0f28b55db54300 drivers/usb/input/hid-core.c Alan Stern 2006-05-15 1002 quirks |= HID_QUIRK_NOGET;
> 0f28b55db54300 drivers/usb/input/hid-core.c Alan Stern 2006-05-15 1003 }
> 0f28b55db54300 drivers/usb/input/hid-core.c Alan Stern 2006-05-15 1004
> c5b7c7c395a34f drivers/usb/input/hid-core.c Dmitry Torokhov 2005-09-15 1005 if (usb_get_extra_descriptor(interface, HID_DT_HID, &hdesc) &&
> c5b7c7c395a34f drivers/usb/input/hid-core.c Dmitry Torokhov 2005-09-15 1006 (!interface->desc.bNumEndpoints ||
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 1007 usb_get_extra_descriptor(&interface->endpoint[0], HID_DT_HID, &hdesc))) {
> 58037eb961f859 drivers/hid/usbhid/hid-core.c Jiri Kosina 2007-05-30 1008 dbg_hid("class descriptor not present\n");
> c500c9714011ed drivers/hid/usbhid/hid-core.c Jiri Slaby 2008-05-16 1009 return -ENODEV;
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 1010 }
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 1011
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 1012 if (hdesc->bLength < sizeof(struct hid_descriptor)) {
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 1013 dbg_hid("hid descriptor is too short\n");
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 1014 return -EINVAL;
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 1015 }
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 1016
> c500c9714011ed drivers/hid/usbhid/hid-core.c Jiri Slaby 2008-05-16 1017 hid->version = le16_to_cpu(hdesc->bcdHID);
> c500c9714011ed drivers/hid/usbhid/hid-core.c Jiri Slaby 2008-05-16 1018 hid->country = hdesc->bCountryCode;
> c500c9714011ed drivers/hid/usbhid/hid-core.c Jiri Slaby 2008-05-16 1019
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 1020 num_descriptors = min_t(int, hdesc->bNumDescriptors,
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 1021 (hdesc->bLength - offset) / sizeof(struct hid_class_descriptor));
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 1022
> f043bfc98c193c drivers/hid/usbhid/hid-core.c Jaejoong Kim 2017-09-28 1023 for (n = 0; n < num_descriptors; n++)
>
> This for loop needs curly braces now.
>
> d3e0d5b253c73b drivers/hid/usbhid/hid-core.c syzbot 2023-12-23 1024 if (n >= ARRAY_SIZE(hdesc->desc))
> d3e0d5b253c73b drivers/hid/usbhid/hid-core.c syzbot 2023-12-23 1025 break;
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 @1026 if (hdesc->desc[n].bDescriptorType == HID_DT_REPORT)
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 1027 rsize = le16_to_cpu(hdesc->desc[n].wDescriptorLength);
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 1028
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 @1029 if (!rsize || rsize > HID_MAX_DESCRIPTOR_SIZE) {
> 58037eb961f859 drivers/hid/usbhid/hid-core.c Jiri Kosina 2007-05-30 1030 dbg_hid("weird size of report descriptor (%u)\n", rsize);
> c500c9714011ed drivers/hid/usbhid/hid-core.c Jiri Slaby 2008-05-16 1031 return -EINVAL;
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 1032 }
> ^1da177e4c3f41 drivers/usb/input/hid-core.c Linus Torvalds 2005-04-16 1033
> 52150c78270db5 drivers/hid/usbhid/hid-core.c Joe Perches 2017-03-01 1034 rdesc = kmalloc(rsize, GFP_KERNEL);
>
> --
> 0-DAY CI Kernel Test Service
> https://github.com/intel/lkp-tests/wiki
>
> --
> You received this message because you are subscribed to the Google Groups "syzkaller-bugs" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to syzkaller-bugs+unsubscribe@xxxxxxxxxxxxxxxx.
> To view this discussion on the web visit https://groups.google.com/d/msgid/syzkaller-bugs/5e68be46-caab-40f4-8e0f-543566fd7c28%40moroto.mountain.

Next message: Maxime Chevallier: "[PATCH ethtool-next 0/3] ethtool: Introduce PHY listing and targetting"
Previous message: Alex Deucher: "Re: [PATCH v2 00/14] Unified cross-architecture kernel-mode FPU API"
In reply to: Dan Carpenter: "Re: [PATCH] usbhid: fix array-index-out-of-bounds in usbhid_parse UBSAN warning"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]