kernel/linux.git/drivers/usb, branch v4.19.77

usb: host: xhci-tegra: Set DMA mask correctly

2019-09-21T05:17:04+00:00

[ Upstream commit 993cc8753453fccfe060a535bbe21fcf1001b626 ] The Falcon microcontroller that runs the XUSB firmware and which is responsible for exposing the XHCI interface can address only 40 bits of memory. Typically that's not a problem because Tegra devices don't have enough system memory to exceed those 40 bits. However, if the ARM SMMU is enable on Tegra186 and later, the addresses passed to the XUSB controller can be anywhere in the 48-bit IOV address space of the ARM SMMU. Since the DMA/IOMMU API starts allocating from the top of the IOVA space, the Falcon microcontroller is not able to load the firmware successfully. Fix this by setting the DMA mask to 40 bits, which will force the DMA API to map the buffer for the firmware to an IOVA that is addressable by the Falcon. Signed-off-by: Nagarjuna Kristam Signed-off-by: Thierry Reding Link: https://lore.kernel.org/r/1566989697-13049-1-git-send-email-nkristam@nvidia.com Signed-off-by: Greg Kroah-Hartman Signed-off-by: Sasha Levin

USB: usbcore: Fix slab-out-of-bounds bug during device reset

2019-09-21T05:16:42+00:00

commit 3dd550a2d36596a1b0ee7955da3b611c031d3873 upstream. The syzbot fuzzer provoked a slab-out-of-bounds error in the USB core: BUG: KASAN: slab-out-of-bounds in memcmp+0xa6/0xb0 lib/string.c:904 Read of size 1 at addr ffff8881d175bed6 by task kworker/0:3/2746 CPU: 0 PID: 2746 Comm: kworker/0:3 Not tainted 5.3.0-rc5+ #28 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Workqueue: usb_hub_wq hub_event Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0xca/0x13e lib/dump_stack.c:113 print_address_description+0x6a/0x32c mm/kasan/report.c:351 __kasan_report.cold+0x1a/0x33 mm/kasan/report.c:482 kasan_report+0xe/0x12 mm/kasan/common.c:612 memcmp+0xa6/0xb0 lib/string.c:904 memcmp include/linux/string.h:400 [inline] descriptors_changed drivers/usb/core/hub.c:5579 [inline] usb_reset_and_verify_device+0x564/0x1300 drivers/usb/core/hub.c:5729 usb_reset_device+0x4c1/0x920 drivers/usb/core/hub.c:5898 rt2x00usb_probe+0x53/0x7af drivers/net/wireless/ralink/rt2x00/rt2x00usb.c:806 The error occurs when the descriptors_changed() routine (called during a device reset) attempts to compare the old and new BOS and capability descriptors. The length it uses for the comparison is the wTotalLength value stored in BOS descriptor, but this value is not necessarily the same as the length actually allocated for the descriptors. If it is larger the routine will call memcmp() with a length that is too big, thus reading beyond the end of the allocated region and leading to this fault. The kernel reads the BOS descriptor twice: first to get the total length of all the capability descriptors, and second to read it along with all those other descriptors. A malicious (or very faulty) device may send different values for the BOS descriptor fields each time. The memory area will be allocated using the wTotalLength value read the first time, but stored within it will be the value read the second time. To prevent this possibility from causing any errors, this patch modifies the BOS descriptor after it has been read the second time: It sets the wTotalLength field to the actual length of the descriptors that were read in and validated. Then the memcpy() call, or any other code using these descriptors, will be able to rely on wTotalLength being valid. Reported-and-tested-by: syzbot+35f4d916c623118d576e@syzkaller.appspotmail.com Signed-off-by: Alan Stern CC: Link: https://lore.kernel.org/r/Pine.LNX.4.44L0.1909041154260.1722-100000@iolanthe.rowland.org Signed-off-by: Greg Kroah-Hartman

usb: typec: tcpm: Try PD-2.0 if sink does not respond to 3.0 source-caps

2019-09-16T06:22:09+00:00

[ Upstream commit 976daf9d1199932df80e7b04546d1a1bd4ed5ece ] PD 2.0 sinks are supposed to accept src-capabilities with a 3.0 header and simply ignore any src PDOs which the sink does not understand such as PPS but some 2.0 sinks instead ignore the entire PD_DATA_SOURCE_CAP message, causing contract negotiation to fail. This commit fixes such sinks not working by re-trying the contract negotiation with PD-2.0 source-caps messages if we don't have a contract after PD_N_HARD_RESET_COUNT hard-reset attempts. The problem fixed by this commit was noticed with a Type-C to VGA dongle. Signed-off-by: Hans de Goede Reviewed-by: Guenter Roeck Cc: stable Signed-off-by: Greg Kroah-Hartman Signed-off-by: Sasha Levin

typec: tcpm: fix a typo in the comparison of pdo_max_voltage

2019-09-06T08:22:18+00:00

commit a684d8fd87182090ee96e34519ecdf009cef093a upstream. There appears to be a typo in the comparison of pdo_max_voltage[i] with the previous value, currently it is checking against the array pdo_min_voltage rather than pdo_max_voltage. I believe this is a typo. Fix this. Addresses-Coverity: ("Copy-paste error") Fixes: 5007e1b5db73 ("typec: tcpm: Validate source and sink caps") Cc: stable Signed-off-by: Colin Ian King Reviewed-by: Guenter Roeck Reviewed-by: Heikki Krogerus Link: https://lore.kernel.org/r/20190822135212.10195-1-colin.king@canonical.com Signed-off-by: Greg Kroah-Hartman

USB: storage: ums-realtek: Whitelist auto-delink support

2019-09-06T08:22:16+00:00

commit 1902a01e2bcc3abd7c9a18dc05e78c7ab4a53c54 upstream. Auto-delink requires writing special registers to ums-realtek devices. Unconditionally enable auto-delink may break newer devices. So only enable auto-delink by default for the original three IDs, 0x0138, 0x0158 and 0x0159. Realtek is working on a patch to properly support auto-delink for other IDs. BugLink: https://bugs.launchpad.net/bugs/1838886 Signed-off-by: Kai-Heng Feng Acked-by: Alan Stern Cc: stable Link: https://lore.kernel.org/r/20190827173450.13572-2-kai.heng.feng@canonical.com Signed-off-by: Greg Kroah-Hartman

USB: storage: ums-realtek: Update module parameter description for auto_delink_en

2019-09-06T08:22:16+00:00

commit f6445b6b2f2bb1745080af4a0926049e8bca2617 upstream. The option named "auto_delink_en" is a bit misleading, as setting it to false doesn't really disable auto-delink but let auto-delink be firmware controlled. Update the description to reflect the real usage of this parameter. Signed-off-by: Kai-Heng Feng Cc: stable Link: https://lore.kernel.org/r/20190827173450.13572-1-kai.heng.feng@canonical.com Signed-off-by: Greg Kroah-Hartman

usb: host: xhci: rcar: Fix typo in compatible string matching

2019-09-06T08:22:16+00:00

commit 636bd02a7ba9025ff851d0cfb92768c8fa865859 upstream. It's spelled "renesas", not "renensas". Due to this typo, RZ/G1M and RZ/G1N were not covered by the check. Fixes: 2dc240a3308b ("usb: host: xhci: rcar: retire use of xhci_plat_type_is()") Signed-off-by: Geert Uytterhoeven Cc: stable Reviewed-by: Yoshihiro Shimoda Link: https://lore.kernel.org/r/20190827125112.12192-1-geert+renesas@glider.be Signed-off-by: Greg Kroah-Hartman

usb: host: ohci: fix a race condition between shutdown and irq

2019-09-06T08:22:15+00:00

commit a349b95d7ca0cea71be4a7dac29830703de7eb62 upstream. This patch fixes an issue that the following error is possible to happen when ohci hardware causes an interruption and the system is shutting down at the same time. [ 34.851754] usb 2-1: USB disconnect, device number 2 [ 35.166658] irq 156: nobody cared (try booting with the "irqpoll" option) [ 35.173445] CPU: 0 PID: 22 Comm: kworker/0:1 Not tainted 5.3.0-rc5 #85 [ 35.179964] Hardware name: Renesas Salvator-X 2nd version board based on r8a77965 (DT) [ 35.187886] Workqueue: usb_hub_wq hub_event [ 35.192063] Call trace: [ 35.194509] dump_backtrace+0x0/0x150 [ 35.198165] show_stack+0x14/0x20 [ 35.201475] dump_stack+0xa0/0xc4 [ 35.204785] __report_bad_irq+0x34/0xe8 [ 35.208614] note_interrupt+0x2cc/0x318 [ 35.212446] handle_irq_event_percpu+0x5c/0x88 [ 35.216883] handle_irq_event+0x48/0x78 [ 35.220712] handle_fasteoi_irq+0xb4/0x188 [ 35.224802] generic_handle_irq+0x24/0x38 [ 35.228804] __handle_domain_irq+0x5c/0xb0 [ 35.232893] gic_handle_irq+0x58/0xa8 [ 35.236548] el1_irq+0xb8/0x180 [ 35.239681] __do_softirq+0x94/0x23c [ 35.243253] irq_exit+0xd0/0xd8 [ 35.246387] __handle_domain_irq+0x60/0xb0 [ 35.250475] gic_handle_irq+0x58/0xa8 [ 35.254130] el1_irq+0xb8/0x180 [ 35.257268] kernfs_find_ns+0x5c/0x120 [ 35.261010] kernfs_find_and_get_ns+0x3c/0x60 [ 35.265361] sysfs_unmerge_group+0x20/0x68 [ 35.269454] dpm_sysfs_remove+0x2c/0x68 [ 35.273284] device_del+0x80/0x370 [ 35.276683] hid_destroy_device+0x28/0x60 [ 35.280686] usbhid_disconnect+0x4c/0x80 [ 35.284602] usb_unbind_interface+0x6c/0x268 [ 35.288867] device_release_driver_internal+0xe4/0x1b0 [ 35.293998] device_release_driver+0x14/0x20 [ 35.298261] bus_remove_device+0x110/0x128 [ 35.302350] device_del+0x148/0x370 [ 35.305832] usb_disable_device+0x8c/0x1d0 [ 35.309921] usb_disconnect+0xc8/0x2d0 [ 35.313663] hub_event+0x6e0/0x1128 [ 35.317146] process_one_work+0x1e0/0x320 [ 35.321148] worker_thread+0x40/0x450 [ 35.324805] kthread+0x124/0x128 [ 35.328027] ret_from_fork+0x10/0x18 [ 35.331594] handlers: [ 35.333862] [<0000000079300c1d>] usb_hcd_irq [ 35.338126] [<0000000079300c1d>] usb_hcd_irq [ 35.342389] Disabling IRQ #156 ohci_shutdown() disables all the interrupt and rh_state is set to OHCI_RH_HALTED. In other hand, ohci_irq() is possible to enable OHCI_INTR_SF and OHCI_INTR_MIE on ohci_irq(). Note that OHCI_INTR_SF is possible to be set by start_ed_unlink() which is called: ohci_irq() -> process_done_list() -> takeback_td() -> start_ed_unlink() So, ohci_irq() has the following condition, the issue happens by &ohci->regs->intrenable = OHCI_INTR_MIE | OHCI_INTR_SF and ohci->rh_state = OHCI_RH_HALTED: /* interrupt for some other device? */ if (ints == 0 || unlikely(ohci->rh_state == OHCI_RH_HALTED)) return IRQ_NOTMINE; To fix the issue, ohci_shutdown() holds the spin lock while disabling the interruption and changing the rh_state flag to prevent reenable the OHCI_INTR_MIE unexpectedly. Note that io_watchdog_func() also calls the ohci_shutdown() and it already held the spin lock, so that the patch makes a new function as _ohci_shutdown(). This patch is inspired by a Renesas R-Car Gen3 BSP patch from Tho Vu. Signed-off-by: Yoshihiro Shimoda Cc: stable Acked-by: Alan Stern Link: https://lore.kernel.org/r/1566877910-6020-1-git-send-email-yoshihiro.shimoda.uh@renesas.com Signed-off-by: Greg Kroah-Hartman

usb: chipidea: udc: don't do hardware access if gadget has stopped

2019-09-06T08:22:15+00:00

commit cbe85c88ce80fb92956a0793518d415864dcead8 upstream. After _gadget_stop_activity is executed, we can consider the hardware operation for gadget has finished, and the udc can be stopped and enter low power mode. So, any later hardware operations (from usb_ep_ops APIs or usb_gadget_ops APIs) should be considered invalid, any deinitializatons has been covered at _gadget_stop_activity. I meet this problem when I plug out usb cable from PC using mass_storage gadget, my callstack like: vbus interrupt->.vbus_session-> composite_disconnect ->pm_runtime_put_sync(&_gadget->dev), the composite_disconnect will call fsg_disable, but fsg_disable calls usb_ep_disable using async way, there are register accesses for usb_ep_disable. So sometimes, I get system hang due to visit register without clock, sometimes not. The Linux Kernel USB maintainer Alan Stern suggests this kinds of solution. See: http://marc.info/?l=linux-usb&m=138541769810983&w=2. Cc: #v4.9+ Signed-off-by: Peter Chen Link: https://lore.kernel.org/r/20190820020503.27080-2-peter.chen@nxp.com Signed-off-by: Greg Kroah-Hartman

usb: hcd: use managed device resources

2019-09-06T08:22:15+00:00

commit 76da906ad727048a74bb8067031ee99fc070c7da upstream. Using managed device resources in usb_hcd_pci_probe() allows devm usage for resource subranges, such as the mmio resource for the platform device created to control host/device mode mux, which is a xhci extended capability, and sits inside the xhci mmio region. If managed device resources are not used then "parent" resource is released before subrange at driver removal as .remove callback is called before the devres list of resources for this device is walked and released. This has been observed with the xhci extended capability driver causing a use-after-free which is now fixed. An additional nice benefit is that error handling on driver initialisation is simplified much. Signed-off-by: Carsten Schmid Tested-by: Carsten Schmid Reviewed-by: Mathias Nyman Fixes: fa31b3cb2ae1 ("xhci: Add Intel extended cap / otg phy mux handling") Cc: # v4.19+ Link: https://lore.kernel.org/r/1566569488679.31808@mentor.com Signed-off-by: Greg Kroah-Hartman