- 14 May, 2019 18 commits
-
-
Boris Ostrovsky authored
commit cae5ec34 upstream s/L1TF/MDS/ Signed-off-by:
Boris Ostrovsky <boris.ostrovsky@oracle.com> Signed-off-by:
Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Reviewed-by:
Tyler Hicks <tyhicks@canonical.com> Reviewed-by:
Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Josh Poimboeuf authored
commit 39226ef0 upstream MDS is vulnerable with SMT. Make that clear with a one-time printk whenever SMT first gets enabled. Signed-off-by:
Jiri Kosina <jkosina@suse.cz> Signed-off-by:
-
Josh Poimboeuf authored
commit 7c3658b2 upstream arch_smt_update() now has a dependency on both Spectre v2 and MDS mitigations. Move its initial call to after all the mitigation decisions have been made. Signed-off-by:
-
Josh Poimboeuf authored
commit d71eb0ce upstream Add the mds=full,nosmt cmdline option. This is like mds=full, but with SMT disabled if the CPU is vulnerable. Signed-off-by:
-
Thomas Gleixner authored
commit 5999bbe7 upstream Add the initial MDS vulnerability documentation. Signed-off-by:
Jon Masters <jcm@redhat.com> Signed-off-by:
-
Thomas Gleixner authored
commit 65fd4cb6 upstream Move L!TF to a separate directory so the MDS stuff can be added at the side. Otherwise the all hardware vulnerabilites have their own top level entry. Should have done that right away. Signed-off-by:
-
Thomas Gleixner authored
commit 22dd8365 upstream In virtualized environments it can happen that the host has the microcode update which utilizes the VERW instruction to clear CPU buffers, but the hypervisor is not yet updated to expose the X86_FEATURE_MD_CLEAR CPUID bit to guests. Introduce an internal mitigation mode VMWERV which enables the invocation of the CPU buffer clearing even if X86_FEATURE_MD_CLEAR is not set. If the system has no updated microcode this results in a pointless execution of the VERW instruction wasting a few CPU cycles. If the microcode is updated, but not exposed to a guest then the CPU buffers will be cleared. That said: Virtual Machines Will Eventually Receive Vaccine Signed-off-by:
Borislav Petkov <bp@suse.de> Reviewed-by:
-
Thomas Gleixner authored
commit 8a4b06d3 upstream Add the sysfs reporting file for MDS. It exposes the vulnerability and mitigation state similar to the existing files for the other speculative hardware vulnerabilities. Signed-off-by:
-
Thomas Gleixner authored
commit bc124170 upstream Now that the mitigations are in place, add a command line parameter to control the mitigation, a mitigation selector function and a SMT update mechanism. This is the minimal straight forward initial implementation which just provides an always on/off mode. The command line parameter is: mds=[full|off] This is consistent with the existing mitigations for other speculative hardware vulnerabilities. The idle invocation is dynamically updated according to the SMT state of the system similar to the dynamic update of the STIBP mitigation. The idle mitigation is limited to CPUs which are only affected by MSBDS and not any other variant, because the other variants cannot be mitigated on SMT enabled systems. Signed-off-by:
-
Thomas Gleixner authored
commit 07f07f55 upstream Add a static key which controls the invocation of the CPU buffer clear mechanism on idle entry. This is independent of other MDS mitigations because the idle entry invocation to mitigate the potential leakage due to store buffer repartitioning is only necessary on SMT systems. Add the actual invocations to the different halt/mwait variants which covers all usage sites. mwaitx is not patched as it's not available on Intel CPUs. The buffer clear is only invoked before entering the C-State to prevent that stale data from the idling CPU is spilled to the Hyper-Thread sibling after the Store buffer got repartitioned and all entries are available to the non idle sibling. When coming out of idle the store buffer is partitioned again so each sibling has half of it available. Now CPU which returned from idle could be speculatively exposed to contents of the sibling, but the buffers are flushed either on exit to user space or on VMENTER. When later on conditional buffer clearing is implemented on top of this, then there is no action required either because before returning to user space the context switch will set the condition flag which causes a flush on the return to user path. Note, that the buffer clearing on idle is only sensible on CPUs which are solely affected by MSBDS and not any other variant of MDS because the other MDS variants cannot be mitigated when SMT is enabled, so the buffer clearing on idle would be a window dressing exercise. This intentionally does not handle the case in the acpi/processor_idle driver which uses the legacy IO port interface for C-State transitions for two reasons: - The acpi/processor_idle driver was replaced by the intel_idle driver almost a decade ago. Anything Nehalem upwards supports it and defaults to that new driver. - The legacy IO port interface is likely to be used on older and therefore unaffected CPUs or on systems which do not receive microcode updates anymore, so there is no point in adding that. Signed-off-by:
Frederic Weisbecker <frederic@kernel.org> Reviewed-by:
-
Thomas Gleixner authored
commit 650b68a0 upstream CPUs which are affected by L1TF and MDS mitigate MDS with the L1D Flush on VMENTER when updated microcode is installed. If a CPU is not affected by L1TF or if the L1D Flush is not in use, then MDS mitigation needs to be invoked explicitly. For these cases, follow the host mitigation state and invoke the MDS mitigation before VMENTER. Signed-off-by:
-
Thomas Gleixner authored
commit 04dcbdb8 upstream Add a static key which controls the invocation of the CPU buffer clear mechanism on exit to user space and add the call into prepare_exit_to_usermode() and do_nmi() right before actually returning. Add documentation which kernel to user space transition this covers and explain why some corner cases are not mitigated. Signed-off-by:
-
Thomas Gleixner authored
commit 6a9e5292 upstream The Microarchitectural Data Sampling (MDS) vulernabilities are mitigated by clearing the affected CPU buffers. The mechanism for clearing the buffers uses the unused and obsolete VERW instruction in combination with a microcode update which triggers a CPU buffer clear when VERW is executed. Provide a inline function with the assembly magic. The argument of the VERW instruction must be a memory operand as documented: "MD_CLEAR enumerates that the memory-operand variant of VERW (for example, VERW m16) has been extended to also overwrite buffers affected by MDS. This buffer overwriting functionality is not guaranteed for the register operand variant of VERW." Documentation also recommends to use a writable data segment selector: "The buffer overwriting occurs regardless of the result of the VERW permission check, as well as when the selector is null or causes a descriptor load segment violation. However, for lowest latency we recommend using a selector that indicates a valid writable data segment." Add x86 specific documentation about MDS and the internal workings of the mitigation. Signed-off-by:
-
Andi Kleen authored
commit 6c4dbbd1 upstream X86_FEATURE_MD_CLEAR is a new CPUID bit which is set when microcode provides the mechanism to invoke a flush of various exploitable CPU buffers by invoking the VERW instruction. Hand it through to guests so they can adjust their mitigations. This also requires corresponding qemu changes, which are available separately. [ tglx: Massaged changelog ] Signed-off-by:
Andi Kleen <ak@linux.intel.com> Signed-off-by:
-
Thomas Gleixner authored
commit e261f209 upstream This bug bit is set on CPUs which are only affected by Microarchitectural Store Buffer Data Sampling (MSBDS) and not by any other MDS variant. This is important because the Store Buffers are partitioned between Hyper-Threads so cross thread forwarding is not possible. But if a thread enters or exits a sleep state the store buffer is repartitioned which can expose data from one thread to the other. This transition can be mitigated. That means that for CPUs which are only affected by MSBDS SMT can be enabled, if the CPU is not affected by other SMT sensitive vulnerabilities, e.g. L1TF. The XEON PHI variants fall into that category. Also the Silvermont/Airmont ATOMs, but for them it's not really relevant as they do not support SMT, but mark them for completeness sake. Signed-off-by:
-
Andi Kleen authored
commit ed5194c2 upstream Microarchitectural Data Sampling (MDS), is a class of side channel attacks on internal buffers in Intel CPUs. The variants are: - Microarchitectural Store Buffer Data Sampling (MSBDS) (CVE-2018-12126) - Microarchitectural Fill Buffer Data Sampling (MFBDS) (CVE-2018-12130) - Microarchitectural Load Port Data Sampling (MLPDS) (CVE-2018-12127) MSBDS leaks Store Buffer Entries which can be speculatively forwarded to a dependent load (store-to-load forwarding) as an optimization. The forward can also happen to a faulting or assisting load operation for a different memory address, which can be exploited under certain conditions. Store buffers are partitioned between Hyper-Threads so cross thread forwarding is not possible. But if a thread enters or exits a sleep state the store buffer is repartitioned which can expose data from one thread to the other. MFBDS leaks Fill Buffer Entries. Fill buffers are used internally to manage L1 miss situations and to hold data which is returned or sent in response to a memory or I/O operation. Fill buffers can forward data to a load operation and also write data to the cache. When the fill buffer is deallocated it can retain the stale data of the preceding operations which can then be forwarded to a faulting or assisting load operation, which can be exploited under certain conditions. Fill buffers are shared between Hyper-Threads so cross thread leakage is possible. MLDPS leaks Load Port Data. Load ports are used to perform load operations from memory or I/O. The received data is then forwarded to the register file or a subsequent operation. In some implementations the Load Port can contain stale data from a previous operation which can be forwarded to faulting or assisting loads under certain conditions, which again can be exploited eventually. Load ports are shared between Hyper-Threads so cross thread leakage is possible. All variants have the same mitigation for single CPU thread case (SMT off), so the kernel can treat them as one MDS issue. Add the basic infrastructure to detect if the current CPU is affected by MDS. [ tglx: Rewrote changelog ] Signed-off-by:
-
Thomas Gleixner authored
commit 36ad3513 upstream The CPU vulnerability whitelists have some overlap and there are more whitelists coming along. Use the driver_data field in the x86_cpu_id struct to denote the whitelisted vulnerabilities and combine all whitelists into one. Suggested-by:
Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by:
-
Thomas Gleixner authored
commit d8eabc37 upstream Greg pointed out that speculation related bit defines are using (1 << N) format instead of BIT(N). Aside of that (1 << N) is wrong as it should use 1UL at least. Clean it up. [ Josh Poimboeuf: Fix tools build ] Reported-by:
-
- 11 May, 2019 22 commits
-
-
Greg Kroah-Hartman authored
-
Will Deacon authored
commit 03110a5c upstream. Our futex implementation makes use of LDXR/STXR loops to perform atomic updates to user memory from atomic context. This can lead to latency problems if we end up spinning around the LL/SC sequence at the expense of doing something useful. Rework our futex atomic operations so that we return -EAGAIN if we fail to update the futex word after 128 attempts. The core futex code will reschedule if necessary and we'll try again later. Cc: <stable@kernel.org> Fixes: 6170a974 ("arm64: Atomic operations") Signed-off-by:
Will Deacon <will.deacon@arm.com> Signed-off-by:
-
Will Deacon authored
commit 6b4f4bc9 upstream. Some futex() operations, including FUTEX_WAKE_OP, require the kernel to perform an atomic read-modify-write of the futex word via the userspace mapping. These operations are implemented by each architecture in arch_futex_atomic_op_inuser() and futex_atomic_cmpxchg_inatomic(), which are called in atomic context with the relevant hash bucket locks held. Although these routines may return -EFAULT in response to a page fault generated when accessing userspace, they are expected to succeed (i.e. return 0) in all other cases. This poses a problem for architectures that do not provide bounded forward progress guarantees or fairness of contended atomic operations and can lead to starvation in some cases. In these problematic scenarios, we must return back to the core futex code so that we can drop the hash bucket locks and reschedule if necessary, much like we do in the case of a page fault. Allow architectures to return -EAGAIN from their implementations of arch_futex_atomic_op_inuser() and futex_atomic_cmpxchg_inatomic(), which will cause the core futex code to reschedule if necessary and return back to the architecture code later on. Cc: <stable@kernel.org> Acked-by:
Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by:
-
Dan Carpenter authored
commit 476c7e1d upstream. The problem here is that addr can be I3C_BROADCAST_ADDR (126). That means we're shifting by (126 * 2) % 64 which is 60. The I3C_ADDR_SLOT_STATUS_MASK is an enum which is an unsigned int in GCC so shifts greater than 31 are undefined. Fixes: 3a379bbc ("i3c: Add core I3C infrastructure") Cc: <stable@vger.kernel.org> Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
Boris Brezillon <boris.brezillon@collabora.com> Signed-off-by:
-
Ross Zwisler authored
commit 0efa3334 upstream. Currently in sst_dsp_new() if we get an error return from sst_dma_new() we just print an error message and then still complete the function successfully. This means that we are trying to run without sst->dma properly set up, which will result in NULL pointer dereference when sst->dma is later used. This was happening for me in sst_dsp_dma_get_channel(): struct sst_dma *dma = dsp->dma; ... dma->ch = dma_request_channel(mask, dma_chan_filter, dsp); This resulted in: BUG: unable to handle kernel NULL pointer dereference at 0000000000000018 IP: sst_dsp_dma_get_channel+0x4f/0x125 [snd_soc_sst_firmware] Fix this by adding proper error handling for the case where we fail to set up DMA. This change only affects Haswell and Broadwell systems. Baytrail systems explicilty opt-out of DMA via sst->pdata->resindex_dma_base being set to -1. Signed-off-by:
Ross Zwisler <zwisler@google.com> Cc: stable@vger.kernel.org Acked-by:
Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by:
Mark Brown <broonie@kernel.org> Signed-off-by:
-
Oliver Neukum authored
commit 3ae62a42 upstream. This is the UAS version of 747668db usb-storage: Set virt_boundary_mask to avoid SG overflows We are not as likely to be vulnerable as storage, as it is unlikelier that UAS is run over a controller without native support for SG, but the issue exists. The issue has been existing since the inception of the driver. Fixes: 115bb1ff ("USB: Add UAS driver") Signed-off-by:
Oliver Neukum <oneukum@suse.com> Cc: stable <stable@vger.kernel.org> Signed-off-by:
-
Chen-Yu Tsai authored
commit 62611abc upstream. The code path for Macs goes through bcm_apple_get_resources(), which skips over the code that sets up the regulator supplies. As a result, the call to regulator_bulk_enable() / regulator_bulk_disable() results in a NULL pointer dereference. This was reported on the kernel.org Bugzilla, bug 202963. Unbreak Broadcom Bluetooth support on Intel Macs by checking if the supplies were set up before enabling or disabling them. The same does not need to be done for the clocks, as the common clock framework API checks for NULL pointers. Fixes: 75d11676 ("Bluetooth: hci_bcm: Add support for regulator supplies") Cc: <stable@vger.kernel.org> # 5.0.x Signed-off-by:
Chen-Yu Tsai <wens@csie.org> Tested-by:
Imre Kaloz <kaloz@openwrt.org> Signed-off-by:
Marcel Holtmann <marcel@holtmann.org> Signed-off-by:
-
Luiz Augusto von Dentz authored
commit ba8f5289 upstream. l2cap_le_flowctl_init was reseting the tx_credits which works only for outgoing connection since that set the tx_credits on the response, for incoming connections that was not the case which leaves the channel without any credits causing it to be suspended. Signed-off-by:
Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Signed-off-by:
-
Marcel Holtmann authored
commit d5bb334a upstream. The minimum encryption key size for LE connections is 56 bits and to align LE with BR/EDR, enforce 56 bits of minimum encryption key size for BR/EDR connections as well. Signed-off-by:
Johan Hedberg <johan.hedberg@intel.com> Cc: stable@vger.kernel.org Signed-off-by:
-
Young Xiao authored
commit a1616a5a upstream. Struct ca is copied from userspace. It is not checked whether the "name" field is NULL terminated, which allows local users to obtain potentially sensitive information from kernel stack memory, via a HIDPCONNADD command. This vulnerability is similar to CVE-2011-1079. Signed-off-by:
Young Xiao <YangX92@hotmail.com> Signed-off-by:
-
Quinn Tran authored
commit 2137490f upstream. This patch fixes issue reported by some of the customers, who discovered that after cable pull scenario the devices disappear and path seems to remain in blocked state. Once the device reappears, driver does not seem to update path to online. This issue appears because of the defer flag creating race condition where the same session reappears. This patch fixes this issue by indicating SCSI-ML of device lost when qlt_free_session_done() is called from qlt_unreg_sess(). Fixes: 41dc529a ("qla2xxx: Improve RSCN handling in driver") Signed-off-by:
Quinn Tran <qtran@marvell.com> Cc: stable@vger.kernel.org #4.19 Signed-off-by:
Himanshu Madhani <hmadhani@marvell.com> Reviewed-by:
Ewan D. Milne <emilne@redhat.com> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by:
-
Giridhar Malavali authored
commit ffc81fc0 upstream. This patch sets remote_port_devloss value to 0. This indicates to FC-NVMe transport that driver is unloading and transport should not retry. Fixes: e476fe8a ("scsi: qla2xxx: Fix unload when NVMe devices are configured") Cc: stable@vger.kernel.org Signed-off-by:
Giridhar Malavali <gmalavali@marvell.com> Signed-off-by:
-
Andrew Vasquez authored
commit 5cbdae10 upstream. Commit e6f77540 ("scsi: qla2xxx: Fix an integer overflow in sysfs code") incorrectly set 'optrom_region_size' to 'start+size', which can overflow option-rom boundaries when 'start' is non-zero. Continue setting optrom_region_size to the proper adjusted value of 'size'. Fixes: e6f77540 ("scsi: qla2xxx: Fix an integer overflow in sysfs code") Cc: stable@vger.kernel.org Signed-off-by:
Andrew Vasquez <andrewv@marvell.com> Signed-off-by:
-
Silvio Cesare authored
commit e7f7b6f3 upstream. Change snprintf to scnprintf. There are generally two cases where using snprintf causes problems. 1) Uses of size += snprintf(buf, SIZE - size, fmt, ...) In this case, if snprintf would have written more characters than what the buffer size (SIZE) is, then size will end up larger than SIZE. In later uses of snprintf, SIZE - size will result in a negative number, leading to problems. Note that size might already be too large by using size = snprintf before the code reaches a case of size += snprintf. 2) If size is ultimately used as a length parameter for a copy back to user space, then it will potentially allow for a buffer overflow and information disclosure when size is greater than SIZE. When the size is used to index the buffer directly, we can have memory corruption. This also means when size = snprintf... is used, it may also cause problems since size may become large. Copying to userspace is mitigated by the HARDENED_USERCOPY kernel configuration. The solution to these issues is to use scnprintf which returns the number of characters actually written to the buffer, so the size variable will never exceed SIZE. Signed-off-by:
Silvio Cesare <silvio.cesare@gmail.com> Signed-off-by:
Willy Tarreau <w@1wt.eu> Signed-off-by:
James Smart <james.smart@broadcom.com> Cc: Dick Kennedy <dick.kennedy@broadcom.com> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Kees Cook <keescook@chromium.org> Cc: Will Deacon <will.deacon@arm.com> Cc: Greg KH <greg@kroah.com> Signed-off-by:
-
Samuel Holland authored
commit a84014e1 upstream. When enabling ARCH_SUNXI from allnoconfig, SUNXI_SRAM is enabled, but not REGMAP_MMIO, so the kernel fails to link with an undefined reference to __devm_regmap_init_mmio_clk. Select REGMAP_MMIO, as suggested in drivers/base/regmap/Kconfig. This creates the following dependency loop: drivers/of/Kconfig:68: symbol OF_IRQ depends on IRQ_DOMAIN kernel/irq/Kconfig:63: symbol IRQ_DOMAIN is selected by REGMAP drivers/base/regmap/Kconfig:7: symbol REGMAP default is visible depending on REGMAP_MMIO drivers/base/regmap/Kconfig:39: symbol REGMAP_MMIO is selected by SUNXI_SRAM drivers/soc/sunxi/Kconfig:4: symbol SUNXI_SRAM is selected by USB_MUSB_SUNXI drivers/usb/musb/Kconfig:63: symbol USB_MUSB_SUNXI depends on GENERIC_PHY drivers/phy/Kconfig:7: symbol GENERIC_PHY is selected by PHY_BCM_NS_USB3 drivers/phy/broadcom/Kconfig:29: symbol PHY_BCM_NS_USB3 depends on MDIO_BUS drivers/net/phy/Kconfig:12: symbol MDIO_BUS default is visible depending on PHYLIB drivers/net/phy/Kconfig:181: symbol PHYLIB is selected by ARC_EMAC_CORE drivers/net/ethernet/arc/Kconfig:18: symbol ARC_EMAC_CORE is selected by ARC_EMAC drivers/net/ethernet/arc/Kconfig:24: symbol ARC_EMAC depends on OF_IRQ To fix the circular dependency, make USB_MUSB_SUNXI select GENERIC_PHY instead of depending on it. This matches the use of GENERIC_PHY by all but two other drivers. Cc: <stable@vger.kernel.org> # 4.19 Fixes: 5828729b ("soc: sunxi: export a regmap for EMAC clock reg on A64") Signed-off-by:
Samuel Holland <samuel@sholland.org> Acked-by:
Maxime Ripard <maxime.ripard@bootlin.com> Signed-off-by:
Bin Liu <b-liu@ti.com> Signed-off-by:
-
Hans de Goede authored
commit c8afd034 upstream. Commit 48402cee ("ACPI / LPSS: Resume BYT/CHT I2C controllers from resume_noirq") makes acpi_lpss_{suspend_late,resume_early}() bail early on BYT/CHT as resume_from_noirq is set. This means that on resume from hibernate dw_i2c_plat_resume() doesn't get called by the restore_early callback, acpi_lpss_resume_early(). Instead it should be called by the restore_noirq callback matching how things are done when resume_from_noirq is set and we are doing a regular resume. Change the restore_noirq callback to acpi_lpss_resume_noirq so that dw_i2c_plat_resume() gets properly called when resume_from_noirq is set and we are resuming from hibernate. Likewise also change the poweroff_noirq callback so that dw_i2c_plat_suspend gets called properly. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=202139 Fixes: 48402cee ("ACPI / LPSS: Resume BYT/CHT I2C controllers from resume_noirq") Reported-by:
Kai-Heng Feng <kai.heng.feng@canonical.com> Signed-off-by:
Hans de Goede <hdegoede@redhat.com> Cc: 4.20+ <stable@vger.kernel.org> # 4.20+ Signed-off-by:
Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by:
-
Gregory CLEMENT authored
commit 8db82563 upstream. The frequency calculation was based on the current(max) frequency of the CPU. However for low frequency, the value used was already the parent frequency divided by a factor of 2. Instead of using this frequency, this fix directly get the frequency from the parent clock. Fixes: 92ce45fb ("cpufreq: Add DVFS support for Armada 37xx") Cc: <stable@vger.kernel.org> Reported-by:
Christian Neubert <christian.neubert.86@gmail.com> Signed-off-by:
Gregory CLEMENT <gregory.clement@bootlin.com> Signed-off-by:
Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by:
-
Bjorn Andersson authored
commit 447ccb4e upstream. The of_device_id table needs to be registered as module alias in order for automatic module loading to pick the kernel module based on the DeviceTree compatible. So add MODULE_DEVICE_TABLE() to make this happen. Fixes: e13d7572 ("iio: adc: Add QCOM SPMI PMIC5 ADC driver") Cc: stable@vger.kernel.org Signed-off-by:
Bjorn Andersson <bjorn.andersson@linaro.org> Signed-off-by:
Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by:
-
Alexander Shishkin authored
commit e60e9a4b upstream. This adds support for Intel TH on Comet Lake. Signed-off-by:
Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: stable <stable@vger.kernel.org> Signed-off-by:
-
Prasad Sodagudi authored
commit 59c39840 upstream. When irq_set_affinity_notifier() replaces the notifier, then the reference count on the old notifier is dropped which causes it to be freed. But nothing ensures that the old notifier is not longer queued in the work list. If it is queued this results in a use after free and possibly in work list corruption. Ensure that the work is canceled before the reference is dropped. Signed-off-by:
Prasad Sodagudi <psodagud@codeaurora.org> Signed-off-by:
-
Alan Stern authored
commit 747668db upstream. The USB subsystem has always had an unusual requirement for its scatter-gather transfers: Each element in the scatterlist (except the last one) must have a length divisible by the bulk maxpacket size. This is a particular issue for USB mass storage, which uses SG lists created by the block layer rather than setting up its own. So far we have scraped by okay because most devices have a logical block size of 512 bytes or larger, and the bulk maxpacket sizes for USB 2 and below are all <= 512. However, USB 3 has a bulk maxpacket size of 1024. Since the xhci-hcd driver includes native SG support, this hasn't mattered much. But now people are trying to use USB-3 mass storage devices with USBIP, and the vhci-hcd driver currently does not have full SG support. The result is an overflow error, when the driver attempts to implement an SG transfer of 63 512-byte blocks as a single 3584-byte (7 blocks) transfer followed by seven 4096-byte (8 blocks) transfers. The device instead sends 31 1024-byte packets followed by a 512-byte packet, and this overruns the first SG buffer. Ideally this would be fixed by adding better SG support to vhci-hcd. But for now it appears we can work around the problem by asking the block layer to respect the maxpacket limitation, through the use of the virt_boundary_mask. Signed-off-by:
Alan Stern <stern@rowland.harvard.edu> Reported-by:
Seth Bollinger <Seth.Bollinger@digi.com> Tested-by:
-
Johan Hovold authored
commit 764478f4 upstream. Fix two long-standing bugs which could potentially lead to memory corruption or leave the port throttled until it is reopened (on weakly ordered systems), respectively, when read-URB completion races with unthrottle(). First, the URB must not be marked as free before processing is complete to prevent it from being submitted by unthrottle() on another CPU. CPU 1 CPU 2 ================ ================ complete() unthrottle() process_urb(); smp_mb__before_atomic(); set_bit(i, free); if (test_and_clear_bit(i, free)) submit_urb(); Second, the URB must be marked as free before checking the throttled flag to prevent unthrottle() on another CPU from failing to observe that the URB needs to be submitted if complete() sees that the throttled flag is set. CPU 1 CPU 2 ================ ================ complete() unthrottle() set_bit(i, free); throttled = 0; smp_mb__after_atomic(); smp_mb(); if (throttled) if (test_and_clear_bit(i, free)) return; submit_urb(); Note that test_and_clear_bit() only implies barriers when the test is successful. To handle the case where the URB is still in use an explicit barrier needs to be added to unthrottle() for the second race condition. Also note that the first race was fixed by 36e59e0d ("cdc-acm: fix race between callback and unthrottle") back in 2015, but the bug was reintroduced a year later. Fixes: 1aba579f ("cdc-acm: handle read pipe errors") Fixes: 088c64f8 ("USB: cdc-acm: re-write read processing") Signed-off-by:
Johan Hovold <johan@kernel.org> Acked-by:
-
