linux/drivers/base, branch v4.20

devres: Align data[] to ARCH_KMALLOC_MINALIGN

2018-11-11T19:40:04Z

Initially we bumped into problem with 32-bit aligned atomic64_t on ARC, see [1]. And then during quite lengthly discussion Peter Z. mentioned ARCH_KMALLOC_MINALIGN which IMHO makes perfect sense. If allocation is done by plain kmalloc() obtained buffer will be ARCH_KMALLOC_MINALIGN aligned and then why buffer obtained via devm_kmalloc() should have any other alignment? This way we at least get the same behavior for both types of allocation. [1] http://lists.infradead.org/pipermail/linux-snps-arc/2018-July/004009.html [2] http://lists.infradead.org/pipermail/linux-snps-arc/2018-July/004036.html Signed-off-by: Alexey Brodkin Cc: Greg Kroah-Hartman Cc: Geert Uytterhoeven Cc: David Laight Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Vineet Gupta Cc: Will Deacon Cc: Greg KH Cc: # 4.8+ Signed-off-by: Greg Kroah-Hartman

mm/memory_hotplug: fix online/offline_pages called w.o. mem_hotplug_lock

2018-10-31T15:54:17Z

There seem to be some problems as result of 30467e0b3be ("mm, hotplug: fix concurrent memory hot-add deadlock"), which tried to fix a possible lock inversion reported and discussed in [1] due to the two locks a) device_lock() b) mem_hotplug_lock While add_memory() first takes b), followed by a) during bus_probe_device(), onlining of memory from user space first took a), followed by b), exposing a possible deadlock. In [1], and it was decided to not make use of device_hotplug_lock, but rather to enforce a locking order. The problems I spotted related to this: 1. Memory block device attributes: While .state first calls mem_hotplug_begin() and the calls device_online() - which takes device_lock() - .online does no longer call mem_hotplug_begin(), so effectively calls online_pages() without mem_hotplug_lock. 2. device_online() should be called under device_hotplug_lock, however onlining memory during add_memory() does not take care of that. In addition, I think there is also something wrong about the locking in 3. arch/powerpc/platforms/powernv/memtrace.c calls offline_pages() without locks. This was introduced after 30467e0b3be. And skimming over the code, I assume it could need some more care in regards to locking (e.g. device_online() called without device_hotplug_lock. This will be addressed in the following patches. Now that we hold the device_hotplug_lock when - adding memory (e.g. via add_memory()/add_memory_resource()) - removing memory (e.g. via remove_memory()) - device_online()/device_offline() We can move mem_hotplug_lock usage back into online_pages()/offline_pages(). Why is mem_hotplug_lock still needed? Essentially to make get_online_mems()/put_online_mems() be very fast (relying on device_hotplug_lock would be very slow), and to serialize against addition of memory that does not create memory block devices (hmm). [1] http://driverdev.linuxdriverproject.org/pipermail/ driverdev-devel/ 2015-February/065324.html This patch is partly based on a patch by Vitaly Kuznetsov. Link: http://lkml.kernel.org/r/20180925091457.28651-4-david@redhat.com Signed-off-by: David Hildenbrand Reviewed-by: Pavel Tatashin Reviewed-by: Rashmica Gupta Reviewed-by: Oscar Salvador Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Michael Ellerman Cc: "Rafael J. Wysocki" Cc: Len Brown Cc: Greg Kroah-Hartman Cc: "K. Y. Srinivasan" Cc: Haiyang Zhang Cc: Stephen Hemminger Cc: Martin Schwidefsky Cc: Heiko Carstens Cc: Boris Ostrovsky Cc: Juergen Gross Cc: Rashmica Gupta Cc: Michael Neuling Cc: Balbir Singh Cc: Kate Stewart Cc: Thomas Gleixner Cc: Philippe Ombredanne Cc: Pavel Tatashin Cc: Vlastimil Babka Cc: Dan Williams Cc: Oscar Salvador Cc: YASUAKI ISHIMATSU Cc: Mathieu Malaterre Cc: John Allen Cc: Jonathan Corbet Cc: Joonsoo Kim Cc: Nathan Fontenot Cc: Rafael J. Wysocki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm/memory_hotplug: make add_memory() take the device_hotplug_lock

2018-10-31T15:54:17Z

add_memory() currently does not take the device_hotplug_lock, however is aleady called under the lock from arch/powerpc/platforms/pseries/hotplug-memory.c drivers/acpi/acpi_memhotplug.c to synchronize against CPU hot-remove and similar. In general, we should hold the device_hotplug_lock when adding memory to synchronize against online/offline request (e.g. from user space) - which already resulted in lock inversions due to device_lock() and mem_hotplug_lock - see 30467e0b3be ("mm, hotplug: fix concurrent memory hot-add deadlock"). add_memory()/add_memory_resource() will create memory block devices, so this really feels like the right thing to do. Holding the device_hotplug_lock makes sure that a memory block device can really only be accessed (e.g. via .online/.state) from user space, once the memory has been fully added to the system. The lock is not held yet in drivers/xen/balloon.c arch/powerpc/platforms/powernv/memtrace.c drivers/s390/char/sclp_cmd.c drivers/hv/hv_balloon.c So, let's either use the locked variants or take the lock. Don't export add_memory_resource(), as it once was exported to be used by XEN, which is never built as a module. If somebody requires it, we also have to export a locked variant (as device_hotplug_lock is never exported). Link: http://lkml.kernel.org/r/20180925091457.28651-3-david@redhat.com Signed-off-by: David Hildenbrand Reviewed-by: Pavel Tatashin Reviewed-by: Rafael J. Wysocki Reviewed-by: Rashmica Gupta Reviewed-by: Oscar Salvador Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Michael Ellerman Cc: "Rafael J. Wysocki" Cc: Len Brown Cc: Greg Kroah-Hartman Cc: Boris Ostrovsky Cc: Juergen Gross Cc: Nathan Fontenot Cc: John Allen Cc: Michal Hocko Cc: Dan Williams Cc: Joonsoo Kim Cc: Vlastimil Babka Cc: Mathieu Malaterre Cc: Pavel Tatashin Cc: YASUAKI ISHIMATSU Cc: Balbir Singh Cc: Haiyang Zhang Cc: Heiko Carstens Cc: Jonathan Corbet Cc: Kate Stewart Cc: "K. Y. Srinivasan" Cc: Martin Schwidefsky Cc: Michael Neuling Cc: Philippe Ombredanne Cc: Stephen Hemminger Cc: Thomas Gleixner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm: remove include/linux/bootmem.h

2018-10-31T15:54:16Z

Move remaining definitions and declarations from include/linux/bootmem.h into include/linux/memblock.h and remove the redundant header. The includes were replaced with the semantic patch below and then semi-automated removal of duplicated '#include @@ @@ - #include + #include [sfr@canb.auug.org.au: dma-direct: fix up for the removal of linux/bootmem.h] Link: http://lkml.kernel.org/r/20181002185342.133d1680@canb.auug.org.au [sfr@canb.auug.org.au: powerpc: fix up for removal of linux/bootmem.h] Link: http://lkml.kernel.org/r/20181005161406.73ef8727@canb.auug.org.au [sfr@canb.auug.org.au: x86/kaslr, ACPI/NUMA: fix for linux/bootmem.h removal] Link: http://lkml.kernel.org/r/20181008190341.5e396491@canb.auug.org.au Link: http://lkml.kernel.org/r/1536927045-23536-30-git-send-email-rppt@linux.vnet.ibm.com Signed-off-by: Mike Rapoport Signed-off-by: Stephen Rothwell Acked-by: Michal Hocko Cc: Catalin Marinas Cc: Chris Zankel Cc: "David S. Miller" Cc: Geert Uytterhoeven Cc: Greentime Hu Cc: Greg Kroah-Hartman Cc: Guan Xuetao Cc: Ingo Molnar Cc: "James E.J. Bottomley" Cc: Jonas Bonn Cc: Jonathan Corbet Cc: Ley Foon Tan Cc: Mark Salter Cc: Martin Schwidefsky Cc: Matt Turner Cc: Michael Ellerman Cc: Michal Simek Cc: Palmer Dabbelt Cc: Paul Burton Cc: Richard Kuo Cc: Richard Weinberger Cc: Rich Felker Cc: Russell King Cc: Serge Semin Cc: Thomas Gleixner Cc: Tony Luck Cc: Vineet Gupta Cc: Yoshinori Sato Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm, proc: add KReclaimable to /proc/meminfo

2018-10-26T23:26:32Z

The vmstat NR_KERNEL_MISC_RECLAIMABLE counter is for kernel non-slab allocations that can be reclaimed via shrinker. In /proc/meminfo, we can show the sum of all reclaimable kernel allocations (including slab) as "KReclaimable". Add the same counter also to per-node meminfo under /sys With this counter, users will have more complete information about kernel memory usage. Non-slab reclaimable pages (currently just the ION allocator) will not be missing from /proc/meminfo, making users wonder where part of their memory went. More precisely, they already appear in MemAvailable, but without the new counter, it's not obvious why the value in MemAvailable doesn't fully correspond with the sum of other counters participating in it. Link: http://lkml.kernel.org/r/20180731090649.16028-6-vbabka@suse.cz Signed-off-by: Vlastimil Babka Acked-by: Roman Gushchin Cc: Christoph Lameter Cc: David Rientjes Cc: Johannes Weiner Cc: Joonsoo Kim Cc: Laura Abbott Cc: Matthew Wilcox Cc: Mel Gorman Cc: Michal Hocko Cc: Sumit Semwal Cc: Vijayanand Jitta Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

Merge tag 'driver-core-4.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core

2018-10-26T15:42:25Z

Pull driver core updates from Greg KH: "Here is a small number of driver core patches for 4.20-rc1. Not much happened here this merge window, only a very tiny number of patches that do: - add BUS_ATTR_WO() for use by drivers - component error path fixes - kernfs range check fix - other tiny error path fixes and const changes All of these have been in linux-next with no reported issues for a while" * tag 'driver-core-4.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core: devres: provide devm_kstrdup_const() mm: move is_kernel_rodata() to asm-generic/sections.h devres: constify p in devm_kfree() driver core: add BUS_ATTR_WO() macro kernfs: Fix range checks in kernfs_get_target_path component: fix loop condition to call unbind() if bind() fails drivers/base/devtmpfs.c: don't pretend path is const in delete_path kernfs: update comment about kernfs_path() return value

Merge branch 'irq-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

2018-10-25T18:43:47Z

Pull irq updates from Thomas Gleixner: "The interrupt brigade came up with the following updates: - Driver for the Marvell System Error Interrupt machinery - Overhaul of the GIC-V3 ITS driver - Small updates and fixes all over the place" * 'irq-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (31 commits) genirq: Fix race on spurious interrupt detection softirq: Fix typo in __do_softirq() comments genirq: Fix grammar s/an /a / irqchip/gic: Unify GIC priority definitions irqchip/gic-v3: Remove acknowledge loop dt-bindings/interrupt-controller: Add documentation for Marvell SEI controller dt-bindings/interrupt-controller: Update Marvell ICU bindings irqchip/irq-mvebu-icu: Add support for System Error Interrupts (SEI) arm64: marvell: Enable SEI driver irqchip/irq-mvebu-sei: Add new driver for Marvell SEI irqchip/irq-mvebu-icu: Support ICU subnodes irqchip/irq-mvebu-icu: Disociate ICU and NSR irqchip/irq-mvebu-icu: Clarify the reset operation of configured interrupts irqchip/irq-mvebu-icu: Fix wrong private data retrieval dt-bindings/interrupt-controller: Fix Marvell ICU length in the example genirq/msi: Allow creation of a tree-based irqdomain for platform-msi dt-bindings: irqchip: renesas-irqc: Document r8a7744 support dt-bindings: irqchip: renesas-irqc: Document R-Car E3 support irqchip/pdc: Setup all edge interrupts as rising edge at GIC irqchip/gic-v3-its: Allow use of LPI tables in reserved memory ...

Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

2018-10-23T14:00:03Z

Pull scheduler updates from Ingo Molnar: "The main changes are: - Migrate CPU-intense 'misfit' tasks on asymmetric capacity systems, to better utilize (much) faster 'big core' CPUs. (Morten Rasmussen, Valentin Schneider) - Topology handling improvements, in particular when CPU capacity changes and related load-balancing fixes/improvements (Morten Rasmussen) - ... plus misc other improvements, fixes and updates" * 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (28 commits) sched/completions/Documentation: Add recommendation for dynamic and ONSTACK completions sched/completions/Documentation: Clean up the document some more sched/completions/Documentation: Fix a couple of punctuation nits cpu/SMT: State SMT is disabled even with nosmt and without "=force" sched/core: Fix comment regarding nr_iowait_cpu() and get_iowait_load() sched/fair: Remove setting task's se->runnable_weight during PELT update sched/fair: Disable LB_BIAS by default sched/pelt: Fix warning and clean up IRQ PELT config sched/topology: Make local variables static sched/debug: Use symbolic names for task state constants sched/numa: Remove unused numa_stats::nr_running field sched/numa: Remove unused code from update_numa_stats() sched/debug: Explicitly cast sched_feat() to bool sched/core: Disable SD_PREFER_SIBLING on asymmetric CPU capacity domains sched/fair: Don't move tasks to lower capacity CPUs unless necessary sched/fair: Set rq->rd->overload when misfit sched/fair: Wrap rq->rd->overload accesses with READ/WRITE_ONCE() sched/core: Change root_domain->overload type to int sched/fair: Change 'prefer_sibling' type to bool sched/fair: Kick nohz balance if rq->misfit_task_load ...

Merge tag 'acpi-4.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm

2018-10-23T09:33:16Z

Pull ACPI updates from Rafael Wysocki: "These fix ACPICA issues related to the handling of module-level AML, fix an ordering issue during ACPI initialization, update ACPICA to upstream revision 20181003 (including fixes mostly), fix issues with system-wide suspend/resume related to the ACPI driver for Intel SoCs (LPSS), fix device enumeration issues on boards with Dollar Cove or Whiskey Cove Intel PMICs, prevent ACPICA from calling ktime_get() in unsuitable conditions, update a few drivers and clean up some code in several places. Specifics: - Fix ACPICA issues related to the handling of module-level AML and make the ACPI initialization code parse ECDT before loading the definition block tables (Erik Schmauss). - Update ACPICA to upstream revision 20181003 including fixes related to the ill-defined "generic serial bus" and the handling of the _REG object (Bob Moore). - Fix some issues with system-wide suspend/resume on Intel BYT/CHT related to the handling of I2C controllers in the ACPI LPSS driver for Intel SoCs (Hans de Goede). - Modify the ACPI namespace scanning code to enumerate INT33FE HID devices as platform devices with I2C resources to avoid device enumeration problems on boards with Dollar Cove or Whiskey Cove Intel PMICs (Hans de Goede). - Prevent ACPICA from using ktime_get() during early resume from system-wide suspend before resuming the timekeeping which generally is unsafe and triggers a warning from the timekeeping code (Bart Van Assche). - Add low-level real time clock support to the ACPI Time and Aalarm Device (TAD) driver (Rafael Wysocki). - Fix the ACPI SBS driver to avoid GPE storms on MacBook Pro and Oopses when removing modules (Ronald Tschalär). - Fix the ACPI PPTT parsing code to handle architecturally unknown cache types properly (Jeffrey Hugo). - Fix initialization issue in the ACPI processor driver (Dou Liyang). - Clean up the code in several places (Andy Shevchenko, Bartlomiej Zolnierkiewicz, David Arcari, zhong jiang)" * tag 'acpi-4.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (33 commits) ACPI / scan: Create platform device for INT33FE ACPI nodes ACPI / OSL: Use 'jiffies' as the time bassis for acpi_os_get_timer() ACPI: probe ECDT before loading AML tables regardless of module-level code flag ACPICA: Remove acpi_gbl_group_module_level_code and only use acpi_gbl_execute_tables_as_methods instead ACPICA: AML Parser: fix parse loop to correctly skip erroneous extended opcodes ACPICA: AML interpreter: add region addresses in global list during initialization ACPI: TAD: Add low-level support for real time capability ACPI: remove redundant 'default n' from Kconfig ACPI / SBS: Fix rare oops when removing modules ACPI / SBS: Fix GPE storm on recent MacBookPro's ACPI/PPTT: Handle architecturally unknown cache types drivers: base: cacheinfo: Do not populate sysfs for unknown cache types ACPICA: Update version to 20181003 ACPICA: Never run _REG on system_memory and system_IO ACPICA: Split large interpreter file ACPICA: Update for field unit access ACPICA: Rename some of the Field Attribute defines ACPICA: Update for generic_serial_bus and attrib_raw_process_bytes protocol ACPI / processor: Fix the return value of acpi_processor_ids_walk() ACPI / LPSS: Resume BYT/CHT I2C controllers from resume_noirq ...

Merge tag 'pm-4.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm

2018-10-23T09:28:21Z

Pull power management updates from Rafael Wysocki: "These make hibernation on 32-bit x86 systems work in all of the cases in which it works on 64-bit x86 ones, update the menu cpuidle governor and the "polling" state to make them more efficient, add more hardware support to cpufreq drivers and fix issues with some of them, fix a bug in the conservative cpufreq governor, fix the operating performance points (OPP) framework and make it more stable, update the devfreq subsystem to take changes in the APIs used by into account and clean up some things all over. Specifics: - Backport hibernation bug fixes from x86-64 to x86-32 and consolidate hibernation handling on x86 to allow 32-bit systems to work in all of the cases in which 64-bit ones work (Zhimin Gu, Chen Yu). - Fix hibernation documentation (Vladimir D. Seleznev). - Update the menu cpuidle governor to fix a couple of issues with it, make it more efficient in some cases and clean it up (Rafael Wysocki). - Rework the cpuidle polling state implementation to make it more efficient (Rafael Wysocki). - Clean up the cpuidle core somewhat (Fieah Lim). - Fix the cpufreq conservative governor to take policy limits into account properly in some cases (Rafael Wysocki). - Add support for retrieving guaranteed performance information to the ACPI CPPC library and make the intel_pstate driver use it to expose the CPU base frequency via sysfs on systems with the hardware-managed P-states (HWP) feature enabled (Srinivas Pandruvada). - Fix clang warning in the CPPC cpufreq driver (Nathan Chancellor). - Get rid of device_node.name printing from cpufreq (Rob Herring). - Remove unnecessary unlikely() from the cpufreq core (Igor Stoppa). - Add support for the r8a7744 SoC to the cpufreq-dt driver (Biju Das). - Update the dt-platdev cpufreq driver to allow RK3399 to have separate tunables per cluster (Dmitry Torokhov). - Fix the dma_alloc_coherent() usage in the tegra186 cpufreq driver (Christoph Hellwig). - Make the imx6q cpufreq driver read OCOTP through nvmem for imx6ul/imx6ull (Anson Huang). - Fix several bugs in the operating performance points (OPP) framework and make it more stable (Viresh Kumar, Dave Gerlach). - Update the devfreq subsystem to take changes in the APIs used by into account, fix some issues with it and make it stop print device_node.name directly (Bjorn Andersson, Enric Balletbo i Serra, Matthias Kaehlcke, Rob Herring, Vincent Donnefort, zhong jiang). - Prepare the generic power domains (genpd) framework for dealing with domains containing CPUs (Ulf Hansson). - Prevent sysfs attributes representing low-power S0 residency counters from being exposed if low-power S0 support is not indicated in ACPI FADT (Rajneesh Bhardwaj). - Get rid of custom CPU features macros for Intel CPUs from the intel_idle and RAPL drivers (Andy Shevchenko). - Update the tasks freezer to list tasks that refused to freeze and caused a system transition to a sleep state to be aborted (Todd Brandt). - Update the pm-graph set of tools to v5.2 (Todd Brandt). - Fix some issues in the cpupower utility (Anders Roxell, Prarit Bhargava)" * tag 'pm-4.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (73 commits) PM / Domains: Document flags for genpd PM / Domains: Deal with multiple states but no governor in genpd PM / Domains: Don't treat zero found compatible idle states as an error cpuidle: menu: Avoid computations when result will be discarded cpuidle: menu: Drop redundant comparison cpufreq: tegra186: don't pass GFP_DMA32 to dma_alloc_coherent() cpufreq: conservative: Take limits changes into account properly Documentation: intel_pstate: Add base_frequency information cpufreq: intel_pstate: Add base_frequency attribute ACPI / CPPC: Add support for guaranteed performance cpuidle: menu: Simplify checks related to the polling state PM / tools: sleepgraph and bootgraph: upgrade to v5.2 PM / tools: sleepgraph: first batch of v5.2 changes cpupower: Fix coredump on VMWare cpupower: Fix AMD Family 0x17 msr_pstate size cpufreq: imx6q: read OCOTP through nvmem for imx6ul/imx6ull cpufreq: dt-platdev: allow RK3399 to have separate tunables per cluster cpuidle: poll_state: Revise loop termination condition cpuidle: menu: Move the latency_req == 0 special case check cpuidle: menu: Avoid computations for very close timers ...