- 07 Apr, 2020 19 commits
-
-
Matthew Wilcox (Oracle) authored
If THP is disabled, find_subpage() can become a no-op by using hpage_nr_pages() instead of compound_nr(). hpage_nr_pages() embeds a check for PageTail, so we can drop the check here. Signed-off-by:
Matthew Wilcox (Oracle) <willy@infradead.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Reviewed-by:
Christoph Hellwig <hch@lst.de> Acked-by:
Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Cc: Pankaj Gupta <pankaj.gupta.linux@gmail.com> Link: http://lkml.kernel.org/r/20200318140253.6141-5-willy@infradead.orgSigned-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
David Rientjes authored
The thp_fault_fallback and thp_file_fallback vmstats are incremented if either the hugepage allocation fails through the page allocator or the hugepage charge fails through mem cgroup. This patch leaves this field untouched but adds two new fields, thp_{fault,file}_fallback_charge, which is incremented only when the mem cgroup charge fails. This distinguishes between attempted hugepage allocations that fail due to fragmentation (or low memory conditions) and those that fail due to mem cgroup limits. That can be used to determine the impact of fragmentation on the system by excluding faults that failed due to memcg usage. Signed-off-by:David Rientjes <rientjes@google.com> Signed-off-by:
Yang Shi <yang.shi@linux.alibaba.com> Acked-by:
-
David Rientjes authored
The existing thp_fault_fallback indicates when thp attempts to allocate a hugepage but fails, or if the hugepage cannot be charged to the mem cgroup hierarchy. Extend this to shmem as well. Adds a new thp_file_fallback to complement thp_file_alloc that gets incremented when a hugepage is attempted to be allocated but fails, or if it cannot be charged to the mem cgroup hierarchy. Additionally, remove the check for CONFIG_TRANSPARENT_HUGE_PAGECACHE from shmem_alloc_hugepage() since it is only called with this configuration option. Signed-off-by:
-
Yang Shi authored
Currently the migration code doesn't migrate PG_readahead flag. Theoretically this would incur slight performance loss as the application might have to ramp its readahead back up again. Even though such problem happens, it might be hidden by something else since migration is typically triggered by compaction and NUMA balancing, any of which should be more noticeable. Migrate the flag after end_page_writeback() since it may clear PG_reclaim flag, which is the same bit as PG_readahead, for the new page. [akpm@linux-foundation.org: tweak comment] Signed-off-by:
-
Wei Yang authored
It can currently happen that we store the status of a page twice: * Once we detect that it is already on the target node * Once we moved a bunch of pages, and a page that's already on the target node is contained in the current interval. Let's simplify the code and always call do_move_pages_to_node() in case we did not queue a page for migration. Note that pages that are already on the target node are not added to the pagelist and are, therefore, ignored by do_move_pages_to_node() - there is no functional change. The status of such a page is now only stored once. [david@redhat.com rephrase changelog] Signed-off-by:
Wei Yang <richardw.yang@linux.intel.com> Signed-off-by:
David Hildenbrand <david@redhat.com> Acked-by:
Michal Hocko <mhocko@suse.com> Link: http://lkml.kernel.org/r/20200214003017.25558-5-richardw.yang@linux.intel.comSigned-off-by:
-
Wei Yang authored
When pagelist is empty, it is not necessary to do the move and store. Also it consolidate the empty list check in one place. Signed-off-by:
-
Wei Yang authored
Usually, do_move_pages_to_node() and store_status() are used in combination. We have three similar call sites. Let's provide a wrapper for both function calls - move_pages_and_store_status - to make the calling code easier to maintain and fix (as noted by Yang Shi, the return value handling of do_move_pages_to_node() has a flaw). [david@redhat.com rephrase changelog] Signed-off-by:
-
Wei Yang authored
Patch series "cleanup on do_pages_move()", v5. The logic in do_pages_move() is a little mess for audience to read and has some potential error on handling the return value. Especially there are three calls on do_move_pages_to_node() and store_status() with almost the same form. This patch set tries to make the code a little friendly for audience by consolidate the calls. This patch (of 4): At this point, we always have i >= start. If i == start, store_status() will return 0. So we can drop the check for i > start. [david@redhat.com rephrase changelog] Signed-off-by:
-
Michal Hocko authored
While it might be really clear to MM developers that gfp reclaim modifiers are applicable only to sleepable allocations (those with __GFP_DIRECT_RECLAIM) it seems that actual users of the API are not always sure. Make it explicit that they are not applicable for GFP_NOWAIT or GFP_ATOMIC allocations which are the most commonly used non-sleepable allocation masks. Signed-off-by:
Joel Fernandes (Google) <joel@joelfernandes.org> Acked-by:
Paul E. McKenney <paulmck@kernel.org> Acked-by:
-
Qiujun Huang authored
There is a typo in comment, fix it. "exeeds" -> "exceeds" Signed-off-by:
Qiujun Huang <hqjagain@gmail.com> Signed-off-by:
-
Anshuman Khandual authored
It is unlikely that an inaccessible VMA without required permission flags will get a page fault. Hence lets just append unlikely() directive to such checks in order to improve performance while also standardizing it across various platforms. Signed-off-by:
Anshuman Khandual <anshuman.khandual@arm.com> Signed-off-by:
-
Anshuman Khandual authored
This replaces all remaining open encodings with vma_is_anonymous(). Signed-off-by:
-
Anshuman Khandual authored
This replaces all remaining open encodings with is_vm_hugetlb_page(). Signed-off-by:
Vlastimil Babka <vbabka@suse.cz> Cc: Paul Mackerras <paulus@ozlabs.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Will Deacon <will@kernel.org> Cc: "Aneesh Kumar K.V" <aneesh.kumar@linux.ibm.com> Cc: Nick Piggin <npiggin@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Ingo Molnar <mingo@redhat.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Andy Lutomirski <luto@kernel.org> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Guo Ren <guoren@kernel.org> Cc: Mel Gorman <mgorman@suse.de> Cc: Paul Burton <paulburton@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Rich Felker <dalias@libc.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Link: http://lkml.kernel.org/r/1582520593-30704-4-git-send-email-anshuman.khandual@arm.comSigned-off-by:
-
Anshuman Khandual authored
Lets move vma_is_accessible() helper to include/linux/mm.h which makes it available for general use. While here, this replaces all remaining open encodings for VMA access check with vma_is_accessible(). Signed-off-by:
Geert Uytterhoeven <geert@linux-m68k.org> Acked-by:
Guo Ren <guoren@kernel.org> Acked-by:
-
Anshuman Khandual authored
Patch series "mm/vma: Use all available wrappers when possible", v2. Apart from adding a VMA flag readable name for trace purpose, this series does some open encoding replacements with availabe VMA specific wrappers. This skips VM_HUGETLB check in vma_migratable() as its already being done with another patch (https://patchwork.kernel.org/patch/11347831/) which is yet to be merged. This patch (of 4): This just adds the missing readable name for VM_SYNC. Signed-off-by:
-
Li Xinhai authored
Set ->vm_next and ->vm_prev to NULL to prevent potential misuse from the new duplicated vma. Currently, only in fork path there are misuse for handling anon_vma. No other bugs been revealed with this patch applied. Signed-off-by:
Li Xinhai <lixinhai.lxh@gmail.com> Signed-off-by:
-
Li Xinhai authored
This reverts commit 4e4a9eb9 ("mm/rmap.c: reuse mergeable anon_vma as parent when fork"). In dup_mmap(), anon_vma_fork() is called for attaching anon_vma and parameter 'tmp' (i.e., the new vma of child) has same ->vm_next and ->vm_prev as its parent vma. That causes the anon_vma used by parent been mistakenly shared by child (In anon_vma_clone(), the code added by that commit will do this reuse work). Besides this issue, the design of reusing anon_vma from vma which has gone through fork should be avoided ([1]). So, this patch reverts that commit and maintains the consistent logic of reusing anon_vma for fork/split/merge vma. Reusing anon_vma within the process is fine. But if a vma has gone through fork(), then that vma's anon_vma should not be shared with its neighbor vma. As explained in [1], when vma gone through fork(), the check for list_is_singular(vma->anon_vma_chain) will be false, and don't share anon_vma. With current issue, one example can clarify more. Parent process do below two steps: 1. p_vma_1 is created and p_anon_vma_1 is prepared; 2. p_vma_2 is created and share p_anon_vma_1; (this is allowed, becaues p_vma_1 didn't gothrough fork()); parent process do fork(): 3. c_vma_1 is dup from p_vma_1, and has its own c_anon_vma_1 prepared; at this point, c_vma_1->anon_vma_chain has two items, one for p_anon_vma_1 and one for c_anon_vma_1; 4. c_vma_2 is dup from p_vma_2, it is not allowed to share c_anon_vma_1, because c_vma_1->anon_vma_chain has two items. [1] commit d0e9fe17 ("Simplify and comment on anon_vma re-use for anon_vma_prepare()") explains the test of "list_is_singular()". Fixes: 4e4a9eb9 ("mm/rmap.c: reuse mergeable anon_vma as parent when fork") Signed-off-by:
-
Li Xinhai authored
Patch series "mm: Fix misuse of parent anon_vma in dup_mmap path". This patchset fixes the misuse of parenet anon_vma, which mainly caused by child vma's vm_next and vm_prev are left same as its parent after duplicate vma. Finally, code reached parent vma's neighbor by referring pointer of child vma and executed wrong logic. The first two patches fix relevant issues, and the third patch sets vm_next and vm_prev to NULL when duplicate vma to prevent potential misuse in future. Effects of the first bug is that causes rmap code to check both parent and child's page table, although a page couldn't be mapped by both parent and child, because child vma has WIPEONFORK so all pages mapped by child are 'new' and not relevant to parent. Effects of the second bug is that the relationship of anon_vma of parent and child are totallyconvoluted. It would cause 'son', 'grandson', ..., etc, to share 'parent' anon_vma, which disobey the design rule of reusing anon_vma (the rule to be followed is that reusing should among vma of same process, and vma should not gone through fork). So, both issues should cause unnecessary rmap walking and have unexpected complexity. These two issues would not be directly visible, I used debugging code to check the anon_vma pointers of parent and child when inspecting the suspicious implementation of issue #2, then find the problem. This patch (of 3): In dup_mmap(), anon_vma_prepare() is called for vma has VM_WIPEONFORK, and parameter 'tmp' (i.e., the new vma of child) has same ->vm_next and ->vm_prev as its parent vma. That allows anon_vma used by parent been mistakenly shared by child (find_mergeable_anon_vma() will do this reuse work). Besides this issue, call anon_vma_prepare() should be avoided because we don't copy page for this vma. Preparing anon_vma will be handled during fault. Fixes: d2cd9ede ("mm,fork: introduce MADV_WIPEONFORK") Signed-off-by:
-
Chris Down authored
The root of the hierarchy cannot have high set, so we will never reclaim based on it. This makes that clearer and avoids another entry. Signed-off-by:
Chris Down <chris@chrisdown.name> Signed-off-by:
Johannes Weiner <hannes@cmpxchg.org> Cc: Tejun Heo <tj@kernel.org> Cc: Roman Gushchin <guro@fb.com> Cc: Michal Hocko <mhocko@kernel.org> Link: http://lkml.kernel.org/r/20200312164137.GA1753625@chrisdown.nameSigned-off-by:
-
- 06 Apr, 2020 11 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pmLinus Torvalds authored
Pull more ACPI updates from Rafael Wysocki: "Additional ACPI updates. These update the ACPICA code in the kernel to the 20200326 upstream revision, fix an ACPI-related CPU hotplug deadlock on x86, update Intel Tiger Lake device IDs in some places, add a new ACPI backlight blacklist entry, update the "acpi_backlight" kernel command line switch documentation and clean up a CPPC library routine. Specifics: - Update the ACPICA code in the kernel to upstream revision 20200326 including: * Fix for a typo in a comment field (Bob Moore) * acpiExec namespace init file fixes (Bob Moore) * Addition of NHLT to the known tables list (Cezary Rojewski) * Conversion of PlatformCommChannel ASL keyword to PCC (Erik Kaneda) * acpiexec cleanup (Erik Kaneda) * WSMT-related typo fix (Erik Kaneda) * sprintf() utility function fix (John Levon) * IVRS IVHD type 11h parsing implementation (Michał Żygowski) * IVRS IVHD type 10h reserved field name fix (Michał Żygowski) - Fix ACPI-related CPU hotplug deadlock on x86 (Qian Cai) - Fix Intel Tiger Lake ACPI device IDs in several places (Gayatri Kammela) - Add ACPI backlight blacklist entry for Acer Aspire 5783z (Hans de Goede) - Fix documentation of the "acpi_backlight" kernel command line switch (Randy Dunlap) - Clean up the acpi_get_psd_map() CPPC library routine (Liguang Zhang)" * tag 'acpi-5.7-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: x86: ACPI: fix CPU hotplug deadlock thermal: int340x_thermal: fix: Update Tiger Lake ACPI device IDs platform/x86: intel-hid: fix: Update Tiger Lake ACPI device ID ACPI: Update Tiger Lake ACPI device IDs ACPI: video: Use native backlight on Acer Aspire 5783z ACPI: video: Docs update for "acpi_backlight" kernel parameter options ACPICA: Update version 20200326 ACPICA: Fixes for acpiExec namespace init file ACPICA: Add NHLT table signature ACPICA: WSMT: Fix typo, no functional change ACPICA: utilities: fix sprintf() ACPICA: acpiexec: remove redeclaration of acpi_gbl_db_opt_no_region_support ACPICA: Change PlatformCommChannel ASL keyword to PCC ACPICA: Fix IVRS IVHD type 10h reserved field name ACPICA: Implement IVRS IVHD type 11h parsing ACPICA: Fix a typo in a comment field ACPI: CPPC: clean up acpi_get_psd_map() -
git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pmLinus Torvalds authored
Pull more power management updates from Rafael Wysocki: "Additional power management updates. These fix a corner-case suspend-to-idle wakeup issue on systems where the ACPI SCI is shared with another wakeup source, add a kernel command line option to set pm_debug_messages via the kernel command line, add a document desctibing system-wide suspend and resume code flows, modify cpufreq Kconfig to choose schedutil as the preferred governor by default in a couple of cases and do some assorted cleanups. Specifics: - Fix corner-case suspend-to-idle wakeup issue on systems where the ACPI SCI is shared with another wakeup source (Hans de Goede). - Add document describing system-wide suspend and resume code flows to the admin guide (Rafael Wysocki). - Add kernel command line option to set pm_debug_messages (Chen Yu). - Choose schedutil as the preferred scaling governor by default on ARM big.LITTLE systems and on x86 systems using the intel_pstate driver in the passive mode (Linus Walleij, Rafael Wysocki). - Drop racy and redundant checks from the PM core's device_prepare() routine (Rafael Wysocki). - Make resume from hibernation take the hibernation_restore() return value into account (Dexuan Cui)" * tag 'pm-5.7-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: platform/x86: intel_int0002_vgpio: Use acpi_register_wakeup_handler() ACPI: PM: Add acpi_[un]register_wakeup_handler() Documentation: PM: sleep: Document system-wide suspend code flows cpufreq: Select schedutil when using big.LITTLE PM: sleep: Add pm_debug_messages kernel command line option PM: sleep: core: Drop racy and redundant checks from device_prepare() PM: hibernate: Propagate the return value of hibernation_restore() cpufreq: intel_pstate: Select schedutil as the default governor -
git://github.com/c-sky/csky-linuxLinus Torvalds authored
Pull csky updates from Guo Ren: - Add kproobes/uprobes support - Add lockdep, rseq, gcov support - Fixup init_fpu - Fixup ftrace_modify deadlock - Fixup speculative execution on IO area * tag 'csky-for-linus-5.7-rc1' of git://github.com/c-sky/csky-linux: csky: Fixup cpu speculative execution to IO area csky: Add uprobes support csky: Add kprobes supported csky: Enable LOCKDEP_SUPPORT csky: Enable the gcov function csky: Fixup get wrong psr value from phyical reg csky/ftrace: Fixup ftrace_modify_code deadlock without CPU_HAS_ICACHE_INS csky: Implement ftrace with regs csky: Add support for restartable sequence csky: Implement ptrace regs and stack API csky: Fixup init_fpu compile warning with __init
-
git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fsLinus Torvalds authored
Pull fsnotify updates from Jan Kara: "This implements the fanotify FAN_DIR_MODIFY event. This event reports the name in a directory under which a change happened and together with the directory filehandle and fstatat() allows reliable and efficient implementation of directory synchronization" * tag 'fsnotify_for_v5.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs: fanotify: Fix the checks in fanotify_fsid_equal fanotify: report name info for FAN_DIR_MODIFY event fanotify: record name info for FAN_DIR_MODIFY event fanotify: Drop fanotify_event_has_fid() fanotify: prepare to report both parent and child fid's fanotify: send FAN_DIR_MODIFY event flavor with dir inode and name fanotify: divorce fanotify_path_event and fanotify_fid_event fanotify: Store fanotify handles differently fanotify: Simplify create_fd() fanotify: fix merging marks masks with FAN_ONDIR fanotify: merge duplicate events on parent and child fsnotify: replace inode pointer with an object id fsnotify: simplify arguments passing to fsnotify_parent() fsnotify: use helpers to access data by data_type fsnotify: funnel all dirent events through fsnotify_name() fsnotify: factor helpers fsnotify_dentry() and fsnotify_file() fsnotify: tidy up FS_ and FAN_ constants
-
git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fsLinus Torvalds authored
Pull ext2/udf updates from Jan Kara: "Cleanups and fixes for ext2 and one cleanup for udf" * tag 'for_v5.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs: ext2: fix empty body warnings when -Wextra is used ext2: fix debug reference to ext2_xattr_cache udf: udf_sb.h: Replace zero-length array with flexible-array member ext2: xattr.h: Replace zero-length array with flexible-array member ext2: Silence lockdep warning about reclaim under xattr_sem
-
git://github.com/martinetd/linuxLinus Torvalds authored
Pull 9p updates from Dominique Martinet: "Not much new, but a few patches for this cycle: - Fix read with O_NONBLOCK to allow incomplete read and return immediately - Rest is just cleanup (indent, unused field in struct, extra semicolon)" * tag '9p-for-5.7' of git://github.com/martinetd/linux: net/9p: remove unused p9_req_t aux field 9p: read only once on O_NONBLOCK 9pnet: allow making incomplete read requests 9p: Remove unneeded semicolon 9p: Fix Kconfig indentation -
git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfsLinus Torvalds authored
Pull vfs pathwalk fix from Al Viro: "Dumb braino in legitimize_path()..." * 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: fix a braino in legitimize_path()
-
Al Viro authored
brown paperbag time... wrong order of arguments ended up confusing the values to check dentry and mount_lock seqcounts against. Reported-by:
kernel test robot <rong.a.chen@intel.com> Fixes: 2aa38470 ("non-RCU analogue of the previous commit") Tested-by:
Al Viro <viro@zeniv.linux.org.uk>
-
Rafael J. Wysocki authored
* acpi-cppc: ACPI: CPPC: clean up acpi_get_psd_map() * acpi-video: ACPI: video: Use native backlight on Acer Aspire 5783z ACPI: video: Docs update for "acpi_backlight" kernel parameter options * acpi-drivers: thermal: int340x_thermal: fix: Update Tiger Lake ACPI device IDs platform/x86: intel-hid: fix: Update Tiger Lake ACPI device ID ACPI: Update Tiger Lake ACPI device IDs
-
Rafael J. Wysocki authored
* acpica: ACPICA: Update version 20200326 ACPICA: Fixes for acpiExec namespace init file ACPICA: Add NHLT table signature ACPICA: WSMT: Fix typo, no functional change ACPICA: utilities: fix sprintf() ACPICA: acpiexec: remove redeclaration of acpi_gbl_db_opt_no_region_support ACPICA: Change PlatformCommChannel ASL keyword to PCC ACPICA: Fix IVRS IVHD type 10h reserved field name ACPICA: Implement IVRS IVHD type 11h parsing ACPICA: Fix a typo in a comment field
-
Rafael J. Wysocki authored
* pm-sleep: Documentation: PM: sleep: Document system-wide suspend code flows PM: sleep: Add pm_debug_messages kernel command line option PM: sleep: core: Drop racy and redundant checks from device_prepare() PM: hibernate: Propagate the return value of hibernation_restore() * pm-cpufreq: cpufreq: Select schedutil when using big.LITTLE cpufreq: intel_pstate: Select schedutil as the default governor
-
- 05 Apr, 2020 10 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/sre/linux-power-supplyLinus Torvalds authored
Pull power supply and reset changes from Sebastian Reichel: "Core: - Nothing Drivers: - at91-reset: cleanups, proper handling for sam9x60 - sc27xx, charger-manager: allow building as module - sc27xx: add support to read current charge capacity - axp288: more quirks for weird hardware - misc fixes" * tag 'for-v5.7' of git://git.kernel.org/pub/scm/linux/kernel/git/sre/linux-power-supply: (26 commits) power: reset: sc27xx: Allow the SC27XX poweroff driver building into a module power: reset: sc27xx: Change to use cpu_down() power: reset: sc27xx: Power off the external subsystems' connection power: twl4030: Use scnprintf() for avoiding potential buffer overflow power: supply: bq27xxx_battery: Silence deferred-probe error power: reset: at91-reset: handle nrst async for sam9x60 power: reset: at91-reset: get rid of at91_reset_data power: reset: at91-reset: keep only one reset function power: reset: at91-reset: make at91sam9g45_restart() generic power: reset: at91-reset: introduce ramc_lpr to struct at91_reset power: reset: at91-reset: use r4 as tmp argument power: reset: at91-reset: introduce args member in at91_reset_data power: reset: at91-reset: introduce struct at91_reset_data power: reset: at91-reset: devm_kzalloc() for at91_reset data structure power: reset: at91-reset: pass rstc base address to at91_reset_status() power: reset: at91-reset: convert reset in pointer to struct at91_reset power: reset: at91-reset: add notifier block to struct at91_reset power: reset: at91-reset: add sclk to struct at91_reset power: reset: at91-reset: add ramc_base[] to struct at91_reset power: reset: at91-reset: introduce struct at91_reset ...
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull more perf updates from Thomas Gleixner: "Perf updates all over the place: core: - Support for cgroup tracking in samples to allow cgroup based analysis tools: - Support for cgroup analysis - Commandline option and hotkey for perf top to change the sort order - A set of fixes all over the place - Various build system related improvements - Updates of the X86 pmu event JSON data - Documentation updates" * tag 'perf-urgent-2020-04-05' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (55 commits) perf python: Fix clang detection to strip out options passed in $CC perf tools: Support Python 3.8+ in Makefile perf script: Fix invalid read of directory entry after closedir() perf script report: Fix SEGFAULT when using DWARF mode perf script: add -S/--symbols documentation perf pmu-events x86: Use CPU_CLK_UNHALTED.THREAD in Kernel_Utilization metric perf events parser: Add missing Intel CPU events to parser perf script: Allow --symbol to accept hexadecimal addresses perf report/top TUI: Fix title line formatting perf top: Support hotkey to change sort order perf top: Support --group-sort-idx to change the sort order perf symbols: Fix arm64 gap between kernel start and module end perf build-test: Honour JOBS to override detection of number of cores perf script: Add --show-cgroup-events option perf top: Add --all-cgroups option perf record: Add --all-cgroups option perf record: Support synthesizing cgroup events perf report: Add 'cgroup' sort key perf cgroup: Maintain cgroup hierarchy perf tools: Basic support for CGROUP event ... -
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull timer fixes from Thomas Gleixner: "Two timer subsystem fixes: - Prevent a use after free in the new lockdep state tracking for hrtimers - Add missing parenthesis in the VF pit timer driver" * tag 'timers-urgent-2020-04-05' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: clocksource/drivers/timer-vf-pit: Add missing parenthesis hrtimer: Don't dereference the hrtimer pointer after the callback -
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull irq fixes from Thomas Gleixner: "Two reverts addressing regressions of the Xilinx interrupt controller driver which affected the PPC users" * tag 'irq-urgent-2020-04-05' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: Revert "irqchip/xilinx: Enable generic irq multi handler" Revert "irqchip/xilinx: Do not call irq_set_default_host()"
-
Linus Torvalds authored
Commit 9255782f ("sysfs: Wrap __compat_only_sysfs_link_entry_to_kobj function to change the symlink name") made this function a wrapper around a new non-underscored function, which is a bit odd. The normal naming convention is the other way around: the underscored function is the wrappee, and the non-underscored function is the wrapper. There's only one single user (well, two call-sites in that user) of the more limited double underscore version of this function, so just remove the oddly named wrapper entirely and just add the extra NULL argument to the user. I considered just doing that in the merge, but that tends to make history really hard to read. Link: https://lore.kernel.org/lkml/CAHk-=wgkkmNV5tMzQDmPAQuNJBuMcry--Jb+h8H1o4RA3kF7QQ@mail.gmail.com/ Cc: Sourabh Jain <sourabhjain@linux.ibm.com> Cc: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linuxLinus Torvalds authored
Pull powerpc updates from Michael Ellerman: "Slightly late as I had to rebase mid-week to insert a bug fix: - A large series from Nick for 64-bit to further rework our exception vectors, and rewrite portions of the syscall entry/exit and interrupt return in C. The result is much easier to follow code that is also faster in general. - Cleanup of our ptrace code to split various parts out that had become badly intertwined with #ifdefs over the years. - Changes to our NUMA setup under the PowerVM hypervisor which should hopefully avoid non-sensical topologies which can lead to warnings from the workqueue code and other problems. - MAINTAINERS updates to remove some of our old orphan entries and update the status of others. - Quite a few other small changes and fixes all over the map. Thanks to: Abdul Haleem, afzal mohammed, Alexey Kardashevskiy, Andrew Donnellan, Aneesh Kumar K.V, Balamuruhan S, Cédric Le Goater, Chen Zhou, Christophe JAILLET, Christophe Leroy, Christoph Hellwig, Clement Courbet, Daniel Axtens, David Gibson, Douglas Miller, Fabiano Rosas, Fangrui Song, Ganesh Goudar, Gautham R. Shenoy, Greg Kroah-Hartman, Greg Kurz, Gustavo Luiz Duarte, Hari Bathini, Ilie Halip, Jan Kara, Joe Lawrence, Joe Perches, Kajol Jain, Larry Finger, Laurentiu Tudor, Leonardo Bras, Libor Pechacek, Madhavan Srinivasan, Mahesh Salgaonkar, Masahiro Yamada, Masami Hiramatsu, Mauricio Faria de Oliveira, Michael Neuling, Michal Suchanek, Mike Rapoport, Nageswara R Sastry, Nathan Chancellor, Nathan Lynch, Naveen N. Rao, Nicholas Piggin, Nick Desaulniers, Oliver O'Halloran, Po-Hsu Lin, Pratik Rajesh Sampat, Rasmus Villemoes, Ravi Bangoria, Roman Bolshakov, Sam Bobroff, Sandipan Das, Santosh S, Sedat Dilek, Segher Boessenkool, Shilpasri G Bhat, Sourabh Jain, Srikar Dronamraju, Stephen Rothwell, Tyrel Datwyler, Vaibhav Jain, YueHaibing" * tag 'powerpc-5.7-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux: (158 commits) powerpc: Make setjmp/longjmp signature standard powerpc/cputable: Remove unnecessary copy of cpu_spec->oprofile_type powerpc: Suppress .eh_frame generation powerpc: Drop -fno-dwarf2-cfi-asm powerpc/32: drop unused ISA_DMA_THRESHOLD powerpc/powernv: Add documentation for the opal sensor_groups sysfs interfaces selftests/powerpc: Fix try-run when source tree is not writable powerpc/vmlinux.lds: Explicitly retain .gnu.hash powerpc/ptrace: move ptrace_triggered() into hw_breakpoint.c powerpc/ptrace: create ppc_gethwdinfo() powerpc/ptrace: create ptrace_get_debugreg() powerpc/ptrace: split out ADV_DEBUG_REGS related functions. powerpc/ptrace: move register viewing functions out of ptrace.c powerpc/ptrace: split out TRANSACTIONAL_MEM related functions. powerpc/ptrace: split out SPE related functions. powerpc/ptrace: split out ALTIVEC related functions. powerpc/ptrace: split out VSX related functions. powerpc/ptrace: drop PARAMETER_SAVE_AREA_OFFSET powerpc/ptrace: drop unnecessary #ifdefs CONFIG_PPC64 powerpc/ptrace: remove unused header includes ... -
git://git.kernel.org/pub/scm/linux/kernel/git/tytso/randomLinus Torvalds authored
Pull /dev/random updates from Ted Ts'o: - Improve getrandom and /dev/random's support for those arm64 architecture variants that have RNG instructions. - Use batched output from CRNG instead of CPU's RNG instructions for better performance. - Miscellaneous bug fixes. * tag 'random_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/random: random: avoid warnings for !CONFIG_NUMA builds random: fix data races at timer_rand_state random: always use batched entropy for get_random_u{32,64} random: Make RANDOM_TRUST_CPU depend on ARCH_RANDOM arm64: add credited/trusted RNG support random: add arch_get_random_*long_early() random: split primary/secondary crng init paths
-
git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4Linus Torvalds authored
Pull ext4 updates from Ted Ts'o: - Replace ext4's bmap and iopoll implementations to use iomap. - Clean up extent tree handling. - Other cleanups and miscellaneous bug fixes * tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4: (31 commits) ext4: save all error info in save_error_info() and drop ext4_set_errno() ext4: fix incorrect group count in ext4_fill_super error message ext4: fix incorrect inodes per group in error message ext4: don't set dioread_nolock by default for blocksize < pagesize ext4: disable dioread_nolock whenever delayed allocation is disabled ext4: do not commit super on read-only bdev ext4: avoid ENOSPC when avoiding to reuse recently deleted inodes ext4: unregister sysfs path before destroying jbd2 journal ext4: check for non-zero journal inum in ext4_calculate_overhead ext4: remove map_from_cluster from ext4_ext_map_blocks ext4: clean up ext4_ext_insert_extent() call in ext4_ext_map_blocks() ext4: mark block bitmap corrupted when found instead of BUGON ext4: use flexible-array member for xattr structs ext4: use flexible-array member in struct fname Documentation: correct the description of FIEMAP_EXTENT_LAST ext4: move ext4_fiemap to use iomap framework ext4: make ext4_ind_map_blocks work with fiemap ext4: move ext4 bmap to use iomap infrastructure ext4: optimize ext4_ext_precache for 0 depth ext4: add IOMAP_F_MERGED for non-extent based mapping ...
-
git://git.kernel.org/pub/scm/linux/kernel/git/clk/linuxLinus Torvalds authored
Pull clk updates from Stephen Boyd: "There's not much to see in the core framework this time around. Instead the majority of the diff is the normal collection of driver additions for new SoCs and non-critical clk data fixes and updates. The framework must be middle aged. The two biggest directories in the diffstat show that the Qualcomm and Unisoc support added a handful of big drivers for new SoCs but that's not really the whole story because those new drivers tend to add large numbers of lines of clk data. There's a handful of AT91 clk drivers added this time around too and a bunch of improvements to drivers like the i.MX driver. All around lots of updates and fixes in various clk drivers which is good to see. The core framework has only one real major change which has been baking in next for the past couple months. It fixes the framework so that it stops caching a clk's phase when the phase clk_op returns an error. Before this change we would consider some negative errno as a phase and that just doesn't make sense. Core: - Don't show clk phase when it is invalid New Drivers: - Add support for Unisoc SC9863A clks - Qualcomm SM8250 RPMh and MSM8976 RPM clks - Qualcomm SM8250 Global Clock Controller (GCC) support - Qualcomm SC7180 Modem Clock Controller (MSS CC) support - EHRPWM's TimeBase clock(TBCLK) for TI AM654 SoCs - Support PMC clks on at91sam9n12, at91rm9200, sama5d3, and at91sam9g45 SoCs Updates: - GPU GX GDSC support on Qualcomm sc7180 - Fixes and improvements for the Marvell MMP2/MMP3 SoC clk drivers - A series from Anson to convert i.MX8 clock bindings to json-schema - Update i.MX pll14xx driver to include new frequency entries for pll1443x table, and return error for invalid PLL type - Add missing of_node_put() call for a number of i.MX clock drivers - Drop flag CLK_IS_CRITICAL from 'A53_CORE' mux clock, as we already have the flag on its child cpu clock - Fix a53 cpu clock for i.MX8 drivers to get it source from ARM PLL via CORE_SEL slice, and source from A53 CCM clk root when we need to change ARM PLL frequency. Thus, we can support core running above 1GHz safely - Update i.MX pfdv2 driver to check zero rate and use determine_rate for getting the best rate - Add CLKO2 for imx8mm, SNVS clock for imx8mn, and PXP clock for imx7d - Remove PMC clks from Tegra clk driver - Improved clock/reset handling for the Renesas R-Car USB2 Clock Selector - Conversion to json-schema of the Renesas CPG/MSSR DT bindings - Add Crypto clocks on Renesas R-Car M3-W/W+, M3-N, E3, and D3 - Add RPC (QSPI/HyperFLASH) clocks on Renesas R-Car H3, M3-W/W+, and M3-N - Update Amlogic audio clock gate hierarchy for meson8 and gxbb - Update Amlogic g12a spicc clock sources - Support for Ingenic X1000 TCU clks" * tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux: (146 commits) clk: sprd: fix to get a correct ibias of pll dt-bindings: imx8mm-clock: Fix the file path dt-bindings: imx8mq-clock: Fix the file path clk: qcom: rpmh: Drop unnecessary semicolons clk: qcom: rpmh: Simplify clk_rpmh_bcm_send_cmd() clk: tegra: Use NULL for pointer initialization clk: sprd: add clocks support for SC9863A clk: sprd: support to get regmap from parent node clk: sprd: Add macros for referencing parents without strings clk: sprd: Add dt-bindings include file for SC9863A dt-bindings: clk: sprd: add bindings for sc9863a clock controller dt-bindings: clk: sprd: rename the common file name sprd.txt to SoC specific clk: sprd: add gate for pll clocks MAINTAINERS: dt: update reference for arm-integrator.txt clk: mmp2: Fix bit masks for LCDC I/O and pixel clocks clk: mmp2: Add clock for fifth SD HCI on MMP3 dt-bindings: marvell,mmp2: Add clock id for the fifth SD HCI on MMP3 clk: mmp2: Add clocks for the thermal sensors dt-bindings: marvell,mmp2: Add clock ids for the thermal sensors clk: mmp2: add the GPU clocks ... -
git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-traceLinus Torvalds authored
Pull tracing updates from Steven Rostedt: "New tracing features: - The ring buffer is no longer disabled when reading the trace file. The trace_pipe file was made to be used for live tracing and reading as it acted like the normal producer/consumer. As the trace file would not consume the data, the easy way of handling it was to just disable writes to the ring buffer. This came to a surprise to the BPF folks who complained about lost events due to reading. This is no longer an issue. If someone wants to keep the old disabling there's a new option "pause-on-trace" that can be set. - New set_ftrace_notrace_pid file. PIDs in this file will not be traced by the function tracer. Similar to set_ftrace_pid, which makes the function tracer only trace those tasks with PIDs in the file, the set_ftrace_notrace_pid does the reverse. - New set_event_notrace_pid file. PIDs in this file will cause events not to be traced if triggered by a task with a matching PID. Similar to the set_event_pid file but will not be traced. Note, sched_waking and sched_switch events may still be traced if one of the tasks referenced by those events contains a PID that is allowed to be traced. Tracing related features: - New bootconfig option, that is attached to the initrd file. If bootconfig is on the command line, then the initrd file is searched looking for a bootconfig appended at the end. - New GPU tracepoint infrastructure to help the gfx drivers to get off debugfs (acked by Greg Kroah-Hartman) And other minor updates and fixes" * tag 'trace-v5.7' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace: (27 commits) tracing: Do not allocate buffer in trace_find_next_entry() in atomic tracing: Add documentation on set_ftrace_notrace_pid and set_event_notrace_pid selftests/ftrace: Add test to test new set_event_notrace_pid file selftests/ftrace: Add test to test new set_ftrace_notrace_pid file tracing: Create set_event_notrace_pid to not trace tasks ftrace: Create set_ftrace_notrace_pid to not trace tasks ftrace: Make function trace pid filtering a bit more exact ftrace/kprobe: Show the maxactive number on kprobe_events tracing: Have the document reflect that the trace file keeps tracing enabled ring-buffer/tracing: Have iterator acknowledge dropped events tracing: Do not disable tracing when reading the trace file ring-buffer: Do not disable recording when there is an iterator ring-buffer: Make resize disable per cpu buffer instead of total buffer ring-buffer: Optimize rb_iter_head_event() ring-buffer: Do not die if rb_iter_peek() fails more than thrice ring-buffer: Have rb_iter_head_event() handle concurrent writer ring-buffer: Add page_stamp to iterator for synchronization ring-buffer: Rename ring_buffer_read() to read_buffer_iter_advance() ring-buffer: Have ring_buffer_empty() not depend on tracing stopped tracing: Save off entry when peeking at next entry ...
-
