- 28 Dec, 2021 23 commits
-
-
Florian Fainelli authored
Now that we can utilize the BCM7038_WDT driver, remove that one which was not converted to the watchdog APIs. There are a couple of notable differences with how the bcm7038_wdt driver proceeds: - bcm63xx_wdt would register with the ad-hoc BCM63xx hardware timer API, but this would only be used in order to catch the interrupt *before* a SoC reset and make the kernel "die" - bcm6xx_wdt would register a software timer and kick it every second in order to pet the watchdog, thus offering a two step watchdog process. This is not something that is brought over to the bcm7038_wdt as it is deemed unnecessary. If user-space cannot pet the watchdog, but a kernel timer can, the system is still in a bad shape anyway. bcm7038_wdt is simpler in its behavior and behaves as a standard watchdog driver and is not making use of any specific platform APIs, therefore making it more maintainable and extensible. Signed-off-by:
Florian Fainelli <f.fainelli@gmail.com> Reviewed-by:
Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211112224636.395101-8-f.fainelli@gmail.comSigned-off-by:
Wim Van Sebroeck <wim@linux-watchdog.org>
-
Florian Fainelli authored
In order to utilize the bcm7038_wdt.c driver which needs to know the clock name to obtain, pass it via platform data using the bcm7038_wdt_platform_data structure. Signed-off-by:
Thomas Bogendoerfer <tsbogend@alpha.franken.de> Reviewed-by:
-
Florian Fainelli authored
In order to phase out bcm63xx_wdt and use bcm7038_wdt instead, introduce a platform_device_id table that allows both names to be matched. Signed-off-by:
-
Florian Fainelli authored
CONFIG_BCM63XX denotes the legacy MIPS-based DSL SoCs which utilize the same piece of hardware as a watchdog, make it possible to select that driver for those platforms. Signed-off-by:
-
Florian Fainelli authored
The BCM7038 watchdog driver needs to be able to obtain a specific clock name on BCM63xx platforms which is the "periph" clock ticking at 50MHz. make it possible to specify the clock name to obtain via platform data. Signed-off-by:
-
Florian Fainelli authored
The BCM7038 watchdog binding is updated to include a "brcm,bcm6345-wdt" compatible string which is the first instance of a DSL (BCM63xx) SoC seeing the integration of such a watchdog timer block. Reviewed-by:
Rob Herring <robh@kernel.org> Signed-off-by:
-
Rafał Miłecki authored
This helps validating DTS files. Signed-off-by:
Rafał Miłecki <rafal@milecki.pl> Acked-by:
-
Artem Lapkin authored
Remove watchdog_stop_on_reboot() The Meson platform still has some hardware drivers problems for some configurations which can freeze devices on shutdown/reboot. Remove watchdog_stop_on_reboot() to catch this situation and ensure that the reboot happens anyway. Users who still want to stop the watchdog on reboot can still do so using the watchdog.stop_on_reboot=1 module parameter. https://lore.kernel.org/linux-watchdog/20210729072308.1908904-1-art@khadas.com/T/#tSigned-off-by:
Artem Lapkin <art@khadas.com> Reviewed-by:
-
Luca Ceresoli authored
Some entries indent their help text with 1 tab + 1 space or 1 tab only instead of 1 tab + 2 spaces. Add the missing spaces. Signed-off-by:
Luca Ceresoli <luca@lucaceresoli.net> Acked-by:
Randy Dunlap <rdunlap@infradead.org> Reviewed-by:
-
Jacky Bai authored
The wdog on i.MX8ULP is derived from i.MX7ULP, it uses two compatible strings, so update the compatible string for i.MX8ULP. Reviewed-by:
Dong Aisheng <aisheng.dong@nxp.com> Acked-by:
Jacky Bai <ping.bai@nxp.com> Signed-off-by:
Peng Fan <peng.fan@nxp.com> Reviewed-by:
-
Sam Protsenko authored
'err' label in probe function is not really need, it just returns. Remove it and replace all 'goto' statements with actual returns in place. No functional change here, just a cleanup patch. Signed-off-by:
Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by:
Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by:
-
Sam Protsenko authored
Right now all devices supported in the driver have the single clock: it acts simultaneously as a bus clock (providing register interface clocking) and source clock (driving watchdog counter). Some newer Exynos chips, like Exynos850, have two separate clocks for that. In that case two clocks will be passed to the driver from the resource provider, e.g. Device Tree. Provide necessary infrastructure to support that case: - use source clock's rate for all timer related calculations - use bus clock to gate/ungate the register interface All devices that use the single clock are kept intact: if only one clock is passed from Device Tree, it will be used for both purposes as before. Signed-off-by:
-
Sam Protsenko authored
Now that PMU enablement code was extended for new Exynos SoCs, it doesn't look very cohesive and consistent anymore. Do a bit of renaming, grouping and style changes, to make it look good again. While at it, add quirks documentation as well. No functional change, just a refactoring commit. Signed-off-by:
-
Sam Protsenko authored
On new Exynos chips (e.g. Exynos850) new CLUSTERx_NONCPU_OUT register is introduced, where CNT_EN_WDT bit must be enabled to make watchdog counter running. Add corresponding quirk and proper infrastructure to handle that register if the quirk is set. This commit doesn't bring any functional change to existing devices, but merely provides an infrastructure for upcoming chips support. Signed-off-by:
-
Sam Protsenko authored
On new Exynos chips (like Exynos850) the MASK_WDT_RESET_REQUEST register is replaced with CLUSTERx_NONCPU_INT_EN, and its mask bit value meaning was reversed: for new register the bit value "1" means "Interrupt enabled", while for MASK_WDT_RESET_REQUEST register "1" means "Mask the interrupt" (i.e. "Interrupt disabled"). Introduce "mask_reset_inv" boolean field in driver data structure; when that field is "true", mask register handling function will invert the value before setting it to the register. This commit doesn't bring any functional change to existing devices, but merely provides an infrastructure for upcoming chips support. Signed-off-by:
-
Sam Protsenko authored
The s3c2410wdt_mask_and_disable_reset() function content is bound to be changed further. Prepare it for upcoming changes by splitting into separate "mask reset" and "disable reset" functions. But keep s3c2410wdt_mask_and_disable_reset() function present as a facade. This commit doesn't bring any functional change to existing devices, but merely provides an infrastructure for upcoming chips support. Signed-off-by:
-
Sam Protsenko authored
On new Exynos chips (e.g. Exynos850 and Exynos9) the AUTOMATIC_WDT_RESET_DISABLE register was removed, and its value can be thought of as "always 0x0". Add correspondig quirk bit, so that the driver can omit accessing it if it's not present. This commit doesn't bring any functional change to existing devices, but merely provides an infrastructure for upcoming chips support. Signed-off-by:
-
Sam Protsenko authored
When "tmr_atboot" module param is set, the watchdog is started in driver's probe. In that case, also set WDOG_HW_RUNNING bit to let watchdog core driver know it's running. This way watchdog core can kick the watchdog for us (if CONFIG_WATCHDOG_HANDLE_BOOT_ENABLED option is enabled), until user space takes control. WDOG_HW_RUNNING bit must be set before registering the watchdog. So the "tmr_atboot" handling code is moved before watchdog registration, to avoid performing the same check twice. This is also logical because WDOG_HW_RUNNING bit makes WDT core expect actually running watchdog. Signed-off-by:
-
Sam Protsenko authored
Driver can't work properly if there no valid timeout was found in s3c2410wdt_set_heartbeat(). Ideally, that function should be reworked in a way that it's always able to find some valid timeout. As a temporary solution let's for now just fail the driver probe in case the valid timeout can't be found in s3c2410wdt_set_heartbeat() function. Signed-off-by:
-
Sam Protsenko authored
Exynos850 SoC has two CPU clusters: - cluster 0: contains CPUs #0, #1, #2, #3 - cluster 1: contains CPUs #4, #5, #6, #7 Each cluster has its own dedicated watchdog timer. Those WDT instances are controlled using different bits in PMU registers, new "samsung,index" property is added to tell the driver which bits to use for defined watchdog node. Also on Exynos850 the peripheral clock and the source clock are two different clocks. Provide a way to specify two clocks in watchdog device tree node. Signed-off-by:
-
Sam Protsenko authored
Exynos7 watchdog driver is clearly indicating that its dts node must define syscon phandle property. That was probably forgotten, so add it. Signed-off-by:
-
AaeonIot authored
This adds watchdog support the Fintek F81966 Super I/O chip. Testing was done on the Aaeon SSE-OPTI Signed-off-by:
AaeonIot <sophiehu@aaeon.com.tw> Signed-off-by:
Chia-Lin Kao (AceLan) <acelan.kao@canonical.com> Reviewed-by:
-
Kevin Hilman authored
Enable CONFIG_MEDIATEK_WATCHDOG when ARCH_MEDIATEK is enabled. On some platforms (e.g. mt8183-pumpkin), watchdog is enabled by bootloader, so kernel driver needs to be enabled to avoid watchdog firing and causing reboot part way through kernel boot. Signed-off-by:
Kevin Hilman <khilman@baylibre.com> Reviewed-by:
-
- 26 Dec, 2021 4 commits
-
-
Linus Torvalds authored
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull x86 fixes from Borislav Petkov: - Prevent potential undefined behavior due to shifting pkey constants into the sign bit - Move the EFI memory reservation code *after* the efi= cmdline parsing has happened - Revert two commits which turned out to be the wrong direction to chase when accommodating early memblock reservations consolidation and command line parameters parsing * tag 'x86_urgent_for_v5.16_rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86/pkey: Fix undefined behaviour with PKRU_WD_BIT x86/boot: Move EFI range reservation after cmdline parsing Revert "x86/boot: Pull up cmdline preparation and early param parsing" Revert "x86/boot: Mark prepare_command_line() __init"
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull objtool fixes from Borislav Petkov: - Prevent clang from reordering the reachable annotation in an inline asm statement without inputs - Fix objtool builds on non-glibc systems due to undefined __always_inline * tag 'objtool_urgent_for_v5.16_rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: compiler.h: Fix annotation macro misplacement with Clang uapi: Fix undefined __always_inline on non-glibc systems
-
git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrlLinus Torvalds authored
Pull pin control fixes from Linus Walleij: "Some hopefully final pin control fixes for the v5.16 kernel: - Fix an out-of-bounds bug in the Mediatek driver - Fix an init order bug in the Broadcom BCM2835 driver - Fix a GPIO offset bug in the STM32 driver" * tag 'pinctrl-v5.16-3' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl: pinctrl: stm32: consider the GPIO offset to expose all the GPIO lines pinctrl: bcm2835: Change init order for gpio hogs pinctrl: mediatek: fix global-out-of-bounds issue
-
- 25 Dec, 2021 12 commits
-
-
Linus Torvalds authored
Merge tag 'hwmon-for-v5.16-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging Pull hwmon fixes from Guenter Roeck: "A couple of lm90 driver fixes. None of them are critical, but they should nevertheless be fixed" * tag 'hwmon-for-v5.16-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging: hwmon: (lm90) Do not report 'busy' status bit as alarm hwmom: (lm90) Fix citical alarm status for MAX6680/MAX6681 hwmon: (lm90) Drop critical attribute support for MAX6654 hwmon: (lm90) Prevent integer overflow/underflow in hysteresis calculations hwmon: (lm90) Fix usage of CONFIG2 register in detect function
-
git://git.kernel.org/pub/scm/linux/kernel/git/dtor/inputLinus Torvalds authored
Pull input fixes from Dmitry Torokhov: "A few small updates to drivers. Of note we are now deferring probes of i8042 on some Asus devices as the controller is not ready to respond to queries first time around when the driver is compiled into the kernel" * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input: Input: elants_i2c - do not check Remark ID on eKTH3900/eKTH5312 Input: atmel_mxt_ts - fix double free in mxt_read_info_block Input: goodix - fix memory leak in goodix_firmware_upload Input: goodix - add id->model mapping for the "9111" model Input: goodix - try not to touch the reset-pin on x86/ACPI devices Input: i8042 - enable deferred probe quirk for ASUS UM325UA Input: elantech - fix stack out of bound access in elantech_change_report_id() Input: iqs626a - prohibit inlining of channel parsing functions Input: i8042 - add deferred probe support
-
Linus Torvalds authored
Merge misc fixes from Andrew Morton: "9 patches. Subsystems affected by this patch series: mm (kfence, mempolicy, memory-failure, pagemap, pagealloc, damon, and memory-failure), core-kernel, and MAINTAINERS" * emailed patches from Andrew Morton <akpm@linux-foundation.org>: mm/hwpoison: clear MF_COUNT_INCREASED before retrying get_any_page() mm/damon/dbgfs: protect targets destructions with kdamond_lock mm/page_alloc: fix __alloc_size attribute for alloc_pages_exact_nid mm: delete unsafe BUG from page_cache_add_speculative() mm, hwpoison: fix condition in free hugetlb page path MAINTAINERS: mark more list instances as moderated kernel/crash_core: suppress unknown crashkernel parameter warning mm: mempolicy: fix THP allocations escaping mempolicy restrictions kfence: fix memory leak when cat kfence objects
-
Liu Shixin authored
Hulk Robot reported a panic in put_page_testzero() when testing madvise() with MADV_SOFT_OFFLINE. The BUG() is triggered when retrying get_any_page(). This is because we keep MF_COUNT_INCREASED flag in second try but the refcnt is not increased. page dumped because: VM_BUG_ON_PAGE(page_ref_count(page) == 0) ------------[ cut here ]------------ kernel BUG at include/linux/mm.h:737! invalid opcode: 0000 [#1] PREEMPT SMP CPU: 5 PID: 2135 Comm: sshd Tainted: G B 5.16.0-rc6-dirty #373 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1ubuntu1.1 04/01/2014 RIP: release_pages+0x53f/0x840 Call Trace: free_pages_and_swap_cache+0x64/0x80 tlb_flush_mmu+0x6f/0x220 unmap_page_range+0xe6c/0x12c0 unmap_single_vma+0x90/0x170 unmap_vmas+0xc4/0x180 exit_mmap+0xde/0x3a0 mmput+0xa3/0x250 do_exit+0x564/0x1470 do_group_exit+0x3b/0x100 __do_sys_exit_group+0x13/0x20 __x64_sys_exit_group+0x16/0x20 do_syscall_64+0x34/0x80 entry_SYSCALL_64_after_hwframe+0x44/0xae Modules linked in: ---[ end trace e99579b570fe0649 ]--- RIP: 0010:release_pages+0x53f/0x840 Link: https://lkml.kernel.org/r/20211221074908.3910286-1-liushixin2@huawei.com Fixes: b94e0282 ("mm,hwpoison: try to narrow window race for free pages") Signed-off-by:Liu Shixin <liushixin2@huawei.com> Reported-by:
Hulk Robot <hulkci@huawei.com> Reviewed-by:
Oscar Salvador <osalvador@suse.de> Acked-by:
Naoya Horiguchi <naoya.horiguchi@nec.com> Cc: <stable@vger.kernel.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
SeongJae Park authored
DAMON debugfs interface iterates current monitoring targets in 'dbgfs_target_ids_read()' while holding the corresponding 'kdamond_lock'. However, it also destructs the monitoring targets in 'dbgfs_before_terminate()' without holding the lock. This can result in a use_after_free bug. This commit avoids the race by protecting the destruction with the corresponding 'kdamond_lock'. Link: https://lkml.kernel.org/r/20211221094447.2241-1-sj@kernel.orgReported-by:
Sangwoo Bae <sangwoob@amazon.com> Fixes: 4bc05954 ("mm/damon: implement a debugfs-based user space interface") Signed-off-by:
SeongJae Park <sj@kernel.org> Cc: <stable@vger.kernel.org> [5.15.x] Signed-off-by:
-
Thibaut Sautereau authored
The second parameter of alloc_pages_exact_nid is the one indicating the size of memory pointed by the returned pointer. Link: https://lkml.kernel.org/r/YbjEgwhn4bGblp//@coeus Fixes: abd58f38 ("mm/page_alloc: add __alloc_size attributes for better bounds checking") Signed-off-by:
Thibaut Sautereau <thibaut.sautereau@ssi.gouv.fr> Acked-by:
Kees Cook <keescook@chromium.org> Cc: Daniel Micay <danielmicay@gmail.com> Cc: Levente Polyak <levente@leventepolyak.net> Signed-off-by:
-
Hugh Dickins authored
It is not easily reproducible, but on 5.16-rc I have several times hit the VM_BUG_ON_PAGE(PageTail(page), page) in page_cache_add_speculative(): usually from filemap_get_read_batch() for an ext4 read, yesterday from next_uptodate_page() from filemap_map_pages() for a shmem fault. That BUG used to be placed where page_ref_add_unless() had succeeded, but now it is placed before folio_ref_add_unless() is attempted: that is not safe, since it is only the acquired reference which makes the page safe from racing THP collapse or split. We could keep the BUG, checking PageTail only when folio_ref_try_add_rcu() has succeeded; but I don't think it adds much value - just delete it. Link: https://lkml.kernel.org/r/8b98fc6f-3439-8614-c3f3-945c659a1aba@google.com Fixes: 020853b6 ("mm: Add folio_try_get_rcu()") Signed-off-by:
Hugh Dickins <hughd@google.com> Acked-by:
Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Reviewed-by:
Matthew Wilcox (Oracle) <willy@infradead.org> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: William Kucharski <william.kucharski@oracle.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Mike Rapoport <rppt@linux.ibm.com> Signed-off-by:
-
Naoya Horiguchi authored
When a memory error hits a tail page of a free hugepage, __page_handle_poison() is expected to be called to isolate the error in 4kB unit, but it's not called due to the outdated if-condition in memory_failure_hugetlb(). This loses the chance to isolate the error in the finer unit, so it's not optimal. Drop the condition. This "(p != head && TestSetPageHWPoison(head)" condition is based on the old semantics of PageHWPoison on hugepage (where PG_hwpoison flag was set on the subpage), so it's not necessray any more. By getting to set PG_hwpoison on head page for hugepages, concurrent error events on different subpages in a single hugepage can be prevented by TestSetPageHWPoison(head) at the beginning of memory_failure_hugetlb(). So dropping the condition should not reopen the race window originally mentioned in commit b985194c ("hwpoison, hugetlb: lock_page/unlock_page does not match for handling a free hugepage") [naoya.horiguchi@linux.dev: fix "HardwareCorrupted" counter] Link: https://lkml.kernel.org/r/20211220084851.GA1460264@u2004 Link: https://lkml.kernel.org/r/20211210110208.879740-1-naoya.horiguchi@linux.devSigned-off-by:
Fei Luo <luofei@unicloud.com> Reviewed-by:
Mike Kravetz <mike.kravetz@oracle.com> Cc: <stable@vger.kernel.org> [5.14+] Signed-off-by:
-
Randy Dunlap authored
Some lists that are moderated are not marked as moderated consistently, so mark them all as moderated. Link: https://lkml.kernel.org/r/20211209001330.18558-1-rdunlap@infradead.orgSigned-off-by:
-
Philipp Rudo authored
When booting with crashkernel= on the kernel command line a warning similar to Kernel command line: ro console=ttyS0 crashkernel=256M Unknown kernel command line parameters "crashkernel=256M", will be passed to user space. is printed. This comes from crashkernel= being parsed independent from the kernel parameter handling mechanism. So the code in init/main.c doesn't know that crashkernel= is a valid kernel parameter and prints this incorrect warning. Suppress the warning by adding a dummy early_param handler for crashkernel=. Link: https://lkml.kernel.org/r/20211208133443.6867-1-prudo@redhat.com Fixes: 86d1919a ("init: print out unknown kernel parameters") Signed-off-by:Philipp Rudo <prudo@redhat.com> Acked-by:
Baoquan He <bhe@redhat.com> Cc: Andrew Halaney <ahalaney@redhat.com> Signed-off-by:
-
Andrey Ryabinin authored
alloc_pages_vma() may try to allocate THP page on the local NUMA node first: page = __alloc_pages_node(hpage_node, gfp | __GFP_THISNODE | __GFP_NORETRY, order); And if the allocation fails it retries allowing remote memory: if (!page && (gfp & __GFP_DIRECT_RECLAIM)) page = __alloc_pages_node(hpage_node, gfp, order); However, this retry allocation completely ignores memory policy nodemask allowing allocation to escape restrictions. The first appearance of this bug seems to be the commit ac5b2c18 ("mm: thp: relax __GFP_THISNODE for MADV_HUGEPAGE mappings"). The bug disappeared later in the commit 89c83fb5 ("mm, thp: consolidate THP gfp handling into alloc_hugepage_direct_gfpmask") and reappeared again in slightly different form in the commit 76e654cc ("mm, page_alloc: allow hugepage fallback to remote nodes when madvised") Fix this by passing correct nodemask to the __alloc_pages() call. The demonstration/reproducer of the problem: $ mount -oremount,size=4G,huge=always /dev/shm/ $ echo always > /sys/kernel/mm/transparent_hugepage/defrag $ cat mbind_thp.c #include <unistd.h> #include <sys/mman.h> #include <sys/stat.h> #include <fcntl.h> #include <assert.h> #include <stdlib.h> #include <stdio.h> #include <numaif.h> #define SIZE 2ULL << 30 int main(int argc, char **argv) { int fd; unsigned long long i; char *addr; pid_t pid; char buf[100]; unsigned long nodemask = 1; fd = open("/dev/shm/test", O_RDWR|O_CREAT); assert(fd > 0); assert(ftruncate(fd, SIZE) == 0); addr = mmap(NULL, SIZE, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0); assert(mbind(addr, SIZE, MPOL_BIND, &nodemask, 2, MPOL_MF_STRICT|MPOL_MF_MOVE)==0); for (i = 0; i < SIZE; i+=4096) { addr[i] = 1; } pid = getpid(); snprintf(buf, sizeof(buf), "grep shm /proc/%d/numa_maps", pid); system(buf); sleep(10000); return 0; } $ gcc mbind_thp.c -o mbind_thp -lnuma $ numactl -H available: 2 nodes (0-1) node 0 cpus: 0 2 node 0 size: 1918 MB node 0 free: 1595 MB node 1 cpus: 1 3 node 1 size: 2014 MB node 1 free: 1731 MB node distances: node 0 1 0: 10 20 1: 20 10 $ rm -f /dev/shm/test; taskset -c 0 ./mbind_thp 7fd970a00000 bind:0 file=/dev/shm/test dirty=524288 active=0 N0=396800 N1=127488 kernelpagesize_kB=4 Link: https://lkml.kernel.org/r/20211208165343.22349-1-arbn@yandex-team.com Fixes: ac5b2c18 ("mm: thp: relax __GFP_THISNODE for MADV_HUGEPAGE mappings") Signed-off-by:Andrey Ryabinin <arbn@yandex-team.com> Acked-by:
Michal Hocko <mhocko@suse.com> Acked-by:
Mel Gorman <mgorman@techsingularity.net> Acked-by:
David Rientjes <rientjes@google.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Baokun Li authored
Hulk robot reported a kmemleak problem: unreferenced object 0xffff93d1d8cc02e8 (size 248): comm "cat", pid 23327, jiffies 4624670141 (age 495992.217s) hex dump (first 32 bytes): 00 40 85 19 d4 93 ff ff 00 10 00 00 00 00 00 00 .@.............. 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ backtrace: seq_open+0x2a/0x80 full_proxy_open+0x167/0x1e0 do_dentry_open+0x1e1/0x3a0 path_openat+0x961/0xa20 do_filp_open+0xae/0x120 do_sys_openat2+0x216/0x2f0 do_sys_open+0x57/0x80 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xa9 unreferenced object 0xffff93d419854000 (size 4096): comm "cat", pid 23327, jiffies 4624670141 (age 495992.217s) hex dump (first 32 bytes): 6b 66 65 6e 63 65 2d 23 32 35 30 3a 20 30 78 30 kfence-#250: 0x0 30 30 30 30 30 30 30 37 35 34 62 64 61 31 32 2d 0000000754bda12- backtrace: seq_read_iter+0x313/0x440 seq_read+0x14b/0x1a0 full_proxy_read+0x56/0x80 vfs_read+0xa5/0x1b0 ksys_read+0xa0/0xf0 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xa9 I find that we can easily reproduce this problem with the following commands: cat /sys/kernel/debug/kfence/objects echo scan > /sys/kernel/debug/kmemleak cat /sys/kernel/debug/kmemleak The leaked memory is allocated in the stack below: do_syscall_64 do_sys_open do_dentry_open full_proxy_open seq_open ---> alloc seq_file vfs_read full_proxy_read seq_read seq_read_iter traverse ---> alloc seq_buf And it should have been released in the following process: do_syscall_64 syscall_exit_to_user_mode exit_to_user_mode_prepare task_work_run ____fput __fput full_proxy_release ---> free here However, the release function corresponding to file_operations is not implemented in kfence. As a result, a memory leak occurs. Therefore, the solution to this problem is to implement the corresponding release function. Link: https://lkml.kernel.org/r/20211206133628.2822545-1-libaokun1@huawei.com Fixes: 0ce20dd8 ("mm: add Kernel Electric-Fence infrastructure") Signed-off-by:Baokun Li <libaokun1@huawei.com> Reported-by:
Marco Elver <elver@google.com> Reviewed-by:
Kefeng Wang <wangkefeng.wang@huawei.com> Cc: Alexander Potapenko <glider@google.com> Cc: Dmitry Vyukov <dvyukov@google.com> Cc: Yu Kuai <yukuai3@huawei.com> Signed-off-by:
-
- 24 Dec, 2021 1 commit
-
-
git://git.armlinux.org.uk/~rmk/linux-armLinus Torvalds authored
Pull ARM fixes from Russell King: - fix nommu after getting rid of mini-stack for ARMv7 - fix Thumb2 bug in iWMMXt exception handling * tag 'for-linus' of git://git.armlinux.org.uk/~rmk/linux-arm: ARM: 9169/1: entry: fix Thumb2 bug in iWMMXt exception handling ARM: 9160/1: NOMMU: Reload __secondary_data after PROCINFO_INITFUNC
-
