kernel/linux.git/fs/eventfd.c, branch v6.12.80

introduce fd_file(), convert all accessors to it.

2024-08-13T02:00:43+00:00

For any changes of struct fd representation we need to turn existing accesses to fields into calls of wrappers. Accesses to struct fd::flags are very few (3 in linux/file.h, 1 in net/socket.c, 3 in fs/overlayfs/file.c and 3 more in explicit initializers). Those can be dealt with in the commit converting to new layout; accesses to struct fd::file are too many for that. This commit converts (almost) all of f.file to fd_file(f). It's not entirely mechanical ('file' is used as a member name more than just in struct fd) and it does not even attempt to distinguish the uses in pointer context from those in boolean context; the latter will be eventually turned into a separate helper (fd_empty()). NOTE: mass conversion to fd_empty(), tempting as it might be, is a bad idea; better do that piecewise in commit that convert from fdget...() to CLASS(...). [conflicts in fs/fhandle.c, kernel/bpf/syscall.c, mm/memcontrol.c caught by git; fs/stat.c one got caught by git grep] [fs/xattr.c conflict] Reviewed-by: Christian Brauner Signed-off-by: Al Viro

eventfd: strictly check the count parameter of eventfd_write to avoid inputting illegal strings

2024-02-08T09:12:26+00:00

Since eventfd's document has clearly stated: A write(2) call adds the 8-byte integer value supplied in its buffer to the counter. However, in the current implementation, the following code snippet did not cause an error: char str[16] = "hello world"; uint64_t value; ssize_t size; int fd; fd = eventfd(0, 0); size = write(fd, &str, strlen(str)); printf("eventfd: test writing a string, size=%ld\n", size); size = read(fd, &value, sizeof(value)); printf("eventfd: test reading as uint64, size=%ld, valus=0x%lX\n", size, value); close(fd); And its output is: eventfd: test writing a string, size=8 eventfd: test reading as uint64, size=8, valus=0x6F77206F6C6C6568 By checking whether count is equal to sizeof(ucnt), such errors could be detected. It also follows the requirements of the manual. Signed-off-by: Wen Yang Link: https://lore.kernel.org/r/tencent_10AAA44731FFFA493F9F5501521F07DD4D0A@qq.com Cc: Alexander Viro Cc: Jens Axboe Cc: Christian Brauner Cc: Jan Kara Cc: David Woodhouse Cc: Matthew Wilcox Cc: Eric Biggers Cc: linux-fsdevel@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Christian Brauner

eventfd: move 'eventfd-count' printing out of spinlock

2024-01-23T16:56:29+00:00

When printing eventfd->count, interrupts will be disabled and a spinlock will be obtained, competing with eventfd_write(). By moving the "eventfd-count" print out of the spinlock and merging multiple seq_printf() into one, it could improve a bit, just like timerfd_show(). Signed-off-by: Wen Yang Link: https://lore.kernel.org/r/tencent_B0B3D2BD9861FD009E03AB18A81783322709@qq.com Cc: Alexander Viro Cc: Jens Axboe Cc: Christian Brauner Cc: Christoph Hellwig Cc: Dylan Yudaken Cc: David Woodhouse Cc: Matthew Wilcox Cc: Eric Biggers Cc: Cc: Signed-off-by: Christian Brauner

eventfd: add a BUILD_BUG_ON() to ensure consistency between EFD_SEMAPHORE and the uapi

2024-01-22T14:33:36+00:00

introduce a BUILD_BUG_ON to check that the EFD_SEMAPHORE is equal to its definition in the uapi file, just like EFD_CLOEXEC and EFD_NONBLOCK. Signed-off-by: Wen Yang Link: https://lore.kernel.org/r/tencent_0BAA2DEAF9208D49987457E6583F9BE79507@qq.com Cc: Alexander Viro Cc: Christian Brauner Cc: Jens Axboe Cc: Jan Kara Cc: Cc: Signed-off-by: Christian Brauner

eventfd: Remove usage of the deprecated ida_simple_xx() API

2023-12-12T13:24:55+00:00

ida_alloc() and ida_free() should be preferred to the deprecated ida_simple_get() and ida_simple_remove(). This is less verbose. Signed-off-by: Christophe JAILLET Link: https://lore.kernel.org/r/575dcecd51097dd30c5515f9f0ed92076b4ef403.1702229520.git.christophe.jaillet@wanadoo.fr Reviewed-by: Jan Kara Signed-off-by: Christian Brauner

eventfd: make eventfd_signal{_mask}() void

2023-11-28T13:08:46+00:00

No caller care about the return value. Link: https://lore.kernel.org/r/20231122-vfs-eventfd-signal-v2-4-bd549b14ce0c@kernel.org Reviewed-by: Jan Kara Reviewed-by: Jens Axboe Signed-off-by: Christian Brauner

eventfd: simplify eventfd_signal_mask()

2023-11-28T13:08:46+00:00

The eventfd_signal_mask() helper was introduced for io_uring and similar to eventfd_signal() it always passed 1 for @n. So don't bother with that argument at all. Link: https://lore.kernel.org/r/20231122-vfs-eventfd-signal-v2-3-bd549b14ce0c@kernel.org Reviewed-by: Jan Kara Reviewed-by: Jens Axboe Signed-off-by: Christian Brauner

eventfd: simplify eventfd_signal()

2023-11-28T13:08:38+00:00

Ever since the eventfd type was introduced back in 2007 in commit e1ad7468c77d ("signal/timer/event: eventfd core") the eventfd_signal() function only ever passed 1 as a value for @n. There's no point in keeping that additional argument. Link: https://lore.kernel.org/r/20231122-vfs-eventfd-signal-v2-2-bd549b14ce0c@kernel.org Acked-by: Xu Yilun Acked-by: Andrew Donnellan # ocxl Acked-by: Eric Farman # s390 Reviewed-by: Jan Kara Reviewed-by: Jens Axboe Signed-off-by: Christian Brauner

eventfd: prevent underflow for eventfd semaphores

2023-07-11T09:41:34+00:00

For eventfd with flag EFD_SEMAPHORE, when its ctx->count is 0, calling eventfd_ctx_do_read will cause ctx->count to overflow to ULLONG_MAX. An underflow can happen with EFD_SEMAPHORE eventfds in at least the following three subsystems: (1) virt/kvm/eventfd.c (2) drivers/vfio/virqfd.c (3) drivers/virt/acrn/irqfd.c where (2) and (3) are just modeled after (1). An eventfd must be specified for use with the KVM_IRQFD ioctl(). This can also be an EFD_SEMAPHORE eventfd. When the eventfd count is zero or has been decremented to zero an underflow can be triggered when the irqfd is shut down by raising the KVM_IRQFD_FLAG_DEASSIGN flag in the KVM_IRQFD ioctl(): // ctx->count == 0 kvm_vm_ioctl() -> kvm_irqfd() -> kvm_irqfd_deassign() -> irqfd_deactivate() -> irqfd_shutdown() -> eventfd_ctx_remove_wait_queue(&cnt) -> eventfd_ctx_do_read(&cnt) Userspace polling on the eventfd wouldn't notice the underflow because 1 is always returned as the value from eventfd_read() while ctx->count would've underflowed. It's not a huge deal because this should only be happening when the irqfd is shutdown but we should still fix it and avoid the spurious wakeup. Fixes: cb289d6244a3 ("eventfd - allow atomic read and waitqueue remove") Signed-off-by: Wen Yang Cc: Alexander Viro Cc: Jens Axboe Cc: Christian Brauner Cc: Christoph Hellwig Cc: Dylan Yudaken Cc: David Woodhouse Cc: Matthew Wilcox Cc: linux-fsdevel@vger.kernel.org Cc: linux-kernel@vger.kernel.org Message-Id: [brauner: rewrite commit message and add explanation how this underflow can happen] Signed-off-by: Christian Brauner

eventfd: show the EFD_SEMAPHORE flag in fdinfo

2023-06-15T07:22:23+00:00

The EFD_SEMAPHORE flag should be displayed in fdinfo, as different value could affect the behavior of eventfd. Suggested-by: Christian Brauner Signed-off-by: Wen Yang Cc: Alexander Viro Cc: Jens Axboe Cc: Christian Brauner Cc: Christoph Hellwig Cc: Dylan Yudaken Cc: David Woodhouse Cc: Matthew Wilcox Cc: Eric Biggers Cc: linux-fsdevel@vger.kernel.org Cc: linux-kernel@vger.kernel.org Message-Id: Signed-off-by: Christian Brauner