The ioctl helper function nilfs_ioctl_wrap_copy, which exchanges a metadata array to/from user space, may copy uninitialized buffer regions to user space memory for read-only ioctl commands NILFS_IOCTL_GET_SUINFO and NILFS_IOCTL_GET_CPINFO. This can occur when the element size of the user space metadata given by the v_size member of the argument nilfs_argv structure is larger than the size of the metadata element (nilfs_suinfo structure or nilfs_cpinfo structure) on the file system side. KMSAN-enabled kernels detect this issue as follows: BUG: KMSAN: kernel-infoleak in instrument_copy_to_user include/linux/instrumented.h:121 [inline] BUG: KMSAN: kernel-infoleak in _copy_to_user+0xc0/0x100 lib/usercopy.c:33 instrument_copy_to_user include/linux/instrumented.h:121 [inline] _copy_to_user+0xc0/0x100 lib/usercopy.c:33 copy_to_user include/linux/uaccess.h:169 [inline] nilfs_ioctl_wrap_copy+0x6fa/0xc10 fs/nilfs2/ioctl.c:99 nilfs_ioctl_get_info fs/nilfs2/ioctl.c:1173 [inline] nilfs_ioctl+0x2402/0x4450 fs/nilfs2/ioctl.c:1290 nilfs_compat_ioctl+0x1b8/0x200 fs/nilfs2/ioctl.c:1343 __do_compat_sys_ioctl fs/ioctl.c:968 [inline] __se_compat_sys_ioctl+0x7dd/0x1000 fs/ioctl.c:910 __ia32_compat_sys_ioctl+0x93/0xd0 fs/ioctl.c:910 do_syscall_32_irqs_on arch/x86/entry/common.c:112 [inline] __do_fast_syscall_32+0xa2/0x100 arch/x86/entry/common.c:178 do_fast_syscall_32+0x37/0x80 arch/x86/entry/common.c:203 do_SYSENTER_32+0x1f/0x30 arch/x86/entry/common.c:246 entry_SYSENTER_compat_after_hwframe+0x70/0x82 Uninit was created at: __alloc_pages+0x9f6/0xe90 mm/page_alloc.c:5572 alloc_pages+0xab0/0xd80 mm/mempolicy.c:2287 __get_free_pages+0x34/0xc0 mm/page_alloc.c:5599 nilfs_ioctl_wrap_copy+0x223/0xc10 fs/nilfs2/ioctl.c:74 nilfs_ioctl_get_info fs/nilfs2/ioctl.c:1173 [inline] nilfs_ioctl+0x2402/0x4450 fs/nilfs2/ioctl.c:1290 nilfs_compat_ioctl+0x1b8/0x200 fs/nilfs2/ioctl.c:1343 __do_compat_sys_ioctl fs/ioctl.c:968 [inline] __se_compat_sys_ioctl+0x7dd/0x1000 fs/ioctl.c:910 __ia32_compat_sys_ioctl+0x93/0xd0 fs/ioctl.c:910 do_syscall_32_irqs_on arch/x86/entry/common.c:112 [inline] __do_fast_syscall_32+0xa2/0x100 arch/x86/entry/common.c:178 do_fast_syscall_32+0x37/0x80 arch/x86/entry/common.c:203 do_SYSENTER_32+0x1f/0x30 arch/x86/entry/common.c:246 entry_SYSENTER_compat_after_hwframe+0x70/0x82 Bytes 16-127 of 3968 are uninitialized ... This eliminates the leak issue by initializing the page allocated as buffer using get_zeroed_page(). Signed-off-by: Ryusuke Konishi <konishi.ryusuke@xxxxxxxxx> Reported-by: syzbot+132fdd2f1e1805fdc591@xxxxxxxxxxxxxxxxxxxxxxxxx Link: https://lkml.kernel.org/r/000000000000a5bd2d05f63f04ae@xxxxxxxxxx Tested-by: Ryusuke Konishi <konishi.ryusuke@xxxxxxxxx> Cc: stable@xxxxxxxxxxxxxxx --- fs/nilfs2/ioctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c index 5ccc638ae92f..1dfbc0c34513 100644 --- a/fs/nilfs2/ioctl.c +++ b/fs/nilfs2/ioctl.c @@ -71,7 +71,7 @@ static int nilfs_ioctl_wrap_copy(struct the_nilfs *nilfs, if (argv->v_index > ~(__u64)0 - argv->v_nmembs) return -EINVAL; - buf = (void *)__get_free_pages(GFP_NOFS, 0); + buf = (void *)get_zeroed_page(GFP_NOFS); if (unlikely(!buf)) return -ENOMEM; maxmembs = PAGE_SIZE / argv->v_size; -- 2.34.1
