pipe_max_size is defined as an unsigned int: unsigned int pipe_max_size = 1048576; but its procfs/sysctl representation is an integer: static struct ctl_table fs_table[] = { ... { .procname = "pipe-max-size", .data = &pipe_max_size, .maxlen = sizeof(int), .mode = 0644, .proc_handler = &pipe_proc_fn, .extra1 = &pipe_min_size, }, ... that is signed: int pipe_proc_fn(struct ctl_table *table, int write, void __user *buf, size_t *lenp, loff_t *ppos) { ... ret = proc_dointvec_minmax(table, write, buf, lenp, ppos) This leads to signed results via procfs for large values of pipe_max_size: % echo 2147483647 >/proc/sys/fs/pipe-max-size % cat /proc/sys/fs/pipe-max-size -2147483648 Use unsigned operations on this variable to avoid such negative values. Reported-by: Mikulas Patocka <mpatocka@xxxxxxxxxx> Signed-off-by: Joe Lawrence <joe.lawrence@xxxxxxxxxx> --- fs/pipe.c | 2 +- kernel/sysctl.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/pipe.c b/fs/pipe.c index 33bb11b0d78e..3b10d39cc5d1 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -1147,7 +1147,7 @@ int pipe_proc_fn(struct ctl_table *table, int write, void __user *buf, mutex_lock(&pipe_max_mutex); orig_pipe_max_size = pipe_max_size; - ret = proc_dointvec_minmax(table, write, buf, lenp, ppos); + ret = proc_douintvec_minmax(table, write, buf, lenp, ppos); if (ret < 0 || !write) { mutex_unlock(&pipe_max_mutex); return ret; diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 6648fbbb8157..c976719bf37a 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -1825,7 +1825,7 @@ static int sysrq_sysctl_handler(struct ctl_table *table, int write, { .procname = "pipe-max-size", .data = &pipe_max_size, - .maxlen = sizeof(int), + .maxlen = sizeof(pipe_max_size), .mode = 0644, .proc_handler = &pipe_proc_fn, .extra1 = &pipe_min_size, -- 1.8.3.1