On Tue, Sep 23, 2025 at 11:43:12AM +0800, Herbert Xu wrote:
On Mon, Sep 15, 2025 at 05:02:15PM +0100, Jonathan McDowell wrote:

diff --git a/drivers/char/hw_random/core.c b/drivers/char/hw_random/core.c
index 018316f54621..1682a9f1b28c 100644
--- a/drivers/char/hw_random/core.c
+++ b/drivers/char/hw_random/core.c
@@ -341,6 +341,10 @@ static ssize_t rng_current_store(struct device *dev,
        if (sysfs_streq(buf, "")) {
                err = enable_best_rng();
+       } else if (sysfs_streq(buf, "none")) {
+               if (current_rng)
+                       cur_rng_set_by_user = 1;

Shouldn't this be set unconditionally?

Yes, I guess so. I'd been guided by enable_best_rng() which only clears it if it changes the current RNG, but re-reading I agree that's probably the wrong thing here.

J.

--
101 things you can't have too much of : 16 - Time.

Reply via email to