I'm seeing the following with kernel-core-5.10.16-200.fc33.x86_64:
$ sudo SYSTEMD_LOG_LEVEL=debug build/systemd-rfkill
Reading struct rfkill_event: got 8 bytes.
A new rfkill device has been added with index 0 and type bluetooth.
Found cgroup2 on /sys/fs/cgroup/, full unified hierarchy
Found container virtualization none.
rfkill0: Operating on rfkill device 'tpacpi_bluetooth_sw'.
Writing struct rfkill_event successful (8 of 9 bytes).
Loaded state '0' from /var/lib/systemd/rfkill/platform-thinkpad_acpi:bluetooth.
Reading struct rfkill_event: got 8 bytes.
A new rfkill device has been added with index 1 and type wwan.
rfkill1: Operating on rfkill device 'tpacpi_wwan_sw'.
Writing struct rfkill_event successful (8 of 9 bytes).
Loaded state '0' from /var/lib/systemd/rfkill/platform-thinkpad_acpi:wwan.
Reading struct rfkill_event: got 8 bytes.
A new rfkill device has been added with index 2 and type bluetooth.
rfkill2: Operating on rfkill device 'hci0'.
Writing struct rfkill_event successful (8 of 9 bytes).
Loaded state '0' from /var/lib/systemd/rfkill/pci-0000:00:14.0-usb-0:7:1.0:bluetooth.
Reading struct rfkill_event: got 8 bytes.
A new rfkill device has been added with index 3 and type wlan.
rfkill3: Operating on rfkill device 'phy0'.
Writing struct rfkill_event successful (8 of 9 bytes).
Loaded state '0' from /var/lib/systemd/rfkill/pci-0000:04:00.0:wlan.
All events read and idle, exiting.
We were expecting a read of exactly RFKILL_EVENT_SIZE_V1==8 bytes. But the
structure has 9 after [1].
[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=
14486c82612a177cb910980c70ba900827ca0894
For some reason the kernel does not accept the full structure size, but cuts
the write short after 8 bytes:
static ssize_t rfkill_fop_write(struct file *file, const char __user *buf,
size_t count, loff_t *pos)
{
struct rfkill_event ev;
/* we don't need the 'hard' variable but accept it */
if (count < RFKILL_EVENT_SIZE_V1 - 1)
return -EINVAL;
/*
* Copy as much data as we can accept into our 'ev' buffer,
* but tell userspace how much we've copied so it can determine
* our API version even in a write() call, if it cares.
*/
count = min(count, sizeof(ev));
if (copy_from_user(&ev, buf, count))
return -EFAULT;
... so it should accept the full size. I'm not sure what is going on here.
But we don't care about the extra fields, so let's accept a write as long as
it's at least RFKILL_EVENT_SIZE_V1.
Fixes #18677.
.op = RFKILL_OP_CHANGE,
.soft = b,
};
+ assert_cc(offsetof(struct rfkill_event, op) < RFKILL_EVENT_SIZE_V1);
+ assert_cc(offsetof(struct rfkill_event, soft) < RFKILL_EVENT_SIZE_V1);
ssize_t l = write(c->rfkill_fd, &we, sizeof we);
if (l < 0)
return log_error_errno(errno, "Failed to restore rfkill state for %i: %m", event->idx);
- if (l != sizeof we)
+ if (l < RFKILL_EVENT_SIZE_V1)
return log_error_errno(SYNTHETIC_ERRNO(EIO),
"Couldn't write rfkill event structure, too short (wrote %zd of %zu bytes).",
l, sizeof we);
+ log_debug("Writing struct rfkill_event successful (%zd of %zu bytes).", l, sizeof we);
log_debug("Loaded state '%s' from %s.", one_zero(b), state_file);
return 0;
}
for (;;) {
- struct rfkill_event event;
+ struct rfkill_event event = {};
ssize_t l = read(c.rfkill_fd, &event, sizeof event);
if (l < 0) {
break;
}
- if (l != RFKILL_EVENT_SIZE_V1)
- return log_error_errno(SYNTHETIC_ERRNO(EIO), "Read event structure of unexpected size (%zd, not %d)",
+ if (l < RFKILL_EVENT_SIZE_V1)
+ return log_error_errno(SYNTHETIC_ERRNO(EIO), "Short read of struct rfkill_event: (%zd < %d)",
l, RFKILL_EVENT_SIZE_V1);
+ log_debug("Reading struct rfkill_event: got %zd bytes.", l);
+
+ /* The event structure has more fields. We only care about the first few, so it's OK if we
+ * don't read the full structure. */
+ assert_cc(offsetof(struct rfkill_event, op) < RFKILL_EVENT_SIZE_V1);
+ assert_cc(offsetof(struct rfkill_event, type) < RFKILL_EVENT_SIZE_V1);
const char *type = rfkill_type_to_string(event.type);
if (!type) {