On Wed, Dec 1, 2021 at 4:05 PM Benjamin Tissoires <benjamin.tissoires@xxxxxxxxxx> wrote:
> > FWIW, I am also working on a test that generates all registered USB
> > vid/pids from the currently running kernel and will try to inject them
> > in the testsuite. This should validate this patch and prevent future
> > mishaps.
>
> That would be great to see. Let me know if you want these broken up
> smaller, and I will go fix the above issues.
https://gitlab.freedesktop.org/bentiss/hid-tools/-/commit/2b04c6cf46c7405c8a916f324988660cb54dadea
This seems sufficient to make an unpatched kernel scream :)
I have now opened https://gitlab.freedesktop.org/libevdev/hid-tools/-/merge_requests/122
with a quicker solution.
This leads apparently to 2 extra crashes:
- one in hid-holtek-kbd -> hid_is_usb() is called *after* hid_parse(), so
report_fixup is actually called for uhid devices.
- another in bigbenff which is fixed by the following patch:
---
diff --git a/drivers/hid/hid-bigbenff.c b/drivers/hid/hid-bigbenff.c
index db6da21ade06..74ad8bf98bfd 100644
--- a/drivers/hid/hid-bigbenff.c
+++ b/drivers/hid/hid-bigbenff.c
@@ -191,7 +191,7 @@ static void bigben_worker(struct work_struct *work)
struct bigben_device, worker);
struct hid_field *report_field = bigben->report->field[0];
- if (bigben->removed)
+ if (bigben->removed || !report_field)
return;
if (bigben->work_led) {
---
Cheers,
Benjamin
