Hi,
On 19-02-18 18:36, Marcel Holtmann wrote:
Hi Hans,
Commit 61f5acea8737 ("Bluetooth: btusb: Restore QCA Rome suspend/resume fix
with a "rewritten" version") applied the USB_QUIRK_RESET_RESUME to all QCA
btusb devices. But it turns out that the resume problems are not caused by
the QCA Rome chipset, on most platforms it resumes fine. The resume
problems are actually a platform problem (likely the platform cutting all
power when suspended).
The USB_QUIRK_RESET_RESUME quirk also disable runtime suspend, so by
matching on usb-ids, we're causing all boards with these chips to use extra
power, to fix resume problems which only happen on some boards.
This commit fixes this by applying the quirk based on DMI matching instead
of on usb-ids, so that we match the platform and not the chipset.
just for the record, can we include the /sys/kernel/debug/usb/devices for that device here.
Will do.
Fixes: 61f5acea8737 ("Bluetooth: btusb: Restore QCA Rome suspend/resume..")
Cc: stable@xxxxxxxxxxxxxxx
Cc: Brian Norris <briannorris@xxxxxxxxxxxx>
Cc: Kai-Heng Feng <kai.heng.feng@xxxxxxxxxxxxx>
Signed-off-by: Hans de Goede <hdegoede@xxxxxxxxxx>
---
drivers/bluetooth/btusb.c | 26 ++++++++++++++++++++------
1 file changed, 20 insertions(+), 6 deletions(-)
diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
index 2a55380ad730..a6023667f3b4 100644
--- a/drivers/bluetooth/btusb.c
+++ b/drivers/bluetooth/btusb.c
@@ -21,6 +21,7 @@
*
*/
+#include <linux/dmi.h>
#include <linux/module.h>
#include <linux/usb.h>
#include <linux/usb/quirks.h>
@@ -379,6 +380,22 @@ static const struct usb_device_id blacklist_table[] = {
{ } /* Terminating entry */
};
+/*
+ * The btusb build into some devices needs to be reset on resume, this is a
Actually “btusb” is a driver name. So “Bluetooth USB modules”
+ * problem with the platform (likely shutting off all power) not with the
+ * btusb chip itself. So we use a DMI list to match known broken platforms.
Here s/btusb/module/
+ */
+static const struct dmi_system_id btusb_plat_needs_reset_resume_list[] = {
I prefer to use _table instead of _list. Also drop the _plat_ part since that seems obvious.
+ {
+ /* Lenovo yoga 920 */
Use “Yoga" please.
I will fix all of the above.
And I would include “(QCA Rome device VID:PID)” so that we have a record of some sorts.
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+ DMI_MATCH(DMI_PRODUCT_VERSION, "Lenovo YOGA 920”),
No DMI_EXACT_MATCH?
The DMI data actually has:
"Lenovo YOGA 920-13IKB"
I'm not using DMI_EXACT_MATCH on purpose here, I think Lenovo
might change the "-13IKB" part and I don't expect them to fix
this bug on newer revisions.
that is fine. What about the “LENOVO” vendor match? Should that be an exact match?
It could be an exact match, in general I tend to not use DMI_EXACT_MATCH
unless necessary though, as vendors sometimes add whitespace padding after
the contents. Esp. Asus is known to do this. But if you want me to change it
over to DMI_EXACT_MATCH I can do that for the non RFC version which I plan
to send tomorrow (I just got test feedback that this version works for the
reporter).
Regards,
Hans