Hi Chris, > I upgraded my macbook from 4.0.2 to 4.1.2, and now I get > messages similar to this bugzilla: > > https://bugzilla.kernel.org/show_bug.cgi?id=100651 > > and this one: > > https://bugzilla.kernel.org/show_bug.cgi?id=101171 > > [21552.711888] Bluetooth: hci0: BCM: Read verbose config info failed (-16) > [21728.019676] Bluetooth: hci0: BCM: Read verbose config info failed (-16) > > A little poking around shows you recently added this check, and from > what I can tell with grep and friends the btbcm_read_verbose_config() > function is completely new? I don't see similar reads being done by > older kernels. > > So, I pushed some code around. It does compile, but printk proves that > only the second hunk is happening on my box, so testing is somewhat > light. > > With this applied, I'm able to use bluetooth again. > > Signed-off-by: Chris Mason <clm@xxxxxx> > > diff --git a/drivers/bluetooth/btbcm.c b/drivers/bluetooth/btbcm.c > index 4bba866..90e7099 100644 > --- a/drivers/bluetooth/btbcm.c > +++ b/drivers/bluetooth/btbcm.c > @@ -294,11 +294,10 @@ int btbcm_setup_patchram(struct hci_dev *hdev) > > /* Read Verbose Config Version Info */ > skb = btbcm_read_verbose_config(hdev); > - if (IS_ERR(skb)) > - return PTR_ERR(skb); > - > - BT_INFO("%s: BCM: chip id %u", hdev->name, skb->data[1]); > - kfree_skb(skb); > + if (!IS_ERR(skb)) { > + BT_INFO("%s: BCM: chip id %u", hdev->name, skb->data[1]); > + kfree_skb(skb); > + } leave this one out of it. Since all Broadcom chips do actually support this as far as I can tell. We really want to fail this one since something went wrong in that case. > > switch ((rev & 0xf000) >> 12) { > case 0: > @@ -378,12 +377,11 @@ int btbcm_setup_apple(struct hci_dev *hdev) > > /* Read Verbose Config Version Info */ > skb = btbcm_read_verbose_config(hdev); > - if (IS_ERR(skb)) > - return PTR_ERR(skb); > - > - BT_INFO("%s: BCM: chip id %u build %4.4u", hdev->name, skb->data[1], > - get_unaligned_le16(skb->data + 5)); > - kfree_skb(skb); > + if (!IS_ERR(skb)) { > + BT_INFO("%s: BCM: chip id %u build %4.4u", hdev->name, skb->data[1], > + get_unaligned_le16(skb->data + 5)); > + kfree_skb(skb); > + } > > set_bit(HCI_QUIRK_STRICT_DUPLICATE_FILTER, &hdev->quirks); This hunk is fine. I would accept a patch for this. I tested 8 Apple Bluetooth modules and they were all fine with this command. It seems there is this one chip in this one MacBook that is not. Actually can you run something like "hciconfig hci0 version" and see if it really is a Broadcom chip in there or if Apple switched vendors and we are accidentally assuming it is Broadcom while in reality it is not. If it is not then the quirk might actually not apply either anymore. Regards Marcel -- To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
