Received-SPF: pass (google.com: domain of linux-bluetooth-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <tedd.an@intel.com> <20210414043857.371176-1-hj.tedd.an@gmail.com>
 <6D9FBA0A-3058-4C96-8CFD-6BE5B585DE8E@holtmann.org> <34494443-5CE2-40F1-95DF-EF05914BD741@intel.com>
In-Reply-To: <34494443-5CE2-40F1-95DF-EF05914BD741@intel.com>
From:   Archie Pusaka <apusaka@google.com>
Date:   Thu, 15 Apr 2021 11:47:55 +0800
Message-ID: <CAJQfnxFirWC+ned2sCrJb7nAiBqjCkO6guMNZU_5NCqkAdKzpg@mail.gmail.com>
Subject: Re: [BlueZ] monitor: Fix the incorrect vendor name
To:     "An, Tedd" <tedd.an@intel.com>
Cc:     Marcel Holtmann <marcel@holtmann.org>,
        "linux-bluetooth@vger.kernel.org" <linux-bluetooth@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Hi Tedd,

On Thu, 15 Apr 2021 at 10:26, An, Tedd <tedd.an@intel.com> wrote:
>
> Hi Marcel,
>
> =EF=BB=BFOn 4/14/21, 3:09 AM, "Marcel Holtmann" <marcel@holtmann.org> wro=
te:
>
>     Hi Tedd,
>
>     > This patch fixes the vendor name is alwasy shown as "Microsoft" eve=
n
>     > though a different vendor.
>     >
>     > < HCI Command: Microsoft Secure Send (0x3f|0x0009) plen 249
>     >        Type: Data fragment (0x01)
>     >> HCI Event: Command Complete (0x0e) plen 4
>     >      Microsoft Secure Send (0x3f|0x0009) ncmd 31
>     >        Status: Success (0x00)
>     > ---
>     > monitor/packet.c | 12 +++---------
>     > 1 file changed, 3 insertions(+), 9 deletions(-)
>     >
>     > diff --git a/monitor/packet.c b/monitor/packet.c
>     > index d729a01cc..91d2294ff 100644
>     > --- a/monitor/packet.c
>     > +++ b/monitor/packet.c
>     > @@ -9325,18 +9325,12 @@ static const char *get_supported_command(in=
t bit)
>     >
>     > static const char *current_vendor_str(void)
>     > {
>     > - uint16_t manufacturer, msft_opcode;
>     > + uint16_t manufacturer;
>     >
>     > - if (index_current < MAX_INDEX) {
>     > + if (index_current < MAX_INDEX)
>     >           manufacturer =3D index_list[index_current].manufacturer;
>     > -         msft_opcode =3D index_list[index_current].msft_opcode;
>     > - } else {
>     > + else
>     >           manufacturer =3D fallback_manufacturer;
>     > -         msft_opcode =3D BT_HCI_CMD_NOP;
>     > - }
>     > -
>     > - if (msft_opcode !=3D BT_HCI_CMD_NOP)
>     > -         return "Microsoft";
>
>     seems we have a bug here, but the fix can not be correct either. If w=
e are running on Intel firmware and the Microsoft extension is used, it sho=
uld show Microsoft and not Intel for the vendor commands.
>
> I submitted v2 and I think I took care of the msft_opcode handling but I =
realized that the msft_event_opcode is also like msft_opcode - each vendor =
will have a different value.
> I know the msft_event_code for Intel, which is 0x50, but don't know for R=
ealtek. (Do you happen to know?)

On my Realtek device the msft_event_code is 8 bytes long: 0x23 0x79
0x54 0x33 0x77 0x88 0x97 0x68.

localhost ~ # hcitool cmd 0x3f 0xf0 0x00
< HCI Command: ogf 0x3f, ocf 0x00f0, plen 1
  00
> HCI Event: 0x0e plen 22
  02 F0 FC 00 00 3F 00 00 00 00 00 00 00 08 23 79 54 33 77 88
  97 68

> I changed the v2 to RFC for your further comments.
>
>     Regards
>
>     Marcel
>
> Regards,
> Tedd
>

Cheers,
Archie