From: Emeltchenko Andrei <Andrei.Emeltchenko.news@gmail.com>
To: linux-bluetooth@vger.kernel.org
Subject: Re: [PATCHv2 1/5] Bluetooth: EWS: extended window size option support
Date: Fri, 7 Oct 2011 13:33:09 +0300 [thread overview]
Message-ID: <20111007103303.GA30817@aemeltch-MOBL1> (raw)
In-Reply-To: <20111006180514.GF5652@joana>
Hi Gustavo,
On Thu, Oct 06, 2011 at 03:05:14PM -0300, Gustavo Padovan wrote:
> > Adds support for EWS option. Code partly based on Qualcomm and Atheros
> > patches sent upstream a year ago.
> >
> > Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
> > ---
> > include/net/bluetooth/l2cap.h | 11 ++++++--
> > net/bluetooth/l2cap_core.c | 53 +++++++++++++++++++++++++++++++++++++---
> > net/bluetooth/l2cap_sock.c | 10 ++++----
> > 3 files changed, 62 insertions(+), 12 deletions(-)
> >
> > diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h
> > index 1f26a39..92eac16 100644
> > --- a/include/net/bluetooth/l2cap.h
> > +++ b/include/net/bluetooth/l2cap.h
> > @@ -31,7 +31,7 @@
> > #define L2CAP_DEFAULT_MTU 672
> > #define L2CAP_DEFAULT_MIN_MTU 48
> > #define L2CAP_DEFAULT_FLUSH_TO 0xffff
> > -#define L2CAP_DEFAULT_TX_WINDOW 63
> > +#define L2CAP_DEFAULT_MAX_TX_WINDOW 63
>
> Just keep this macro as is.
>
> > #define L2CAP_DEFAULT_MAX_TX 3
> > #define L2CAP_DEFAULT_RETRANS_TO 2000 /* 2 seconds */
> > #define L2CAP_DEFAULT_MONITOR_TO 12000 /* 12 seconds */
> > @@ -42,6 +42,8 @@
> > #define L2CAP_DEFAULT_SDU_ARRIVAL_TIME 0xFFFFFFFF
> > #define L2CAP_DEFAULT_ACCESS_LATENCY 0xFFFFFFFF
> >
> > +#define L2CAP_DEFAULT_MAX_EXT_WINDOW 0x3FFF
> > +
>
> and remove MAX from this one.
>
> > #define L2CAP_CONN_TIMEOUT (40000) /* 40 seconds */
> > #define L2CAP_INFO_TIMEOUT (4000) /* 4 seconds */
> >
> > @@ -240,6 +242,7 @@ struct l2cap_conf_opt {
> > #define L2CAP_CONF_RFC 0x04
> > #define L2CAP_CONF_FCS 0x05
> > #define L2CAP_CONF_EFS 0x06
> > +#define L2CAP_CONF_EWS 0x07
> >
> > #define L2CAP_CONF_MAX_SIZE 22
> >
> > @@ -357,7 +360,7 @@ struct l2cap_chan {
> >
> > __u8 fcs;
> >
> > - __u8 tx_win;
> > + __u16 tx_win;
> > __u8 max_tx;
> > __u16 retrans_timeout;
> > __u16 monitor_timeout;
> > @@ -381,7 +384,7 @@ struct l2cap_chan {
> > struct sk_buff *sdu;
> > struct sk_buff *sdu_last_frag;
> >
> > - __u8 remote_tx_win;
> > + __u16 remote_tx_win;
> > __u8 remote_max_tx;
> > __u16 remote_mps;
> >
> > @@ -488,6 +491,7 @@ enum {
> > CONF_STATE2_DEVICE,
> > CONF_LOCAL_PEND,
> > CONF_REMOTE_PEND,
> > + CONF_EWS_RECV,
> > };
> >
> > #define L2CAP_CONF_MAX_CONF_REQ 2
> > @@ -508,6 +512,7 @@ enum {
> > /* Definitions for flags in l2cap_chan */
> > enum {
> > FLAG_EFS_ENABLE,
> > + FLAG_EXT_CTRL,
> > };
> >
> > #define __set_chan_timer(c, t) l2cap_set_timer(c, &c->chan_timer, (t))
> > diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c
> > index 7aad856..69e974a 100644
> > --- a/net/bluetooth/l2cap_core.c
> > +++ b/net/bluetooth/l2cap_core.c
> > @@ -1906,6 +1906,22 @@ static inline bool __l2cap_efs_supported(struct l2cap_chan *chan)
> > return enable_hs && chan->conn->feat_mask & L2CAP_FEAT_EXT_FLOW;
> > }
> >
> > +static inline bool __l2cap_ews_supported(struct l2cap_chan *chan)
> > +{
> > + return enable_hs && chan->conn->feat_mask & L2CAP_FEAT_EXT_WINDOW;
> > +}
> > +
> > +static inline void l2cap_txwin_setup(struct l2cap_chan *chan)
> > +{
> > + if (chan->tx_win > L2CAP_DEFAULT_MAX_TX_WINDOW &&
> > + __l2cap_ews_supported(chan))
> > + /* use extended control field */
> > + set_bit(FLAG_EXT_CTRL, &chan->flags);
> > + else
> > + chan->tx_win = min_t(u16, chan->tx_win,
> > + L2CAP_DEFAULT_MAX_TX_WINDOW);
> > +}
> > +
> > static int l2cap_build_conf_req(struct l2cap_chan *chan, void *data)
> > {
> > struct l2cap_conf_req *req = data;
> > @@ -1957,7 +1973,6 @@ done:
> >
> > case L2CAP_MODE_ERTM:
> > rfc.mode = L2CAP_MODE_ERTM;
> > - rfc.txwin_size = chan->tx_win;
> > rfc.max_transmit = chan->max_tx;
> > rfc.retrans_timeout = 0;
> > rfc.monitor_timeout = 0;
> > @@ -1965,6 +1980,11 @@ done:
> > if (L2CAP_DEFAULT_MAX_PDU_SIZE > chan->conn->mtu - 10)
> > rfc.max_pdu_size = cpu_to_le16(chan->conn->mtu - 10);
> >
> > + l2cap_txwin_setup(chan);
> > +
> > + rfc.txwin_size = min_t(u16, chan->tx_win,
> > + L2CAP_DEFAULT_MAX_TX_WINDOW);
>
> rfc.txwin_size is exactly chan->tx_win.
chan->tx_win might be bigger then L2CAP_DEFAULT_MAX_TX_WINDOW since we
allow to set bigger value via socket (see below).
Current idea is to allow user to set txwin < EXT_WINDOW and if ews_supported
and txwin > TX_WINDOW then we enable extended control field.
> Also I would like to see a patch cleaning up the names we give to the tx
> window. Some places is tx_win, others txwin_size or even txwin.
This is true, I can create clean up patch later.
>
> > +
> > l2cap_add_conf_opt(&ptr, L2CAP_CONF_RFC, sizeof(rfc),
> > (unsigned long) &rfc);
> >
> > @@ -1996,6 +2016,10 @@ done:
> > l2cap_add_conf_opt(&ptr, L2CAP_CONF_FCS, 1, chan->fcs);
> > }
> >
> > + if (__l2cap_ews_supported(chan) &&
> > + test_bit(FLAG_EXT_CTRL, &chan->flags))
>
> if EXT_CTRL is true there is no need to to check for __l2cap_ews_supported().
> It is true as well.
will fix
>
> > + l2cap_add_conf_opt(&ptr, L2CAP_CONF_EWS, 2,
> > + chan->tx_win);
> > break;
> >
> > case L2CAP_MODE_STREAMING:
> > @@ -2033,7 +2057,10 @@ done:
> > chan->fcs = L2CAP_FCS_NONE;
> > l2cap_add_conf_opt(&ptr, L2CAP_CONF_FCS, 1, chan->fcs);
> > }
> > -
> > +
> > + if (__l2cap_ews_supported(chan) &&
> > + test_bit(FLAG_EXT_CTRL, &chan->flags))
> > + l2cap_add_conf_opt(&ptr, L2CAP_CONF_EWS, 2, 0);
>
> same here.
OK, btw in streaming mode we do not use txwin so I do not know how to
trigger EXT_CTRL from user.
I think I will remove this code at all.
>
> > break;
> > }
> >
> > @@ -2093,6 +2120,12 @@ static int l2cap_parse_conf_req(struct l2cap_chan *chan, void *data)
> > memcpy(&efs, (void *) val, olen);
> > break;
> >
> > + case L2CAP_CONF_EWS:
> > + set_bit(FLAG_EXT_CTRL, &chan->flags);
> > + chan->remote_tx_win = val;
> > + set_bit(CONF_EWS_RECV, &chan->conf_state);
> > + break;
> > +
> > default:
> > if (hint)
> > break;
> > @@ -2183,7 +2216,11 @@ done:
> > break;
> >
> > case L2CAP_MODE_ERTM:
> > - chan->remote_tx_win = rfc.txwin_size;
> > + if (!test_bit(CONF_EWS_RECV, &chan->conf_state))
> > + chan->remote_tx_win = rfc.txwin_size;
> > + else
> > + rfc.txwin_size = L2CAP_DEFAULT_MAX_TX_WINDOW;
> > +
>
> This seems better:
> if (test_bit(CONF_EWS_RECV)
> rfc.txwin_size = L2CAP_DEFAULT_MAX_TX_WINDOW
>
> chan->remote_tx_win = rfc.txwin_size;
remote_tx_win and rfc.txwin_size are different in size. remote_tx_win
might be bigger then L2CAP_DEFAULT_MAX_TX_WINDOW.
> Actually why does this code exist? Why do you want to set tx_win to
> L2CAP_DEFAULT_MAX_TX_WINDOW when EWS is received.
I thought it would be better to put this value despite RFC field will be
ignored. Otherwise the field would be the same as for streaming mode.
> > chan->remote_max_tx = rfc.max_transmit;
> >
> > if (le16_to_cpu(rfc.max_pdu_size) > chan->conn->mtu - 10)
> > @@ -2304,6 +2341,13 @@ static int l2cap_parse_conf_rsp(struct l2cap_chan *chan, void *rsp, int len, voi
> > l2cap_add_conf_opt(&ptr, L2CAP_CONF_EFS,
> > sizeof(efs), (unsigned long) &efs);
> > break;
> > +
> > + case L2CAP_CONF_EWS:
> > + chan->tx_win = val < L2CAP_DEFAULT_MAX_EXT_WINDOW ?
> > + val : L2CAP_DEFAULT_MAX_EXT_WINDOW;
>
> This should be <=, but the end result is the same.
I change it to:
chan->tx_win = min_t(u16, val, L2CAP_DEFAULT_EXT_WINDOW);
Best regards
Andrei Emeltchenko
next prev parent reply other threads:[~2011-10-07 10:33 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-09-16 13:39 [PATCHv2 0/5] EWS: extended window size and extended control field support Emeltchenko Andrei
2011-09-16 13:39 ` [PATCHv2 1/5] Bluetooth: EWS: extended window size option support Emeltchenko Andrei
2011-10-06 18:05 ` Gustavo Padovan
2011-10-07 10:33 ` Emeltchenko Andrei [this message]
2011-10-07 20:26 ` Gustavo Padovan
2011-10-10 8:02 ` Emeltchenko Andrei
2011-09-16 13:39 ` [PATCHv2 2/5] Bluetooth: EWS: l2cap extended control field support Emeltchenko Andrei
2011-10-06 18:19 ` Gustavo Padovan
2011-09-16 13:39 ` [PATCHv2 3/5] Bluetooth: EWS: support extended seq numbers Emeltchenko Andrei
2011-09-16 13:39 ` [PATCHv2 4/5] Bluetooth: EWS: remove magic numbers in l2cap Emeltchenko Andrei
2011-09-16 13:39 ` [PATCHv2 5/5] Bluetooth: EWS: fix max_pdu calculation Emeltchenko Andrei
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20111007103303.GA30817@aemeltch-MOBL1 \
--to=andrei.emeltchenko.news@gmail.com \
--cc=linux-bluetooth@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.