Return-Path: From: "Savoy, Pavan" To: "Gustavo F. Padovan" CC: "marcel@holtmann.org" , "linux-bluetooth@vger.kernel.org" , "linux-kernel@vger.kernel.org" Date: Tue, 19 Oct 2010 01:23:54 +0530 Subject: RE: [PATCH] Bluetooth: btwilink driver Message-ID: <19F8576C6E063C45BE387C64729E739404AA4E759A@dbde02.ent.ti.com> References: <1287176299-19313-1-git-send-email-pavan_savoy@ti.com> <20101018194811.GG2468@vigoh> In-Reply-To: <20101018194811.GG2468@vigoh> Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 List-ID: > -----Original Message----- > From: Gustavo F. Padovan [mailto:pao@profusion.mobi] On Behalf Of Gustavo= F. > Padovan > Sent: Monday, October 18, 2010 2:48 PM > To: Savoy, Pavan > Cc: marcel@holtmann.org; linux-bluetooth@vger.kernel.org; linux- > kernel@vger.kernel.org > Subject: Re: [PATCH] Bluetooth: btwilink driver > > Hi Pavan, > > * pavan_savoy@ti.com [2010-10-15 16:58:19 -0400]: > > > From: Pavan Savoy > > > > Gustavo, Marcel, > > > > Renaming the patch, since the driver is renamed to btwilink. > > Thanks for the comments, > > Please review and provide your comments on this version of patch, > > > > -- patch description -- > > > > This is the bluetooth protocol driver for the TI WiLink7 chipsets. > > Texas Instrument's WiLink chipsets combine wireless technologies > > like BT, FM, GPS and WLAN onto a single chip. > > > > This Bluetooth driver works on top of the TI_ST shared transport > > line discipline driver which also allows other drivers like > > FM V4L2 and GPS character driver to make use of the same UART interface= . > > > > Kconfig and Makefile modifications to enable the Bluetooth > > driver for Texas Instrument's WiLink 7 chipset. > > > > Signed-off-by: Pavan Savoy > > --- > > drivers/bluetooth/Kconfig | 10 + > > drivers/bluetooth/Makefile | 1 + > > drivers/bluetooth/btwilink.c | 424 > ++++++++++++++++++++++++++++++++++++++++++ > > 3 files changed, 435 insertions(+), 0 deletions(-) > > create mode 100644 drivers/bluetooth/btwilink.c > > > > diff --git a/drivers/bluetooth/Kconfig b/drivers/bluetooth/Kconfig > > index 02deef4..e0d67dd 100644 > > --- a/drivers/bluetooth/Kconfig > > +++ b/drivers/bluetooth/Kconfig > > @@ -219,4 +219,14 @@ config BT_ATH3K > > Say Y here to compile support for "Atheros firmware download driv= er" > > into the kernel or say M to compile it as module (ath3k). > > > > +config BT_WILINK > > + tristate "BlueZ bluetooth driver for TI ST" > > + depends on TI_ST > > + help > > + This enables the Bluetooth driver for Texas Instrument's BT/FM/GP= S > > + combo devices. This makes use of shared transport line discipline > > + core driver to communicate with the BT core of the combo chip. > > + > > + Say Y here to compile support for Texas Instrument's WiLink7 driv= er > > + into the kernel or say M to compile it as module. > > endmenu > > diff --git a/drivers/bluetooth/Makefile b/drivers/bluetooth/Makefile > > index 71bdf13..f4460f4 100644 > > --- a/drivers/bluetooth/Makefile > > +++ b/drivers/bluetooth/Makefile > > @@ -18,6 +18,7 @@ obj-$(CONFIG_BT_HCIBTSDIO) +=3D btsdio.o > > obj-$(CONFIG_BT_ATH3K) +=3D ath3k.o > > obj-$(CONFIG_BT_MRVL) +=3D btmrvl.o > > obj-$(CONFIG_BT_MRVL_SDIO) +=3D btmrvl_sdio.o > > +obj-$(CONFIG_BT_WILINK) +=3D btwilink.o > > > > btmrvl-y :=3D btmrvl_main.o > > btmrvl-$(CONFIG_DEBUG_FS) +=3D btmrvl_debugfs.o > > diff --git a/drivers/bluetooth/btwilink.c b/drivers/bluetooth/btwilink.= c > > new file mode 100644 > > index 0000000..f67791f > > --- /dev/null > > +++ b/drivers/bluetooth/btwilink.c > > @@ -0,0 +1,424 @@ > > +/* > > + * Texas Instrument's Bluetooth Driver For Shared Transport. > > + * > > + * Bluetooth Driver acts as interface between HCI CORE and > > + * TI Shared Transport Layer. > > + * > > + * Copyright (C) 2009-2010 Texas Instruments > > + * Author: Raja Mani > > + * Pavan Savoy > > + * > > + * This program is free software; you can redistribute it and/or modi= fy > > + * it under the terms of the GNU General Public License version 2 as > > + * published by the Free Software Foundation. > > + * > > + * This program is distributed in the hope that it will be useful, > > + * but WITHOUT ANY WARRANTY; without even the implied warranty of > > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > > + * GNU General Public License for more details. > > + * > > + * You should have received a copy of the GNU General Public License > > + * along with this program; if not, write to the Free Software > > + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-13= 07 > USA > > + * > > + */ > > + > > +#include > > +#include > > +#include > > + > > +#include > > + > > +/* Bluetooth Driver Version */ > > +#define VERSION "1.0" > > + > > +/* Defines number of seconds to wait for reg completion > > + * callback getting called from ST (in case,registration > > + * with ST returns PENDING status) > > + */ > > +#define BT_REGISTER_TIMEOUT 6000 /* 6 sec */ > > + > > +/** > > + * struct ti_st - BT driver operation structure > > + * @hdev: hci device pointer which binds to bt driver > > + * @flags: used locally,to maintain various BT driver status > > + * @streg_cbdata: to hold ST registration callback status > > + * @st_write: write function pointer of ST driver > > + * @wait_reg_completion - completion sync between ti_st_open > > + * and ti_st_registration_completion_cb. > > + */ > > +struct ti_st { > > + struct hci_dev *hdev; > > + char streg_cbdata; > > + long (*st_write) (struct sk_buff *); > > + struct completion wait_reg_completion; > > +}; > > + > > +static int reset; > > + > > +/* Increments HCI counters based on pocket ID (cmd,acl,sco) */ > > +static inline void ti_st_tx_complete(struct ti_st *hst, int pkt_type) > > +{ > > + struct hci_dev *hdev; > > + hdev =3D hst->hdev; > > + > > + /* Update HCI stat counters */ > > + switch (pkt_type) { > > + case HCI_COMMAND_PKT: > > + hdev->stat.cmd_tx++; > > + break; > > + > > + case HCI_ACLDATA_PKT: > > + hdev->stat.acl_tx++; > > + break; > > + > > + case HCI_SCODATA_PKT: > > + hdev->stat.sco_tx++; > > + break; > > + } > > +} > > + > > +/* ------- Interfaces to Shared Transport ------ */ > > + > > +/* Called by ST layer to indicate protocol registration completion > > + * status.ti_st_open() function will wait for signal from this > > + * API when st_register() function returns ST_PENDING. > > + */ > > +static void st_registration_completion_cb(void *priv_data, char data) > > +{ > > + struct ti_st *lhst =3D (struct ti_st *)priv_data; > > Blank line here. > > > + /* ti_st_open() function needs value of 'data' to know > > + * the registration status(success/fail),So have a back > > + * up of it. > > + */ > > + lhst->streg_cbdata =3D data; > > + > > + /* Got a feedback from ST for BT driver registration > > + * request.Wackup ti_st_open() function to continue > > + * it's open operation. > > + */ > > + complete(&lhst->wait_reg_completion); > > +} > > + > > +/* Called by Shared Transport layer when receive data is > > + * available */ > > +static long st_receive(void *priv_data, struct sk_buff *skb) > > +{ > > + int err; > > + struct ti_st *lhst =3D (struct ti_st *)priv_data; > > + > > + if (!skb) > > + return -EFAULT; > > + > > + if (!lhst) { > > + kfree_skb(skb); > > + return -EFAULT; > > + } > > + > > + skb->dev =3D (struct net_device *)lhst->hdev; > > + > > + /* Forward skb to HCI CORE layer */ > > + err =3D hci_recv_frame(skb); > > + if (err) { > > + kfree_skb(skb); > > + BT_ERR("Unable to push skb to HCI CORE(%d)", err); > > s/CORE/core/ > > > + return err; > > + } > > + > > + lhst->hdev->stat.byte_rx +=3D skb->len; > > + return 0; > > +} > > + > > +/* ------- Interfaces to HCI layer ------ */ > > +/* protocol structure registered with shared transport */ > > +static struct st_proto_s ti_st_proto =3D { > > + .type =3D ST_BT, > > + .recv =3D st_receive, > > + .reg_complete_cb =3D st_registration_completion_cb, > > + .priv_data =3D NULL, > > +}; > > + > > +/* Called from HCI core to initialize the device */ > > +static int ti_st_open(struct hci_dev *hdev) > > +{ > > + unsigned long timeleft; > > + struct ti_st *hst; > > + int err; > > + > > + BT_DBG("%s %p", hdev->name, hdev); > > + > > + /* provide contexts for callbacks from ST */ > > + hst =3D hdev->driver_data; > > + ti_st_proto.priv_data =3D hst; > > + > > + err =3D st_register(&ti_st_proto); > > + if (err =3D=3D -EINPROGRESS) { > > + /* Prepare wait-for-completion handler data structures. > > + * Needed to syncronize this and st_registration_completion= _cb() > > + * functions. > > + */ > > + init_completion(&hst->wait_reg_completion); > > + > > + /* Reset ST registration callback status flag , this value > > + * will be updated in ti_st_registration_completion_cb() > > + * function whenever it called from ST driver. > > + */ > > + hst->streg_cbdata =3D -EINPROGRESS; > > + > > + /* ST is busy with other protocol registration(may be busy = with > > + * firmware download).So,Wait till the registration callbac= k > > + * (passed as a argument to st_register() function) getting > > + * called from ST. > > + */ > > + BT_DBG(" waiting for reg completion signal from ST"); > > + > > + timeleft =3D wait_for_completion_timeout > > + (&hst->wait_reg_completion, > > + msecs_to_jiffies(BT_REGISTER_TIMEOUT)); > > + if (!timeleft) { > > + BT_ERR("Timeout(%d sec),didn't get reg" > > + "completion signal from ST", > > + BT_REGISTER_TIMEOUT / 1000); > > How does this get printed. "...regcompletion..." ? > > > + return -ETIMEDOUT; > > + } > > + > > + /* Is ST registration callback called with ERROR value? */ > > + if (hst->streg_cbdata !=3D 0) { > > + BT_ERR("ST reg completion CB called with invalid" > > + "status %d", hst->streg_cbdata); > > + return -EAGAIN; > > + } > > + err =3D 0; > > + } else if (err =3D=3D -EPERM) { > > + BT_ERR("st_register failed %d", err); > > + return -EAGAIN; > > Why? if -EPERM return -EAGAIN? > > > + } > > + > > + /* Do we have proper ST write function? */ > > + if (ti_st_proto.write !=3D NULL) { > > + /* We need this pointer for sending any Bluetooth pkts */ > > + hst->st_write =3D ti_st_proto.write; > > I asked in the other e-mail: Who sets ti_st_proto.write()? I didn't get > this. The write function is set by the TI-ST driver. This was in order to avoid another EXPORT_SYMBOL(st_write) or so. So, as soon as some protocol driver registers to shared transport driver, t= he registration function inside the driver will set it's write function to thi= s function pointer. Yes, I will fix up the rest and post another version, Thanks. > > + } else { > > + BT_ERR("failed to get ST write func pointer"); > > + > > + /* Undo registration with ST */ > > + err =3D st_unregister(ST_BT); > > + if (err) > > + BT_ERR("st_unregister failed %d", err); > > + > > + hst->st_write =3D NULL; > > + return -EAGAIN; > > return err; instead > > > + } > > + > > + /* Registration with ST layer is completed successfully, > > + * now chip is ready to accept commands from HCI CORE. > > + * Mark HCI Device flag as RUNNING > > + */ > > + set_bit(HCI_RUNNING, &hdev->flags); > > + return err; > > +} > > + > > +/* Close device */ > > +static int ti_st_close(struct hci_dev *hdev) > > +{ > > + int err =3D 0; > > do not set err to 0 here. > > > + struct ti_st *hst; > > you can set hst to hdev->.... here. > > > + > > + hst =3D hdev->driver_data; > > + > > + /* Clear HCI device RUNNING flag */ > > + if (!test_and_clear_bit(HCI_RUNNING, &hdev->flags)) > > + BT_ERR("HCI not RUNNING?"); > > > You didn't fix the things I told you. This is not a error. Just return 0 > here. > > > + > > + /* continue to unregister from transport */ > > + err =3D st_unregister(ST_BT); > > + if (err) > > + BT_ERR("st_unregister failed %d", err); > > + > > + hst->st_write =3D NULL; > > + return err; > > +} > > + > > +/* Called from HCI CORE , Sends frames to Shared Transport */ > > +static int ti_st_send_frame(struct sk_buff *skb) > > +{ > > + struct hci_dev *hdev; > > + struct ti_st *hst; > > + long len; > > + > > + if (!skb) > > + return -ENOMEM; > > + > > + hdev =3D (struct hci_dev *)skb->dev; > > + if (!hdev) > > + return -ENODEV; > > + > > + if (!test_bit(HCI_RUNNING, &hdev->flags)) > > + return -EBUSY; > > + > > + hst =3D (struct ti_st *)hdev->driver_data; > > + > > + /* Prepend skb with frame type */ > > + memcpy(skb_push(skb, 1), &bt_cb(skb)->pkt_type, 1); > > + > > + BT_DBG(" %s: type %d len %d", hdev->name, bt_cb(skb)->pkt_type, > > + skb->len); > > + > > + /* Insert skb to shared transport layer's transmit queue. > > + * Freeing skb memory is taken care in shared transport layer, > > + * so don't free skb memory here. > > + */ > > + if (!hst->st_write) { > > + kfree_skb(skb); > > + BT_ERR(" Can't write to ST, st_write null?"); > > + return -EAGAIN; > > + } > > + > > + len =3D hst->st_write(skb); > > + if (len < 0) { > > + kfree_skb(skb); > > + BT_ERR(" ST write failed (%ld)", len); > > + return -EAGAIN; > > + } > > Explain to me why it is worth return -EAGAIN in both cases. > > > + > > + /* ST accepted our skb. So, Go ahead and do rest */ > > + hdev->stat.byte_tx +=3D len; > > + ti_st_tx_complete(hst, bt_cb(skb)->pkt_type); > > + > > + return 0; > > +} > > + > > +static void ti_st_destruct(struct hci_dev *hdev) > > +{ > > + if (!hdev) > > + BT_ERR("Destruct called with invalid HCI Device" > > + "(hdev=3DNULL)"); > > + > > + BT_DBG("%s", hdev->name); > > + > > + /* free ti_st memory */ > > + kfree(hdev->driver_data); > > + return; > > +} > > + > > +/* Creates new HCI device */ > > +static int ti_st_register_dev(struct ti_st *hst) > > +{ > > + struct hci_dev *hdev; > > + > > + /* Initialize and register HCI device */ > > + hdev =3D hci_alloc_dev(); > > + if (!hdev) > > + return -ENOMEM; > > + > > + BT_DBG(" HCI device allocated. hdev=3D %p", hdev); > > just BT_DBG("hdev=3D %p", hdev); > > > + > > + hst->hdev =3D hdev; > > + hdev->bus =3D HCI_UART; > > + hdev->driver_data =3D hst; > > + hdev->open =3D ti_st_open; > > + hdev->close =3D ti_st_close; > > + hdev->flush =3D NULL; > > + hdev->send =3D ti_st_send_frame; > > + hdev->destruct =3D ti_st_destruct; > > + hdev->owner =3D THIS_MODULE; > > + > > + if (reset) > > + set_bit(HCI_QUIRK_NO_RESET, &hdev->quirks); > > + > > + if (hci_register_dev(hdev) < 0) { > > err =3D hci_register_dev() > if (err < 0) > ... > ... > return err; > > > + BT_ERR("Can't register HCI device"); > > + hci_free_dev(hdev); > > + return -ENODEV; > > + } > > + > > + BT_DBG(" HCI device registered. hdev=3D %p", hdev); > > + return 0; > > +} > > + > > + > > +static int bt_ti_probe(struct platform_device *pdev) > > +{ > > + int err; > > + static struct ti_st *hst; > > + err =3D 0; > > + > > + BT_DBG(" Bluetooth Driver Version %s", VERSION); > > + > > + hst =3D kzalloc(sizeof(struct ti_st), GFP_KERNEL); > > + if (!hst) { > > + BT_ERR("Can't allocate control structure"); > > remove this BT_ERR. > > > + return -ENOMEM; > > + } > > + > > + /* Expose "hciX" device to user space */ > > + err =3D ti_st_register_dev(hst); > > + if (err) { > > + kfree(hst); > > + BT_ERR("Unable to expose hciX device(%d)", err); > > Ok, if you don't know how to extract the hci id, remove the whole error > message then. > > > + return err; > > + } > > + > > + dev_set_drvdata(&pdev->dev, hst); > > + return err; > > +} > > + > > +static int bt_ti_remove(struct platform_device *pdev) > > +{ > > + struct ti_st *hst; > > + > > + hst =3D dev_get_drvdata(&pdev->dev); > > + /* Deallocate local resource's memory */ > > + if (hst) { > > + struct hci_dev *hdev =3D hst->hdev; > > + if (!hdev) { > > + BT_ERR("Invalid hdev memory"); > > + kfree(hst); > > + } else { > > + ti_st_close(hdev); > > + hci_unregister_dev(hdev); > > + /* Free HCI device memory */ > > + hci_free_dev(hdev); > > + } > > + } > > + return 0; > > +} > > + > > +static struct platform_driver btwilink_driver =3D { > > + .probe =3D bt_ti_probe, > > + .remove =3D bt_ti_remove, > > + .driver =3D { > > + .name =3D "btwilink", > > + .owner =3D THIS_MODULE, > > + }, > > +}; > > + > > +/* ------- Module Init/Exit interfaces ------ */ > > +static int __init bt_drv_init(void) > > +{ > > + long ret; > > + > > + ret =3D platform_driver_register(&btwilink_driver); > > + if (ret !=3D 0) { > > + BT_ERR("btwilink platform drv registration failed"); > > + return -EPERM; > > + } > > + return 0; > > if (ret) > BT_ERR("btwilink platform drv registration failed"); > > return ret; > > and s/drv/driver/ > > > Please fix all the issues, then we go to another round of review. ;) > > > -- > Gustavo F. Padovan > ProFUSION embedded systems - http://profusion.mobi