Re: [PATCH v7] Bluetooth: btwilink driver

From: Pavan Savoy
Date: Tue Nov 30 2010 - 02:29:58 EST


On Fri, Nov 26, 2010 at 2:50 PM, <pavan_savoy@xxxxxx> wrote:
> From: Pavan Savoy <pavan_savoy@xxxxxx>
>
Marcel, Gustavo,

Please find some time to comment and also answer some of the
concerns I have below,

Thanks & Regards,
Pavan Savoy.

> comments attended to from v5 and v6,
>
> 1. Inside ti_st_open, I previously only checked for EINPROGRESS & EPERM,
> Now I handle for EINPROGRESS - which is not really an error and
> return during all other error cases.
>
> 2. _write is still a function pointer and not an exported function, I
> need to change the underlying driver's code for this.
> However, previous lkml comments on the underlying driver's code
> suggested it to be kept as a function pointer and not EXPORT.
> Gustavo, Marcel - Please comment on this.
> Is this absolutely required? If so why?
>
> 3. test_and_set_bit of HCI_RUNNING is done at beginning of
> ti_st_open, and did not see issues during firmware download.
> However ideally I would still like to set HCI_RUNNING once the firmware
> download is done, because I don't want to allow a _send_frame during
> firmware download - Marcel, Gustavo - Please comment.
>
> 4. test_and_clear of HCI_RUNNING now done @ beginning of close.
>
> 5. EAGAIN on failure of st_write is to suggest to try and write again.
> I have never this happen - However only if UART goes bad this case may
> occur.
>
> 6. ti_st_tx_complete is very similar to hci_ldisc's tx_complete - in
> fact the code is pretty much borrowed from there.
> Marcel, Gustavo - Please suggest where should it be done? If not here.
>
> 7. comments cleaned-up + hst memory leak fixed when hci_alloc_dev fails.
>
> 8. platform_driver registration inside module_init now is similar to
> other drivers.
>
> 9. Dan Carpenter's comments on leaking hst memory on failed
> hci_register_dev and empty space after quotes in debug statements
> fixed.
>
> Thanks for the comments...
> Sorry, for previously not being very clear on which comments were
> handled and which were not.
>
> -- 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 <pavan_savoy@xxxxxx>
> ---
> Âdrivers/bluetooth/Kconfig  Â|  10 ++
> Âdrivers/bluetooth/Makefile  |  Â1 +
> Âdrivers/bluetooth/btwilink.c | Â363 ++++++++++++++++++++++++++++++++++++++++++
> Â3 files changed, 374 insertions(+), 0 deletions(-)
> Âcreate mode 100644 drivers/bluetooth/btwilink.c
>
> diff --git a/drivers/bluetooth/Kconfig b/drivers/bluetooth/Kconfig
> index 02deef4..8e0de9a 100644
> --- a/drivers/bluetooth/Kconfig
> +++ b/drivers/bluetooth/Kconfig
> @@ -219,4 +219,14 @@ config BT_ATH3K
> Â Â Â Â ÂSay Y here to compile support for "Atheros firmware download driver"
> Â Â Â Â Âinto the kernel or say M to compile it as module (ath3k).
>
> +config BT_WILINK
> + Â Â Â tristate "Texas Instruments WiLink7 driver"
> + Â Â Â depends on TI_ST
> + Â Â Â help
> + Â Â Â Â This enables the Bluetooth driver for Texas Instrument's BT/FM/GPS
> + Â Â Â Â combo devices. This makes use of shared transport line discipline
> + Â Â Â Â core driver to communicate with the BT core of the combo chip.
> +
> + Â Â Â Â Say Y here to compile support for Texas Instrument's WiLink7 driver
> + Â Â Â Â into the kernel or say M to compile it as module.
> Âendmenu
> 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-$(CONFIG_BT_HCIBTSDIO) Â Â+= btsdio.o
> Âobj-$(CONFIG_BT_ATH3K) Â Â Â Â += ath3k.o
> Âobj-$(CONFIG_BT_MRVL) Â Â Â Â Â+= btmrvl.o
> Âobj-$(CONFIG_BT_MRVL_SDIO) Â Â += btmrvl_sdio.o
> +obj-$(CONFIG_BT_WILINK) Â Â Â Â Â Â Â Â+= btwilink.o
>
> Âbtmrvl-y            := btmrvl_main.o
> Âbtmrvl-$(CONFIG_DEBUG_FS) Â Â Â+= btmrvl_debugfs.o
> diff --git a/drivers/bluetooth/btwilink.c b/drivers/bluetooth/btwilink.c
> new file mode 100644
> index 0000000..71e69f8
> --- /dev/null
> +++ b/drivers/bluetooth/btwilink.c
> @@ -0,0 +1,363 @@
> +/*
> + * ÂTexas Instrument's Bluetooth Driver For Shared Transport.
> + *
> + * ÂBluetooth Driver acts as interface between HCI core and
> + * ÂTI Shared Transport Layer.
> + *
> + * ÂCopyright (C) 2009-2010 Texas Instruments
> + * ÂAuthor: Raja Mani <raja_mani@xxxxxx>
> + * Â Â Pavan Savoy <pavan_savoy@xxxxxx>
> + *
> + * ÂThis program is free software; you can redistribute it and/or modify
> + * Âit under the terms of the GNU General Public License version 2 as
> + * Âpublished by the Free Software Foundation.
> + *
> + * ÂThis program is distributed in the hope that it will be useful,
> + * Âbut WITHOUT ANY WARRANTY; without even the implied warranty of
> + * ÂMERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. ÂSee the
> + * ÂGNU General Public License for more details.
> + *
> + * ÂYou should have received a copy of the GNU General Public License
> + * Âalong with this program; if not, write to the Free Software
> + * ÂFoundation, Inc., 59 Temple Place, Suite 330, Boston, MA Â02111-1307 ÂUSA
> + *
> + */
> +
> +#include <linux/platform_device.h>
> +#include <net/bluetooth/bluetooth.h>
> +#include <net/bluetooth/hci_core.h>
> +
> +#include <linux/ti_wilink_st.h>
> +
> +/* Bluetooth Driver Version */
> +#define VERSION Â Â Â Â Â Â Â "1.0"
> +
> +/* Number of seconds to wait for registration completion
> + * when ST returns PENDING status.
> + */
> +#define BT_REGISTER_TIMEOUT Â 6000 Â Â /* 6 sec */
> +
> +/**
> + * struct ti_st - driver operation structure
> + * @hdev: hci device pointer which binds to bt driver
> + * @reg_status: ST registration callback status
> + * @st_write: write function provided by the ST driver
> + * Â Â to be used by the driver during send_frame.
> + * @wait_reg_completion - completion sync between ti_st_open
> + * Â Â and ti_st_registration_completion_cb.
> + */
> +struct ti_st {
> + Â Â Â struct hci_dev *hdev;
> + Â Â Â char reg_status;
> + Â Â Â long (*st_write) (struct sk_buff *);
> + Â Â Â struct completion wait_reg_completion;
> +};
> +
> +/* Increments HCI counters based on pocket ID (cmd,acl,sco) */
> +static inline void ti_st_tx_complete(struct ti_st *hst, int pkt_type)
> +{
> + Â Â Â struct hci_dev *hdev = hst->hdev;
> +
> + Â Â Â /* Update HCI stat counters */
> + Â Â Â switch (pkt_type) {
> + Â Â Â case HCI_COMMAND_PKT:
> + Â Â Â Â Â Â Â hdev->stat.cmd_tx++;
> + Â Â Â Â Â Â Â break;
> +
> + Â Â Â case HCI_ACLDATA_PKT:
> + Â Â Â Â Â Â Â hdev->stat.acl_tx++;
> + Â Â Â Â Â Â Â break;
> +
> + Â Â Â case HCI_SCODATA_PKT:
> + Â Â Â Â Â Â Â hdev->stat.sco_tx++;
> + Â Â Â Â Â Â Â break;
> + Â Â Â }
> +}
> +
> +/* ------- 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)
> +{
> + Â Â Â struct ti_st *lhst = priv_data;
> +
> + Â Â Â /* Save registration status for use in ti_st_open() */
> + Â Â Â lhst->reg_status = data;
> + Â Â Â /* complete the wait in ti_st_open() */
> + Â Â Â 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)
> +{
> + Â Â Â struct ti_st *lhst = priv_data;
> + Â Â Â int err;
> +
> + Â Â Â if (!skb)
> + Â Â Â Â Â Â Â return -EFAULT;
> +
> + Â Â Â if (!lhst) {
> + Â Â Â Â Â Â Â kfree_skb(skb);
> + Â Â Â Â Â Â Â return -EFAULT;
> + Â Â Â }
> +
> + Â Â Â skb->dev = (void *) lhst->hdev;
> +
> + Â Â Â /* Forward skb to HCI core layer */
> + Â Â Â err = hci_recv_frame(skb);
> + Â Â Â if (err < 0) {
> + Â Â Â Â Â Â Â BT_ERR("Unable to push skb to HCI core(%d)", err);
> + Â Â Â Â Â Â Â return err;
> + Â Â Â }
> +
> + Â Â Â lhst->hdev->stat.byte_rx += skb->len;
> +
> + Â Â Â return 0;
> +}
> +
> +/* ------- Interfaces to HCI layer ------ */
> +/* protocol structure registered with shared transport */
> +static struct st_proto_s ti_st_proto = {
> + Â Â Â .type = ST_BT,
> + Â Â Â .recv = st_receive,
> + Â Â Â .reg_complete_cb = st_registration_completion_cb,
> +};
> +
> +/* Called from HCI core to initialize the device */
> +static int ti_st_open(struct hci_dev *hdev)
> +{
> + Â Â Â unsigned long timeleft;
> + Â Â Â struct ti_st *hst;
> + Â Â Â int err;
> +
> + Â Â Â BT_DBG("%s %p", hdev->name, hdev);
> + Â Â Â if (test_and_set_bit(HCI_RUNNING, &hdev->flags)) {
> + Â Â Â Â Â Â Â BT_ERR("btwilink already opened");
> + Â Â Â Â Â Â Â return -EBUSY;
> + Â Â Â }
> +
> + Â Â Â /* provide contexts for callbacks from ST */
> + Â Â Â hst = hdev->driver_data;
> + Â Â Â ti_st_proto.priv_data = hst;
> +
> + Â Â Â err = st_register(&ti_st_proto);
> + Â Â Â if (err == -EINPROGRESS) {
> + Â Â Â Â Â Â Â /* ST is busy with either protocol registration or firmware
> + Â Â Â Â Â Â Â Â* download.
> + Â Â Â Â Â Â Â Â*/
> + Â Â Â Â Â Â Â /* Prepare wait-for-completion handler data structures.
> + Â Â Â Â Â Â Â Â*/
> + Â Â Â Â Â Â Â init_completion(&hst->wait_reg_completion);
> +
> + Â Â Â Â Â Â Â /* Reset ST registration callback status flag , this value
> + Â Â Â Â Â Â Â Â* will be updated in ti_st_registration_completion_cb()
> + Â Â Â Â Â Â Â Â* function whenever it called from ST driver.
> + Â Â Â Â Â Â Â Â*/
> + Â Â Â Â Â Â Â hst->reg_status = -EINPROGRESS;
> +
> + Â Â Â Â Â Â Â BT_DBG("waiting for registration completion signal from ST");
> + Â Â Â Â Â Â Â timeleft = wait_for_completion_timeout
> + Â Â Â Â Â Â Â Â Â Â Â (&hst->wait_reg_completion,
> + Â Â Â Â Â Â Â Â Â Â Â Âmsecs_to_jiffies(BT_REGISTER_TIMEOUT));
> + Â Â Â Â Â Â Â if (!timeleft) {
> + Â Â Â Â Â Â Â Â Â Â Â clear_bit(HCI_RUNNING, &hdev->flags);
> + Â Â Â Â Â Â Â Â Â Â Â BT_ERR("Timeout(%d sec),didn't get reg "
> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â "completion signal from ST",
> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â BT_REGISTER_TIMEOUT / 1000);
> + Â Â Â Â Â Â Â Â Â Â Â return -ETIMEDOUT;
> + Â Â Â Â Â Â Â }
> +
> + Â Â Â Â Â Â Â /* Is ST registration callback called with ERROR status? */
> + Â Â Â Â Â Â Â if (hst->reg_status != 0) {
> + Â Â Â Â Â Â Â Â Â Â Â clear_bit(HCI_RUNNING, &hdev->flags);
> + Â Â Â Â Â Â Â Â Â Â Â BT_ERR("ST registration completed with invalid "
> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â "status %d", hst->reg_status);
> + Â Â Â Â Â Â Â Â Â Â Â return -EAGAIN;
> + Â Â Â Â Â Â Â }
> + Â Â Â Â Â Â Â err = 0;
> + Â Â Â } else if (err != 0) {
> + Â Â Â Â Â Â Â clear_bit(HCI_RUNNING, &hdev->flags);
> + Â Â Â Â Â Â Â BT_ERR("st_register failed %d", err);
> + Â Â Â Â Â Â Â return err;
> + Â Â Â }
> +
> + Â Â Â /* ti_st_proto.write is filled up by the underlying shared
> + Â Â Â Â* transport driver upon registration
> + Â Â Â Â*/
> + Â Â Â hst->st_write = ti_st_proto.write;
> + Â Â Â if (!hst->st_write) {
> + Â Â Â Â Â Â Â BT_ERR("undefined ST write function");
> + Â Â Â Â Â Â Â clear_bit(HCI_RUNNING, &hdev->flags);
> +
> + Â Â Â Â Â Â Â /* Undo registration with ST */
> + Â Â Â Â Â Â Â err = st_unregister(ST_BT);
> + Â Â Â Â Â Â Â if (err)
> + Â Â Â Â Â Â Â Â Â Â Â BT_ERR("st_unregister() failed with error %d", err);
> +
> + Â Â Â Â Â Â Â hst->st_write = NULL;
> + Â Â Â Â Â Â Â return err;
> + Â Â Â }
> +
> + Â Â Â return err;
> +}
> +
> +/* Close device */
> +static int ti_st_close(struct hci_dev *hdev)
> +{
> + Â Â Â int err;
> + Â Â Â struct ti_st *hst = hdev->driver_data;
> +
> + Â Â Â if (!test_and_clear_bit(HCI_RUNNING, &hdev->flags))
> + Â Â Â Â Â Â Â return 0;
> +
> + Â Â Â /* continue to unregister from transport */
> + Â Â Â err = st_unregister(ST_BT);
> + Â Â Â if (err)
> + Â Â Â Â Â Â Â BT_ERR("st_unregister() failed with error %d", err);
> +
> + Â Â Â hst->st_write = NULL;
> +
> + Â Â Â return err;
> +}
> +
> +static int ti_st_send_frame(struct sk_buff *skb)
> +{
> + Â Â Â struct hci_dev *hdev;
> + Â Â Â struct ti_st *hst;
> + Â Â Â long len;
> +
> + Â Â Â hdev = (struct hci_dev *)skb->dev;
> +
> + Â Â Â if (!test_bit(HCI_RUNNING, &hdev->flags))
> + Â Â Â Â Â Â Â return -EBUSY;
> +
> + Â Â Â hst = hdev->driver_data;
> +
> + Â Â Â /* Prepend skb with frame type */
> + Â Â Â memcpy(skb_push(skb, 1), &bt_cb(skb)->pkt_type, 1);
> +
> + Â Â Â BT_DBG("%s: type %d len %d", hdev->name, bt_cb(skb)->pkt_type,
> + Â Â Â Â Â Â Â Â Â Â Â skb->len);
> +
> + Â Â Â /* Insert skb to shared transport layer's transmit queue.
> + Â Â Â Â* Freeing skb memory is taken care in shared transport layer,
> + Â Â Â Â* so don't free skb memory here.
> + Â Â Â Â*/
> + Â Â Â len = hst->st_write(skb);
> + Â Â Â if (len < 0) {
> + Â Â Â Â Â Â Â kfree_skb(skb);
> + Â Â Â Â Â Â Â BT_ERR("ST write failed (%ld)", len);
> + Â Â Â Â Â Â Â /* Try Again, would only fail if UART has gone bad */
> + Â Â Â Â Â Â Â return -EAGAIN;
> + Â Â Â }
> +
> + Â Â Â /* ST accepted our skb. So, Go ahead and do rest */
> + Â Â Â hdev->stat.byte_tx += len;
> + Â Â Â ti_st_tx_complete(hst, bt_cb(skb)->pkt_type);
> +
> + Â Â Â return 0;
> +}
> +
> +static void ti_st_destruct(struct hci_dev *hdev)
> +{
> + Â Â Â BT_DBG("%s", hdev->name);
> + Â Â Â kfree(hdev->driver_data);
> +}
> +
> +static int bt_ti_probe(struct platform_device *pdev)
> +{
> + Â Â Â static struct ti_st *hst;
> + Â Â Â struct hci_dev *hdev;
> + Â Â Â int err;
> +
> + Â Â Â hst = kzalloc(sizeof(struct ti_st), GFP_KERNEL);
> + Â Â Â if (!hst)
> + Â Â Â Â Â Â Â return -ENOMEM;
> +
> + Â Â Â /* Expose "hciX" device to user space */
> + Â Â Â hdev = hci_alloc_dev();
> + Â Â Â if (!hdev) {
> + Â Â Â Â Â Â Â kfree(hst);
> + Â Â Â Â Â Â Â return -ENOMEM;
> + Â Â Â }
> +
> + Â Â Â BT_DBG("hdev %p", hdev);
> +
> + Â Â Â hst->hdev = hdev;
> + Â Â Â hdev->bus = HCI_UART;
> + Â Â Â hdev->driver_data = hst;
> + Â Â Â hdev->open = ti_st_open;
> + Â Â Â hdev->close = ti_st_close;
> + Â Â Â hdev->flush = NULL;
> + Â Â Â hdev->send = ti_st_send_frame;
> + Â Â Â hdev->destruct = ti_st_destruct;
> + Â Â Â hdev->owner = THIS_MODULE;
> +
> + Â Â Â err = hci_register_dev(hdev);
> + Â Â Â if (err < 0) {
> + Â Â Â Â Â Â Â BT_ERR("Can't register HCI device error %d", err);
> + Â Â Â Â Â Â Â kfree(hst);
> + Â Â Â Â Â Â Â hci_free_dev(hdev);
> + Â Â Â Â Â Â Â return err;
> + Â Â Â }
> +
> + Â Â Â BT_DBG("HCI device registered (hdev %p)", hdev);
> +
> + Â Â Â dev_set_drvdata(&pdev->dev, hst);
> + Â Â Â return err;
> +}
> +
> +static int bt_ti_remove(struct platform_device *pdev)
> +{
> + Â Â Â struct hci_dev *hdev;
> + Â Â Â struct ti_st *hst = dev_get_drvdata(&pdev->dev);
> +
> + Â Â Â if (!hst)
> + Â Â Â Â Â Â Â return -EFAULT;
> +
> + Â Â Â hdev = hst->hdev;
> + Â Â Â ti_st_close(hdev);
> + Â Â Â hci_unregister_dev(hdev);
> +
> + Â Â Â hci_free_dev(hdev);
> + Â Â Â kfree(hst);
> +
> + Â Â Â dev_set_drvdata(&pdev->dev, NULL);
> + Â Â Â return 0;
> +}
> +
> +static struct platform_driver btwilink_driver = {
> + Â Â Â .probe = bt_ti_probe,
> + Â Â Â .remove = bt_ti_remove,
> + Â Â Â .driver = {
> + Â Â Â Â Â Â Â .name = "btwilink",
> + Â Â Â Â Â Â Â .owner = THIS_MODULE,
> + Â Â Â },
> +};
> +
> +/* ------- Module Init/Exit interfaces ------ */
> +static int __init btwilink_init(void)
> +{
> + Â Â Â BT_INFO("Bluetooth Driver for TI WiLink - Version %s", VERSION);
> +
> + Â Â Â return platform_driver_register(&btwilink_driver);
> +}
> +
> +static void __exit btwilink_exit(void)
> +{
> + Â Â Â platform_driver_unregister(&btwilink_driver);
> +}
> +
> +module_init(btwilink_init);
> +module_exit(btwilink_exit);
> +
> +/* ------ Module Info ------ */
> +
> +MODULE_AUTHOR("Raja Mani <raja_mani@xxxxxx>");
> +MODULE_DESCRIPTION("Bluetooth Driver for TI Shared Transport" VERSION);
> +MODULE_VERSION(VERSION);
> +MODULE_LICENSE("GPL");
> --
> 1.5.6.3
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at Âhttp://vger.kernel.org/majordomo-info.html
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/