Return-Path: <cyrus@holtmann.org>
Content-Type: text/plain; charset=us-ascii
Mime-Version: 1.0 (Mac OS X Mail 8.2 \(2104\))
Subject: Re: [PATCH v2 2/4] Bluetooth: hci_bcm: Fix IRQ polarity for T100
From: Marcel Holtmann <marcel@holtmann.org>
In-Reply-To: <1441373747-23042-3-git-send-email-frederic.danis@linux.intel.com>
Date: Fri, 4 Sep 2015 12:13:20 -0700
Cc: linux-bluetooth@vger.kernel.org
Message-Id: <16F8AF0D-420C-45C4-A042-1FCA3BD9DD1C@holtmann.org>
References: <1441373747-23042-1-git-send-email-frederic.danis@linux.intel.com> <1441373747-23042-3-git-send-email-frederic.danis@linux.intel.com>
To: Frederic Danis <frederic.danis@linux.intel.com>
Sender: linux-bluetooth-owner@vger.kernel.org
List-ID: <linux-bluetooth.vger.kernel.org>

Hi Fred,

> ACPI table for BCM2E39 of T100TA is not correct.
> Invert irq_polarity for this device.
> 
> Signed-off-by: Frederic Danis <frederic.danis@linux.intel.com>
> ---
> drivers/bluetooth/hci_bcm.c | 21 +++++++++++++++++++++
> 1 file changed, 21 insertions(+)
> 
> diff --git a/drivers/bluetooth/hci_bcm.c b/drivers/bluetooth/hci_bcm.c
> index f306541..efb9566 100644
> --- a/drivers/bluetooth/hci_bcm.c
> +++ b/drivers/bluetooth/hci_bcm.c
> @@ -32,6 +32,7 @@
> #include <linux/gpio/consumer.h>
> #include <linux/tty.h>
> #include <linux/interrupt.h>
> +#include <linux/dmi.h>
> 
> #include <net/bluetooth/bluetooth.h>
> #include <net/bluetooth/hci_core.h>
> @@ -546,6 +547,20 @@ static int bcm_resource(struct acpi_resource *ares, void *data)
> 	return 1;
> }
> 
> +/* IRQ polarity of some chipset are not defined correctly in ACPI table. */
> +static const struct dmi_system_id bcm_wrong_irq_dmi_table[] = {
> +#if defined(CONFIG_DMI) && defined(CONFIG_X86)
> +	{
> +		/* Asus T100TA */

I think instead of a comment you could just fill in .ident here. Or is that suppose to be used for something else?

> +		.matches = {
> +			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
> +			DMI_MATCH(DMI_PRODUCT_NAME, "T100TA"),

Why are we not using DMI_EXACT_MATCH here in the first place. We really know which one is broken, correct?

> +		},
> +	},
> +#endif

What is this #ifdef buying us? Is include/linux/dmi.h in some way that we can not have this defined all the time?

> +	{ }
> +};
> +
> static int bcm_acpi_probe(struct bcm_device *dev)
> {
> 	struct platform_device *pdev = dev->pdev;
> @@ -608,6 +623,12 @@ static int bcm_acpi_probe(struct bcm_device *dev)
> 
> 	acpi_dev_get_resources(adev, &resources, bcm_resource, dev);
> 
> +	if (strstr(id->id, "BCM2E39") &&

Lets not bother with that check and just always run through the DMI table here. Especially when using DMI_EXACT_MATCH that should not be a problem.

If however that causes a problem, then I prefer we actually set .driver_data in the ACPI module table and base the check on the .driver_data instead of checking the string here once more.

I really don't know if BCM2E39 is specific to a single design or manufacture or platform. I am not sure how good we are in not accidentally re-using these IDs.

> +	    dmi_check_system(bcm_wrong_irq_dmi_table)) {
> +		bt_dev_dbg(dev, "Fix irq polarity");
> +		dev->irq_polarity = !dev->irq_polarity;
> +	}
> +
> 	return 0;
> }
> #else

Regards

Marcel