[PATCH v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

[PATCH v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

[Nobuaki Tsunashima]

From: Nobuaki Tsunashima <Nobuaki.Tsunashima@infineon.com>

CYW4373 ROM FW has an issue that it claims LE_Read_Transmit_Power command
as supported in a response of Read_Local_Supported_Command command but
rejects the LE_Read_Transmit_Power command with "Unknown HCI Command"
status. Because Bluetooth driver of kernel 5.11 added sending the
LE_Read_Transmit_Power command in initialize phase, hci up fails due to the
issue.

Especially in USB i/f case, it would be difficult to download patch FW that
includes its fix unless hci is up.

The driver already contains infrastructure to apply the quirk for the
issue, but currently it only supports DMI based matching. Add support to
match by chip id and baseline FW version to detect CYW4373 ROM FW build
in generic system.

Fixes: 7c395ea521e6 ("Bluetooth: Query LE tx power on startup")
Signed-off-by: Nobuaki Tsunashima <Nobuaki.Tsunashima@infineon.com>
---
V3 -> V4: Fix a few coding style warnings and refine comments for clarify.
V2 -> V3: Fix a few coding style warnings and change the subject as more specific.
V1 -> V2: Fix several coding style warnings.

 drivers/bluetooth/btbcm.c | 32 +++++++++++++++++++++++++++++++-
 drivers/bluetooth/btusb.c |  4 ++++
 2 files changed, 35 insertions(+), 1 deletion(-)

diff --git a/drivers/bluetooth/btbcm.c b/drivers/bluetooth/btbcm.c
index 0a5445ac5e1b..29e3f83a19fa 100644
--- a/drivers/bluetooth/btbcm.c
+++ b/drivers/bluetooth/btbcm.c
@@ -437,18 +437,48 @@ static const struct dmi_system_id disable_broken_read_transmit_power[] = {
 	{ }
 };
 
+struct bcm_chip_version_table {
+	u8 chip_id;			/* Chip ID */
+	u16 baseline;		/* Baseline version of patch FW */
+};
+#define BCM_ROMFW_BASELINE_NUM	0xFFFF
+static const struct bcm_chip_version_table disable_broken_read_transmit_power_by_chip_ver[] = {
+	{ 0x87, BCM_ROMFW_BASELINE_NUM }		/* CYW4373/4373E */
+};
+static bool btbcm_is_disable_broken_read_tx_power_by_chip_ver(u8 chip_id, u16 baseline)
+{
+	int i;
+	size_t table_size = ARRAY_SIZE(disable_broken_read_transmit_power_by_chip_ver);
+	const struct bcm_chip_version_table *entry =
+						&disable_broken_read_transmit_power_by_chip_ver[0];
+
+	for (i = 0 ; i < table_size ; i++, entry++)	{
+		if ((chip_id == entry->chip_id) && (baseline == entry->baseline))
+			return true;
+	}
+
+	return false;
+}
+
 static int btbcm_read_info(struct hci_dev *hdev)
 {
 	struct sk_buff *skb;
+	u8 chip_id;
+	u16 baseline;
 
 	/* Read Verbose Config Version Info */
 	skb = btbcm_read_verbose_config(hdev);
 	if (IS_ERR(skb))
 		return PTR_ERR(skb);
-
+	chip_id = skb->data[1];
+	baseline = skb->data[3] | (skb->data[4] << 8);
 	bt_dev_info(hdev, "BCM: chip id %u", skb->data[1]);
 	kfree_skb(skb);
 
+	/* Check Chip ID and disable broken Read LE Min/Max Tx Power */
+	if (btbcm_is_disable_broken_read_tx_power_by_chip_ver(chip_id, baseline))
+		set_bit(HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER, &hdev->quirks);
+
 	return 0;
 }
 
diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
index d31edad7a056..52561c8d8828 100644
--- a/drivers/bluetooth/btusb.c
+++ b/drivers/bluetooth/btusb.c
@@ -142,6 +142,10 @@ static const struct usb_device_id btusb_table[] = {
 	{ USB_VENDOR_AND_INTERFACE_INFO(0x04ca, 0xff, 0x01, 0x01),
 	  .driver_info = BTUSB_BCM_PATCHRAM },
 
+	/* Cypress devices with vendor specific id */
+	{ USB_VENDOR_AND_INTERFACE_INFO(0x04b4, 0xff, 0x01, 0x01),
+	  .driver_info = BTUSB_BCM_PATCHRAM },
+
 	/* Broadcom devices with vendor specific id */
 	{ USB_VENDOR_AND_INTERFACE_INFO(0x0a5c, 0xff, 0x01, 0x01),
 	  .driver_info = BTUSB_BCM_PATCHRAM },
-- 
2.25.1

RE: [v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

[bluez.test.bot]

[-- Attachment #1: Type: text/plain, Size: 2953 bytes --]

This is automated email and please do not reply to this email!

Dear submitter,

Thank you for submitting the patches to the linux bluetooth mailing list.
This is a CI test results with your patch series:
PW Link:https://patchwork.kernel.org/project/bluetooth/list/?series=855541

---Test result---

Test Summary:
CheckPatch                    PASS      0.70 seconds
GitLint                       FAIL      0.45 seconds
SubjectPrefix                 PASS      0.06 seconds
BuildKernel                   PASS      30.18 seconds
CheckAllWarning               PASS      32.50 seconds
CheckSparse                   PASS      38.18 seconds
CheckSmatch                   FAIL      35.54 seconds
BuildKernel32                 PASS      29.01 seconds
TestRunnerSetup               PASS      517.53 seconds
TestRunner_l2cap-tester       PASS      20.46 seconds
TestRunner_iso-tester         PASS      28.92 seconds
TestRunner_bnep-tester        PASS      4.89 seconds
TestRunner_mgmt-tester        PASS      109.37 seconds
TestRunner_rfcomm-tester      PASS      7.40 seconds
TestRunner_sco-tester         PASS      14.96 seconds
TestRunner_ioctl-tester       PASS      7.81 seconds
TestRunner_mesh-tester        PASS      5.87 seconds
TestRunner_smp-tester         PASS      6.87 seconds
TestRunner_userchan-tester    PASS      5.02 seconds
IncrementalBuild              PASS      27.34 seconds

Details
##############################
Test: GitLint - FAIL
Desc: Run gitlint
Output:
[v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

WARNING: I3 - ignore-body-lines: gitlint will be switching from using Python regex 'match' (match beginning) to 'search' (match anywhere) semantics. Please review your ignore-body-lines.regex option accordingly. To remove this warning, set general.regex-style-search=True. More details: https://jorisroovers.github.io/gitlint/configuration/#regex-style-search
23: B1 Line exceeds max length (82>80): "V2 -> V3: Fix a few coding style warnings and change the subject as more specific."
##############################
Test: CheckSmatch - FAIL
Desc: Run smatch tool with source
Output:

Segmentation fault (core dumped)
make[4]: *** [scripts/Makefile.build:244: net/bluetooth/hci_core.o] Error 139
make[4]: *** Deleting file 'net/bluetooth/hci_core.o'
make[3]: *** [scripts/Makefile.build:485: net/bluetooth] Error 2
make[2]: *** [scripts/Makefile.build:485: net] Error 2
make[2]: *** Waiting for unfinished jobs....
Segmentation fault (core dumped)
make[4]: *** [scripts/Makefile.build:244: drivers/bluetooth/bcm203x.o] Error 139
make[4]: *** Deleting file 'drivers/bluetooth/bcm203x.o'
make[4]: *** Waiting for unfinished jobs....
make[3]: *** [scripts/Makefile.build:485: drivers/bluetooth] Error 2
make[2]: *** [scripts/Makefile.build:485: drivers] Error 2
make[1]: *** [/github/workspace/src/src/Makefile:1919: .] Error 2
make: *** [Makefile:240: __sub-make] Error 2


---
Regards,
Linux Bluetooth

Re: [PATCH v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

[Luiz Augusto von Dentz]

Hi Nobuaki,

On Thu, May 23, 2024 at 9:31 PM Nobuaki Tsunashima
<nobuaki.tsunashima@infineon.com> wrote:
>
> From: Nobuaki Tsunashima <Nobuaki.Tsunashima@infineon.com>
>
> CYW4373 ROM FW has an issue that it claims LE_Read_Transmit_Power command
> as supported in a response of Read_Local_Supported_Command command but
> rejects the LE_Read_Transmit_Power command with "Unknown HCI Command"
> status. Because Bluetooth driver of kernel 5.11 added sending the
> LE_Read_Transmit_Power command in initialize phase, hci up fails due to the
> issue.
>
> Especially in USB i/f case, it would be difficult to download patch FW that
> includes its fix unless hci is up.
>
> The driver already contains infrastructure to apply the quirk for the
> issue, but currently it only supports DMI based matching. Add support to
> match by chip id and baseline FW version to detect CYW4373 ROM FW build
> in generic system.
>
> Fixes: 7c395ea521e6 ("Bluetooth: Query LE tx power on startup")
> Signed-off-by: Nobuaki Tsunashima <Nobuaki.Tsunashima@infineon.com>
> ---
> V3 -> V4: Fix a few coding style warnings and refine comments for clarify.
> V2 -> V3: Fix a few coding style warnings and change the subject as more specific.
> V1 -> V2: Fix several coding style warnings.
>
>  drivers/bluetooth/btbcm.c | 32 +++++++++++++++++++++++++++++++-
>  drivers/bluetooth/btusb.c |  4 ++++
>  2 files changed, 35 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/bluetooth/btbcm.c b/drivers/bluetooth/btbcm.c
> index 0a5445ac5e1b..29e3f83a19fa 100644
> --- a/drivers/bluetooth/btbcm.c
> +++ b/drivers/bluetooth/btbcm.c
> @@ -437,18 +437,48 @@ static const struct dmi_system_id disable_broken_read_transmit_power[] = {
>         { }
>  };
>
> +struct bcm_chip_version_table {
> +       u8 chip_id;                     /* Chip ID */
> +       u16 baseline;           /* Baseline version of patch FW */
> +};
> +#define BCM_ROMFW_BASELINE_NUM 0xFFFF
> +static const struct bcm_chip_version_table disable_broken_read_transmit_power_by_chip_ver[] = {
> +       { 0x87, BCM_ROMFW_BASELINE_NUM }                /* CYW4373/4373E */
> +};
> +static bool btbcm_is_disable_broken_read_tx_power_by_chip_ver(u8 chip_id, u16 baseline)
> +{
> +       int i;
> +       size_t table_size = ARRAY_SIZE(disable_broken_read_transmit_power_by_chip_ver);
> +       const struct bcm_chip_version_table *entry =
> +                                               &disable_broken_read_transmit_power_by_chip_ver[0];
> +
> +       for (i = 0 ; i < table_size ; i++, entry++)     {
> +               if ((chip_id == entry->chip_id) && (baseline == entry->baseline))
> +                       return true;
> +       }
> +
> +       return false;
> +}
> +
>  static int btbcm_read_info(struct hci_dev *hdev)
>  {
>         struct sk_buff *skb;
> +       u8 chip_id;
> +       u16 baseline;
>
>         /* Read Verbose Config Version Info */
>         skb = btbcm_read_verbose_config(hdev);
>         if (IS_ERR(skb))
>                 return PTR_ERR(skb);
> -
> +       chip_id = skb->data[1];
> +       baseline = skb->data[3] | (skb->data[4] << 8);

This is not really safe, you shouldn't attempt to access skb->data
without first checking skb->len, actually it would be much better that
you would use skb_pull_data which does skb->len check before pulling
data.

>         bt_dev_info(hdev, "BCM: chip id %u", skb->data[1]);
>         kfree_skb(skb);
>
> +       /* Check Chip ID and disable broken Read LE Min/Max Tx Power */
> +       if (btbcm_is_disable_broken_read_tx_power_by_chip_ver(chip_id, baseline))
> +               set_bit(HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER, &hdev->quirks);
> +
>         return 0;
>  }
>
> diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
> index d31edad7a056..52561c8d8828 100644
> --- a/drivers/bluetooth/btusb.c
> +++ b/drivers/bluetooth/btusb.c
> @@ -142,6 +142,10 @@ static const struct usb_device_id btusb_table[] = {
>         { USB_VENDOR_AND_INTERFACE_INFO(0x04ca, 0xff, 0x01, 0x01),
>           .driver_info = BTUSB_BCM_PATCHRAM },
>
> +       /* Cypress devices with vendor specific id */
> +       { USB_VENDOR_AND_INTERFACE_INFO(0x04b4, 0xff, 0x01, 0x01),
> +         .driver_info = BTUSB_BCM_PATCHRAM },
> +
>         /* Broadcom devices with vendor specific id */
>         { USB_VENDOR_AND_INTERFACE_INFO(0x0a5c, 0xff, 0x01, 0x01),
>           .driver_info = BTUSB_BCM_PATCHRAM },
> --
> 2.25.1
>


-- 
Luiz Augusto von Dentz

RE: [PATCH v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

[Nobuaki.Tsunashima]

Hi Luiz,

Thanks for your review.

>>  static int btbcm_read_info(struct hci_dev *hdev)  {
>>         struct sk_buff *skb;
>> +       u8 chip_id;
>> +       u16 baseline;
>>
>>         /* Read Verbose Config Version Info */
>>         skb = btbcm_read_verbose_config(hdev);
>>         if (IS_ERR(skb))
>>                 return PTR_ERR(skb);
>> -
>> +       chip_id = skb->data[1];
>> +       baseline = skb->data[3] | (skb->data[4] << 8);
>
>This is not really safe, you shouldn't attempt to access skb->data without first checking skb->len, actually it would be much better that >you would use skb_pull_data which does skb->len check before pulling data.

I think it could be safe because its length is checked inside btbcm_read_verbose_config() as below.
Please let me know if further checking is needed.

>>>
static struct sk_buff *btbcm_read_verbose_config(struct hci_dev *hdev)
{
	struct sk_buff *skb;

	skb = __hci_cmd_sync(hdev, 0xfc79, 0, NULL, HCI_INIT_TIMEOUT);
	if (IS_ERR(skb)) {
		bt_dev_err(hdev, "BCM: Read verbose config info failed (%ld)",
			   PTR_ERR(skb));
		return skb;
	}

	if (skb->len != 7) {
		bt_dev_err(hdev, "BCM: Verbose config length mismatch");
		kfree_skb(skb);
		return ERR_PTR(-EIO);
	}

	return skb;
}
<<<

Best Regards,
Nobuaki Tsunashima

RE: [PATCH v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

[Nobuaki.Tsunashima]

Hello,

Please let me know if there are any additional actions needed from my end to facilitate progress on this patch.
Your feedback would be greatly appreciated.

Thank you for your attention to this matter.

Best Regards,
Nobuaki Tsunashima

-----Original Message-----
From: Tsunashima Nobuaki (SMD C3 JP RM WLS AE) 
Sent: Monday, May 27, 2024 10:59 AM
To: 'Luiz Augusto von Dentz' <luiz.dentz@gmail.com>
Cc: Marcel Holtmann <marcel@holtmann.org>; linux-bluetooth@vger.kernel.org; linux-kernel@vger.kernel.org
Subject: RE: [PATCH v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

Hi Luiz,

Thanks for your review.

>>  static int btbcm_read_info(struct hci_dev *hdev)  {
>>         struct sk_buff *skb;
>> +       u8 chip_id;
>> +       u16 baseline;
>>
>>         /* Read Verbose Config Version Info */
>>         skb = btbcm_read_verbose_config(hdev);
>>         if (IS_ERR(skb))
>>                 return PTR_ERR(skb);
>> -
>> +       chip_id = skb->data[1];
>> +       baseline = skb->data[3] | (skb->data[4] << 8);
>
>This is not really safe, you shouldn't attempt to access skb->data without first checking skb->len, actually it would be much better that >you would use skb_pull_data which does skb->len check before pulling data.

I think it could be safe because its length is checked inside btbcm_read_verbose_config() as below.
Please let me know if further checking is needed.

>>>
static struct sk_buff *btbcm_read_verbose_config(struct hci_dev *hdev) {
	struct sk_buff *skb;

	skb = __hci_cmd_sync(hdev, 0xfc79, 0, NULL, HCI_INIT_TIMEOUT);
	if (IS_ERR(skb)) {
		bt_dev_err(hdev, "BCM: Read verbose config info failed (%ld)",
			   PTR_ERR(skb));
		return skb;
	}

	if (skb->len != 7) {
		bt_dev_err(hdev, "BCM: Verbose config length mismatch");
		kfree_skb(skb);
		return ERR_PTR(-EIO);
	}

	return skb;
}
<<<

Best Regards,
Nobuaki Tsunashima

Re: [PATCH v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

[Luiz Augusto von Dentz]

Hi,

On Sun, May 26, 2024 at 9:59 PM <Nobuaki.Tsunashima@infineon.com> wrote:
>
> Hi Luiz,
>
> Thanks for your review.
>
> >>  static int btbcm_read_info(struct hci_dev *hdev)  {
> >>         struct sk_buff *skb;
> >> +       u8 chip_id;
> >> +       u16 baseline;
> >>
> >>         /* Read Verbose Config Version Info */
> >>         skb = btbcm_read_verbose_config(hdev);
> >>         if (IS_ERR(skb))
> >>                 return PTR_ERR(skb);
> >> -
> >> +       chip_id = skb->data[1];
> >> +       baseline = skb->data[3] | (skb->data[4] << 8);
> >
> >This is not really safe, you shouldn't attempt to access skb->data without first checking skb->len, actually it would be much better that >you would use skb_pull_data which does skb->len check before pulling data.
>
> I think it could be safe because its length is checked inside btbcm_read_verbose_config() as below.
> Please let me know if further checking is needed.
>
> >>>
> static struct sk_buff *btbcm_read_verbose_config(struct hci_dev *hdev)
> {
>         struct sk_buff *skb;
>
>         skb = __hci_cmd_sync(hdev, 0xfc79, 0, NULL, HCI_INIT_TIMEOUT);
>         if (IS_ERR(skb)) {
>                 bt_dev_err(hdev, "BCM: Read verbose config info failed (%ld)",
>                            PTR_ERR(skb));
>                 return skb;
>         }
>
>         if (skb->len != 7) {
>                 bt_dev_err(hdev, "BCM: Verbose config length mismatch");
>                 kfree_skb(skb);
>                 return ERR_PTR(-EIO);
>         }
>
>         return skb;
> }
> <<<

Ok, but I still consider reworking these to use skb_pull_data.

> Best Regards,
> Nobuaki Tsunashima
>


-- 
Luiz Augusto von Dentz

RE: [PATCH v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

[Nobuaki.Tsunashima]

Hi Luiz,

Thanks for your comment.

> Ok, but I still consider reworking these to use skb_pull_data.

Now, I reconsider and found the skb_pull_data is more convenient rather than directly accessing to skb->data.
As I am on business trip on a few days, I will submit new patch after I come back.

Regards,
Nobuaki Tsunashima

-----Original Message-----
From: Luiz Augusto von Dentz <luiz.dentz@gmail.com> 
Sent: Friday, June 28, 2024 10:29 PM
To: Tsunashima Nobuaki (SMD C3 JP RM WLS AE) <Nobuaki.Tsunashima@infineon.com>
Cc: marcel@holtmann.org; linux-bluetooth@vger.kernel.org; linux-kernel@vger.kernel.org
Subject: Re: [PATCH v4] Bluetooth: btbcm: Apply HCI_QUIRK_BROKEN_READ_TRANSMIT_POWER to CYW4373

Caution: This e-mail originated outside Infineon Technologies. Please be cautious when sharing information or opening attachments especially from unknown senders. Refer to our intranet guide<https://intranet-content.infineon.com/explore/aboutinfineon/rules/informationsecurity/ug/SocialEngineering/Pages/SocialEngineeringElements_en.aspx> to help you identify Phishing email.



Hi,

On Sun, May 26, 2024 at 9:59 PM <Nobuaki.Tsunashima@infineon.com> wrote:
>
> Hi Luiz,
>
> Thanks for your review.
>
> >>  static int btbcm_read_info(struct hci_dev *hdev)  {
> >>         struct sk_buff *skb;
> >> +       u8 chip_id;
> >> +       u16 baseline;
> >>
> >>         /* Read Verbose Config Version Info */
> >>         skb = btbcm_read_verbose_config(hdev);
> >>         if (IS_ERR(skb))
> >>                 return PTR_ERR(skb);
> >> -
> >> +       chip_id = skb->data[1];
> >> +       baseline = skb->data[3] | (skb->data[4] << 8);
> >
> >This is not really safe, you shouldn't attempt to access skb->data without first checking skb->len, actually it would be much better that >you would use skb_pull_data which does skb->len check before pulling data.
>
> I think it could be safe because its length is checked inside btbcm_read_verbose_config() as below.
> Please let me know if further checking is needed.
>
> >>>
> static struct sk_buff *btbcm_read_verbose_config(struct hci_dev *hdev) 
> {
>         struct sk_buff *skb;
>
>         skb = __hci_cmd_sync(hdev, 0xfc79, 0, NULL, HCI_INIT_TIMEOUT);
>         if (IS_ERR(skb)) {
>                 bt_dev_err(hdev, "BCM: Read verbose config info failed (%ld)",
>                            PTR_ERR(skb));
>                 return skb;
>         }
>
>         if (skb->len != 7) {
>                 bt_dev_err(hdev, "BCM: Verbose config length mismatch");
>                 kfree_skb(skb);
>                 return ERR_PTR(-EIO);
>         }
>
>         return skb;
> }
> <<<

Ok, but I still consider reworking these to use skb_pull_data.

> Best Regards,
> Nobuaki Tsunashima
>


--
Luiz Augusto von Dentz