public inbox for linux-bluetooth@vger.kernel.org
 help / color / mirror / Atom feed
From: Robin Candau <antiz@archlinux.org>
To: Xiao Yao <xiaokeqinhealth@126.com>, linux-bluetooth@vger.kernel.org
Cc: Xiao Yao <xiaoyao@rock-chips.com>
Subject: Re: [PATCH 1/2] adapter: Fix addr_type for smp_irk/ltk_info/link_key
Date: Sat, 16 Dec 2023 01:59:51 +0100	[thread overview]
Message-ID: <55903fdb-e970-4b89-b620-daa93bad7811@archlinux.org> (raw)
In-Reply-To: <20231211162729.1183207-1-xiaokeqinhealth@126.com>


[-- Attachment #1.1.1.1: Type: text/plain, Size: 4200 bytes --]

On 12/11/23 17:27, Xiao Yao wrote:
> From: Xiao Yao<xiaoyao@rock-chips.com>
>
> According to BLUETOOTH CORE SPECIFICATION Version 5.4 | Vol 3,
> Part H, 2.4.24/2.4.25, The LE LTK and BR/EDR link keys can be
> converted between each other, so the addr type can be either
> BREDR or LE, so fix it.
>
> Signed-off-by: Xiao Yao<xiaoyao@rock-chips.com>
> ---
>   src/adapter.c | 20 ++++++++++++++------
>   1 file changed, 14 insertions(+), 6 deletions(-)
>
> diff --git a/src/adapter.c b/src/adapter.c
> index 86fff72bc..ee70b00d2 100644
> --- a/src/adapter.c
> +++ b/src/adapter.c
> @@ -170,6 +170,7 @@ static GSList *conn_fail_list = NULL;
>   
>   struct link_key_info {
>   	bdaddr_t bdaddr;
> +	uint8_t bdaddr_type;
>   	unsigned char key[16];
>   	uint8_t type;
>   	uint8_t pin_len;
> @@ -3964,7 +3965,9 @@ static bool is_blocked_key(uint8_t key_type, uint8_t *key_value)
>   	return false;
>   }
>   
> -static struct link_key_info *get_key_info(GKeyFile *key_file, const char *peer)
> +static struct link_key_info *get_key_info(GKeyFile *key_file, const char *peer,
> +					uint8_t bdaddr_type)
> +
>   {
>   	struct link_key_info *info = NULL;
>   	char *str;
> @@ -3976,6 +3979,7 @@ static struct link_key_info *get_key_info(GKeyFile *key_file, const char *peer)
>   	info = g_new0(struct link_key_info, 1);
>   
>   	str2ba(peer, &info->bdaddr);
> +	info->bdaddr_type = bdaddr_type;
>   
>   	if (!strncmp(str, "0x", 2))
>   		str2buf(&str[2], info->key, sizeof(info->key));
> @@ -4343,7 +4347,7 @@ static void load_link_keys(struct btd_adapter *adapter, GSList *keys,
>   		struct link_key_info *info = l->data;
>   
>   		bacpy(&key->addr.bdaddr, &info->bdaddr);
> -		key->addr.type = BDADDR_BREDR;
> +		key->addr.type = info->bdaddr_type;
>   		key->type = info->type;
>   		memcpy(key->val, info->key, 16);
>   		key->pin_len = info->pin_len;
> @@ -4598,14 +4602,18 @@ static void load_conn_params(struct btd_adapter *adapter, GSList *params)
>   		btd_error(adapter->dev_id, "Load connection parameters failed");
>   }
>   
> -static uint8_t get_le_addr_type(GKeyFile *keyfile)
> +static uint8_t get_addr_type(GKeyFile *keyfile)
>   {
>   	uint8_t addr_type;
>   	char *type;
>   
> +	/* The AddressType is written to file only When dev->le is
> +	 * set to true, as referenced in the update_technologies().
> +	 * Therefore, When type is NULL, it default to BDADDR_BREDR.
> +	 */
>   	type = g_key_file_get_string(keyfile, "General", "AddressType", NULL);
>   	if (!type)
> -		return BDADDR_LE_PUBLIC;
> +		return BDADDR_BREDR;
>   
>   	if (g_str_equal(type, "public"))
>   		addr_type = BDADDR_LE_PUBLIC;
> @@ -4914,9 +4922,9 @@ static void load_devices(struct btd_adapter *adapter)
>   			g_clear_error(&gerr);
>   		}
>   
> -		key_info = get_key_info(key_file, entry->d_name);
> +		bdaddr_type = get_addr_type(key_file);
>   
> -		bdaddr_type = get_le_addr_type(key_file);
> +		key_info = get_key_info(key_file, entry->d_name, bdaddr_type);
>   
>   		ltk_info = get_ltk_info(key_file, entry->d_name, bdaddr_type);
>   

Hello,

It seems like the above commit introduced a regression where the initial 
auto-connect for Bluetooth devices doesn't work anymore.

Indeed, at system startup, starting a Bluetooth device will cause it to 
go in a "connected/disconnected" state loop, requiring to initialize a 
manual connection first (with sometimes multiple attempts needed) before 
getting it connected correctly and working as intended.

`systemctl status bluetooth` reports the following error:

[...]
déc. 15 11:03:18 Arch-Desktop bluetoothd[592]: Failed to load link keys 
for hci0: Invalid Parameters (0x0d)
[...]

**I bisected the bug with `git bisect` and it pointed out this commit 
[1] as the "faulty" one.
I can confirm that reverting it solves the issue.

I reported this bug including details in the GitHub repo [2].

I remain available if any additional information are needed.

[1] 
https://github.com/bluez/bluez/commit/d5536e0cd431e22be9a1132be6d4af2445590633
[2] https://github.com/bluez/bluez/issues/686|
|

-- 
Regards,
Robin Candau / Antiz


[-- Attachment #1.1.1.2: Type: text/html, Size: 4984 bytes --]

[-- Attachment #1.1.2: OpenPGP public key --]
[-- Type: application/pgp-keys, Size: 3203 bytes --]

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 840 bytes --]

  parent reply	other threads:[~2023-12-16  1:06 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-12-11 16:27 [PATCH 1/2] adapter: Fix addr_type for smp_irk/ltk_info/link_key Xiao Yao
2023-12-11 16:27 ` [PATCH 2/2] device: Add notifications when the bdaddr_type changes Xiao Yao
2023-12-11 18:17 ` [1/2] adapter: Fix addr_type for smp_irk/ltk_info/link_key bluez.test.bot
2023-12-11 18:40 ` [PATCH 1/2] " patchwork-bot+bluetooth
2023-12-16  0:59 ` Robin Candau [this message]
2024-08-26 21:12   ` Luiz Augusto von Dentz
2024-08-27  7:39     ` Yao Xiao
2024-08-27 12:52       ` Luiz Augusto von Dentz
2024-08-27 14:33         ` Yao Xiao
2024-08-27 15:47           ` Luiz Augusto von Dentz
2024-08-27 18:01             ` Yao Xiao
2024-08-27 18:30               ` Luiz Augusto von Dentz
2024-08-28 10:01                 ` Yao Xiao
2024-08-28 14:37                   ` Luiz Augusto von Dentz
2024-08-29 13:27                     ` Yao Xiao

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=55903fdb-e970-4b89-b620-daa93bad7811@archlinux.org \
    --to=antiz@archlinux.org \
    --cc=linux-bluetooth@vger.kernel.org \
    --cc=xiaokeqinhealth@126.com \
    --cc=xiaoyao@rock-chips.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox