[PATCH BlueZ] set: Abandon the use of the existing set gatt-db

[PATCH BlueZ] set: Abandon the use of the existing set gatt-db

[zhangchen200426]

From: Chen Zhang <zhangchen01@kylinos.cn>

When pairing and connecting a pair of earbuds, if you attempt to connect the
second earbud while one earbud is already connected, reusing the existing
set gatt-db will interrupt the GATT procedure. Subsequently, the attribute
values will no longer be read, and the ASE state will not switch. This will
eventually prevent the second earbud from creating a CIS, resulting in no sound
output from the later-connected earbud during music playback.

Signed-off-by: Zhang Chen <zhangchen01@kylinos.cn>
---
 src/set.c | 14 --------------
 1 file changed, 14 deletions(-)

diff --git a/src/set.c b/src/set.c
index 4ca2f78c3..0f2e9613c 100644
--- a/src/set.c
+++ b/src/set.c
@@ -282,20 +282,6 @@ static void foreach_rsi(void *data, void *user_data)
 	if (memcmp(ad->data, res, sizeof(res)))
 		return;
 
-	/* Attempt to use existing gatt_db from set if device has never been
-	 * connected before.
-	 *
-	 * If dbs don't really match bt_gatt_client will attempt to rediscover
-	 * the ranges that don't match.
-	 */
-	if (gatt_db_isempty(btd_device_get_gatt_db(set->device))) {
-		struct btd_device *device;
-
-		device = queue_get_entries(set->devices)->data;
-		btd_device_set_gatt_db(set->device,
-					btd_device_get_gatt_db(device));
-	}
-
 	device_connect_le(set->device);
 }
 
-- 
2.25.1

RE: [BlueZ] set: Abandon the use of the existing set gatt-db

[bluez.test.bot]

[-- Attachment #1: Type: text/plain, Size: 1781 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=1113188

---Test result---

Test Summary:
CheckPatch                    FAIL      0.43 seconds
GitLint                       PASS      0.33 seconds
BuildEll                      PASS      20.57 seconds
BluezMake                     PASS      667.03 seconds
CheckSmatch                   PASS      355.38 seconds
bluezmakeextell               PASS      184.89 seconds
IncrementalBuild              PASS      654.82 seconds
ScanBuild                     PASS      1045.56 seconds

Details
##############################
Test: CheckPatch - FAIL
Desc: Run checkpatch.pl script
Output:
[BlueZ] set: Abandon the use of the existing set gatt-db
WARNING:COMMIT_LOG_LONG_LINE: Possible unwrapped commit description (prefer a maximum 75 chars per line)
#65: 
When pairing and connecting a pair of earbuds, if you attempt to connect the

/github/workspace/src/patch/14634422.patch total: 0 errors, 1 warnings, 20 lines checked

NOTE: For some of the reported defects, checkpatch may be able to
      mechanically convert to the typical style using --fix or --fix-inplace.

/github/workspace/src/patch/14634422.patch has style problems, please review.

NOTE: Ignored message types: COMMIT_MESSAGE COMPLEX_MACRO CONST_STRUCT FILE_PATH_CHANGES MISSING_SIGN_OFF PREFER_PACKED SPDX_LICENSE_TAG SPLIT_STRING SSCANF_TO_KSTRTO

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.




https://github.com/bluez/bluez/pull/2238

---
Regards,
Linux Bluetooth

Re: [PATCH BlueZ] set: Abandon the use of the existing set gatt-db

[Luiz Augusto von Dentz]

Hi,

On Wed, Jun 17, 2026 at 10:53 PM <zhangchen200426@163.com> wrote:
>
> From: Chen Zhang <zhangchen01@kylinos.cn>
>
> When pairing and connecting a pair of earbuds, if you attempt to connect the
> second earbud while one earbud is already connected, reusing the existing
> set gatt-db will interrupt the GATT procedure. Subsequently, the attribute
> values will no longer be read, and the ASE state will not switch. This will
> eventually prevent the second earbud from creating a CIS, resulting in no sound
> output from the later-connected earbud during music playback.

Why would it prevent any operation of the ASE though? Perhaps we have
a different problem and this change is just a workaround. Anyway, the
intent is to prevent only the discovery procedure not ASE procedures.

> Signed-off-by: Zhang Chen <zhangchen01@kylinos.cn>
> ---
>  src/set.c | 14 --------------
>  1 file changed, 14 deletions(-)
>
> diff --git a/src/set.c b/src/set.c
> index 4ca2f78c3..0f2e9613c 100644
> --- a/src/set.c
> +++ b/src/set.c
> @@ -282,20 +282,6 @@ static void foreach_rsi(void *data, void *user_data)
>         if (memcmp(ad->data, res, sizeof(res)))
>                 return;
>
> -       /* Attempt to use existing gatt_db from set if device has never been
> -        * connected before.
> -        *
> -        * If dbs don't really match bt_gatt_client will attempt to rediscover
> -        * the ranges that don't match.
> -        */
> -       if (gatt_db_isempty(btd_device_get_gatt_db(set->device))) {
> -               struct btd_device *device;
> -
> -               device = queue_get_entries(set->devices)->data;
> -               btd_device_set_gatt_db(set->device,
> -                                       btd_device_get_gatt_db(device));
> -       }
> -
>         device_connect_le(set->device);
>  }
>
> --
> 2.25.1
>
>


-- 
Luiz Augusto von Dentz