[PATCH BlueZ] hog-lib: Fix HoG discovering characteristics order

[PATCH BlueZ] hog-lib: Fix HoG discovering characteristics order

[Frédéric Danis]

PTS test HOGP/RH/HGRF/BV-01-I fails as inconclusive after waiting for
second report map read request with message 'Please send ReadRequest to
read Report Map characteristic with handle = 0x00E7'.
Handle which has already been read, as it can be seen in btmon traces.

This commit change the recursion order to read instance characteristics
before recurring to next instances.

Closes: https://github.com/bluez/bluez/issues/657
---
 profiles/input/hog-lib.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/profiles/input/hog-lib.c b/profiles/input/hog-lib.c
index 7ff1ede3d..db62a3c9f 100644
--- a/profiles/input/hog-lib.c
+++ b/profiles/input/hog-lib.c
@@ -1780,12 +1780,6 @@ bool bt_hog_attach(struct bt_hog *hog, void *gatt)
 
 	queue_foreach(hog->bas, (void *) bt_bas_attach, gatt);
 
-	for (l = hog->instances; l; l = l->next) {
-		struct bt_hog *instance = l->data;
-
-		bt_hog_attach(instance, gatt);
-	}
-
 	if (!hog->uhid_created) {
 		DBG("HoG discovering characteristics");
 		if (hog->attr)
@@ -1798,6 +1792,12 @@ bool bt_hog_attach(struct bt_hog *hog, void *gatt)
 					char_discovered_cb, hog);
 	}
 
+	for (l = hog->instances; l; l = l->next) {
+		struct bt_hog *instance = l->data;
+
+		bt_hog_attach(instance, gatt);
+	}
+
 	if (!hog->uhid_created)
 		return true;
 
-- 
2.34.1

RE: [BlueZ] hog-lib: Fix HoG discovering characteristics order

[bluez.test.bot]

[-- Attachment #1: Type: text/plain, Size: 946 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=809574

---Test result---

Test Summary:
CheckPatch                    PASS      0.28 seconds
GitLint                       PASS      0.21 seconds
BuildEll                      PASS      24.29 seconds
BluezMake                     PASS      717.00 seconds
MakeCheck                     PASS      11.80 seconds
MakeDistcheck                 PASS      157.76 seconds
CheckValgrind                 PASS      215.39 seconds
CheckSmatch                   PASS      324.50 seconds
bluezmakeextell               PASS      102.73 seconds
IncrementalBuild              PASS      672.22 seconds
ScanBuild                     PASS      912.77 seconds



---
Regards,
Linux Bluetooth

Re: [PATCH BlueZ] hog-lib: Fix HoG discovering characteristics order

[Luiz Augusto von Dentz]

Hi Frédéric,

On Wed, Dec 13, 2023 at 3:50 AM Frédéric Danis
<frederic.danis@collabora.com> wrote:
>
> PTS test HOGP/RH/HGRF/BV-01-I fails as inconclusive after waiting for
> second report map read request with message 'Please send ReadRequest to
> read Report Map characteristic with handle = 0x00E7'.
> Handle which has already been read, as it can be seen in btmon traces.
>
> This commit change the recursion order to read instance characteristics
> before recurring to next instances.
>
> Closes: https://github.com/bluez/bluez/issues/657
> ---
>  profiles/input/hog-lib.c | 12 ++++++------
>  1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/profiles/input/hog-lib.c b/profiles/input/hog-lib.c
> index 7ff1ede3d..db62a3c9f 100644
> --- a/profiles/input/hog-lib.c
> +++ b/profiles/input/hog-lib.c
> @@ -1780,12 +1780,6 @@ bool bt_hog_attach(struct bt_hog *hog, void *gatt)
>
>         queue_foreach(hog->bas, (void *) bt_bas_attach, gatt);
>
> -       for (l = hog->instances; l; l = l->next) {
> -               struct bt_hog *instance = l->data;
> -
> -               bt_hog_attach(instance, gatt);
> -       }
> -
>         if (!hog->uhid_created) {
>                 DBG("HoG discovering characteristics");
>                 if (hog->attr)
> @@ -1798,6 +1792,12 @@ bool bt_hog_attach(struct bt_hog *hog, void *gatt)
>                                         char_discovered_cb, hog);
>         }
>
> +       for (l = hog->instances; l; l = l->next) {
> +               struct bt_hog *instance = l->data;
> +
> +               bt_hog_attach(instance, gatt);
> +       }
> +
>         if (!hog->uhid_created)
>                 return true;

While I don't oppose this change, but this seems to be a limitation on
the PTS side which expects a certain order of the operation which
doesn't seem to be a requirement coming from the spec, so perhaps we
should file an issue to PTS.

> --
> 2.34.1
>
>


-- 
Luiz Augusto von Dentz

Re: [PATCH BlueZ] hog-lib: Fix HoG discovering characteristics order

[Frédéric Danis]

Hi Luiz,

On 13/12/2023 18:27, Luiz Augusto von Dentz wrote:
> Hi Frédéric,
>
> On Wed, Dec 13, 2023 at 3:50 AM Frédéric Danis
> <frederic.danis@collabora.com> wrote:
>> PTS test HOGP/RH/HGRF/BV-01-I fails as inconclusive after waiting for
>> second report map read request with message 'Please send ReadRequest to
>> read Report Map characteristic with handle = 0x00E7'.
>> Handle which has already been read, as it can be seen in btmon traces.
>>
>> This commit change the recursion order to read instance characteristics
>> before recurring to next instances.
>>
>> Closes: https://github.com/bluez/bluez/issues/657
>> ---
>>   profiles/input/hog-lib.c | 12 ++++++------
>>   1 file changed, 6 insertions(+), 6 deletions(-)
>>
>> diff --git a/profiles/input/hog-lib.c b/profiles/input/hog-lib.c
>> index 7ff1ede3d..db62a3c9f 100644
>> --- a/profiles/input/hog-lib.c
>> +++ b/profiles/input/hog-lib.c
>> @@ -1780,12 +1780,6 @@ bool bt_hog_attach(struct bt_hog *hog, void *gatt)
>>
>>          queue_foreach(hog->bas, (void *) bt_bas_attach, gatt);
>>
>> -       for (l = hog->instances; l; l = l->next) {
>> -               struct bt_hog *instance = l->data;
>> -
>> -               bt_hog_attach(instance, gatt);
>> -       }
>> -
>>          if (!hog->uhid_created) {
>>                  DBG("HoG discovering characteristics");
>>                  if (hog->attr)
>> @@ -1798,6 +1792,12 @@ bool bt_hog_attach(struct bt_hog *hog, void *gatt)
>>                                          char_discovered_cb, hog);
>>          }
>>
>> +       for (l = hog->instances; l; l = l->next) {
>> +               struct bt_hog *instance = l->data;
>> +
>> +               bt_hog_attach(instance, gatt);
>> +       }
>> +
>>          if (!hog->uhid_created)
>>                  return true;
> While I don't oppose this change, but this seems to be a limitation on
> the PTS side which expects a certain order of the operation which
> doesn't seem to be a requirement coming from the spec, so perhaps we
> should file an issue to PTS.

Yes, I opened an issue to PTS.
Let's see what they will reply.

Regards,

Fred

-- 

Frédéric Danis
Senior Software Engineer

Collabora Ltd.
Platinum Building, St John's Innovation Park, Cambridge CB4 0DS, United Kingdom
Registered in England & Wales, no. 5513718