Re-run ext adv after completion (IDFGH-9373)

[kevinhikaruevans]

Both extended and normal advertising should be reran when complete in this Nimble example. Without this, only a single BLE client is able to connect when CONFIG_EXAMPLE_EXTENDED_ADV=1

[rahult-github]

sha=35926387738ba28c0c84d29c6e803ed15c5a8ae7

[rahult-github]

Hi @kevinhikaruevans ,

I found some issue with the patch. If connection was success , for extended adv , adv complete event is posted by stack. If we restart the ext adv here, then when the connection is disconnected, it will again try to restart ext_adv as it will invoke ext_bleprph_advertise in original code. . This results in crash since the previous instance of ext adv was already started by the new function call introduced in this patch. This may need more handling , i can check later. Have you not observed any issue in your testing ?

`I (595) NimBLE_BLE_PRPH: BLE Host Task Started
I (605) NimBLE: Device Address:
I (605) NimBLE: 60:55:f9:f6:05:4a
I (615) NimBLE:

I (625) uart: queue free spaces: 8
I (635) main_task: Returned from app_main()
I (5075) NimBLE: connection established; status=0
I (5075) NimBLE: handle=1 our_ota_addr_type=0 our_ota_addr=
I (5085) NimBLE: 60:55:f9:f6:05:4a
I (5085) NimBLE: our_id_addr_type=0 our_id_addr=
I (5095) NimBLE: 60:55:f9:f6:05:4a
I (5095) NimBLE: peer_ota_addr_type=1 peer_ota_addr=
I (5105) NimBLE: 45:40:e1:10:0f:73
I (5105) NimBLE: peer_id_addr_type=1 peer_id_addr=
I (5115) NimBLE: 45:40:e1:10:0f:73
I (5115) NimBLE: conn_itvl=36 conn_latency=0 supervision_timeout=500 encrypted=0 authenticated=0 bonded=0

I (5125) NimBLE:

I (5135) NimBLE: advertise complete; reason=0
I (7035) NimBLE: encryption change event; status=1292
I (7035) NimBLE: handle=1 our_ota_addr_type=0 our_ota_addr=
I (7045) NimBLE: 60:55:f9:f6:05:4a
I (7045) NimBLE: our_id_addr_type=0 our_id_addr=
I (7045) NimBLE: 60:55:f9:f6:05:4a
I (7055) NimBLE: peer_ota_addr_type=1 peer_ota_addr=
I (7055) NimBLE: 45:40:e1:10:0f:73
I (7065) NimBLE: peer_id_addr_type=1 peer_id_addr=
I (7065) NimBLE: 45:40:e1:10:0f:73
I (7075) NimBLE: conn_itvl=36 conn_latency=0 supervision_timeout=500 encrypted=0 authenticated=0 bonded=0

I (7085) NimBLE:

I (10055) NimBLE: disconnect; reason=531
I (10055) NimBLE: handle=1 our_ota_addr_type=0 our_ota_addr=
I (10055) NimBLE: 60:55:f9:f6:05:4a
I (10055) NimBLE: our_id_addr_type=0 our_id_addr=
I (10065) NimBLE: 60:55:f9:f6:05:4a
I (10065) NimBLE: peer_ota_addr_type=1 peer_ota_addr=
I (10075) NimBLE: 45:40:e1:10:0f:73
I (10075) NimBLE: peer_id_addr_type=1 peer_id_addr=
I (10085) NimBLE: 45:40:e1:10:0f:73
I (10085) NimBLE: conn_itvl=36 conn_latency=0 supervision_timeout=500 encrypted=0 authenticated=0 bonded=0

I (10095) NimBLE:

assert failed: ext_bleprph_advertise main.c:116 (rc == 0)
Stack dump detected
Core 0 register dump:
MEPC : 0x408005fa RA : 0x4080919a SP : 0x4081b6c0 GP : 0x4080edf0
0x408005fa: panic_abort at /esp-idf/components/esp_system/panic.c:452

0x4080919a: __ubsan_include at /esp-idf/components/esp_system/ubsan.c:313

TP : 0x40804eec T0 : 0x37363534 T1 : 0x7271706f T2 : 0x33323130
0x40804eec: ble_lll_sched_insert_after at ??:?

S0/FP : 0x0000008e S1 : 0x00000001 A0 : 0x4081b6fc A1 : 0x4080f7d5
A2 : 0x00000001 A3 : 0x00000029 A4 : 0x00000001 A5 : 0x40811000
A6 : 0x00000000 A7 : 0x76757473 S2 : 0x00000009 S3 : 0x4081b809
S4 : 0x4080f7d4 S5 : 0x00000000 S6 : 0x00000000 S7 : 0x00000000
S8 : 0x00000000 S9 : 0x00000000 S10 : 0x00000000 S11 : 0x00000000
T3 : 0x6e6d6c6b T4 : 0x6a696867 T5 : 0x66656463 T6 : 0x62613938
MSTATUS : 0x00001881 MTVEC : 0x40800001 MCAUSE : 0x00000007 MTVAL : 0x00000000
0x40800001: _vector_table at ??:?

MHARTID : 0x00000000

Backtrace:

panic_abort (details=details@entry=0x4081b6fc "assert failed: ext_bleprph_advertise main.c:116 (rc == 0)") at /esp-idf/components/esp_system/panic.c:452
452 *((volatile int *) 0) = 0; // NOLINT(clang-analyzer-core.NullDereference) should be an invalid operation on targets
#0 panic_abort (details=details@entry=0x4081b6fc "assert failed: ext_bleprph_advertise main.c:116 (rc == 0)") at /esp-idf/components/esp_system/panic.c:452
#1 0x4080919a in esp_system_abort (details=details@entry=0x4081b6fc "assert failed: ext_bleprph_advertise main.c:116 (rc == 0)") at /esp-idf/components/esp_system/port/esp_system_chip.c:77
#2 0x4080d4e6 in __assert_func (file=file@entry=0x420635ed "", line=line@entry=116, func=, func@entry=0x42063be4 <func.2> "", expr=expr@entry=0x420635dc "") at /esp-idf/components/newlib/assert.c:81
#3 0x42006cb4 in ext_bleprph_advertise () at ../main/main.c:116
#4 0x42006f10 in bleprph_gap_event (event=0x4081b8a0, event@entry=, arg=) at ../main/main.c:308
`

[kevinhikaruevans]

Hi @rahult-github, I see the crash now. I was previously checking if the device was busy in ext_bleprph_advertise, but this is probably not a good way to handle this. This is what I was including in my test code:

    rc = ble_gap_ext_adv_configure(instance, &params, NULL,
                                   bleprph_gap_event, NULL);
    
    if (rc == BLE_HS_EBUSY) {
        return;
    }

Just wondering, are you able to run multiple ble clients with ext adv without the patch?

[rahult-github]

Hi @kevinhikaruevans ,

Please help check if attached patch works for you
0001-Nimble-Updated-bleprph-example-to-enable-ext-adv-aft.txt

[kevinhikaruevans]

@rahult-github Thank you! Yes, that patch worked for me. I've updated the PR with that patch.

[rahult-github]

sha=0b921fda149c69510258ba879e34cbeb1829cce0

[rahult-github]

Hi @kevinhikaruevans ,

I have below suggestions:

1. Can you please merge the two commits and make one. You can have the commit in your name as contributor.
2. Please rebase to latest master

[kevinhikaruevans]

Hi @rahult-github

1. I've squashed the two commits
2. I've rebased the branch with master

[rahult-github]

Changes part of master tree . Thanks for your contribution.