LE LL Priavcy support enabled
Subsystems (2)
cover LE LL Priavcy support enabled
With this LE LL Priavcy feature available if BT controller supports LL Privacy when privacy is enabled.
The patches are verified in below case.
Pairing - where the IRK is added to device Resolving List after identity info & IRK is shared in SMP.
Unpair - Delete the resolving entry from the Device resolving list Local IRK change - Whenever there is change in local IRK (privacy) the resolving list entries in the device is updated.
Whitelist - filter policies for whitelist is updated based on LL_adress resolution support.
Reload Resolving list - device reset/sysmte reboot the Resolving list reloaded in to the device.
Sathish Narsimman (7):
Bluetooth: LL Privacy Delete Store Resolving list
Bluetooth: Add device to Resolving List
Bluetooth: Update the Resolving list when Local IRK changed
Bluetooth: LL_PRIVACY re-load resolving list
Bluetooth: LL Privacy Check and Update the Resolving List
Bluetooth: Enable LL Privacy
Bluetooth LL Privacy update White list
include/net/bluetooth/hci.h | 3 + include/net/bluetooth/hci_core.h | 7 + net/bluetooth/hci_conn.c | 32 ++++ net/bluetooth/hci_core.c | 20 ++- net/bluetooth/hci_event.c | 32 +++- net/bluetooth/hci_request.c | 251 ++++++++++++++++++++++++++++++- net/bluetooth/hci_request.h | 2 + net/bluetooth/mgmt.c | 10 +- net/bluetooth/smp.c | 11 ++ 9 files changed, 354 insertions(+), 14 deletions(-)
Signature · 1 line
2.17.1
PATCH 1/7 Bluetooth: LL Privacy Delete Store Resolving list
This patch help to delte the resolving list stored in the BT Controller w.r.t BD_ADDR.
--- include/net/bluetooth/hci.h | 1 + include/net/bluetooth/hci_core.h | 2 ++ net/bluetooth/hci_request.c | 33 ++++++++++++++++++++++++++++++++ 3 files changed, 36 insertions(+)
include/net/bluetooth/hci.h +1 -0
@@ -457,6 +457,7 @@ enum {#define HCI_LE_DATA_LEN_EXT 0x20#define HCI_LE_PHY_2M 0x01#define HCI_LE_PHY_CODED 0x08 +#define HCI_LE_LL_PRIVACY 0x40#define HCI_LE_EXT_ADV 0x10#define HCI_LE_EXT_SCAN_POLICY 0x80#define HCI_LE_PHY_2M 0x01
include/net/bluetooth/hci_core.h +2 -0
@@ -1657,6 +1657,8 @@ void hci_le_start_enc(struct hci_conn *conn, __le16 ediv, __le64 rand,void hci_copy_identity_address (struct hci_dev *hdev,bdaddr_t *bdaddr,u8 *bdaddr_type); +void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type, +bdaddr_t *bdaddr);#define SCO_AIRMODE_MASK 0x0003#define SCO_AIRMODE_CVSD 0x0000
net/bluetooth/hci_request.c +33 -0
@@ -885,6 +885,39 @@ static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval, } } +void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr) +{ +struct hci_cp_le_del_from_resolv_list cp; +struct hci_request req; +struct bdaddr_list_with_irk *irk; + +BT_DBG ("" ); + +/* Nothing to be done if LL privacy is not supported */ +if ( !(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) ) +return ; + +if ( !hci_dev_test_flag (hdev,HCI_LL_RPA_RESOLUTION ) && + (hci_dev_test_flag (hdev,HCI_LE_ADV ) || +hci_dev_test_flag (hdev,HCI_LE_SCAN ) || +hci_lookup_le_connect (hdev) ) ) +return ; + +/* If the device is not present in resolving list +* no need to send HCI command. +*/ + irk =hci_bdaddr_list_lookup_with_irk (&hdev->le_resolv_list, bdaddr, addr_type); +if (!irk) +return ; + +hci_req_init (&req, hdev); + cp.bdaddr_type = addr_type; +bacpy (&cp.bdaddr, bdaddr); + +hci_req_add (&req,HCI_OP_LE_DEL_FROM_RESOLV_LIST ,sizeof (cp), &cp); +hci_req_run (&req,NULL ); +} +void hci_req_add_le_passive_scan (struct hci_request *req) {struct hci_dev *hdev = req->hdev;
Signature · 1 line
2.17.1
Marcel Holtmann
maintainerRe: [PATCH 1/7] Bluetooth: LL Privacy Delete Store Resolving list
Hi Sathish,
Quoted 50 lines
This patch help to delte the resolving list stored in the BT Controller w.r.t
BD_ADDR.Signed-off-by Sathish Narsimman <[email protected]>---
include/net/bluetooth/hci.h | 1 + include/net/bluetooth/hci_core.h | 2 ++ net/bluetooth/hci_request.c | 33 ++++++++++++++++++++++++++++++++ 3 files changed, 36 insertions(+)include/net/bluetooth/hci.h +1
-0@@ -457,6 +457,7 @@ enum {#define HCI_LE_DATA_LEN_EXT 0x20#define HCI_LE_PHY_2M 0x01#define HCI_LE_PHY_CODED 0x08 +#define HCI_LE_LL_PRIVACY 0x40#define HCI_LE_EXT_ADV 0x10#define HCI_LE_EXT_SCAN_POLICY 0x80#define HCI_LE_PHY_2M 0x01include/net/bluetooth/hci_core.h +2
-0@@ -1657,6 +1657,8 @@ void hci_le_start_enc(struct hci_conn *conn, __le16 ediv, __le64 rand,void hci_copy_identity_address (struct hci_dev *hdev,bdaddr_t *bdaddr,u8 *bdaddr_type); +void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type, +bdaddr_t *bdaddr);#define SCO_AIRMODE_MASK 0x0003#define SCO_AIRMODE_CVSD 0x0000net/bluetooth/hci_request.c +7
-0@@ -885,6 +885,39 @@ static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval, } } +void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr) +{ +struct hci_cp_le_del_from_resolv_list cp; +struct hci_request req; +struct bdaddr_list_with_irk *irk; + +BT_DBG ("" );
net/bluetooth/hci_request.c static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval, · line 904
Quoted 10 lines
net/bluetooth/hci_request.c +10
-0+ +/* Nothing to be done if LL privacy is not supported */ +if ( !(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) ) +return ; + +if ( !hci_dev_test_flag (hdev,HCI_LL_RPA_RESOLUTION ) && + (hci_dev_test_flag (hdev,HCI_LE_ADV ) || +hci_dev_test_flag (hdev,HCI_LE_SCAN ) || +hci_lookup_le_connect (hdev) ) ) +return ;
please don’t invent your own coding style. Follow the netdev convention.
Regards
Marcel
Re: [PATCH 1/7] Bluetooth: LL Privacy Delete Store Resolving list
Hi Marcel
Quoted 57 lines · Marcel Holtmann
Hi Sathish,
Quoted 54 lines
This patch help to delte the resolving list stored in the BT Controller w.r.t
BD_ADDR.Signed-off-by Sathish Narsimman <[email protected]>---
include/net/bluetooth/hci.h | 1 + include/net/bluetooth/hci_core.h | 2 ++ net/bluetooth/hci_request.c | 33 ++++++++++++++++++++++++++++++++ 3 files changed, 36 insertions(+)include/net/bluetooth/hci.h +1
-0@@ -457,6 +457,7 @@ enum {#define HCI_LE_DATA_LEN_EXT 0x20#define HCI_LE_PHY_2M 0x01#define HCI_LE_PHY_CODED 0x08 +#define HCI_LE_LL_PRIVACY 0x40#define HCI_LE_EXT_ADV 0x10#define HCI_LE_EXT_SCAN_POLICY 0x80#define HCI_LE_PHY_2M 0x01include/net/bluetooth/hci_core.h +2
-0@@ -1657,6 +1657,8 @@ void hci_le_start_enc(struct hci_conn *conn, __le16 ediv, __le64 rand,void hci_copy_identity_address (struct hci_dev *hdev,bdaddr_t *bdaddr,u8 *bdaddr_type); +void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type, +bdaddr_t *bdaddr);#define SCO_AIRMODE_MASK 0x0003#define SCO_AIRMODE_CVSD 0x0000net/bluetooth/hci_request.c +11
-0@@ -885,6 +885,39 @@ static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval, } } +void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr) +{ +struct hci_cp_le_del_from_resolv_list cp; +struct hci_request req; +struct bdaddr_list_with_irk *irk; + +BT_DBG ("" ); + +/* Nothing to be done if LL privacy is not supported */ +if ( !(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) ) +return ;
Quoted 8 lines
net/bluetooth/hci_request.c static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval, · line 904
Quoted 6 lines
net/bluetooth/hci_request.c +6
-0+ +if ( !hci_dev_test_flag (hdev,HCI_LL_RPA_RESOLUTION ) && + (hci_dev_test_flag (hdev,HCI_LE_ADV ) || +hci_dev_test_flag (hdev,HCI_LE_SCAN ) || +hci_lookup_le_connect (hdev) ) ) +return ;please don’t invent your own coding style. Follow the netdev convention.
Sure will correct the coding style in next version of the patch
Quoted 5 lines
Regards
Marcel
PATCH 2/7 Bluetooth: Add device to Resolving List
In reply to Sathish NarsimmanThe Patch is used to add the device to resolving list. The patch will look for the local IRK list and if the device is present in the local resolving list it will be updated and the same updated w.r.t to BT controller by first deleting the existing and adding the new IRK w.r.t bd_addr.
--- include/net/bluetooth/hci_core.h | 2 + net/bluetooth/hci_request.c | 80 ++++++++++++++++++++++++++++++++ 2 files changed, 82 insertions(+)
include/net/bluetooth/hci_core.h +2 -0
@@ -1657,6 +1657,8 @@ void hci_le_start_enc(struct hci_conn *conn, __le16 ediv, __le64 rand,void hci_copy_identity_address (struct hci_dev *hdev,bdaddr_t *bdaddr,u8 *bdaddr_type); +int hci_req_update_resolving_list (struct hci_dev *hdev,u8 addr_type, +bdaddr_t *bdaddr,u8 irk[16 ]);void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr);
net/bluetooth/hci_request.c +80 -0
86 lines not on this page · raw message
Signature · 1 line
2.17.1
Re: [PATCH 2/7] Bluetooth: Add device to Resolving List
Hi,
Quoted 56 lines · On Thursday, 12 March 2020 11:07:49 CET Sathish Narsimman wrote:
The Patch is used to add the device to resolving list. The patch will look for the local IRK list and if the device is present in the local resolving list it will be updated and the same updated w.r.t to BT controller by first deleting the existing and adding the new IRK w.r.t
bd_addr.Signed-off-by Sathish Narsimman <[email protected]>Signed-off-by Joy Shermin <[email protected]>--- include/net/bluetooth/hci_core.h | 2 + net/bluetooth/hci_request.c | 80 ++++++++++++++++++++++++++++++++ 2 files changed, 82 insertions(+)(quoted diff) +0
-0
b/include/net/bluetooth/hci_core.hindex5f04ef88da35..e5e09d530ce7100644include/net/bluetooth/hci_core.h +2
-0@@ -1657,6 +1657,8 @@ void hci_le_start_enc(struct hci_conn *conn, __le16 ediv, __le64 rand, void hci_copy_identity_address(struct hci_dev *hdev, bdaddr_t *bdaddr,u8 *bdaddr_type); +int hci_req_update_resolving_list (struct hci_dev *hdev,u8 addr_type, +bdaddr_t *bdaddr,u8 irk[16 ]);void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr);net/bluetooth/hci_request.c +20
-0@@ -885,6 +885,86 @@ static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval, } } +int hci_req_update_resolving_list(struct hci_dev *hdev, u8 type, bdaddr_t *bdaddr, + u8 irk_val[16]) +{ + struct hci_request req; +struct hci_cp_le_add_to_resolv_list cp; +struct bdaddr_list_with_irk *irk; +u8 entries; + +BT_DBG ("" ); + +/* Nothing to be done if LL privacy is not supported */ +if ( !(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) ) +return -EPROTONOSUPPORT ; + +/* Resolving List cannot be updated if address resolution +* in the controller is enabled and advertisement or scanning +* or create connection command is ongoing. +*/ + if ( !hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION) && + ( hci_dev_test_flag(hdev, HCI_LE_ADV) || + hci_dev_test_flag(hdev, HCI_LE_SCAN) ||
net/bluetooth/hci_request.c static void hci_req_start_scan(struct hci_request *req, · line 918
Quoted 10 lines
net/bluetooth/hci_request.c +9
-0+ hci_lookup_le_connect(hdev) ) ) +return -EINVAL ; + +hci_req_init (&req, hdev); + + irk =hci_bdaddr_list_lookup_with_irk (&hdev->le_resolv_list, bdaddr, type); + if (irk) { +/* Device is present in resolving list.*/ +if (memcmp (irk->peer_irk, irk_val,16 ) ==0 ) { +/* Device present in resolving list with same
IRK.
net/bluetooth/hci_request.c static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval,
Quoted 4 lines
net/bluetooth/hci_request.c +4
-0+* No HCI communication is required.Sort the +* Kernel list. +*/ + hci_bdaddr_list_add_with_irk(&irk->list, bdaddr,
type, irk_val, NULL);
Quoted 28 lines
net/bluetooth/hci_request.c +27
-0+return 0 ; + + } +/* IRK has changed for the device in resolving list +* Queue up commands to delete the existing entry and +* add new one. Sorting will be done when command complete +* for add command is receieved. +*/ + goto remote; + } +/* Device is not present in resolving list.If resolving list +* is not full add the device to resolving list. +*/ + entries =0 ; +list_for_each_entry (irk, &hdev->le_resolv_list, list) { + entries++; + } + +if (entries < hdev->le_resolv_list_size) + goto add; + +/* If the resolving list is full, queue up HCI command to delete +* the entry that was used least recently to make space for the +* new device. Kernel list will be updated when command complete +* is received. +*/ + irk = list_first_entry_or_null(&hdev->le_resolv_list, struct
bdaddr_list_with_irk, list); +
net/bluetooth/hci_request.c static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval, · line 952
Quoted 2 lines
net/bluetooth/hci_request.c +2
-0+if (!irk) +return -1 ;
I'm not sure about this... we should have some fallback mechanism when HW RL is full. Maybe some rotation of devies on list or simply disable address resolution in controller? Otherwise we won't be able to connect to some of paired devices.
Quoted 19 lines
net/bluetooth/hci_request.c +3
-0+ +remote: + hci_req_del_from_resolving_list(hdev, irk->bdaddr_type, &irk-bdaddr);
net/bluetooth/hci_request.c +12
-0+ +add: + cp.bdaddr_type = type; +bacpy (&cp.bdaddr, bdaddr); +memcpy (cp.peer_irk, irk_val,16 ); +memcpy (cp.local_irk, hdev->irk,16 ); + +hci_req_add (&req,HCI_OP_LE_ADD_TO_RESOLV_LIST ,sizeof (cp), &cp); +hci_req_run (&req,NULL ); +return 0 ; +} + void hci_req_del_from_resolving_list(struct hci_dev *hdev, u8 addr_type,
bdaddr_t*bdaddr) {
structhci_cp_le_del_from_resolv_listcp;
Signature · 2 lines
pozdrawiam Szymon Janc
Re: [PATCH 2/7] Bluetooth: Add device to Resolving List
Hi Szymon,
Quoted 103 lines · Szymon Janc
Hi,
Quoted 56 lines · On Thursday, 12 March 2020 11:07:49 CET Sathish Narsimman wrote:
The Patch is used to add the device to resolving list. The patch will look for the local IRK list and if the device is present in the local resolving list it will be updated and the same updated w.r.t to BT controller by first deleting the existing and adding the new IRK w.r.t
bd_addr.Signed-off-by Sathish Narsimman <[email protected]>Signed-off-by Joy Shermin <[email protected]>--- include/net/bluetooth/hci_core.h | 2 + net/bluetooth/hci_request.c | 80 ++++++++++++++++++++++++++++++++ 2 files changed, 82 insertions(+)(quoted diff) +0
-0
b/include/net/bluetooth/hci_core.hindex5f04ef88da35..e5e09d530ce7100644include/net/bluetooth/hci_core.h +2
-0@@ -1657,6 +1657,8 @@ void hci_le_start_enc(struct hci_conn *conn, __le16 ediv, __le64 rand, void hci_copy_identity_address(struct hci_dev *hdev, bdaddr_t *bdaddr,u8 *bdaddr_type); +int hci_req_update_resolving_list (struct hci_dev *hdev,u8 addr_type, +bdaddr_t *bdaddr,u8 irk[16 ]);void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr);net/bluetooth/hci_request.c +20
-0@@ -885,6 +885,86 @@ static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval, } } +int hci_req_update_resolving_list(struct hci_dev *hdev, u8 type, bdaddr_t *bdaddr, + u8 irk_val[16]) +{ + struct hci_request req; +struct hci_cp_le_add_to_resolv_list cp; +struct bdaddr_list_with_irk *irk; +u8 entries; + +BT_DBG ("" ); + +/* Nothing to be done if LL privacy is not supported */ +if ( !(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) ) +return -EPROTONOSUPPORT ; + +/* Resolving List cannot be updated if address resolution +* in the controller is enabled and advertisement or scanning +* or create connection command is ongoing. +*/ + if ( !hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION) && + ( hci_dev_test_flag(hdev, HCI_LE_ADV) || + hci_dev_test_flag(hdev, HCI_LE_SCAN) ||net/bluetooth/hci_request.c static void hci_req_start_scan(struct hci_request *req, · line 918
Quoted 10 lines
net/bluetooth/hci_request.c +9
-0+ hci_lookup_le_connect(hdev) ) ) +return -EINVAL ; + +hci_req_init (&req, hdev); + + irk =hci_bdaddr_list_lookup_with_irk (&hdev->le_resolv_list, bdaddr, type); + if (irk) { +/* Device is present in resolving list.*/ +if (memcmp (irk->peer_irk, irk_val,16 ) ==0 ) { +/* Device present in resolving list with same IRK.
net/bluetooth/hci_request.c static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval,
Quoted 4 lines
net/bluetooth/hci_request.c +4
-0+* No HCI communication is required.Sort the +* Kernel list. +*/ + hci_bdaddr_list_add_with_irk(&irk->list, bdaddr,type,
irk_val, NULL);Quoted 27 lines
net/bluetooth/hci_request.c +27
-0+return 0 ; + + } +/* IRK has changed for the device in resolving list +* Queue up commands to delete the existing entry and +* add new one. Sorting will be done when command complete +* for add command is receieved. +*/ + goto remote; + } +/* Device is not present in resolving list.If resolving list +* is not full add the device to resolving list. +*/ + entries =0 ; +list_for_each_entry (irk, &hdev->le_resolv_list, list) { + entries++; + } + +if (entries < hdev->le_resolv_list_size) + goto add; + +/* If the resolving list is full, queue up HCI command to delete +* the entry that was used least recently to make space for the +* new device. Kernel list will be updated when command complete +* is received. +*/ + irk = list_first_entry_or_null(&hdev->le_resolv_list, struct
Quoted 8 lines
net/bluetooth/hci_request.c static void hci_req_start_scan(struct hci_request *req, u8 type, u16 interval, · line 952
Quoted 3 lines
bdaddr_list_with_irk, list); +net/bluetooth/hci_request.c +2
-0+if (!irk) +return -1 ;I'm not sure about this... we should have some fallback mechanism when HW RL is full. Maybe some rotation of devies on list or simply disable address resolution in controller? Otherwise we won't be able to connect to some of paired devices.
What I understood was
if (HW RL is full) {
don't delete the device from the kernel resolving list. delete the device from HW RL list(least used)
but we have to use a method to denote that it is deleted from HW RL list. }
from the above method at what scenario we need to add the deleted device back to the HW RL list?
And yes disabling HW RL list much simpler.
Quoted 17 lines
Quoted 17 lines
net/bluetooth/hci_request.c +3
-0+ +remote: + hci_req_del_from_resolving_list(hdev, irk->bdaddr_type, &irk-bdaddr);
net/bluetooth/hci_request.c +12
-0+ +add: + cp.bdaddr_type = type; +bacpy (&cp.bdaddr, bdaddr); +memcpy (cp.peer_irk, irk_val,16 ); +memcpy (cp.local_irk, hdev->irk,16 ); + +hci_req_add (&req,HCI_OP_LE_ADD_TO_RESOLV_LIST ,sizeof (cp), &cp); +hci_req_run (&req,NULL ); +return 0 ; +} +
Quoted 10 lines
Quoted 3 lines
void hci_req_del_from_resolving_list(struct
hci_dev*hdev, u8addr_type,bdaddr_t *bdaddr) { struct hci_cp_le_del_from_resolv_list cp;--
pozdrawiam
Szymon Janc
Regards
Sathish N
PATCH 3/7 Bluetooth: Update the Resolving list when Local IRK changed
In reply to Sathish NarsimmanWhenever the Local IRK changes. i.e privacy mode is toggled this patch helps to change the IRK list in the bluetooth controller
--- include/net/bluetooth/hci_core.h | 2 ++ net/bluetooth/hci_request.c | 50 ++++++++++++++++++++++++++++++++ 2 files changed, 52 insertions(+)
include/net/bluetooth/hci_core.h +2 -0
@@ -1659,6 +1659,8 @@ void hci_copy_identity_address(struct hci_dev *hdev, bdaddr_t *bdaddr, u8 *bdaddr_type);int hci_req_update_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr,u8 irk[16 ]); +void hci_req_update_resolving_list_local_irk (struct hci_dev *hdev); +void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr);
net/bluetooth/hci_request.c +50 -0
56 lines not on this page · raw message
Signature · 1 line
2.17.1
Marcel Holtmann
maintainerRe: [PATCH 3/7] Bluetooth: Update the Resolving list when Local IRK changed
Hi Sathish,
Quoted 36 lines
Whenever the Local IRK changes. i.e privacy mode is toggled this patch helps to change the IRK list in the bluetooth controller
Signed-off-by Sathish Narsimman <[email protected]>Signed-off-by Joy Shermin <[email protected]>---
include/net/bluetooth/hci_core.h| 2 ++net/bluetooth/hci_request.c| 50 ++++++++++++++++++++++++++++++++2 files changed, 52 insertions(+)include/net/bluetooth/hci_core.h +2
-0@@ -1659,6 +1659,8 @@ void hci_copy_identity_address(struct hci_dev *hdev, bdaddr_t *bdaddr, u8 *bdaddr_type);int hci_req_update_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr,u8 irk[16 ]); +void hci_req_update_resolving_list_local_irk (struct hci_dev *hdev); +void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr);net/bluetooth/hci_request.c +4
-0@@ -965,6 +965,56 @@ int hci_req_update_resolving_list(struct hci_dev *hdev, u8 type, bdaddr_t *bdaddreturn 0 ; } +void hci_req_update_resolving_list_local_irk (struct hci_dev *hdev) +{ +struct bdaddr_list_with_irk *irk; +struct hci_request req;
net/bluetooth/hci_request.c int hci_req_update_resolving_list(struct hci_dev *hdev, u8 type, bdaddr_t *bdadd · line 981
Quoted 10 lines
net/bluetooth/hci_request.c +10
-0+ +BT_DBG ("" ); + +/* Nothing to be done if LL privacy is not supported. */ +if (!(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY )) +return ; + +/* If resolving list is empty, nothing is to be done.*/ +if (list_empty (&hdev->le_resolv_list)) +return ;
actually the le_resolv_list represents the current list programmed into the controller.
net/bluetooth/hci_request.c int hci_req_update_resolving_list(struct hci_dev *hdev, u8 type, bdaddr_t *bdadd · line 991
Quoted 10 lines
net/bluetooth/hci_request.c +10
-0+ +/* Resolving List cannot be updated if address resolution +* in the controller is enabled and advertisement or scanning +* or create connection command is ongoing. +*/ +if ( !hci_dev_test_flag (hdev,HCI_LL_RPA_RESOLUTION ) && + (hci_dev_test_flag (hdev,HCI_LE_ADV ) || +hci_dev_test_flag (hdev,HCI_LE_SCAN ) || +hci_lookup_le_connect (hdev) ) ) +return ;
Coding style needs to be correct.
Quoted 12 lines
net/bluetooth/hci_request.c +12
-0+ +/* If resolving list is not empty, then clear the resolving list +* and add back each entry with updated local IRK. +*/ +hci_req_init (&req, hdev); + +list_for_each_entry (irk, &hdev->le_resolv_list, list) { +struct hci_cp_le_add_to_resolv_list cp; +struct hci_cp_le_del_from_resolv_list cp1; + + cp1.bdaddr_type = irk->bdaddr_type; +bacpy (&cp1.bdaddr, &irk->bdaddr);
net/bluetooth/hci_request.c int hci_req_update_resolving_list(struct hci_dev *hdev, u8 type, bdaddr_t *bdadd · line 1013
Quoted 10 lines
net/bluetooth/hci_request.c +10
-0+hci_req_add (&req,HCI_OP_LE_DEL_FROM_RESOLV_LIST , +sizeof (cp1), &cp1); + + cp.bdaddr_type = irk->bdaddr_type; +bacpy (&cp.bdaddr, &irk->bdaddr); +memcpy (cp.peer_irk, irk->peer_irk,16 ); +memcpy (cp.local_irk, hdev->irk,16 ); +hci_req_add (&req,HCI_OP_LE_ADD_TO_RESOLV_LIST , +sizeof (cp), &cp); + }
Can we please do this the smart way, no need to remove and entry to just add it back later.
Regards
Marcel
PATCH 4/7 Bluetooth: LL_PRIVACY re-load resolving list
In reply to Sathish NarsimmanThe patch is used to load the entries in the local IRK to the Bluetooth LE controller resolving list.
--- include/net/bluetooth/hci_core.h | 1 + net/bluetooth/hci_request.c | 29 +++++++++++++++++++++++++++++ 2 files changed, 30 insertions(+)
include/net/bluetooth/hci_core.h +1 -0
@@ -1659,6 +1659,7 @@ void hci_copy_identity_address(struct hci_dev *hdev, bdaddr_t *bdaddr, u8 *bdaddr_type);int hci_req_update_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr,u8 irk[16 ]); +void hci_load_resolving_list (struct hci_dev *hdev);void hci_req_update_resolving_list_local_irk (struct hci_dev *hdev); void hci_req_del_from_resolving_list(struct hci_dev *hdev, u8 addr_type,
net/bluetooth/hci_request.c +29 -0
@@ -1015,6 +1015,35 @@ void hci_req_update_resolving_list_local_irk(struct hci_dev *hdev)hci_req_run (&req,NULL ); } +void hci_load_resolving_list (struct hci_dev *hdev) +{ +struct smp_irk *irk; +u8 num =0 ; + +/* Nothing to be done if LL privacy is not supported */ +if ( !(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) ) +return ; + +if ( !hci_dev_test_flag (hdev,HCI_LL_RPA_RESOLUTION ) && + (hci_dev_test_flag (hdev,HCI_LE_ADV ) || +hci_dev_test_flag (hdev,HCI_LE_SCAN ) || +hci_lookup_le_connect (hdev) ) ) +return ; + +/* Load the first le_resolving_list_size entries from IRK +* list in to resolving list. +*/ +rcu_read_lock (); +list_for_each_entry_rcu (irk, &hdev->identity_resolving_keys, list) { +if (num++ >= hdev->le_resolv_list_size) +return ; + +hci_req_update_resolving_list (hdev, irk->addr_type, &irk->bdaddr, irk->val); + } + +rcu_read_unlock (); +} +void hci_req_del_from_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr) {struct hci_cp_le_del_from_resolv_list cp;
Signature · 1 line
2.17.1
Marcel Holtmann
maintainerRe: [PATCH 4/7] Bluetooth: LL_PRIVACY re-load resolving list
Hi Sathish,
Quoted 45 lines
The patch is used to load the entries in the local IRK to the Bluetooth LE controller resolving list.
Signed-off-by Sathish Narsimman <[email protected]>Signed-off-by Joy Shermin <[email protected]>---
include/net/bluetooth/hci_core.h| 1 +net/bluetooth/hci_request.c| 29 +++++++++++++++++++++++++++++2 files changed, 30 insertions(+)include/net/bluetooth/hci_core.h +1
-0@@ -1659,6 +1659,7 @@ void hci_copy_identity_address(struct hci_dev *hdev, bdaddr_t *bdaddr, u8 *bdaddr_type);int hci_req_update_resolving_list (struct hci_dev *hdev,u8 addr_type,bdaddr_t *bdaddr,u8 irk[16 ]); +void hci_load_resolving_list (struct hci_dev *hdev);void hci_req_update_resolving_list_local_irk (struct hci_dev *hdev); void hci_req_del_from_resolving_list(struct hci_dev *hdev, u8 addr_type,net/bluetooth/hci_request.c +15
-0@@ -1015,6 +1015,35 @@ void hci_req_update_resolving_list_local_irk(struct hci_dev *hdev)hci_req_run (&req,NULL ); } +void hci_load_resolving_list (struct hci_dev *hdev) +{ +struct smp_irk *irk; +u8 num =0 ; + +/* Nothing to be done if LL privacy is not supported */ +if ( !(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) ) +return ; + +if ( !hci_dev_test_flag (hdev,HCI_LL_RPA_RESOLUTION ) && + (hci_dev_test_flag (hdev,HCI_LE_ADV ) || +hci_dev_test_flag (hdev,HCI_LE_SCAN ) || +hci_lookup_le_connect (hdev) ) ) +return ; +
net/bluetooth/hci_request.c void hci_req_update_resolving_list_local_irk(struct hci_dev *hdev) · line 1042
Quoted 10 lines
net/bluetooth/hci_request.c +10
-0+/* Load the first le_resolving_list_size entries from IRK +* list in to resolving list. +*/ +rcu_read_lock (); +list_for_each_entry_rcu (irk, &hdev->identity_resolving_keys, list) { +if (num++ >= hdev->le_resolv_list_size) +return ; + +hci_req_update_resolving_list (hdev, irk->addr_type, &irk->bdaddr, irk->val); + }
I comparison to the le_whitelist, the le_resolv_list_size does not have to match the actual available memory of the controller. It is just some guesstimate. I think we have to design this in a way that we add the IRK anyway and handle the error case gracefully.
I am also confused why we are not actually checking here if an entry is already present and just loading the whole list. I have seen code being duplicated in at least 3 places now.
Regards
Marcel
PATCH 5/7 Bluetooth: LL Privacy Check and Update the Resolving List
In reply to Sathish NarsimmanThis patch is an helper function which will be usefull to check the Local Resolving list and update the controller Resolving list during le_create_conn , le_directed_adv.
--- net/bluetooth/hci_request.c | 21 +++++++++++++++++++++ net/bluetooth/hci_request.h | 2 ++ 2 files changed, 23 insertions(+)
net/bluetooth/hci_request.c +21 -0
@@ -1015,6 +1015,27 @@ void hci_req_update_resolving_list_local_irk(struct hci_dev *hdev)hci_req_run (&req,NULL ); } +void hci_req_check_and_update_resolving_list (struct hci_dev *hdev,u8 type,bdaddr_t *bdaddr) +{ +struct smp_irk *irk; + +BT_DBG ("" ); + +/* Nothing to be done if LL privacy is not supported */ +if ( !(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) ) +return ; + +if ( !hci_dev_test_flag (hdev,HCI_LL_RPA_RESOLUTION ) && + (hci_dev_test_flag (hdev,HCI_LE_ADV ) || +hci_dev_test_flag (hdev,HCI_LE_SCAN ) || +hci_lookup_le_connect (hdev) ) ) +return ; + + irk =hci_find_irk_by_addr (hdev, bdaddr, type); +if (irk) +hci_req_update_resolving_list (hdev, irk->addr_type, &irk->bdaddr, irk->val); +} +void hci_load_resolving_list (struct hci_dev *hdev) {struct smp_irk *irk;
net/bluetooth/hci_request.h +2 -0
@@ -90,6 +90,8 @@ void __hci_req_clear_ext_adv_sets(struct hci_request *req);int hci_get_random_address (struct hci_dev *hdev,bool require_privacy,bool use_rpa,struct adv_info *adv_instance,u8 *own_addr_type,bdaddr_t *rand_addr); +void hci_req_check_and_update_resolving_list (struct hci_dev *hdev,u8 addr_type, +bdaddr_t *bdaddr);void __hci_req_update_class (struct hci_request *req);
Signature · 1 line
2.17.1
PATCH 6/7 Bluetooth: Enable LL Privacy
In reply to Sathish NarsimmanThis enables Link Layer Privacy and sets RPA timeout to 900sec
< HCI Command: LE Set Address Resolution Enable (0x08|0x002d) plen 1
Address resolution: Enabled (0x01)
Quoted 1 line
HCI Event: Command Complete (
0x0e) plen 4
LE Set Address Resolution Enable (0x08|0x002d) ncmd 1d
Status: Success (0x00)< HCI Command: LE Set Resolvable Private Address Timeout (0x08|0x002e) plen 2
Timeout: 900 seconds
Quoted 1 line
HCI Event: Command Complete (
0x0e) plen 4
LE Set Resolvable Private Address Timeout (0x08|0x002e) ncmd 1
Status: Success (0x00)During Successfull pairing adding the information to device Resolving list
< HCI Command: LE Add Device To Resolving List (0x08|0x0027) plen 39
Address type: Random (0x01)
Address: D1:79:92:B1:83:DF (Static)
Peer identity resolving key: ecc317056e5b385cda281b9c24a453e9
Local identity resolving key: 00000000000000000000000000000000< ACL Data TX: Handle 3585 flags 0x00 dlen 21
SMP: Signing Information (0x0a) len 16
Signature key: 37b6c45656c7e8b3ec8db062c3e0c2b2Quoted 1 line
ACL Data RX: Handle 3585 flags
0x02dlen 9
ATT: Error Response (0x01) len 4
Read By Group Type Request (0x10)
Handle: 0x0001
Error: Attribute Not Found (0x0a)Quoted 1 line
ACL Data RX: Handle 3585 flags
0x02dlen 11
ATT: Read By Group Type Request (0x10) len 6
Handle range: 0x000a-0xffff
Attribute group type: Primary Service (0x2800)Quoted 1 line
HCI Event: Command Complete (
0x0e) plen 4
LE Add Device To Resolving List (0x08|0x0027) ncmd 1
Status: Success (0x00)--- include/net/bluetooth/hci.h | 2 ++ net/bluetooth/hci_conn.c | 32 ++++++++++++++++++++++++++++++++ net/bluetooth/hci_core.c | 20 +++++++++++++++++++- net/bluetooth/hci_event.c | 32 +++++++++++++++++++++++++------- net/bluetooth/hci_request.c | 16 +++++++++++++--- net/bluetooth/mgmt.c | 10 +++++++++- net/bluetooth/smp.c | 11 +++++++++++ 7 files changed, 111 insertions(+), 12 deletions(-)
include/net/bluetooth/hci.h +2 -0
@@ -1617,6 +1617,8 @@ struct hci_rp_le_read_resolv_list_size {#define HCI_OP_LE_SET_ADDR_RESOLV_ENABLE 0x202d +#define HCI_OP_LE_SET_RPA_TIMEOUT 0x202e +#define HCI_OP_LE_READ_MAX_DATA_LEN 0x202fstruct hci_rp_le_read_max_data_len {__u8 status;
net/bluetooth/hci_conn.c +32 -0
@@ -830,6 +830,14 @@ static void hci_req_add_le_create_conn(struct hci_request *req, plen =sizeof (*cp); +/* If the own_addr_type is 0x03 or 0x02 and peer is present +* in Kernel IRK list, but not in resolving add the peer to +* the resolving list. +*/ +if (own_addr_type ==0x03 || own_addr_type ==0x02 ) +hci_req_check_and_update_resolving_list (hdev, conn->dst_type, + &conn->dst); +if (scan_1m (hdev)) { cp->phys |=LE_SCAN_PHY_1M ;set_ext_conn_params (conn, p);
@@ -866,6 +874,14 @@ static void hci_req_add_le_create_conn(struct hci_request *req, cp.scan_interval =cpu_to_le16 (hdev->le_scan_interval); cp.scan_window = cp.scan_interval; +/* If the own_addr_type is 0x03 or 0x02 and peer is present +* in Kernel IRK list, but not in resolving add the peer to +* the resolving list. +*/ +if (own_addr_type ==0x03 || own_addr_type ==0x02 ) +hci_req_check_and_update_resolving_list (hdev, conn->dst_type, + &conn->dst); +bacpy (&cp.peer_addr, &conn->dst); cp.peer_addr_type = conn->dst_type; cp.own_address_type = own_addr_type;
@@ -901,6 +917,14 @@ static void hci_req_directed_advertising(struct hci_request *req, &own_addr_type, &random_addr) < 0)return ; +/* If the own_addr_type is 0x03 or 0x02 and peer is present +* in Kernel IRK list, but not in resolving add the peer to +* the resolving list. +*/ +if (own_addr_type ==0x03 || own_addr_type ==0x03 ) +hci_req_check_and_update_resolving_list (hdev, conn->dst_type, + &conn->dst); +memset (&cp,0 ,sizeof (cp)); cp.evt_properties =cpu_to_le16 (LE_LEGACY_ADV_DIRECT_IND );
@@ -959,6 +983,14 @@ static void hci_req_directed_advertising(struct hci_request *req, &own_addr_type) < 0)return ; +/* If the own_addr_type is 0x03 or 0x02 and peer is present +* in Kernel IRK list, but not in resolving add the peer to +* the resolving list. +*/ +if (own_addr_type ==0x03 || own_addr_type ==0x02 ) +hci_req_check_and_update_resolving_list (hdev, conn->dst_type, + &conn->dst); +memset (&cp,0 ,sizeof (cp));/* Some controllers might reject command if intervals are not
net/bluetooth/hci_core.c +19 -1
@@ -713,7 +713,7 @@ static int hci_init3_req(struct hci_request *req, unsigned long opt)/* If the controller supports the LE Extended Create Connection * command, enable the corresponding event. */ -if (use_ext_conn (hdev)) +if (use_ext_conn (hdev) || hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) events[1 ] |=0x02 ;/* LE Enhanced Connection * Complete */
@@ -762,6 +762,24 @@ static int hci_init3_req(struct hci_request *req, unsigned long opt)hci_req_add (req,HCI_OP_LE_CLEAR_RESOLV_LIST ,0 ,NULL ); } +if (hdev->commands[35 ] &0x02 ) { +u8 addr_resoln_enable =0x01 ; + +/* Enable Address Resolution in controller */ +hci_req_add (req,HCI_OP_LE_SET_ADDR_RESOLV_ENABLE , +sizeof (addr_resoln_enable), + &addr_resoln_enable); + } + +if (hdev->commands[35 ] &0x40 ) { +__le16 rpa_timeout =cpu_to_le16 (hdev->rpa_timeout); + +/* Set RPA timeout */ +hci_req_add (req,HCI_OP_LE_SET_RPA_TIMEOUT ,2 , + &rpa_timeout); + + } +if (hdev->le_features[0 ] &HCI_LE_DATA_LEN_EXT ) {/* Read LE Maximum Data Length */ hci_req_add (req,HCI_OP_LE_READ_MAX_DATA_LEN ,0 ,NULL );
net/bluetooth/hci_event.c +25 -7
@@ -1684,7 +1684,11 @@ static void hci_cc_set_adv_param(struct hci_dev *hdev, struct sk_buff *skb)return ;hci_dev_lock (hdev); - hdev->adv_addr_type = cp->own_address_type; + +if (cp->own_address_type ==ADDR_LE_DEV_PUBLIC ) + hdev->adv_addr_type =ADDR_LE_DEV_PUBLIC ; +else + hdev->adv_addr_type =ADDR_LE_DEV_RANDOM ;hci_dev_unlock (hdev); }
@@ -1704,7 +1708,12 @@ static void hci_cc_set_ext_adv_param(struct hci_dev *hdev, struct sk_buff *skb)return ;hci_dev_lock (hdev); - hdev->adv_addr_type = cp->own_addr_type; + +if (cp->own_addr_type ==ADDR_LE_DEV_PUBLIC ) + hdev->adv_addr_type =ADDR_LE_DEV_PUBLIC ; +else + hdev->adv_addr_type =ADDR_LE_DEV_RANDOM ; +if (!hdev->cur_adv_instance) {/* Store in hdev for instance 0 */ hdev->adv_tx_power = rp->tx_power;
@@ -2267,11 +2276,13 @@ static void cs_le_create_conn(struct hci_dev *hdev, bdaddr_t *peer_addr,* is needed for SMP. These values will not change during the * lifetime of the connection. */ - conn->init_addr_type = own_address_type; -if (own_address_type ==ADDR_LE_DEV_RANDOM ) +if (own_address_type ==0x03 || own_address_type ==0x01 ) { + conn->init_addr_type =ADDR_LE_DEV_RANDOM ;bacpy (&conn->init_addr, &hdev->random_addr); -else + }else { + conn->init_addr_type =ADDR_LE_DEV_PUBLIC ;bacpy (&conn->init_addr, &hdev->bdaddr); + } conn->resp_addr_type = peer_addr_type;bacpy (&conn->resp_addr, peer_addr);
@@ -5141,6 +5152,8 @@ static void le_conn_complete_evt(struct hci_dev *hdev, u8 status, } } +hci_req_check_and_update_resolving_list (hdev, (conn->dst_type), &conn->dst); + unlock:hci_update_background_scan (hdev);hci_dev_unlock (hdev);
@@ -5365,8 +5378,13 @@ static void process_adv_report(struct hci_dev *hdev, u8 type, bdaddr_t *bdaddr,/* Only resolvable random addresses are valid for these * kind of reports and others can be ignored. */ -if (!hci_bdaddr_is_rpa (direct_addr, direct_addr_type)) -return ; +if (!hci_bdaddr_is_rpa (direct_addr, direct_addr_type)) { +/* Controller with LL Privacy Supported tend to resolve +* the RPA. In that case we should ignore this condition +*/ +if ( !(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY )) +return ; + }/* If the controller is not using resolvable random * addresses, then this report can be ignored.
net/bluetooth/hci_request.c +13 -3
@@ -1916,7 +1916,12 @@ int hci_get_random_address(struct hci_dev *hdev, bool require_privacy,if (use_rpa) {int to; - *own_addr_type =ADDR_LE_DEV_RANDOM ; +if (hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) { +/*if there is no resolving list used in the controller +* use the bd_addr from latest Le_set_random_address*/ + *own_addr_type =0x03 ; + }else + *own_addr_type =ADDR_LE_DEV_RANDOM ;if (adv_instance) { if (!adv_instance->rpa_expired &&
@@ -2074,7 +2079,7 @@ int __hci_req_setup_ext_adv_instance(struct hci_request *req, u8 instance)hci_req_add (req,HCI_OP_LE_SET_EXT_ADV_PARAMS ,sizeof (cp), &cp); -if (own_addr_type ==ADDR_LE_DEV_RANDOM && +if (((own_addr_type ==ADDR_LE_DEV_RANDOM ) || (own_addr_type ==0x03 )) &&bacmp (&random_addr,BDADDR_ANY )) {struct hci_cp_le_set_adv_set_rand_addr cp;
@@ -2341,7 +2346,12 @@ int hci_update_random_address(struct hci_request *req, bool require_privacy,if (use_rpa) {int to; - *own_addr_type =ADDR_LE_DEV_RANDOM ; +if (hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) { +/*if there is no resolving list used in the controller +* use the bd_addr from latest Le_set_random_address*/ + *own_addr_type =0x03 ; + }else + *own_addr_type =ADDR_LE_DEV_RANDOM ;if (!hci_dev_test_and_clear_flag (hdev,HCI_RPA_EXPIRED ) && !bacmp (&hdev->random_addr, &hdev->rpa))
net/bluetooth/mgmt.c +9 -1
@@ -5084,8 +5084,10 @@ static int set_debug_keys(struct sock *sk, struct hci_dev *hdev,if (err <0 ) goto unlock; -if (changed) +if (changed) { err =new_settings (hdev, sk); +hci_req_update_resolving_list_local_irk (hdev); + } unlock:hci_dev_unlock (hdev);
@@ -5227,6 +5229,8 @@ static int load_irks(struct sock *sk, struct hci_dev *hdev, void *cp_data, BDADDR_ANY); } +/* Load the resolving list with entries from IRK list.*/ +hci_load_resolving_list (hdev);hci_dev_set_flag (hdev,HCI_RPA_RESOLVING ); err =mgmt_cmd_complete (sk, hdev->id,MGMT_OP_LOAD_IRKS ,0 ,NULL ,0 );
@@ -7213,6 +7217,10 @@ void mgmt_power_on(struct hci_dev *hdev, int err)if (match.sk)sock_put (match.sk); +/* If load _irk was called when controller was powered down, +* then the resolving list has to be updated now. +*/ +hci_load_resolving_list (hdev);hci_dev_unlock (hdev); }
net/bluetooth/smp.c +11 -0
@@ -2425,6 +2425,8 @@ int smp_cancel_and_remove_pairing(struct hci_dev *hdev, bdaddr_t *bdaddr,struct smp_chan *smp;int err; +hci_req_del_from_resolving_list (hdev, addr_type, bdaddr); + err =hci_remove_ltk (hdev, bdaddr, addr_type);hci_remove_irk (hdev, bdaddr, addr_type);
@@ -2619,6 +2621,15 @@ static int smp_cmd_ident_addr_info(struct l2cap_conn *conn, smp->remote_irk =hci_add_irk (conn->hcon->hdev, &smp->id_addr, smp->id_addr_type, smp->irk, &rpa); +/* Add the device to the resolving list.*/ +hci_req_update_resolving_list (conn->hcon->hdev, + smp->remote_irk->addr_type, + &smp->remote_irk->bdaddr, + smp->remote_irk->val); + + +/*TODO: Check remote device Supports LL Privacy. If not set the device +* with IRK to Device-privacy Mode - Should be discussed - Based on */ distribute:if (!(smp->remote_key_dist &KEY_DIST_MASK ))smp_distribute_keys (smp);
Signature · 1 line
2.17.1
Marcel Holtmann
maintainerRe: [PATCH 6/7] Bluetooth: Enable LL Privacy
Hi Sathish,
Quoted 69 lines
This enables Link Layer Privacy and sets RPA timeout to 900sec
< HCI Command: LE Set Address Resolution Enable (0x08|0x002d) plen 1
Address resolution: Enabled (0x01)Quoted 1 line
HCI Event: Command Complete (
0x0e) plen 4LE Set Address Resolution Enable (0x08|0x002d) ncmd 1d Status: Success (0x00)< HCI Command: LE Set Resolvable Private Address Timeout (0x08|0x002e) plen 2
Timeout: 900 secondsQuoted 1 line
HCI Event: Command Complete (
0x0e) plen 4LE Set Resolvable Private Address Timeout (0x08|0x002e) ncmd 1 Status: Success (0x00)During Successfull pairing adding the information to device Resolving list
< HCI Command: LE Add Device To Resolving List (0x08|0x0027) plen 39
Address type: Random (0x01) Address: D1:79:92:B1:83:DF (Static) Peer identity resolving key: ecc317056e5b385cda281b9c24a453e9 Local identity resolving key: 00000000000000000000000000000000< ACL Data TX: Handle 3585 flags
0x00dlen 21SMP: Signing Information (0x0a) len 16 Signature key: 37b6c45656c7e8b3ec8db062c3e0c2b2Quoted 1 line
ACL Data RX: Handle 3585 flags
0x02dlen 9ATT: Error Response (0x01) len 4 Read By Group Type Request (0x10) Handle: 0x0001 Error: Attribute Not Found (0x0a)Quoted 1 line
ACL Data RX: Handle 3585 flags
0x02dlen 11ATT: Read By Group Type Request (0x10) len 6 Handle range: 0x000a-0xffff Attribute group type: Primary Service (0x2800)Quoted 1 line
HCI Event: Command Complete (
0x0e) plen 4LE Add Device To Resolving List (0x08|0x0027) ncmd 1 Status: Success (0x00)Signed-off-by Sathish Narsimman <[email protected]>---
include/net/bluetooth/hci.h | 2 ++ net/bluetooth/hci_conn.c | 32 ++++++++++++++++++++++++++++++++ net/bluetooth/hci_core.c | 20 +++++++++++++++++++- net/bluetooth/hci_event.c | 32 +++++++++++++++++++++++++------- net/bluetooth/hci_request.c | 16 +++++++++++++--- net/bluetooth/mgmt.c | 10 +++++++++- net/bluetooth/smp.c | 11 +++++++++++ 7 files changed, 111 insertions(+), 12 deletions(-)include/net/bluetooth/hci.h +2
-0@@ -1617,6 +1617,8 @@ struct hci_rp_le_read_resolv_list_size {#define HCI_OP_LE_SET_ADDR_RESOLV_ENABLE 0x202d +#define HCI_OP_LE_SET_RPA_TIMEOUT 0x202e +#define HCI_OP_LE_READ_MAX_DATA_LEN 0x202fstruct hci_rp_le_read_max_data_len {__u8 status;net/bluetooth/hci_conn.c +0
-0@@ -830,6 +830,14 @@ static void hci_req_add_le_create_conn(struct hci_request *req,
net/bluetooth/hci_conn.c static void hci_req_add_le_create_conn(struct hci_request *req, · line 840
Quoted 10 lines
net/bluetooth/hci_conn.c +8
-0plen =sizeof (*cp); +/* If the own_addr_type is 0x03 or 0x02 and peer is present +* in Kernel IRK list, but not in resolving add the peer to +* the resolving list. +*/ +if (own_addr_type ==0x03 || own_addr_type ==0x02 ) +hci_req_check_and_update_resolving_list (hdev, conn->dst_type, + &conn->dst); +
I am failing to understand this part. The address should have been in the resolving list already. Either via Load Identify Resolving Keys mgmt command or from the SMP pairing. Doing that at time when calling Create Connection seems wrong. It should be independent from that.
Quoted 35 lines
net/bluetooth/hci_conn.c +16
-0if (scan_1m (hdev)) { cp->phys |=LE_SCAN_PHY_1M ;set_ext_conn_params (conn, p);@@ -866,6 +874,14 @@ static void hci_req_add_le_create_conn(struct hci_request *req, cp.scan_interval =cpu_to_le16 (hdev->le_scan_interval); cp.scan_window = cp.scan_interval; +/* If the own_addr_type is 0x03 or 0x02 and peer is present +* in Kernel IRK list, but not in resolving add the peer to +* the resolving list. +*/ +if (own_addr_type ==0x03 || own_addr_type ==0x02 ) +hci_req_check_and_update_resolving_list (hdev, conn->dst_type, + &conn->dst); +bacpy (&cp.peer_addr, &conn->dst); cp.peer_addr_type = conn->dst_type; cp.own_address_type = own_addr_type;@@ -901,6 +917,14 @@ static void hci_req_directed_advertising(struct hci_request *req, &own_addr_type, &random_addr) < 0)return ; +/* If the own_addr_type is 0x03 or 0x02 and peer is present +* in Kernel IRK list, but not in resolving add the peer to +* the resolving list. +*/ +if (own_addr_type ==0x03 || own_addr_type ==0x03 ) +hci_req_check_and_update_resolving_list (hdev, conn->dst_type, + &conn->dst); +memset (&cp,0 ,sizeof (cp)); cp.evt_properties =cpu_to_le16 (LE_LEGACY_ADV_DIRECT_IND );@@ -959,6 +983,14 @@ static void hci_req_directed_advertising(struct hci_request *req, &own_addr_type) < 0)
net/bluetooth/hci_conn.c static void hci_req_directed_advertising(struct hci_request *req, · line 993
Quoted 10 lines
net/bluetooth/hci_conn.c +8
-0return ; +/* If the own_addr_type is 0x03 or 0x02 and peer is present +* in Kernel IRK list, but not in resolving add the peer to +* the resolving list. +*/ +if (own_addr_type ==0x03 || own_addr_type ==0x02 ) +hci_req_check_and_update_resolving_list (hdev, conn->dst_type, + &conn->dst); +
We just keep repeating this code. It doesn’t look right.
Quoted 8 lines
net/bluetooth/hci_conn.c +0
-0memset (&cp,0 ,sizeof (cp));/* Some controllers might reject command if intervals are not net/bluetooth/hci_core.c +0
-0
net/bluetooth/hci_core.c static int hci_init3_req(struct hci_request *req, unsigned long opt) · line 719
Quoted 8 lines
net/bluetooth/hci_core.c +1
-1@@ -713,7 +713,7 @@ static int hci_init3_req(struct hci_request *req, unsigned long opt)/* If the controller supports the LE Extended Create Connection * command, enable the corresponding event. */ -if (use_ext_conn (hdev)) +if (use_ext_conn (hdev) || hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) events[1 ] |=0x02 ;/* LE Enhanced Connection * Complete */
I do not get this change.
net/bluetooth/hci_core.c static int hci_init3_req(struct hci_request *req, unsigned long opt) · line 766
Quoted 6 lines
net/bluetooth/hci_core.c +2
-0@@ -762,6 +762,24 @@ static int hci_init3_req(struct hci_request *req, unsigned long opt)hci_req_add (req,HCI_OP_LE_CLEAR_RESOLV_LIST ,0 ,NULL ); } +if (hdev->commands[35 ] &0x02 ) { +u8 addr_resoln_enable =0x01 ;
I don’t know what resoln is suppose to stand for.
Quoted 6 lines
net/bluetooth/hci_core.c +6
-0+ +/* Enable Address Resolution in controller */ +hci_req_add (req,HCI_OP_LE_SET_ADDR_RESOLV_ENABLE , +sizeof (addr_resoln_enable), + &addr_resoln_enable); + }
net/bluetooth/hci_core.c static int hci_init3_req(struct hci_request *req, unsigned long opt) · line 782
Quoted 10 lines
net/bluetooth/hci_core.c +10
-0+ +if (hdev->commands[35 ] &0x40 ) { +__le16 rpa_timeout =cpu_to_le16 (hdev->rpa_timeout); + +/* Set RPA timeout */ +hci_req_add (req,HCI_OP_LE_SET_RPA_TIMEOUT ,2 , + &rpa_timeout); + + } +
I think you should be adding RPA timeout handling in a separate patch. It is also important to note that there is a debugfs setting that should be capable of changing this value at runtime.
Quoted 8 lines
net/bluetooth/hci_core.c +0
-0if (hdev->le_features[0 ] &HCI_LE_DATA_LEN_EXT ) {/* Read LE Maximum Data Length */ hci_req_add (req,HCI_OP_LE_READ_MAX_DATA_LEN ,0 ,NULL );net/bluetooth/hci_event.c +0
-0
net/bluetooth/hci_event.c static void hci_cc_set_adv_param(struct hci_dev *hdev, struct sk_buff *skb) · line 1691
Quoted 9 lines
net/bluetooth/hci_event.c +5
-1@@ -1684,7 +1684,11 @@ static void hci_cc_set_adv_param(struct hci_dev *hdev, struct sk_buff *skb)return ;hci_dev_lock (hdev); - hdev->adv_addr_type = cp->own_address_type; + +if (cp->own_address_type ==ADDR_LE_DEV_PUBLIC ) + hdev->adv_addr_type =ADDR_LE_DEV_PUBLIC ; +else + hdev->adv_addr_type =ADDR_LE_DEV_RANDOM ;
I don’t get this change.
Quoted 4 lines
net/bluetooth/hci_event.c +0
-0hci_dev_unlock (hdev); }
net/bluetooth/hci_event.c static void hci_cc_set_ext_adv_param(struct hci_dev *hdev, struct sk_buff *skb) · line 1716
Quoted 10 lines
net/bluetooth/hci_event.c +6
-1@@ -1704,7 +1708,12 @@ static void hci_cc_set_ext_adv_param(struct hci_dev *hdev, struct sk_buff *skb)return ;hci_dev_lock (hdev); - hdev->adv_addr_type = cp->own_addr_type; + +if (cp->own_addr_type ==ADDR_LE_DEV_PUBLIC ) + hdev->adv_addr_type =ADDR_LE_DEV_PUBLIC ; +else + hdev->adv_addr_type =ADDR_LE_DEV_RANDOM ; +
Same as above. I fail to see why this is important.
Quoted 7 lines
net/bluetooth/hci_event.c +0
-0if (!hdev->cur_adv_instance) {/* Store in hdev for instance 0 */ hdev->adv_tx_power = rp->tx_power;@@ -2267,11 +2276,13 @@ static void cs_le_create_conn(struct hci_dev *hdev, bdaddr_t *peer_addr,* is needed for SMP. These values will not change during the * lifetime of the connection. */
net/bluetooth/hci_event.c static void cs_le_create_conn(struct hci_dev *hdev, bdaddr_t *peer_addr, · line 2285
Quoted 10 lines
net/bluetooth/hci_event.c +5
-3- conn->init_addr_type = own_address_type; -if (own_address_type ==ADDR_LE_DEV_RANDOM ) +if (own_address_type ==0x03 || own_address_type ==0x01 ) { + conn->init_addr_type =ADDR_LE_DEV_RANDOM ;bacpy (&conn->init_addr, &hdev->random_addr); -else + }else { + conn->init_addr_type =ADDR_LE_DEV_PUBLIC ;bacpy (&conn->init_addr, &hdev->bdaddr); + }
This requires an appropriate comment change as well.
Quoted 14 lines
net/bluetooth/hci_event.c +2
-0conn->resp_addr_type = peer_addr_type;bacpy (&conn->resp_addr, peer_addr);@@ -5141,6 +5152,8 @@ static void le_conn_complete_evt(struct hci_dev *hdev, u8 status, } } +hci_req_check_and_update_resolving_list (hdev, (conn->dst_type), &conn->dst); +unlock:
hci_update_background_scan(hdev); hci_dev_unlock(hdev);net/bluetooth/hci_event.c +0
-0@@ -5365,8 +5378,13 @@ static void process_adv_report(struct hci_dev *hdev, u8 type, bdaddr_t *bdaddr,/* Only resolvable random addresses are valid for these
net/bluetooth/hci_event.c static void process_adv_report(struct hci_dev *hdev, u8 type, bdaddr_t *bdaddr, · line 5386
Quoted 10 lines
net/bluetooth/hci_event.c +6
-2* kind of reports and others can be ignored. */ -if (!hci_bdaddr_is_rpa (direct_addr, direct_addr_type)) -return ; +if (!hci_bdaddr_is_rpa (direct_addr, direct_addr_type)) { +/* Controller with LL Privacy Supported tend to resolve +* the RPA. In that case we should ignore this condition +*/ +if ( !(hdev->le_features[0 ] &HCI_LE_LL_PRIVACY )) +return ;
Please obey the coding style.
And just checking the existence of feature bit is not enough. We also need to check that the feature is active and actually used in this case.
Quoted 9 lines
net/bluetooth/hci_event.c +1
-0+ }/* If the controller is not using resolvable random * addresses, then this report can be ignored. net/bluetooth/hci_request.c +0
-0
net/bluetooth/hci_request.c int hci_get_random_address(struct hci_dev *hdev, bool require_privacy, · line 1924
Quoted 10 lines
net/bluetooth/hci_request.c +6
-1@@ -1916,7 +1916,12 @@ int hci_get_random_address(struct hci_dev *hdev, bool require_privacy,if (use_rpa) {int to; - *own_addr_type =ADDR_LE_DEV_RANDOM ; +if (hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) { +/*if there is no resolving list used in the controller +* use the bd_addr from latest Le_set_random_address*/ + *own_addr_type =0x03 ; + }else + *own_addr_type =ADDR_LE_DEV_RANDOM ;
I really don’t see how this is going to work. Local privacy can off and maybe remote device don’t use privacy. As noted above the le_features bit is not an indicate if LL Privacy is in use.
In addition, please get the coding style right.
net/bluetooth/hci_request.c int __hci_req_setup_ext_adv_instance(struct hci_request *req, u8 instance) · line 2082
Quoted 9 lines
net/bluetooth/hci_request.c +1
-1if (adv_instance) { if (!adv_instance->rpa_expired &&@@ -2074,7 +2079,7 @@ int __hci_req_setup_ext_adv_instance(struct hci_request *req, u8 instance)hci_req_add (req,HCI_OP_LE_SET_EXT_ADV_PARAMS ,sizeof (cp), &cp); - if (own_addr_type == ADDR_LE_DEV_RANDOM && + if (((own_addr_type == ADDR_LE_DEV_RANDOM) || (own_addr_type == 0x03)) &&
The extra (x == y) are not needed and it might be a good idea to introduce constants for 0x02 and 0x03 early on.
Quoted 4 lines
net/bluetooth/hci_request.c +0
-0bacmp(&random_addr, BDADDR_ANY)) {struct hci_cp_le_set_adv_set_rand_addr cp;
net/bluetooth/hci_request.c int hci_update_random_address(struct hci_request *req, bool require_privacy, · line 2354
Quoted 10 lines
net/bluetooth/hci_request.c +6
-1@@ -2341,7 +2346,12 @@ int hci_update_random_address(struct hci_request *req, bool require_privacy,if (use_rpa) {int to; - *own_addr_type =ADDR_LE_DEV_RANDOM ; +if (hdev->le_features[0 ] &HCI_LE_LL_PRIVACY ) { +/*if there is no resolving list used in the controller +* use the bd_addr from latest Le_set_random_address*/ + *own_addr_type =0x03 ; + }else + *own_addr_type =ADDR_LE_DEV_RANDOM ;
I really need to stop reviewing patches where the coding style mistakes are this many.
Quoted 55 lines
net/bluetooth/hci_request.c +0
-0if (!hci_dev_test_and_clear_flag (hdev,HCI_RPA_EXPIRED ) && !bacmp (&hdev->random_addr, &hdev->rpa))net/bluetooth/mgmt.c +3
-1@@ -5084,8 +5084,10 @@ static int set_debug_keys(struct sock *sk, struct hci_dev *hdev,if (err <0 ) goto unlock; -if (changed) +if (changed) { err =new_settings (hdev, sk); +hci_req_update_resolving_list_local_irk (hdev); + }unlock:
hci_dev_unlock(hdev);net/bluetooth/mgmt.c +6
-0@@ -5227,6 +5229,8 @@ static int load_irks(struct sock *sk, struct hci_dev *hdev, void *cp_data, BDADDR_ANY); } +/* Load the resolving list with entries from IRK list.*/ +hci_load_resolving_list (hdev);hci_dev_set_flag (hdev,HCI_RPA_RESOLVING ); err =mgmt_cmd_complete (sk, hdev->id,MGMT_OP_LOAD_IRKS ,0 ,NULL ,0 );@@ -7213,6 +7217,10 @@ void mgmt_power_on(struct hci_dev *hdev, int err)if (match.sk)sock_put (match.sk); +/* If load _irk was called when controller was powered down, +* then the resolving list has to be updated now. +*/ +hci_load_resolving_list (hdev);hci_dev_unlock (hdev); }net/bluetooth/smp.c +2
-0@@ -2425,6 +2425,8 @@ int smp_cancel_and_remove_pairing(struct hci_dev *hdev, bdaddr_t *bdaddr,struct smp_chan *smp;int err; +hci_req_del_from_resolving_list (hdev, addr_type, bdaddr); + err =hci_remove_ltk (hdev, bdaddr, addr_type);hci_remove_irk (hdev, bdaddr, addr_type);@@ -2619,6 +2621,15 @@ static int smp_cmd_ident_addr_info(struct l2cap_conn *conn, smp->remote_irk =hci_add_irk (conn->hcon->hdev, &smp->id_addr, smp->id_addr_type, smp->irk, &rpa);
net/bluetooth/smp.c static int smp_cmd_ident_addr_info(struct l2cap_conn *conn, · line 2632
Quoted 10 lines
net/bluetooth/smp.c +9
-0+/* Add the device to the resolving list.*/ +hci_req_update_resolving_list (conn->hcon->hdev, + smp->remote_irk->addr_type, + &smp->remote_irk->bdaddr, + smp->remote_irk->val); + + +/*TODO: Check remote device Supports LL Privacy. If not set the device +* with IRK to Device-privacy Mode - Should be discussed - Based on */
I rather have this fixed instead of keeping a TODO in here.
Quoted 3 lines
distribute:
if (!(smp->remote_key_dist & KEY_DIST_MASK)) smp_distribute_keys(smp);
Regards
Marcel
PATCH 7/7 Bluetooth LL Privacy update White list
In reply to Sathish NarsimmanUpdate white list filter policy of if LL Privacy is enabled When Address Resolution is enabled and device is already in the white list then follow the filter policy
--- net/bluetooth/hci_request.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-)
net/bluetooth/hci_request.c +20 -2
@@ -711,7 +711,16 @@ static int add_to_white_list(struct hci_request *req,/* White list can not be used with RPAs */ if (!allow_rpa &&hci_find_irk_by_addr (hdev, ¶ms->addr, params->addr_type)) { -return -1 ; +/* White list can not be used with RPAs if address +* resolution is disabled in the controller +*/ +if (!hci_dev_test_flag (hdev,HCI_LL_RPA_RESOLUTION )) +return -1 ; + +/* if the peer device is not present in the resolving list */ +if (!hci_bdaddr_list_lookup_with_irk (&hdev->le_resolv_list, &b->bdaddr, + b->bdaddr_type)) +return -1 ; }/* During suspend, only wakeable devices can be in whitelist */
@@ -768,7 +777,16 @@ static u8 update_white_list(struct hci_request *req)/* White list can not be used with RPAs */ if (!allow_rpa &&hci_find_irk_by_addr (hdev, &b->bdaddr, b->bdaddr_type)) { -return 0x00 ; +/* White list can not be used with RPAs if address +* resolution is disabled in the controller +*/ +if (!hci_dev_test_flag (hdev,HCI_LL_RPA_RESOLUTION )) +return 0x00 ; + +/* if the peer device is not present in the resolving list */ +if (!hci_bdaddr_list_lookup_with_irk (&hdev->le_resolv_list, &b->bdaddr, + b->bdaddr_type)) +return 0x00 ; } num_entries++;
Signature · 1 line
2.17.1
kbuild test robot
build botRe: [PATCH 7/7] Bluetooth LL Privacy update White list
Thank you for the patch! Yet something to improve:
Hi Sathish,
Thank you for the patch! Yet something to improve:
[auto build test ERROR on bluetooth-next/master]
[also build test ERROR on next-20200312]
[cannot apply to v5.6-rc5]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system. BTW, we also suggest to use '--base' option to specify the base tree in git format-patch, please see https://stackoverflow.com/a/37406982]
url: https://github.com/0day-ci/linux/commits/Sathish-Narsimman/LE-LL-Priavcy-support-enabled/20200313-033706 base: https://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next.git master config: parisc-allyesconfig (attached as .config) compiler: hppa-linux-gcc (GCC) 9.2.0
reproduce:
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# save the attached .config to linux build tree
GCC_VERSION=9.2.0 make.cross ARCH=parisc
If you fix the issue, kindly add following tag
All errors (new ones prefixed by >>):
net//bluetooth/hci_request.c: In function 'add_to_white_list':
>> net//bluetooth/hci_request.c:721:64: error: 'b' undeclared (first use in this function); did you mean 'mb'?
721 | if (!hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, &b->bdaddr,
| ^
| mb net//bluetooth/hci_request.c:721:64: note: each undeclared identifier is reported only once for each function it appears in
vim +721 net//bluetooth/hci_request.c
693
694 /* Adds connection to white list if needed. On error, returns -1. */
695 static int add_to_white_list(struct hci_request *req,
696 struct hci_conn_params *params, u8 *num_entries,
697 bool allow_rpa)
698 {
699 struct hci_cp_le_add_to_white_list cp;
700 struct hci_dev *hdev = req->hdev;
701
702 /* Already in white list */
703 if (hci_bdaddr_list_lookup(&hdev->le_white_list, ¶ms->addr,
704 params->addr_type))
705 return 0;
706
707 /* Select filter policy to accept all advertising */
708 if (*num_entries >= hdev->le_white_list_size)
709 return -1;
710
711 /* White list can not be used with RPAs */
712 if (!allow_rpa &&
713 hci_find_irk_by_addr(hdev, ¶ms->addr, params->addr_type)) {
714 /* White list can not be used with RPAs if address
715 * resolution is disabled in the controller
716 */
717 if (!hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION))
718 return -1;
719
720 /* if the peer device is not present in the resolving list */
> 721 if (!hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, &b->bdaddr,
722 b->bdaddr_type))
723 return -1;
724 }
725
726 /* During suspend, only wakeable devices can be in whitelist */
727 if (hdev->suspended && !params->wakeable)
728 return 0;
729
730 *num_entries += 1;
731 cp.bdaddr_type = params->addr_type;
732 bacpy(&cp.bdaddr, ¶ms->addr);
733
734 bt_dev_dbg(hdev, "Add %pMR (0x%x) to whitelist", &cp.bdaddr,
735 cp.bdaddr_type);
736 hci_req_add(req, HCI_OP_LE_ADD_TO_WHITE_LIST, sizeof(cp), &cp);
737
738 return 0;
739 }
740
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org
kbuild test robot
build botRe: [PATCH 7/7] Bluetooth LL Privacy update White list
In reply to Sathish NarsimmanThank you for the patch! Yet something to improve:
Hi Sathish,
Thank you for the patch! Yet something to improve:
[auto build test ERROR on bluetooth-next/master]
[also build test ERROR on next-20200312]
[cannot apply to bluetooth/master v5.6-rc5]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system. BTW, we also suggest to use '--base' option to specify the base tree in git format-patch, please see https://stackoverflow.com/a/37406982]
url: https://github.com/0day-ci/linux/commits/Sathish-Narsimman/LE-LL-Priavcy-support-enabled/20200313-033706 base: https://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next.git master config: x86_64-randconfig-s0-20200313 (attached as .config) compiler: gcc-5 (Ubuntu 5.5.0-12ubuntu1) 5.5.0 20171010 reproduce: # save the attached .config to linux build tree make ARCH=x86_64
If you fix the issue, kindly add following tag
All errors (new ones prefixed by >>):
net//bluetooth/hci_request.c: In function 'add_to_white_list':
>> net//bluetooth/hci_request.c:721:64: error: 'b' undeclared (first use in this function)
if (!hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, &b->bdaddr,
^ net//bluetooth/hci_request.c:721:64: note: each undeclared identifier is reported only once for each function it appears in
vim +/b +721 net//bluetooth/hci_request.c
693
694 /* Adds connection to white list if needed. On error, returns -1. */
695 static int add_to_white_list(struct hci_request *req,
696 struct hci_conn_params *params, u8 *num_entries,
697 bool allow_rpa)
698 {
699 struct hci_cp_le_add_to_white_list cp;
700 struct hci_dev *hdev = req->hdev;
701
702 /* Already in white list */
703 if (hci_bdaddr_list_lookup(&hdev->le_white_list, ¶ms->addr,
704 params->addr_type))
705 return 0;
706
707 /* Select filter policy to accept all advertising */
708 if (*num_entries >= hdev->le_white_list_size)
709 return -1;
710
711 /* White list can not be used with RPAs */
712 if (!allow_rpa &&
713 hci_find_irk_by_addr(hdev, ¶ms->addr, params->addr_type)) {
714 /* White list can not be used with RPAs if address
715 * resolution is disabled in the controller
716 */
717 if (!hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION))
718 return -1;
719
720 /* if the peer device is not present in the resolving list */
> 721 if (!hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, &b->bdaddr,
722 b->bdaddr_type))
723 return -1;
724 }
725
726 /* During suspend, only wakeable devices can be in whitelist */
727 if (hdev->suspended && !params->wakeable)
728 return 0;
729
730 *num_entries += 1;
731 cp.bdaddr_type = params->addr_type;
732 bacpy(&cp.bdaddr, ¶ms->addr);
733
734 bt_dev_dbg(hdev, "Add %pMR (0x%x) to whitelist", &cp.bdaddr,
735 cp.bdaddr_type);
736 hci_req_add(req, HCI_OP_LE_ADD_TO_WHITE_LIST, sizeof(cp), &cp);
737
738 return 0;
739 }
740
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org
kbuild test robot
build botRe: [PATCH 7/7] Bluetooth LL Privacy update White list
In reply to Sathish NarsimmanThank you for the patch! Perhaps something to improve:
Hi Sathish,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on bluetooth-next/master]
[also build test WARNING on next-20200312]
[cannot apply to bluetooth/master v5.6-rc5]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system. BTW, we also suggest to use '--base' option to specify the base tree in git format-patch, please see https://stackoverflow.com/a/37406982]
url: https://github.com/0day-ci/linux/commits/Sathish-Narsimman/LE-LL-Priavcy-support-enabled/20200313-033706 base: https://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next.git master config: i386-randconfig-d001-20200312 (attached as .config) compiler: gcc-7 (Debian 7.5.0-5) 7.5.0 reproduce: # save the attached .config to linux build tree make ARCH=i386
If you fix the issue, kindly add following tag
All warnings (new ones prefixed by >>):
In file included from include/linux/export.h:43:0,
from include/linux/linkage.h:7,
from include/linux/kernel.h:8,
from include/linux/list.h:9,
from include/linux/rculist.h:10,
from include/linux/sched/signal.h:5,
from net/bluetooth/hci_request.c:24:
net/bluetooth/hci_request.c: In function 'add_to_white_list':
net/bluetooth/hci_request.c:721:64: error: 'b' undeclared (first use in this function); did you mean 'mb'?
if (!hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, &b->bdaddr,
^
include/linux/compiler.h:58:52: note: in definition of macro '__trace_if_var'
#define __trace_if_var(cond) (__builtin_constant_p(cond) ? (cond) : __trace_if_value(cond))
^~~~
>> net/bluetooth/hci_request.c:721:3: note: in expansion of macro 'if'
if (!hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, &b->bdaddr,
^~
net/bluetooth/hci_request.c:721:64: note: each undeclared identifier is reported only once for each function it appears in
if (!hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, &b->bdaddr,
^
include/linux/compiler.h:58:52: note: in definition of macro '__trace_if_var'
#define __trace_if_var(cond) (__builtin_constant_p(cond) ? (cond) : __trace_if_value(cond))
^~~~
>> net/bluetooth/hci_request.c:721:3: note: in expansion of macro 'if'
if (!hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, &b->bdaddr,
^~vim +/if +721 net/bluetooth/hci_request.c
693
694 /* Adds connection to white list if needed. On error, returns -1. */
695 static int add_to_white_list(struct hci_request *req,
696 struct hci_conn_params *params, u8 *num_entries,
697 bool allow_rpa)
698 {
699 struct hci_cp_le_add_to_white_list cp;
700 struct hci_dev *hdev = req->hdev;
701
702 /* Already in white list */
703 if (hci_bdaddr_list_lookup(&hdev->le_white_list, ¶ms->addr,
704 params->addr_type))
705 return 0;
706
707 /* Select filter policy to accept all advertising */
708 if (*num_entries >= hdev->le_white_list_size)
709 return -1;
710
711 /* White list can not be used with RPAs */
712 if (!allow_rpa &&
713 hci_find_irk_by_addr(hdev, ¶ms->addr, params->addr_type)) {
714 /* White list can not be used with RPAs if address
715 * resolution is disabled in the controller
716 */
717 if (!hci_dev_test_flag(hdev, HCI_LL_RPA_RESOLUTION))
718 return -1;
719
720 /* if the peer device is not present in the resolving list */
> 721 if (!hci_bdaddr_list_lookup_with_irk(&hdev->le_resolv_list, &b->bdaddr,
722 b->bdaddr_type))
723 return -1;
724 }
725
726 /* During suspend, only wakeable devices can be in whitelist */
727 if (hdev->suspended && !params->wakeable)
728 return 0;
729
730 *num_entries += 1;
731 cp.bdaddr_type = params->addr_type;
732 bacpy(&cp.bdaddr, ¶ms->addr);
733
734 bt_dev_dbg(hdev, "Add %pMR (0x%x) to whitelist", &cp.bdaddr,
735 cp.bdaddr_type);
736 hci_req_add(req, HCI_OP_LE_ADD_TO_WHITE_LIST, sizeof(cp), &cp);
737
738 return 0;
739 }
740
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org
Hi
A gentle reminder to review the patches
Quoted 38 lines · Sathish Narsimman
With this LE LL Priavcy feature available if BT controller supports LL Privacy when privacy is enabled.
The patches are verified in below case.
Pairing - where the IRK is added to device Resolving List after identity info & IRK is shared in SMP.
Unpair - Delete the resolving entry from the Device resolving list Local IRK change - Whenever there is change in local IRK (privacy) the resolving list entries in the device is updated.
Whitelist - filter policies for whitelist is updated based onLL_adressresolution support.
Reload Resolving list - device reset/sysmte reboot the Resolving list reloaded in to the device.Sathish Narsimman (7):
Bluetooth: LL Privacy Delete Store Resolving list
Bluetooth: Add device to Resolving List
Bluetooth: Update the Resolving list when Local IRK changed
Bluetooth:LL_PRIVACYre-load resolving list
Bluetooth: LL Privacy Check and Update the Resolving List
Bluetooth: Enable LL Privacy
Bluetooth LL Privacy update White listinclude/net/bluetooth/hci.h | 3 + include/net/bluetooth/hci_core.h | 7 + net/bluetooth/hci_conn.c | 32 ++++ net/bluetooth/hci_core.c | 20 ++- net/bluetooth/hci_event.c | 32 +++- net/bluetooth/hci_request.c | 251 ++++++++++++++++++++++++++++++- net/bluetooth/hci_request.h | 2 + net/bluetooth/mgmt.c | 10 +- net/bluetooth/smp.c | 11 ++ 9 files changed, 354 insertions(+), 14 deletions(-)--
2.17.1
Hi Satish,
Looking through the patches in this series, it looks like you are adding the IRK for all connected devices and I'm not sure that's the best method.
The resolv list seems to be useful in the same way as the le
whitelist: to filter incoming advertisements for devices we care about.
Thus, to simplify your design, could we not do the same thing as the le whitelist:
* Update the resolv list anytime passive/background scan is being enabled
* Only keep entries in the resolv list that are part of the pend_le_conn or pend_le_report lists
Then, you would only need to update the resolv list in hci_req_add_le_passive_scan and any IRK changes would just disable passive scan, remove IRKs if existing and re-enable passive scan (which would add it back with the new one).
Thanks,
Abhishek
Quoted 44 lines · Sathish Narasimman
Hi
A gentle reminder to review the patches
Quoted 38 lines · Sathish Narsimman
With this LE LL Priavcy feature available if BT controller supports LL Privacy when privacy is enabled.
The patches are verified in below case.
Pairing - where the IRK is added to device Resolving List after identity info & IRK is shared in SMP.
Unpair - Delete the resolving entry from the Device resolving list Local IRK change - Whenever there is change in local IRK (privacy) the resolving list entries in the device is updated.
Whitelist - filter policies for whitelist is updated based onLL_adressresolution support.
Reload Resolving list - device reset/sysmte reboot the Resolving list reloaded in to the device.Sathish Narsimman (7):
Bluetooth: LL Privacy Delete Store Resolving list
Bluetooth: Add device to Resolving List
Bluetooth: Update the Resolving list when Local IRK changed
Bluetooth:LL_PRIVACYre-load resolving list
Bluetooth: LL Privacy Check and Update the Resolving List
Bluetooth: Enable LL Privacy
Bluetooth LL Privacy update White listinclude/net/bluetooth/hci.h | 3 + include/net/bluetooth/hci_core.h | 7 + net/bluetooth/hci_conn.c | 32 ++++ net/bluetooth/hci_core.c | 20 ++- net/bluetooth/hci_event.c | 32 +++- net/bluetooth/hci_request.c | 251 ++++++++++++++++++++++++++++++- net/bluetooth/hci_request.h | 2 + net/bluetooth/mgmt.c | 10 +- net/bluetooth/smp.c | 11 ++ 9 files changed, 354 insertions(+), 14 deletions(-)--
2.17.1
Marcel Holtmann
maintainerHi Abhishek,
Quoted 13 lines
Looking through the patches in this series, it looks like you are adding the IRK for all connected devices and I'm not sure that's the best method.
The resolv list seems to be useful in the same way as the le
whitelist: to filter incoming advertisements for devices we care about.Thus, to simplify your design, could we not do the same thing as the le whitelist:
* Update the resolv list anytime passive/background scan is being enabled
* Only keep entries in the resolv list that are part of thepend_le_connorpend_le_reportlists
Quoted 4 lines
Then, you would only need to update the resolv list in
hci_req_add_le_passive_scanand any IRK changes would just disable passive scan, remove IRKs if existing and re-enable passive scan (which would add it back with the new one).
so I have been looking at this again and yes, we should just put IRKs in the resolving list for devices that we also put in the whitelist. And we only use the whitelist for background scanning.
This means I would only focus on enabling background scanning. For everything else, we can just let the host do the resolving.
Enable passive scanning -> Enable resolving list if privacy device in whitelist -> Set Scan Parameters -> Set Scan Enable Disable passive scanning -> Set Scan Disable -> Disable resolving list if enabled
And when updating the whitelist, also add update the resolving list with needed entries for the whitelist. This means if the privacy enabled device goes into the whitelist, add the IRK to the resolving list. Remove all no longer needed IRKs.
Regards
Marcel