Changes the behavior of the message cache to optimize for cache
capacity. Previously, the message cache's primary function was to avoid
decrypting messages multiple times, although the cache's main function
in the spec is to avoid message rebroadcasting. Optimizing for minimal
decryption causes us to fill the network cache faster, which in turn
causes more cache misses, potentially outweighing the advantage.
Now stores src + seq in message cache instead of field hash value. This
cuts cache size in two, while including more of the sequence number than
before.
Adds messages to the cache only after the packet is successfully
decrypted. This reduces noise in the cache, and ensures that no
invalid deobfuscations are added.
Additionally, this fixes a bug where multiple calls to net_decrypt with
the same packet failed, as the message cache found its own entry from
the previous call.
Implements several changes to the transport layer segmented tx to
improve group message performance:
- Moves retransmit counter to tx context instead of per packet. As every
unacked packet is sent every retransmit, the retransmit counters would
be the same in each segment. This makes it easier to control progress.
- Delays the scheduling of the retransmit until the completion of the
last segment by adding a seg_pending counter. This is essentially the
same as the old behavior, except that the old behavior might retrigger
the sending before all segments are finished if the advertising is
slow.
- Allows the group transmits to stop as soon as all retransmits have
been exhausted, instead of timing out waiting for acks that won't
come. This allows group tx to finish without error.
- Fixes a bug where a failed TX would block IV update.
- Cancels any pending transmissions of acked segments.
- Reduces log level for several common group tx scenarios that aren't
erronous.
Backport of https://github.com/apache/mynewt-nimble/pull/724
Mesh spec 1.0.1 changes proxy disabling behavior to only affect the
relaying from proxy nodes. Previously, disabling proxy would shut down
all proxy and node activity.
Tweaks from the original commit:
- Removed redundant call to bt_mesh_adv_update() in gatt_proxy_set()
- Removed invalid ref to 4.2.11.1 in node_identity_set()
---
According to Mesh Profile Spec 1.0.1, Section 4.2.11:
"If the Proxy feature is disabled, a GATT client device can connect
over GATT to that node for configuration and control. Messages from
the GATT bearer are not relayed to the advertising bearer."
Moreover some notes have been removed from the spec compared to
version 1.0:
Mesh Profile Spec 1.0, Section 4.2.11:
"Upon transition from GATT Proxy state 0x01 to GATT Proxy state 0x00
the GATT Bearer Server shall disconnect all GATT Bearer Clients."
"The Configuration Client should turn off the Proxy state as the last
step in the configuration process."
Mesh Profile Spec 1.0, Section 4.2.11.1:
"When the GATT Proxy state is set to 0x00, the Node Identity state
for all subnets shall be set to 0x00 and shall not be changed."
Until now the choice of reliable sending (segmented messages with
acks) was implicitly dependent on the size of the payload. Add a new
member to the bt_mesh_model_pub to force using segment acks even when
the payload would fit a single unsegmented message.
When PB-GATT support has been enabled the provisioning code "borrows"
the buffer from the proxy code. However, the way that initialization
was happening the proxy buffers were initialized only after
provisioning initialization, resulting in a corrupted buffer with
buf->data pointing to NULL. Reorder the initialization calls so that
proxy is done first and provisioning only after it.
Allow models to skip a periodic publish interval by returning an error
from the publish update callback.
Previously, an error return from publish update would cancel periodic
publishing. This can't be recovered from, and as such, no valid model
implementation could return an error from this callback, and there was
no way to skip a periodic publish.
The function bt_mesh_ctl_send() used to support maximum length of
11 bytes. The segmentation complies with the BLE Mesh Standard.
The ack is disabled in case of non unicast address.
1. fix the bug for softAP update second channel wrong
2. change wifi_scan_time_t from union to struct
3. query country code wrong when policy change from manual to auto
4. fix the timeout issue of EAPOL four-way handshake
5. fix the crash issue when fetch peer after delete the cur_peer
It was observed that after sending credentials, if the first
Wi-Fi conection attempt fails, the Wi-Fi provisioning managager
ignores subsequent Wi-Fi events. So, even if the device eventually
connects to the AP, the provisioning stays on indefinitely.
Changing a check so that subsequent events are also captured and
provisioning finishes successfully.
1. Buffers for SAE messages are not freed after the handshake.
This causes memory leak, free buffers after SAE handshake.
2. SAE global data is not freed until the next WPA3 connection
takes place, holding up heap space without reason. Free theis
data after SAE handshake is complete or event fails.
3. Update wifi lib which includes memory leak fix during BIP
encryption/decryption operations.
When fast provisioning is enabled, Provisioner shall not
ignore messages from the nodes whose addresses are not in
the provisioning database. Because other nodes which are
not provisioned by the Primary Provisioner will send node
address messages to the Primary Provisioner.
Previously only mesh node info is supported to be stored
in flash. So when trying to reset the node, we only need
to judge if the BLE_MESH_VALID flag is set.
Currently we support storing both node & Provisioner info
in flash, when trying to erase the node info from flash,
the BLE_MESH_NODE flag will be checked. So we need to set
bt_mesh.flags to 0 when all the erase operations are done.
EAP reauth frames are dropped at various stages due to current
implementation of WPA2 ENT states and EAP SM init/deinit logic.
Route EAPOL frames based on EAP pkt type and maintain EAP SM
to facilitate EAP re-authentication process.
Added WPA3 Testcases support for -
1. Anti-Clogging Token Request support
2. Return correct status from SAE modules for invalid scenarios
3. Add PMK Caching support for WPA3
wifi lib includes fixes for below PMF Certification issues -
1. Check return status of decrypt operation. Fixes 5.3.3.1.
2. Allow PMF negotiation for WPA2-Enterprise. Fixes 5.3.3.2, 5.3.3.4.
3. Add NULL check on key before encrypting PMF, fixes crash.
1. Add changes in 4-way handshake path to allow SAE key mgmt.
2. Support for configuring WAP3 at init time, added Kconfig option.
3. Handle and propagate error conditions properly.
Under WPA3-Personal, SAE authentication is used to derive PMK
which is more secure and immune to offline dictionary attacks.
1. Add modules to generate SAE commit/confirm for the handshake
2. Add modules that build and parse SAE data in Auth frames
3. Add WPA3 association and key mgmt definitions
4. Invert y-bit while solving for ECC co-ordinate -
Once an X co-ordinate is obtained, solving for Y co-ordinate
using an elliptical curve equation results in 2 possible values,
Y and (P - Y), where p is the prime number. The co-ordinates are
used for deriving keys in SAE handshake. As par the 802.11 spec
if LSB of X is same as LSB of Y then Y is chosen, (P - Y) otherwise.
This is not what is implemented, so fix this behavior to obtain the
correct Y co-ordinate.
This change ports SAE(Simultaneous Authentication of Equals)
feature from wpa_supplicant and makes it work with mbedtls
crypto APIs. Currently only group 19 is supported. A sample
SAE handshake is included in the testcase. Other minor
changes for DH groups are also included.
1. Add CCMP, AES crypto modules for unicast protected Mgmt frames
2. Add support for computing SHA256 MIC on Bcast Mgmt frames
3. Add support for storing iGTK during 4-way handshake.
4. Provide APIs to MLME for utilizing the SW crypto modules
1. Add APIs for configuring PMF through set config.
2. Map Supplicant and Wifi Cipher types.
3. Add support for PMF negotiation while generating RSN IE.
1) Added PMK caching module from wpa_supplicant.
2) Modified wpa_sm to
a) Add entry to PMK cache when first time associated to an AP.
b) Maintain entry across the associations.
c) Clear current PMKSA when deauth happens.
d) Search for an entry when re-associating to the same AP and
set it as current PMKSA
e) Wait for msg 1/4 from AP instead of starting EAP authentication.
f) Check PMKID in msg 1 with current PMKSA/cache.
g) Use the cached PMK to complete 4-way handshake.
3) Remove config_bss callback as it was redundant and used to cause
problems for PMK caching flow.
Closes IDF-969
During BLE Mesh Provisioner initialization, the stack will restore
the nodes information if settings storage is enabled.
Previously when a failure happens (e.g. found the same uuid) during
the restore procedure, the information of the following nodes will
not be restored and error will be directly returned.
But this will introduce some problem with user experience, because
some newly provisioned nodes information will not be restored and
Provisioner will not be able to control those nodes.
So we change the operation here, when a failure happens during the
restore procedure, Provisioner will only ignore the information of
the current node and continue restoring other nodes information.
With this change, if a Provisioner has provisioned the maximum
number of nodes, it can still report the unprovisioned device
beacon from other nodes to the application layer. And this will
be more reasonable compared with the previous implementation.
Previously when the node array of Provisioner is full, no beacon
from unprovisioned devices will be reported, only some warning
logs will be given.
Previously only check the node address when it is assigned by the
application layer. Here we also check the address when the address
is allocated internally. And this will be useful when some mesh
internal tests are performed.
Previously the BLE_MESH_MAX_STORED_NODES option is added for
internal mesh test, which will be a little confusing for the
users to understand.
Here we remove this option, instead the BLE_MESH_MAX_PROV_NODES
will be used for all the cases. For mesh internal test, when
the test function is called to add some nodes info, the info
will be stored in the array of provisioned nodes directly.
The replay protection list of Provisioner should be at least equal
to the number of nodes with the precondition that each node contains
only one element.
The help information of replay protection list is updated, and the
maximum number of nodes for Provisioner is adjusted based on the
replay protection list size.
On Xtensa, backtrace can not recover the two most significant bits of
the address, as the window call size is encoded in these bits.
Because of this, __builtin_return_address modifies these MSBs to
match those of the callee, "fixing" the address. An unfortunate side
effect is that the zero return address, which usually terminates the
backtrace, gets converted to 0x40000000. While there is a valid
instruction at this address, its occurrence in the backtrace is
highly unlikely: this is the first instruction of WindowOverflow4
vector, and IDF apps switch VECBASE to an IRAM location very early at
startup.
Commit fc03161f updated esp_ota_get_app_elf_sha256 to store and return
X number of bytes of the hash, but the test case still expected 64 bytes.
Updated test case to use CONFIG value for expected length.
This is a regression from earlier commit related to TLSV12 which used
sha functions that are currently declared static.
Solution: Follow upstream code structure and resolve the errors.
1. add enable PSRAM 2T mode function
2. abort when himem and 2T mode are enabled meanwhile
3. set SPIRAM_2T_MODE as "n" by default, enable it when needed
1. Fix WiFi log print level
2. Optimize WiFi/BT coexist sleep
- If any data tx/rx in WiFi slice, fore wakeup next TBTT
- Increase active timeout time to coex TBTT interval to avoid sleep in WiFi slice
3. Fix esp_wifi_stop() crash
esp_wifi_stop() crashes if it's called when WiFi is not initialized.
4. Fix esp_wifi_stop() leads to memory leak
5. esp_wifi_sta_get_ap_info() returns actual values of group ciphers
It is not possible to write to Flash when the stack is located in
PSRAM, and it is not possible to write PSRAM buffers into Flash when
malloc can not be used.
and all ECDSA to be disabled if secure boot is not enabled
Previously if ECDSA disabled in config then secure_boot_signatures.c would
fail to build (whether or not secure boot was enabled).
To avoid breaking apps that might be using the signature scheme with custom OTA
without enabling secure boot signatures in config, this change just disables
this functionality if unavailable in mbedTLS config.
Possible fix for root cause of https://github.com/espressif/esp-idf/pull/3703
Closes https://github.com/espressif/esp-idf/issues/4758
Wifi_provisioning: Change service_uuid to non standard 128 bit UUID and add retries in test script (backport v4.0)
See merge request espressif/esp-idf!8163
Using the ble mesh white list test functions, a node can choose to
only receive mesh messages from a specific node and relay the
messages for it. Messages from other nodes will be ignored.
1,Reduce WiFi bin size
2,Add TX packets size check
3,Fix scan get rssi error
4,Add wifi stop check at WiFi deinit entry
5,coex adjust scheme when bt is connected status
6,Return fail when setting AP's channel is out of range
7,Fix the bug for setting channel when WiFi in NULL mode
When a handler instance is the last one in the list und unregisters
itself, the handler iterator will be invalidated by entering free'd
memory. Same applies for event base and id, if they become empty.
Merges https://github.com/espressif/esp-idf/pull/4139
Problem:
mbedtls_ctr_drbg_context was initialized in crypto_ec_point_mul. This
was okay in releases before 2.16.4 as entropy_len used to get set to
MBEDTLS_CTR_DRBG_ENTROPY_LEN in function mbedtls_ctr_drbg_seed. The
function is now changed to set the length to
MBEDTLS_CTR_DRBG_ENTROPY_LEN if previous length is 0 and hence the bug.
Solution:
Initialize mbedtls_ctr_drbg_context in crypto_ec_point_mul.
components/esp_https_ota: Changes in public requirement list for CMake and esp_ota_ops.h is included in esp_https_ota.c
bootloader_support is added in public requirements of CMake
bootloader_common.h is included in esp_https_ota.h
Closes https://github.com/espressif/esp-idf/issues/4198
Closes IDFGH-2066
Current implimentation of esp_http_ota does not perform any error-checking in the data writing phase calls `esp_ota_get_next_update_partition()` irrespetive of the received state of the image. A few additional error checking mechanism have now been added inside the esp_https_ota which returns the control in case an invalid header is received and a wrapper to the function `esp_http_client_is_complete_data_received()` of `esp_http_client` has been added.
While downloading OTA firmware, if their is a Origin Respnse Timeout or the binary is only partially downloaded, OTA failure is observed. Checking binary size can also be helpful for simple http client applications.
Closes https://github.com/espressif/esp-idf/issues/3004
The NimBLE host buffers that consume a significant amount of memory are
now allocated dynamically.
The advantage is that, the memory can be reclaimed in cases where BLE
is turned off and not required for the current boot cycle
- Add menuconfig option for NimBLE host flow control
- Include changes in `blecent` example from upstream PR!702
- add ble_hs_lock in ble_gap_unpair Upstream PR!584
- ble_hs_hci_rx_evt, upstream PR!738
Closes https://github.com/espressif/esp-idf/issues/4243
1. Bugfix for AG audio crash (change the return position)
2. Fix the error macro name and error return in hfp_hf demo
3. Fix the annotation error using UTF-8 ' (from Github)
4. Change or remove the log in SCO related code region.
5. Correct error of introduction of a function.
ledc.h includes two similar enums, ledc_clk_src_t & ledc_clk_cfg_t.
The two enums do different things but there are two similar names: LEDC_REF_TICK / LEDC_USE_REF_TICK
and LEDC_APB_CLK / LEDC_USE_APB_CLK.
Because C will accept any enum or integer value for an enum argument, there's no easy way to check
the correct enum is passed without using static analysis.
To avoid accidental errors, make the numeric values for the two similarly named enums the same.,
Noticed when looking into https://github.com/espressif/esp-idf/issues/4476
Allocation of the temporary internal buffer will now repeat until a small enough buffer can be
allocated, and only fail if less than a 256 byte block of internal RAM is free.
Adds unit test for the same, and generic test utility for creating memory pressure.
Previously would try allocate buffer of minimum size 16KB not maximum size 16KB, causing
out of memory errors for any large reads, or if less than 16KB contiguous free heap.
Also, if using legacy API and internal allocation failed then implementation would abort()
instead of returning the error to the caller.
Added test for using large buffers in PSRAM.
Closes https://github.com/espressif/esp-idf/issues/4769
Also reported on forum: https://esp32.com/viewtopic.php?f=13&t=14304&p=55972
enabled. The global interrupts will be disabled before taking current
time for scheduling further bluetooth events. Earlier disabling
interrupts was taking some time because of which WiFi events were
interrupting in current bluetooth scheduling.
Also fixes ble crash issue triggered by ble event irq miss(0x20000) (819979b9)
transport_ws can now be read multiple times in a row to read frames larger than the buffer.
Added reporting of total payload length and offset to the user in websocket_client.
Added local example test for long messages.
Closes IDF-1083
1. Add an API to set Provisioner static oob value
2. Add an API to deinit BLE Mesh stack
3. Add an API to set Provisioner unicast address
4. Add an API to provision devices with fixed address
5. Add an API to store node composition data
6. Add an API to get node with device uuid
7. Add an API to get node with unicast address
8. Add an API to delete node with device uuid
9. Add an API to delete node with unicast address
10. Add an API for Provisioner to update local AppKey
11. Add an API for Provisioner to update local NetKey
12. Support Provisioner persistent functionality
13. Fix Provisioner entering IV Update procedure
14. Fix an issue which may cause client failing to send msg
15. Use bt_mesh.flags to indicate device role
16. Remove several useless macros
17. Callback RSSI of received mesh provisioning packets
18. Modify the Provisioner disable function
19. Change some log level from debug to info
20. Add parameters to Provisioner bind AppKey completion event
21. Fix node ignoring relay messages issue
22. Support using a specific partition for BLE Mesh
23. Fix compile warning when proxy related macros are disabled
24. Clean up BLE Mesh stack included header files
25. NULL can be input if client message needs no parameters
26. Fix compile warning when BT log is disabled
27. Initilize BLE Mesh stack local variables
28. Support using PSRAM for BLE Mesh mutex, queue and task
29. Add a menuconfig option to enable using memory from PSRAM
30. Clean up sdkconfig.defaults of BLE Mesh examples
Fix an error of mesh deinit lock.
Fix mesh send returns TIMEOUT.
Increase root window close time to 120s.
Clear ie.child when the child connected or fully associated.
Update rc_rssi when roots direct/forward.
ix assoc in mesh ie wrong.
Mesh packets use AC_VI instead of AC_BE.
For SoftAP provisioning, if we start Wi-Fi connection as soon as the
apply config request is received, it has been observed that the clients
(Phone apps mostly) do not get the response for this as we switch the
Wi-Fi channel, causing a momentary disconnection.
Delaying the connection calls helps mitigate the issue.
Signed-off-by: Piyush Shah <piyush@espressif.com>
this issue is mainly exposed when using larger (4096) client key in TLS mutual auth,
since it uses multiplications > 2048 when mbedtls_mpi_mul_mpi is used in recursion,
which works only if both operands point to different location than result since
mpi_mult_mpi_overlong() called mbedtls_mpi_grow() to reallocate buffers used in previous
pointer arithmetics and thus corrupting it. Fixed by growing the mpi buffer before
calling mpi_mult_mpi_overlong()
1. The fault signal 3 for unit 2 is corrected to the right value
2. Now `mcpwm_fault_set_oneshot_mode` will clear the status before, no
need to reset the peripheral.
3. The capture feature relies on the prescaler, but it's only
initialized in the mcpwm_init funciton, which is used to initialize a
PWM channel. This means, the capture may not work correctly if no PWM
channel is enabled.
Now the prescaler is also updated when `mcpwm_capture_enable` is
called.
Fix following WiFi issues:
1. Fix WiFi buffer reload issue
2. Fix AMSDU decrypt issue
3. Fix some WiFi timer issues
4. Fix the crash caused by too big of association request RSN information
5. Fix the crash caused by block scan
6. Fix the bug for getting channel and bandwidth
7. Fix some Sniffer bugs
8. Fix some ESP-NOW issues
1> fix the bug when modifying the channel info of peer node
2> fix the crash when modifying peer node between unencrypted and encrypted
3> fix the bug for fetch peer
4> modify the esp_wifi_set_channel() function
5> fix the bug that the channel parameter doesn't work when adding peer node
Closes https://github.com/espressif/esp-idf/issues/2833
Closes https://github.com/espressif/esp-idf/issues/4311
When node is being reset, the init functions of each sig-defined
models will be invoked again, this will cause memory leak because
some model internal data will be allocated again.
Hence before trying to allocate memory for them, we add some check
to make sure no memory has been allocated previously.
And for client model, when the init functions are invoked again,
we will clear the list items.
Even if frame->exccause != PANIC_RSN_CACHEERR, it is possible that
the cache error interrupt status is set. For example, this may happen
due to an invalid cache access in the panic handler itself.
Check cache error interrupt status instead of frame->exccause to
decide whether to do CPU reset or digital reset.
Also remove unnecessary esp_dport_access_int_pause from
esp_cache_err_get_cpuid, since the panic handler already calls
esp_dport_access_int_abort on entry.
