Commit Graph

12588 Commits

Author SHA1 Message Date
liminyang
be80b6f270 docs:perfect 128-bit UUID description
The previous description is not easy for the reader to understand, some changees have been made to perfect it.
Closes https://github.com/espressif/esp-idf/issues/5057
2020-06-18 14:44:28 +08:00
Nachiket Kukade
ee70736a03 esp_wifi: Update wifi lib
1. Fix WPA3 Auth timeout during BLE coex by increasing CNX timeout
2. Extend Coex phase for SAE Handshake
3. Define and utilize WPA3 Mixed Authmode, update esp_wifi_types.h
2020-06-18 05:19:26 +00:00
Angus Gratton
9e55345c43 Merge branch 'bugfix/efuse_logs_v4.1' into 'release/v4.1'
esp32: Reduce using ESP_EARLY_LOGx and move some code after the stdout initialization in startup code (v4.1)

See merge request espressif/esp-idf!9206
2020-06-18 07:11:42 +08:00
Jiang Jiang Jian
6b66850dc1 Merge branch 'bugfix/ble_conn_fail_v4.1' into 'release/v4.1'
Fix bugs from coex test v4.1

See merge request espressif/esp-idf!9136
2020-06-17 18:17:15 +08:00
KonstantinKondrashov
e0e391dce3 esp32s2: Move some code after the stdout initialization 2020-06-17 15:53:45 +08:00
KonstantinKondrashov
0a1919abff esp32: Reduce using ESP_EARLY_LOGx and move some code after the stdout initialization
After setting _GLOBAL_REENT, ESP_LOGIx can be used instead of ESP_EARLY_LOGx.

Closes: https://github.com/espressif/esp-idf/issues/5343
2020-06-17 15:53:45 +08:00
yuanjm
af82516cfb feat(lwip): make LWIP TCP rto time configurable 2020-06-17 15:29:25 +08:00
Piyush Shah
6baa5f0d7b protocomm_httpd: Restart security session if request is received on a new session
This commit fixes a bug as well as changes a behaviour.

Bugfix: During softap/httpd based provisioning, if a session was closed
midway and a new one started, it would never proceed if the http server
assigns same socket number to the new session (which happens almost always).
Now, if a session is closed, using the http callbacks, the older session
data is cleared so that a new one can be created.

Behavioural change: If a client (mobile app particularly) does not use
persistent http session i.e. all provisioning communication on the same
socket, the provisioning may fail. Earlier, since the session context was
not getting cleared, even if the client closed a session and continued
on a new one, it would go through if the socket number assigned was same
(which happens almost always).

Ideally, from a security perspective, all communication related
to secure provisioning must happen on the same socket, and so, this
change is required.
2020-06-17 05:34:11 +00:00
Jiang Jiang Jian
c5bbe87425 Merge branch 'bugfix/airkiss_find_hidden_ap_fail_v4.1' into 'release/v4.1'
fix the bug that airkiss find hidden AP fail (backport v4.1)

See merge request espressif/esp-idf!9176
2020-06-17 12:26:56 +08:00
Jiang Jiang Jian
5e52e5f6f5 Merge branch 'bugfix/blufi_buffer_overflow_report_v4.1' into 'release/v4.1'
Bugfix/blufi buffer overflow report v4.1

See merge request espressif/esp-idf!9178
2020-06-16 15:59:49 +08:00
Jiang Jiang Jian
f82010fa6a Merge branch 'bugfix/att_buffer_count_fixed_v4.1' into 'release/v4.1'
Bugfix/att buffer count fixed v4.1

See merge request espressif/esp-idf!9160
2020-06-16 15:59:34 +08:00
xiongyu
cf3b2df4a4 bugfix(uart): fix esp32 fifo_cnt bug
When using DPort to read fifo, fifo_cnt is not credible, we need to calculate the real cnt based on the fifo read and write pointer. When using AHB to read FIFO, we can use fifo_cnt to indicate the data length in fifo.
2020-06-16 14:48:02 +08:00
zhangyanjiao
08e7bce4ea esp_wifi:
1. fix the bug that airkiss find hidden AP fail
2. addba fix and send bar in first ampdu
2020-06-16 10:48:48 +08:00
Michael (XIAO Xufeng)
a945ae8a9d Merge branch 'bugfix/add_return_task_awoken_for_timer_callback_v4.1' into 'release/v4.1'
timer: add return task awoken for timer callback (backport v4.1)

See merge request espressif/esp-idf!9148
2020-06-16 10:38:43 +08:00
Jiang Jiang Jian
b33828e254 Merge branch 'bugfix/authmode_degrade_fix_v4.1' into 'release/v4.1'
esp_wifi: Remove auth_changed event and fix wpa2_ent authmode threshold (backport v4.1)

See merge request espressif/esp-idf!8804
2020-06-15 23:18:55 +08:00
wangcheng
ae64d9e738 blufi: When the format of the received data packet is wrong, reply with an error response 2020-06-15 19:37:38 +08:00
Wang Cheng
6861f7d5d8 components/bt: Add a detailed description for the user to distinguish the function of the query buffer api. 2020-06-15 19:19:14 +08:00
Geng Yu Chao
c5a6fc9c7e Update the limited of function esp_ble_get_sendable_packets_num. 2020-06-15 19:17:27 +08:00
wangcheng
3b8b057894 components/bt: Consider link_xmit_data_q in congest events to prevent excessive memory usage 2020-06-13 15:07:28 +08:00
wangcheng
e69ee31e1a components/bt: Fixed the problem of incomplete data packets caused by disordered acl data packets 2020-06-13 15:06:28 +08:00
wangcheng
846a2f95f4 components/bt: Add api to query the number of available buffers for the current connection 2020-06-13 15:06:14 +08:00
chenjianqiang
74623ac9d5 timer: add return task awoken for timer callback 2020-06-12 14:33:21 +08:00
Angus Gratton
37c155c359 msys2: Remove python version consideration in pacman commands 2020-06-12 10:22:10 +10:00
Angus Gratton
194a3b09dc windows: Update MSYS2 pre-compiled legacy build environment
* Bundled with esp2020r2 toolchain
* Built with Python 3 & latest MSYS2 packages
2020-06-12 10:22:10 +10:00
baohongde
1b8905e544 components/bt: Fix ble disconnect due to instant of channel map update is too small 2020-06-11 16:55:36 +08:00
baohongde
d0093c7675 Set AFH in time(075dd35)
Send LMP_channelClassification immediately
Fix first LMP_set_AFH not using host value
2020-06-11 16:55:26 +08:00
baohongde
de6f21936f components/coex: Fix ble connecting fail in coexist 2020-06-11 16:55:12 +08:00
kapil.gupta
46486a863c esp_wifi: Remove auth_changed event and fix wpa2-ent authmode threshold
Added following as part of this change
1. Removed auth_changed event
2. Updated doc regarding usage of  WPA/WEP secuiry modes
3. Fixed WPA2 enterprise authmode threshold.
2020-06-11 13:40:58 +05:30
Jiang Jiang Jian
29bc65e719 Merge branch 'feature/toolchain_2020r2_v4.1' into 'release/v4.1'
Update toolchain to esp-2020r2  (backport v4.1)

See merge request espressif/esp-idf!9000
2020-06-11 15:44:47 +08:00
Jiang Jiang Jian
7de0c391ac Merge branch 'bugfix/fix_some_wifi_bugs_0522_v4.1' into 'release/v4.1'
bugfix: fix some wifi bugs (backport v4.1)

See merge request espressif/esp-idf!8880
2020-06-11 15:41:06 +08:00
Jiang Jiang Jian
326b42560d Merge branch 'bugfix/collection_of_confirmed_fixes_from_baidu_project_4.1' into 'release/v4.1'
Backport MR for baidu project for release/v4.1

See merge request espressif/esp-idf!8987
2020-06-11 14:46:22 +08:00
Angus Gratton
6f768b74e4 Merge branch 'bugfix/parttool_get_info_wo_info_option_v4.1' into 'release/v4.1'
partition_table: Fix parttool.py crashes when retrieving partition info (v4.1)

See merge request espressif/esp-idf!9063
2020-06-10 16:02:03 +08:00
Angus Gratton
aa6481837f Merge branch 'bugfix/bootloader_header_align_v4.1' into 'release/v4.1'
bootloader: Force bootloader_image_hdr to be word aligned (v4.1)

See merge request espressif/esp-idf!9068
2020-06-05 07:28:35 +08:00
Ivan Grokhotkov
bd3caffdcd bootloader_support: force alignment of flash_read argument 2020-06-04 21:01:56 +10:00
Angus Gratton
a5b4fda207 bootloader: Force bootloader_image_hdr to be word aligned
Possible due to linker order for this file to be placed unaligned,
causing failure from bootloader_flash_read() function.
2020-06-04 20:53:46 +10:00
KonstantinKondrashov
7070fea6e4 partition_table: Fix parttool.py crashes when retrieving partition info
Closes: IDFGH-3279
Closes: https://github.com/espressif/esp-idf/issues/5271
2020-06-04 16:31:05 +08:00
Angus Gratton
d35226c656 Merge branch 'bugfix/wpa_supplicant_no_hw_mpi_v4.1' into 'release/v4.1'
wpa_supplicant: Allow building with mbedTLS integration but no hardware MPI (v4.1)

See merge request espressif/esp-idf!9013
2020-06-04 16:22:14 +08:00
aleks
4be2170250 freemodbus: add comments for limitation in current version of freemodbus 2020-06-04 09:46:04 +02:00
Alexander Borsuk
29d3c40a25 sdmmc: Correctly check function parameters before using them
Closes https://github.com/espressif/esp-idf/pull/5000
2020-06-04 11:49:40 +08:00
baohongde
7cdc2f2b81 Backport MR for baidu proj 2020-06-04 11:07:30 +08:00
Angus Gratton
50319bb03c wpa_supplicant: Allow building with mbedTLS integration but no hardware MPI
Also disable the relevant function in bignum.h based on config, so fails at
compile not link time.

Closes https://github.com/espressif/esp-idf/issues/5321
2020-06-03 15:11:21 +10:00
Angus Gratton
0ea6d39686 Merge branch 'bugfix/secure_boot_v2_fixes_v4.1' into 'release/v4.1'
Small secure boot v2 fixes (v4.1)

See merge request espressif/esp-idf!9019
2020-06-03 12:34:55 +08:00
Anton Maklakov
298f23c958 Update toolchain to esp-2020r2
Fixes PSRAM issues, volatile loads with -O2, std::locale, C++ exception crashs, FDE sorting switchable;

Includes a 64-bit toolchain for Windows (win64);

Closes https://github.com/espressif/esp-idf/issues/5090
Closes https://github.com/espressif/esp-idf/issues/5112
Closes https://github.com/espressif/esp-idf/issues/5150
2020-06-02 13:19:00 +07:00
Ivan Grokhotkov
1e12199832 esp32: use ccomp_timer in SHA test 2020-06-02 13:15:14 +07:00
Ivan Grokhotkov
36a66a7243 newlib: use --specs=nano.specs to enable "nano" version of C library
esp2020r1 toolchain includes a nano.specs file, which instructs GCC to
substitute libc.a with libc_nano.a.

In the build system, this simplifies handling of the nano formatting
option, eliminating LIBC and LIBM global variables.
2020-06-02 13:15:08 +07:00
Jeroen Domburg
4eec4fbc99 Add fixes for gcc8 psram fix improvement 2020-06-02 13:14:29 +07:00
Angus Gratton
b00f38f91c secure boot v2: Add anti-FI check that secure boot not enabled yet
Prevent a fault from causing bootloader to trust the provided signature incorrectly.
2020-06-02 16:14:01 +10:00
Angus Gratton
3c6b1b4c0a secure boot v2: Don't check efuse BLK2 if only boot-time signature verification is enabled 2020-06-02 16:14:01 +10:00
Angus Gratton
2c531d5bb3 secure boot v2: Don't log warnings when BLK2 is empty as expected
If BLK2 is empty then it's OK to continue with a warning (otherwise it may spook users into thinking
something this is wrong, but this is the expected workflow.)

If BLK2 is not empty and doesn't match then we need to fail because it won't be possible to
trust the signature.
2020-06-02 16:14:01 +10:00
lly
1bd7944316 ble_mesh: Rework prov pdu length check 2020-06-02 10:13:51 +08:00