search

stanleyhuangyc / Freematics

Official source code repository for Freematics
https://freematics.com
425 stars 346 forks source link

webserver hangs in v5 datalogger #52

Closed AndreaBorgia-Abo closed 5 years ago

AndreaBorgia-Abo commented 6 years ago

Up to, say, a couple of months ago the internal webserver in datalogger worked ok and would be available both in AP and in station mode.

With today's build, the connection hangs and no output is returned: Motion:0.58 Wakeup File:/DATA/71.CSV OBD...NO Standby ACC Bias:0.00/-0.00/-1.27 [56] Client IP: 192.168.4.2 [56] Socket closed, 0 connections

This happens also when only AP mode is enabled; tested with Firefox, wget and also raw telnet. Can anyone else verify?

AndreaBorgia-Abo commented 6 years ago

This is still happening with today's commit. Wanted to "git bisect" it but versions prior to commit 4a4a2d30171796498f7848655e8a10e39d9faee6 fail to build for a deprecated bta_api.h reference. Need to edit platformio.ini and force the use of an older version: [env:esp32git102] platform=espressif32@1.0.2 board=esp32dev framework=arduino monitor_baud=115200

AndreaBorgia-Abo commented 6 years ago

Found it: first bad commit is f8e709ac003a26dc9e1945af52838ee8c46bdcdb I can provide log file and also config.h / platformio.ini for each step, since each step used the most up-to-date platform available when the corresponding commit was logged.

Given how inconsequential this specific commit is, I'm inclined to think the issue lies elsewhere: (bad) f8e709ac003a26dc9e1945af52838ee8c46bdcdb used v.1.1.0 (good) 32516b9c3c2c457f6bf5157f442187b737ed1c7b used v.1.0.2

Sure enough, building f8e709ac003a26dc9e1945af52838ee8c46bdcdb with 1.0.2 gives me a webserver which does not hang.

AndreaBorgia-Abo commented 6 years ago

Escalating the issue to Platform first and, perhaps, ESP32 as well.

AndreaBorgia-Abo commented 5 years ago

After the first and so far only merge of a PR (#76) on this project, I wondered if it applied to my bug as well.

To test, I rebuilt f8e709ac003a26dc9e1945af52838ee8c46bdcdb first as is and verified the bug was still reproducible: it was. Then I applied the patch, rebuilt and tried again: bingo, now the webserver actually replies.

Closing this and related tickets / posts.