[ESP8266] new Uint8Array -> reset

[mikabytes]

Previously I was on version 2.14 and all was fine. But with 2.18 and 2.19 I have this error:

>new Uint8Array(3)

=F 0

 ets Jan  8 2013,rst cause:2, boot mode:(3,7)

load 0x40100000, len 2408, room 16 
tail 8
chksum 0xe5
load 0x3ffe8000, len 776, room 0 
tail 8
chksum 0x84
load 0x3ffe8310, len 632, room 0 
tail 8
chksum 0xd8
csum 0xd8

2nd boot version : 1.6
  SPI Speed      : 80MHz
  SPI Mode       : DIO
  SPI Flash Size & Map: 32Mbit(512KB+512KB)
jump to run user1 @ 1000

nÇrn
    l`rlÄlþLoading 5717 bytes from flash...

[gfwilliams]

This works in the normal build, so what hardware are you running this on? And is it a build we supply or one you did yourself?

[mikabytes]

This is running on an ESP8266 with firmware downloaded from the releases page (here to be specific)

I tried it on a second chip just to rule out bad hardware. But I get the same issue there.

[mikabytes]

Also tried on a completely fresh install, no JS loaded whatsoever. Just cli using Espruino binary from NPM

[gfwilliams]

Ok, thanks!

Also tried on a completely fresh install,

Are you sure? Because it says Loading 5717 bytes from flash... like you might have some code already in there which might be causing some kind of issue? You could try running reset(1) which should hopefully get rid of that

[mikabytes]

Yes, initial bug report was indeed on a chip that had other JS loaded on it. Here's the output after running reset(1)

>reset(1)
Erasing saved code.
 ____                 _ 
|  __|___ ___ ___ _ _|_|___ ___ 
|  __|_ -| . |  _| | | |   | . |
|____|___|  _|_| |___|_|_|_|___|
         |_| espruino.com
 2v19 (c) 2021 G.Williams

Espruino is Open Source. Our work is supported
only by sales of official boards and donations:
http://espruino.com/Donate
Flash map 4MB:512/512, manuf 0xd8 chip 0x4016

>new Uint8Array(3)
=new 
 ets Jan  8 2013,rst cause:2, boot mode:(3,7)

load 0x40100000, len 2408, room 16 
tail 8
chksum 0xe5
load 0x3ffe8000, len 776, room 0 
tail 8
chksum 0x84
load 0x3ffe8310, len 632, room 0 
tail 8
chksum 0xd8
csum 0xd8

2nd boot version : 1.6
  SPI Speed      : 80MHz
  SPI Mode       : DIO
  SPI Flash Size & Map: 32Mbit(512KB+512KB)
jump to run user1 @ 1000

nÇrn
    l`rlÄlþ
 ____                 _ 
|  __|___ ___ ___ _ _|_|___ ___ 
|  __|_ -| . |  _| | | |   | . |
|____|___|  _|_| |___|_|_|_|___|
         |_| espruino.com
 2v19 (c) 2021 G.Williams

Espruino is Open Source. Our work is supported
only by sales of official boards and donations:
http://espruino.com/Donate
Flash map 4MB:512/512, manuf 0xd8 chip 0x4016

[gfwilliams]

Ok, thanks - it's interesting this one lacks the F 0 - I'll try and take a look when I get a second. I know ESP8266 is very picky about accessing non-word aligned RAM areas and it's possible something's crept in which causes a problem

[mikabytes]

Another interesting observation; it will allows me to create signed integers once, but if I do it twice then it will reset. Reproducible every time.

>reset(1)
Erasing saved code.
 ____                 _ 
|  __|___ ___ ___ _ _|_|___ ___ 
|  __|_ -| . |  _| | | |   | . |
|____|___|  _|_| |___|_|_|_|___|
         |_| espruino.com
 2v19 (c) 2021 G.Williams

Espruino is Open Source. Our work is supported
only by sales of official boards and donations:
http://espruino.com/Donate
Flash map 4MB:512/512, manuf 0xd8 chip 0x4016

>new Int8Array(3)
=new Int8Array([111, 114, 121])
>new Int8Array(3)
=F 0

 ets Jan  8 2013,rst cause:2, boot mode:(3,7)

load 0x40100000, len 2408, room 16 
tail 8
chksum 0xe5
load 0x3ffe8000, len 776, room 0 
tail 8
chksum 0x84
load 0x3ffe8310, len 632, room 0 
tail 8
chksum 0xd8
csum 0xd8

2nd boot version : 1.6
  SPI Speed      : 80MHz
  SPI Mode       : DIO
  SPI Flash Size & Map: 32Mbit(512KB+512KB)
jump to run user1 @ 1000

[gfwilliams]

This appears to be fixed now - I'm pretty sure it was due to an overflow in the JsVar union for arraybuffers, which only showed up on devices with little RAM

[mikabytes]

Great, thanks!

I can confirm. Tested my old code on v2.20 and it's running fine now.