Running Espruino on Win10 WSL?

Hi,

I was wondering if anyone has tried to compile Espruino on Windows Subsystem for Linux?

I have a few long flights from Europe back to Australia and I thought it would be great to be able to continue testing some things I'm working on for my Picos but on Windows 10? I should have bought a Pico with me and then I could just use that but only thought about this project while I have been away.

So far I have got Debian installed with WSL. I have installed all the required build tools, cloned from git and then make. Everything looks ok. Then I try run Espruino and I get the following...

Interactive mode.
Size of JsVar is now 32 bytes
Size of JsVarRef is now 4 bytes
Added SIGINT hook
Added SIGHUP hook
Added SIGTERM hook
ASSERT(strlen(name)<=8) FAILED AT src/jsflash.c:56
  #1[r1,l2] Object {
    #2[r1,l2] Name String [1 blocks] "\xFF"      #3[r1,l1] Object {
        #6[r1,l2] Name String [1 blocks] "timers"          #7[r2,l0] Array(0) [ ]
        #8[r1,l2] Name String [1 blocks] "watches"          #9[r2,l0] Array(0) [ ]
      }
  }
EXITING.

I had a look at jsflash.c but I'll admit its a bit over my head.

Anybody else tried this? Is it even possible?

Cheers
Ryan

So you compiled this to run natively on WSL? Basically just running make?

It definitely should work - pretty sure I've done it before. Any chance you could run:

gdb espruino
s (for start)
c (for continue)
(it should now break at the assertion)
bt (backtrace - and paste up what it says)
q (quit)

That should tell us exactly which bit of code tried to access a file with a name that was too long.

Hi Gordon,

Thanks for that. See below. I had to run bt before continue as seemed to exit and provide no info, only No stack. Hopefully this is what you're after.

(gdb) start
Temporary breakpoint 2 at 0x806abc7: file targets/linux/main.c, line 252.
Starting program: /usr/local/bin/espruino
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".

Temporary breakpoint 2, main (argc=1, argv=0x7ffffffde3a8) at targets/linux/main.c:252
252       int i, args = 0;
(gdb) bt
#0  main (argc=1, argv=0x7ffffffde3a8) at targets/linux/main.c:252
(gdb) continue
Continuing.
Interactive mode.
Size of JsVar is now 32 bytes
Size of JsVarRef is now 4 bytes
Added SIGINT hook
Added SIGHUP hook
Added SIGTERM hook
[New Thread 0x7ffffe570700 (LWP 12967)]
ASSERT(strlen(name)<=8) FAILED AT src/jsflash.c:56
  #1[r1,l2] Object {
    #2[r1,l2] Name String [1 blocks] "\xFF"      #3[r1,l1] Object {
        #6[r1,l2] Name String [1 blocks] "timers"          #7[r2,l0] Array(0) [ ]
        #8[r1,l2] Name String [1 blocks] "watches"          #9[r2,l0] Array(0) [ ]
      }
  }
EXITING.
[Thread 0x7fffff7b0740 (LWP 12966) exited]
[Inferior 1 (process 12966) exited with code 01]
(gdb) q

Cheers

Ahh, ok - can you try it again, but with break jsAssertFail after s but before c - it should then stop on the assertion itself and you can backtrace it.

I don't know if it is the same problem, or this could help
but I compile on windows 7 (not 10 WSL) the ESP32 version
I flash it on an esp32 board.
I can make some test
but I can't use the reset() function because I have :

>reset()
=undefined
WARNING: Scan stop failed
WARNING: set rssi scan not implemeted yet
ASSERT(strlen(name)<=8) FAILED AT src/jsflash.c:56
  #1[r1,l2] Object {
    #2[r1,l2] Name String [1 blocks] "\xFF"      #3[r1,l1] Object {
        #6[r1,l2] Name String [2 blocks] "timers"          #8[r2,l0] Array(0) [ ]
        #9[r1,l2] Name String [2 blocks] "watches"          #11[r2,l0] Array(0) [ ]
        #14[r1,l2] Name String [1 blocks] "net"          #12[r1,l1] String [2 blocks] "\x05\x00\x00\x00\x00\xFF\xFF\xFF\x00\x00\x00\x00"
      }
  }
HALTING.

it look's similar no ?

well after modifying

JsfFileName jsfNameFromString(const char *name) {
  jsiConsolePrintf("jsfNameFromString [%s]%d\n",name, strlen(name));
  assert(strlen(name)<=8);

at boot I have :

ets Jun  8 2016 00:22:57
rst:0x1 (POWERON_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT)
configsip: 0, SPIWP:0xee
clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00
mode:DIO, clock div:2
load:0x3fff0018,len:4
load:0x3fff001c,len:2364
load:0x40078000,len:0
load:0x40078000,len:10880
entry 0x40078c4c
jsfNameFromString [.varimg]7
jsfNameFromString [.bootcde]8
jsfNameFromString [.bootrst]8
jsfNameFromString [.boot0ý?]8
jsfNameFromString [.boot1ý?]8
jsfNameFromString [.boot2ý?]8
jsfNameFromString [.boot3ý?]8
jsfNameFromString [.bootrst]8
jsfNameFromString [.bootcde]8

and after reset() I have :

>reset()
=undefined
WARNING: Scan stop failed
WARNING: set rssi scan not implemeted yet
jsfNameFromString [.boot0ý?
]9

so the problem is of '\0' not well placed

bool jsfLoadBootCodeFromFlash(bool isReset) {
  // Load code in .boot0/1/2/3
  char filename[6] = ".bootX";
  for (int i=0;i<4;i++) {

filename 6 no place for '\0' it's a bug ?

changing 6 by 7

bool jsfLoadBootCodeFromFlash(bool isReset) {
  // Load code in .boot0/1/2/3
  char filename[7] = ".bootX";

no more problem everything is ok know

@ryan42 can you do the same change ?

(One good point for this stupid windows : it detects some bugs ...)

Thanks! I just added that change to GitHub so if you pull now it should work.

That's really handy - I guess on most platforms it aligns the stack to the nearest 4 bytes, so it never got noticed.

Hi jbgagnon,

Yes that did it! Thanks for that.

@Gordon I do have that backtrace if you’d like it but I guess it’s all good?

Now for a 14 hour flight back to Australia, at least I can keep myself entertained.

Cheers!

I do have that backtrace if you’d like it

It would have been helpful, but @jbgagnon beat us to it and fixed the problem - so we're fine now :)