[9front] 9front on spim

[9front] 9front on spim

[adventures in9]

I'll make this a separate thread for anyone who wants to talk about spim
specifically.

I saw spim mentioned in the Port Status thread.  I haven't had any trouble
building for spim, and some of the missing parts might just be because it
is building off code stored in "mips" but does it with with the "spim"
compiler, which is really just a script with the mips compiler set with a
little-endian flag.

As for hardware, I've been using the Onion Omega 2.  I've also found
several Chinese router boards that also use the MT7688.  I have got it to
start booting on the Onion, but haven't finished it up yet.  I haven't
touched it in a couple months, but what I have done so far is here:

https://github.com/adventuresin9/9front-mt7688

I'm pretty sure the Loongson and Ingenic mips chips are also little endian,
so there are a few systems out there that would use spim.

My goal was to use the Onion Omega 2 kits to set up i2c sensors and control
lights and such around my house.  I also got a MT7688 based router, to make
a 9front based wifi router with.  And a MT7621 based router, which is
similar but dual core.

So, if anyone wants to help, I do have some little endian mips hardware to
test on.

Re: [9front] 9front on spim

[ori]

Quoth adventures in9 <adventuresin9@gmail.com>:
> I'll make this a separate thread for anyone who wants to talk about spim
> specifically.
> 
> I saw spim mentioned in the Port Status thread.  I haven't had any trouble
> building for spim, and some of the missing parts might just be because it
> is building off code stored in "mips" but does it with with the "spim"
> compiler, which is really just a script with the mips compiler set with a
> little-endian flag.
> 
> As for hardware, I've been using the Onion Omega 2.  I've also found
> several Chinese router boards that also use the MT7688.  I have got it to
> start booting on the Onion, but haven't finished it up yet.  I haven't
> touched it in a couple months, but what I have done so far is here:
> 
> https://github.com/adventuresin9/9front-mt7688
> 
> I'm pretty sure the Loongson and Ingenic mips chips are also little endian,
> so there are a few systems out there that would use spim.
> 
> My goal was to use the Onion Omega 2 kits to set up i2c sensors and control
> lights and such around my house.  I also got a MT7688 based router, to make
> a 9front based wifi router with.  And a MT7621 based router, which is
> similar but dual core.
> 
> So, if anyone wants to help, I do have some little endian mips hardware to
> test on.

Awesome. Please send patches when you're ready!

Re: [9front] 9front on spim

[Frank D. Engel, Jr.]

For reference, I am actually doing the build from an arm64 system 
(Raspberry Pi) and getting "up to date" when running sysupdate, and when 
I cd /sys/src then objtype=spim mk all I am getting a bunch of the 
typical working stuff followed by variations of the following at the end:


pcc -c -Fw -D_POSIX_SOURCE -D_PLAN9_SOURCE -I. ../../../libc/9sys/idn.c
cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ctime.c:1 Could not find 
include file <stdarg.h>
pcc: 0c: cpp 201201: error
mk: pcc -c -Fw ...  : exit status=rc 201186: pcc 201193: 0c: cpp 201201: 
error<eof> not a function
<eof> syntax error, last name: ex

cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ../../../libc/9sys/idn.c:2 
Could not find include file <stdarg.h>
cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ../../../libc/port/crypt.c:11 
Could not find include file <stdarg.h>
pcc: 0c: cpp 201197: error
pcc: 0c: cpp 201200: error
cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 
../../../libc/port/getfields.c:2 Could not find include file <stdarg.h>
<eof> syntax error, last name: Rune
<eof> not a function
<eof> syntax error, last name: ext
pcc: 0c: cpp 201204: error
mk: for (i in ...  : exit status=rc 201180: rc 201182: mk 201183: error
mk: cd lib mk all  : exit status=rc 201174: mk 201176: error
mk: mk lib.all mk ...  : exit status=rc 201171: mk 201173: error
mk: date for (i ...  : exit status=rc 200132: rc 201169: mk 201170: error
glenda@brains1% <eof> not a function
<eof> syntax error, last name: runestrstr



On 7/4/22 7:21 PM, ori@eigenstate.org wrote:
> Quoth adventures in9 <adventuresin9@gmail.com>:
>> I'll make this a separate thread for anyone who wants to talk about spim
>> specifically.
>>
>> I saw spim mentioned in the Port Status thread.  I haven't had any trouble
>> building for spim, and some of the missing parts might just be because it
>> is building off code stored in "mips" but does it with with the "spim"
>> compiler, which is really just a script with the mips compiler set with a
>> little-endian flag.
>>
>> As for hardware, I've been using the Onion Omega 2.  I've also found
>> several Chinese router boards that also use the MT7688.  I have got it to
>> start booting on the Onion, but haven't finished it up yet.  I haven't
>> touched it in a couple months, but what I have done so far is here:
>>
>> https://github.com/adventuresin9/9front-mt7688
>>
>> I'm pretty sure the Loongson and Ingenic mips chips are also little endian,
>> so there are a few systems out there that would use spim.
>>
>> My goal was to use the Onion Omega 2 kits to set up i2c sensors and control
>> lights and such around my house.  I also got a MT7688 based router, to make
>> a 9front based wifi router with.  And a MT7621 based router, which is
>> similar but dual core.
>>
>> So, if anyone wants to help, I do have some little endian mips hardware to
>> test on.
> Awesome. Please send patches when you're ready!
>
>

Re: [9front] 9front on spim

[adventures in9]

I just tried a /sys/src mk all and it is giving me similar errors too.

cd 9src
mk all
0l  -o 0.stty stty.0
0l  -o 0.tar tar.0
0l  -o 0.cc cc.0
lock: undefined: _tas
(856)    JAL    ,_tas+0(SB)
lock: undefined: _tas
(856)    JAL    ,_tas+0(SB)
lock: undefined: _tas
(856)    JAL    ,_tas+0(SB)
mk: 0l  -o ...  : exit status=rc 324833: vl 324839: error
mk: cd 9src mk all  : exit status=rc 324830: mk 324832: error
mk: mk lib.all mk ...  : exit status=rc 324142: mk 324829: error
mk: date for (i ...  : exit status=rc 323149: rc 324140: mk 324141: error

I also tried it with objtype=mips, and it builds without error.

I ran objtype=spim and mk all again, and got a similar error.

Then just to be sure, I did mk clean, and ran mk all again, and this
time it it finished building.

0c -FTVw cpufp.c
warning: v8e.c:89 param declared and not used: file
0c -FTVw mem.c
0c -FTVw sys.c
0l  -o 0.out v8e.0 cpu.0 cpubcd.0 cpufp.0 mem.0 sys.0
Mon Jul  4 17:33:38 PDT 2022
cpu% echo $objtype
spim
cpu%

So it might just be an issue of something getting left behind from
building for other architectures.  Like I mentioned before, spim is
sort of an odd one because the 0c compiler is just a script that calls
"exec vc -l $*"

Try mk clean and then build just objtype=spim

On Mon, Jul 4, 2022 at 5:02 PM Frank D. Engel, Jr. <fde101@fjrhome.net> wrote:
>
> For reference, I am actually doing the build from an arm64 system
> (Raspberry Pi) and getting "up to date" when running sysupdate, and when
> I cd /sys/src then objtype=spim mk all I am getting a bunch of the
> typical working stuff followed by variations of the following at the end:
>
>
> pcc -c -Fw -D_POSIX_SOURCE -D_PLAN9_SOURCE -I. ../../../libc/9sys/idn.c
> cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ctime.c:1 Could not find
> include file <stdarg.h>
> pcc: 0c: cpp 201201: error
> mk: pcc -c -Fw ...  : exit status=rc 201186: pcc 201193: 0c: cpp 201201:
> error<eof> not a function
> <eof> syntax error, last name: ex
>
> cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ../../../libc/9sys/idn.c:2
> Could not find include file <stdarg.h>
> cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ../../../libc/port/crypt.c:11
> Could not find include file <stdarg.h>
> pcc: 0c: cpp 201197: error
> pcc: 0c: cpp 201200: error
> cpp: /sys/include/ape/fmt.h:26 ./libc.h:21
> ../../../libc/port/getfields.c:2 Could not find include file <stdarg.h>
> <eof> syntax error, last name: Rune
> <eof> not a function
> <eof> syntax error, last name: ext
> pcc: 0c: cpp 201204: error
> mk: for (i in ...  : exit status=rc 201180: rc 201182: mk 201183: error
> mk: cd lib mk all  : exit status=rc 201174: mk 201176: error
> mk: mk lib.all mk ...  : exit status=rc 201171: mk 201173: error
> mk: date for (i ...  : exit status=rc 200132: rc 201169: mk 201170: error
> glenda@brains1% <eof> not a function
> <eof> syntax error, last name: runestrstr
>
>
>
> On 7/4/22 7:21 PM, ori@eigenstate.org wrote:
> > Quoth adventures in9 <adventuresin9@gmail.com>:
> >> I'll make this a separate thread for anyone who wants to talk about spim
> >> specifically.
> >>
> >> I saw spim mentioned in the Port Status thread.  I haven't had any trouble
> >> building for spim, and some of the missing parts might just be because it
> >> is building off code stored in "mips" but does it with with the "spim"
> >> compiler, which is really just a script with the mips compiler set with a
> >> little-endian flag.
> >>
> >> As for hardware, I've been using the Onion Omega 2.  I've also found
> >> several Chinese router boards that also use the MT7688.  I have got it to
> >> start booting on the Onion, but haven't finished it up yet.  I haven't
> >> touched it in a couple months, but what I have done so far is here:
> >>
> >> https://github.com/adventuresin9/9front-mt7688
> >>
> >> I'm pretty sure the Loongson and Ingenic mips chips are also little endian,
> >> so there are a few systems out there that would use spim.
> >>
> >> My goal was to use the Onion Omega 2 kits to set up i2c sensors and control
> >> lights and such around my house.  I also got a MT7688 based router, to make
> >> a 9front based wifi router with.  And a MT7621 based router, which is
> >> similar but dual core.
> >>
> >> So, if anyone wants to help, I do have some little endian mips hardware to
> >> test on.
> > Awesome. Please send patches when you're ready!
> >
> >
>

Re: [9front] 9front on spim

[Frank D. Engel, Jr.]

I tried this:

objtype=spim mk clean

objtype=spim mk all


I continue to get errors related to <stdarg.h> being missing.

0a mips/getcallerpc.s
cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ctime.c:1 Could not find 
include file <stdarg.h>
0a mips/getfcr.s
pcc: 0c: cpp 221059: error
<eof> not a function
<eof> syntax error, last name: ex
cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ../../../libc/port/crypt.c:11 
Could not find include file <stdarg.h>
mk: pcc -c -Fw ...  : exit status=rc 221046: pcc 221051: 0c: cpp 221059: 
errorpcc: 0c: cpp 221057: error

<eof> not a function
<eof> syntax error, last name: ext
cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 rerrstr.c:1 Could not find 
include file <stdarg.h>
pcc: 0c: cpp 221071: error
mk: for (i in ...  : exit status=rc 221036: rc 221038: mk 221039: error
mk: cd lib mk all  : exit status=rc 221030: mk 221032: error
mk: mk lib.all mk ...  : exit status=rc 221027: mk 221029: error
mk: date for (i ...  : exit status=rc 217389: rc 221025: mk 221026: error


On 7/4/22 8:40 PM, adventures in9 wrote:
> I just tried a /sys/src mk all and it is giving me similar errors too.
>
> cd 9src
> mk all
> 0l  -o 0.stty stty.0
> 0l  -o 0.tar tar.0
> 0l  -o 0.cc cc.0
> lock: undefined: _tas
> (856)    JAL    ,_tas+0(SB)
> lock: undefined: _tas
> (856)    JAL    ,_tas+0(SB)
> lock: undefined: _tas
> (856)    JAL    ,_tas+0(SB)
> mk: 0l  -o ...  : exit status=rc 324833: vl 324839: error
> mk: cd 9src mk all  : exit status=rc 324830: mk 324832: error
> mk: mk lib.all mk ...  : exit status=rc 324142: mk 324829: error
> mk: date for (i ...  : exit status=rc 323149: rc 324140: mk 324141: error
>
> I also tried it with objtype=mips, and it builds without error.
>
> I ran objtype=spim and mk all again, and got a similar error.
>
> Then just to be sure, I did mk clean, and ran mk all again, and this
> time it it finished building.
>
> 0c -FTVw cpufp.c
> warning: v8e.c:89 param declared and not used: file
> 0c -FTVw mem.c
> 0c -FTVw sys.c
> 0l  -o 0.out v8e.0 cpu.0 cpubcd.0 cpufp.0 mem.0 sys.0
> Mon Jul  4 17:33:38 PDT 2022
> cpu% echo $objtype
> spim
> cpu%
>
> So it might just be an issue of something getting left behind from
> building for other architectures.  Like I mentioned before, spim is
> sort of an odd one because the 0c compiler is just a script that calls
> "exec vc -l $*"
>
> Try mk clean and then build just objtype=spim
>
> On Mon, Jul 4, 2022 at 5:02 PM Frank D. Engel, Jr. <fde101@fjrhome.net> wrote:
>> For reference, I am actually doing the build from an arm64 system
>> (Raspberry Pi) and getting "up to date" when running sysupdate, and when
>> I cd /sys/src then objtype=spim mk all I am getting a bunch of the
>> typical working stuff followed by variations of the following at the end:
>>
>>
>> pcc -c -Fw -D_POSIX_SOURCE -D_PLAN9_SOURCE -I. ../../../libc/9sys/idn.c
>> cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ctime.c:1 Could not find
>> include file <stdarg.h>
>> pcc: 0c: cpp 201201: error
>> mk: pcc -c -Fw ...  : exit status=rc 201186: pcc 201193: 0c: cpp 201201:
>> error<eof> not a function
>> <eof> syntax error, last name: ex
>>
>> cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ../../../libc/9sys/idn.c:2
>> Could not find include file <stdarg.h>
>> cpp: /sys/include/ape/fmt.h:26 ./libc.h:21 ../../../libc/port/crypt.c:11
>> Could not find include file <stdarg.h>
>> pcc: 0c: cpp 201197: error
>> pcc: 0c: cpp 201200: error
>> cpp: /sys/include/ape/fmt.h:26 ./libc.h:21
>> ../../../libc/port/getfields.c:2 Could not find include file <stdarg.h>
>> <eof> syntax error, last name: Rune
>> <eof> not a function
>> <eof> syntax error, last name: ext
>> pcc: 0c: cpp 201204: error
>> mk: for (i in ...  : exit status=rc 201180: rc 201182: mk 201183: error
>> mk: cd lib mk all  : exit status=rc 201174: mk 201176: error
>> mk: mk lib.all mk ...  : exit status=rc 201171: mk 201173: error
>> mk: date for (i ...  : exit status=rc 200132: rc 201169: mk 201170: error
>> glenda@brains1% <eof> not a function
>> <eof> syntax error, last name: runestrstr
>>
>>
>>
>> On 7/4/22 7:21 PM, ori@eigenstate.org wrote:
>>> Quoth adventures in9 <adventuresin9@gmail.com>:
>>>> I'll make this a separate thread for anyone who wants to talk about spim
>>>> specifically.
>>>>
>>>> I saw spim mentioned in the Port Status thread.  I haven't had any trouble
>>>> building for spim, and some of the missing parts might just be because it
>>>> is building off code stored in "mips" but does it with with the "spim"
>>>> compiler, which is really just a script with the mips compiler set with a
>>>> little-endian flag.
>>>>
>>>> As for hardware, I've been using the Onion Omega 2.  I've also found
>>>> several Chinese router boards that also use the MT7688.  I have got it to
>>>> start booting on the Onion, but haven't finished it up yet.  I haven't
>>>> touched it in a couple months, but what I have done so far is here:
>>>>
>>>> https://github.com/adventuresin9/9front-mt7688
>>>>
>>>> I'm pretty sure the Loongson and Ingenic mips chips are also little endian,
>>>> so there are a few systems out there that would use spim.
>>>>
>>>> My goal was to use the Onion Omega 2 kits to set up i2c sensors and control
>>>> lights and such around my house.  I also got a MT7688 based router, to make
>>>> a 9front based wifi router with.  And a MT7621 based router, which is
>>>> similar but dual core.
>>>>
>>>> So, if anyone wants to help, I do have some little endian mips hardware to
>>>> test on.
>>> Awesome. Please send patches when you're ready!
>>>
>>>

Re: [9front] 9front on spim

[Lyndon Nerenberg (VE7TFX/VE6BBM)]

That's interesting.  I have a couple of those Omega 2s lying around
here, somewhere.  I never found them to be very useful (to me) in
their native form, but having a Plan 9 on them makes them very viable
for remote sensing type applications.

If I can figure out which box they are lurking in I'll dig them out
and take your code for a spin.

--lyndon

Fwd: [9front] 9front on spim

[adventures in9]

---------- Forwarded message ---------
From: adventures in9 <adventuresin9@gmail.com>
Date: Mon, Jul 4, 2022 at 5:55 PM
Subject: Re: [9front] 9front on spim
To: Lyndon Nerenberg (VE7TFX/VE6BBM) <lyndon@orthanc.ca>


That would be cool.  I have the dock for mine that does uart over usb,
and I got that working.  I have not got around to drivers for the wifi
or ethernet adapter, because right now it crashes during boot.  The
video I made covering what I did so far can be found here;
https://youtu.be/m75mf-j4V5c

There was work on an open source wifi driver, so I was hoping to
borrow from it to get a 9front one working.  It also has i2c, and I
wanted to hook up various sensors over that.  I've done it with
Raspberry Pi's, and the Plan 9 way of remotely mounting sensor data as
files over the network is nice.

On Mon, Jul 4, 2022 at 5:11 PM Lyndon Nerenberg (VE7TFX/VE6BBM)
<lyndon@orthanc.ca> wrote:
>
> That's interesting.  I have a couple of those Omega 2s lying around
> here, somewhere.  I never found them to be very useful (to me) in
> their native form, but having a Plan 9 on them makes them very viable
> for remote sensing type applications.
>
> If I can figure out which box they are lurking in I'll dig them out
> and take your code for a spin.
>
> --lyndon

Re: Fwd: [9front] 9front on spim

[Lyndon Nerenberg (VE7TFX/VE6BBM)]

It's generally frowned upon to forward private correspondance to
a mailing list.

--lyndon