Building mega65-core

Snoopy · Dec 10th 2019

Today I used some time trying to build the mega65-core from github (with Xubuntu 18.04.3 LTS 64bit).

The very good build documentation describes it step by step.

After some "make"s I did, I got stuck with the following error:

Code

Running first pass to get size of routines...
sh: 1: Ophis/bin/ophis: not found
Reading sizes of routines...
src/preprocess.c:170:main():Could not open temp.map for reading. Did Ophis fail due to an error in one of the assembly files?
Makefile:45: recipe for target 'c64/kernal/OUT.BIN' failed
make[1]: *** [c64/kernal/OUT.BIN] Error 255
make[1]: Verzeichnis „/home/snoopy/git_root/mega65-core/src/open-roms“ wird verlassen
Makefile:514: recipe for target 'src/open-roms/newc65' failed
make: *** [src/open-roms/newc65] Error 2

What is this "Ophis/bin/ophis"? It isn't mentioned in the build documentation?

Here is the full log of the last "make" run:

Code

snoopy@snoopy-vb:~/git_root/mega65-core$ make
( cd src/open-roms ; make newc65 )
make[1]: Verzeichnis „/home/snoopy/git_root/mega65-core/src/open-roms“ wird betreten
src/preprocess -d c64/kernal -l e4d3 -h ffff
Found 113 source files
function 'ioinit' should be placed at $FDA3
function 's' should be placed at $EA81
function 's' should be placed at $FFAB
function 's' should be placed at $FFEA
function 's' should be placed at $FF81
function 's' should be placed at $FFA8
function 's' should be placed at $FFB1
function 's' should be placed at $FF93
function 's' should be placed at $FF96
function 's' should be placed at $FFCC
function 's' should be placed at $FFA2
function 's' should be placed at $FFCF
function 's' should be placed at $FFF3
function 's' should be placed at $FFC6
function 's' should be placed at $FF90
function 's' should be placed at $FFAE
function 's' should be placed at $FFB4
function 's' should be placed at $FFF0
function 's' should be placed at $FFE1
function 's' should be placed at $FFD2
function 'setup_vicii' should be placed at $E5A0
function 's' should be placed at $FF9F
function 's' should be placed at $FFFA
function 's' should be placed at $FFE4
function 's' should be placed at $FFB7
function 's' should be placed at $FFBD
function 's' should be placed at $FF99
function 's' should be placed at $FFA5
function 's' should be placed at $FFDE
function 's' should be placed at $FFBA
function 's' should be placed at $FFC0
function 's' should be placed at $FFFE
function 'default_irq_handler' should be placed at $EA31
function 's' should be placed at $FF84
function 's' should be placed at $FFD8
function 's' should be placed at $FF8A
function 's' should be placed at $EA7E
function 's' should be placed at $FF9C
function 'clear_screen' should be placed at $E544
function 's' should be placed at $FF87
function 's' should be placed at $FFE7
function 's' should be placed at $FFED
function 's' should be placed at $FFC9
function 's' should be placed at $FFD5
function 's' should be placed at $FFC3
function 's' should be placed at $FFDB
function 's' should be placed at $FF8D
function 's' should be placed at $FFFC
Got fixed routine addresses
Sorted.
$FFFFFFFF : ,aliases.s
$FFFFFFFF : acptr.s
$FFFFFFFF : blink_cursor.s
$FFFFFFFF : cartridge_check.s
$FFFFFFFF : chkin.s
$FFFFFFFF : chkout.s
$FFFFFFFF : chrin.s
$FFFFFFFF : chrout.s
$FFFFFFFF : cint.s
$FFFFFFFF : ciout.s
$FFFFFFFF : close.s
$FFFFFFFF : close_all_files.s
$FFFFFFFF : clrchn.s
$FFFFFFFF : errors.s
$FFFFFFFF : get_screen_size.s
$FFFFFFFF : getin.s
$FFFFFFFF : home_cursor.s
$FFFFFFFF : iec_assert_atn.s
$FFFFFFFF : iec_assert_clk_and_data.s
$FFFFFFFF : iec_assert_clk_release_data.s
$FFFFFFFF : iec_assert_data.s
$FFFFFFFF : iec_release_atn.s
$FFFFFFFF : iec_release_clk.s
$FFFFFFFF : iec_release_clk_and_data.s
$FFFFFFFF : iec_rx_byte.s
$FFFFFFFF : iec_turnaround_to_listen.s
$FFFFFFFF : iec_tx_byte.s
$FFFFFFFF : iec_wait20us.s
$FFFFFFFF : iec_wait60us.s
$FFFFFFFF : iec_wait_for_clock_assert.s
$FFFFFFFF : iec_wait_for_clock_release.s
$FFFFFFFF : iec_wait_for_data_release.s
$FFFFFFFF : irq_handler.s
$FFFFFFFF : keyboard_matrix.s
$FFFFFFFF : listen.s
$FFFFFFFF : load.s
$FFFFFFFF : membot.s
$FFFFFFFF : memtop.s
$FFFFFFFF : nmi_handler.s
$FFFFFFFF : open.s
$FFFFFFFF : plot.s
$FFFFFFFF : printf.s
$FFFFFFFF : ramtas.s
$FFFFFFFF : rdtim.s
$FFFFFFFF : readst.s
$FFFFFFFF : reset_entry.s
$FFFFFFFF : restor.s
$FFFFFFFF : save.s
$FFFFFFFF : scan_keyboard.s
$FFFFFFFF : second.s
$FFFFFFFF : set_pointer_to_current_screen_line.s
$FFFFFFFF : setlfs.s
$FFFFFFFF : setmsg.s
$FFFFFFFF : setnam.s
$FFFFFFFF : settim.s
$FFFFFFFF : settmo.s
$FFFFFFFF : stop.s
$FFFFFFFF : talk.s
$FFFFFFFF : tksa.s
$FFFFFFFF : under_rom_routines.s
$FFFFFFFF : unlsn.s
$FFFFFFFF : untlk.s
$FFFFFFFF : update_jiffy_clock.s
$FFFFFFFF : vector.s
$FFFFFFFF : wait1ms.s
$E544 : e544.clear_screen.s
$E5A0 : e5a0.setup_vicii.s
$EA31 : ea31.default_irq_handler.s
$EA7E : ea7e.s
$EA81 : ea81.s
$FDA3 : fda3.ioinit.s
$FF81 : ff81.s
$FF84 : ff84.s
$FF87 : ff87.s
$FF8A : ff8a.s
$FF8D : ff8d.s
$FF90 : ff90.s
$FF93 : ff93.s
$FF96 : ff96.s
$FF99 : ff99.s
$FF9C : ff9c.s
$FF9F : ff9f.s
$FFA2 : ffa2.s
$FFA5 : ffa5.s
$FFA8 : ffa8.s
$FFAB : ffab.s
$FFAE : ffae.s
$FFB1 : ffb1.s
$FFB4 : ffb4.s
$FFB7 : ffb7.s
$FFBA : ffba.s
$FFBD : ffbd.s
$FFC0 : ffc0.s
$FFC3 : ffc3.s
$FFC6 : ffc6.s
$FFC9 : ffc9.s
$FFCC : ffcc.s
$FFCF : ffcf.s
$FFD2 : ffd2.s
$FFD5 : ffd5.s
$FFD8 : ffd8.s
$FFDB : ffdb.s
$FFDE : ffde.s
$FFE1 : ffe1.s
$FFE4 : ffe4.s
$FFE7 : ffe7.s
$FFEA : ffea.s
$FFED : ffed.s
$FFF0 : fff0.s
$FFF3 : fff3.s
$FFFA : fffa.s
$FFFC : fffc.s
$FFFE : fffe.s
Running first pass to get size of routines...
sh: 1: Ophis/bin/ophis: not found
Reading sizes of routines...
src/preprocess.c:170:main():Could not open temp.map for reading. Did Ophis fail due to an error in one of the assembly files?
Makefile:45: recipe for target 'c64/kernal/OUT.BIN' failed
make[1]: *** [c64/kernal/OUT.BIN] Error 255
make[1]: Verzeichnis „/home/snoopy/git_root/mega65-core/src/open-roms“ wird verlassen
Makefile:514: recipe for target 'src/open-roms/newc65' failed
make: *** [src/open-roms/newc65] Error 2
snoopy@snoopy-vb:~/git_root/mega65-core$

Display More

spasskl · Dec 10th 2019

Have you cloned "https://github.com/gardners/Ophis" to your mega65-core directory ?

Look at .gitmodules -> Submodules

Snoopy · Dec 10th 2019

Quote from spasskl

Have you cloned "https://github.com/gardners/Ophis" to your mega65-core directory ?

Look at .gitmodules -> Submodules

Thanks for the hint!

I only cloned "fpgajtag" as mentioned in the build doc as the only 3rd party program.

I'll try it with the Ophis and report about it

FeralChild · Dec 10th 2019

Probably you have to execute:

Code

git submodule init
git submodule update

in the cloned repository. This should fetch Ophis and other needed submodules.

Snoopy · Dec 10th 2019

Thanks for the answers.

I tried both hints:

Code

snoopy@snoopy-vb:~/git_root/mega65-core$ git clone https://github.com/gardners/Ophis.git
fatal: Zielpfad 'Ophis' existiert bereits und ist kein leeres Verzeichnis.
snoopy@snoopy-vb:~/git_root/mega65-core$ git submodule init
snoopy@snoopy-vb:~/git_root/mega65-core$ git submodule update

If I try "make" again I get the same errors as mentioned in my first posting.

Maybe something is missing in my cloned git-rep?

Code

snoopy@snoopy-vb:~/git_root/mega65-core$ ll
insgesamt 428
drwxrwxr-x 19 snoopy snoopy 4096 Dez 10 18:07 ./
drwxrwxr-x 4 snoopy snoopy 4096 Dez 10 23:07 ../
drwxrwxr-x 2 snoopy snoopy 4096 Dez 10 15:31 65xx_and_c64_docs/
drwxrwxr-x 3 snoopy snoopy 4096 Dez 10 16:24 assets/
drwxrwxr-x 2 snoopy snoopy 4096 Dez 10 18:14 bin/
drwxrwxr-x 2 snoopy snoopy 4096 Dez 10 17:42 cbmconvert/
drwxrwxr-x 17 snoopy snoopy 4096 Dez 10 17:38 cc65/
drwxrwxr-x 2 snoopy snoopy 4096 Dez 10 15:31 d81-files/
drwxrwxr-x 3 snoopy snoopy 4096 Dez 10 16:24 docs/
drwxrwxr-x 8 snoopy snoopy 4096 Dez 10 17:37 ghdl/
drwxrwxr-x 9 snoopy snoopy 4096 Dez 10 18:07 .git/
-rw-rw-r-- 1 snoopy snoopy 1679 Dez 10 16:24 .gitignore
-rw-rw-r-- 1 snoopy snoopy 726 Dez 10 16:24 .gitmodules
-rw-rw-r-- 1 snoopy snoopy 74043 Dez 10 16:24 iomap.txt
drwxrwxr-x 4 snoopy snoopy 4096 Dez 10 16:24 ipcore_dir/
drwxrwxr-x 2 snoopy snoopy 4096 Dez 10 15:31 isework/
drwxrwxr-x 28 snoopy snoopy 12288 Dez 10 18:06 iverilog/
drwxrwxr-x 2 snoopy snoopy 4096 Dez 10 16:24 KickAss/
-rw-rw-r-- 1 snoopy snoopy 7785 Dez 10 15:31 LICENSE.txt
-rw-rw-r-- 1 snoopy snoopy 28901 Dez 10 16:24 Makefile
-rwxrwxr-x 1 snoopy snoopy 20246 Dez 10 15:31 Makefile.ise*
-rw-rw-r-- 1 snoopy snoopy 47699 Dez 10 15:31 mega65-ddr.xise
-rw-rw-r-- 1 snoopy snoopy 47702 Dez 10 15:31 mega65-nonddr.xise
-rwxrwxr-x 1 snoopy snoopy 294 Dez 10 16:24 megaphone-write-flash.sh*
-rw-rw-r-- 1 snoopy snoopy 2335 Dez 10 16:24 megaphone-write-flash.tcl
-rw-rw-r-- 1 snoopy snoopy 36440 Dez 10 18:07 monitor.list
-rw-rw-r-- 1 snoopy snoopy 19970 Dez 10 18:07 monitor.map
-rwxrwxr-x 1 snoopy snoopy 89 Dez 10 15:31 nexys4ddr-run-bitstream.sh*
-rwxrwxr-x 1 snoopy snoopy 294 Dez 10 15:31 nexys4ddr-write-flash.sh*
-rw-rw-r-- 1 snoopy snoopy 2335 Dez 10 15:31 nexys4ddr-write-flash.tcl
drwxrwxr-x 9 snoopy snoopy 4096 Dez 10 17:37 Ophis/
-rw-rw-r-- 1 snoopy snoopy 591 Dez 10 15:31 README.md
-rwxrwxr-x 1 snoopy snoopy 955 Dez 10 15:31 record-m65*
-rwxrwxr-x 1 snoopy snoopy 6717 Dez 10 15:31 run_ise*
drwxrwxr-x 2 snoopy snoopy 4096 Dez 10 18:19 sdcard-files/
drwxrwxr-x 14 snoopy snoopy 4096 Dez 10 18:07 src/
-rw-rw-r-- 1 snoopy snoopy 576 Dez 10 15:31 TRADEMARKS.md
drwxrwxr-x 2 snoopy snoopy 4096 Dez 10 16:24 vivado/
-rwxrwxr-x 1 snoopy snoopy 224 Dez 10 16:24 vivado_wrapper*
-rwxrwxr-x 1 snoopy snoopy 738 Dez 10 15:31 watch-m65*
snoopy@snoopy-vb:~/git_root/mega65-core$

Display More

That's my first try with the build and I don't know where I should start with the search for the error.

Any hint is welcome!

mega65 · Dec 10th 2019

We should add the git submodule commmands as a dummy dependency for Ophis in the Makefile, so that this is more bullet-proof.

LG

Paul

darkvision · Dec 10th 2019

Hast Du es mal mit

Code

git clone --recurse-submodules https://github.com/MEGA65/mega65-core.git
cd mega*
make -j8

versucht? Das lädt hier alles inkl. aller (auch unnötiger) submodules in ein neues repo.

Ein make -j8 (Zahl hängt von den Prozessor-Kernen ab...) compiliert hier ein paar viele Dateien...

Läuft noch... keine Ahnung wie lange... cmake wäre da ausgabefreudiger

P.S. zu früh gefreut...

Code

# Now remove the dummy definitions of UART_TX_CTRL and uart_rx, as we will use the actual VHDL implementations of them.
cat src/vhdl/uart_monitor.vhdl.tmp | awk 'BEGIN { echo=1; } {if ($1=="--"&&$2=="Generated"&&$3=="from"&&$4=="Verilog") { if ($6=="UART_TX_CTRL"||$6=="uart_rx") echo=0; else echo=1; } if (echo) print; }' > src/vhdl/uart_monitor.vhdl
echo MOOSE vivado/mega65r1.xpr from vivado/mega65r1_gen.tcl
echo MOOSE vivado/nexys4.xpr from vivado/nexys4_gen.tcl
echo MOOSE vivado/nexys4ddr.xpr from vivado/nexys4ddr_gen.tcl
MOOSE vivado/mega65r1.xpr from vivado/mega65r1_gen.tcl
./vivado_wrapper -mode batch -source vivado/mega65r1_gen.tcl
MOOSE vivado/nexys4.xpr from vivado/nexys4_gen.tcl
./vivado_wrapper -mode batch -source vivado/nexys4_gen.tcl
MOOSE vivado/nexys4ddr.xpr from vivado/nexys4ddr_gen.tcl
./vivado_wrapper -mode batch -source vivado/nexys4ddr_gen.tcl
ls: Zugriff auf '/opt/Xilinx/Vivado/*' nicht möglich: Datei oder Verzeichnis nicht gefunden
ls: Zugriff auf '/opt/Xilinx/Vivado/*' nicht möglich: Datei oder Verzeichnis nicht gefunden
./vivado_wrapper: Zeile 3: /settings64.sh: Datei oder Verzeichnis nicht gefunden
vivado -mode batch -source vivado/mega65r1_gen.tcl
./vivado_wrapper: Zeile 3: /settings64.sh: Datei oder Verzeichnis nicht gefunden
vivado -mode batch -source vivado/nexys4_gen.tcl
./vivado_wrapper: Zeile 5: vivado: Kommando nicht gefunden.
./vivado_wrapper: Zeile 5: vivado: Kommando nicht gefunden.
make: *** [Makefile:614: vivado/mega65r1.xpr] Fehler 127
make: *** Es wird auf noch nicht beendete Prozesse gewartet....
make: *** [Makefile:614: vivado/nexys4.xpr] Fehler 127
ls: Zugriff auf '/opt/Xilinx/Vivado/*' nicht möglich: Datei oder Verzeichnis nicht gefunden
./vivado_wrapper: Zeile 3: /settings64.sh: Datei oder Verzeichnis nicht gefunden
vivado -mode batch -source vivado/nexys4ddr_gen.tcl
./vivado_wrapper: Zeile 5: vivado: Kommando nicht gefunden.
make: *** [Makefile:614: vivado/nexys4ddr.xpr] Fehler 127
markus@jupiter:~/m65/mega65-core$

Display More

Sieht so aus als fehlt mir noch etwas in /opt ... mal sehn...

Snoopy · Dec 10th 2019

Maybe there is an error in the build doc?

There are two(!) git rep mentioned to be cloned, but only in the first command line is a git clone command.

What is the second git rep to be cloned?

darkvision · Dec 10th 2019

Sorry for the German response... looks like me is missing the XILINX SDK... maybe download all submodules will also download maybe unneeded submodules.

Looks like the XILINX SDK takes up to 26Gb... way to much for my slow DSL connection for now... maybe i will have a look at it on weekend...

mega65 · Dec 10th 2019

Meanwhile, can you please do:

find mega65-core -ls > fileforpaul.txt

from outside the directory, then zip that up and send it to me. I'll try to figure out what is going on.

LG

Paul.

darkvision · Dec 11th 2019

Snoopy ...

Code

git clone --recurse-submodules https://github.com/MEGA65/mega65-core.git
cd mega*
rm -rf src/vhdl/*
git clone https://github.com/cambridgehackers/fpgajtag.git
make -j8

This will end the assembly without error, but maybe the VHDL code is still missing. Maybe that will help you, anyway, 26Gb is way to much download for me here right now so i cannot help testing the VHDL stuff...

Snoopy · Dec 11th 2019

Quote from darkvision

Sorry for the German response...

Wegen mir kein Problem! Ich bin deutscher Muttersprachler, soweit man das von jemanden sagen kann, der in der niederbayrischen Provinz aufgewachsen ist.

Ich wollte die Frage hier nur auf englisch stellen, damit möglichst viele Interessierte und - vor allem auch Paul - mitlesen und eventuell mithelfen können.

I'll try the hints as soon as I have enough sparetime for them and will report about it. Thanks for all the help!

Snoopy · Dec 11th 2019

Quote from darkvision

Snoopy ...

Code

git clone --recurse-submodules https://github.com/MEGA65/mega65-core.git

cd mega*

rm -rf src/vhdl/*

git clone https://github.com/cambridgehackers/fpgajtag.git

make -j8

This will end the assembly without error, but maybe the VHDL code is still missing. Maybe that will help you, anyway, 26Gb is way to much download for me here right now so i cannot help testing the VHDL stuff...

Perfect! This worked! Many thanks!

The "Xilinx Vivado WebPACK edition" I hadn't downloaded yet. This will be the next step.

Snoopy · Dec 11th 2019

Quote from mega65

Meanwhile, can you please do:

find mega65-core -ls > fileforpaul.txt

from outside the directory, then zip that up and send it to me. I'll try to figure out what is going on.

LG

Paul.

Display More

Hey Paul, I have tried this hint from darkvision and that works.

For your interest I have attached the txt file (created as mentioned above). Maybe you can see what is wrong with my first attempt?

Thanks for all!

mega65 · Dec 11th 2019

Thanks -- I'll take a look.

LG

Paul.

mega65 · Dec 11th 2019

Quick look: It seems in the open-roms repo, that the Ophis sub-module hadn't been checked out.

So it is the open-roms Makefile that needs to do the git submodule stuff, if Ophis/bin/ophis doesn't exist.

That said, I think open-roms is moving, if not has already moved, to using KickAss instead.

LG

Paul.

Snoopy · Dec 11th 2019

I just tried again to build mega65-core as described in the build doc.

Here is the log of it:

Code

snoopy@snoopy-vb:~/git__root$ git clone https://github.com/MEGA65/mega65-core.git
Klone nach 'mega65-core' ...
remote: Enumerating objects: 1103, done.
remote: Counting objects: 100% (1103/1103), done.
remote: Compressing objects: 100% (406/406), done.
remote: Total 33882 (delta 813), reused 972 (delta 697), pack-reused 32779
Empfange Objekte: 100% (33882/33882), 24.37 MiB | 2.88 MiB/s, Fertig.
Löse Unterschiede auf: 100% (25432/25432), Fertig.
snoopy@snoopy-vb:~/git__root$ cd mega65-core/
snoopy@snoopy-vb:~/git__root/mega65-core$ git checkout development
Branch 'development' folgt nun Remote-Branch 'development' von 'origin'.
Zu neuem Branch 'development' gewechselt
snoopy@snoopy-vb:~/git__root/mega65-core$ git status
Auf Branch development
Ihr Branch ist auf demselben Stand wie 'origin/development'.
nichts zu committen, Arbeitsverzeichnis unverändert
snoopy@snoopy-vb:~/git__root/mega65-core$ git pull
Bereits aktuell.
snoopy@snoopy-vb:~/git__root/mega65-core$ cd ..
snoopy@snoopy-vb:~/git__root$ git clone https://github.com/cambridgehackers/fpgajtag.git
Klone nach 'fpgajtag' ...
remote: Enumerating objects: 3142, done.
remote: Total 3142 (delta 0), reused 0 (delta 0), pack-reused 3142
Empfange Objekte: 100% (3142/3142), 800.27 KiB | 1.43 MiB/s, Fertig.
Löse Unterschiede auf: 100% (2303/2303), Fertig.
snoopy@snoopy-vb:~/git__root$ cd fpgajtag
snoopy@snoopy-vb:~/git__root/fpgajtag$ make
make -C src
make[1]: Verzeichnis „/home/snoopy/git__root/fpgajtag/src“ wird betreten
cc -Wall -fPIC -I/usr/include/libusb-1.0 -I/opt/local/include/libusb-1.0 -I/usr/local//Cellar/libusb/1.0.18/include/libusb-1.0/ -g -c -o fpgajtag.o fpgajtag.c
cc -Wall -fPIC -I/usr/include/libusb-1.0 -I/opt/local/include/libusb-1.0 -I/usr/local//Cellar/libusb/1.0.18/include/libusb-1.0/ -g -c -o process.o process.c
cc -Wall -fPIC -I/usr/include/libusb-1.0 -I/opt/local/include/libusb-1.0 -I/usr/local//Cellar/libusb/1.0.18/include/libusb-1.0/ -g -c -o cortex.o cortex.c
cc -Wall -fPIC -I/usr/include/libusb-1.0 -I/opt/local/include/libusb-1.0 -I/usr/local//Cellar/libusb/1.0.18/include/libusb-1.0/ -g -c -o util.o util.c
gcc -g -o fpgajtag fpgajtag.o process.o cortex.o util.o -lpthread -lusb-1.0 -lz
make[1]: Verzeichnis „/home/snoopy/git__root/fpgajtag/src“ wird verlassen
snoopy@snoopy-vb:~/git__root/fpgajtag$ sudo cp src/fpgajtag /usr/local/bin
[sudo] Passwort für snoopy:
snoopy@snoopy-vb:~/git__root/fpgajtag$ cd ..
snoopy@snoopy-vb:~/git__root$ cd mega65-core
snoopy@snoopy-vb:~/git__root/mega65-core$ fpgajtag --version
Usage fpgajtag [ -a ][ -x ] [ -l ] [ -m ] [ -t ] [ -s <serialno> ] [ -i <index> ] [ -I interface ] [ -r ] <filename>
Programs Xilinx FPGA from a bitstream. The bitstream may be compressed and it may be contained an ELF executable.
If filename is an ELF executable, reads the data from the fpgajtag section of the file, otherwise it reads the whole file.
If the data is compressed with gzip, it is uncompressed.
If the data has a .bit header, the header is removed.
Unless using /dev/xdevcfg, scans USB for devices whose IDCODE matches the bitstream
and programs the device whose position matches index. Index defaults to 0.
When using /dev/xdevcfg, programs the device by writing the bitstream to /dev/xdevcfg.
A bitstream may be embedded into ELF executable application.elf via the following command:
objcopy --add-section fpgadata=system.bin.gz application.elf
Optional arguments:
-a Match any idcode
-x Write input file to /dev/xdevcfg on Zynq devices
-l Display a list of all FPGA jtag interfaces discovered on USB
-m Use FPGA Manager
-s <serialno> Use the jtag interface with the given serial number
-i <index> Program the 'index' device in the jtag chain that matches the IDCODE in the input file
-I <0|1> Which interface of FT2232 to use
-t Trace usb programming traffic
snoopy@snoopy-vb:~/git__root/mega65-core$ make
[... many messages lines removed ....]
Running first pass to get size of routines...
sh: 1: Ophis/bin/ophis: not found
Reading sizes of routines...
src/preprocess.c:170:main():Could not open temp.map for reading. Did Ophis fail due to an error in one of the assembly files?
Makefile:45: recipe for target 'c64/kernal/OUT.BIN' failed
make[1]: *** [c64/kernal/OUT.BIN] Error 255
make[1]: Verzeichnis „/home/snoopy/git__root/mega65-core/src/open-roms“ wird verlassen
Makefile:514: recipe for target 'src/open-roms/newc65' failed
make: *** [src/open-roms/newc65] Error 2
snoopy@snoopy-vb:~/git__root/mega65-core$

Display More

It's seems to be the same error as mentioned in my first post here.

Trying the steps described by darkvision in this post all works fine.

spasskl · Dec 11th 2019

70: running first pass to get size of routines...

72: sh: 1: Ophis/bin/ophis: not found

copy of Ophis an KickAssembler in the right places ?

PS: I think you need openjdk-11-jre installed to

FeralChild · Dec 11th 2019

mega65 Yes, Open ROMs has moved to KickAssembler completely (but from the logs I can see that some old version is being compiled) and does not need any submodules anymore (but requires Java now to run the KickAssembler).

The build system has changed since then - to build the newc65.rom file type 'make build/newc65.rom' (or 'make all' if you want all the builds). On my development branch you will find pre-built file (with release date and number embedded in ROM) in 'bin' subdirectory (this change is not merged yet).

Be warned, however - this ROM has a problem under XEMU, once we try to print a character in the top-right corner of the screen it damages screen memory and crashes. Reason unknown yet, the bug does not reproduce under VICE (I'm curious if this happens on a real Mega65 too...). Unfortunately, the startup banner from Retrofan (default for Mega65 build) does exactly this - I think I'll change it temporarily until the problem is resolved, so the ROM is at least usable under XEMU. Hypothesis: we might accidentally trigger one of the Mega65 extended opcodes.

spasskl Ophis needs Python 2.x only (not JRE), no Open ROMs version from the past required both Ophis and KickAssembler.

Unfortunately, I will be rather busy for the upcoming few days and I can't look into this right now.