Catalina 4.1 is now available

RossH

Catalina 4.1 is now available on Sourceforge - https://sourceforge.net/projects/catalina-c/

Catalina 4.1 is the latest release of Catalina, supporting both the Propeller 1 and the Propeller 2. It replaces all previous releases and is validated on the Propeller 2 P2_EVAL board for all memory model, library, optimization level and floating point combinations. It is also validated on many Propeller 1 platforms.

This release adds Multi-Memory Model support, allowing C programs that use different memory models to be run simultaneously. It also allows program overlays to be loaded and run from XMM RAM (Propeller 1) or SD Card (Propeller 1 and 2).

This release is intended to provide a baseline release for further incremental updates, which will mainly now concentrate on adding new plugins to Catalina's Propeller 2 support, until Catalina on the Propeller 2 has as rich a range of plugins as Catalina already has on the Propeller 1.

Catalina 4.1 includes both Windows and Linux releases.

This release will be supported for both the Propeller 1 and the Propeller 2.

Attached is the README file. Refer to it for more details.

Ross.

RossH

After struggling yet again with Code::Blocks in an attempt to reduce some of its complexity for new users, I have decided to introduce a new, simpler Integrated Development Environment (IDE) for Catalina.

This is the Catalina Geany IDE, based (as the name implies) on the Geany IDE (see https://www.geany.org/).

I decided to do this after realizing how flexible and simple Geany was to use. It is the first full-function IDE I have ever seen that tempts me to use it over a simple command-line interface. It didn't require any customization to use with Catalina, but I added some anyway - mainly because the number of configurable commands in the off-the-shelf version was not sufficient for use in an embedded environment where you typically need to do many more things than just "build" and "execute".

I have added a new set of releases to SourceForge called "4.1_with_IDE" for both Windows and Linux (see https://sourceforge.net/projects/catalina-c/files/releases/4.1_with_IDE/). As the name implies, this is exactly the same release as Catalina 4.1, but the Windows installer gives you the option of also installing the new IDE, and the Linux package also has a precompiled version included.


Some initial documentation for the new IDE is attached.

Ross.

P.S. I have no plans to drop Code::Blocks support, but it may no longer progress from the current version.

RossH

I've just uploaded a couple of small bug fixes for Catalina 4.1. If you don't use the BlackBox debugger, or use Catalina on Linux, you probably don't need to worry about it. Just copy the files included in the errata over any existing Catalina 4.1 release.

The fixes will be included in the next release.

It is available here - https://sourceforge.net/projects/catalina-c/files/errata/Catalina_4.1_Errata.zip/download

AntoineDoinel

Ross,

First, let me thank you for your work.

I've tried the precompiled catalyst sd card files. It mostly works, but I found a problem while trying to launch large binaries from the card. It seems to be affected by file size, so ls.bin cat.bin etc. works, sst.bin pcom.bin etc doesn't (they apparently just freeze, but not really... see below).

I don't think it's the binaries "per se", and neither the fs library routines, but more some kind of strange interaction between size of binary and use of the filesystem, maybe? Because I can load pcom.bin from serial, get it to compile a .pas pulled from the card into a .p5. Which in turn executes just fine if I load pcom.bin from serial and specify it.

Second test, I compiled reSound moduleplayer (using spin2gui) with an embedded .mod song of increasing size: it works when the binary is up to 64KByte or slightly above, it doesn't when it's larger than that. All those binaries are working fine when loaded serially.

Third test, which left me even more confused, I tried a somewhat large binary (more than 300KB because of beefy data padding, also made using spin2gui) which first prints a countdown, and then it tries to access the card: I can indeed see the countdown, so it only seems to freeze when it tries to access the SD card.

I'm using a P2ES board with rev. A chip, don't know if that may be the cause.

Alessandro

RossH

AntoineDoinel wrote: »

I'm using a P2ES board with rev. A chip, don't know if that may be the cause.

Odd. I have certainly had all the large Catalyst demo programs (pcom, sst, Lua etc) working on the P2 Rev A. However, it is possible the Rev A chip is the problem. Catalina no longer supports the Rev A chip "out of the box", because there are some internal instruction changes. You have to make a couple of "tweaks" - as described in the README.WhatsNew file:

The -v33 flag to p2asm was not being specified in the file Makefile.inc
    in the source\lib directory. This may have prevented some multithreading 
    functions from compiling correctly for the P2 RevB silicon. Note that
    this flag will have to be removed again from both this file, and also
    from the files p2_asm (Linux) and p2_asm.bat (Windows) in the bin
    directory to rebuild Catalina for the P2 RevA silicon. Out of the box,
    Catalina now only supports the Propeller 2 RevB (or later) silicon.

However, before you bother rebuilding Catalina, let me do some investigation. I still have a rev A board, and I will try to reproduce your problem.

Ross.

RossH

Hello @AntoineDoinel

No, it does not seem to be the -v33 flag. I just recompiled catalyst and all the demo programs with Catalina 4.1, and all the example programs work correctly on my Rev A board. I don't think they use the particular instructions that are different between Rev A and Rev B silicon.

I compiled catalyst using the following command (in the folder catalina\catalyst) ...

build_all P2_EVAL

I also tried the precompiled versions in the P2_EVAL.ZIP file in the catalina directory. Are these what you used? If not, could you try them?

What SD Card do you have? I am using a Sandisk 8Gb SD Card, formatted as FAT32.

EDIT: Also, can you tell me where you have set the switches on your board? You have to have "P59^" set to OFF to run programs from the SD Card, but ON to load programs serially.

Perhaps if you can zip up and post here one of the binaries of the program that doesn't work (if it is not one of the precompiled versions), I will see if it works on my Rev A board.

Ross.

AntoineDoinel

The card format/sector was it!

I was sure I had the card freshly formatted to FAT32, but didn't check the partition size, which was only 64MByte, being a recycled card from a Rasperry Pi. This probably meant I had small sectors.

Wiped everything and formatted the whole 8GByte, this time it asked for sector size which I set to 32KByte, and now everything works fine.

Thanks, and sorry for the false alarm.

RossH

AntoineDoinel wrote: »

The card format/sector was it!

Glad you found it - I should have remembered that the cluster size affects the size of programs Catalyst can load. This is from the Catalina_SD_Loader ...

' 
' The maximum size of programs that can be loaded by Catalyst is determined 
' by the size of the cluster list and the cluster size itself. The size of
' the cluster list is set in Catalina_Common.spin and catalyst.h (they must
' match!) for a maximum program size of 4Mb but this will only be achieved
' when using a cluster size of 32k. For other cluster sizes, see the table
' below:
' 
' Cluster Size    Max Program Size
' ============    ================
' 512 bytes        64 kbyte
'   1 kbyte       128 kbyte
'   2 kbyte       256 kbyte
'   4 kbyte       512 kbyte
'   8 kbyte         1 Mbyte
'  16 kbyte         2 Mbyte
'  32 kbyte         4 Mbyte
' 
'