z88dk forums

As a bit of background (to save background reading), this code is called from the NMI and so maskable interrupts are disabled. As far as I can tell this was fixed in UNIDOS. I do remember receiving updated system files from various people so someone might have fixed it. Unfortunately it doesn't look...

In the first writebytes, it writes 4 bytes at position 128 which should add another sector - but presumably it goes into the cache and isn't written to the file yet. This could be why when it tries to position to the end of the file that it ends up thinking 128 is the end of the file instead of 256...

1. I don't know why I did that initial write, maybe it was so that read+write files behave sensibly and that there actually is an soft EOF on disc for when the cache is bypassed? Possibly. Since all reads in CP/M are 128 byte records, the important thing is that if a sector is not readable that it ...

I think you've uncovered the behaviour of the CP/M cache that I put in a couple of years ago for you! 1. I don't know why I did that initial write, maybe it was so that read+write files behave sensibly and that there actually is an soft EOF on disc for when the cache is bypassed? 2. fflush() is as y...

Yeah, dzasm decodes for the z88 rst parameters - I was just using it as an example of a decoder.

I pushed something last night, though I've not had the time to test it so if you don't mind trying it out.

There's a zip utils application for the z88 here: https://bitbucket.org/cambridge/z88/src ... /ziputils/ which might be interesting to compare against.The main routine is here: https://bitbucket.org/cambridge/z88/src ... nflate.asm

Yes, the TI calculator is an interpreter of numeric operations, for what I could understand TI is not interested in allowing asm programs interfacing with their FP engine, but I can be wrong. Page 102 of this document: https://ia800606.us.archive.org/16/items/83psdk/sdk83pguide.pdf As another examp...

Just for the record, the continuation is handled by the preprocessor and the exact same behaviour occurs with clang.

Nice spot, open() checked for O_WRONLY|O_RDWR instead of O_CREAT

I do like the idea - unfortunately I don't think we pass enough information to the linker to do it there (length isn't available) so would have to be done on an individual compilation unit basis by the compiler. Given that the files for RAGE are auto-generated, that's probably the best place for it ...

Oh this is a nice gotcha.

Load the file into a syntax highlighting editor and you'll spot it straightaway.
Your comment line ends in a \ so the following for line becomes part of the comment. As a result the break isn't in a for loop!

That's good to hear, thanks for the success report!

Yeah, the problem was down in parsefcb - that assembler code is not exactly easy to follow! Hopefully fixed and nothing else broken now.

Sorry about that, that's a really odd behaviour that I can't completely explain. As background, some jp optimisation has been added to cut down on jumps to jumps. For _freopen1.c (and only that file), the optimisation was put into the wrong section. This meant that instead of jumping to another part...

Thanks, I've reproduced it (not on windows) and raised this issue: https://github.com/z88dk/z88dk/issues/2417 Hopefully Paulo will figure out a fix. As a workaround, uou should be able to redefine the temporary directory using the environment variable TMP which will hopefully unstick you for the mom...

Can you give me the terminal output for running zcc with -v? Would be helpful to know which tool is complaining!

I fixed up some issues earlier this year (see this post: viewtopic.php?p=21364#p21364) so if you're using the last release, could you try with a nightly build?

gcc 13.2 would suffice, however: "however no standard library is available."

Which is a bit unfortunate!

Hot on the heels of Rabbit 4000, we've added support for the Kawasaki Steel KC160 CPU. The what? This is an almost forgotten-to-time z80 clone and extension from Kawasaki back in 1994. It was discontinued in 2004 and references to it online are very elusive. However, why let that be a barrier? We fo...

We had a contribution earlier this week with some patches for building for Mint, so it was fairly easy to get the nightly build creating a kit. I've not been able to test it (for some reason the keyboard is mangled on the emulator), but it's there if anyone is interested. For the sake of history I t...

Thanks. Let me see if I understand you. You want me to download the macOS version No,I was trying to say that the z88dk-latest.tgz version on the nightly site had the libraries in it already though reading back it's not particularly clear. and then do "./build.sh -l -z" That's still the c...

No it's a good question! The libraries do take a substantial time to build The tgz and the zip files for Windows + macOS contain the already built libraries. So if you download that then ./build.sh -l -z should just build the binaries and skip over all the tests (which means the need for the perl gu...

You'll just have to overwrite the one that's in lib/coleco/classic/adam_bootstrap.asm - once you have another one I'll add a mechanism.

Following another heroic effort from Paulo, Rabbit 4000/5000 support has been added to z88dk. The 4000 was a follow up on the 3000 adding two sets of 32 bit index registers, new 16 bit registers (jk and jk') and the ability to operate on 32 bit register pairs (jkhl and bcde) as well as bunch of new ...