beneroth changed the topic of #picolisp to: PicoLisp language | The scalpel of software development | Channel Log: https://libera.irclog.whitequark.org/picolisp | Check www.picolisp.com for more information
ygrek has joined #picolisp
ygrek has quit [Remote host closed the connection]
<abu[7]> tankf33der: I posted a comment to Github with a code example demonstrating the bug
<tankf33der> morning all
<tankf33der> thanks.
<abu[7]> o/
<tankf33der> installed adelie linux ppc64 on docker and llvm 18.1.8 have the same issue. ppc64 is big endian
<abu[7]> ok
<abu[7]> Stupid Github removed indentation and spaces from my code, so the example is rather unreadable :(
<tankf33der> yeah.
<abu[7]> needs some markup probably
<abu[7]> <pre> ?
<tankf33der> ```
<tankf33der> code
<tankf33der> ```
<abu[7]> ok
<tankf33der> this is md, iirc
<abu[7]> I try
<tankf33der> try edit small part
<abu[7]> Cool! Much better now
<abu[7]> bbl
ygrek has joined #picolisp
<tankf33der> abu[7]: I am sure you will see the same issues in the code on s390x as soon as you take a look, and it will immediately become clear where how it fails.
<abu[7]> Yeah, I expect that too, but I don't know s390x assembly
<abu[7]> It shows the same symptoms like PPC, right?
<abu[7]> So perhaps even the same bug
<tankf33der> I don't know much and don't understand any assembler, but I can send something if you're interested.
<bjorkintosh> tankf33der: aren't you running your ppc in VM anyway?
<tankf33der> docker+qemu
<abu[7]> Not currently
<abu[7]> is s390x also ppc ?
<tankf33der> ppc64le - little
<tankf33der> ppc64 - big
<abu[7]> ah
<tankf33der> s390x - big
<abu[7]> I thought it is a different CPU
<tankf33der> different
<abu[7]> Only the endianess does not make much differece in asm
<abu[7]> Let's wait if the github issue results in some fix
<abu[7]> Perhaps s390x is fixed then too
<tankf33der> very optimistic.
<tankf33der> slackers
<tankf33der> they are slackers
<abu[7]> hmm
<abu[7]> But *they* must fix the backend, we can't
<abu[7]> I think for Pil ppc64le vs. ppc64 does not matter. Should be agnostic about endianess.
<tankf33der> let me check statistics again
<abu[7]> Pil64 needed to know about endianess
<abu[7]> but pil21 not
<tankf33der> backend:Powerpc tag - 192 open, 936 closed
<abu[7]> not bad ;)
<tankf33der> not bad
<tankf33der> a lot of pull request merged
<abu[7]> Perhaps it is not directly ppc, but LLVM levels above it
<abu[7]> afk
corecheckno has quit [Remote host closed the connection]
ygrek has quit [Remote host closed the connection]
corecheckno has joined #picolisp
corecheckno has quit [Client Quit]
<abu[7]>