Crashing when using Stockfish 15 about stockfish

[johndoknjas]

There should be a '15' string instead of '/lib/x86_64-linux-gnu/libc'.

This error should explain why there's issues with SF 15 and not 14. The new SF 15 seems to use an updated glibc version, and if the user doesn't have it on their linux comp then there may be problems running the engine.

These links discuss the error, and funnily enough the answer in the second one actually mentions SF 15:
https://superuser.com/questions/537683/how-to-fix-lib-x86-64-linux-gnu-libc-so-6-version-glibc-2-14-not-found
https://stackoverflow.com/questions/71940179/error-lib-x86-64-linux-gnu-libc-so-6-version-glibc-2-34-not-found

[doodledood]

@johndoknjas Compiling Stockfish 15 directly from the source worked like a charm. Thanks!

[johndoknjas]

Hi, for the function crashing do you mean the "test_get_stockfish_major_version" function in test_models.py crashes?

[doodledood]

Haven't tried running the tests, just used the main module as it is. It crashed when trying to get the major version, which fails for Stockfish 15 (at least on my machine).

[johndoknjas]

@doodledood Hmm, alright. For the test failing it made sense since 15 for the major version wasn't included yet as one of the acceptable outputs. For the actual function though it works for me, and running SF 15 by itself seems to use the same output format as SF 14. Do you know if it works for you on SF 14?

[doodledood]

Yeah, I'm running it perfectly with stockfish 14.1

…

On Thu, 5 May 2022 at 22:01 John Doknjas ***@***.***> wrote: @doodledood <https://github.com/doodledood> Hmm, alright. For the test failing it made sense since 15 for the major version wasn't included yet as one of the acceptable outputs. For the actual function though it works for me, and running SF 15 by itself seems to use the same output format as SF 14. Do you know if it works for you on SF 14? — Reply to this email directly, view it on GitHub <#90 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAGY7DKI2TAPV2JUG67XYQLVIQLJHANCNFSM5UOPER6A> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[janrnc]

For the actual function though it works for me

Same.

It crashed when trying to get the major version, which fails for Stockfish 15 (at least on my machine).

@doodledood do you have some additional info as returned errors or similar that can help understanding where is the problem?

[doodledood]

So using the latest stockfish 15 (stockfish_15_linux_x64_avx2) this is what I'm getting when running it on Debian-based Linux: `ValueError: invalid literal for int() with base 10: '/lib/x86_64-linux-gnu/libc'` In this file: `File "/usr/local/lib/python3.9/site-packages/stockfish/models.py", line 48, in __init__ self._stockfish_major_version: int = int(...` As a side note, I get the same error for all stockfish 15 Linux versions, not just avx2

…

On Fri, May 6, 2022 at 10:50 PM Jan Franchini ***@***.***> wrote: For the actual function though it works for me Same. It crashed when trying to get the major version, which fails for Stockfish 15 (at least on my machine). @doodledood <https://github.com/doodledood> do you have some additional info as returned errors or similar that can help understanding where is the problem? — Reply to this email directly, view it on GitHub <#90 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAGY7DOAES4275Z64LCARGLVIVZW5ANCNFSM5UOPER6A> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[janrnc]

There should be a '15' string instead of '/lib/x86_64-linux-gnu/libc'. These strings are coming from the first output line of the stockfish engine (stderr gets here too since stderr=subprocess.STDOUT option is set).

Have you tried to run the engine itself from the command line? The first output line should be something like this:

Stockfish 15 by the Stockfish developers (see AUTHORS file)

Maybe you're getting an error as a first line and this wrapper tries to parse it without getting what it would expect.