Re: v2.26.0 fails to start with return code -1073741569 (0xC00000FD)

--=-8C2XN/0vXuAzlujob/CL
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Sat, 2026-04-25 at 23:41 -0500, [email protected] wrote:
> So it's always nice to have ARM binaries though Frescobaldi and
> LilyPond have been working under emulation nicely up until now.

Just a quick update, native Windows on ARM binaries for LilyPond might
be a bit tricky because of support in the GCC toolchain:
https://linaro.atlassian.net/wiki/spaces/WOAR/pages/28802842658/MinGW+GNU+T=
oolchain
Clang is working, but would need to investigate how much that would
change...

> I tried those variables you mentioned.  First, I set GUILE_JIT_LOG=3D1
> and there was a ton of diagnostic output (attached) before it crashed
> as before.  So then I tried also setting GUILE_JIT_THRESHOLD=3D-1 and
> retested -- this time there was none of that lengthy diagnostic
> output... but the result was the same

Thanks for testing, I think this rules out the JIT as the (immediate)
problem. When it crashes, do you get a popup that would allow you to
get a stack trace / crash report? These might just be hexadecimal
numbers (0x...) without much sense, but I should have the necessary
build information to turn this back into something useful. Then at
least we would know in which part of the code it crashes...

Jonas

--=-8C2XN/0vXuAzlujob/CL
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----

iHUEABYKAB0WIQRiF4DLRvon1vw5biCcVOmY/6OEDwUCae3C7AAKCRCcVOmY/6OE
D/G2AQCSlPiXWQqVnzQ+a2XrLE0iBVNSOS8X/eZDyMplpD8Z5wD+NAvfOmipKylo
EKqgF5+DKfGTJJMLNrpxsOUEa/glzgo=
=33ph
-----END PGP SIGNATURE-----

--=-8C2XN/0vXuAzlujob/CL--