Can't use PythonCall in Custom System Image

**Affects:** PythonCall

**Describe the bug**
I'm trying to build a Julia system image that uses PythonCall.jl. I can build the system image, but when I try to launch Julia with it it crashes on initialisation.

As a MWE, I've set up a package [here](https://github.com/BenCurran98/CondaIssue.jl) that just uses PythonCall.jl. You can then run the following code to create a system image with this package:

```julia
using Pkg
Pkg.add(url = "https://github.com/BenCurran98/CondaIssue.jl")
Pkg.add("PackageCompiler.jl")
using PackageCompiler
create_sysimage(["CondaIssue"]; sysimage_path = "CondaSysImage.so")
```

This builds successfully, but when I run it I get the following error:
```
julia -JCondaSysImage.so --project=.
```

```julia
CondaPkg Found dependencies: /home/ben/.julia/packages/PythonCall/WMWY0/CondaPkg.toml
fatal: error thrown and no exception handler available.
InitError(mod=:C, error=MethodError(f=Base.Cmd, args=(Array{Nothing, 1}(dims=(1,), mem=Memory{Nothing}(1, 0x7183379f3380)[nothing]),), world=0x0000000000006869))
jl_method_error_bare at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/gf.c:2254
jl_method_error at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/gf.c:2272
jl_lookup_generic_ at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/gf.c:3106 [inlined]
ijl_apply_generic at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/gf.c:3121
jl_apply at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/julia.h:2157 [inlined]
do_call at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/interpreter.c:126
eval_value at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/interpreter.c:223
eval_stmt_value at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/interpreter.c:174 [inlined]
eval_body at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/interpreter.c:670
jl_fptr_interpret_call at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/interpreter.c:760
#pixi#4 at /home/ben/.julia/packages/JLLWrappers/GfYNv/src/products/executable_generators.jl:81
jl_apply at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/julia.h:2157 [inlined]
do_call at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/interpreter.c:126
eval_value at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/interpreter.c:223
eval_stmt_value at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/interpreter.c:174 [inlined]
eval_body at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/interpreter.c:670
jl_fptr_interpret_call at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/interpreter.c:760
pixi at /home/ben/.julia/packages/JLLWrappers/GfYNv/src/products/executable_generators.jl:72
jl_apply at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/julia.h:2157 [inlined]
jl_f__call_latest at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/builtins.c:875
#invokelatest#2 at ./essentials.jl:1055 [inlined]
invokelatest at ./essentials.jl:1052
unknown function (ip: 0x71833b906212)
#pixi_cmd#7 at /home/ben/.julia/packages/CondaPkg/pEObH/src/backend.jl:96
pixi_cmd at /home/ben/.julia/packages/CondaPkg/pEObH/src/backend.jl:93
#70 at /home/ben/.julia/packages/CondaPkg/pEObH/src/resolve.jl:916
cd at ./file.jl:112
unknown function (ip: 0x71833b97ccd6)
#resolve#64 at /home/ben/.julia/packages/CondaPkg/pEObH/src/resolve.jl:904
resolve at /home/ben/.julia/packages/CondaPkg/pEObH/src/resolve.jl:682
envdir at /home/ben/.julia/packages/CondaPkg/pEObH/src/env.jl:70
init_context at /home/ben/.julia/packages/PythonCall/WMWY0/src/C/context.jl:63
unknown function (ip: 0x71833b902cbf)
tojlinvoke103745 at /home/ben/Testing/Julia/CondaSysImage.so (unknown line)
j_init_context_102425 at /home/ben/Testing/Julia/CondaSysImage.so (unknown line)
__init__ at /home/ben/.julia/packages/PythonCall/WMWY0/src/C/C.jl:23
jfptr___init___102423 at /home/ben/Testing/Julia/CondaSysImage.so (unknown line)
jl_apply at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/julia.h:2157 [inlined]
jl_module_run_initializer at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/toplevel.c:76
_finish_julia_init at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/init.c:902
julia_init at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/init.c:843
jl_repl_entrypoint at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/src/jlapi.c:1053
main at /cache/build/builder-amdci5-5/julialang/julia-release-1-dot-11/cli/loader_exe.c:58
unknown function (ip: 0x71834382a1c9)
__libc_start_main at /lib/x86_64-linux-gnu/libc.so.6 (unknown line)
unknown function (ip: 0x4010b8)
```

I've seen similar issues with system images posted here, but they seem more related to the python path rather than this error which I can't identify the cause of. I would appreciate any help you can offer! Thanks :)

**Your system**
- Operating system: Ubuntu 24.04
- Julia version: 1.11.4
- PythonCall version: 0.9.24

```julia
julia> versioninfo()
Julia Version 1.11.4
Commit 8561cc3d68d (2025-03-10 11:36 UTC)
Build Info:
Official https://julialang.org/ release
Platform Info:
OS: Linux (x86_64-linux-gnu)
CPU: 28 × 13th Gen Intel(R) Core(TM) i7-13850HX
WORD_SIZE: 64
LLVM: libLLVM-16.0.6 (ORCJIT, alderlake)
Threads: 16 default, 0 interactive, 8 GC (on 28 virtual cores)

julia> Pkg.status()
Project CondaIssue v0.1.0
Status `~/Testing/Julia/CondaIssue.jl/Project.toml`
[6099a3de] PythonCall v0.9.24

julia> CondaPkg.status()
CondaPkg Status /home/ben/Testing/Julia/CondaIssue.jl/CondaPkg.toml
Environment
/home/ben/Testing/Julia/CondaIssue.jl/.CondaPkg/.pixi/envs/default
Packages
numpy v2.1.0
```

Start by reproducing the failure with the provided PackageCompiler.create_sysimage command and inspect PythonCall's src/C/context.jl together with CondaPkg's src/backend.jl around the reported stack frames. Done means the generated system image launches without the initialization crash while retaining the reported PythonCall and CondaPkg setup.

由索引模型根据 Issue 内容生成。