segmentation fault when trying to use an array return type for a pool_processor callback

[ttytm]

Describe the bug

import sync.pool

pub struct Foo {
	page string
}

mut pp := pool.new_pool_processor(
	// array return types cause `SIGSEGV`
	callback: fn (mut pp pool.PoolProcessor, idx int, wid int) []Foo {
		page := pp.get_item[int](idx)
		return [Foo{page.str()}]
	}
)

mut pp_ := pool.new_pool_processor(
	// non-array return types work.
	callback: fn (mut pp pool.PoolProcessor, idx int, wid int) &Foo {
		page := pp.get_item[int](idx)
		return &Foo{page.str()}
	}
)

pages := []int{len: 2, init: index}

// Comment out block and uncomment below for working version.
pp.work_on_items(pages)
res := pp.get_results[[]Foo]().map(it)
dump(res)

/* pp_.work_on_items(pages)
res := pp_.get_results[Foo]().map(it)
dump(res) */

Reproduction Steps

above

Expected Behavior

works, or error.

Current Behavior

segfault

Possible Solution

No response

Additional Information/Context

Workaround is to use another struct with an array field:

import sync.pool

pub struct Foo {
	page string
}

struct Workaround {
	foo_arr []Foo
}

mut pp := pool.new_pool_processor(
	callback: fn (mut pp pool.PoolProcessor, idx int, wid int) &Workaround {
		page := pp.get_item[int](idx)
		res := &Workaround{
			foo_arr: [Foo{page.str()}]
		}
		return res
	}
)

pages := []int{len: 2, init: index}
pp.work_on_items(pages)
res := pp.get_results[[]Foo]().map(it)
dump(res)

V version

v0.4.6

Environment details (OS name and version, etc.)

linux amd64

Note

You can use the 👍 reaction to increase the issue's priority for developers.

Please note that only the 👍 reaction to the issue itself counts as a vote.
Other reactions and those to comments will not be taken into account.