SimpleSpanProcessor.Shutdown panics when constructed with a nil exporter

[pellared]

Description

NewSimpleSpanProcessor(nil) returns a non-nil processor, but calling Shutdown on that processor panics.

The constructor stores the supplied exporter directly, and the existing TestNewSimpleSpanProcessorWithNilExporter explicitly verifies that a processor can be constructed with a nil exporter. In Shutdown, the exporter is captured and cleared under the mutex, then exp.Shutdown(ctx) is called in a goroutine without checking whether exp is nil.

The resulting panic terminates the host process during telemetry shutdown.

Environment

• OS: Linux
• Architecture: amd64
• Go Version: go1.26.4
• opentelemetry-go version: 0de413a318cb52629baefb89a1554a905e105aa3

Steps To Reproduce

Add this regression test to sdk/trace/simple_span_processor_test.go:

func TestSimpleSpanProcessorShutdownWithNilExporter(t *testing.T) {
	ssp := NewSimpleSpanProcessor(nil)
	require.NoError(t, ssp.Shutdown(t.Context()))
}

Then run:

cd sdk
go test ./trace -run '^TestSimpleSpanProcessorShutdownWithNilExporter$' -count=1

On current main, this panics in the shutdown goroutine at simple_span_processor.go:92 with a nil pointer dereference.

Expected behavior

Shutting down a SimpleSpanProcessor constructed with a nil exporter should not panic. Given the constructor's existing tested behavior, Shutdown should treat the absent exporter as a no-op and return nil.

_{Tip: React with 👍 to help prioritize this issue. Please use comments to provide useful context, avoiding +1 or me too, to help us triage it. Learn more here.}

[ShreeyashPatil2708]

I would like to work on this, will open a PR shortly.