Fix shutdown ordering in TestDnssd.

[bzbarsky-apple]

For init it does, in order: PlatformMgr().InitChipStack(), ChipDnssdInit, RunEventLoop.

For shutdown it should do them in exactly the opposite order.

Fixes #26023

[github-actions]

PR #26024: Size comparison from 8676d06 to 48bc4e9

Decreases (1 build for cc32xx)

platform	target	config	section	8676d06	48bc4e9	change	% change
cc32xx	lock	CC3235SF_LAUNCHXL	.debug_info	20302991	20302990	-1	-0.0

Full report (1 build for cc32xx)

platform	target	config	section	8676d06	48bc4e9	change	% change
cc32xx	lock	CC3235SF_LAUNCHXL		0	0	0	0.0
			(read only)	642745	642745	0	0.0
			(read/write)	203848	203848	0	0.0
			.ARM.attributes	44	44	0	0.0
			.ARM.exidx	8	8	0	0.0
			.bss	197248	197248	0	0.0
			.comment	194	194	0	0.0
			.data	1480	1480	0	0.0
			.debug_abbrev	933119	933119	0	0.0
			.debug_aranges	87608	87608	0	0.0
			.debug_frame	301328	301328	0	0.0
			.debug_info	20302991	20302990	-1	-0.0
			.debug_line	2679759	2679759	0	0.0
			.debug_loc	2824679	2824679	0	0.0
			.debug_ranges	286200	286200	0	0.0
			.debug_str	3039406	3039406	0	0.0
			.ramVecs	780	780	0	0.0
			.resetVecs	64	64	0	0.0
			.rodata	104289	104289	0	0.0
			.shstrtab	232	232	0	0.0
			.stab	204	204	0	0.0
			.stabstr	441	441	0	0.0
			.stack	2048	2048	0	0.0
			.strtab	377422	377422	0	0.0
			.symtab	256768	256768	0	0.0
			.text	536336	536336	0	0.0

[arkq]

If you've got some time, please also take a look at the exit(0) in the HandleResolve() function. When fixing this issue: #25466 I had to comment out lines 42-44, otherwise, in some test runs Shutdown() was called twice which has caused "strange" errors due to use after free/double free issue. If I remember correctly the exit(0) triggers handler registered by atexit() which also does something with platform shutdown. However, I've got no time to dive dipper into that.

[bzbarsky-apple]

If I remember correctly the exit(0) triggers handler registered by atexit() which also does something with platform shutdown.

There's no atexit handler like that in the unit tests.

[arkq]

There's no atexit handler like that in the unit tests.

Yes, it's not in the unit test. I've got some troubles to determine where the handler was registered, but I do remember that it was "atexit() > ... > Shutdown() > ...." in the callstack...