Re: [PATCH v14 14/14] selftests/sgx: Add scripts for EPC cgroup testing

["Jarkko Sakkinen" <jarkko@kernel.org>]

On Sat Jun 1, 2024 at 1:26 AM EEST, Haitao Huang wrote:
> With different cgroups, the script starts one or multiple concurrent SGX
> selftests (test_sgx), each to run the unclobbered_vdso_oversubscribed
> test case, which loads an enclave of EPC size equal to the EPC capacity
> available on the platform. The script checks results against the
> expectation set for each cgroup and reports success or failure.
>
> The script creates 3 different cgroups at the beginning with following
> expectations:
>
> 1) small - intentionally small enough to fail the test loading an
> enclave of size equal to the capacity.
> 2) large - large enough to run up to 4 concurrent tests but fail some if
> more than 4 concurrent tests are run. The script starts 4 expecting at
> least one test to pass, and then starts 5 expecting at least one test
> to fail.
> 3) larger - limit is the same as the capacity, large enough to run lots of
> concurrent tests. The script starts 8 of them and expects all pass.
> Then it reruns the same test with one process randomly killed and
> usage checked to be zero after all processes exit.
>
> The script also includes a test with low mem_cg limit and large sgx_epc
> limit to verify that the RAM used for per-cgroup reclamation is charged
> to a proper mem_cg. For this test, it turns off swapping before start,
> and turns swapping back on afterwards.
>
> Add README to document how to run the tests.
>
> Signed-off-by: Haitao Huang <haitao.huang@linux.intel.com>
> Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
> Tested-by: Jarkko Sakkinen <jarkko@kernel.org>

Reorg:

void sgx_cgroup_init(void)
{
	struct workqueue_struct *wq;

	/* eagerly allocate the workqueue: */
	wq = alloc_workqueue("sgx_cg_wq", wq_unbound | wq_freezable, wq_unbound_max_active);
	if (!wq) {
		pr_warn("sgx_cg_wq creation failed\n");
		return;
	}

	misc_cg_set_ops(MISC_CG_RES_SGX_EPC, &sgx_cgroup_ops);
	sgx_cgroup_misc_init(misc_cg_root(), &sgx_cg_root);

	/* Depending on misc state, keep or destory the workqueue: */
	if (cgroup_subsys_enabled(misc_cgrp_subsys))
		sgx_cg_wq = wq;
	else
		destroy_workqueue(wq);
}

BTW, why two previous operations are performed if subsystem is not
enabled?

I.e. why not instead:

void sgx_cgroup_init(void)
{
	struct workqueue_struct *wq;

	/* Eagerly allocate the workqueue: */
	wq = alloc_workqueue("sgx_cg_wq", wq_unbound | wq_freezable, wq_unbound_max_active);
	if (!wq) {
		pr_warn("sgx_cg_wq creation failed\n");
		return;
	}

	if (!cgroup_subsys_enabled(misc_cgrp_subsys)) {
		destroy_workqueue(wq);
		return;
	}

	misc_cg_set_ops(MISC_CG_RES_SGX_EPC, &sgx_cgroup_ops);
	sgx_cgroup_misc_init(misc_cg_root(), &sgx_cg_root);
	sgx_cg_wq = wq;
}

Finally, why this does not have __init? And neither sgx_cgroup_misc_init().

The names for these are also somewhat confusing, maybe something like:

* __sgx_cgroups_misc_init()
* sgx_cgroups_misc_init()

And both with __init.

I just made a trivial checkpatch run as a final check, and spotted the
warning on BUG_ON(), and noticed that this can't be right as it is but
please comment and correct where I might have gotten something wrong.

With "--strict" flag I also catched these:

CHECK: spinlock_t definition without comment
#1308: FILE: arch/x86/kernel/cpu/sgx/sgx.h:122:
+	spinlock_t lock;

CHECK: multiple assignments should be avoided
#444: FILE: kernel/cgroup/misc.c:450:
+		parent_cg = cg = &root_cg;

BR, Jarkko