Corrected check for in-kernel driver

[ScottR-Intel]

Description of the changes

The in-kernel driver by itself doesn't create the /dev/sgx folder any longer. This is done by the PSW installers now. So, if the PSW isn't yet installed when running the tool, it will fail the in-kernel driver check. Modifying the tool to check for /dev/sgx_provision /dev/sgx_enclave which is what the in-kernel driver creates.

---

This change is 

[woju]

Reviewed 1 of 1 files at r1.
Reviewable status: all files reviewed, 2 unresolved discussions, not enough approvals from maintainers (3 more required), not enough approvals from different teams (2 more required, approved so far: ) (waiting on @ScottR-Intel)

---

Pal/src/host/Linux-SGX/tools/is-sgx-available/is_sgx_available.cpp, line 169 at r1 (raw file):

bool sgx_driver_loaded() {
    // /dev/isgx is for LKM version, /dev/sgx_enclave and /dev/sgx_provision are for in-kernel support.
    return file_exists("/dev/isgx") || (file_exists("/dev/sgx_enclave") && file_exists("/dev/sgx_provision"));

This will return false on DCAP < 1.10. Do we care about that?

---

Pal/src/host/Linux-SGX/tools/is-sgx-available/is_sgx_available.cpp, line 169 at r1 (raw file):

bool sgx_driver_loaded() {
    // /dev/isgx is for LKM version, /dev/sgx_enclave and /dev/sgx_provision are for in-kernel support.
    return file_exists("/dev/isgx") || (file_exists("/dev/sgx_enclave") && file_exists("/dev/sgx_provision"));

We don't need to check for sgx_provision.

[dimakuv]

Reviewed 1 of 1 files at r1.
Reviewable status: all files reviewed, 2 unresolved discussions, not enough approvals from maintainers (3 more required), not enough approvals from different teams (2 more required, approved so far: ) (waiting on @woju)

---

Pal/src/host/Linux-SGX/tools/is-sgx-available/is_sgx_available.cpp, line 169 at r1 (raw file):

Previously, woju (Wojtek Porczyk) wrote…

This will return false on DCAP < 1.10. Do we care about that?

+1 to woju's request. Why not keep the older /dev/sgx as a third option here?

---

Pal/src/host/Linux-SGX/tools/is-sgx-available/is_sgx_available.cpp, line 169 at r1 (raw file):

Previously, woju (Wojtek Porczyk) wrote…

We don't need to check for sgx_provision.

It doesn't hurt to check for it, it's a part of the SGX driver installation. I would keep it.

[mkow]

Reviewed 1 of 1 files at r1.
Reviewable status: all files reviewed, 2 unresolved discussions, not enough approvals from maintainers (3 more required), not enough approvals from different teams (2 more required, approved so far: ) (waiting on @woju)

[woju]

Reviewable status: all files reviewed, 2 unresolved discussions, not enough approvals from maintainers (3 more required), not enough approvals from different teams (2 more required, approved so far: ) (waiting on @dimakuv and @woju)

---

Pal/src/host/Linux-SGX/tools/is-sgx-available/is_sgx_available.cpp, line 169 at r1 (raw file):

Previously, dimakuv (Dmitrii Kuvaiskii) wrote…

It doesn't hurt to check for it, it's a part of the SGX driver installation. I would keep it.

IDK, we don't use it, right? But no strong opinion, unblocking.

[dimakuv]

Reviewable status: all files reviewed, 2 unresolved discussions, not enough approvals from maintainers (3 more required), not enough approvals from different teams (2 more required, approved so far: ) (waiting on @woju)

---

Pal/src/host/Linux-SGX/tools/is-sgx-available/is_sgx_available.cpp, line 169 at r1 (raw file):

Previously, woju (Wojtek Porczyk) wrote…

IDK, we don't use it, right? But no strong opinion, unblocking.

Graphene doesn't use it. But the tool in question (is_sgx_available) is unrelated to Graphene -- it just so happened that it is built and shipped with Graphene (because one of the ITL people wrote this handy tool, and we just have it now).

[mkow]

Reviewable status: all files reviewed, 3 unresolved discussions, not enough approvals from maintainers (2 more required), not enough approvals from different teams (2 more required, approved so far: ) (waiting on @ScottR-Intel and @woju)

a discussion (no related file):
Whoops, I forgot about this PR and made my own fix for this issue (#2438). Sorry @ScottR-Intel, I didn't intend to take over this PR :)
Anyways, I prefer my version, because it doesn't remove the checks for /dev/sgx, which AFAIK still has users. But if you want to introduce a similar change here, then we can keep your PR and I'll close mine. If you want to do so, please also reformat the comment like I did in #2438.

---

[dimakuv]

Reviewable status: all files reviewed, 3 unresolved discussions, not enough approvals from maintainers (2 more required), not enough approvals from different teams (2 more required, approved so far: ) (waiting on @ScottR-Intel and @woju)

a discussion (no related file):

Previously, mkow (Michał Kowalczyk) wrote…

Whoops, I forgot about this PR and made my own fix for this issue (#2438). Sorry @ScottR-Intel, I didn't intend to take over this PR :)
Anyways, I prefer my version, because it doesn't remove the checks for /dev/sgx, which AFAIK still has users. But if you want to introduce a similar change here, then we can keep your PR and I'll close mine. If you want to do so, please also reformat the comment like I did in #2438.

I suggest to go with Michal's #2438.

---