[tip: x86/urgent] x86/sgx: Silence softlockup detection when releasing large enclaves

[tip-bot2 for Reinette Chatre]

The following commit has been merged into the x86/urgent branch of tip:

Commit-ID:     8795359e35bc33bf86b6d0765aa7f37431db3b9c
Gitweb:        https://git.kernel.org/tip/8795359e35bc33bf86b6d0765aa7f37431db3b9c
Author:        Reinette Chatre <reinette.chatre@xxxxxxxxx>
AuthorDate:    Tue, 08 Feb 2022 10:48:07 -08:00
Committer:     Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
CommitterDate: Thu, 10 Feb 2022 15:58:14 -08:00

x86/sgx: Silence softlockup detection when releasing large enclaves

Vijay reported that the "unclobbered_vdso_oversubscribed" selftest
triggers the softlockup detector.

Actual SGX systems have 128GB of enclave memory or more.  The
"unclobbered_vdso_oversubscribed" selftest creates one enclave which
consumes all of the enclave memory on the system. Tearing down such a
large enclave takes around a minute, most of it in the loop where
the EREMOVE instruction is applied to each individual 4k enclave page.

Spending one minute in a loop triggers the softlockup detector.

Add a cond_resched() to give other tasks a chance to run and placate
the softlockup detector.

Cc: stable@xxxxxxxxxxxxxxx
Fixes: 1728ab54b4be ("x86/sgx: Add a page reclaimer")
Reported-by: Vijay Dhanraj <vijay.dhanraj@xxxxxxxxx>
Signed-off-by: Reinette Chatre <reinette.chatre@xxxxxxxxx>
Signed-off-by: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
Reviewed-by: Jarkko Sakkinen <jarkko@xxxxxxxxxx>
Acked-by: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
Tested-by: Jarkko Sakkinen <jarkko@xxxxxxxxxx>  (kselftest as sanity check)
Link: https://lkml.kernel.org/r/ced01cac1e75f900251b0a4ae1150aa8ebd295ec.1644345232.git.reinette.chatre@xxxxxxxxx
---
 arch/x86/kernel/cpu/sgx/encl.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/arch/x86/kernel/cpu/sgx/encl.c b/arch/x86/kernel/cpu/sgx/encl.c
index 001808e..48afe96 100644
--- a/arch/x86/kernel/cpu/sgx/encl.c
+++ b/arch/x86/kernel/cpu/sgx/encl.c
@@ -410,6 +410,8 @@ void sgx_encl_release(struct kref *ref)
 		}
 
 		kfree(entry);
+		/* Invoke scheduler to prevent soft lockups. */
+		cond_resched();
 	}
 
 	xa_destroy(&encl->page_array);