#asahi · 2021-05-07 12:34
marshmallowright :-)#asahi · 2021-05-07 12:22
marshmallowright, so secure enclave is even better than EL3 since it runs in a completely separated processor#asahi · 2021-05-07 12:13
marshmallow(considered that it involves external security coprocessor stuff)#asahi · 2021-05-07 12:12
marshmallowsven: but isn't the secure enclave used as sort of replacement for the secure world?#asahi · 2021-05-06 12:34
marshmallowso basically Secure Enclave replaces EL3 TrustZone in M1?#asahi · 2021-05-06 12:34
marshmallowsven: very nice writeup!#asahi · 2021-02-25 12:41
marshmallowif a function pointer in the kernel points to userspace code, and the area is marked as executable-only, then PXN fits in. if the area is marked as writable-only, that is prohibited 'by default'.#asahi · 2021-02-25 12:39
marshmallowk, so the PXN bit works independently of the protection attributes of the EL0 translation tables, correct?#asahi · 2021-02-25 12:33
marshmallowwell, PXN is useful only in kernel mode#asahi · 2021-02-25 12:01
marshmallowif PXN is architecturally enforced, why does even PXN exist?#asahi · 2021-02-09 13:54
marshmallowdavidrysk[m]: blackbird?