cpu/sgx/encl.c

27  * reclaimer_writing_to_pcmd() - Query if any enclave page associated with
29  * @encl:        Enclave to which PCMD page belongs
30  * @start_addr:  Address of enclave page using first entry within the PCMD page
32  * When an enclave page is reclaimed some Paging Crypto MetaData (PCMD) is
33  * stored. The PCMD data of a reclaimed enclave page contains enough
35  * it is loaded back into the Enclave Page Cache (EPC).
37  * The backing storage to which enclave pages are reclaimed is laid out as
39  * Encrypted enclave pages:SECS page:PCMD pages
42  * PAGE_SIZE/sizeof(struct sgx_pcmd) enclave pages.
46  * a check if an enclave page sharing the PCMD page is in the process of being
50  * intends to reclaim that enclave page - it means that the PCMD page
51  * associated with that enclave page is about to get some data and thus
54  * Context: Enclave mutex (&sgx_encl->lock) must be held.
79 		 * started and completed with enclave mutex held so  in reclaimer_writing_to_pcmd()
105  * Calculate byte offset of a PCMD struct associated with an enclave page. PCMD's
107  * visible enclave pages, there's one extra page slot for SECS, before PCMD
152 	 * Address of enclave page using the first entry within the PCMD page.  in __sgx_encl_eldu()
240  * to protect the enclave states including SECS and ensure the SECS page is
315  * sgx_encl_eaug_page() - Dynamically add page to initialized enclave
317  * @encl:	enclave accessing the page
320  * When an initialized enclave accesses a page with no backing EPC page
345 	 * phase. The enclave decides the permissions by the means of  in sgx_encl_eaug_page()
450 	 * The page_array keeps track of all enclave pages, whether they  in sgx_vma_fault()
453 	 * a new enclave page. This is only possible for an initialized  in sgx_vma_fault()
454 	 * enclave that will be checked for right away.  in sgx_vma_fault()
506  * @encl:		an enclave pointer
511  * Iterate through the enclave pages contained within [@start, @end) to verify
514  * permissions of any of the enclave pages within the given address range.
516  * An enclave creator must declare the strongest permissions that will be
517  * needed for each enclave page. This ensures that mappings have the identical
532 	/* Disallow mapping outside enclave's address range. */  in sgx_encl_may_map()
539 	 * conflict with the enclave page permissions.  in sgx_encl_may_map()
604  * Load an enclave page to EPC if required, and take encl->lock.
694  * sgx_encl_release - Destroy an enclave instance
698  * enclave and the instance itself.
781 	 * The enclave itself can remove encl_mm.  Note, objects can't be moved  in sgx_mmu_notifier_release()
842 	 * Even though a single enclave may be mapped into an mm more than once,  in sgx_encl_mm_add()
883  * sgx_encl_cpumask() - Query which CPUs might be accessing the enclave
884  * @encl: the enclave
888  * copies a page from the enclave page cache to regular main memory but
893  * when an enclave is exited via ENCLU[EEXIT] or an Asynchronous Enclave
894  * Exit (AEX). Exiting an enclave will thus ensure cached linear-to-physical
899  * When the ENCLS[ETRACK] function is issued on an enclave the hardware
900  * tracks threads operating inside the enclave at that time. The SGX
902  * exited the enclave in order to flush the mappings before a function such
909  *    accessing the enclave.
910  * 3) Send IPI to identified CPUs, kicking them out of the enclave and
917  *          enclave with fresh linear-to-physical address mappings.
921  *          of the above flow with the enclave's mutex.
960  * @encl:	an enclave pointer
961  * @page_index:	enclave page index
965  * Crypto MetaData (PCMD) of an enclave page.
997  * in the enclave's mm_list. When not called from ksgxd, just returns
1015 	 * Search the enclave's mm_list to find an mm associated with  in sgx_encl_get_mem_cgroup()
1016 	 * this enclave to charge the allocation to.  in sgx_encl_get_mem_cgroup()
1035 	 * the enclave, set memcg to the current active mem_cgroup.  in sgx_encl_get_mem_cgroup()
1047  * @encl:	an enclave pointer
1048  * @page_index:	enclave page index
1052  * mms in the enclave's mm_list prior to any backing page allocation,
1054  * enclave.  Create a backing page for loading data back into an EPC page with
1079  * @encl:	an enclave pointer
1080  * @page_index:	enclave page index
1128  * @page:	enclave page to be tested for recent access
1130  * Checks the Access (A) bit from the PTE corresponding to the enclave page and
1191  * sgx_zap_enclave_ptes() - remove PTEs mapping the address from enclave
1192  * @encl: the enclave
1195  * Multiple VMAs may have an enclave page mapped. Remove the PTE mapping
1235  * @reclaim: Reclaim EPC pages directly if none available. Enclave
1307  * sgx_encl_free_epc_page - free an EPC page assigned to an enclave
1310  * Free an EPC page assigned to an enclave. It does EREMOVE for the page, and