On Mon, Sep 16, 2019 at 01:17:59PM +0300, Jarkko Sakkinen wrote: > Make sgx_reclaim_evict() idempotennt and rename it to sgx_can_reclaim() and > set SGX_ENCL_PAGE_RECLAIMED in the call site. > > Cc: Sean Christopherson <sean.j.christopherson@xxxxxxxxx> > Cc: Shay Katz-zamir <shay.katz-zamir@xxxxxxxxx> > Cc: Serge Ayoun <serge.ayoun@xxxxxxxxx> > Signed-off-by: Jarkko Sakkinen <jarkko.sakkinen@xxxxxxxxxxxxxxx> > --- > arch/x86/kernel/cpu/sgx/reclaim.c | 32 +++++++++++++++++-------------- > 1 file changed, 18 insertions(+), 14 deletions(-) > > diff --git a/arch/x86/kernel/cpu/sgx/reclaim.c b/arch/x86/kernel/cpu/sgx/reclaim.c > index 5a4d44dd02d7..cc3155b61513 100644 > --- a/arch/x86/kernel/cpu/sgx/reclaim.c > +++ b/arch/x86/kernel/cpu/sgx/reclaim.c > @@ -121,7 +121,19 @@ void sgx_mark_page_reclaimable(struct sgx_epc_page *page) > spin_unlock(&sgx_active_page_list_lock); > } > > -static bool sgx_reclaimer_evict(struct sgx_epc_page *epc_page) > +/** > + * sgx_can_reclaim() - Filter out the pages that should not be reclaimed > + * @epc_page: a candidate EPC page > + * > + * Do not reclaim this page if it has been recently accessed by any mm_struct > + * *and* if the enclave is still alive. No need to take the enclave's lock, > + * worst case scenario reclaiming pages from a dead enclave is delayed slightly. > + * A live enclave with a recently accessed page is more common and avoiding lock > + * contention in that case is a boon to performance. > + * > + * Return: true if the page should be reclaimed > + */ > +static bool sgx_can_reclaim(struct sgx_epc_page *epc_page) Belated thought. The param can be 'struct sgx_encl_page *encl_page'. The caller already has the encl_page, and the function only uses the epc_page to retrieve the encl_page.
