From patchwork Tue Mar 8 11:28:33 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jarkko Sakkinen X-Patchwork-Id: 12773546 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by smtp.lore.kernel.org (Postfix) with ESMTP id 32D80C433EF for ; Tue, 8 Mar 2022 11:29:25 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id B63328D0006; Tue, 8 Mar 2022 06:29:24 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id AEA988D0001; Tue, 8 Mar 2022 06:29:24 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 9652B8D0006; Tue, 8 Mar 2022 06:29:24 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0073.hostedemail.com [216.40.44.73]) by kanga.kvack.org (Postfix) with ESMTP id 871A88D0001 for ; Tue, 8 Mar 2022 06:29:24 -0500 (EST) Received: from smtpin30.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay05.hostedemail.com (Postfix) with ESMTP id 3AF04180FF4C0 for ; Tue, 8 Mar 2022 11:29:24 +0000 (UTC) X-FDA: 79220998248.30.34BAA7B Received: from dfw.source.kernel.org (dfw.source.kernel.org [139.178.84.217]) by imf15.hostedemail.com (Postfix) with ESMTP id B49B2A0004 for ; Tue, 8 Mar 2022 11:29:23 +0000 (UTC) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 20CDC61661; Tue, 8 Mar 2022 11:29:23 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 928A8C340EC; Tue, 8 Mar 2022 11:29:22 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1646738962; bh=Cgl0qA66PzKkXnlxny3eTwo2XWjpLBhZQNxtn6qkKoU=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=VubbAgcJ5C1kTfkMRe5DOVPs3CVCfg5RBHj6uVzV5thQwyrl70oV606uxiFAsxiea YBDSE79E1vqFWdsRxSTTPj5n6v3BmmsbgXT5uccdWArP4L34OGL6S6Gmmg1V7xOPkt byS9Zoc7bJ9Wj4lrQuwQ4XfqdCKK1ERpX0cB3UVTfGxz7k16Xm5gEtDDC2JIdQNt8W BCz70rcPSZ0Kwf/Nk7mm+kOeNmWViJhxHwV5fNYK5V0AHESKLyztFG8Uidv3uZWyn7 Nl8DsM47fYJdtBjIh2G/TtktzsxvgSyYyya+rZrc7STHiBlTYT4PAHkawU6Z0L7+wM JnU6eOwA4GSnQ== From: Jarkko Sakkinen To: linux-mm@kvack.org, linux-fsdevel@vger.kernel.org Cc: Dave Hansen , Nathaniel McCallum , Reinette Chatre , Alexander Viro , linux-sgx@vger.kernel.org, linux-kernel@vger.kernel.org, Andrew Morton , Jarkko Sakkinen Subject: [PATCH RFC v3 3/3] x86/sgx: Implement EAUG population with MAP_POPULATE Date: Tue, 8 Mar 2022 13:28:33 +0200 Message-Id: <20220308112833.262805-4-jarkko@kernel.org> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220308112833.262805-1-jarkko@kernel.org> References: <20220308112833.262805-1-jarkko@kernel.org> MIME-Version: 1.0 X-Rspamd-Queue-Id: B49B2A0004 X-Stat-Signature: b98oxzhudihyebgi9y15n5tf9ewgphss X-Rspam-User: Authentication-Results: imf15.hostedemail.com; dkim=pass header.d=kernel.org header.s=k20201202 header.b=VubbAgcJ; spf=pass (imf15.hostedemail.com: domain of jarkko@kernel.org designates 139.178.84.217 as permitted sender) smtp.mailfrom=jarkko@kernel.org; dmarc=pass (policy=none) header.from=kernel.org X-Rspamd-Server: rspam07 X-HE-Tag: 1646738963-754736 X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: With SGX1 an enclave needs to be created with its maximum memory demands pre-allocated. Pages cannot be added to an enclave after it is initialized. SGX2 introduces a new function, ENCLS[EAUG] for adding pages to an initialized enclave. Add support for dynamically adding pages to an initialized enclave with mmap() by populating pages with EAUG. Use f_ops->populate() callback to achieve this behaviour. Signed-off-by: Jarkko Sakkinen --- arch/x86/kernel/cpu/sgx/driver.c | 128 +++++++++++++++++++++++++++++++ 1 file changed, 128 insertions(+) diff --git a/arch/x86/kernel/cpu/sgx/driver.c b/arch/x86/kernel/cpu/sgx/driver.c index aa9b8b868867..848938334e8a 100644 --- a/arch/x86/kernel/cpu/sgx/driver.c +++ b/arch/x86/kernel/cpu/sgx/driver.c @@ -9,6 +9,7 @@ #include #include "driver.h" #include "encl.h" +#include "encls.h" u64 sgx_attributes_reserved_mask; u64 sgx_xfrm_reserved_mask = ~0x3; @@ -101,6 +102,132 @@ static int sgx_mmap(struct file *file, struct vm_area_struct *vma) return 0; } +static int sgx_encl_augment_page(struct sgx_encl *encl, unsigned long offset) +{ + struct sgx_pageinfo pginfo = {0}; + struct sgx_encl_page *encl_page; + struct sgx_epc_page *epc_page; + struct sgx_va_page *va_page; + u64 secinfo_flags; + int ret; + + /* + * Ignore internal permission checking for dynamically added pages. + * They matter only for data added during the pre-initialization phase. + * The enclave decides the permissions by the means of EACCEPT, + * EACCEPTCOPY and EMODPE. + */ + secinfo_flags = SGX_SECINFO_R | SGX_SECINFO_W | SGX_SECINFO_X; + encl_page = sgx_encl_page_alloc(encl, offset, secinfo_flags); + if (IS_ERR(encl_page)) + return PTR_ERR(encl_page); + + epc_page = sgx_alloc_epc_page(encl_page, true); + if (IS_ERR(epc_page)) { + ret = PTR_ERR(epc_page); + goto err_alloc_epc_page; + } + + va_page = sgx_encl_grow(encl); + if (IS_ERR(va_page)) { + ret = PTR_ERR(va_page); + goto err_grow; + } + + mutex_lock(&encl->lock); + + /* + * Adding to encl->va_pages must be done under encl->lock. Ditto for + * deleting (via sgx_encl_shrink()) in the error path. + */ + if (va_page) + list_add(&va_page->list, &encl->va_pages); + + /* + * Insert prior to EADD in case of OOM. EADD modifies MRENCLAVE, i.e. + * can't be gracefully unwound, while failure on EADD/EXTEND is limited + * to userspace errors (or kernel/hardware bugs). + */ + ret = xa_insert(&encl->page_array, PFN_DOWN(encl_page->desc), + encl_page, GFP_KERNEL); + + /* + * If ret == -EBUSY then page was created in another flow while + * running without encl->lock + */ + if (ret) + goto err_xa_insert; + + pginfo.secs = (unsigned long)sgx_get_epc_virt_addr(encl->secs.epc_page); + pginfo.addr = encl_page->desc & PAGE_MASK; + pginfo.metadata = 0; + + ret = __eaug(&pginfo, sgx_get_epc_virt_addr(epc_page)); + if (ret) + goto err_eaug; + + encl_page->encl = encl; + encl_page->epc_page = epc_page; + encl_page->type = SGX_PAGE_TYPE_REG; + encl->secs_child_cnt++; + + sgx_mark_page_reclaimable(encl_page->epc_page); + + mutex_unlock(&encl->lock); + + return 0; + +err_eaug: + xa_erase(&encl->page_array, PFN_DOWN(encl_page->desc)); + +err_xa_insert: + sgx_encl_shrink(encl, va_page); + mutex_unlock(&encl->lock); + +err_grow: + sgx_encl_free_epc_page(epc_page); + +err_alloc_epc_page: + kfree(encl_page); + + return VM_FAULT_SIGBUS; +} + +/* + * Add new pages to the enclave sequentially with ENCLS[EAUG]. Note that + * sgx_mmap() validates that the given VMA is within the enclave range. Calling + * here sgx_encl_may_map() second time would too time consuming. + */ +static int sgx_populate(struct file *file, unsigned long start, unsigned long end) +{ + struct sgx_encl *encl = file->private_data; + unsigned long length = end - start; + unsigned long pos; + int ret; + + /* EAUG works only for initialized enclaves. */ + if (!test_bit(SGX_ENCL_INITIALIZED, &encl->flags)) + return -EINVAL; + + for (pos = 0 ; pos < length; pos += PAGE_SIZE) { + if (signal_pending(current)) { + if (!pos) + ret = -ERESTARTSYS; + + break; + } + + if (need_resched()) + cond_resched(); + + ret = sgx_encl_augment_page(encl, start + pos); + if (ret) + break; + } + + return ret; +} + static unsigned long sgx_get_unmapped_area(struct file *file, unsigned long addr, unsigned long len, @@ -133,6 +260,7 @@ static const struct file_operations sgx_encl_fops = { .compat_ioctl = sgx_compat_ioctl, #endif .mmap = sgx_mmap, + .populate = sgx_populate, .get_unmapped_area = sgx_get_unmapped_area, };