From patchwork Mon Dec 16 21:35:23 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ben Gardon X-Patchwork-Id: 11295329 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 1DA7C6C1 for ; Mon, 16 Dec 2019 21:35:49 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id EFACD24685 for ; Mon, 16 Dec 2019 21:35:48 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="cLMpQGkx" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727549AbfLPVfs (ORCPT ); Mon, 16 Dec 2019 16:35:48 -0500 Received: from mail-pg1-f201.google.com ([209.85.215.201]:40384 "EHLO mail-pg1-f201.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726610AbfLPVfs (ORCPT ); Mon, 16 Dec 2019 16:35:48 -0500 Received: by mail-pg1-f201.google.com with SMTP id z12so5962816pgf.7 for ; Mon, 16 Dec 2019 13:35:46 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20161025; h=date:message-id:mime-version:subject:from:to:cc; bh=flMQx0RQ+tn1L0OZWy96vlecDFYT8pwNVqw/lok2bBY=; b=cLMpQGkx99eeWLH4P/unEkS15oJ0X5gTpKr0Ot5VrGyjOzB4dA0ifIETD5xPnL915T rubBQ1Z1iW4qP8ZPAFTNHDSpa1zHhugbJ2QxmSBGJ8YFQSbzxHjfKNHTdoja8vQT2IGs ZJ9gn+5y5Ern0CrQyWiOXApsJgUmCFVLDPffnLDVVzBJuidyMrKqQVRQnBM8tnVfAagh afrkhSTDsBVhP4jgntZuieMwkliL0ZvcoC6Mlo8Oj52LTnAulg5T+H8eQl5xK6xNjN2+ jtpvhLNqUrXAu9q2trjCcLueT55qoIgDlBNl9WNltONfRe97/+D2FG+d2akzZiwxjJPq fs/A== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:date:message-id:mime-version:subject:from:to:cc; bh=flMQx0RQ+tn1L0OZWy96vlecDFYT8pwNVqw/lok2bBY=; b=AOsLSKOeM6k9qBiEZqf4VXs3u/tbb/nYt+7Aito/t67OyHlv2isEhOToJ3Kw7KH7zJ b3bTjQO8tiCNnmr7D6abV0Ubi4KJmFcVPXukm+27ZlCqQBXGQn55gHhnIyhdtIcPMPE2 NNU90m/+XtsWmvICKmEC3GFBX0mqR73IuzyXDxfa/2hm/l9umlltABiVmIUapq0qSSa8 0tdaNylzoCsUvd0au1RGMeS8DrKc27gysPQuiOeXsiLPhDYzG9T4zuZmXU5DKL6fmnzs FJXv2zMWidHz9QKetIGrYXiqgbjtmlIKu654iDYpyHuJ0wye2kzaaTg9K963znCPpjQv 8A3A== X-Gm-Message-State: APjAAAWBif9qrFORDSc4LNKQpi97Z6qKEfVEZX0oh34Dijo/2P//vkUZ SJk2654BiRDOYwJh4ilKByTIhvs8A42b X-Google-Smtp-Source: APXvYqy1p58uD7ILOXRlPnATDG1C2VyIQV8EfLRJI2x+ZgT/v8x5ksiTotwGpJiUF6/nvEUtv7a0zUt3NmVO X-Received: by 2002:a63:b642:: with SMTP id v2mr4217832pgt.126.1576532145784; Mon, 16 Dec 2019 13:35:45 -0800 (PST) Date: Mon, 16 Dec 2019 13:35:23 -0800 Message-Id: <20191216213532.91237-1-bgardon@google.com> Mime-Version: 1.0 X-Mailer: git-send-email 2.24.1.735.g03f4e72817-goog Subject: [PATCH 0/9] Create a userfaultfd demand paging test From: Ben Gardon To: linux-kernel@vger.kernel.org, kvm@vger.kernel.org, linux-kselftest@vger.kernel.org Cc: Paolo Bonzini , Cannon Matthews , Peter Xu , Andrew Jones , Ben Gardon Sender: linux-kselftest-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kselftest@vger.kernel.org When handling page faults for many vCPUs during demand paging, KVM's MMU lock becomes highly contended. This series creates a test with a naive userfaultfd based demand paging implementation to demonstrate that contention. This test serves both as a functional test of userfaultfd and a microbenchmark of demand paging performance with a variable number of vCPUs and memory per vCPU. The test creates N userfaultfd threads, N vCPUs, and a region of memory with M pages per vCPU. The N userfaultfd polling threads are each set up to serve faults on a region of memory corresponding to one of the vCPUs. Each of the vCPUs is then started, and touches each page of its disjoint memory region, sequentially. In response to faults, the userfaultfd threads copy a static buffer into the guest's memory. This creates a worst case for MMU lock contention as we have removed most of the contention between the userfaultfd threads and there is no time required to fetch the contents of guest memory. This test was run successfully on Intel Haswell, Broadwell, and Cascadelake hosts with a variety of vCPU counts and memory sizes. This test was adapted from the dirty_log_test. The series can also be viewed in Gerrit here: https://linux-review.googlesource.com/c/virt/kvm/kvm/+/1464 (Thanks to Dmitry Vyukov for setting up the Gerrit instance) Ben Gardon (9): KVM: selftests: Create a demand paging test KVM: selftests: Add demand paging content to the demand paging test KVM: selftests: Add memory size parameter to the demand paging test KVM: selftests: Pass args to vCPU instead of using globals KVM: selftests: Support multiple vCPUs in demand paging test KVM: selftests: Time guest demand paging KVM: selftests: Add parameter to _vm_create for memslot 0 base paddr KVM: selftests: Support large VMs in demand paging test Add static flag tools/testing/selftests/kvm/.gitignore | 1 + tools/testing/selftests/kvm/Makefile | 4 +- .../selftests/kvm/demand_paging_test.c | 610 ++++++++++++++++++ tools/testing/selftests/kvm/dirty_log_test.c | 2 +- .../testing/selftests/kvm/include/kvm_util.h | 3 +- tools/testing/selftests/kvm/lib/kvm_util.c | 7 +- 6 files changed, 621 insertions(+), 6 deletions(-) create mode 100644 tools/testing/selftests/kvm/demand_paging_test.c