From patchwork Fri Feb 24 03:31:43 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jitendra Kolhe X-Patchwork-Id: 9589257 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id 27FBB60578 for ; Fri, 24 Feb 2017 03:32:42 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 10CB22876A for ; Fri, 24 Feb 2017 03:32:42 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 056B52877A; Fri, 24 Feb 2017 03:32:42 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00,RCVD_IN_DNSWL_HI autolearn=ham version=3.3.1 Received: from lists.gnu.org (lists.gnu.org [208.118.235.17]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (No client certificate requested) by mail.wl.linuxfoundation.org (Postfix) with ESMTPS id 1EBD72876A for ; Fri, 24 Feb 2017 03:32:41 +0000 (UTC) Received: from localhost ([::1]:34438 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ch6ca-0007hG-92 for patchwork-qemu-devel@patchwork.kernel.org; Thu, 23 Feb 2017 22:32:40 -0500 Received: from eggs.gnu.org ([2001:4830:134:3::10]:41334) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ch6Zt-0005DU-1u for qemu-devel@nongnu.org; Thu, 23 Feb 2017 22:29:54 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1ch6Zp-0000YQ-0n for qemu-devel@nongnu.org; Thu, 23 Feb 2017 22:29:53 -0500 Received: from g2t2353.austin.hpe.com ([15.233.44.26]:36776) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1ch6Zo-0000YG-QD for qemu-devel@nongnu.org; Thu, 23 Feb 2017 22:29:48 -0500 Received: from hpvmrhel1.in.rdlabs.hpecorp.net (unknown [15.213.178.32]) by g2t2353.austin.hpe.com (Postfix) with ESMTP id 224FE54; Fri, 24 Feb 2017 03:29:40 +0000 (UTC) From: Jitendra Kolhe To: qemu-devel@nongnu.org Date: Fri, 24 Feb 2017 09:01:43 +0530 Message-Id: <1487907103-32350-1-git-send-email-jitendra.kolhe@hpe.com> X-Mailer: git-send-email 1.8.3.1 X-detected-operating-system: by eggs.gnu.org: GNU/Linux 3.x [fuzzy] X-Received-From: 15.233.44.26 Subject: [Qemu-devel] [PATCH v4] mem-prealloc: reduce large guest start-up and migration time. X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.21 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: kwolf@redhat.com, peter.maydell@linaro.org, ehabkost@redhat.com, crosthwaite.peter@gmail.com, sw@weilnetz.de, quintela@redhat.com, pl@kamp.de, armbru@redhat.com, dgilbert@redhat.com, renganathan.meenakshisundaram@hpe.com, mohan_parthasarathy@hpe.com, pbonzini@redhat.com, imammedo@redhat.com, jitendra.kolhe@hpe.com Errors-To: qemu-devel-bounces+patchwork-qemu-devel=patchwork.kernel.org@nongnu.org Sender: "Qemu-devel" X-Virus-Scanned: ClamAV using ClamSMTP Using "-mem-prealloc" option for a large guest leads to higher guest start-up and migration time. This is because with "-mem-prealloc" option qemu tries to map every guest page (create address translations), and make sure the pages are available during runtime. virsh/libvirt by default, seems to use "-mem-prealloc" option in case the guest is configured to use huge pages. The patch tries to map all guest pages simultaneously by spawning multiple threads. Currently limiting the change to QEMU library functions on POSIX compliant host only, as we are not sure if the problem exists on win32. Below are some stats with "-mem-prealloc" option for guest configured to use huge pages. ------------------------------------------------------------------------ Idle Guest | Start-up time | Migration time ------------------------------------------------------------------------ Guest stats with 2M HugePage usage - single threaded (existing code) ------------------------------------------------------------------------ 64 Core - 4TB | 54m11.796s | 75m43.843s 64 Core - 1TB | 8m56.576s | 14m29.049s 64 Core - 256GB | 2m11.245s | 3m26.598s ------------------------------------------------------------------------ Guest stats with 2M HugePage usage - map guest pages using 8 threads ------------------------------------------------------------------------ 64 Core - 4TB | 5m1.027s | 34m10.565s 64 Core - 1TB | 1m10.366s | 8m28.188s 64 Core - 256GB | 0m19.040s | 2m10.148s ----------------------------------------------------------------------- Guest stats with 2M HugePage usage - map guest pages using 16 threads ----------------------------------------------------------------------- 64 Core - 4TB | 1m58.970s | 31m43.400s 64 Core - 1TB | 0m39.885s | 7m55.289s 64 Core - 256GB | 0m11.960s | 2m0.135s ----------------------------------------------------------------------- Changed in v2: - modify number of memset threads spawned to min(smp_cpus, 16). - removed 64GB memory restriction for spawning memset threads. Changed in v3: - limit number of threads spawned based on min(sysconf(_SC_NPROCESSORS_ONLN), 16, smp_cpus) - implement memset thread specific siglongjmp in SIGBUS signal_handler. Changed in v4 - remove sigsetjmp/siglongjmp and SIGBUS unblock/block for main thread as main thread no longer touches any pages. - simplify code my returning memset_thread_failed status from touch_all_pages. Signed-off-by: Jitendra Kolhe --- backends/hostmem.c | 4 +- exec.c | 2 +- include/qemu/osdep.h | 3 +- util/oslib-posix.c | 108 +++++++++++++++++++++++++++++++++++++++++---------- util/oslib-win32.c | 3 +- 5 files changed, 94 insertions(+), 26 deletions(-) diff --git a/backends/hostmem.c b/backends/hostmem.c index 7f5de70..162c218 100644 --- a/backends/hostmem.c +++ b/backends/hostmem.c @@ -224,7 +224,7 @@ static void host_memory_backend_set_prealloc(Object *obj, bool value, void *ptr = memory_region_get_ram_ptr(&backend->mr); uint64_t sz = memory_region_size(&backend->mr); - os_mem_prealloc(fd, ptr, sz, &local_err); + os_mem_prealloc(fd, ptr, sz, smp_cpus, &local_err); if (local_err) { error_propagate(errp, local_err); return; @@ -328,7 +328,7 @@ host_memory_backend_memory_complete(UserCreatable *uc, Error **errp) */ if (backend->prealloc) { os_mem_prealloc(memory_region_get_fd(&backend->mr), ptr, sz, - &local_err); + smp_cpus, &local_err); if (local_err) { goto out; } diff --git a/exec.c b/exec.c index 8b9ed73..53afcd2 100644 --- a/exec.c +++ b/exec.c @@ -1379,7 +1379,7 @@ static void *file_ram_alloc(RAMBlock *block, } if (mem_prealloc) { - os_mem_prealloc(fd, area, memory, errp); + os_mem_prealloc(fd, area, memory, smp_cpus, errp); if (errp && *errp) { goto error; } diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h index 56c9e22..fb1d22b 100644 --- a/include/qemu/osdep.h +++ b/include/qemu/osdep.h @@ -401,7 +401,8 @@ unsigned long qemu_getauxval(unsigned long type); void qemu_set_tty_echo(int fd, bool echo); -void os_mem_prealloc(int fd, char *area, size_t sz, Error **errp); +void os_mem_prealloc(int fd, char *area, size_t sz, int smp_cpus, + Error **errp); int qemu_read_password(char *buf, int buf_size); diff --git a/util/oslib-posix.c b/util/oslib-posix.c index f631464..7e87c87 100644 --- a/util/oslib-posix.c +++ b/util/oslib-posix.c @@ -55,6 +55,21 @@ #include "qemu/error-report.h" #endif +#define MAX_MEM_PREALLOC_THREAD_COUNT (MIN(sysconf(_SC_NPROCESSORS_ONLN), 16)) + +struct MemsetThread { + char *addr; + uint64_t numpages; + uint64_t hpagesize; + QemuThread pgthread; + sigjmp_buf env; +}; +typedef struct MemsetThread MemsetThread; + +static MemsetThread *memset_thread; +static int memset_num_threads; +static bool memset_thread_failed; + int qemu_get_thread_id(void) { #if defined(__linux__) @@ -316,18 +331,83 @@ char *qemu_get_exec_dir(void) return g_strdup(exec_dir); } -static sigjmp_buf sigjump; - static void sigbus_handler(int signal) { - siglongjmp(sigjump, 1); + int i; + if (memset_thread) { + for (i = 0; i < memset_num_threads; i++) { + if (qemu_thread_is_self(&memset_thread[i].pgthread)) { + siglongjmp(memset_thread[i].env, 1); + } + } + } +} + +static void *do_touch_pages(void *arg) +{ + MemsetThread *memset_args = (MemsetThread *)arg; + char *addr = memset_args->addr; + uint64_t numpages = memset_args->numpages; + uint64_t hpagesize = memset_args->hpagesize; + sigset_t set, oldset; + int i = 0; + + /* unblock SIGBUS */ + sigemptyset(&set); + sigaddset(&set, SIGBUS); + pthread_sigmask(SIG_UNBLOCK, &set, &oldset); + + if (sigsetjmp(memset_args->env, 1)) { + memset_thread_failed = true; + } else { + for (i = 0; i < numpages; i++) { + memset(addr, 0, 1); + addr += hpagesize; + } + } + pthread_sigmask(SIG_SETMASK, &oldset, NULL); + return NULL; +} + +static bool touch_all_pages(char *area, size_t hpagesize, size_t numpages, + int smp_cpus) +{ + uint64_t numpages_per_thread, size_per_thread; + char *addr = area; + int i = 0; + + memset_thread_failed = false; + memset_num_threads = MIN(smp_cpus, MAX_MEM_PREALLOC_THREAD_COUNT); + memset_thread = g_new0(MemsetThread, memset_num_threads); + numpages_per_thread = (numpages / memset_num_threads); + size_per_thread = (hpagesize * numpages_per_thread); + for (i = 0; i < memset_num_threads; i++) { + memset_thread[i].addr = addr; + memset_thread[i].numpages = (i == (memset_num_threads - 1)) ? + numpages : numpages_per_thread; + memset_thread[i].hpagesize = hpagesize; + qemu_thread_create(&memset_thread[i].pgthread, "touch_pages", + do_touch_pages, &memset_thread[i], + QEMU_THREAD_JOINABLE); + addr += size_per_thread; + numpages -= numpages_per_thread; + } + for (i = 0; i < memset_num_threads; i++) { + qemu_thread_join(&memset_thread[i].pgthread); + } + g_free(memset_thread); + memset_thread = NULL; + + return memset_thread_failed; } -void os_mem_prealloc(int fd, char *area, size_t memory, Error **errp) +void os_mem_prealloc(int fd, char *area, size_t memory, int smp_cpus, + Error **errp) { int ret; struct sigaction act, oldact; - sigset_t set, oldset; + size_t hpagesize = qemu_fd_getpagesize(fd); + size_t numpages = DIV_ROUND_UP(memory, hpagesize); memset(&act, 0, sizeof(act)); act.sa_handler = &sigbus_handler; @@ -340,23 +420,10 @@ void os_mem_prealloc(int fd, char *area, size_t memory, Error **errp) return; } - /* unblock SIGBUS */ - sigemptyset(&set); - sigaddset(&set, SIGBUS); - pthread_sigmask(SIG_UNBLOCK, &set, &oldset); - - if (sigsetjmp(sigjump, 1)) { + /* touch pages simultaneously */ + if (touch_all_pages(area, hpagesize, numpages, smp_cpus)) { error_setg(errp, "os_mem_prealloc: Insufficient free host memory " "pages available to allocate guest RAM\n"); - } else { - int i; - size_t hpagesize = qemu_fd_getpagesize(fd); - size_t numpages = DIV_ROUND_UP(memory, hpagesize); - - /* MAP_POPULATE silently ignores failures */ - for (i = 0; i < numpages; i++) { - memset(area + (hpagesize * i), 0, 1); - } } ret = sigaction(SIGBUS, &oldact, NULL); @@ -365,7 +432,6 @@ void os_mem_prealloc(int fd, char *area, size_t memory, Error **errp) perror("os_mem_prealloc: failed to reinstall signal handler"); exit(1); } - pthread_sigmask(SIG_SETMASK, &oldset, NULL); } diff --git a/util/oslib-win32.c b/util/oslib-win32.c index 0b1890f..80e4668 100644 --- a/util/oslib-win32.c +++ b/util/oslib-win32.c @@ -541,7 +541,8 @@ int getpagesize(void) return system_info.dwPageSize; } -void os_mem_prealloc(int fd, char *area, size_t memory, Error **errp) +void os_mem_prealloc(int fd, char *area, size_t memory, int smp_cpus, + Error **errp) { int i; size_t pagesize = getpagesize();