From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from lindbergh.monkeyblade.net ([23.128.96.19]:54264 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S240873AbhDMMA0 (ORCPT ); Tue, 13 Apr 2021 08:00:26 -0400 Received: from desiato.infradead.org (desiato.infradead.org [IPv6:2001:8b0:10b:1:d65d:64ff:fe57:4e05]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id BF653C061574 for ; Tue, 13 Apr 2021 05:00:06 -0700 (PDT) Received: from [65.144.74.35] (helo=kernel.dk) by desiato.infradead.org with esmtpsa (Exim 4.94 #2 (Red Hat Linux)) id 1lWHi7-0099ew-RX for fio@vger.kernel.org; Tue, 13 Apr 2021 12:00:04 +0000 Subject: Recent changes (master) From: Jens Axboe Message-Id: <20210413120001.A6EFA1BC0143@kernel.dk> Date: Tue, 13 Apr 2021 06:00:01 -0600 (MDT) List-Id: fio@vger.kernel.org To: fio@vger.kernel.org The following changes since commit 1588c8f571f67a004571e51cdbb5de97c3e4f457: Merge branch 'wip-rados-dont-zerowrite' of https://github.com/aclamk/fio (2021-04-10 11:46:30 -0600) are available in the Git repository at: git://git.kernel.dk/fio.git master for you to fetch changes up to 9b6253bc6af3b38d4677f7470f42a1ff22492ef3: t/zbd: test repeated async write with block size unaligned to zone size (2021-04-12 06:56:29 -0600) ---------------------------------------------------------------- Shin'ichiro Kawasaki (2): zbd: avoid zone reset during asynchronous IOs in-flight t/zbd: test repeated async write with block size unaligned to zone size t/zbd/test-zbd-support | 18 ++++++++++++++++++ zbd.c | 23 +++++++---------------- 2 files changed, 25 insertions(+), 16 deletions(-) --- Diff of recent changes: diff --git a/t/zbd/test-zbd-support b/t/zbd/test-zbd-support index be129615..26aff373 100755 --- a/t/zbd/test-zbd-support +++ b/t/zbd/test-zbd-support @@ -1201,6 +1201,24 @@ test56() { >> "${logfile}.${test_number}" 2>&1 || return $? } +# Test that repeated async write job does not cause zone reset during writes +# in-flight, when the block size is not a divisor of the zone size. +test57() { + local bs off + + require_zbd || return $SKIP_TESTCASE + + bs=$((4096 * 7)) + off=$((first_sequential_zone_sector * 512)) + + run_fio --name=job --filename="${dev}" --rw=randwrite --bs="${bs}" \ + --offset="${off}" --size=$((4 * zone_size)) --iodepth=256 \ + "$(ioengine "libaio")" --time_based=1 --runtime=30s \ + --zonemode=zbd --direct=1 --zonesize="${zone_size}" \ + ${job_var_opts[@]} \ + >> "${logfile}.${test_number}" 2>&1 || return $? +} + SECONDS=0 tests=() dynamic_analyzer=() diff --git a/zbd.c b/zbd.c index d16b890f..eed796b3 100644 --- a/zbd.c +++ b/zbd.c @@ -842,16 +842,13 @@ static void zbd_close_zone(struct thread_data *td, const struct fio_file *f, * @f: fio file for which to reset zones * @zb: first zone to reset. * @ze: first zone not to reset. - * @all_zones: whether to reset all zones or only those zones for which the - * write pointer is not a multiple of td->o.min_bs[DDIR_WRITE]. */ static int zbd_reset_zones(struct thread_data *td, struct fio_file *f, struct fio_zone_info *const zb, - struct fio_zone_info *const ze, bool all_zones) + struct fio_zone_info *const ze) { struct fio_zone_info *z; const uint32_t min_bs = td->o.min_bs[DDIR_WRITE]; - bool reset_wp; int res = 0; assert(min_bs); @@ -864,16 +861,10 @@ static int zbd_reset_zones(struct thread_data *td, struct fio_file *f, if (!z->has_wp) continue; zone_lock(td, f, z); - if (all_zones) { - pthread_mutex_lock(&f->zbd_info->mutex); - zbd_close_zone(td, f, nz); - pthread_mutex_unlock(&f->zbd_info->mutex); - - reset_wp = z->wp != z->start; - } else { - reset_wp = z->wp % min_bs != 0; - } - if (reset_wp) { + pthread_mutex_lock(&f->zbd_info->mutex); + zbd_close_zone(td, f, nz); + pthread_mutex_unlock(&f->zbd_info->mutex); + if (z->wp != z->start) { dprint(FD_ZBD, "%s: resetting zone %u\n", f->file_name, zbd_zone_nr(f, z)); if (zbd_reset_zone(td, f, z) < 0) @@ -996,8 +987,8 @@ void zbd_file_reset(struct thread_data *td, struct fio_file *f) * writing any data to avoid that a zone reset has to be issued while * writing data, which causes data loss. */ - zbd_reset_zones(td, f, zb, ze, td->o.verify != VERIFY_NONE && - td->runstate != TD_VERIFYING); + if (td->o.verify != VERIFY_NONE && td->runstate != TD_VERIFYING) + zbd_reset_zones(td, f, zb, ze); zbd_reset_write_cnt(td, f); }