From patchwork Tue Sep 19 16:57:57 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrey Drobyshev X-Patchwork-Id: 13391610 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 lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 46B91CE79A8 for ; Tue, 19 Sep 2023 17:00:17 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qie3H-0003iw-OF; Tue, 19 Sep 2023 12:58:19 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie3B-0003eH-CY; Tue, 19 Sep 2023 12:58:13 -0400 Received: from relay.virtuozzo.com ([130.117.225.111]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie37-0002Mb-Ev; Tue, 19 Sep 2023 12:58:13 -0400 Received: from [130.117.225.1] (helo=dev005.ch-qa.vzint.dev) by relay.virtuozzo.com with esmtp (Exim 4.96) (envelope-from ) id 1qidzZ-00DUte-2m; Tue, 19 Sep 2023 18:57:54 +0200 To: qemu-block@nongnu.org Cc: qemu-devel@nongnu.org, hreitz@redhat.com, kwolf@redhat.com, eblake@redhat.com, andrey.drobyshev@virtuozzo.com, den@virtuozzo.com Subject: [PATCH v3 1/8] qemu-img: rebase: stop when reaching EOF of old backing file Date: Tue, 19 Sep 2023 19:57:57 +0300 Message-Id: <20230919165804.439110-2-andrey.drobyshev@virtuozzo.com> X-Mailer: git-send-email 2.39.3 In-Reply-To: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> References: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> MIME-Version: 1.0 Received-SPF: pass client-ip=130.117.225.111; envelope-from=andrey.drobyshev@virtuozzo.com; helo=relay.virtuozzo.com X-Spam_score_int: -18 X-Spam_score: -1.9 X-Spam_bar: - X-Spam_report: (-1.9 / 5.0 requ) BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-to: Andrey Drobyshev X-Patchwork-Original-From: Andrey Drobyshev via From: Andrey Drobyshev Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org In case when we're rebasing within one backing chain, and when target image is larger than old backing file, bdrv_is_allocated_above() ends up setting *pnum = 0. As a result, target offset isn't getting incremented, and we get stuck in an infinite for loop. Let's detect this case and proceed further down the loop body, as the offsets beyond the old backing size need to be explicitly zeroed. Signed-off-by: Andrey Drobyshev Reviewed-by: Denis V. Lunev Reviewed-by: Hanna Czenczek --- qemu-img.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/qemu-img.c b/qemu-img.c index a48edb7101..50660ba920 100644 --- a/qemu-img.c +++ b/qemu-img.c @@ -3805,6 +3805,8 @@ static int img_rebase(int argc, char **argv) } if (prefix_chain_bs) { + uint64_t bytes = n; + /* * If cluster wasn't changed since prefix_chain, we don't need * to take action @@ -3817,9 +3819,18 @@ static int img_rebase(int argc, char **argv) strerror(-ret)); goto out; } - if (!ret) { + if (!ret && n) { continue; } + if (!n) { + /* + * If we've reached EOF of the old backing, it means that + * offsets beyond the old backing size were read as zeroes. + * Now we will need to explicitly zero the cluster in + * order to preserve that state after the rebase. + */ + n = bytes; + } } /* From patchwork Tue Sep 19 16:57:58 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrey Drobyshev X-Patchwork-Id: 13391606 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 lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 1D21FCE79AC for ; Tue, 19 Sep 2023 16:59:29 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qie3G-0003iI-I3; Tue, 19 Sep 2023 12:58:18 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie3B-0003eL-GR; Tue, 19 Sep 2023 12:58:13 -0400 Received: from relay.virtuozzo.com ([130.117.225.111]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie37-0002Mi-EL; Tue, 19 Sep 2023 12:58:13 -0400 Received: from [130.117.225.1] (helo=dev005.ch-qa.vzint.dev) by relay.virtuozzo.com with esmtp (Exim 4.96) (envelope-from ) id 1qidzZ-00DUte-2p; Tue, 19 Sep 2023 18:57:54 +0200 To: qemu-block@nongnu.org Cc: qemu-devel@nongnu.org, hreitz@redhat.com, kwolf@redhat.com, eblake@redhat.com, andrey.drobyshev@virtuozzo.com, den@virtuozzo.com Subject: [PATCH v3 2/8] qemu-iotests: 024: add rebasing test case for overlay_size > backing_size Date: Tue, 19 Sep 2023 19:57:58 +0300 Message-Id: <20230919165804.439110-3-andrey.drobyshev@virtuozzo.com> X-Mailer: git-send-email 2.39.3 In-Reply-To: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> References: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> MIME-Version: 1.0 Received-SPF: pass client-ip=130.117.225.111; envelope-from=andrey.drobyshev@virtuozzo.com; helo=relay.virtuozzo.com X-Spam_score_int: -18 X-Spam_score: -1.9 X-Spam_bar: - X-Spam_report: (-1.9 / 5.0 requ) BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-to: Andrey Drobyshev X-Patchwork-Original-From: Andrey Drobyshev via From: Andrey Drobyshev Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Before previous commit, rebase was getting infitely stuck in case of rebasing within the same backing chain and when overlay_size > backing_size. Let's add this case to the rebasing test 024 to make sure it doesn't break again. Signed-off-by: Andrey Drobyshev Reviewed-by: Denis V. Lunev Reviewed-by: Hanna Czenczek --- tests/qemu-iotests/024 | 57 ++++++++++++++++++++++++++++++++++++++ tests/qemu-iotests/024.out | 30 ++++++++++++++++++++ 2 files changed, 87 insertions(+) diff --git a/tests/qemu-iotests/024 b/tests/qemu-iotests/024 index 25a564a150..98a7c8fd65 100755 --- a/tests/qemu-iotests/024 +++ b/tests/qemu-iotests/024 @@ -199,6 +199,63 @@ echo # $BASE_OLD and $BASE_NEW) $QEMU_IMG map "$OVERLAY" | _filter_qemu_img_map +# Check that rebase within the chain is working when +# overlay_size > old_backing_size +# +# base_new <-- base_old <-- overlay +# +# Backing (new): 11 11 11 11 11 +# Backing (old): 22 22 22 22 +# Overlay: -- -- -- -- -- +# +# As a result, overlay should contain data identical to base_old, with the +# last cluster remaining unallocated. + +echo +echo "=== Test rebase within one backing chain ===" +echo + +echo "Creating backing chain" +echo + +TEST_IMG=$BASE_NEW _make_test_img $(( CLUSTER_SIZE * 5 )) +TEST_IMG=$BASE_OLD _make_test_img -b "$BASE_NEW" -F $IMGFMT \ + $(( CLUSTER_SIZE * 4 )) +TEST_IMG=$OVERLAY _make_test_img -b "$BASE_OLD" -F $IMGFMT \ + $(( CLUSTER_SIZE * 5 )) + +echo +echo "Fill backing files with data" +echo + +$QEMU_IO "$BASE_NEW" -c "write -P 0x11 0 $(( CLUSTER_SIZE * 5 ))" \ + | _filter_qemu_io +$QEMU_IO "$BASE_OLD" -c "write -P 0x22 0 $(( CLUSTER_SIZE * 4 ))" \ + | _filter_qemu_io + +echo +echo "Check the last cluster is zeroed in overlay before the rebase" +echo +$QEMU_IO "$OVERLAY" -c "read -P 0x00 $(( CLUSTER_SIZE * 4 )) $CLUSTER_SIZE" \ + | _filter_qemu_io + +echo +echo "Rebase onto another image in the same chain" +echo + +$QEMU_IMG rebase -b "$BASE_NEW" -F $IMGFMT "$OVERLAY" + +echo "Verify that data is read the same before and after rebase" +echo + +# Verify the first 4 clusters are still read the same as in the old base +$QEMU_IO "$OVERLAY" -c "read -P 0x22 0 $(( CLUSTER_SIZE * 4 ))" \ + | _filter_qemu_io +# Verify the last cluster still reads as zeroes +$QEMU_IO "$OVERLAY" -c "read -P 0x00 $(( CLUSTER_SIZE * 4 )) $CLUSTER_SIZE" \ + | _filter_qemu_io + +echo # success, all done echo "*** done" diff --git a/tests/qemu-iotests/024.out b/tests/qemu-iotests/024.out index 973a5a3711..245fe8b1d1 100644 --- a/tests/qemu-iotests/024.out +++ b/tests/qemu-iotests/024.out @@ -171,4 +171,34 @@ read 65536/65536 bytes at offset 196608 Offset Length File 0 0x30000 TEST_DIR/subdir/t.IMGFMT 0x30000 0x10000 TEST_DIR/subdir/t.IMGFMT.base_new + +=== Test rebase within one backing chain === + +Creating backing chain + +Formatting 'TEST_DIR/subdir/t.IMGFMT.base_new', fmt=IMGFMT size=327680 +Formatting 'TEST_DIR/subdir/t.IMGFMT.base_old', fmt=IMGFMT size=262144 backing_file=TEST_DIR/subdir/t.IMGFMT.base_new backing_fmt=IMGFMT +Formatting 'TEST_DIR/subdir/t.IMGFMT', fmt=IMGFMT size=327680 backing_file=TEST_DIR/subdir/t.IMGFMT.base_old backing_fmt=IMGFMT + +Fill backing files with data + +wrote 327680/327680 bytes at offset 0 +320 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 262144/262144 bytes at offset 0 +256 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +Check the last cluster is zeroed in overlay before the rebase + +read 65536/65536 bytes at offset 262144 +64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +Rebase onto another image in the same chain + +Verify that data is read the same before and after rebase + +read 262144/262144 bytes at offset 0 +256 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 65536/65536 bytes at offset 262144 +64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + *** done From patchwork Tue Sep 19 16:57:59 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrey Drobyshev X-Patchwork-Id: 13391607 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 lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 1AD96CE79AB for ; Tue, 19 Sep 2023 16:59:29 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qie3E-0003hn-Mo; Tue, 19 Sep 2023 12:58:17 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie3B-0003eM-Gf; Tue, 19 Sep 2023 12:58:13 -0400 Received: from relay.virtuozzo.com ([130.117.225.111]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie37-0002Md-G3; Tue, 19 Sep 2023 12:58:13 -0400 Received: from [130.117.225.1] (helo=dev005.ch-qa.vzint.dev) by relay.virtuozzo.com with esmtp (Exim 4.96) (envelope-from ) id 1qidzZ-00DUte-2z; Tue, 19 Sep 2023 18:57:54 +0200 To: qemu-block@nongnu.org Cc: qemu-devel@nongnu.org, hreitz@redhat.com, kwolf@redhat.com, eblake@redhat.com, andrey.drobyshev@virtuozzo.com, den@virtuozzo.com Subject: [PATCH v3 3/8] qemu-img: rebase: use backing files' BlockBackend for buffer alignment Date: Tue, 19 Sep 2023 19:57:59 +0300 Message-Id: <20230919165804.439110-4-andrey.drobyshev@virtuozzo.com> X-Mailer: git-send-email 2.39.3 In-Reply-To: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> References: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> MIME-Version: 1.0 Received-SPF: pass client-ip=130.117.225.111; envelope-from=andrey.drobyshev@virtuozzo.com; helo=relay.virtuozzo.com X-Spam_score_int: -18 X-Spam_score: -1.9 X-Spam_bar: - X-Spam_report: (-1.9 / 5.0 requ) BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-to: Andrey Drobyshev X-Patchwork-Original-From: Andrey Drobyshev via From: Andrey Drobyshev Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Since commit bb1c05973cf ("qemu-img: Use qemu_blockalign"), buffers for the data read from the old and new backing files are aligned using BlockDriverState (or BlockBackend later on) referring to the target image. However, this isn't quite right, because buf_new is only being used for reading from the new backing, while buf_old is being used for both reading from the old backing and writing to the target. Let's take that into account and use more appropriate values as alignments. Signed-off-by: Andrey Drobyshev --- qemu-img.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/qemu-img.c b/qemu-img.c index 50660ba920..4dc91505bf 100644 --- a/qemu-img.c +++ b/qemu-img.c @@ -3750,8 +3750,13 @@ static int img_rebase(int argc, char **argv) int64_t n; float local_progress = 0; - buf_old = blk_blockalign(blk, IO_BUF_SIZE); - buf_new = blk_blockalign(blk, IO_BUF_SIZE); + if (blk_old_backing && bdrv_opt_mem_align(blk_bs(blk_old_backing)) > + bdrv_opt_mem_align(blk_bs(blk))) { + buf_old = blk_blockalign(blk_old_backing, IO_BUF_SIZE); + } else { + buf_old = blk_blockalign(blk, IO_BUF_SIZE); + } + buf_new = blk_blockalign(blk_new_backing, IO_BUF_SIZE); size = blk_getlength(blk); if (size < 0) { From patchwork Tue Sep 19 16:58:00 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrey Drobyshev X-Patchwork-Id: 13391613 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 lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id AA274CE79A9 for ; Tue, 19 Sep 2023 17:00:32 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qie3H-0003iu-Ks; Tue, 19 Sep 2023 12:58:19 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie3C-0003go-MQ; Tue, 19 Sep 2023 12:58:14 -0400 Received: from relay.virtuozzo.com ([130.117.225.111]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie37-0002Mg-Fq; Tue, 19 Sep 2023 12:58:14 -0400 Received: from [130.117.225.1] (helo=dev005.ch-qa.vzint.dev) by relay.virtuozzo.com with esmtp (Exim 4.96) (envelope-from ) id 1qidzZ-00DUte-39; Tue, 19 Sep 2023 18:57:54 +0200 To: qemu-block@nongnu.org Cc: qemu-devel@nongnu.org, hreitz@redhat.com, kwolf@redhat.com, eblake@redhat.com, andrey.drobyshev@virtuozzo.com, den@virtuozzo.com Subject: [PATCH v3 4/8] qemu-img: add chunk size parameter to compare_buffers() Date: Tue, 19 Sep 2023 19:58:00 +0300 Message-Id: <20230919165804.439110-5-andrey.drobyshev@virtuozzo.com> X-Mailer: git-send-email 2.39.3 In-Reply-To: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> References: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> MIME-Version: 1.0 Received-SPF: pass client-ip=130.117.225.111; envelope-from=andrey.drobyshev@virtuozzo.com; helo=relay.virtuozzo.com X-Spam_score_int: -18 X-Spam_score: -1.9 X-Spam_bar: - X-Spam_report: (-1.9 / 5.0 requ) BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-to: Andrey Drobyshev X-Patchwork-Original-From: Andrey Drobyshev via From: Andrey Drobyshev Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Add @chsize param to the function which, if non-zero, would represent the chunk size to be used for comparison. If it's zero, then BDRV_SECTOR_SIZE is used as default chunk size, which is the previous behaviour. In particular, we're going to use this param in img_rebase() to make the write requests aligned to a predefined alignment value. Signed-off-by: Andrey Drobyshev Reviewed-by: Eric Blake Reviewed-by: Hanna Czenczek --- qemu-img.c | 24 +++++++++++++++--------- 1 file changed, 15 insertions(+), 9 deletions(-) diff --git a/qemu-img.c b/qemu-img.c index 4dc91505bf..0f67b021f7 100644 --- a/qemu-img.c +++ b/qemu-img.c @@ -1274,23 +1274,29 @@ static int is_allocated_sectors_min(const uint8_t *buf, int n, int *pnum, } /* - * Compares two buffers sector by sector. Returns 0 if the first - * sector of each buffer matches, non-zero otherwise. + * Compares two buffers chunk by chunk, where @chsize is the chunk size. + * If @chsize is 0, default chunk size of BDRV_SECTOR_SIZE is used. + * Returns 0 if the first chunk of each buffer matches, non-zero otherwise. * - * pnum is set to the sector-aligned size of the buffer prefix that - * has the same matching status as the first sector. + * @pnum is set to the size of the buffer prefix aligned to @chsize that + * has the same matching status as the first chunk. */ static int compare_buffers(const uint8_t *buf1, const uint8_t *buf2, - int64_t bytes, int64_t *pnum) + int64_t bytes, uint64_t chsize, int64_t *pnum) { bool res; - int64_t i = MIN(bytes, BDRV_SECTOR_SIZE); + int64_t i; assert(bytes > 0); + if (!chsize) { + chsize = BDRV_SECTOR_SIZE; + } + i = MIN(bytes, chsize); + res = !!memcmp(buf1, buf2, i); while (i < bytes) { - int64_t len = MIN(bytes - i, BDRV_SECTOR_SIZE); + int64_t len = MIN(bytes - i, chsize); if (!!memcmp(buf1 + i, buf2 + i, len) != res) { break; @@ -1559,7 +1565,7 @@ static int img_compare(int argc, char **argv) ret = 4; goto out; } - ret = compare_buffers(buf1, buf2, chunk, &pnum); + ret = compare_buffers(buf1, buf2, chunk, 0, &pnum); if (ret || pnum != chunk) { qprintf(quiet, "Content mismatch at offset %" PRId64 "!\n", offset + (ret ? 0 : pnum)); @@ -3878,7 +3884,7 @@ static int img_rebase(int argc, char **argv) int64_t pnum; if (compare_buffers(buf_old + written, buf_new + written, - n - written, &pnum)) + n - written, 0, &pnum)) { if (buf_old_is_zero) { ret = blk_pwrite_zeroes(blk, offset + written, pnum, 0); From patchwork Tue Sep 19 16:58:01 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrey Drobyshev X-Patchwork-Id: 13391605 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 lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 0CB1BCE79A9 for ; Tue, 19 Sep 2023 16:59:28 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qie3J-0003jI-Ns; Tue, 19 Sep 2023 12:58:22 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie3D-0003h7-7W; Tue, 19 Sep 2023 12:58:15 -0400 Received: from relay.virtuozzo.com ([130.117.225.111]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie37-0002Mf-PS; Tue, 19 Sep 2023 12:58:14 -0400 Received: from [130.117.225.1] (helo=dev005.ch-qa.vzint.dev) by relay.virtuozzo.com with esmtp (Exim 4.96) (envelope-from ) id 1qidza-00DUte-06; Tue, 19 Sep 2023 18:57:54 +0200 To: qemu-block@nongnu.org Cc: qemu-devel@nongnu.org, hreitz@redhat.com, kwolf@redhat.com, eblake@redhat.com, andrey.drobyshev@virtuozzo.com, den@virtuozzo.com Subject: [PATCH v3 5/8] qemu-img: rebase: avoid unnecessary COW operations Date: Tue, 19 Sep 2023 19:58:01 +0300 Message-Id: <20230919165804.439110-6-andrey.drobyshev@virtuozzo.com> X-Mailer: git-send-email 2.39.3 In-Reply-To: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> References: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> MIME-Version: 1.0 Received-SPF: pass client-ip=130.117.225.111; envelope-from=andrey.drobyshev@virtuozzo.com; helo=relay.virtuozzo.com X-Spam_score_int: -18 X-Spam_score: -1.9 X-Spam_bar: - X-Spam_report: (-1.9 / 5.0 requ) BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-to: Andrey Drobyshev X-Patchwork-Original-From: Andrey Drobyshev via From: Andrey Drobyshev Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org When rebasing an image from one backing file to another, we need to compare data from old and new backings. If the diff between that data happens to be unaligned to the target cluster size, we might end up doing partial writes, which would lead to copy-on-write and additional IO. Consider the following simple case (virtual_size == cluster_size == 64K): base <-- inc1 <-- inc2 qemu-io -c "write -P 0xaa 0 32K" base.qcow2 qemu-io -c "write -P 0xcc 32K 32K" base.qcow2 qemu-io -c "write -P 0xbb 0 32K" inc1.qcow2 qemu-io -c "write -P 0xcc 32K 32K" inc1.qcow2 qemu-img rebase -f qcow2 -b base.qcow2 -F qcow2 inc2.qcow2 While doing rebase, we'll write a half of the cluster to inc2, and block layer will have to read the 2nd half of the same cluster from the base image inc1 while doing this write operation, although the whole cluster is already read earlier to perform data comparison. In order to avoid these unnecessary IO cycles, let's make sure every write request is aligned to the overlay subcluster boundaries. Using subcluster size is universal as for the images which don't have them this size equals to the cluster size. so in any case we end up aligning to the smallest unit of allocation. Signed-off-by: Andrey Drobyshev --- qemu-img.c | 74 +++++++++++++++++++++++++++++++++++++++--------------- 1 file changed, 54 insertions(+), 20 deletions(-) diff --git a/qemu-img.c b/qemu-img.c index 0f67b021f7..a2d6241648 100644 --- a/qemu-img.c +++ b/qemu-img.c @@ -3523,6 +3523,7 @@ static int img_rebase(int argc, char **argv) uint8_t *buf_new = NULL; BlockDriverState *bs = NULL, *prefix_chain_bs = NULL; BlockDriverState *unfiltered_bs; + BlockDriverInfo bdi = {0}; char *filename; const char *fmt, *cache, *src_cache, *out_basefmt, *out_baseimg; int c, flags, src_flags, ret; @@ -3533,6 +3534,7 @@ static int img_rebase(int argc, char **argv) bool quiet = false; Error *local_err = NULL; bool image_opts = false; + int64_t write_align; /* Parse commandline parameters */ fmt = NULL; @@ -3656,6 +3658,20 @@ static int img_rebase(int argc, char **argv) } } + /* + * We need overlay subcluster size to make sure write requests are + * aligned. + */ + ret = bdrv_get_info(unfiltered_bs, &bdi); + if (ret < 0) { + error_report("could not get block driver info"); + goto out; + } else if (bdi.subcluster_size == 0) { + bdi.subcluster_size = 1; + } + + write_align = bdi.subcluster_size; + /* For safe rebasing we need to compare old and new backing file */ if (!unsafe) { QDict *options = NULL; @@ -3753,7 +3769,7 @@ static int img_rebase(int argc, char **argv) int64_t old_backing_size = 0; int64_t new_backing_size = 0; uint64_t offset; - int64_t n; + int64_t n, n_old = 0, n_new = 0; float local_progress = 0; if (blk_old_backing && bdrv_opt_mem_align(blk_bs(blk_old_backing)) > @@ -3799,7 +3815,8 @@ static int img_rebase(int argc, char **argv) } for (offset = 0; offset < size; offset += n) { - bool buf_old_is_zero = false; + bool old_backing_eof = false; + int64_t n_alloc; /* How many bytes can we handle with the next read? */ n = MIN(IO_BUF_SIZE, size - offset); @@ -3844,33 +3861,46 @@ static int img_rebase(int argc, char **argv) } } + /* + * At this point we know that the region [offset; offset + n) + * is unallocated within the target image. This region might be + * unaligned to the target image's (sub)cluster boundaries, as + * old backing may have smaller clusters (or have subclusters). + * We extend it to the aligned boundaries to avoid CoW on + * partial writes in blk_pwrite(), + */ + n += offset - QEMU_ALIGN_DOWN(offset, write_align); + offset = QEMU_ALIGN_DOWN(offset, write_align); + n += QEMU_ALIGN_UP(offset + n, write_align) - (offset + n); + n = MIN(n, size - offset); + assert(!bdrv_is_allocated(unfiltered_bs, offset, n, &n_alloc) && + n_alloc == n); + + /* + * Much like with the target image, we'll try to read as much + * of the old and new backings as we can. + */ + n_old = MIN(n, MAX(0, old_backing_size - (int64_t) offset)); + n_new = MIN(n, MAX(0, new_backing_size - (int64_t) offset)); + /* * Read old and new backing file and take into consideration that * backing files may be smaller than the COW image. */ - if (offset >= old_backing_size) { - memset(buf_old, 0, n); - buf_old_is_zero = true; + memset(buf_old + n_old, 0, n - n_old); + if (!n_old) { + old_backing_eof = true; } else { - if (offset + n > old_backing_size) { - n = old_backing_size - offset; - } - - ret = blk_pread(blk_old_backing, offset, n, buf_old, 0); + ret = blk_pread(blk_old_backing, offset, n_old, buf_old, 0); if (ret < 0) { error_report("error while reading from old backing file"); goto out; } } - if (offset >= new_backing_size || !blk_new_backing) { - memset(buf_new, 0, n); - } else { - if (offset + n > new_backing_size) { - n = new_backing_size - offset; - } - - ret = blk_pread(blk_new_backing, offset, n, buf_new, 0); + memset(buf_new + n_new, 0, n - n_new); + if (n_new) { + ret = blk_pread(blk_new_backing, offset, n_new, buf_new, 0); if (ret < 0) { error_report("error while reading from new backing file"); goto out; @@ -3884,11 +3914,12 @@ static int img_rebase(int argc, char **argv) int64_t pnum; if (compare_buffers(buf_old + written, buf_new + written, - n - written, 0, &pnum)) + n - written, write_align, &pnum)) { - if (buf_old_is_zero) { + if (old_backing_eof) { ret = blk_pwrite_zeroes(blk, offset + written, pnum, 0); } else { + assert(written + pnum <= IO_BUF_SIZE); ret = blk_pwrite(blk, offset + written, pnum, buf_old + written, 0); } @@ -3900,6 +3931,9 @@ static int img_rebase(int argc, char **argv) } written += pnum; + if (offset + written >= old_backing_size) { + old_backing_eof = true; + } } qemu_progress_print(local_progress, 100); } From patchwork Tue Sep 19 16:58:02 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrey Drobyshev X-Patchwork-Id: 13391611 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 lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 125C2CE79A9 for ; Tue, 19 Sep 2023 17:00:20 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qie3H-0003iz-U8; Tue, 19 Sep 2023 12:58:20 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie3D-0003h8-7g; Tue, 19 Sep 2023 12:58:16 -0400 Received: from relay.virtuozzo.com ([130.117.225.111]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie37-0002Mh-IL; Tue, 19 Sep 2023 12:58:14 -0400 Received: from [130.117.225.1] (helo=dev005.ch-qa.vzint.dev) by relay.virtuozzo.com with esmtp (Exim 4.96) (envelope-from ) id 1qidza-00DUte-08; Tue, 19 Sep 2023 18:57:54 +0200 To: qemu-block@nongnu.org Cc: qemu-devel@nongnu.org, hreitz@redhat.com, kwolf@redhat.com, eblake@redhat.com, andrey.drobyshev@virtuozzo.com, den@virtuozzo.com Subject: [PATCH v3 6/8] iotests/{024, 271}: add testcases for qemu-img rebase Date: Tue, 19 Sep 2023 19:58:02 +0300 Message-Id: <20230919165804.439110-7-andrey.drobyshev@virtuozzo.com> X-Mailer: git-send-email 2.39.3 In-Reply-To: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> References: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> MIME-Version: 1.0 Received-SPF: pass client-ip=130.117.225.111; envelope-from=andrey.drobyshev@virtuozzo.com; helo=relay.virtuozzo.com X-Spam_score_int: -18 X-Spam_score: -1.9 X-Spam_bar: - X-Spam_report: (-1.9 / 5.0 requ) BAYES_00=-1.9, PDS_OTHER_BAD_TLD=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-to: Andrey Drobyshev X-Patchwork-Original-From: Andrey Drobyshev via From: Andrey Drobyshev Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org As the previous commit changes the logic of "qemu-img rebase" (it's using write alignment now), let's add a couple more test cases which would ensure it works correctly. In particular, the following scenarios: 024: add test case for rebase within one backing chain when the overlay cluster size > backings cluster size; 271: add test case for rebase images that contain subclusters. Check that no extra allocations are being made. Signed-off-by: Andrey Drobyshev Reviewed-by: Hanna Czenczek --- tests/qemu-iotests/024 | 60 ++++++++++++++++++++++++++++++++++ tests/qemu-iotests/024.out | 43 +++++++++++++++++++++++++ tests/qemu-iotests/271 | 66 ++++++++++++++++++++++++++++++++++++++ tests/qemu-iotests/271.out | 42 ++++++++++++++++++++++++ 4 files changed, 211 insertions(+) diff --git a/tests/qemu-iotests/024 b/tests/qemu-iotests/024 index 98a7c8fd65..285f17e79f 100755 --- a/tests/qemu-iotests/024 +++ b/tests/qemu-iotests/024 @@ -257,6 +257,66 @@ $QEMU_IO "$OVERLAY" -c "read -P 0x00 $(( CLUSTER_SIZE * 4 )) $CLUSTER_SIZE" \ echo +# Check that rebase within the chain is working when +# overlay cluster size > backings cluster size +# (here overlay cluster size == 2 * backings cluster size) +# +# base_new <-- base_old <-- overlay +# +# Backing (new): -- -- -- -- -- -- +# Backing (old): -- 11 -- -- 22 -- +# Overlay: |-- --|-- --|-- --| +# +# We should end up having 1st and 3rd cluster allocated, and their halves +# being read as zeroes. + +echo +echo "=== Test rebase with different cluster sizes ===" +echo + +echo "Creating backing chain" +echo + +TEST_IMG=$BASE_NEW _make_test_img $(( CLUSTER_SIZE * 6 )) +TEST_IMG=$BASE_OLD _make_test_img -b "$BASE_NEW" -F $IMGFMT \ + $(( CLUSTER_SIZE * 6 )) +CLUSTER_SIZE=$(( CLUSTER_SIZE * 2 )) TEST_IMG=$OVERLAY \ + _make_test_img -b "$BASE_OLD" -F $IMGFMT $(( CLUSTER_SIZE * 6 )) + +TEST_IMG=$OVERLAY _img_info + +echo +echo "Fill backing files with data" +echo + +$QEMU_IO "$BASE_OLD" -c "write -P 0x11 $CLUSTER_SIZE $CLUSTER_SIZE" \ + -c "write -P 0x22 $(( CLUSTER_SIZE * 4 )) $CLUSTER_SIZE" \ + | _filter_qemu_io + +echo +echo "Rebase onto another image in the same chain" +echo + +$QEMU_IMG rebase -b "$BASE_NEW" -F $IMGFMT "$OVERLAY" + +echo "Verify that data is read the same before and after rebase" +echo + +$QEMU_IO "$OVERLAY" -c "read -P 0x00 0 $CLUSTER_SIZE" \ + -c "read -P 0x11 $CLUSTER_SIZE $CLUSTER_SIZE" \ + -c "read -P 0x00 $(( CLUSTER_SIZE * 2 )) $(( CLUSTER_SIZE * 2 ))" \ + -c "read -P 0x22 $(( CLUSTER_SIZE * 4 )) $CLUSTER_SIZE" \ + -c "read -P 0x00 $(( CLUSTER_SIZE * 5 )) $CLUSTER_SIZE" \ + | _filter_qemu_io + +echo +echo "Verify that untouched cluster remains unallocated" +echo + +$QEMU_IMG map "$OVERLAY" | _filter_qemu_img_map + +echo + # success, all done echo "*** done" rm -f $seq.full diff --git a/tests/qemu-iotests/024.out b/tests/qemu-iotests/024.out index 245fe8b1d1..e1e8eea863 100644 --- a/tests/qemu-iotests/024.out +++ b/tests/qemu-iotests/024.out @@ -201,4 +201,47 @@ read 262144/262144 bytes at offset 0 read 65536/65536 bytes at offset 262144 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +=== Test rebase with different cluster sizes === + +Creating backing chain + +Formatting 'TEST_DIR/subdir/t.IMGFMT.base_new', fmt=IMGFMT size=393216 +Formatting 'TEST_DIR/subdir/t.IMGFMT.base_old', fmt=IMGFMT size=393216 backing_file=TEST_DIR/subdir/t.IMGFMT.base_new backing_fmt=IMGFMT +Formatting 'TEST_DIR/subdir/t.IMGFMT', fmt=IMGFMT size=393216 backing_file=TEST_DIR/subdir/t.IMGFMT.base_old backing_fmt=IMGFMT +image: TEST_DIR/subdir/t.IMGFMT +file format: IMGFMT +virtual size: 384 KiB (393216 bytes) +cluster_size: 131072 +backing file: TEST_DIR/subdir/t.IMGFMT.base_old +backing file format: IMGFMT + +Fill backing files with data + +wrote 65536/65536 bytes at offset 65536 +64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 65536/65536 bytes at offset 262144 +64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +Rebase onto another image in the same chain + +Verify that data is read the same before and after rebase + +read 65536/65536 bytes at offset 0 +64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 65536/65536 bytes at offset 65536 +64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 131072/131072 bytes at offset 131072 +128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 65536/65536 bytes at offset 262144 +64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 65536/65536 bytes at offset 327680 +64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +Verify that untouched cluster remains unallocated + +Offset Length File +0 0x20000 TEST_DIR/subdir/t.IMGFMT +0x40000 0x20000 TEST_DIR/subdir/t.IMGFMT + *** done diff --git a/tests/qemu-iotests/271 b/tests/qemu-iotests/271 index c7c2cadda0..e243f57ba7 100755 --- a/tests/qemu-iotests/271 +++ b/tests/qemu-iotests/271 @@ -899,6 +899,72 @@ _concurrent_io | $QEMU_IO | _filter_qemu_io | \ sed -e 's/\(20480\|40960\)/OFFSET/' _concurrent_verify | $QEMU_IO | _filter_qemu_io +############################################################ +############################################################ +############################################################ + +echo +echo "### Rebase of qcow2 images with subclusters ###" +echo + +l2_offset=$((0x400000)) + +# Check that rebase operation preserve holes between allocated subclusters +# within one cluster (i.e. does not allocate extra space). Check that the +# data is preserved as well. +# +# Base (new backing): -- -- -- ... -- -- -- +# Mid (old backing): -- 11 -- ... -- 22 -- +# Top: -- -- -- ... -- -- -- + +echo "### Preservation of unallocated holes after rebase ###" +echo + +echo "# create backing chain" +echo + +TEST_IMG="$TEST_IMG.base" _make_test_img -o cluster_size=1M,extended_l2=on 1M +TEST_IMG="$TEST_IMG.mid" _make_test_img -o cluster_size=1M,extended_l2=on \ + -b "$TEST_IMG.base" -F qcow2 1M +TEST_IMG="$TEST_IMG.top" _make_test_img -o cluster_size=1M,extended_l2=on \ + -b "$TEST_IMG.mid" -F qcow2 1M + +echo +echo "# fill old backing with data (separate subclusters within cluster)" +echo + +$QEMU_IO -c "write -P 0x11 32k 32k" \ + -c "write -P 0x22 $(( 30 * 32 ))k 32k" \ + "$TEST_IMG.mid" | _filter_qemu_io + +echo +echo "# rebase topmost image onto the new backing" +echo + +$QEMU_IMG rebase -b "$TEST_IMG.base" -F qcow2 "$TEST_IMG.top" + +echo "# verify that data is read the same before and after rebase" +echo + +$QEMU_IO -c "read -P 0x00 0 32k" \ + -c "read -P 0x11 32k 32k" \ + -c "read -P 0x00 64k $(( 28 * 32 ))k" \ + -c "read -P 0x22 $(( 30 * 32 ))k 32k" \ + -c "read -P 0x00 $(( 31 * 32 ))k 32k" \ + "$TEST_IMG.top" | _filter_qemu_io + +echo +echo "# verify that only selected subclusters remain allocated" +echo + +$QEMU_IMG map "$TEST_IMG.top" | _filter_testdir + +echo +echo "# verify image bitmap" +echo + +TEST_IMG="$TEST_IMG.top" alloc="1 30" zero="" _verify_l2_bitmap 0 + # success, all done echo "*** done" rm -f $seq.full diff --git a/tests/qemu-iotests/271.out b/tests/qemu-iotests/271.out index 5be780de76..c335a6c608 100644 --- a/tests/qemu-iotests/271.out +++ b/tests/qemu-iotests/271.out @@ -723,4 +723,46 @@ wrote 2048/2048 bytes at offset OFFSET 2 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) wrote 2048/2048 bytes at offset OFFSET 2 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +### Rebase of qcow2 images with subclusters ### + +### Preservation of unallocated holes after rebase ### + +# create backing chain + +Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=1048576 +Formatting 'TEST_DIR/t.IMGFMT.mid', fmt=IMGFMT size=1048576 backing_file=TEST_DIR/t.IMGFMT.base backing_fmt=IMGFMT +Formatting 'TEST_DIR/t.IMGFMT.top', fmt=IMGFMT size=1048576 backing_file=TEST_DIR/t.IMGFMT.mid backing_fmt=IMGFMT + +# fill old backing with data (separate subclusters within cluster) + +wrote 32768/32768 bytes at offset 32768 +32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 32768/32768 bytes at offset 983040 +32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +# rebase topmost image onto the new backing + +# verify that data is read the same before and after rebase + +read 32768/32768 bytes at offset 0 +32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 32768/32768 bytes at offset 32768 +32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 917504/917504 bytes at offset 65536 +896 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 32768/32768 bytes at offset 983040 +32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 32768/32768 bytes at offset 1015808 +32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +# verify that only selected subclusters remain allocated + +Offset Length Mapped to File +0x8000 0x8000 0x508000 TEST_DIR/t.qcow2.top +0xf0000 0x8000 0x5f0000 TEST_DIR/t.qcow2.top + +# verify image bitmap + +L2 entry #0: 0x8000000000500000 0000000040000002 *** done From patchwork Tue Sep 19 16:58:03 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrey Drobyshev X-Patchwork-Id: 13391608 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 lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 18E52CE79AA for ; Tue, 19 Sep 2023 16:59:29 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qie3C-0003gi-Nq; Tue, 19 Sep 2023 12:58:14 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie3B-0003e5-2g; Tue, 19 Sep 2023 12:58:13 -0400 Received: from relay.virtuozzo.com ([130.117.225.111]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie36-0002Ma-Rg; Tue, 19 Sep 2023 12:58:11 -0400 Received: from [130.117.225.1] (helo=dev005.ch-qa.vzint.dev) by relay.virtuozzo.com with esmtp (Exim 4.96) (envelope-from ) id 1qidza-00DUte-0J; Tue, 19 Sep 2023 18:57:54 +0200 To: qemu-block@nongnu.org Cc: qemu-devel@nongnu.org, hreitz@redhat.com, kwolf@redhat.com, eblake@redhat.com, andrey.drobyshev@virtuozzo.com, den@virtuozzo.com Subject: [PATCH v3 7/8] qemu-img: add compression option to rebase subcommand Date: Tue, 19 Sep 2023 19:58:03 +0300 Message-Id: <20230919165804.439110-8-andrey.drobyshev@virtuozzo.com> X-Mailer: git-send-email 2.39.3 In-Reply-To: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> References: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> MIME-Version: 1.0 Received-SPF: pass client-ip=130.117.225.111; envelope-from=andrey.drobyshev@virtuozzo.com; helo=relay.virtuozzo.com X-Spam_score_int: -18 X-Spam_score: -1.9 X-Spam_bar: - X-Spam_report: (-1.9 / 5.0 requ) BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-to: Andrey Drobyshev X-Patchwork-Original-From: Andrey Drobyshev via From: Andrey Drobyshev Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org If we rebase an image whose backing file has compressed clusters, we might end up wasting disk space since the copied clusters are now uncompressed. In order to have better control over this, let's add "--compress" option to the "qemu-img rebase" command. Note that this option affects only the clusters which are actually being copied from the original backing file. The clusters which were uncompressed in the target image will remain so. Signed-off-by: Andrey Drobyshev Reviewed-by: Denis V. Lunev Reviewed-by: Hanna Czenczek --- docs/tools/qemu-img.rst | 6 ++++-- qemu-img-cmds.hx | 4 ++-- qemu-img.c | 26 ++++++++++++++++++++------ 3 files changed, 26 insertions(+), 10 deletions(-) diff --git a/docs/tools/qemu-img.rst b/docs/tools/qemu-img.rst index ca5a2773cf..4459c065f1 100644 --- a/docs/tools/qemu-img.rst +++ b/docs/tools/qemu-img.rst @@ -667,7 +667,7 @@ Command description: List, apply, create or delete snapshots in image *FILENAME*. -.. option:: rebase [--object OBJECTDEF] [--image-opts] [-U] [-q] [-f FMT] [-t CACHE] [-T SRC_CACHE] [-p] [-u] -b BACKING_FILE [-F BACKING_FMT] FILENAME +.. option:: rebase [--object OBJECTDEF] [--image-opts] [-U] [-q] [-f FMT] [-t CACHE] [-T SRC_CACHE] [-p] [-u] [-c] -b BACKING_FILE [-F BACKING_FMT] FILENAME Changes the backing file of an image. Only the formats ``qcow2`` and ``qed`` support changing the backing file. @@ -694,7 +694,9 @@ Command description: In order to achieve this, any clusters that differ between *BACKING_FILE* and the old backing file of *FILENAME* are merged - into *FILENAME* before actually changing the backing file. + into *FILENAME* before actually changing the backing file. With the + ``-c`` option specified, the clusters which are being merged (but not + the entire *FILENAME* image) are compressed when written. Note that the safe mode is an expensive operation, comparable to converting an image. It only works if the old backing file still diff --git a/qemu-img-cmds.hx b/qemu-img-cmds.hx index 1b1dab5b17..068692d13e 100644 --- a/qemu-img-cmds.hx +++ b/qemu-img-cmds.hx @@ -88,9 +88,9 @@ SRST ERST DEF("rebase", img_rebase, - "rebase [--object objectdef] [--image-opts] [-U] [-q] [-f fmt] [-t cache] [-T src_cache] [-p] [-u] -b backing_file [-F backing_fmt] filename") + "rebase [--object objectdef] [--image-opts] [-U] [-q] [-f fmt] [-t cache] [-T src_cache] [-p] [-u] [-c] -b backing_file [-F backing_fmt] filename") SRST -.. option:: rebase [--object OBJECTDEF] [--image-opts] [-U] [-q] [-f FMT] [-t CACHE] [-T SRC_CACHE] [-p] [-u] -b BACKING_FILE [-F BACKING_FMT] FILENAME +.. option:: rebase [--object OBJECTDEF] [--image-opts] [-U] [-q] [-f FMT] [-t CACHE] [-T SRC_CACHE] [-p] [-u] [-c] -b BACKING_FILE [-F BACKING_FMT] FILENAME ERST DEF("resize", img_resize, diff --git a/qemu-img.c b/qemu-img.c index a2d6241648..9d7a4a3566 100644 --- a/qemu-img.c +++ b/qemu-img.c @@ -3527,11 +3527,13 @@ static int img_rebase(int argc, char **argv) char *filename; const char *fmt, *cache, *src_cache, *out_basefmt, *out_baseimg; int c, flags, src_flags, ret; + BdrvRequestFlags write_flags = 0; bool writethrough, src_writethrough; int unsafe = 0; bool force_share = false; int progress = 0; bool quiet = false; + bool compress = false; Error *local_err = NULL; bool image_opts = false; int64_t write_align; @@ -3548,9 +3550,10 @@ static int img_rebase(int argc, char **argv) {"object", required_argument, 0, OPTION_OBJECT}, {"image-opts", no_argument, 0, OPTION_IMAGE_OPTS}, {"force-share", no_argument, 0, 'U'}, + {"compress", no_argument, 0, 'c'}, {0, 0, 0, 0} }; - c = getopt_long(argc, argv, ":hf:F:b:upt:T:qU", + c = getopt_long(argc, argv, ":hf:F:b:upt:T:qUc", long_options, NULL); if (c == -1) { break; @@ -3598,6 +3601,9 @@ static int img_rebase(int argc, char **argv) case 'U': force_share = true; break; + case 'c': + compress = true; + break; } } @@ -3650,6 +3656,14 @@ static int img_rebase(int argc, char **argv) unfiltered_bs = bdrv_skip_filters(bs); + if (compress && !block_driver_can_compress(unfiltered_bs->drv)) { + error_report("Compression not supported for this file format"); + ret = -1; + goto out; + } else if (compress) { + write_flags |= BDRV_REQ_WRITE_COMPRESSED; + } + if (out_basefmt != NULL) { if (bdrv_find_format(out_basefmt) == NULL) { error_report("Invalid format name: '%s'", out_basefmt); @@ -3659,18 +3673,18 @@ static int img_rebase(int argc, char **argv) } /* - * We need overlay subcluster size to make sure write requests are - * aligned. + * We need overlay subcluster size (or cluster size in case writes are + * compressed) to make sure write requests are aligned. */ ret = bdrv_get_info(unfiltered_bs, &bdi); if (ret < 0) { error_report("could not get block driver info"); goto out; } else if (bdi.subcluster_size == 0) { - bdi.subcluster_size = 1; + bdi.cluster_size = bdi.subcluster_size = 1; } - write_align = bdi.subcluster_size; + write_align = compress ? bdi.cluster_size : bdi.subcluster_size; /* For safe rebasing we need to compare old and new backing file */ if (!unsafe) { @@ -3921,7 +3935,7 @@ static int img_rebase(int argc, char **argv) } else { assert(written + pnum <= IO_BUF_SIZE); ret = blk_pwrite(blk, offset + written, pnum, - buf_old + written, 0); + buf_old + written, write_flags); } if (ret < 0) { error_report("Error while writing to COW image: %s", From patchwork Tue Sep 19 16:58:04 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrey Drobyshev X-Patchwork-Id: 13391612 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 lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 39571CE79A9 for ; Tue, 19 Sep 2023 17:00:28 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qie3M-0003jf-9M; Tue, 19 Sep 2023 12:58:24 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie3D-0003hF-Qe; Tue, 19 Sep 2023 12:58:16 -0400 Received: from relay.virtuozzo.com ([130.117.225.111]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qie37-0002Me-G5; Tue, 19 Sep 2023 12:58:15 -0400 Received: from [130.117.225.1] (helo=dev005.ch-qa.vzint.dev) by relay.virtuozzo.com with esmtp (Exim 4.96) (envelope-from ) id 1qidza-00DUte-0L; Tue, 19 Sep 2023 18:57:54 +0200 To: qemu-block@nongnu.org Cc: qemu-devel@nongnu.org, hreitz@redhat.com, kwolf@redhat.com, eblake@redhat.com, andrey.drobyshev@virtuozzo.com, den@virtuozzo.com Subject: [PATCH v3 8/8] iotests: add tests for "qemu-img rebase" with compression Date: Tue, 19 Sep 2023 19:58:04 +0300 Message-Id: <20230919165804.439110-9-andrey.drobyshev@virtuozzo.com> X-Mailer: git-send-email 2.39.3 In-Reply-To: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> References: <20230919165804.439110-1-andrey.drobyshev@virtuozzo.com> MIME-Version: 1.0 Received-SPF: pass client-ip=130.117.225.111; envelope-from=andrey.drobyshev@virtuozzo.com; helo=relay.virtuozzo.com X-Spam_score_int: -18 X-Spam_score: -1.9 X-Spam_bar: - X-Spam_report: (-1.9 / 5.0 requ) BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-to: Andrey Drobyshev X-Patchwork-Original-From: Andrey Drobyshev via From: Andrey Drobyshev Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org The test cases considered so far: 314 (new test suite): 1. Check that compression mode isn't compatible with "-f raw" (raw format doesn't support compression). 2. Check that rebasing an image onto no backing file preserves the data and writes the copied clusters actually compressed. 3. Same as 2, but with a raw backing file (i.e. the clusters copied from the backing are originally uncompressed -- we check they end up compressed after being merged). 4. Remove a single delta from a backing chain, perform the same checks as in 2. 5. Check that even when backing and overlay are initially uncompressed, copied clusters end up compressed when rebase with compression is performed. 271: 1. Check that when target image has subclusters, rebase with compression will make an entire cluster containing the written subcluster compressed. Signed-off-by: Andrey Drobyshev Reviewed-by: Hanna Czenczek --- tests/qemu-iotests/271 | 65 +++++++++++++++ tests/qemu-iotests/271.out | 40 +++++++++ tests/qemu-iotests/314 | 165 +++++++++++++++++++++++++++++++++++++ tests/qemu-iotests/314.out | 75 +++++++++++++++++ 4 files changed, 345 insertions(+) create mode 100755 tests/qemu-iotests/314 create mode 100644 tests/qemu-iotests/314.out diff --git a/tests/qemu-iotests/271 b/tests/qemu-iotests/271 index e243f57ba7..59a6fafa2f 100755 --- a/tests/qemu-iotests/271 +++ b/tests/qemu-iotests/271 @@ -965,6 +965,71 @@ echo TEST_IMG="$TEST_IMG.top" alloc="1 30" zero="" _verify_l2_bitmap 0 +# Check that rebase with compression works correctly with images containing +# subclusters. When compression is enabled and we allocate a new +# subcluster within the target (overlay) image, we expect the entire cluster +# containing that subcluster to become compressed. +# +# Here we expect 1st and 3rd clusters of the top (overlay) image to become +# compressed after the rebase, while cluster 2 to remain unallocated and +# be read from the base (new backing) image. +# +# Base (new backing): |-- -- .. -- --|11 11 .. 11 11|-- -- .. -- --| +# Mid (old backing): |-- -- .. -- 22|-- -- .. -- --|33 -- .. -- --| +# Top: |-- -- .. -- --|-- -- -- -- --|-- -- .. -- --| + +echo +echo "### Rebase with compression for images with subclusters ###" +echo + +echo "# create backing chain" +echo + +TEST_IMG="$TEST_IMG.base" _make_test_img -o cluster_size=1M,extended_l2=on 3M +TEST_IMG="$TEST_IMG.mid" _make_test_img -o cluster_size=1M,extended_l2=on \ + -b "$TEST_IMG.base" -F qcow2 3M +TEST_IMG="$TEST_IMG.top" _make_test_img -o cluster_size=1M,extended_l2=on \ + -b "$TEST_IMG.mid" -F qcow2 3M + +echo +echo "# fill old and new backing with data" +echo + +$QEMU_IO -c "write -P 0x11 1M 1M" "$TEST_IMG.base" | _filter_qemu_io +$QEMU_IO -c "write -P 0x22 $(( 31 * 32 ))k 32k" \ + -c "write -P 0x33 $(( 64 * 32 ))k 32k" \ + "$TEST_IMG.mid" | _filter_qemu_io + +echo +echo "# rebase topmost image onto the new backing, with compression" +echo + +$QEMU_IMG rebase -c -b "$TEST_IMG.base" -F qcow2 "$TEST_IMG.top" + +echo "# verify that the 1st and 3rd clusters've become compressed" +echo + +$QEMU_IMG map --output=json "$TEST_IMG.top" | _filter_testdir + +echo +echo "# verify that data is read the same before and after rebase" +echo + +$QEMU_IO -c "read -P 0x22 $(( 31 * 32 ))k 32k" \ + -c "read -P 0x11 1M 1M" \ + -c "read -P 0x33 $(( 64 * 32 ))k 32k" \ + "$TEST_IMG.top" | _filter_qemu_io + +echo +echo "# verify image bitmap" +echo + +# For compressed clusters bitmap is always 0. For unallocated cluster +# there should be no entry at all, thus bitmap is also 0. +TEST_IMG="$TEST_IMG.top" alloc="" zero="" _verify_l2_bitmap 0 +TEST_IMG="$TEST_IMG.top" alloc="" zero="" _verify_l2_bitmap 1 +TEST_IMG="$TEST_IMG.top" alloc="" zero="" _verify_l2_bitmap 2 + # success, all done echo "*** done" rm -f $seq.full diff --git a/tests/qemu-iotests/271.out b/tests/qemu-iotests/271.out index c335a6c608..0b24d50159 100644 --- a/tests/qemu-iotests/271.out +++ b/tests/qemu-iotests/271.out @@ -765,4 +765,44 @@ Offset Length Mapped to File # verify image bitmap L2 entry #0: 0x8000000000500000 0000000040000002 + +### Rebase with compression for images with subclusters ### + +# create backing chain + +Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=3145728 +Formatting 'TEST_DIR/t.IMGFMT.mid', fmt=IMGFMT size=3145728 backing_file=TEST_DIR/t.IMGFMT.base backing_fmt=IMGFMT +Formatting 'TEST_DIR/t.IMGFMT.top', fmt=IMGFMT size=3145728 backing_file=TEST_DIR/t.IMGFMT.mid backing_fmt=IMGFMT + +# fill old and new backing with data + +wrote 1048576/1048576 bytes at offset 1048576 +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 32768/32768 bytes at offset 1015808 +32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 32768/32768 bytes at offset 2097152 +32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +# rebase topmost image onto the new backing, with compression + +# verify that the 1st and 3rd clusters've become compressed + +[{ "start": 0, "length": 1048576, "depth": 0, "present": true, "zero": false, "data": true, "compressed": true}, +{ "start": 1048576, "length": 1048576, "depth": 1, "present": true, "zero": false, "data": true, "compressed": false, "offset": 5242880}, +{ "start": 2097152, "length": 1048576, "depth": 0, "present": true, "zero": false, "data": true, "compressed": true}] + +# verify that data is read the same before and after rebase + +read 32768/32768 bytes at offset 1015808 +32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 1048576/1048576 bytes at offset 1048576 +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 32768/32768 bytes at offset 2097152 +32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) + +# verify image bitmap + +L2 entry #0: 0x4008000000500000 0000000000000000 +L2 entry #1: 0x0000000000000000 0000000000000000 +L2 entry #2: 0x400800000050040b 0000000000000000 *** done diff --git a/tests/qemu-iotests/314 b/tests/qemu-iotests/314 new file mode 100755 index 0000000000..96d7b4d258 --- /dev/null +++ b/tests/qemu-iotests/314 @@ -0,0 +1,165 @@ +#!/usr/bin/env bash +# group: rw backing auto quick +# +# Test qemu-img rebase with compression +# +# Copyright (c) 2023 Virtuozzo International GmbH. +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program. If not, see . +# + +# creator +owner=andrey.drobyshev@virtuozzo.com + +seq=`basename $0` +echo "QA output created by $seq" + +status=1 # failure is the default! + +_cleanup() +{ + _cleanup_test_img + _rm_test_img "$TEST_IMG.base" + _rm_test_img "$TEST_IMG.itmd" +} +trap "_cleanup; exit \$status" 0 1 2 3 15 + +# get standard environment, filters and checks +. ./common.rc +. ./common.filter + +_supported_fmt qcow2 +_supported_proto file +_supported_os Linux + +# Want the size divisible by 2 and 3 +size=$(( 48 * 1024 * 1024 )) +half_size=$(( size / 2 )) +third_size=$(( size / 3 )) + +# 1. "qemu-img rebase -c" should refuse working with any format which doesn't +# support compression. We only check "-f raw" here. +echo +echo "=== Testing compressed rebase format compatibility ===" +echo + +$QEMU_IMG create -f raw "$TEST_IMG" "$size" | _filter_img_create +$QEMU_IMG rebase -c -f raw -b "" "$TEST_IMG" + +# 2. Write the 1st half of $size to backing file (compressed), 2nd half -- to +# the top image (also compressed). Rebase the top image onto no backing file, +# with compression (i.e. "qemu-img -c -b ''"). Check that the resulting image +# has the written data preserved, and "qemu-img check" reports 100% clusters +# as compressed. +echo +echo "=== Testing rebase with compression onto no backing file ===" +echo + +TEST_IMG="$TEST_IMG.base" _make_test_img $size +_make_test_img -b "$TEST_IMG.base" -F $IMGFMT $size + +$QEMU_IO -c "write -c -P 0xaa 0 $half_size" "$TEST_IMG.base" | _filter_qemu_io +$QEMU_IO -c "write -c -P 0xbb $half_size $half_size" "$TEST_IMG" \ + | _filter_qemu_io + +$QEMU_IMG rebase -c -f $IMGFMT -b "" "$TEST_IMG" + +$QEMU_IO -c "read -P 0xaa 0 $half_size" "$TEST_IMG" | _filter_qemu_io +$QEMU_IO -c "read -P 0xbb $half_size $half_size" "$TEST_IMG" | _filter_qemu_io + +$QEMU_IMG check "$TEST_IMG" | _filter_testdir + +# 3. Same as the previous one, but with raw backing file (hence write to +# the backing is uncompressed). +echo +echo "=== Testing rebase with compression with raw backing file ===" +echo + +$QEMU_IMG create -f raw "$TEST_IMG.base" "$half_size" | _filter_img_create +_make_test_img -b "$TEST_IMG.base" -F raw $size + +$QEMU_IO -f raw -c "write -P 0xaa 0 $half_size" "$TEST_IMG.base" \ + | _filter_qemu_io +$QEMU_IO -c "write -c -P 0xbb $half_size $half_size" \ + "$TEST_IMG" | _filter_qemu_io + +$QEMU_IMG rebase -c -f $IMGFMT -b "" "$TEST_IMG" + +$QEMU_IO -c "read -P 0xaa 0 $half_size" "$TEST_IMG" | _filter_qemu_io +$QEMU_IO -c "read -P 0xbb $half_size $half_size" "$TEST_IMG" | _filter_qemu_io + +$QEMU_IMG check "$TEST_IMG" | _filter_testdir + +# 4. Create a backing chain base<--itmd<--img, filling 1st, 2nd and 3rd +# thirds of them, respectively (with compression). Rebase img onto base, +# effectively deleting itmd from the chain, and check that written data is +# preserved in the resulting image. Also check that "qemu-img check" reports +# 100% clusters as compressed. +echo +echo "=== Testing compressed rebase removing single delta from the chain ===" +echo + +TEST_IMG="$TEST_IMG.base" _make_test_img $size +TEST_IMG="$TEST_IMG.itmd" _make_test_img -b "$TEST_IMG.base" -F $IMGFMT $size +_make_test_img -b "$TEST_IMG.itmd" -F $IMGFMT $size + +$QEMU_IO -c "write -c -P 0xaa 0 $third_size" \ + "$TEST_IMG.base" | _filter_qemu_io +$QEMU_IO -c "write -c -P 0xbb $third_size $third_size" \ + "$TEST_IMG.itmd" | _filter_qemu_io +$QEMU_IO -c "write -c -P 0xcc $((third_size * 2 )) $third_size" \ + "$TEST_IMG" | _filter_qemu_io + +$QEMU_IMG rebase -c -f $IMGFMT -b "$TEST_IMG.base" -F $IMGFMT "$TEST_IMG" + +$QEMU_IO -c "read -P 0xaa 0 $third_size" "$TEST_IMG" | _filter_qemu_io +$QEMU_IO -c "read -P 0xbb $third_size $third_size" \ + "$TEST_IMG" | _filter_qemu_io +$QEMU_IO -c "read -P 0xcc $(( third_size * 2 )) $third_size" \ + "$TEST_IMG" | _filter_qemu_io + +$QEMU_IMG check "$TEST_IMG" | _filter_testdir + +# 5. Create one-cluster backing and overlay images, and fill only the first +# (half - 1) bytes of the backing with data (uncompressed). Rebase the +# overlay onto no backing file with compression. Check that data is still +# read correctly, and that cluster is now really compressed ("qemu-img check" +# reports 100% clusters as compressed. +echo +echo "=== Testing compressed rebase with unaligned unmerged data ===" +echo + +CLUSTER_SIZE=65536 + +TEST_IMG="$TEST_IMG.base" _make_test_img $CLUSTER_SIZE +_make_test_img -b "$TEST_IMG.base" -F $IMGFMT $CLUSTER_SIZE + +$QEMU_IO -c "write -P 0xaa 0 $(( CLUSTER_SIZE / 2 - 1 ))" $TEST_IMG.base \ + | _filter_qemu_io + +$QEMU_IMG rebase -c -f $IMGFMT -b "" "$TEST_IMG" + +$QEMU_IO -c "read -P 0xaa 0 $(( CLUSTER_SIZE / 2 - 1 ))" "$TEST_IMG" \ + | _filter_qemu_io +$QEMU_IO -c \ + "read -P 0x00 $(( CLUSTER_SIZE / 2 - 1 )) $(( CLUSTER_SIZE / 2 + 1 ))" \ + "$TEST_IMG" | _filter_qemu_io + +$QEMU_IMG check "$TEST_IMG" | _filter_testdir + +# success, all done +echo +echo '*** done' +rm -f $seq.full +status=0 diff --git a/tests/qemu-iotests/314.out b/tests/qemu-iotests/314.out new file mode 100644 index 0000000000..ac9337a543 --- /dev/null +++ b/tests/qemu-iotests/314.out @@ -0,0 +1,75 @@ +QA output created by 314 + +=== Testing compressed rebase format compatibility === + +Formatting 'TEST_DIR/t.IMGFMT', fmt=raw size=50331648 +qemu-img: Compression not supported for this file format + +=== Testing rebase with compression onto no backing file === + +Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=50331648 +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=50331648 backing_file=TEST_DIR/t.IMGFMT.base backing_fmt=IMGFMT +wrote 25165824/25165824 bytes at offset 0 +24 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 25165824/25165824 bytes at offset 25165824 +24 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 25165824/25165824 bytes at offset 0 +24 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 25165824/25165824 bytes at offset 25165824 +24 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +No errors were found on the image. +768/768 = 100.00% allocated, 100.00% fragmented, 100.00% compressed clusters +Image end offset: 458752 + +=== Testing rebase with compression with raw backing file === + +Formatting 'TEST_DIR/t.IMGFMT.base', fmt=raw size=25165824 +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=50331648 backing_file=TEST_DIR/t.IMGFMT.base backing_fmt=raw +wrote 25165824/25165824 bytes at offset 0 +24 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 25165824/25165824 bytes at offset 25165824 +24 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 25165824/25165824 bytes at offset 0 +24 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 25165824/25165824 bytes at offset 25165824 +24 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +No errors were found on the image. +768/768 = 100.00% allocated, 100.00% fragmented, 100.00% compressed clusters +Image end offset: 458752 + +=== Testing compressed rebase removing single delta from the chain === + +Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=50331648 +Formatting 'TEST_DIR/t.IMGFMT.itmd', fmt=IMGFMT size=50331648 backing_file=TEST_DIR/t.IMGFMT.base backing_fmt=IMGFMT +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=50331648 backing_file=TEST_DIR/t.IMGFMT.itmd backing_fmt=IMGFMT +wrote 16777216/16777216 bytes at offset 0 +16 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 16777216/16777216 bytes at offset 16777216 +16 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 16777216/16777216 bytes at offset 33554432 +16 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 16777216/16777216 bytes at offset 0 +16 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 16777216/16777216 bytes at offset 16777216 +16 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 16777216/16777216 bytes at offset 33554432 +16 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +No errors were found on the image. +512/768 = 66.67% allocated, 100.00% fragmented, 100.00% compressed clusters +Image end offset: 458752 + +=== Testing compressed rebase with unaligned unmerged data === + +Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=65536 +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=65536 backing_file=TEST_DIR/t.IMGFMT.base backing_fmt=IMGFMT +wrote 32767/32767 bytes at offset 0 +31.999 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 32767/32767 bytes at offset 0 +31.999 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +read 32769/32769 bytes at offset 32767 +32.001 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +No errors were found on the image. +1/1 = 100.00% allocated, 100.00% fragmented, 100.00% compressed clusters +Image end offset: 393216 + +*** done