From: Max Reitz <mreitz@redhat.com>
To: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>,
"qemu-block@nongnu.org" <qemu-block@nongnu.org>
Cc: Kevin Wolf <kwolf@redhat.com>,
"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>
Subject: Re: [Qemu-devel] [PATCH v6 35/42] block: Fix check_to_replace_node()
Date: Thu, 15 Aug 2019 19:01:48 +0200 [thread overview]
Message-ID: <39c41687-e192-e63a-984f-6d60b075df1a@redhat.com> (raw)
In-Reply-To: <92ded75c-caed-c674-f2ef-72b15b4405b4@virtuozzo.com>
[-- Attachment #1.1: Type: text/plain, Size: 10509 bytes --]
On 15.08.19 17:21, Vladimir Sementsov-Ogievskiy wrote:
> 09.08.2019 19:14, Max Reitz wrote:
>> Currently, check_to_replace_node() only allows mirror to replace a node
>> in the chain of the source node, and only if it is the first non-filter
>> node below the source. Well, technically, the idea is that you can
>> exactly replace a quorum child by mirroring from quorum.
>>
>> This has (probably) two reasons:
>> (1) We do not want to create loops.
>> (2) @replaces and @device should have exactly the same content so
>> replacing them does not cause visible data to change.
>>
>> This has two issues:
>> (1) It is overly restrictive. It is completely fine for @replaces to be
>> a filter.
>> (2) It is not restrictive enough. You can create loops with this as
>> follows:
>>
>> $ qemu-img create -f qcow2 /tmp/source.qcow2 64M
>> $ qemu-system-x86_64 -qmp stdio
>> {"execute": "qmp_capabilities"}
>> {"execute": "object-add",
>> "arguments": {"qom-type": "throttle-group", "id": "tg0"}}
>> {"execute": "blockdev-add",
>> "arguments": {
>> "node-name": "source",
>> "driver": "throttle",
>> "throttle-group": "tg0",
>> "file": {
>> "node-name": "filtered",
>> "driver": "qcow2",
>> "file": {
>> "driver": "file",
>> "filename": "/tmp/source.qcow2"
>> } } } }
>> {"execute": "drive-mirror",
>> "arguments": {
>> "job-id": "mirror",
>> "device": "source",
>> "target": "/tmp/target.qcow2",
>> "format": "qcow2",
>> "node-name": "target",
>> "sync" :"none",
>> "replaces": "filtered"
>> } }
>> {"execute": "block-job-complete", "arguments": {"device": "mirror"}}
>>
>> And qemu crashes because of a stack overflow due to the loop being
>> created (target's backing file is source, so when it replaces filtered,
>> it points to itself through source).
>>
>> (blockdev-mirror can be broken similarly.)
>>
>> So let us make the checks for the two conditions above explicit, which
>> makes the whole function exactly as restrictive as it needs to be.
>>
>> Signed-off-by: Max Reitz <mreitz@redhat.com>
>> ---
>> include/block/block.h | 1 +
>> block.c | 83 +++++++++++++++++++++++++++++++++++++++----
>> blockdev.c | 34 ++++++++++++++++--
>> 3 files changed, 110 insertions(+), 8 deletions(-)
>>
>> diff --git a/include/block/block.h b/include/block/block.h
>> index 6ba853fb90..8da706cd89 100644
>> --- a/include/block/block.h
>> +++ b/include/block/block.h
>> @@ -404,6 +404,7 @@ bool bdrv_is_first_non_filter(BlockDriverState *candidate);
>>
>> /* check if a named node can be replaced when doing drive-mirror */
>> BlockDriverState *check_to_replace_node(BlockDriverState *parent_bs,
>> + BlockDriverState *backing_bs,
>> const char *node_name, Error **errp);
>>
>> /* async block I/O */
>> diff --git a/block.c b/block.c
>> index 915b80153c..4858d3e718 100644
>> --- a/block.c
>> +++ b/block.c
>> @@ -6290,7 +6290,59 @@ bool bdrv_is_first_non_filter(BlockDriverState *candidate)
>> return false;
>> }
>>
>> +static bool is_child_of(BlockDriverState *child, BlockDriverState *parent)
>> +{
>> + BdrvChild *c;
>> +
>> + if (!parent) {
>> + return false;
>> + }
>> +
>> + QLIST_FOREACH(c, &parent->children, next) {
>> + if (c->bs == child || is_child_of(child, c->bs)) {
>> + return true;
>> + }
>> + }
>> +
>> + return false;
>> +}
>> +
>> +/*
>> + * Return true if there are only filters in [@top, @base). Note that
>> + * this may include quorum (which bdrv_chain_contains() cannot
>> + * handle).
>
> More presizely: return true if exists chain of filters from top to base or if
> top == base.
>
> I keep in mind backup-top filter:
>
> [backup-top]
> | \target
backup-top can’t be a filter if it has two children with different
contents, though.
(commit-top and mirror-top aren’t filters either.)
That’s why there must be a unique chain [@top, @base).
I should probably not that it will return true if top == base, though, yes.
> |backing -------->[target]
> V /
> [source] <---------/backing
>
>> + */
>> +static bool is_filtered_child(BlockDriverState *top, BlockDriverState *base)
>> +{
>> + BdrvChild *c;
>> +
>> + if (!top) {
>> + return false;
>> + }
>> +
>> + if (top == base) {
>> + return true;
>> + }
>> +
>> + if (!top->drv->is_filter) {
>> + return false;
>> + }
>> +
>> + QLIST_FOREACH(c, &top->children, next) {
>> + if (is_filtered_child(c->bs, base)) {
>> + return true;
>> + }
>> + }
>
> interesting, how much is it better to somehow reuse DFS search written in should_update_child()..
> [just note, don't do it in these series please]
>
>> +
>> + return false;
>> +}
>> +
>> +/*
>> + * @parent_bs is mirror's source BDS, @backing_bs is the BDS which
>> + * will be attached to the target when mirror completes.
>> + */
>> BlockDriverState *check_to_replace_node(BlockDriverState *parent_bs,
>> + BlockDriverState *backing_bs,
>> const char *node_name, Error **errp)
>> {
>> BlockDriverState *to_replace_bs = bdrv_find_node(node_name);
>> @@ -6309,13 +6361,32 @@ BlockDriverState *check_to_replace_node(BlockDriverState *parent_bs,
>> goto out;
>> }
>>
>> - /* We don't want arbitrary node of the BDS chain to be replaced only the top
>> - * most non filter in order to prevent data corruption.
>> - * Another benefit is that this tests exclude backing files which are
>> - * blocked by the backing blockers.
>> + /*
>> + * If to_replace_bs is (recursively) a child of backing_bs,
>> + * replacing it may create a loop. We cannot allow that.
>> */
>> - if (!bdrv_recurse_is_first_non_filter(parent_bs, to_replace_bs)) {
>> - error_setg(errp, "Only top most non filter can be replaced");
>> + if (to_replace_bs == backing_bs || is_child_of(to_replace_bs, backing_bs)) {
>
> first condition is covered by second, so first may be omitted.
It is not. is_child_of() does not return true if child == parent.
>> + error_setg(errp, "Replacing this node would result in a loop");
>> + to_replace_bs = NULL;
>> + goto out;
>> + }
>> +
>> + /*
>> + * Mirror is designed in such a way that when it completes, the
>> + * source BDS is seamlessly replaced.
>
> Not source but to_replace_bs is replaced?
It has originally been designed to replace the source. If it could
replace any arbitrary BDS, all of this would be moot.
>> It is therefore not allowed
>> + * to replace a BDS where this condition would be violated, as that
>> + * would defeat the purpose of mirror and could lead to data
>> + * corruption.
>> + * Therefore, between parent_bs and to_replace_bs there may be
>> + * only filters (and the one on top must be a filter, too), so
>> + * their data always stays in sync and mirror can complete and
>> + * replace to_replace_bs without any possible corruptions.
>> + */
>> + if (!is_filtered_child(parent_bs, to_replace_bs) &&
>> + !is_filtered_child(to_replace_bs, parent_bs))
>> + {
>> + error_setg(errp, "The node to be replaced must be connected to the "
>> + "source through filter nodes only");
>
> "and the one on top must be a filter, too" not mentioned in the error..
Well, unless the source node is the node to be replaced. Hm... This
gets very hard to express. I think I’d prefer to keep this as it is,
even though it is not quite correct, unless you have a better suggestion
of what to report. :-/
>> to_replace_bs = NULL;
>> goto out;
>> }
>> diff --git a/blockdev.c b/blockdev.c
>> index 4e72f6f701..758e0b5431 100644
>> --- a/blockdev.c
>> +++ b/blockdev.c
>> @@ -3887,7 +3887,7 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
>> }
>>
>> if (has_replaces) {
>> - BlockDriverState *to_replace_bs;
>> + BlockDriverState *to_replace_bs, *backing_bs;
>> AioContext *replace_aio_context;
>> int64_t bs_size, replace_size;
>>
>> @@ -3897,7 +3897,37 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
>> return;
>> }
>>
>> - to_replace_bs = check_to_replace_node(bs, replaces, errp);
>> + if (backing_mode == MIRROR_SOURCE_BACKING_CHAIN ||
>> + backing_mode == MIRROR_OPEN_BACKING_CHAIN)
>> + {
>> + /*
>> + * While we do not quite know what OPEN_BACKING_CHAIN
>> + * (used for mode=existing) will yield, it is probably
>> + * best to restrict it exactly like SOURCE_BACKING_CHAIN,
>> + * because that is our best guess.
>> + */
>> + switch (sync) {
>> + case MIRROR_SYNC_MODE_FULL:
>> + backing_bs = NULL;
>> + break;
>> +
>> + case MIRROR_SYNC_MODE_TOP:
>> + backing_bs = bdrv_filtered_cow_bs(bdrv_skip_rw_filters(bs));
>
> why not bdrv_backing_chain_next(bs) like in mirror_start?
Good question. I suppose it should be
bdrv_filtered_cow_bs(bdrv_backing_chain_next(bs)) in mirror_start()?
Because with sync=top, we just want to remove the topmost COW node (and
filters on top), but keep filters behind it.
Max
>> + break;
>> +
>> + case MIRROR_SYNC_MODE_NONE:
>> + backing_bs = bs;
>> + break;
>> +
>> + default:
>> + abort();
>> + }
>> + } else {
>> + assert(backing_mode == MIRROR_LEAVE_BACKING_CHAIN);
>> + backing_bs = bdrv_filtered_cow_bs(bdrv_skip_rw_filters(target));
>> + }
>> +
>> + to_replace_bs = check_to_replace_node(bs, backing_bs, replaces, errp);
>> if (!to_replace_bs) {
>> return;
>> }
>>
>
>
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
next prev parent reply other threads:[~2019-08-15 17:07 UTC|newest]
Thread overview: 136+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-09 16:13 [Qemu-devel] [PATCH v6 00/42] block: Deal with filters Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 01/42] block: Mark commit and mirror as filter drivers Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 02/42] copy-on-read: Support compressed writes Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 03/42] throttle: " Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 04/42] block: Add child access functions Max Reitz
2019-08-09 16:56 ` Eric Blake
2019-09-04 16:16 ` Kevin Wolf
2019-09-09 7:56 ` Max Reitz
2019-09-09 9:36 ` Kevin Wolf
2019-09-09 14:04 ` Max Reitz
2019-09-09 16:13 ` Kevin Wolf
2019-09-10 9:14 ` Max Reitz
2019-09-10 10:47 ` Kevin Wolf
2019-09-10 11:36 ` Max Reitz
2019-09-10 12:48 ` Kevin Wolf
2019-09-10 12:59 ` Max Reitz
2019-09-10 13:10 ` Kevin Wolf
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 05/42] block: Add chain helper functions Max Reitz
2019-08-09 17:01 ` Eric Blake
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 06/42] qcow2: Implement .bdrv_storage_child() Max Reitz
2019-08-09 17:07 ` Eric Blake
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 07/42] block: *filtered_cow_child() for *has_zero_init() Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 08/42] block: bdrv_set_backing_hd() is about bs->backing Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 09/42] block: Include filters when freezing backing chain Max Reitz
2019-08-10 13:32 ` Vladimir Sementsov-Ogievskiy
2019-08-12 12:56 ` Max Reitz
2019-09-05 13:05 ` Kevin Wolf
2019-09-09 8:02 ` Max Reitz
2019-09-09 9:40 ` Kevin Wolf
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 10/42] block: Drop bdrv_is_encrypted() Max Reitz
2019-08-10 13:42 ` Vladimir Sementsov-Ogievskiy
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 11/42] block: Add bdrv_supports_compressed_writes() Max Reitz
2019-09-05 13:11 ` Kevin Wolf
2019-09-09 8:09 ` Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 12/42] block: Use bdrv_filtered_rw* where obvious Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 13/42] block: Use CAFs in block status functions Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 14/42] block: Use CAFs when working with backing chains Max Reitz
2019-08-10 15:19 ` Vladimir Sementsov-Ogievskiy
2019-09-05 14:05 ` Kevin Wolf
2019-09-09 8:25 ` Max Reitz
2019-09-09 9:55 ` Kevin Wolf
2019-09-09 14:08 ` Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 15/42] block: Re-evaluate backing file handling in reopen Max Reitz
2019-08-10 16:05 ` Vladimir Sementsov-Ogievskiy
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 16/42] block: Flush all children in generic code Max Reitz
2019-08-10 15:36 ` Vladimir Sementsov-Ogievskiy
2019-08-12 12:58 ` Max Reitz
2019-09-05 16:24 ` Kevin Wolf
2019-09-09 8:31 ` Max Reitz
2019-09-09 10:01 ` Kevin Wolf
2019-09-09 14:15 ` Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 17/42] block: Use CAFs in bdrv_refresh_limits() Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 18/42] block: Use CAFs in bdrv_refresh_filename() Max Reitz
2019-08-10 16:22 ` Vladimir Sementsov-Ogievskiy
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 19/42] block: Use CAF in bdrv_co_rw_vmstate() Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 20/42] block/snapshot: Fix fallback Max Reitz
2019-08-10 16:34 ` Vladimir Sementsov-Ogievskiy
2019-08-12 13:06 ` Max Reitz
2019-09-10 11:56 ` Kevin Wolf
2019-09-10 12:04 ` Max Reitz
2019-09-10 12:49 ` Kevin Wolf
2019-09-10 13:06 ` Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 21/42] block: Use CAFs for debug breakpoints Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 22/42] block: Fix bdrv_get_allocated_file_size's fallback Max Reitz
2019-08-10 16:41 ` Vladimir Sementsov-Ogievskiy
2019-08-12 13:09 ` Max Reitz
2019-08-12 17:14 ` Vladimir Sementsov-Ogievskiy
2019-08-12 19:15 ` Max Reitz
2019-09-10 14:52 ` Kevin Wolf
2019-09-11 6:20 ` Max Reitz
2019-09-11 6:55 ` Kevin Wolf
2019-09-11 7:37 ` Max Reitz
2019-09-11 8:27 ` Kevin Wolf
2019-09-11 10:00 ` Max Reitz
2019-09-11 10:31 ` Kevin Wolf
2019-09-11 11:00 ` Max Reitz
2019-09-12 10:34 ` Kevin Wolf
2019-11-14 13:11 ` Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 23/42] blockdev: Use CAF in external_snapshot_prepare() Max Reitz
2019-09-10 15:02 ` Kevin Wolf
2019-09-11 6:21 ` Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 24/42] block: Use child access functions for QAPI queries Max Reitz
2019-08-10 16:57 ` Vladimir Sementsov-Ogievskiy
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 25/42] mirror: Deal with filters Max Reitz
2019-08-12 11:09 ` Vladimir Sementsov-Ogievskiy
2019-08-12 13:26 ` Max Reitz
2019-08-14 15:17 ` Vladimir Sementsov-Ogievskiy
2019-08-31 9:57 ` Vladimir Sementsov-Ogievskiy
2019-09-02 14:35 ` Max Reitz
2019-09-03 8:32 ` Vladimir Sementsov-Ogievskiy
2019-09-09 7:41 ` Max Reitz
2019-09-13 12:55 ` Kevin Wolf
2019-09-16 10:26 ` Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 26/42] backup: " Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 27/42] commit: " Max Reitz
2019-08-31 10:44 ` Vladimir Sementsov-Ogievskiy
2019-09-02 14:55 ` Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 28/42] stream: " Max Reitz
2019-08-12 11:55 ` Vladimir Sementsov-Ogievskiy
2019-09-13 14:16 ` Kevin Wolf
2019-09-16 9:52 ` Max Reitz
2019-09-16 14:47 ` Kevin Wolf
2019-12-11 12:52 ` Max Reitz
2019-12-11 15:52 ` Kevin Wolf
2019-12-11 16:12 ` Max Reitz
2019-12-11 16:35 ` Kevin Wolf
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 29/42] nbd: Use CAF when looking for dirty bitmap Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 30/42] qemu-img: Use child access functions Max Reitz
2019-08-12 12:14 ` Vladimir Sementsov-Ogievskiy
2019-08-12 13:28 ` Max Reitz
2019-08-14 16:04 ` Vladimir Sementsov-Ogievskiy
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 31/42] block: Drop backing_bs() Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 32/42] block: Make bdrv_get_cumulative_perm() public Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 33/42] blockdev: Fix active commit choice Max Reitz
2019-08-09 16:13 ` [Qemu-devel] [PATCH v6 34/42] block: Inline bdrv_co_block_status_from_*() Max Reitz
2019-08-09 16:14 ` [Qemu-devel] [PATCH v6 35/42] block: Fix check_to_replace_node() Max Reitz
2019-08-15 15:21 ` Vladimir Sementsov-Ogievskiy
2019-08-15 17:01 ` Max Reitz [this message]
2019-08-16 11:01 ` Vladimir Sementsov-Ogievskiy
2019-08-16 13:30 ` Max Reitz
2019-08-16 14:24 ` Vladimir Sementsov-Ogievskiy
2019-08-09 16:14 ` [Qemu-devel] [PATCH v6 36/42] iotests: Add tests for mirror @replaces loops Max Reitz
2019-08-09 16:14 ` [Qemu-devel] [PATCH v6 37/42] block: Leave BDS.backing_file constant Max Reitz
2019-08-16 16:16 ` Vladimir Sementsov-Ogievskiy
2019-08-09 16:14 ` [Qemu-devel] [PATCH v6 38/42] iotests: Let complete_and_wait() work with commit Max Reitz
2019-08-23 5:59 ` Vladimir Sementsov-Ogievskiy
2019-08-09 16:14 ` [Qemu-devel] [PATCH v6 39/42] iotests: Add filter commit test cases Max Reitz
2019-08-31 11:41 ` Vladimir Sementsov-Ogievskiy
2019-09-02 15:06 ` Max Reitz
2019-08-31 12:35 ` Vladimir Sementsov-Ogievskiy
2019-09-02 15:09 ` Max Reitz
2019-08-09 16:14 ` [Qemu-devel] [PATCH v6 40/42] iotests: Add filter mirror " Max Reitz
2019-08-31 12:35 ` Vladimir Sementsov-Ogievskiy
2019-08-09 16:14 ` [Qemu-devel] [PATCH v6 41/42] iotests: Add test for commit in sub directory Max Reitz
2019-08-09 16:14 ` [Qemu-devel] [PATCH v6 42/42] iotests: Test committing to overridden backing Max Reitz
2019-09-03 9:18 ` Vladimir Sementsov-Ogievskiy
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=39c41687-e192-e63a-984f-6d60b075df1a@redhat.com \
--to=mreitz@redhat.com \
--cc=kwolf@redhat.com \
--cc=qemu-block@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=vsementsov@virtuozzo.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).