[libvirt] [PATCH v3 1/3] util: Introduce virDevMapperGetTargets

Peter Krempa pkrempa at redhat.com
Wed Apr 4 15:13:19 UTC 2018


On Tue, Apr 03, 2018 at 11:03:25 +0200, Michal Privoznik wrote:
> This helper fetches dependencies for given device mapper target.
> 
> At the same time, we need to provide a dummy log function because
> by default libdevmapper prints out error messages to stderr which
> we need to suppress.
> 
> Signed-off-by: Michal Privoznik <mprivozn at redhat.com>
> ---
>  src/libvirt_private.syms |   4 +
>  src/util/Makefile.inc.am |   2 +
>  src/util/virdevmapper.c  | 200 +++++++++++++++++++++++++++++++++++++++++++++++
>  src/util/virdevmapper.h  |  31 ++++++++
>  4 files changed, 237 insertions(+)
>  create mode 100644 src/util/virdevmapper.c
>  create mode 100644 src/util/virdevmapper.h

[...]

> diff --git a/src/util/virdevmapper.c b/src/util/virdevmapper.c
> new file mode 100644
> index 0000000000..9491848f48
> --- /dev/null
> +++ b/src/util/virdevmapper.c
> @@ -0,0 +1,200 @@
> +/*
> + * virdevmapper.c: Functions for handling devmapper

device mapper

> + *
> + * Copyright (C) 2018 Red Hat, Inc.
> + *
> + * This library is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2.1 of the License, or (at your option) any later version.
> + *
> + * This library 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
> + * Lesser General Public License for more details.
> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with this library.  If not, see
> + * <http://www.gnu.org/licenses/>.
> + *
> + * Authors:
> + *     Michal Privoznik <mprivozn at redhat.com>
> + */
> +
> +#include <config.h>

[...]

> +static int
> +virDevMapperGetTargetsImpl(const char *path,
> +                           char ***devPaths_ret,
> +                           unsigned int ttl)
> +{
> +    struct dm_task *dmt = NULL;
> +    struct dm_deps *deps;
> +    struct dm_info info;
> +    char **devPaths = NULL;
> +    char **recursiveDevPaths = NULL;
> +    size_t i;
> +    int ret = -1;
> +
> +    *devPaths_ret = NULL;
> +
> +    if (virDevMapperInitialize() < 0)
> +        return ret;
> +
> +    if (!ttl) {

if (ttl == 0)

> +        errno = ELOOP;
> +        return ret;
> +    }
> +
> +    if (!(dmt = dm_task_create(DM_DEVICE_DEPS)))
> +        return ret;
> +
> +    if (!dm_task_set_name(dmt, path)) {
> +        if (errno == ENOENT) {
> +            /* It's okay, @path is not managed by devmapper =>
> +             * not a devmapper device. */
> +            ret = 0;
> +        }
> +        goto cleanup;
> +    }
> +
> +    dm_task_no_open_count(dmt);
> +
> +    if (!dm_task_run(dmt))
> +        goto cleanup;
> +
> +    if (!dm_task_get_info(dmt, &info))
> +        goto cleanup;
> +
> +    if (!info.exists) {
> +        ret = 0;
> +        goto cleanup;
> +    }
> +
> +    if (!(deps = dm_task_get_deps(dmt)))
> +        goto cleanup;
> +
> +    if (VIR_ALLOC_N_QUIET(devPaths, deps->count + 1) < 0)
> +        goto cleanup;
> +
> +    for (i = 0; i < deps->count; i++) {
> +        if (virAsprintfQuiet(&devPaths[i], "/dev/block/%u:%u",
> +                             major(deps->device[i]),
> +                             minor(deps->device[i])) < 0)
> +            goto cleanup;
> +    }
> +
> +    recursiveDevPaths = NULL;
> +    for (i = 0; i < deps->count; i++) {
> +        char **tmpPaths;
> +
> +        if (virDevMapperGetTargetsImpl(devPaths[i], &tmpPaths, ttl - 1) < 0)
> +            goto cleanup;
> +
> +        if (tmpPaths &&
> +            virStringListMerge(&recursiveDevPaths, &tmpPaths) < 0) {

So these report libvirt errors, but don't mess them up. And only in case
of OOM, so this is fine.

> +            virStringListFree(tmpPaths);
> +            goto cleanup;
> +        }
> +    }
> +
> +    if (virStringListMerge(&devPaths, &recursiveDevPaths) < 0)
> +        goto cleanup;
> +
> +    VIR_STEAL_PTR(*devPaths_ret, devPaths);
> +    ret = 0;
> + cleanup:
> +    virStringListFree(recursiveDevPaths);
> +    virStringListFree(devPaths);
> +    dm_task_destroy(dmt);
> +    return ret;
> +}
> +
> +
> +/**
> + * virDevMapperGetTargets:
> + * @path: devmapper target
> + * @devPaths: returned string list of devices
> + *
> + * For given @path figure out its targets, and store them in
> + * @devPaths array. Note, @devPaths is a string list so it's NULL
> + * terminated.
> + *
> + * If @path is not a devmapper device, @devPaths is set to NULL and
> + * success is returned.
> + *
> + * If @path consists of yet another devmapper targets these are
> + * consulted recursively.
> + *
> + * If we don't have permissions to talk to kernel, -1 is returned
> + * and errno is set to EBADF.
> + *
> + * Returns 0 on success,
> + *        -1 otherwise (with errno set, no libvirt error is
> + *        reported)
> + */
> +int
> +virDevMapperGetTargets(const char *path,
> +                       char ***devPaths)
> +{
> +    const unsigned int ttl = 4;

So even trivial setups (dmcrypt + LVM) would exhaust half of your limit
for the device mapper block devices, and would still recurse one more
level to check the backing block device. I think we need some more
headroom here. I'd go with 16 or 32.

> +
> +    /* Arbitrary limit on recursion level. A devmapper target can
> +     * consist of devices or yet another targets. If that's the
> +     * case, we have to stop recursion somewhere. My dice chose
> +     * four. */

Drop the last sentence of this comment.

> +
> +    return virDevMapperGetTargetsImpl(path, devPaths, ttl);
> +}
> +

ACK if you fix the condition syntax, raise the recursion level
significantly, and tweak the comment.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/libvir-list/attachments/20180404/13884311/attachment-0001.sig>


More information about the libvir-list mailing list