From: Wenchao Xia <xiawenc@linux.vnet.ibm.com>
To: Stefan Hajnoczi <stefanha@gmail.com>
Cc: kwolf@redhat.com, pbonzini@redhat.com, qemu-devel@nongnu.org,
dietmar@proxmox.com
Subject: Re: [Qemu-devel] [PATCH V3 5/5] block: make all steps in qmp_transaction() as callback
Date: Thu, 25 Apr 2013 16:13:22 +0800 [thread overview]
Message-ID: <5178E5A2.4010602@linux.vnet.ibm.com> (raw)
In-Reply-To: <20130424092532.GA24635@stefanha-thinkpad.redhat.com>
于 2013-4-24 17:25, Stefan Hajnoczi 写道:
> On Fri, Apr 19, 2013 at 08:57:10AM +0800, Wenchao Xia wrote:
>> diff --git a/blockdev.c b/blockdev.c
>> index 051be98..b336794 100644
>> --- a/blockdev.c
>> +++ b/blockdev.c
>> @@ -779,14 +779,41 @@ void qmp_blockdev_snapshot_sync(const char *device, const char *snapshot_file,
>>
>>
>> /* New and old BlockDriverState structs for group snapshots */
>> -typedef struct BlkTransactionStates {
>> +
>> +typedef struct BlkTransactionStates BlkTransactionStates;
>> +
>> +/* Only prepare() may fail. In a single transaction, only one of commit() or
>> + rollback() will be called. */
>
> Please document that clean() is always called - after either commit() or
> rollback().
>
>> +const BdrvActionOps external_snapshot_ops = {
>
> static
>
>> @@ -909,32 +950,36 @@ void qmp_transaction(BlockdevActionList *dev_list, Error **errp)
>> /* We don't do anything in this loop that commits us to the snapshot */
>> while (NULL != dev_entry) {
>> BlockdevAction *dev_info = NULL;
>> + ExternalSnapshotStates *ext;
>>
>> dev_info = dev_entry->value;
>> dev_entry = dev_entry->next;
>>
>> - states = g_malloc0(sizeof(BlkTransactionStates));
>> - QSIMPLEQ_INSERT_TAIL(&snap_bdrv_states, states, entry);
>> -
>> switch (dev_info->kind) {
>> case BLOCKDEV_ACTION_KIND_BLOCKDEV_SNAPSHOT_SYNC:
>> - external_snapshot_prepare(dev_info, states, errp);
>> - if (error_is_set(&local_err)) {
>> - error_propagate(errp, local_err);
>> - goto delete_and_fail;
>> - }
>> + ext = g_malloc0(sizeof(ExternalSnapshotStates));
>> + states = &ext->common;
>> + states->ops = &external_snapshot_ops;
>> break;
>> default:
>> abort();
>> }
>
> Code duplication can be avoided like this:
>
> typedef struct BdrvActionOps {
> /* Size of state struct, in bytes */
> size_t instance_size;
> /* Prepare the work, must NOT be NULL. */
> void (*prepare)(BlkTransactionStates *common, Error **errp);
> /* Commit the changes, must NOT be NULL. */
> void (*commit)(BlkTransactionStates *common);
> /* Rollback the changes on fail, can be NULL. */
> void (*rollback)(BlkTransactionStates *common);
> /* Clean up resource in the end, can be NULL. */
> void (*clean)(BlkTransactionStates *common);
> } BdrvActionOps;
>
> static const BdrvActionOps actions[] = {
> [BLOCKDEV_ACTION_KIND_BLOCKDEV_SNAPSHOT_SYNC] = {
> .instance_size = sizeof(ExternalSnapshotStates),
> .prepare = external_snapshot_prepare,
> .commit = external_snapshot_commit,
> .rollback = external_snapshot_rollback,
> },
> };
>
> Then the state struct is allocated as follows:
>
> assert(dev_info->kind < ARRAY_SIZE(actions));
> const BdrvActionOps *ops = &actions[dev_info->kind];
> states = g_malloc0(ops->instance_size);
> states->ops = ops;
>
> No switch statement is necessary and the states setup doesn't need to be
> duplicated when new actions are added.
>
It is better, will use it, thanks.
--
Best Regards
Wenchao Xia
prev parent reply other threads:[~2013-04-25 8:14 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-04-19 0:57 [Qemu-devel] [PATCH V3 0/5] block: make qmp_transaction extendable Wenchao Xia
2013-04-19 0:57 ` [Qemu-devel] [PATCH V3 1/5] block: package preparation code in qmp_transaction() Wenchao Xia
2013-04-19 0:57 ` [Qemu-devel] [PATCH V3 2/5] block: move input parsing " Wenchao Xia
2013-04-19 0:57 ` [Qemu-devel] [PATCH V3 3/5] block: package committing " Wenchao Xia
2013-04-19 0:57 ` [Qemu-devel] [PATCH V3 4/5] block: package rollback " Wenchao Xia
2013-04-19 0:57 ` [Qemu-devel] [PATCH V3 5/5] block: make all steps in qmp_transaction() as callback Wenchao Xia
2013-04-24 9:25 ` Stefan Hajnoczi
2013-04-25 8:13 ` Wenchao Xia [this message]
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=5178E5A2.4010602@linux.vnet.ibm.com \
--to=xiawenc@linux.vnet.ibm.com \
--cc=dietmar@proxmox.com \
--cc=kwolf@redhat.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@gmail.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.