[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [libvirt] [PATCHv2 08/16] storage: get entire metadata chain in one call



On 10/16/2012 10:58 AM, Eric Blake wrote:
> On 10/16/2012 08:43 AM, Laine Stump wrote:
>> On 10/13/2012 06:00 PM, Eric Blake wrote:
>>> Previously, no one was using virStorageFileGetMetadata, and for good
>>> reason - it couldn't support root-squash NFS.  Change the signature
>>> and make it useful to future patches, including enhancing the metadata
>>> to recursively track the entire chain.
>>>
>>> +
>>> +    if (VIR_CLOSE(fd) < 0)
>>> +        virReportSystemError(errno, _("could not close file %s"), path);
>> If this isn't fatal to the operation, shouldn't it be a warning instead
>> of an error? (And if it is fatal, it should free the remaining resources
>> and return NULL)
> Later in the series, I convert virDomainDiskDefForeachPath, which had an
> argument that said whether to ignore open failures while recursing
> through a chain.  That implies that there are some callers that care
> about being able to find every file in a chain, while other callers want
> to know as much as possible but don't care once they can't get further.
>  I debated about how best to represent this in the new recursive data
> structure, and ended up with: if meta->backingStoreIsFile is set but
> meta->backingStoreMeta is clear, the recursion failed partway through,
> and it is up to the caller to decide if an aborted recursion is fatal.
>
> I also debated about whether I should have the helper function return an
> int instead of a pointer; returning a pointer or NULL is a boolean
> operation, and does not allow for as much flexibility.  Returning an int
> would let me do things like returning 0 if recursion ended normally,
> returning 1 if recursion ended because a file could not be opened, and
> returning -1 if recursion ended due to OOM or other fatal problem.  If
> you want me to revise along those lines I can, but for now, I think I'll
> just change this to be a VIR_WARN instead of a virReportSystemError().

Either is fine with me.

>
>>> +
>>> +    if (ret->backingStoreIsFile) {
>>> +        if (ret->backingStoreFormat == VIR_STORAGE_FILE_AUTO && !allow_probe)
>>> +            ret->backingStoreFormat = VIR_STORAGE_FILE_RAW;
>>> +        else if (ret->backingStoreFormat == VIR_STORAGE_FILE_AUTO_SAFE)
>>> +            ret->backingStoreFormat = VIR_STORAGE_FILE_AUTO;
>>> +        format = ret->backingStoreFormat;
>> Why are you bothering to set this, instead of just using
>> ret->backingStoreFormat in the args to the following function call?
>> (It's harmless, though, so no big deal)
> I did that solely to fit within 80 columns.
>
>> ACK, assuming acceptable response to my above questions.


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]