[libvirt] [PATCH 8/9] Simplified version of volume wiping based on feedback from the list.

[Laine Stump]

On 03/17/2010 10:49 AM, Laine Stump wrote:
>
>>> +    if (ret == -1) {
>>> +        virReportSystemError(ret,
>>> +                             _("Failed to truncate volume with "
>>> +                               "path '%s' to %ju bytes: '%s'\n"),
>>> +                             vol->target.path, (intmax_t)size,
>>> +                             virStrerror(errno, errbuf, 
>>> sizeof(errbuf)));
> +    ret = ftruncate(fd, size);
>
> Likewise, this should be:
>
>          virReportSystemError(errno,
>                               _("Failed to truncate volume with "
>                              "path '%s' to %ju bytes),
>                              vol->target.path, (intmax_t)size);
>                              virStrerror(errno, errbuf, sizeof(errbuf)));

Oops - that last line is (probably obviously) an artifact of cut-paste 
that I meant to delete.

>
>>> +    fd = open(def->target.path, O_RDWR);
>>> +    if (fd == -1) {
>>> +        VIR_ERROR("Failed to open storage volume with path '%s': 
>>> '%s'",
>>> +                  def->target.path,
>>> +                  virStrerror(errno, errbuf, sizeof(errbuf)));
> +
>
> Not sure why you're using VIR_ERROR() + manually adding virStrerror() 
> - isn't this the same thing as virReportSystemError?

I had meant to mention that I've seen this in at least one other place 
as well. Is there any reason for using VIR_ERROR like this, or is it 
just a historical artifact?