Libguestfs - Jun 2015 - Re: [PATCH 2/3] inspection: Add support for CoreOS

On 02/06/15 17:10, Richard W.M. Jones wrote:
Hello,
> On Fri, May 29, 2015 at 12:24:58PM +0300, Nikos Skalkotos wrote:
>> +  if (collect_coreos_inspection_info (g)) {
>> +    guestfs_int_free_inspect_info (g);
>> +    return NULL;
>> +  }
> Although this is stylistic, I think it's easier to understand if
> you change the if condition to:
>
>   if (collect_coreos_inspection_info (g) == -1) {
>     ...
>
Since we 'll be using safe_realloc, I don't think there is a need for
any of:
 * guestfs_int_merge_fs_inspections()
 * collect_coreos_inspection_info()
to be returning a value at all. So, I'll remove the if check completely.

Nikos

P.S. You have a check like this one on
guestfs_int_check_for_filesystem_on() a few lines above :-)
> Rich.
>
>
>

On Tue, Jun 02, 2015 at 06:18:38PM +0300, Nikos Skalkotos
wrote:
> On 02/06/15 17:10, Richard W.M. Jones wrote:
> Hello,
> 
> > On Fri, May 29, 2015 at 12:24:58PM +0300, Nikos Skalkotos wrote:
> >> +  if (collect_coreos_inspection_info (g)) {
> >> +    guestfs_int_free_inspect_info (g);
> >> +    return NULL;
> >> +  }
> > Although this is stylistic, I think it's easier to understand if
> > you change the if condition to:
> >
> >   if (collect_coreos_inspection_info (g) == -1) {
> >     ...
> >
> Since we 'll be using safe_realloc, I don't think there is a need
for
> any of:
>  * guestfs_int_merge_fs_inspections()
>  * collect_coreos_inspection_info()
> to be returning a value at all. So, I'll remove the if check
completely.
> 
> Nikos
> 
> P.S. You have a check like this one on
> guestfs_int_check_for_filesystem_on() a few lines above :-)
In extend_fses?  That's a bug ..

Rich.

-- 
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming and virtualization blog: http://rwmj.wordpress.com
Fedora Windows cross-compiler. Compile Windows programs, test, and
build Windows installers. Over 100 libraries supported.
http://fedoraproject.org/wiki/MinGW

Sorry, I don't get it. In inspect.c line 67 you have this:

  for (fs = fses; *fs; fs += 2) {
    if (guestfs_int_check_for_filesystem_on (g, *fs)) {
      guestfs_int_free_inspect_info (g);
      return NULL;
    }
  }

I don't see the bug.

On 02/06/15 18:30, Richard W.M. Jones wrote:
> On Tue, Jun 02, 2015 at 06:18:38PM +0300, Nikos Skalkotos wrote:
>> On 02/06/15 17:10, Richard W.M. Jones wrote:
>> Hello,
>>
>>> On Fri, May 29, 2015 at 12:24:58PM +0300, Nikos Skalkotos wrote:
>>>> +  if (collect_coreos_inspection_info (g)) {
>>>> +    guestfs_int_free_inspect_info (g);
>>>> +    return NULL;
>>>> +  }
>>> Although this is stylistic, I think it's easier to understand
if
>>> you change the if condition to:
>>>
>>>   if (collect_coreos_inspection_info (g) == -1) {
>>>     ...
>>>
>> Since we 'll be using safe_realloc, I don't think there is a
need for
>> any of:
>>  * guestfs_int_merge_fs_inspections()
>>  * collect_coreos_inspection_info()
>> to be returning a value at all. So, I'll remove the if check
completely.
>>
>> Nikos
>>
>> P.S. You have a check like this one on
>> guestfs_int_check_for_filesystem_on() a few lines above :-)
> In extend_fses?  That's a bug ..
>
> Rich.
>