[libvirt] [PATCH 4/9] iptables.c: Use virStrerror, not strerror.

[Jim Meyering]

"Daniel P. Berrange" <berrange at redhat.com> wrote:
> On Wed, Feb 04, 2009 at 01:07:49PM +0100, Jim Meyering wrote:
...
>> Here's an incremental diff that also includes additional changes, e.g.,
             ^^^^^^^^^^^^^^^^
>>   - enable the syntax-check for strerror
>>   - change more "out of error" reports to use virReportOOMError
>>   - use QEMUD_ERROR consistently
...
>> diff --git a/Makefile.nonreentrant b/Makefile.nonreentrant
>> index 13fa59d..b567f31 100644
>> --- a/Makefile.nonreentrant
>> +++ b/Makefile.nonreentrant
>> @@ -79,7 +79,7 @@ NON_REENTRANT += setstate
>>  NON_REENTRANT += sgetspent
>>  NON_REENTRANT += srand48
>>  NON_REENTRANT += srandom
>> -# NON_REENTRANT += strerror
>> +NON_REENTRANT += strerror
>>  NON_REENTRANT += strtok
>>  NON_REENTRANT += tmpnam
>>  NON_REENTRANT += ttyname
>> diff --git a/qemud/Makefile.am b/qemud/Makefile.am
>> index 372b931..a0c161a 100644
>> --- a/qemud/Makefile.am
>> +++ b/qemud/Makefile.am
>> @@ -107,7 +107,6 @@ libvirtd_LDADD =					\
>>  if ! WITH_DRIVER_MODULES
>>  if WITH_QEMU
>>  libvirtd_LDADD += ../src/libvirt_driver_qemu.la
>> -libvirtd_LDADD += ../src/libvirt_util.la
>>  endif
>>
>>  if WITH_LXC
>
> Someting seems not right here - we shouldn't need this change,
> since this shouldn't be added in the first place.

That's the *incremental* change, against what you reviewed the
first time.

>> diff --git a/src/network_driver.c b/src/network_driver.c
>> index 8d7340e..4138939 100644
>> --- a/src/network_driver.c
>> +++ b/src/network_driver.c
>> @@ -904,16 +904,17 @@ static int networkStartNetworkDaemon(virConnectPtr conn,
>>   err_delbr2:
>>      networkRemoveIptablesRules(driver, network);
>>
>> - err_delbr1:;
>> -    char ebuf[1024];
>> + err_delbr1:
>>      if (network->def->ipAddress &&
>>          (err = brSetInterfaceUp(driver->brctl, network->def->bridge, 0))) {
>> +        char ebuf[1024];
>>          networkLog(NETWORK_WARN, _("Failed to bring down bridge '%s' : %s\n"),
>>                   network->def->bridge, virStrerror(err, ebuf, sizeof ebuf));
>>      }
>
> This also appears to be changing code which doesn't exist.

Same as above.

> [snip more of the same]
>
> Have you got the actual patch against current CVS, since this doesn't
> appear to be it ?

Yep.  It was attached at the end, after this:

>> Here's the 10-change-set sequence: