On 10/22/2013 10:16 AM, Joe MacDonald wrote:
> [Re: [PATCH 1/2] libselinux: support building on older PCRE libs] On 13.10.21 (Mon 16:27) Stephen Smalley wrote:
>
>> On 10/21/2013 03:16 PM, Joe MacDonald wrote:
>>> Versions of PCRE prior to 8.20 did not have pcre_free_study(). In its
>>> absence, use pcre_free() instead.
>>>
>>> Signed-off-by: Joe MacDonald <joe@xxxxxxxxxxxx>
>>> ---
>>> libselinux/src/label_file.c | 4 ++++
>>> libselinux/utils/sefcontext_compile.c | 4 ++++
>>> 2 files changed, 8 insertions(+)
>>>
>>> diff --git a/libselinux/src/label_file.c b/libselinux/src/label_file.c
>>> index 9b0d6b0..51520cf 100644
>>> --- a/libselinux/src/label_file.c
>>> +++ b/libselinux/src/label_file.c
>>> @@ -561,7 +561,11 @@ static void closef(struct selabel_handle *rec)
>>> free(spec->type_str);
>>> if (spec->regcomp) {
>>> pcre_free(spec->regex);
>>> +#if (PCRE_MAJOR >= 8 && PCRE_MINOR >= 20)
>>> pcre_free_study(spec->sd);
>>> +#else
>>> + pcre_free(spec->sd);
>>
>> We try to avoid #if / #ifdef scattered throughout the code.
>> The preferred approach is to conditionally define pcre_free_study() to
>> pcre_free() if building with the older libpcre.
>
> Sure thing. I'll resubmit. Given that this is present in only two
> files with the only obvious common #include (within the selinux repo
> itself) being label_file.h, I'll put the #define there. Does that seem
> reasonable?
Yes, that looks right to me.
--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@xxxxxxxxxxxxx with
the words "unsubscribe selinux" without quotes as the message.
