This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: [PATCH][libstdc++-v3 parallel mode] Fix memory leaks in multiway_merge
- From: Johannes Singler <singler at ira dot uka dot de>
- To: Paolo Carlini <paolo dot carlini at oracle dot com>
- Cc: libstdc++ <libstdc++ at gcc dot gnu dot org>, gcc-patches at gcc dot gnu dot org
- Date: Fri, 11 Sep 2009 16:31:50 +0200
- Subject: Re: [PATCH][libstdc++-v3 parallel mode] Fix memory leaks in multiway_merge
- References: <4AA8C9DF.2040703@ira.uka.de> <4AA9285F.6080203@oracle.com>
Paolo Carlini wrote:
> Johannes Singler wrote:
>> This patch fixes two memory leaks (and non-called destructors) in
>> multiway_merge.
>>
>> Tested x86_64-unknown-linux-gnu: No regressions
>>
>> Please approve for mainline and gcc-4_4-branch.
>>
> Ok, thanks. Out of curiosity, are you using valgrind to chase those
> leaks?
Yes, these leaks were detected by valgrind.
> I'm trusting you both about the issue proper and the fix.
Sorry, I don't understand this phrase grammatically.
>> 2009-09-10 Johannes Singler <singler@ira.uka.de>
>>
>> * include/parallel/multiway_merge.h
>> (multiway_merge_exact_splitting): Deallocate borders correctly.
>> (parallel_multiway_merge): Remove unnecessarily complicated
>> allocation, random access iterators are default-constructible.
>>
>> Deallocate ne_seqs correctly.
>>
> A semicolon instead of the blank line, I guess. Like this, isn't a
> properly formatted ChangeLog entry, for sure ;)
Somehow, this blank line sneaked in. Better like this? Or still add
some colon?
2009-09-11 Johannes Singler <singler@ira.uka.de>
* include/parallel/multiway_merge.h
(multiway_merge_exact_splitting): Deallocate borders correctly.
(parallel_multiway_merge): Remove unnecessarily complicated
allocation, random access iterators are default-constructible.
Deallocate ne_seqs correctly.
-- Johannes