Hi all,
I found now that PdfOutlines.cpp lines 82 and 90 are highly suspect memory
leaks, not anywhere I touched.
Best regards, mabri
> Matthew Brincke <ma...@mailbox.org> hat am 12. März 2017 um 00:48 geschrieben:
>
>
> Hi all,
>
> one addendum: that's what I've come up with up to now,
> I've just seen there are still some indirect memory leaks
> claimed by g++ 5.2 -fsanitize=address in the non-happypath case
> in my tests (just partial-linking to my PdfOutlines version,
> otherwise using the Debian stretch libpodofo0.9.4 due to security policy).
> I don't know if I'll have time to work on those, if I'm not posting
> next week or so, could you please take it up from there?
>
> Best regards, mabri
>
> > Matthew Brincke <ma...@mailbox.org> has written on 12 March 2017 at 00:38:
>
> Hi all,
>
> zyx <z...@litepdf.cz> has written on 2 March 2017 at 07:48:
>
> On Thu, 2017-03-02 at 00:00 +0100, Matthew Brincke wrote:
>
> I'm sorry that have only build-tested it a bit yet, still I'm submitting
> the patch to PdfOutlineItem to correct memory management for review.
>
> Hi,
> you know, such statements sound quite discouraging. If you are not
> willing to test your own patches, then why should do it anyone else? I
> can understand it for a straightforward changes, but not for anything
> more complicated. By the way, as a bonus, writing unit test for the
> change(s) would be more than welcome.
>
> Now the item given to InsertChild() is copied and the pointers are set to
> NULL so that only that bare item is inserted, no deep copying or problems
> with referenced items being reachable/erasable from two different trees.
>
> This doesn't sound correct to me, but I cannot tell yet. The attached
> patch was incomplete, caused a build break at the best.
>
> the attached patch on this e-mail should be complete now. Then I think it's
> correct, I have tested it now (test code also attached).
>
> Please consider review/testing/when appropriate applying to the public
> repository.
>
> No need to mention this, it's somehow understood on its own (at least
> by me, on this list).
> Bye,
> zyx
>
> I may be away from this list for ca. 3 months for finishing my studies with
> a thesis.
>
> Best regards, mabri
>
> ------------------------------------------------------------------------------
> Announcing the Oxford Dictionaries API! The API offers world-renowned
> dictionary content that is easy and intuitive to access. Sign up for an
> account today to start using our lexical data to power your apps and
> projects. Get started today and enter our developer competition.
> http://sdm.link/oxford
> _______________________________________________
> Podofo-users mailing list
> Podofo-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/podofo-users
------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Podofo-users mailing list
Podofo-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/podofo-users
