[poppler] invalid option can cause SEGV in pdftocairo
suzuki toshiya
mpsuzuki at hiroshima-u.ac.jp
Tue Nov 26 17:26:39 PST 2013
Oops, I'm quite sorry for that my lazyness made you do it.
Thank you very much.
Regards,
mpsuzuki
Albert Astals Cid wrote (2013/11/27 8:06):
> El Dijous, 21 de novembre de 2013, a les 08:51:17, suzuki toshiya va escriure:
>> Albert Astals Cid wrote (2013/11/21 4:46):
>>> "Wrong page range is given: the last page {0:d} must be after the first
>>> page {1:d}."
>> Oops! You're right. Sorry for my misunderstanding the situation.
>>
>>> is not correct, they can also be the same page, what about
>>>
>>> "Wrong page range given: the first page can not be after the last page"
>> Exactly. I will rework.
>
> I just pushed it.
>
> Cheers,
> Albert
>
>> Regards,
>> mpsuzuki
>>
>>> ?
>>>
>>> Cheers,
>>>
>>> Albert
>>>> Please comment.
>>>>
>>>> Regards,
>>>> mpsuzuki
>>>>
>>>> suzuki toshiya wrote:
>>>>> Albert Astals Cid wrote:
>>>>>>> If endDocument() is invoked without beginDocument(), uninitialized
>>>>>>> surface and output_file are referred, therefore, SEGV is caused.
>>>>>>>
>>>>>>> The easiest fix would be NULL-initialization (not needed in C++?) of
>>>>>>> surface & output_file, and checking NULL before referring them in
>>>>>>> endDocument(). Patch is attached.
>>>>>> I'd prefer an if check for lastPage < firstPage that tells the user to
>>>>>> not
>>>>>> give the program bad parameters. Can you work on that?
>>>>> OK, I will do so. Although SEGV does not occur, other utilities
>>>>> (e.g. pdftotext, pdftoppm, pdftops etc) do not check such. I will
>>>>> make a patchset for all pdftoXXX utilities.
>>>>>
>>>>>> Cheers.
>>>>>>
>>>>>> Albert
>>>>>>> Regards,
>>>>>>> mpsuzuki
>>>>> _______________________________________________
>>>>> poppler mailing list
>>>>> poppler at lists.freedesktop.org
>>>>> http://lists.freedesktop.org/mailman/listinfo/poppler
>> _______________________________________________
>> poppler mailing list
>> poppler at lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/poppler
>
More information about the poppler
mailing list