<AWT Dev> [9] Review request for 8077982: GIFLIB upgrade

Alexander Zvegintsev alexander.zvegintsev at oracle.com
Wed Apr 29 09:10:25 UTC 2015


Hi Anton,

this change is intentional:
http://sourceforge.net/p/giflib/code/ci/dfc5b4de5b1cc619d90afc8d2731d31145897aee/tree/lib/dgif_lib.c?diff=63352c4fa592db103f7300790a3383d638dd1edc

Thanks,
Alexander.

On 29/04/15 10:49, Anton V. Tarasov wrote:
> Hi Alexander,
>
> I've just checked a diff b/w the original gif lib sources and your 
> updated sources.
>
> The only diff (except the bool and headers changes) is this:
>
> dgif_lib.c
>
> +     /* Sanity check for corrupted file */
> +     if (GifFile->ImageCount == 0) {
> +         GifFile->Error = D_GIF_ERR_NO_IMAG_DSCR;
> +         return(GIF_ERROR);
> +     }
>
> Is this an intentional insert?
>
> Regards,
> Anton.
>
>
> On 17.04.2015 21:10, Alexander Zvegintsev wrote:
>> Hello,
>>
>> please review the fix
>> http://cr.openjdk.java.net/~azvegint/jdk/9/8077982/00/
>> for the issue
>> https://bugs.openjdk.java.net/browse/JDK-8077982
>>
>> This fix is GIFLIB upgrade to the latest version 5.1.1 [1]
>>
>> webrev against GIFLIB 5.1.1 available here [2]
>>
>> Please see some explanations below:
>>
>> GIFLIB 5.0.0+ supports interlaced images properly,
>> so we don't need to handle it in splashscreen_gif.c anymore (pass = 
>> 4;  npass = 5).
>>
>> We compile JDK with Microsoft's compilers on Windows, there is no 
>> unistd.h[3] (but Cygwin/MinGW has).
>>
>> stdbool.h also isn't supported [4]. However I got a lot of strange 
>> build errors on Solaris with included stdbool.h and -std=gnu99.
>> So it was replaced with own definition of bool.
>>
>>
>> [1] 
>> http://sourceforge.net/projects/giflib/files/giflib-5.1.1.tar.gz/download
>> [2] http://cr.openjdk.java.net/~azvegint/jdk/9/8077982/giflib/00/
>> [3] 
>> http://stackoverflow.com/questions/341817/is-there-a-replacement-for-unistd-h-for-windows-visual-c/1759731#1759731
>> [4] 
>> http://stackoverflow.com/questions/8548521/trying-to-use-include-stdbool-h-in-vs-2010/8549206#8549206
>>
>



More information about the awt-dev mailing list