Re: libpng-12.58rc01, 1.4.21rc01, 1.5.29rc01, 1.6.32rc01, and 1.7.0beta89 are available

compiling 1.7.0b89 with visual studio...
generates quite a few warnings

...\libpng-1.7.0b89\pngchunk.h(47): warning C4028: formal parameter 2
different from declaration (compiling source file
...\libpng-1.7.0b89\pngrutil.c)

This is caused because png_inforp (what is used in png_handle_IHDR) is not
exactly the same as png_infop which is used for the function declaration....


(line 627 in png.h)
typedef png_info * PNG_RESTRICT png_inforp;

(line 610 in png.h)
typedef png_info * png_infop;


These can be fixed by adding PNG_RESTRICT macro to png_infop...
typedef png_info * PNG_RESTRICT png_infop;


All of these lines should be updated...
(line 586&587 png.h)
typedef const png_struct * PNG_RESTRICT png_const_structp;
typedef png_struct * PNG_RESTRICT png_structp;

(line 610&611 png.h)
typedef png_info * PNG_RESTRICT png_infop;
typedef const png_info * PNG_RESTRICT png_const_infop;

The others are only used in tests... I don't build those, so I only know by
searching for the symbol.

(612)
typedef png_info * * png_infopp;

-----------------------
...\libpng-1.7.0b89\pngrutil.c(4251): warning C4018: '>=': signed/unsigned
mismatch
                        pass >= PNG_LAST_PASS(width, height))

pass is unsigned... the macro is
(line 3094, png.h)

#define PNG_LAST_PASS(width, height) ((height) > 1 ? 6 : ((width) > 1 ? 5 :
0))

and should be

#define PNG_LAST_PASS(width, height) ((height) > 1 ? 6U : ((width) > 1 ? 5U
: 0))

----
This sort of expression

pixel_block & -pixel_block

is used quite a bit, it generates warning
...\libpng-1.7.0b89\pngwutil.c(859): warning C4146: unary minus operator
applied to unsigned type, result still unsigned

Maybe such a test for (LOWEST_SET_BIT) could be made as a macro so a one
place fix could be made... (though I don't know a good fix)

-----

...libpng-1.7.0b89\pngwutil.c(1564): warning C4334: '<<': result of 32-bit
shift implicitly converted to 64 bits (was 64-bit shift
            while (windowBits < 15 && (1U << windowBits) < test_size)

This sort of warning I really hate.  It's range tested to be under 32 bits,
and there's no way to make visual studio keep it as 32 bit... even applying
like (windowBits & 0xF) on the value.  And maybe it's because it's a 64 bit
build...

-----
The final bit of warnings are

...\libpng-1.7.0b89\pngrtran.c(610): warning C4146: unary minus operator
applied to unsigned type, result still unsigned
...\libpng-1.7.0b89\pngrtran.c(626): warning C4146: unary minus operator
applied to unsigned type, result still unsigned
...\libpng-1.7.0b89\pngrtran.c(642): warning C4146: unary minus operator
applied to unsigned type, result still unsigned
...\libpng-1.7.0b89\pngrtran.c(909): warning C4146: unary minus operator
applied to unsigned type, result still unsigned
...\libpng-1.7.0b89\pngrtran.c(932): warning C4146: unary minus operator
applied to unsigned type, result still unsigned
...\libpng-1.7.0b89\pngrtran.c(966): warning C4146: unary minus operator
applied to unsigned type, result still unsigned
...\libpng-1.7.0b89\pngrtran.c(5781): warning C4146: unary minus operator
applied to unsigned type, result still unsigned

(for example)

unsigned int shift = 7U & -(tc->width << 2)/*overflow ok*/;

no suggestion on a fix for this...

-----
Just reporting; I don't have a preference whether these are fixed or not.









On Fri, Aug 18, 2017 at 7:13 PM, Glenn Randers-Pehrson <[email protected]>
wrote:

> libpng-1..2.58rc01, 1.4.21rc01, 1.5.29rc01, 1.6.32rc01, and 1.70beta89
> are available from
> https://ftp-osl.osuosl.org/pub/libpng/src/beta/
> ftp://ftp-osl.osuosl.org/pub/libpng/src/beta/
> http://libpng.download/src/beta/
> and from
> http://libpng.sf.net
>
> version 1.2.58rc01 [August 19, 2017]
>   Check for 0 return from png_get_rowbytes() and added some (size_t)
> typecasts
>     in contrib/pngminus/*.c to stop some Coverity issues (162705, 162706,
>     and 162707).
>
> version 1.4.21rc01 [August 19, 2017]
>   Check for 0 return from png_get_rowbytes() and added some (size_t)
> typecasts
>     in contrib/pngminus/*.c to stop some Coverity issues (162705, 162706,
>     and 162707).
>
> version 1.5.29rc01 [August 19, 2017]
>   Check for 0 return from png_get_rowbytes() in contrib/pngminus/*.c to
> stop
>     some Coverity issues (162705, 162706, and 162707).
>
> Version 1.6.32rc01 [August 18, 2017]
>   Added a set of "huge_xxxx_chunk.png" files to contrib/testpngs/crashers,
>     one for each known chunk type, with length = 2GB-1.
>   Check for 0 return from png_get_rowbytes() and added some (size_t)
> typecasts
>     in contrib/pngminus/*.c to stop some Coverity issues (162705, 162706,
>     and 162707).
>   Renamed chunks in contrib/testpngs/crashers to avoid having files whose
>     names differ only in case; this causes problems with some platforms
>     (github issue #172).
>
> Version 1.7.0beta89 [August 19, 2017]
>   Check for 0 return from png_get_rowbytes() in contrib/pngminus/*.c to
> stop
>     some Coverity issues (162705, 162706, and 162707).
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> png-mng-implement mailing list
> png-mng-implement-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/png-mng-implement
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot

_______________________________________________
png-mng-implement mailing list
png-mng-implement-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/png-mng-implement