[Bug 908114] Review Request: python-pillow - Python image processing library

[bugzilla at redhat.com]

Product: Fedora
https://bugzilla.redhat.com/show_bug.cgi?id=908114

Toshio Kuratomi <tkuratom at redhat.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |NEW

--- Comment #10 from Toshio Kuratomi <tkuratom at redhat.com> ---
> - Concerning the no-strict-aliasing option, it would be nice to know why the flag was added last November.

Okay -- some spelunking finds this:

$ cd Imaging-1.1.7/libImaging
$ gcc -Wall `python2.7-config --includes` `rpm --eval '%{optflags}'`  -c
Quant.c

Quant.c: In function 'rehash_collide':
Quant.c:154:4: warning: dereferencing type-punned pointer will break
strict-aliasing rules [-Wstrict-aliasing]
Quant.c:154:10: warning: cast to pointer from integer of different size
[-Wint-to-pointer-cast]
Quant.c: In function 'hash_to_list':
Quant.c:247:4: warning: dereferencing type-punned pointer will break
strict-aliasing rules [-Wstrict-aliasing]

Those two functions have been updated in pillow to fix that error:

$ cd python-imaging-Pillow-e09ff61/libImaging
$ gcc -Wall `python2.7-config --includes` `rpm --eval '%{optflags}'`  -c
Quant.c

Quant.c: In function 'rehash_collide':
Quant.c:154:23: warning: cast from pointer to integer of different size
[-Wpointer-to-int-cast]
Quant.c:154:39: warning: cast from pointer to integer of different size
[-Wpointer-to-int-cast]
Quant.c:154:13: warning: cast to pointer from integer of different size
[-Wint-to-pointer-cast]
Quant.c: In function 'hash_to_list':
Quant.c:247:14: warning: cast from pointer to integer of different size
[-Wpointer-to-int-cast]

Note that these new warnings don't occur on x86, just on x86_64.  I'm thinking
they're real bugs... but looking at the source code I'm not sure what needs
changing.... It seems like some of them might be saving the address of a
pointer when it should be saving the value pointed to by the pointer instead. 
It's hard to believe that that kind of bug wouldn't have shown up and been
fixed long ago, though....

-- 
You are receiving this mail because:
You are on the CC list for the bug.
Unsubscribe from this bug https://bugzilla.redhat.com/token.cgi?t=p6Bb6NgMhD&a=cc_unsubscribe