Re: Quartz regression in release 1.12

cu <[email protected]>
Newsgroups gmane.comp.lib.cairo
Message-ID <[email protected]>
I am able to reproduce both bugs on the MacOS version of Cairo built
from the branch.

Font bug is now isolated, attached please see the source that triggers
it. When executed, the surface status shows as 13 ("type mismatch") and
png is not saved. For comparison, the same test works fine with cairo
1.9.8 - surface status is 0 and png is properly saved.

Note that I am building Cairo with the following options:

./configure  --disable-xcb --disable-xcb-shm --disable-script
--disable-win32 --disable-win32-font --disable-beos --disable-os2
--disable-glitz --disable-directfb --disable-gl --enable-pthread
--disable-svg --disable-xml --disable-test-surfaces --disable-script
--enable-interpreter=no --disable-full-testing --enable-ft
--disable-shared --disable-xlib --disable-xlib-xrender --enable-pdf
--disable-ps --disable-fc --enable-quartz --disable-quartz-font
--disable-quartz-image

In particular please note that Quartz surface is turned enabled, Quartz
fonts are disabled and Freetype fonts are enabled.

I am still working on trying to isolate the context memory corruption
issue, but it is at least consistent (i.e. switching from image surface
to Quartz surface immediately triggers memory corruption when attempting
to render to a surface in a background thread). Will try to come up with
a test case shortly.




Andrea Canciani wrote:
> On Wed, Apr 11, 2012 at 8:11 PM, cu <[email protected]> wrote:
>   
>> Apologize for "self-bumping" this, but is there anyone that may be able
>> to take a look at Quartz issues?
>>     
>
> I might be able to do that, but it looks like I need some help to
> reproduce them.
> Using the branch from:
> http://cgit.freedesktop.org/~ranma42/cairo/log/?h=wip/map-quartz
> I can run the testsuite getting no crash (there are some failures, but
> I'm already working on them).
>
> Could you please try if this branch fixes the crashes you were getting
> (point #2)?
> In the testsuite ft-text-vertical-layout-type{1,3} look fine, can you
> confirm whether they work in your env?
> Could you provide a small code snippet which triggers your #1 issue?
>
> Thank you for your report, sorry for the late reply
>
> Andrea
>
>   
>> cu wrote:
>>     
>>> Dear all,
>>>
>>>  There appear to be a few fairly severe bugs in Quartz surface when used on
>>> iOS/ARM in 1.12 release. These issues do not exist in version 1.9.8.
>>>
>>>       
>> --
>> cairo mailing list
>> [email protected]
>> http://lists.cairographics.org/mailman/listinfo/cairo
>>     
> --
> cairo mailing list
> [email protected]
> http://lists.cairographics.org/mailman/listinfo/cairo
>

--
cairo mailing list
[email protected]
http://lists.cairographics.org/mailman/listinfo/cairo
bug.c (text/plain, 568 B)
#include <cairo.h>
#include <cairo-quartz.h>
#include <stdio.h>

int main()
{
   cairo_surface_t   *s1;
   cairo_t *cr;

   s1 = cairo_quartz_surface_create(CAIRO_FORMAT_ARGB32, 1000, 600);
   cr = cairo_create(s1);

   cairo_set_source_rgb(cr, 1, 0, 0);
   cairo_paint(cr);

   cairo_set_font_size(cr, 16);

   cairo_move_to(cr, 100, 100);
   cairo_set_source_rgb(cr, 0, 0, 0);
   cairo_show_text(cr, "TEST");
   
   cairo_destroy(cr);
   cairo_surface_write_to_png(s1, "s1.png");
   printf("Status: %d\n", cairo_surface_status(s1));
   cairo_surface_destroy(s1);
}
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.