On Sun, Jan 15, 2012 at 1:42 PM, devendra.aaru wrote:
>>
>> valgrind lua cairo test4oo.lua?? :)
>>>
>>> Segmentation fault
>>
>> atleast you will get some inputs by running valgrind .
No Valgrind on ARM. And running same test with Valgrind on x86 doesn't
find anything.
I have dmalloc linked in an
On Mon, Jan 16, 2012 at 12:11 AM, devendra.aaru wrote:
>
>
> On Sun, Jan 15, 2012 at 11:39 PM, jonsm...@gmail.com
> wrote:
>
>> First run builds the fontconfig cache.
>> Second run uses the cache and seg faults.
>> I delete it, and program runs again.
>>
>> The gdb data from the segfault is usele
On Sun, Jan 15, 2012 at 11:39 PM, jonsm...@gmail.com wrote:
> First run builds the fontconfig cache.
> Second run uses the cache and seg faults.
> I delete it, and program runs again.
>
> The gdb data from the segfault is useless since the seg fault is
> caused by application data getting corrupte
First run builds the fontconfig cache.
Second run uses the cache and seg faults.
I delete it, and program runs again.
The gdb data from the segfault is useless since the seg fault is
caused by application data getting corrupted by some other source I
have not been able to track down.
I have tried
After more testing updating the font file didn't fix the corruption,
it just moved it to a different spot which makes a different test
program fail. Now I really don't know what is messing up memory but it
is definitely related to fonts.
--
Jon Smirl
jonsm...@gmail.com
___
Index: font/liberation-fonts-ttf/Makefile
===
--- font/liberation-fonts-ttf/Makefile (revision 29747)
+++ font/liberation-fonts-ttf/Makefile (working copy)
@@ -6,14 +6,14 @@
#
include $(TOPDIR)/rules.mk
-PKG_NAME:=liberation-fonts
On 2012-01-15 4:40 PM, jonsm...@gmail.com wrote:
> I finally found the bug. One of the glyphs in the 1.04 liberation font
> package has a bad index. That index causes the higher layers to
> corrupt a few bytes of memory when they cache it. Of course corrupting
> a few bytes of memory may, but not a
I finally found the bug. One of the glyphs in the 1.04 liberation font
package has a bad index. That index causes the higher layers to
corrupt a few bytes of memory when they cache it. Of course corrupting
a few bytes of memory may, but not always, lead to a bunch of
unrelated failures later. In t
I've convinced myself that this memory corruption bug was introduce in
Cairo 10.x and later, not the lua wrapper. This C program hits it too.
#include
#include
#include
#include
#include
#include
#include
typedef struct _cairo_linuxfb_device {
int fb_fd;
char *fb_data;
Minimal failing test program...
local cs = cairo.ImageSurface(CAIRO.FORMAT_RGB24, 320, 240)
local cr = cairo.Context(cs)
cr:select_font_face("Sans", CAIRO.FONT_SLANT_NORMAL, CAIRO.FONT_WEIGHT_BOLD)
cr:set_font_size(8)
cr:show_text("Hs")
cr:select_font_face("Sans", CAIRO.FONT_SLANT_NORMAL, CAIRO.
I have verified now that the luaCairo wrapper works without issue on
x86. So this is something OpenWRT or ARM specific.
--
Jon Smirl
jonsm...@gmail.com
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/mailman/list
I'm working on a Wifi to 802.15.4 router project. 802.15.4 runs IPv6
using 6lowpan/ROLL. The hardware and software are open source, main
site is http://mc1322x.devl.org/. Most users are universities doing
research into sensor networks.
The router is currently under development. It is based on a lp
12 matches
Mail list logo
