Bug#805323: Fwd: Bug #805323: graphite2 ARM HP PA FTBFS

Hi Rene,

I also have an RPi2 here so I've used that to test this round of fixes, this bug should be fixed for good now as I've re-factored the problem code to a more straightforward form.

I've also fixed the CMake policy warnings.

Thanks for the log that was a big help. I noticed in the log that the cmake invocation is being passed:

-DGRAPHITE2_NTRACING:bool=1

Whilst it works I believe that is a typo, and it should be:

-DGRAPHITE2_NTRACING:BOOL=ON

Regards,

Tim

On Fri, Dec 18, 2015 at 3:29 AM, Rene Engelhard <rene@debian.org> wrote:

forwarded 805323 Tim Eves <tim_eves@sil.org>
thanks

Hi Tim,

On Thu, Dec 17, 2015 at 11:37:20AM +0700, Tim Eves wrote:
> The bug has been fixed in commit 594d92c0c0efbd3a87a32260f57e502200c351a7
> "Fix misaligned access on ARM and HP PA processors during collision".
>
> It was a non-32 bit aligned 32bit read, some processors (ARM) are stricter
> about this than others (Intel), which could explain the fact it built fine
> on other BE archs.

Unfortunately the patch doesn't work for me, I still get a build failure on armhf.
(See attached buildlog). gdb'ing core gives

(cowbuilder-unstable)rene@raspberrypi:~/tmp/graphite2-1.3.4/build/tests$ gdb --core=core
GNU gdb (Debian 7.10-1) 7.10
Copyright (C) 2015 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "arm-linux-gnueabihf".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word".
[New LWP 14630]
Core was generated by `/usr/bin/python /data/tmp/graphite2-1.3.4/tests/fnttxtrender --graphite_library'.
Program terminated with signal SIGBUS, Bus error.
#0 0x768e8314 in ?? ()
(gdb) bt
#0 0x768e8314 in ?? ()
#1 0x768e8306 in ?? ()
Backtrace stopped: previous frame identical to this frame (corrupt stack?)
(gdb)

which is not exactly helpful but the same I get without the patch. (and yes, the patch IS applied.)

> We have been regularly testing against armhf on the Qemu Versatile Express
> emulator, under which it turns out this bug is not reproducible. Once we
> tried with real hardware (RPi B+) it showed up.

RPi 2 here.

Regards,

Rene

Tim Eves

NRSI - SIL International