Anyone else who wants to try this, I've uploaded the build scripts and my Xcode
project to [email protected]:chris838/dirac-ios.git
>>
>> If I try and run the code in testsuite/encode.c (on the phone) I get errors
>> at the end of orc_memcpy when calling func = c->exec. Occasionally it will
>> get as far as pushing one or two frames before this occurs. Any ideas?
This problem no longer seems to occur when compiling with --with-thread=none.
I'm now also running the code in schro-test.c from the orc repo. I get a bunch
of warnings of the form:
program X failed to compile, reason: no code generation rule for Y on
target arm
> No clue. If you run with 'ORC_CODE=backup' or 'ORC_CODE=emulate' in the
> environment, it will run a lot slower, but tends to be more debuggable.
> If it works with either of those, that is a strong indication of bad
> code generation on Arm/Neon, which is unlikely, but happens.
With either of these switches the warnings in schro-test.c now become:
program test_schro_0 failed to compile, reason: Compilation disabled,
using emulation
In both cases, the code from encode.c pushes 19 frames before crashing - this
time in orc_downsample_vert_u8. It appears that the the pointer
_orc_code_orc_downsample_vert_u8 is set to 0x0, causing an access error when
this is dereferenced in the line func = c->exec;
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Schrodinger-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/schrodinger-devel
