Kalle, Thank you for the hints.
I noticed that when building the lib/kernel package, both clang-3.3/3.2 complained heavily on "-target armv7", saying "unrecognized flag, use -mfloat=soft" instead. I hacked the Makefile a little bit, and use "-target armv7l-unknown-linux-gnueabi" instead. The original warning all go away. However, the produced tests still fail. E.g. czhao@panda0:~/ResearchTools/Compiler/POCL/pocl-0.8/robj32-rc7/examples/example1/.libs$ ./lt-example1 ERROR (0.000000, 0.000000, 0.000000, 0.000000) . (0.000000, 0.000000, 0.000000, 0.000000) = -0.000007 FAIL czhao@panda0:~/ResearchTools/Compiler/POCL/pocl-0.8/robj32-rc7/examples/example1/.libs$ I am debugging now. Any hints/suggestions are highly appreciated. Chuck On 7/19/2013 10:40 AM, Kalle Raiskila wrote: > On 17.07.2013 23:44, Chuck Zhao wrote: >> I did a test build on my PandaBoard (ARM/Ubuntu 12.04 + LLVM/CLANG-3.3 + >> gcc/g++-4.5.3, pocl-0.8-rc7), the build is fine, but "make check" seems >> to have some issues with the basic tests. > Strange. I have been keeping the "expected failures" quite up to date > both on the trunk and 0.8 branches. And the platform I test on is a > pretty vanilla Ubuntu 12.04 on a pandaboard. > > One notable difference is that I have ocl-icd 2.0.2 installed on the > system - but the ICD-less build is tested on other platforms quite > sufficiently, so that should not be the issue. > > Also, Sumit Semwal sent in a few patches to fix some issues on > ARM/Ubuntu 13.04. I'm not sure what the status of latest pocl is on that > platform, but some time since the 0.7, all checks worked as expected, I > believe. (Sumit, got an update on this?) > >> Please let me know when pocl-0.8's code base is "stable/mature" enough, >> so that I can dive in to debug some ARM-specific issues. > Now. Patches welcome. Especially since I have not seen the problems you > posted later in this chain. > > One pointer where to start looking is the target triple passed to the > kernel compiler. This has caused lots of issues in the past. Turns out > (at least) clang 3.2 didn't understand the default triple that Ubuntu > had (something like armv7l-...), and reverted (silently!) to using the > ARMv4 ISA. Which then caused issues with linking. Though this should not > effect the clCreateBuffer (if indeed the failure comes from there). > > > kalle > > ------------------------------------------------------------------------------ > See everything from the browser to the database with AppDynamics > Get end-to-end visibility with application monitoring from AppDynamics > Isolate bottlenecks and diagnose root cause in seconds. > Start your free trial of AppDynamics Pro today! > http://pubads.g.doubleclick.net/gampad/clk?id=48808831&iu=/4140/ostg.clktrk > _______________________________________________ > pocl-devel mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/pocl-devel > ------------------------------------------------------------------------------ See everything from the browser to the database with AppDynamics Get end-to-end visibility with application monitoring from AppDynamics Isolate bottlenecks and diagnose root cause in seconds. Start your free trial of AppDynamics Pro today! http://pubads.g.doubleclick.net/gampad/clk?id=48808831&iu=/4140/ostg.clktrk _______________________________________________ pocl-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/pocl-devel
