[PATCH] XCore target: use TEST_TARGET_FLAGS to override target's memory size.

Daniel Dunbar daniel at zuster.org
Tue Mar 11 16:04:52 PDT 2014


It's annoying to have the same fragment in a bunch of places, but I
understand the problem.

Would it be possible to make a new variable for this (XCORE_MEMORY_SIZE or
something), and then put all the extra shared bits of the fragment in a
common location?

 - Daniel



On Tue, Mar 11, 2014 at 12:31 PM, robert lytton <robert at xmos.com> wrote:

> Hi ddunbar,
>
> The tests build XCore-image's with a default memory map of 32MB - passed
> in the target flags.
> It should be noted that the XCore emulator's speed is affected by the size
> of memory.
> Also the amount of host memory required is several times the XCore-image
> memory size.
> Thus running several large xcore-images in parallel may result in
> thrashing!
>
> The RUNTIMELIMIT=7200 is passed in on the commandline.
> Make sure only tests that need longer over-write it.
>
>
> http://llvm-reviews.chandlerc.com/D3048
>
> Files:
>   MultiSource/Applications/JM/ldecod/Makefile
>   MultiSource/Applications/lua/Makefile
>   MultiSource/Applications/minisat/Makefile
>   MultiSource/Applications/spiff/Makefile
>   MultiSource/Benchmarks/ASCI_Purple/SMG2000/Makefile
>   MultiSource/Benchmarks/Bullet/Makefile
>   MultiSource/Benchmarks/Fhourstones-3.1/Makefile
>   MultiSource/Benchmarks/NPB-serial/is/Makefile
>   MultiSource/Benchmarks/Olden/health/Makefile
>   MultiSource/Benchmarks/Olden/perimeter/Makefile
>   MultiSource/Benchmarks/Olden/treeadd/Makefile
>   MultiSource/Benchmarks/Olden/tsp/Makefile
>   MultiSource/Benchmarks/Olden/voronoi/Makefile
>   MultiSource/Benchmarks/SciMark2-C/Makefile
>   MultiSource/Benchmarks/Trimaran/netbench-crc/Makefile
>   MultiSource/Benchmarks/mafft/Makefile
>   MultiSource/Benchmarks/nbench/Makefile
>   SingleSource/Benchmarks/CoyoteBench/Makefile
>   SingleSource/Benchmarks/Misc-C++/Makefile
>   SingleSource/Benchmarks/Polybench/linear-algebra/kernels/2mm/Makefile
>   SingleSource/Benchmarks/Polybench/linear-algebra/kernels/3mm/Makefile
>   SingleSource/Benchmarks/Polybench/linear-algebra/kernels/atax/Makefile
>   SingleSource/Benchmarks/Polybench/linear-algebra/kernels/bicg/Makefile
>   SingleSource/Benchmarks/Polybench/linear-algebra/kernels/doitgen/Makefile
>   SingleSource/Benchmarks/Polybench/linear-algebra/kernels/gemver/Makefile
>   SingleSource/Benchmarks/Polybench/linear-algebra/kernels/gesummv/Makefile
>   SingleSource/Benchmarks/Polybench/linear-algebra/kernels/mvt/Makefile
>   SingleSource/Benchmarks/Polybench/linear-algebra/kernels/trisolv/Makefile
>   SingleSource/Benchmarks/Polybench/linear-algebra/solvers/durbin/Makefile
>   SingleSource/Benchmarks/Shootout-C++/Makefile
>   SingleSource/Benchmarks/Shootout/Makefile
>   SingleSource/Benchmarks/SmallPT/Makefile
>   SingleSource/UnitTests/Makefile
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20140311/6d48aca0/attachment.html>


More information about the llvm-commits mailing list