RB tree VM regions bench test

[Davide Libenzi]

I ran the mmap validation utest, before and after the RB tree change.

Before:

/ $ ./bin/tests/utest/mmap_stress_test 

<-- BEGIN_USERSPACE_mmap_TESTS -->

test_run_fragment_test: Success in mmap at allocating 20480 pages: addr=0x400000608000..0x400005608000

test_run_fragment_test: Run time: 1517.030000 ms

        PASSED   [test_run_fragment_test](1.519759s)   

<-- END_USERSPACE_mmap_TESTS -->

After:

/ $ ./bin/tests/utest/mmap_stress_test 

<-- BEGIN_USERSPACE_mmap_TESTS -->

test_run_fragment_test: Success in mmap at allocating 20480 pages: addr=0x400000608000..0x400005608000

test_run_fragment_test: Run time: 87.544000 ms

        PASSED   [test_run_fragment_test](0.090364s)   

<-- END_USERSPACE_mmap_TESTS -->

[Kevin Klues]

Was this done on hardware?  If so, what were the machine specs?  I assume this is in reference to the code under review at:

https://groups.google.com/forum/#!topic/akaros/mKmbiQxnKDc

[Davide Libenzi]

Sorry I did not specify. That was under KVM, in my workstation.

I doubt VMM skewed the result, as the number of faults generated by the test (in trying to access unmapped memory to verify VMA handling consistency) do not change.

I can/will repeat the tests on real HW, but I am not expecting large relative differences.

--
You received this message because you are subscribed to the Google Groups "Akaros" group.
To unsubscribe from this group and stop receiving emails from it, send an email to akaros+un...@googlegroups.com.
To post to this group, send email to aka...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.