Re: rfc: OMAP2: micro benchmarks

["Andy Shevchenko" <andy.shevchenko%gmail.com@localhost>]

Hello.

Some news and progress in the libMicro on OMAP/H4.

First off, my configuration is the current NetBSD/ARM (yesterday or
little bit older snapshot) with applied IRQ patches from [1] series.
Configuration includes LOCKDEBUG option.

I've apply patches [2-5] to be built under NetBSD-current/ARM .
Some tests are hung up (cascade_<smth>_200), other have problem with
NPROC value (fork_1000, exit_1000), and 'close_tcp' provides lock
collision [6].
'connection' test case provides several errors and in result somehow
is hung up. So, the patch [7] cuts their from the list.

The results for all passed tests are [8-9].

[1] http://mail-index.netbsd.org/port-arm/2008/06/27/msg000276.html
[2] http://koowaldah.org/people/ash/netbsd/micro1/libMicro-0.3.0-netbsd.patch
[3] 
http://koowaldah.org/people/ash/netbsd/micro1/0001-MICRO-inject-NetBSD-s-Makefile-for-libMicro.patch
[4] 
http://koowaldah.org/people/ash/netbsd/micro1/0002-MICRO-don-t-use-arch-binary.patch
[5] 
http://koowaldah.org/people/ash/netbsd/micro1/0003-MICRO-don-t-build-cachetocache-test-case.patch
[6] http://mail-index.netbsd.org/tech-kern/2008/07/11/msg002025.html
[7] 
http://koowaldah.org/people/ash/netbsd/micro1/0004-MICRO-don-t-run-broken-test-cases.patch
[8] http://koowaldah.org/people/ash/netbsd/micro1/benchmarks-200807161421
[9] http://koowaldah.org/people/ash/netbsd/micro1/benchmarks-200807170653

P.S. More detailed description about bugs is coming soon...

-- 
With Best Regards,
Andy Shevchenko