On 09/13/2010 12:31 PM, Frank Mehnert wrote:
On Monday 13 September 2010, walt wrote:
On 09/13/2010 05:49 AM, Frank Mehnert wrote:
On Sunday 12 September 2010, walt wrote:
I'm using the latest OSE.svn, and I'm seeing Windows 7 guests running at
one-third the usual speed (linux host).
We also had a check of the time which is used for a Windows 7
installation and were not able to observe any differences between 3.2.8
and current trunk. We tested this on an Intel Xeon E5345 (2 x 4 cores) on
Windows Server 2008.
I found the guilty commit:
------------------------------------------------------------------------
r32410 | vboxsync | 2010-09-10 07:19:33 -0700 (Fri, 10 Sep 2010) | 2 lines
Got rid of unnecessary dirty page flush in InvalidatePage. (NOTE: high risk
change)
------------------------------------------------------------------------
Thanks for this information! Does your host have an Intel CPU or an AMD CPU?
AMD Athlon(tm) 64 X2 Dual Core Processor 4200+
I've found an obvious bug in git-svn during this whole process, and I'd like
to file a helpful bug report. If you can spare a minute or two to give me
your educated guess about it, I'd be very grateful.
The clone I made of your svn repo using git-svn won't compile, and to me the
error message doesn't make any sense:
kBuild: Compiling RuntimeR3 -
/U/wa1ter/src/vbox.git/src/VBox/Runtime/common/misc/buildconfig.cpp
/U/wa1ter/src/vbox.git/src/VBox/Runtime/common/misc/buildconfig.cpp: In
function ‘uint32_t RTBldCfgRevision()’:
/U/wa1ter/src/vbox.git/src/VBox/Runtime/common/misc/buildconfig.cpp:39: error:
‘ev’ was not declared in this scope
The source file is identical between your svn repo and my git-svn repo, so it
seems git-svn has messed up some other file(s) somewhere.
Any wild guesses about where I should look in my git-svn repo to find the real
source of the breakage?
Many thanks for your help!
_______________________________________________
vbox-dev mailing list
[email protected]
http://vbox.innotek.de/mailman/listinfo/vbox-dev
