November 2007 - Wine-devel - list.winehq.org

Re: advapi32/service.c -- remove untriggerable check
by Uwe Bonnes 18 Nov '07

18 Nov '07

>>>>> "Gerald" == Gerald Pfeifer <gerald(a)pfeifer.com> writes: Gerald> n is of type DWORD which is unsigned, so n < 0 always will Gerald> evaluate to false. Gerald> Gerald Gerald> ChangeLog: Remove untriggerable check. Gerald> Index: dlls/advapi32/service.c Gerald> =================================================================== Gerald> RCS file: /home/wine/wine/dlls/advapi32/service.c,v retrieving Gerald> revision 1.160 diff -u -3 -p -r1.160 service.c --- Gerald> dlls/advapi32/service.c 15 Oct 2007 16:29:59 -0000 1.160 +++ Gerald> dlls/advapi32/service.c 18 Nov 2007 06:01:28 -0000 @@ -2107,9 Gerald> +2107,6 @@ QueryServiceConfigW( SC_HANDLE hService, n -= Gerald> sizeof(WCHAR); } Gerald> - if( n < 0 ) - ERR("Buffer overflow!\n"); - TRACE("Image path = Gerald> %s\n", debugstr_w(lpServiceConfig->lpBinaryPathName) ); Gerald> TRACE("Group = %s\n", Gerald> debugstr_w(lpServiceConfig->lpLoadOrderGroup) ); Gerald> TRACE("Dependencies = %s\n", Gerald> debugstr_w(lpServiceConfig->lpDependencies) ); Is dropping the check the right answer? Shouldn't the check test for high values like > 0xff00 and report a possible problem? -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

1 0

Valgrind results for Nov 12 & 13
by Dan Kegel 18 Nov '07

18 Nov '07

It seems that valgrind doesn't like Ubuntu 7.10; I had to drop back to my Feisty system to generate good valgrind stack dumps. And the tests didn't hang once! Results for the last two days are at http://kegel.com/wine/valgrind/20071112/ http://kegel.com/wine/valgrind/20071113/ The logs are trimmed a bit more to make diffs less noisy. Diff now shows clearly that there were significant improvements in three tests between those two runs: ole32/compobj, ole32/marshal, and user32/dde Errors still remain in all three tests, though. - Dan

5 7

Re: bits: Added framework for testing the BackgroundCopyJob interface (25/27)
by Detlef Riekenberg 18 Nov '07

18 Nov '07

On Fr, 2007-11-16 at 16:21 -0800, Roy Shea wrote: > +/* Generic test setup */ > +static BOOL setup(void) unused code > +/* Generic test cleanup */ > +static void teardown(void) unused code +START_TEST(job) > +{ > + CoInitialize(NULL); > + CoUninitialize(); > +} Do you want to call "setup()" and "teardown()" here? -- By by ... Detlef

1 0

Re: bits: INF and resource file for qmgr registration (8/27)
by Detlef Riekenberg 18 Nov '07

18 Nov '07

On Fr, 2007-11-16 at 16:16 -0800, Roy Shea wrote: > dlls/qmgr/qmgr.inf | 32 ++++++++++++++++++++++++++++++++ > dlls/qmgr/rsrc.rc | 21 +++++++++++++++++++++ > 2 files changed, 53 insertions(+), 0 deletions(-) The change for Makefile.in to add rsrc.rc is missing here. (found in the next patch) -- By by ... Detlef

1 0

Re: svchost: Added base implementation of svchost (1/27)
by Robert Shearman 18 Nov '07

18 Nov '07

Roy Shea wrote: > +TOPSRCDIR = @top_srcdir@ > +TOPOBJDIR = ../.. > +SRCDIR = @srcdir@ > +VPATH = @srcdir@ > +MODULE = svchost.exe > +APPMODE = -municode > +IMPORTS = advapi32 kernel32 ole32 > I can't see any reason why svchost needs to import ole32. -- Rob Shearman

1 0

Hi, and a regression in bug 10486
by Paweł Różański 17 Nov '07

17 Nov '07

Hi I have found a regression in wine. a strange one. What is a policy about posting such regressions, here on malling list? (Or is regression is worth enough to post it here?) I don't play that particular game on wine and it's "minor" bug - but i have found (via bitsect) that "user32: Cache the window class name on the client side." produce strange effects on directx code. I have faith that regressions are supposed to post here as well. More info in bug #10486 Take care & Hope that helps -- Paweł Różański, a wine user witch occasionally post something on appdb/bugzilla

1 0

Regression testing and cc-ing the author
by Paul Vriens 17 Nov '07

17 Nov '07

Hi, I've opened a bug (which if fixed now btw) and the regression testing page at http://wiki.winehq.org/RegressionTesting said "Also be sure to add the author of the patch to the CC". So this is what I did. Should that line be removed from the wiki or do we generally want the author to be CC-ed? -- Cheers, Paul.

4 4

Re: Removing Windows and Solaris from the Downloads page
by Chris McKillroy 17 Nov '07

17 Nov '07

I compiled wine on Solaris using a .spec provided by the people of http://pkgbuild.sourceforge.net/spec-files-extra/ Maybe contact them ? My version runs quite nice as far as i can see. Cheers

1 0

Re: Add tests for RSA_AES provider in rsaenh
by Reece Dunn 17 Nov '07

17 Nov '07

On 16/11/2007, Vijay Kiran Kamuju <infyquest(a)gmail.com> wrote: > ChangeLog > -------------- > Add tests for RSA_AES provider in rsaenh > > Note: > ------- > Some tests might fail as there is no RSA_AES provider in wine. > These will be fixed once the implementation is in place, which will be > sent next. The recommended way to do this is mark the parts of the tests that fail in Wine with todo_wine. - Reece

3 3

Re: ReadProcessMemory question
by Robert Shearman 17 Nov '07

17 Nov '07

Dave Cartright wrote: > I am trying to read the memory of another process with > ReadProcessmemory and I can only get it to work when I commit the > entire region I want to read with VirtualAllocEx - but as you probably > know when I do this it initializes the entire memory region to 0s. > So my question is how do I get ReadProcessmemory to read from the region?? It sounds to me like you're trying to read the previous contents of areas of memory. However, you don't seem to realise either that virtual memory is not physical memory or that uncommitted virtual memory is not backed by a page of memory in swap or physical memory, so calling ReadProcessMemory on it doesn't make sense as there is no data there to be read! Furthermore, when a page of virtual memory is deallocated then it free to be allocated again by any process in the system, and as you say it is initialised to all 0's as a security measure and you can't get that same page back again. -- Rob Shearman

1 0