October 2012 - Wine-devel - list.winehq.org

Re: [PATCH 1/4] jscript: Use custom string container instead of BSTR.
by Charles Davis 12 Oct '12

12 Oct '12

On Oct 11, 2012, at 4:16 AM, Jacek Caban wrote: > This patch alone makes SunSpider 0.9 17x faster. Seems to me that something is really wrong with our BSTR implementation if replacing it with a home-grown implementation speeds this up by a factor of 17. Chip

2 1

Re: [1/3] scrrun: Implement IDictionary_Add and Count (try 2)
by Henri Verbeet 12 Oct '12

12 Oct '12

On 9 October 2012 11:26, Alistair Leslie-Hughes <leslie_alistair(a)hotmail.com> wrote: > Changelog: > scrrun: Implement IDictionary_Add and Count > Actually, does it ever make sense to use a linked list at all to implement a dictionary? It seems to me that if you only ever have a handful of entries a simple array would be better suited, while if you have more than that you'd want some kind of search tree or a hash table.

3 2

Fwd: [1/3] scrrun: Implement IDictionary_Add and Count (try 2)
by Per Johansson 12 Oct '12

12 Oct '12

On Fri, Oct 12, 2012 at 8:48 AM, Alistair Leslie-Hughes <leslie_alistair(a)hotmail.com> wrote: > I will have a look at some sort of hash table. > I'd use wine existing tree implementation, unless there's some specific reason to use a hash table. Hash tables are good in theory but with in practice trees are just as effective in my experience (especially AVL trees). Regards, -- Per Johansson

1 0

Re: [PATCH] amstream: Display new ref in AddRef and Release.
by Nikolay Sivov 11 Oct '12

11 Oct '12

On 10/12/2012 00:46, Christian Costa wrote: > --- > dlls/amstream/mediastreamfilter.c | 14 +++++++++----- > 1 file changed, 9 insertions(+), 5 deletions(-) > > diff --git a/dlls/amstream/mediastreamfilter.c b/dlls/amstream/mediastreamfilter.c > index ee9d970..0e56a68 100644 > --- a/dlls/amstream/mediastreamfilter.c > +++ b/dlls/amstream/mediastreamfilter.c > @@ -133,17 +133,21 @@ static HRESULT WINAPI MediaStreamFilterImpl_QueryInterface(IMediaStreamFilter *i > > static ULONG WINAPI MediaStreamFilterImpl_AddRef(IMediaStreamFilter *iface) > { > - return BaseFilterImpl_AddRef((IBaseFilter*)iface); > + ULONG ref = BaseFilterImpl_AddRef((IBaseFilter*)iface); > + > + TRACE("(%p)->(): new ref = %u\n", iface, ref); > + > + return ref; > } You should get rid of casts like that. Also I don't see a reason to print anything but refcount itself life (%p)->(%u). > > static ULONG WINAPI MediaStreamFilterImpl_Release(IMediaStreamFilter *iface) > { > IMediaStreamFilterImpl *This = impl_from_IMediaStreamFilter(iface); > - ULONG refCount = BaseFilterImpl_Release((IBaseFilter*)iface); > + ULONG ref = BaseFilterImpl_Release((IBaseFilter*)iface); > > - TRACE("(%p)->() Release from %d\n", iface, refCount + 1); > + TRACE("(%p)->(): new ref = %u\n", iface, ref); > > - if (!refCount) > + if (!ref) > { > int i; > for (i = 0; i < This->nb_streams; i++) > @@ -154,7 +158,7 @@ static ULONG WINAPI MediaStreamFilterImpl_Release(IMediaStreamFilter *iface) > HeapFree(GetProcessHeap(), 0, This); > } > > - return refCount; > + return ref; > } > > /*** IPersist methods ***/ > > > >

2 1

Re: [PATCH] amstream: Check params in IAMMultiMediaStreamImpl_OpenFile.
by Nikolay Sivov 11 Oct '12

11 Oct '12

On 10/12/2012 00:59, Christian Costa wrote: > --- > dlls/amstream/amstream.c | 12 +++++++++--- > 1 file changed, 9 insertions(+), 3 deletions(-) > > > - TRACE("(%p/%p)->(%s,%x)\n", This, iface, debugstr_w(pszFileName), dwFlags); > + TRACE("(%p/%p)->(%s,%x)\n", This, iface, debugstr_w(filename), flags); > + > + if (!filename) > + return E_POINTER; > + > + if (flags & ~(AMMSF_NOCLOCK | AMMSF_NORENDER | AMMSF_RENDERALLSTREAMS | AMMSF_RENDERTOEXISTING | AMMSF_RUN)) > + return E_INVALIDARG; This needs tests.

1 0

Re: [1/3] cmd/tests: Show name of current test file on standard output
by Alexandre Julliard 11 Oct '12

11 Oct '12

Frédéric Delanoy <frederic.delanoy(a)gmail.com> writes: > When there are several test files, indicates which file error line numbers relate to, making > "debugging" easier. That's what trace() is for. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: gdi32: use HandleToULong inline function to convert handle to index instead of casting
by Dmitry Timoshkov 11 Oct '12

11 Oct '12

Daniel Lehman <dlehman(a)esri.com> wrote: > 0xffffffff000026a8 becomes 0xc00009aa instead of just 0x9aa It can't, the code explicitly casts handle to unsigned before the shift. -- Dmitry.

3 2

Re: [2/2] shlwapi/tests: Add more PathCreateFromUrl() tests.
by Marvin 11 Oct '12

11 Oct '12

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=22127 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 1/4] jscript: Use custom string container instead of BSTR.
by Henri Verbeet 11 Oct '12

11 Oct '12

On 11 October 2012 12:16, Jacek Caban <jacek(a)codeweavers.com> wrote: > > +struct _jsstr_t { > + unsigned length_flags; > + unsigned ref; > + WCHAR str[1]; > +}; It's probably much too late for jscript, and perhaps Wine in general, but strictly speaking the _t suffix is reserved by POSIX.

2 1

Re: [PATCH 5/6] mshtml: Added createStyleSheet and get_styleSheet tests
by Marvin 11 Oct '12

11 Oct '12

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=22123 Your paranoid android. === WVISTAADM (32 bit dom) === Failure running script in VM: The specified guest user must be logged in interactively to perform this operation

1 0