February 2011 - Wine-Devel - list.winehq.org

Tray icon not showing up if wine was run previously with DISPLAY unset?

by Dan Kegel

So, I've got this shell script that does something like h=`wine winepath -w $HOME` ... wine autohotkey run-setup.ahk Works great. However, if the user has selected to run wine in a virtual desktop, running winepath will start up and then shut down the virtual desktop, which is visually annoying. Second try: unset DISPLAY when running winepath, e.g. h=`DISPLAY= wine winepath -w $HOME` ... wine autohotkey run-setup.ahk Works great; now winepath doesn't pop up the desktop. Except now autohotkey's tray icon doesn't show up at all, which makes it hard to debug the autohotkey script. (The setup that autohotkey runs can access X just fine, so the problem seems confined to the tray icon.) Is the lack of tray icon a wine bug, or a user error? Third try: wait for wineserver to exit before starting autohotkey h=`DISPLAY= wine winepath -w $HOME` ... wineserver -w wine autohotkey run-setup.ahk but that's slower, and will hang if the user is running a wine app. I guess I could do "sleep 5" but that's an awful kludge. Probably I'll go back to letting the ugly virtual desktop pop up. So, what's the proper way to run one app under wine *without* X while letting other apps run wine *with* X? - Dan

Feb. 4, 2011

Re: loader: Build the preloader for ARM

by Peter Rosin

Den 2011-02-04 20:36 skrev Peter Rosin: > Den 2011-02-04 17:41 skrev André Hentschel: >> % and / both lead to external dependencies here... >> maybe the euclidean algorithm helps, because -Os would need ugly changes to configure i guess. >> > > Finding gcd is a bit over the top... > > How about something like the following? > > Cheers, > Peter > > unsigned int wld_modulo( const unsigned int a, const Elf32_Word b ) > { > #ifdef __ARM_EABI__ > unsigned int r = 0; > int i; > > for (i = sizeof(a) * 8 - 1; i >= 0; --i) { > r <<= 1; > if (a & (1U << i)) > r |= 1; > if (r > b) Ooops, bug. Should be "if (r >= b)" > r -= b; > } > > return r; > ... > > Cheers, Peter

Feb. 4, 2011

Re: loader: Build the preloader for ARM

by Peter Rosin

Den 2011-02-04 17:41 skrev André Hentschel: > Am 04.02.2011 17:26, schrieb Marcus Meissner: >> On Fri, Feb 04, 2011 at 04:51:46PM +0100, André Hentschel wrote: >>> Am 04.02.2011 16:31, schrieb Alexandre Julliard: >>>> André Hentschel <nerv(a)dawncrow.de> writes: >>>> >>>>> +unsigned int wld_modulo( const unsigned int a, const Elf32_Word b ) >>>>> +{ >>>>> +#ifdef __ARM_EABI__ >>>>> + unsigned int i; >>>>> + if(b > a) return a; >>>>> + for(i = 0; i * b < a + b; i++) >>>>> + if(i * b > a - b) return (a - i * b); >>>>> + return 0; >>>> >>>> That's very inefficient, there are better ways... >>>> >>> >>> Thanks for having a look at it. >>> The Problem is that i also can't use '/', otherwise it wants to link to __aeabi_uidiv . >>> That leads to some Problems, so i'll try to find some better math solution and if that fails i maybe try it the assembler way... >> >> modulo would be "%" btw, >> >> return a%b; >> >> but this might generate external functions too... hmm, try -Os as option? >> >> Ciao, Marcus > > % and / both lead to external dependencies here... > maybe the euclidean algorithm helps, because -Os would need ugly changes to configure i guess. > Finding gcd is a bit over the top... How about something like the following? Cheers, Peter unsigned int wld_modulo( const unsigned int a, const Elf32_Word b ) { #ifdef __ARM_EABI__ unsigned int r = 0; int i; for (i = sizeof(a) * 8 - 1; i >= 0; --i) { r <<= 1; if (a & (1U << i)) r |= 1; if (r > b) r -= b; } return r; ...

Feb. 4, 2011

Re: the patches around ok()

by André Hentschel

Am 04.02.2011 20:00, schrieb Joris Huizer: > Hello Henri Verbeet, and André Hentschel > > I wasn't (actively) on the mailing list, that's why I couldn't reply directly to your messages. > I will try and make sure I have better titles for such patches in future! > > Regards, > Joris > Great, but you should also resend them this time. I guess something like that would be ok: "imagehlp/tests/integrity.c: Don't test function directly when reporting GetLastError() (resend)" -- Best Regards, André Hentschel

Feb. 4, 2011

the patches around ok()

by Joris Huizer

Hello Henri Verbeet, and André Hentschel I wasn't (actively) on the mailing list, that's why I couldn't reply directly to your messages. I will try and make sure I have better titles for such patches in future! Regards, Joris

Feb. 4, 2011

Pulling Patch

by James McKenzie

Since my Mac is dying I have decided to return to the Windows world. Please pull any and all patches. I have envoked the right to copyright and none of my code can or will be used in Wine. Thank you. James McKenzie

Feb. 4, 2011

Re: kernel32/tests/thread.c: Don't call in ok()

by André Hentschel

Am 04.02.2011 17:24, schrieb Henri Verbeet: > On 4 February 2011 17:15, Joris Huizer <joris_huizer(a)yahoo.com> wrote: >> Subject: kernel32/tests/thread.c: Don't call in ok() > Why not? > > Because of the GetLastError() calls. But on the other hand Joris really should mention that in the subject. Joris, please have a look at my older patches like 85386c2c6eaed6ea89eefef6c8948169323bbc28 -- Best Regards, André Hentschel

Feb. 4, 2011

Re: loader: Build the preloader for ARM

by André Hentschel

Am 04.02.2011 17:26, schrieb Marcus Meissner: > On Fri, Feb 04, 2011 at 04:51:46PM +0100, André Hentschel wrote: >> Am 04.02.2011 16:31, schrieb Alexandre Julliard: >>> André Hentschel <nerv(a)dawncrow.de> writes: >>> >>>> +unsigned int wld_modulo( const unsigned int a, const Elf32_Word b ) >>>> +{ >>>> +#ifdef __ARM_EABI__ >>>> + unsigned int i; >>>> + if(b > a) return a; >>>> + for(i = 0; i * b < a + b; i++) >>>> + if(i * b > a - b) return (a - i * b); >>>> + return 0; >>> >>> That's very inefficient, there are better ways... >>> >> >> Thanks for having a look at it. >> The Problem is that i also can't use '/', otherwise it wants to link to __aeabi_uidiv . >> That leads to some Problems, so i'll try to find some better math solution and if that fails i maybe try it the assembler way... > > modulo would be "%" btw, > > return a%b; > > but this might generate external functions too... hmm, try -Os as option? > > Ciao, Marcus % and / both lead to external dependencies here... maybe the euclidean algorithm helps, because -Os would need ugly changes to configure i guess. -- Best Regards, André Hentschel

Feb. 4, 2011

Re: loader: Build the preloader for ARM

by Marcus Meissner

On Fri, Feb 04, 2011 at 04:51:46PM +0100, André Hentschel wrote: > Am 04.02.2011 16:31, schrieb Alexandre Julliard: > > André Hentschel <nerv(a)dawncrow.de> writes: > > > >> +unsigned int wld_modulo( const unsigned int a, const Elf32_Word b ) > >> +{ > >> +#ifdef __ARM_EABI__ > >> + unsigned int i; > >> + if(b > a) return a; > >> + for(i = 0; i * b < a + b; i++) > >> + if(i * b > a - b) return (a - i * b); > >> + return 0; > > > > That's very inefficient, there are better ways... > > > > Thanks for having a look at it. > The Problem is that i also can't use '/', otherwise it wants to link to __aeabi_uidiv . > That leads to some Problems, so i'll try to find some better math solution and if that fails i maybe try it the assembler way... modulo would be "%" btw, return a%b; but this might generate external functions too... hmm, try -Os as option? Ciao, Marcus

Feb. 4, 2011

Re: kernel32/tests/thread.c: Don't call in ok()

by Henri Verbeet

On 4 February 2011 17:15, Joris Huizer <joris_huizer(a)yahoo.com> wrote: > Subject: kernel32/tests/thread.c: Don't call in ok() Why not?

Feb. 4, 2011

Re: Try to Implement my first Stub function - AbortPrinter().

by Henri Verbeet

On 4 February 2011 15:30, Charles Davis <cdavis(a)mymail.mines.edu> wrote: > Probably. Every UNIX I've seen uses CUPS (short for Common UNIX Printing CUPS is probably pretty common these days, but LPD has been around for a long time. If it's really CUPS specific it should check for CUPS though, not OS X. Also note that the code now has unused variables in the #else case.

Feb. 4, 2011

Re: msvcr90: Fixed a typo

by André Hentschel

Am 04.02.2011 16:51, schrieb Nikolay Sivov: > On 2/4/2011 18:33, André Hentschel wrote: >> --- >> dlls/msvcr90/msvcr90.c | 2 +- >> 1 files changed, 1 insertions(+), 1 deletions(-) >> >> diff --git a/dlls/msvcr90/msvcr90.c b/dlls/msvcr90/msvcr90.c >> index 2c811cb..d2a61dd 100644 >> --- a/dlls/msvcr90/msvcr90.c >> +++ b/dlls/msvcr90/msvcr90.c >> @@ -315,7 +315,7 @@ const char * __thiscall MSVCRT_type_info_name_internal_method(type_info * _this, >> if (!_this->name) >> { >> /* Create and set the demangled name */ >> - /* Nota: mangled name in type_info struct always start with a '.', while >> + /* Note: mangled name in type_info struct always start with a '.', while >> * it isn't valid for mangled name. >> * Is this '.' really part of the mangled name, or has it some other meaning ? >> */ > Maybe 'starts with' ? True... -- Best Regards, André Hentschel

Feb. 4, 2011

Re: loader: Build the preloader for ARM

by André Hentschel

Am 04.02.2011 16:31, schrieb Alexandre Julliard: > André Hentschel <nerv(a)dawncrow.de> writes: > >> +unsigned int wld_modulo( const unsigned int a, const Elf32_Word b ) >> +{ >> +#ifdef __ARM_EABI__ >> + unsigned int i; >> + if(b > a) return a; >> + for(i = 0; i * b < a + b; i++) >> + if(i * b > a - b) return (a - i * b); >> + return 0; > > That's very inefficient, there are better ways... > Thanks for having a look at it. The Problem is that i also can't use '/', otherwise it wants to link to __aeabi_uidiv . That leads to some Problems, so i'll try to find some better math solution and if that fails i maybe try it the assembler way... -- Best Regards, André Hentschel

Feb. 4, 2011

Re: msvcr90: Fixed a typo

by Nikolay Sivov

On 2/4/2011 18:33, André Hentschel wrote: > --- > dlls/msvcr90/msvcr90.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/dlls/msvcr90/msvcr90.c b/dlls/msvcr90/msvcr90.c > index 2c811cb..d2a61dd 100644 > --- a/dlls/msvcr90/msvcr90.c > +++ b/dlls/msvcr90/msvcr90.c > @@ -315,7 +315,7 @@ const char * __thiscall MSVCRT_type_info_name_internal_method(type_info * _this, > if (!_this->name) > { > /* Create and set the demangled name */ > - /* Nota: mangled name in type_info struct always start with a '.', while > + /* Note: mangled name in type_info struct always start with a '.', while > * it isn't valid for mangled name. > * Is this '.' really part of the mangled name, or has it some other meaning ? > */ Maybe 'starts with' ?

Feb. 4, 2011

Re: loader: Build the preloader for ARM

by Alexandre Julliard

André Hentschel <nerv(a)dawncrow.de> writes: > +unsigned int wld_modulo( const unsigned int a, const Elf32_Word b ) > +{ > +#ifdef __ARM_EABI__ > + unsigned int i; > + if(b > a) return a; > + for(i = 0; i * b < a + b; i++) > + if(i * b > a - b) return (a - i * b); > + return 0; That's very inefficient, there are better ways... -- Alexandre Julliard julliard(a)winehq.org

Feb. 4, 2011

Re: Try to Implement my first Stub function - AbortPrinter().

by Charles Davis

On 2/4/11 6:53 AM, Loïc Maury wrote: > On 03/02/11 12:02, Nikolay Sivov wrote: >> Why it's Mac specific? > In fact I don't know if I need this, but I am on Mac with CUPS, I don't > know if it the same > issue for the other platform ? Probably. Every UNIX I've seen uses CUPS (short for Common UNIX Printing System). There's also HPLIP (Hewlett-Packard Linux Imaging and Printing System), but I've never seen any programs that use that directly, and I'm not even sure if it still exists anymore :). Chip

Feb. 4, 2011

Re: Try to Implement my first Stub function - AbortPrinter().

by Loïc Maury

Hello Mr.Sivov and Wine community, On 03/02/11 12:02, Nikolay Sivov wrote: > On 2/3/2011 13:51, Loïc Maury wrote: >> Hello, >> >> I try to implement my first stub function - *AbortPrinter()*. >> But before to make a patch, I need your advice. >> >> For what I understood, AbortPrinter(), remove the document >> spool file for a printer, created by StartDocPrinter(), who indicate >> that a document was spooled. >> >> I saw that the API printer functions are in dlls/winspool.drv/info.c >> >> I understood that there are 4 structures used, *opened_printer_t* for >> manage a printer, >> *jobqueue_t* for a list of job for a printer, *started_doc_t* who >> manage a document, >> and *job_t* for manage a job. >> >> I am on Mac Os 10.6, with CUPS. >> >> Here is the code, what do you think ? > Patch needs to be a diff, not a copy of a call. Some useful links > http://wiki.winehq.org/GitWine, http://wiki.winehq.org/SubmittingPatches. OK I understand. > >> BOOL WINAPI AbortPrinter(HANDLE hPrinter) >> { >> PRINTER_INFO_5W *pi5 = NULL; >> BOOL ret = FALSE; >> opened_printer_t *printer; >> struct list *cursor, *cursor2; >> job_t *job = 0; > No need to init 'job' it seems. Ok >> DWORD jobDocId; >> DWORD needed; >> LPWSTR portname; >> >> #if defined(__APPLE__) > Why it's Mac specific? In fact I don't know if I need this, but I am on Mac with CUPS, I don't know if it the same issue for the other platform ? >> >> EnterCriticalSection(&printer_handles_cs); >> >> printer = get_opened_printer(hPrinter); >> >> if(!printer) >> { >> ERR("The handle for the printer is invalid.\n"); >> SetLastError(ERROR_INVALID_HANDLE); > If you set last error you should add a test for it. Ok >> goto end; >> } >> >> TRACE("(%s, %s, %p, %d, %d)\n",debugstr_w(printer->name) >> ,debugstr_w(printer->printername) >> ,printer->backend_printer >> ,printer->queue->ref >> ,list_count(&printer->queue->jobs)); >> >> /* No jobs to manage. */ >> if(list_count(&printer->queue->jobs) == 0) >> { >> ERR("No job in the list.\n"); > It's not an error to have no jobs. Yes, I can remove this. >> SetLastError(ERROR_SPL_NO_STARTDOC); >> goto end; >> } >> >> if(printer->doc) { >> TRACE("Document inside for job id : %d\n", printer->doc->job_id); >> jobDocId = printer->doc->job_id; >> } >> else { >> ERR("No document.\n"); >> SetLastError(ERROR_SPL_NO_STARTDOC); >> goto end; >> } >> >> /* For each jobs, get the job document in the double linked list */ >> LIST_FOR_EACH_SAFE(cursor, cursor2, &printer->queue->jobs) >> { >> /* Take a job. */ >> job = LIST_ENTRY(cursor, job_t, entry); >> >> TRACE("(job id : %d, filename : %s, portname : %s, document >> title : %s, printer name %s)\n",job->job_id >> >> ,debugstr_w(job->filename) >> >> ,debugstr_w(job->portname) >> >> ,debugstr_w(job->document_title) >> >> ,debugstr_w(job->printer_name)); >> if(jobDocId == job->job_id) >> { >> TRACE("(hf : %p, job id : %d)\n",printer->doc->hf >> ,printer->doc->job_id); >> >> /* Get portname. */ >> if (!job->portname) >> { >> GetPrinterW(hPrinter, 5, NULL, 0, &needed); >> pi5 = HeapAlloc(GetProcessHeap(), 0, needed); >> GetPrinterW(hPrinter, 5, (LPBYTE)pi5, needed, &needed); >> portname = pi5->pPortName; >> } >> >> TRACE("portname : %s\n", debugstr_w(portname)); >> >> /* Cups Port */ >> if(!strncmpW(portname, CUPS_Port, strlenW(CUPS_Port))) >> { >> TRACE("Remove job from the list.\n"); >> list_remove(cursor); >> CloseHandle(printer->doc->hf); >> DeleteFileW(job->filename); >> HeapFree(GetProcessHeap(), 0, job->document_title); >> HeapFree(GetProcessHeap(), 0, job->printer_name); >> HeapFree(GetProcessHeap(), 0, job->portname); >> HeapFree(GetProcessHeap(), 0, job->filename); >> HeapFree(GetProcessHeap(), 0, job->devmode); >> HeapFree(GetProcessHeap(), 0, job); > >> job = 0; > What is it for? Yes, sorry it's not necessary. >> ret = TRUE; >> >> if(pi5) >> HeapFree(GetProcessHeap(), 0, pi5); > No need to check point here. Ok I will move it after the "end:" label. >> } >> else >> FIXME("only CUPS for now.\n"); >> } >> } >> >> end: >> LeaveCriticalSection(&printer_handles_cs); >> >> TRACE("return %d\n", ret); >> return ret; >> >> #else >> TRACE("Only CUPS for Mac Os is managed for now.\n"); >> return FALSE; >> #endif > This should stay as FIXME obviously. Ok I will change it to FIXME. Thank you Loïc >> } >> >> Merci >> >> Loïc >> >> >> >

Feb. 4, 2011

Re: user32: Add a test for reparenting a WS_POPUP window to a WS_CHILD parent.

by Alexandre Julliard

Dmitry Timoshkov <dmitry(a)codeweavers.com> writes: > + SetLastError(0xdeadbeef); > +todo_wine > + ok(SetForegroundWindow(popup), "SetForegroundWindow() error %d\n", GetLastError()); It would be interesting to test what the foreground window is actually set to. > + SetLastError(0xdeadbeef); > + ok(DestroyWindow(parent), "DestroyWindow() error %d\n", GetLastError()); Please don't add invalid uses of GetLastError again, this was cleaned up recently. -- Alexandre Julliard julliard(a)winehq.org

Feb. 4, 2011

Re: user32: Add a test for reparenting a WS_POPUP window to a WS_CHILD parent.

by Dmitry Timoshkov

Joris Huizer <joris_huizer(a)yahoo.com> wrote: > > - ret = DestroyWindow(parent); > > - ok( ret, "DestroyWindow() error %d\n", > > GetLastError()); > > + ok(DestroyWindow(parent), "DestroyWindow() > > error %d\n", GetLastError()); > > Please don't merge the call with the ok() statement; the GetLastError() call might be executed before the function to be tested, thus giving the old value. (There has been some cleaning up for this issue recently) > > The same goes for a few tests you added: That's no the point of the test, and actually ok() calls are just a sinity check in this case, GetLastError() results are not interesting, since the previous calls should never fail. There are many other places which do that already, feel free to send patches to correct them. -- Dmitry.

Feb. 4, 2011

Re: user32: Add a test for reparenting a WS_POPUP window to a WS_CHILD parent.

by Joris Huizer

Hello Dmitry Timoshkov, I noticed a small issue in your patch: > > - ret = DestroyWindow(parent); > - ok( ret, "DestroyWindow() error %d\n", > GetLastError()); > + ok(DestroyWindow(parent), "DestroyWindow() > error %d\n", GetLastError()); Please don't merge the call with the ok() statement; the GetLastError() call might be executed before the function to be tested, thus giving the old value. (There has been some cleaning up for this issue recently) The same goes for a few tests you added: > + SetLastError(0xdeadbeef); > + ok(SetForegroundWindow(popup), > "SetForegroundWindow() error %d\n", GetLastError()); > + SetLastError(0xdeadbeef); > +todo_wine > + ok(SetForegroundWindow(popup), > "SetForegroundWindow() error %d\n", GetLastError()); > + > + SetLastError(0xdeadbeef); > + ok(DestroyWindow(parent), "DestroyWindow() > error %d\n", GetLastError()); > + HTH, Joris

Feb. 4, 2011

RE: Correction to crash inside RtlCaptureStackBackTrace() + test case

by Janne Hakonen

> You can't use exception handlers in tests I'm not sure why not, but I'll take you word of this. Without exception handling the test cases for RtlCaptureStackBackTrace would also always crash on Win 2000/NT so that makes all of the tests useless, I think. Well, at least I learned how to (/ not to) write test cases. :) Thanks, Janne

Feb. 4, 2011

Re: wined3d/arb_program_shader: Small optimization in state_texfactor_arbfp().

by Stefan Dösinger

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Actually I am afraid the patch isn't quite correct. If e.g. ps constant 220 is dirty, and no others are, then highest_dirty_ps_const is 220, and pshader_const_dirty[ARB_FFP_CONST_TFACTOR] is 0. After setting the texfactor you still have to reload constant nr. ARB_FFP_CONST_TFACTOR, so you have to set it to 1 in the array. If managing the dirty constant array is expensive you could try to replace it by a bitmap where you just store 1 bit per dirty constant. That reduces the memory size(helps with the CPU cache) and you can skip 8, 16 or 32 clean constants at once in shader_arb_load_constantsF but it adds some overhead because extra operations are needed to test for single bits. > + /* TODO: this call is actually very expensive, figure out a better way... */ > GL_EXTCALL(glProgramEnvParameter4fvARB(GL_FRAGMENT_PROGRAM_ARB, ARB_FFP_CONST_TFACTOR, col)); There's no real alternative, except hardcoding the texfactor into the shader, and I doubt that is cheaper. Maybe local constants are cheaper, but I doubt this as well. That also means you have to reload the texfactor every time you switch between fixed function pipeline replacement shaders. Am 04.02.2011 um 10:27 schrieb Stefan Dösinger: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > > Am 03.02.2011 um 23:01 schrieb Adam Martinson: > >> >> Oddly enough this gives us an average 1.8% boost in the 3DMark06 batch size tests. > Impressive. > > One more thing you could investigate(but that doesn't affect this patch): Try to figure out why this code is called in the first place. The only situation when this should happen is if the app is using fixed function fragment processing. I wouldn't expect that 3Dmark06 does that. > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG/MacGPG2 v2.0.17 (Darwin) > > iQIcBAEBAgAGBQJNS8aYAAoJEN0/YqbEcdMwI1AQAJder+8LZILzdT+8tA2YiwZ+ > ZEVOi4S2voFASHUVJbYuWIXkzzVu6fQlZXw/KCyvdQaslpTWQFFc/cOMXcCVjVs5 > puJJDydjCyI8FZF7ec8dCIjCHqBH4S91kxDjRmWGifdp1Qd6h/f7o1xyhOHG62xG > XZbY1xx3DhQ7lRxRtsdqqq22MM5cOqR4mCen2YPGsd04rlwsxW6zuxZ32oBof0rO > ggzBW1F2WdDHeT6rcDK5Iv7R4zdSCtl6IXfsMaT1frTcd8Mo/9jLj7GaLMnQ/UtG > JkyQbCP59AOtEKU8nwrISn71YMRkCBHOttaIAVhr6Nq02eNmcxWomSbIsb293LzP > fQHFDPJ2aO161AimO/7Nk6DbGniUtlgRGhPhrK+YJvrAhftrtZ2gFUaN3QLJgcKA > HHhkU00NsBXOJvz8YuGFDRVmpPpCLcGM7gvMIyjYy3x3W1FWO4giNxnj6CxqHmLs > ohkAbsLEnJohhpTyVEaQKTG0Zq49s5WcvkjQbmncqKdP2D9NHmi50FI8MbJJxXAu > mMW6sA6oExn31Fnm1/pQOWCZk8hidK1hRN7f25nSRa58q7bSvNT/Gw3kk6S8T5CU > yL3r5UIb64ss4QmtXY8SPIR9IDNvmnYUX4/Av/iKpZrSRHCHzpYX7Eecp2jdV/ta > jgKpR4VaDyemyij093oW > =LXas > -----END PGP SIGNATURE----- > > -----BEGIN PGP SIGNATURE----- Version: GnuPG/MacGPG2 v2.0.17 (Darwin) iQIcBAEBAgAGBQJNS8nSAAoJEN0/YqbEcdMwJpQP/jAISQAlW/44UoMrUWtOio+S Hb7ut/9LDm2nBvQ6AaicehMkGj7V8+07D61/jPc/9DkYjhEDCr+D1kulXZ9p4nxs 7YLWw1Wc3Z45z+3YznUbnHD4kzQIhLntEIKdl80PBBNcHpfouNMHOO5J31ncrbzc q0wvjxYJJ80pskdzbIzQv1sO7+OrwgHVIJ8lNLW36PQuadBsOrSy6vlxO7brflB+ aG0/qIyN30M83MKzdrvYZjcck+CpPwzFdsfU+3niL+hlv5JoBuZMoxqVllLdtwDG 9fe1rOF8lDDsYXhrqZdgndKSpbWn7aY9LY0ya4YQEtTiNfkSGCf+iMVBhGr0Jy7c MGa+ElXLZKq/ZbSIWVVRXdYt3O5Znvtyta2tThxW/QK7HyK3Jq3l7P+EZgzfWlQt kjkJgVXVtPox441cvr3R8oV2M57iwBhsYWwXsusC0BCD/WKFCRb/FDwY6f2G6L2p pFkKMhhC7nIaB6CBjPailQHfMC9tlhq/vB951Ir3UVIOAUuEHrKHK9L0jercnTfd QPym37IygBdAFjNKOvPI2b8DmVFKUAkkDGdaNZSU1Z8IMCogzs9bqf6L2JFF97U3 mH1PG6/tJnsBMUfUhnA8xreg7FyqIpmn53XV14q2XFtiAWY8IRhF97p8N71oSAUJ jhUnwEL1Tw51/p6GIIu7 =dmun -----END PGP SIGNATURE-----

Feb. 4, 2011

Re: wined3d/arb_program_shader: Small optimization in state_texfactor_arbfp().

by Stefan Dösinger

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Am 03.02.2011 um 23:01 schrieb Adam Martinson: > > Oddly enough this gives us an average 1.8% boost in the 3DMark06 batch size tests. Impressive. One more thing you could investigate(but that doesn't affect this patch): Try to figure out why this code is called in the first place. The only situation when this should happen is if the app is using fixed function fragment processing. I wouldn't expect that 3Dmark06 does that. -----BEGIN PGP SIGNATURE----- Version: GnuPG/MacGPG2 v2.0.17 (Darwin) iQIcBAEBAgAGBQJNS8aYAAoJEN0/YqbEcdMwI1AQAJder+8LZILzdT+8tA2YiwZ+ ZEVOi4S2voFASHUVJbYuWIXkzzVu6fQlZXw/KCyvdQaslpTWQFFc/cOMXcCVjVs5 puJJDydjCyI8FZF7ec8dCIjCHqBH4S91kxDjRmWGifdp1Qd6h/f7o1xyhOHG62xG XZbY1xx3DhQ7lRxRtsdqqq22MM5cOqR4mCen2YPGsd04rlwsxW6zuxZ32oBof0rO ggzBW1F2WdDHeT6rcDK5Iv7R4zdSCtl6IXfsMaT1frTcd8Mo/9jLj7GaLMnQ/UtG JkyQbCP59AOtEKU8nwrISn71YMRkCBHOttaIAVhr6Nq02eNmcxWomSbIsb293LzP fQHFDPJ2aO161AimO/7Nk6DbGniUtlgRGhPhrK+YJvrAhftrtZ2gFUaN3QLJgcKA HHhkU00NsBXOJvz8YuGFDRVmpPpCLcGM7gvMIyjYy3x3W1FWO4giNxnj6CxqHmLs ohkAbsLEnJohhpTyVEaQKTG0Zq49s5WcvkjQbmncqKdP2D9NHmi50FI8MbJJxXAu mMW6sA6oExn31Fnm1/pQOWCZk8hidK1hRN7f25nSRa58q7bSvNT/Gw3kk6S8T5CU yL3r5UIb64ss4QmtXY8SPIR9IDNvmnYUX4/Av/iKpZrSRHCHzpYX7Eecp2jdV/ta jgKpR4VaDyemyij093oW =LXas -----END PGP SIGNATURE-----

Feb. 4, 2011

Re: wintrust/tests: make sure return values are used (LLVM/Clang) (resend)

by Austin English

On Fri, Feb 4, 2011 at 00:42, Paul Vriens <paul.vriens.wine(a)gmail.com> wrote: > On 02/04/2011 09:32 AM, Austin English wrote: >> >> - ret = funcs->pfnObjectTrust(&data); >> + funcs->pfnObjectTrust(&data); >> + ok(ret == S_FALSE, "Expected S_FALSE, got %08x\n", ret); > > This doesn't look correct (you are removing the assignment and then check > it). I know that this code path will never be taken, but still. Fixed, thanks. Bedtime now, before I make any more typos :-). -- -Austin

Feb. 4, 2011

Re: wintrust/tests: make sure return values are used (LLVM/Clang) (resend)

by Paul Vriens

On 02/04/2011 09:32 AM, Austin English wrote: > - ret = funcs->pfnObjectTrust(&data); > + funcs->pfnObjectTrust(&data); > + ok(ret == S_FALSE, "Expected S_FALSE, got %08x\n", ret); This doesn't look correct (you are removing the assignment and then check it). I know that this code path will never be taken, but still. -- Cheers, Paul.

Feb. 4, 2011