January 2002 - Wine-Devel - list.winehq.org

Re: crypto patch

by Travis Michielsen

Hmm, forgot this part of the part of the patch. - Travis On Wednesday, January 09, 2002 5:27 pm, you wrote: > Hi, > > This patch implements the cryptographic API functions existing in > advapi32.dll. However, there are still issues with some of the M$ > providers and the pstores.exe service, so I'm not sure how functional this > will be until something can be done about this service. > > Changelog: > Travis Michielsen <tjmichielsen(a)yahoo.com> > Implement advapi32 portion of the CryptoAPI.

Jan. 10, 2002

Debugger set breakpoint with full filename

by Bill Medland

I've spent time looking for the answer and trying the (to me) obvious options and now I give in. How, in the debugger, do I specify "Set a break point at DllMain of E:\BIN.50A\..MYDLL.DLL" without the debugger complaining about syntax etc.? If someone answers then I am prepared to add the info to debugger.sgml Bill

Jan. 10, 2002

How about sponsoring from IBM?

by Roland

Hello, it seems that IBM is becoming a major Linux advocate. What would be the possibilities of having it sponsor the WINE project? I mean 1 Million dollars is not much for IBM, but it certainly would mean a lot to the development of WINE. IBM would also profit of this. Since Linux runs on their supercomputers, it would be nice if you could run all Windows programs there(you need another Pentium emulator then, but that would not be a big problem, would it?). Just thinking about this... Roland

Jan. 10, 2002

Re: Lindows screenshots AND THE WINE LICENSE

by Roland

At 12:14 AM 1/10/02 +0000, James Tabor wrote: >Personally I would be pissed off if someone is making money after all the >work I had but into a "Free Project" assuming it would be free for everyone >else to use, including the source. To prevent this type of thing from happening is that the GPL was invented. But that doesn't mean that the WINE license is bad. Lets be honest, if WINE was GPL, then Lindows would not exist, and WINE would be in the same developing state that it is now. Since WINE is not GPL, someone took his chance and created Lindows, and WINE is also in the same developing state, since that company didn't contribute anything back. But, as you notice, this didn't make any difference for the status of WINE. As you see, with or without GPL(or LGPL) the developing state of the WINE project wouldn't change. But there are more dangers for the WINE license. It would be possible for a company like Lindows hire away all WINE developers, effectively hijacking the project. Alternatively if Lindows becomes a success it will be able to hire more programmers and keep improving their own version of WINE, so that it would always be much better than the free WINE. I wonder why this didn't happen to FreeBSD. Anyway, I'm not unhappy with the WINE license. In fact, since most important open source projects are GPL, or LGPL, I think it is interesting to have some projects under a different license to see the practical implications this will have. I think WINE is a good study case for that. It will help us all to see the advantages and disadvantages of the GPL(LGPL) and draw conclusions for the future from that. Suppose that in a few years another Opensource project faces the question of what license it will need, then we can point back and say: look at the WINE project and what license they had and what happened to them, etc... Just my 0.05$ Roland

Jan. 10, 2002

Re: Babystep: Testing framework

by Andriy Palamarchuk

--- Jeremy White <jwhite(a)codeweavers.com> wrote: > However, having the diff feature allows us to more > rapdily adapt existing programs to become tests. > Since it's done (and it's trivial code), I don't see > the harm in leaving in the feature. We can hide > it/discourage in in the (as yet unwritten) doco if you like. Some disadvantages of the diff approach was discussed before. I just realized the problem which will make using this approach practically impossible. The problem - variations of output as result of: 1) using different Windows versions. 2) using TODO tests. The problem becomes even worse if more than one Win32 implementation project (e.g. ODIN) starts to use the test, because list of TODOs is project-specific 3) I'm thinking about introducing WONTFIX feature. This is when we choose, e.g. WinNT behavior and don't implement Win95 behaviour. These WONTFIX calls are also project-specific. Now, imagin combinations of these :-) Explicit check, on other hand, nicely comments all these conditions in one place - in code. Andriy Palamarchuk __________________________________________________ Do You Yahoo!? Send FREE video emails in Yahoo! Mail! http://promo.yahoo.com/videomail/

Jan. 10, 2002

Re: Babystep: Testing framework

by Alexandre Julliard

Jeremy White <jwhite(a)codeweavers.com> writes: > Um...the way I constructed it, there is no additional effort required > in the Makefile whatsoever. Now, if you get rid of my .test file, > yes, then, doing this all in the Makefile is a pain in the neck. There is no effort unless you want to get the dependencies right; then you are in for a lot of trouble. With your proposal the only dependency is on the .test file, which will most likely never change over the lifetime of a test. OTOH the Perl/C source will change frequently, but make test will never notice it and won't re-run the test. That defeats the whole purpose of using make. > If the goal is to make it easy for a newcomer to add a test, then requiring > them to hack the Makefile is going to be a barrier. Asking them to > copy and cut/paste a 'samples/foo.test' file will not be. Sorry if I wasn't clear. My comment was about your example of adding a new category of tests, like support for .so tests. That's when you need some serious hacking. If all you do is add a Perl or C test, you only have to add a line in the Makefile, just like with your solution. > I think either choice is valid, but let me lay down the gauntlet: > replace my 'winetest/samples' directory with a Make based equivalent > that allows for the same sorts of tests, and makes it similarly obvious > to a newcomer how to add a new test. You are better with > Make than I am; you may persuade me readily with a patch. It would look pretty much the same. Just replace this: TESTS = sample1.test sample2.test sample3.test in your makefile by: PLTESTS = sample2.pl CTESTS = sample3.c and delete the .test files (there is no SHTESTS for shell tests, but I think that's a feature if we want portability). Creating a new test involves simply adding its name on the right line; no .test to write, no .spec file, no new Makefile. (if you'd like to try it I have committed the rules for PLTESTS so this should work now; CTESTS will still need a bit more work) -- Alexandre Julliard julliard(a)winehq.com

Jan. 9, 2002

Re: Babystep: Testing framework

by Jeremy White

Alexandre Julliard wrote: >Jeremy White <jwhite(a)codeweavers.com> writes: > >> The normal behaviour of the tests is that they are >>oblidged to provide an 0 exit code on success; I think that >>is an excellent standard. >> However, having the diff feature allows us to more >>rapdily adapt existing programs to become tests. >>Since it's done (and it's trivial code), I don't see >>the harm in leaving in the feature. We can hide it/discourage >>in in the (as yet unwritten) doco if you like. >> > >The problem is that it's hard to support both types of tests from the >makefile, so it would force tests without output to generate dummy >output and reference files. I think it's better to make the default be >no output and have tests that need output diffs do the extra work. > Um...the way I constructed it, there is no additional effort required in the Makefile whatsoever. Now, if you get rid of my .test file, yes, then, doing this all in the Makefile is a pain in the neck. >"Normal" Wine use should definitely include running regression >tests. I don't want to be the only one to run them. > Point taken; but I still believe that this is an issue not relevant to my proposed patch, but is a problem with using C tests as a whole. A solution to this problem will be just as appropriate with or without my patch. > >But without your .test file it's even easier to do, all you have to do >is to hack the makefile. Plus you can get the dependencies right (how >can make guess what file your .test is referencing?) > If the goal is to make it easy for a newcomer to add a test, then requiring them to hack the Makefile is going to be a barrier. Asking them to copy and cut/paste a 'samples/foo.test' file will not be. I do concede the point that my .test file hides dependencies, and a Makefile only approach can do that better. >You can do that from a Perl test too, but I don't think this should be >encouraged. Tests should not depend on external programs any more than >strictly necessary, otherwise people will soon need a complete Linux >distribution on their Windows machine to run our tests. > I threw that out as an example, not necessarily a serious case. I think the bottom line question is whether or not the test scripts should be built such that we depend on a shell script (runtest) and a configuration file for each test (foo.test), or whether we rely completely on the Makefiles to hold the test configuration information. I think either choice is valid, but let me lay down the gauntlet: replace my 'winetest/samples' directory with a Make based equivalent that allows for the same sorts of tests, and makes it similarly obvious to a newcomer how to add a new test. You are better with Make than I am; you may persuade me readily with a patch. Jer p.s. If you're just rejecting the .test files because they have long horizontal comment lines, we could negotiate that...<g>

Jan. 9, 2002

Re: Immediate IO in ReadFile / WriteFile

by Dan Kegel

Mike McCormack wrote: > As Lawson said, some programs use WaitForComm event to wait for data, > then GetCommError to check how many bytes arrived, then ReadFile to > read it. They then expect ReadFile to return TRUE. > > One such program is VBTerm, the visual basic terminal program, or at > least the VBX in it. > > Testing applications that use serial ports is difficult, and debugging > often must be done remotely. (eg. send me a trace like this, apply > this patch and send me another trace, etc.) For what it's worth, I documented some overlapped serial I/O quirks some time ago at http://www.alumni.caltech.edu/~dank/overlap.htm I also have a little running sample code there. - Dan

Jan. 9, 2002

Re: Lindows screenshots

by J.Brown (Ender/Amigo)

> Personally I would be pissed off if someone is making money after all the > work I had but into a "Free Project" assuming it would be free for everyone > else to use, including the source. > > I guess the project should change the license to LGPL and make it > retroactive and enforce it. > James > > Well, I wouldn't go that far. But alas the Lindows work will probably never see the light of day, according to the interview linked from slashdot.. "While Lindows will be constructed on top of open-source software, some of its key elements will be proprietary -- a strategy that will not make some open-source advocates happy but which Robertson says is essential to making the business work." Similar to Transgaming, only it looks like they've made major progress in making just everyday apps work.

Jan. 9, 2002

Re: Babystep: Testing framework

by Alexandre Julliard

Jeremy White <jwhite(a)codeweavers.com> writes: > The normal behaviour of the tests is that they are > oblidged to provide an 0 exit code on success; I think that > is an excellent standard. > However, having the diff feature allows us to more > rapdily adapt existing programs to become tests. > Since it's done (and it's trivial code), I don't see > the harm in leaving in the feature. We can hide it/discourage > in in the (as yet unwritten) doco if you like. The problem is that it's hard to support both types of tests from the makefile, so it would force tests without output to generate dummy output and reference files. I think it's better to make the default be no output and have tests that need output diffs do the extra work. > The entire samples directory that I submitted consisted > of 2,705 bytes. For all CVS checkouts and 'normal' Wine use, > there will be an insignificant penalty. "Normal" Wine use should definitely include running regression tests. I don't want to be the only one to run them. > For example, say we have 70 Winelib/C tests that are big > and bloated and we want to fix that. We could define a > new make target, say a '.test.so'. We could modify winetest > to be able to load a .test.so file and call main() with > the right parameters and suddenly all of our bloat is gone. > We'd just need a rule to build a .test.so file from > a .test.c file, and then we'd need to tweak our foo.test > file to add a new command 'so_test=foo.test.so' instead > of 'invoke=foo' and we're done. We could probably even > eliminate the .spec file (only for console apps, obviously). But without your .test file it's even easier to do, all you have to do is to hack the makefile. Plus you can get the dependencies right (how can make guess what file your .test is referencing?) > And another thing, what if I have an app that creates a > graphical image and then uses Gimp or some other scripting > process to take screen shots and then compares them to a known > reference? In a makefile only solution, it'd be a pain > to add a target to call some fancy script to perform > a regression test. This case is already handled by > the .test file. You can do that from a Perl test too, but I don't think this should be encouraged. Tests should not depend on external programs any more than strictly necessary, otherwise people will soon need a complete Linux distribution on their Windows machine to run our tests. -- Alexandre Julliard julliard(a)winehq.com

Jan. 9, 2002

Re: Immediate IO in ReadFile / WriteFile

by Mike McCormack

Hey Martin, As Lawson said, some programs use WaitForComm event to wait for data, then GetCommError to check how many bytes arrived, then ReadFile to read it. They then expect ReadFile to return TRUE. One such program is VBTerm, the visual basic terminal program, or at least the VBX in it. Testing applications that use serial ports is difficult, and debugging often must be done remotely. (eg. send me a trace like this, apply this patch and send me another trace, etc.) After making a change, i usually get a terminal program (such as KoalaTerm) and send a couple of "AT" commands to it... :-) Mike Original message from: lawson_whitney(a)juno.com > >On Wed, 9 Jan 2002, Martin Wilck wrote: >.... >> How do you check if data is ready in Windows? I didn't find a select() >> equivalent, nor is it possible to wait on file handles. >> >GetCommError and some other functions, I think, returns lpStat->cbInQue, >which is in effect the number of characters immediately available to be >read. > >> Martin >> >Lawson ------------------------------------------ mailto:Mike_McCormack(a)start.com.au ph +82 16 430 0425 __________________________________________________________________ Get your free Australian email account at http://www.Looksmart.com.au

Jan. 9, 2002

Re: Babystep: Testing framework

by Jeremy White

> Francois Gouget <fgouget(a)free.fr> writes: > > > So the test has to do the diff itself. And no calling 'diff' because > > it's not portable :-( > > But you shouldn't need diff at all. You just have to write the test > slightly differently to have checks instead of printfs. The normal behaviour of the tests is that they are oblidged to provide an 0 exit code on success; I think that is an excellent standard. However, having the diff feature allows us to more rapdily adapt existing programs to become tests. Since it's done (and it's trivial code), I don't see the harm in leaving in the feature. We can hide it/discourage in in the (as yet unwritten) doco if you like. > > > A C test should just be a .spec file and a .c file. If they are all > > handled the same way it's actually quite simple. In 'Programming Windows > > 98' I have 148 executables/libraries and 171 C files. It's quite > > manageable. > > I'm not sure about that; if you run make in libtest/ you end up with > 70 files and more than 7Mb of disk space, for just 10 nearly empty > programs. And the makefile is a big mess too. The entire samples directory that I submitted consisted of 2,705 bytes. For all CVS checkouts and 'normal' Wine use, there will be an insignificant penalty. And yes, in the framework I submitted, if you choose to run 'make test', it chews disk space (and the Winelib tests are fairly slow to launch). But the Makefile.in is, IMO, clean and tight, and easy to replicate. And we don't *have* to use Winelib/C tests - the Perl tests work fine. Further, if we had 70 regression tests, that would be exactly 70 more than we have now! We can fix the big bloatedness later. Finally, the more I think about my structure of using the '.test' files, the more I believe it is the right solution, because I think it gives us great long term flexibility. For example, say we have 70 Winelib/C tests that are big and bloated and we want to fix that. We could define a new make target, say a '.test.so'. We could modify winetest to be able to load a .test.so file and call main() with the right parameters and suddenly all of our bloat is gone. We'd just need a rule to build a .test.so file from a .test.c file, and then we'd need to tweak our foo.test file to add a new command 'so_test=foo.test.so' instead of 'invoke=foo' and we're done. We could probably even eliminate the .spec file (only for console apps, obviously). And another thing, what if I have an app that creates a graphical image and then uses Gimp or some other scripting process to take screen shots and then compares them to a known reference? In a makefile only solution, it'd be a pain to add a target to call some fancy script to perform a regression test. This case is already handled by the .test file. And don't forget the Julian fries! The very best feature of the framework! <g> Jer

Jan. 9, 2002

Re: int $3 and pstores service

by Francois Gouget

On Wed, 9 Jan 2002, Travis Michielsen wrote: [...] > Immediately after the call to CreateProcessA, a call to interrupt 3 seems to > be made but doesn't appear to return. Anyone know what this interupt does or > sopposed to do? Is there any support for it in wine? 'int 3' is a software breakpoint. I know that I used to insert such software breakpoints in my code on Windows. It made it possible to implement pretty intelligent breakpoints and was sometimes pretty useful. But here I suspect it is there to prevent debugging of pstores... -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ The nice thing about meditation is that it makes doing nothing quite respectable -- Paul Dean

Jan. 9, 2002

Re: Lindows screenshots

by Steve Langasek

On Thu, 10 Jan 2002, James Tabor wrote: > Bang Jun-Young wrote: > > On Fri, Jan 04, 2002 at 08:06:11AM -0800, Bill Medland wrote: > > > I wouldn't hold my breath; why would they contribute the valuable stuff > > > back? > > What do you think of people that have contributed their invaluable > > time/money/energy/passion/etc to Wine for last years? :P > Personally I would be pissed off if someone is making money after all the > work I had but into a "Free Project" assuming it would be free for everyone > else to use, including the source. > I guess the project should change the license to LGPL and make it > retroactive and enforce it. Once a license has been granted, it can't be unilaterally revoked. Although going forward Wine could be distributed under a different license than the present one, past decisions will stand. Steve Langasek postmodern programmer

Jan. 9, 2002

int $3 and pstores service

by Travis Michielsen

Hello, I have been tracing into pstores.exe the protected storage sevice provider in windows (win98). This service may be necessary to get the cryptographic API to function using native dlls (no wine providers currently exist). I noticed that the service was being started using the following calls: 1) The program calls CryptAcquireContext. 2) CryptAcquireContext calls CPAcquireContext of the apropriate service provider 3) Any MS provider I tested would call PStoreCreateInstance in pstorec.dll 4) pstore.exe would be started by a call to CreateProcessA Immediately after the call to CreateProcessA, a call to interrupt 3 seems to be made but doesn't appear to return. Anyone know what this interupt does or sopposed to do? Is there any support for it in wine? A few other notes: -pstores appears to execute fine when started directly under wine. -when started from above it seems to crash (randomly?) depending on the provider? usually tries to execute the instruction at 0x00000000. -some other calls in the API appear to rely on this service I *think* through psbase.dll (but maybe pstores.exe somehow). I haven't verified this yet though. Anyone have some suggestions? - Travis _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com

Jan. 9, 2002

Re: Lindows screenshots

by James Tabor

Bang Jun-Young wrote: > > On Fri, Jan 04, 2002 at 08:06:11AM -0800, Bill Medland wrote: > > I wouldn't hold my breath; why would they contribute the valuable stuff > > back? > > What do you think of people that have contributed their invaluable > time/money/energy/passion/etc to Wine for last years? :P > > Jun-Young > > -- > Bang Jun-Young <junyoung(a)mogua.com> Personally I would be pissed off if someone is making money after all the work I had but into a "Free Project" assuming it would be free for everyone else to use, including the source. I guess the project should change the license to LGPL and make it retroactive and enforce it. James

Jan. 9, 2002

Re: Babystep: Testing framework

by Alexandre Julliard

Francois Gouget <fgouget(a)free.fr> writes: > So the test has to do the diff itself. And no calling 'diff' because > it's not portable :-( But you shouldn't need diff at all. You just have to write the test slightly differently to have checks instead of printfs. > A C test should just be a .spec file and a .c file. If they are all > handled the same way it's actually quite simple. In 'Programming Windows > 98' I have 148 executables/libraries and 171 C files. It's quite > manageable. I'm not sure about that; if you run make in libtest/ you end up with 70 files and more than 7Mb of disk space, for just 10 nearly empty programs. And the makefile is a big mess too. -- Alexandre Julliard julliard(a)winehq.com

Jan. 9, 2002

Re: Babystep: Testing framework

by Francois Gouget

On 9 Jan 2002, Alexandre Julliard wrote: > Francois Gouget <fgouget(a)free.fr> writes: > > > I think we should definitely be able to check the test output against > > a reference version. > > The idea is that there is no test output (except for debugging); > everything is checked right inside the test. So the test has to do the diff itself. And no calling 'diff' because it's not portable :-( > > Why can't you have one executable per test? Or is there some > > test/check confusion here? One check per executable, or perl file for > > that matter would be crazy. > > Because a Winelib executable is too heavyweight; you need a spec file, > multiple object files, a lot of disk space, etc. You cannot > realistically have hundreds of them without creating a horrible mess. A C test should just be a .spec file and a .c file. If they are all handled the same way it's actually quite simple. In 'Programming Windows 98' I have 148 executables/libraries and 171 C files. It's quite manageable. -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ Good judgment comes from experience, and experience comes from bad judgment -- Barry LePatner

Jan. 9, 2002

Re: Is W really UTF-16?

by Bill Medland

Thanks for the repsonse "Ove Kaaven" <ovehk(a)ping.uio.no> wrote in message news:Pine.LNX.4.21.0201092154570.4461-100000(a)mizar.ping.uio.no... > > On Wed, 9 Jan 2002, Bill Medland wrote: > > > While I was working on the DrawText functions over the past many months I > > started wondering about when it would fail. (I'm pedantic and such things > > fascinate me!). The main concern I have is how to walk a W string > > correctly. For example while "ellipsifying" text we will need to "move the > > pointer to the previous character" which is currently done by decrementing > > the pointer by 1. But from what I currently understand that won't work if > > there are surrogate pairs. > > If you're concerned about that, surrogate pairs are the least of your > worries. You should also be concerned about Unicode combining (or > composite) characters. I think they might be identified with ctypes > C3_NONSPACING and C3_DIACRITIC and that kind of stuff... Good point. > > > 1. Does anyone know under what circumstances CharNextW isn't +1 (apart from > > when pointing at the terminating 0) > > Have you tried low surrogate followed by high surrogate, on a Microsoft OS > recent enough that Microsoft *might* have thought about preparing it for > dealing with surrogates? Well, that's the complication. I am lazy so I don't fancy the work involved in learning enough to put together a font that actually uses a surrogate pair so that I can test it with ExtTextOut, which is why I took the easy route of assuming that was what CharNextW was for. I guess I'll have to do the hard work since that family of functions are the ones I have seen that seem to suggest they are UTF-16 compatible. > > > 2. Is e.g. XP really using UTF-16 or is it actually still UCS2? > > I don't know. But it probably ought to be UTF16. > > > 3 Have we thought about how we should handle walking along a W string (in a > > fashion that doesn't reduced the speed to a crawl). I guess that in the > > short term I am expecting some sort of macro or inline. > > With p++, perhaps? There aren't very many circumstances where that is > going to be a problem (where unicode composite characters are not also), > is there? No, so they should both probably be handled together. Ah well, when I am next in there I'll probably just add a FIXME UTF-16 comment or something. Thanks again

Jan. 9, 2002

Re: Immediate IO in ReadFile / WriteFile

by lawson_whitney＠juno.com

On Wed, 9 Jan 2002, Martin Wilck wrote: ... > How do you check if data is ready in Windows? I didn't find a select() > equivalent, nor is it possible to wait on file handles. > GetCommError and some other functions, I think, returns lpStat->cbInQue, which is in effect the number of characters immediately available to be read. > Martin > Lawson ---oof---

Jan. 9, 2002

A new async patch - feedback, please :-)

by Martin Wilck

Hi Mike, all, Unfortunately this patch has become big, because i came across so many problems today and I just needed to go forward. I am sorry for this, but it has become very late tonight, and I want people to see this stuff to stop me before it's too late (if I'm on the wrong track). It compiles without warnings, but I have not tested it - may crash wine right away, I don't know. However it should illustrate my ideas well: * Full separation of async handling and file io-specific stuff. * Do not rely on LPOVERLAPPED for status data in generic code, it may be NULL. File IO code may rely on it, here it is required to be non-NULL. * Use GetOverlappedResult() to check for immediate completion (UNTESTED!) Please give me comments telling me which parts you don't like, so that I can adjust for further development. Mike, can you tell me about those apps that were causing trouble in the past ? Where can I obtain them? Do they need special hardware (modem, mouse) for test runs ? Cheers, Martin -- Martin Wilck Phone: +49 5251 8 15113 Fujitsu Siemens Computers Fax: +49 5251 8 20409 Heinz-Nixdorf-Ring 1 mailto:Martin.Wilck(a)Fujitsu-Siemens.com D-33106 Paderborn http://www.fujitsu-siemens.com/primergy ChangeLog: server/protocol.def: cancel_async(): Define new reqest type. server/async.c: cancel_async(): Implement new reqest type. include/file.h: struct async_ops: 4 methods for async_private: get_status, set_status, get_count, call_completion. struct async_private: contains only those elements which are absolutely necessary for queueing and server interaction struct async_fileio: contains all other fields. scheduler/synchro.c: register_async(): replacement for FILE_StartAsync(). Uses STATUS_USER_APC as (pseudo-) indicator to queue a new request. Hooks new requests in NtCurrentTeb()->pending_list (formerly done in calling functions). Not used by CancelIo() anymore. Finishes requests that are not in PENDING state after scheduling. finish_async(): Assume status is already set correctly. Queue call_completion unconditionally (No completion_func check needed) check_async(): Use register_async() for rescheduling. files/file.c: GetOverlappedResult(): Return ERROR_IO_INCOMPLETE if IO still pending (MSDN docs) FILE_StartAsync(): -> scheduler/synchro.c: register_async CancelIo(): Use new cancel_async server request (I felt former StartAsync() was too overloaded). Skip requests with wrong handle. fileio_call_completion_func(): Now called unconditionally, check if completion_func is non-NULL. FILE_AsyncReadService(): type adjustments according to new API FILE_AsyncWriteService(): dito FILE_ReadFileEx(): use new async API. Call GetOverlappedResult() to check for immediate completion. FILE_WriteFileEx(): dito. Now exactly equivalent to FILE_ReadFileEx() kernel/comm.c: COMM_WaitCommEventService(): Use new async API. diff -ruX diffignore CVS/wine/dlls/kernel/comm.c MW/wine/dlls/kernel/comm.c --- CVS/wine/dlls/kernel/comm.c Tue Jan 8 18:53:41 2002 +++ MW/wine/dlls/kernel/comm.c Wed Jan 9 21:20:37 2002 @@ -1514,12 +1514,13 @@ */ static void COMM_WaitCommEventService(async_private *ovp) { - LPOVERLAPPED lpOverlapped = ovp->lpOverlapped; + async_fileio *fileio = (async_fileio*) ovp; + LPOVERLAPPED lpOverlapped = fileio->lpOverlapped; TRACE("overlapped %p\n",lpOverlapped); /* FIXME: detect other events */ - *ovp->buffer = EV_RXCHAR; + *fileio->buffer = EV_RXCHAR; lpOverlapped->Internal = STATUS_SUCCESS; } @@ -1536,7 +1537,7 @@ LPOVERLAPPED lpOverlapped) /* [in/out] for Asynchronous waiting */ { int fd,ret; - async_private *ovp; + async_fileio *ovp; if(!lpOverlapped) { @@ -1547,52 +1548,35 @@ if(NtResetEvent(lpOverlapped->hEvent,NULL)) return FALSE; - lpOverlapped->Internal = STATUS_PENDING; - lpOverlapped->InternalHigh = 0; - lpOverlapped->Offset = 0; - lpOverlapped->OffsetHigh = 0; - fd = FILE_GetUnixHandle( hFile, GENERIC_WRITE ); if(fd<0) return FALSE; - ovp = (async_private *) HeapAlloc(GetProcessHeap(), 0, sizeof (async_private)); + ovp = (async_fileio *) HeapAlloc(GetProcessHeap(), 0, sizeof (async_fileio)); if(!ovp) { close(fd); return FALSE; } + + ovp->async.ops = &fileio_async_ops; + ovp->async.handle = hFile; + ovp->async.fd = fd; + ovp->async.type = ASYNC_TYPE_WAIT; + ovp->async.func = COMM_WaitCommEventService; + ovp->lpOverlapped = lpOverlapped; - ovp->func = COMM_WaitCommEventService; ovp->buffer = (char *)lpdwEvents; - ovp->fd = fd; ovp->count = 0; ovp->completion_func = 0; - ovp->type = ASYNC_TYPE_WAIT; - ovp->handle = hFile; - - ovp->next = NtCurrentTeb()->pending_list; - ovp->prev = NULL; - if(ovp->next) - ovp->next->prev=ovp; - NtCurrentTeb()->pending_list = ovp; - /* start an ASYNCHRONOUS WaitCommEvent */ - SERVER_START_REQ( register_async ) - { - req->handle = hFile; - req->overlapped = lpOverlapped; - req->type = ASYNC_TYPE_WAIT; - req->count = 0; - req->func = check_async_list; - req->status = STATUS_PENDING; - - ret=wine_server_call_err(req); - } - SERVER_END_REQ; + lpOverlapped->Internal = STATUS_USER_APC; + lpOverlapped->InternalHigh = 0; + lpOverlapped->Offset = 0; + lpOverlapped->OffsetHigh = 0; - if (!ret) - SetLastError(ERROR_IO_PENDING); + if (register_async (&ovp->async)) + SetLastError(ERROR_IO_PENDING); return FALSE; } diff -ruX diffignore CVS/wine/files/file.c MW/wine/files/file.c --- CVS/wine/files/file.c Tue Jan 8 18:57:47 2002 +++ MW/wine/files/file.c Wed Jan 9 20:55:09 2002 @@ -60,6 +60,20 @@ static HANDLE dos_handles[DOS_TABLE_SIZE]; +/* Async operations struct (see file.h) */ + +static DWORD fileio_get_async_status (async_private *ovp); +static DWORD fileio_get_async_count (async_private *ovp); +static void fileio_set_async_status (async_private *ovp, DWORD status); +static void CALLBACK fileio_call_completion_func (ULONG_PTR data); + +async_ops fileio_async_ops = +{ + fileio_get_async_status, /* get_status */ + fileio_set_async_status, /* set_status */ + fileio_get_async_count, /* get_count */ + fileio_call_completion_func /* call_completion */ +}; /*********************************************************************** * FILE_ConvertOFMode @@ -1244,35 +1258,13 @@ if(lpTransferred) *lpTransferred = lpOverlapped->InternalHigh; - SetLastError(lpOverlapped->Internal); + /* This is what the function should return accirding to MSDN specs */ + if ( lpOverlapped->Internal == STATUS_PENDING ) + SetLastError (ERROR_IO_INCOMPLETE); return (r==WAIT_OBJECT_0); } - -/*********************************************************************** - * FILE_StartAsync (INTERNAL) - * - * type==ASYNC_TYPE_NONE means cancel the indicated overlapped operation - * lpOverlapped==NULL means all overlappeds match - */ -BOOL FILE_StartAsync(HANDLE hFile, LPOVERLAPPED lpOverlapped, DWORD type, DWORD count, DWORD status) -{ - BOOL ret; - SERVER_START_REQ(register_async) - { - req->handle = hFile; - req->overlapped = lpOverlapped; - req->type = type; - req->count = count; - req->func = check_async_list; - req->status = status; - ret = wine_server_call( req ); - } - SERVER_END_REQ; - return !ret; -} - /*********************************************************************** * CancelIo (KERNEL32.@) */ @@ -1282,21 +1274,60 @@ TRACE("handle = %x\n",handle); - ovp = NtCurrentTeb()->pending_list; - while(ovp) + + for (ovp = NtCurrentTeb()->pending_list; ovp; ovp = t) { t = ovp->next; - if(FILE_StartAsync(handle, ovp->lpOverlapped, ovp->type, 0, STATUS_CANCELLED)) + if (ovp->handle != handle) continue; + + SERVER_START_REQ ( cancel_async ) { - TRACE("overlapped = %p\n",ovp->lpOverlapped); - finish_async(ovp, STATUS_CANCELLED); + req->handle = handle; + req->type = ovp->type; + req->overlapped = ovp; + wine_server_call ( req ); + if (reply->cancelled) + { + TRACE ("cancelling request: %p\n", ovp); + ovp->ops->set_status ( ovp, STATUS_CANCELLED ); + finish_async (ovp); + } } - ovp = t; + SERVER_END_REQ; } WaitForMultipleObjectsEx(0,NULL,FALSE,1,TRUE); return TRUE; } +static DWORD fileio_get_async_status (async_private *ovp) +{ + return ((async_fileio*) ovp)->lpOverlapped->Internal; +} + +static void fileio_set_async_status (async_private *ovp, DWORD status) +{ + ((async_fileio*) ovp)->lpOverlapped->Internal = status; +} + +static DWORD fileio_get_async_count (async_private *ovp) +{ + async_fileio *fileio = (async_fileio*) ovp; + DWORD ret = fileio->count - fileio->lpOverlapped->InternalHigh; + return (ret < 0 ? 0 : ret); +} + +static void CALLBACK fileio_call_completion_func (ULONG_PTR data) +{ + async_fileio *ovp = (async_fileio*) data; + TRACE ("data: %p\n", ovp); + + if (ovp->completion_func) + ovp->completion_func(ovp->lpOverlapped->Internal, + ovp->lpOverlapped->InternalHigh, + ovp->lpOverlapped); + HeapFree(GetProcessHeap(), 0, ovp); +} + /*********************************************************************** * FILE_AsyncReadService (INTERNAL) * @@ -1305,18 +1336,19 @@ */ static void FILE_AsyncReadService(async_private *ovp) { - LPOVERLAPPED lpOverlapped = ovp->lpOverlapped; + async_fileio *fileio = (async_fileio*) ovp; + LPOVERLAPPED lpOverlapped = fileio->lpOverlapped; int result, r; int already = lpOverlapped->InternalHigh; - TRACE("%p %p\n", lpOverlapped, ovp->buffer ); + TRACE("%p %p\n", lpOverlapped, fileio->buffer ); /* check to see if the data is ready (non-blocking) */ - result = pread (ovp->fd, &ovp->buffer[already], ovp->count - already, + result = pread (ovp->fd, &fileio->buffer[already], fileio->count - already, OVERLAPPED_OFFSET (lpOverlapped) + already); if ((result < 0) && (errno == ESPIPE)) - result = read (ovp->fd, &ovp->buffer[already], ovp->count - already); + result = read (ovp->fd, &fileio->buffer[already], fileio->count - already); if ( (result<0) && ((errno == EAGAIN) || (errno == EINTR))) { @@ -1334,9 +1366,9 @@ } lpOverlapped->InternalHigh += result; - TRACE("read %d more bytes %ld/%d so far\n",result,lpOverlapped->InternalHigh,ovp->count); + TRACE("read %d more bytes %ld/%d so far\n",result,lpOverlapped->InternalHigh,fileio->count); - if(lpOverlapped->InternalHigh < ovp->count) + if(lpOverlapped->InternalHigh < fileio->count) r = STATUS_PENDING; else r = STATUS_SUCCESS; @@ -1352,7 +1384,7 @@ LPOVERLAPPED overlapped, LPOVERLAPPED_COMPLETION_ROUTINE lpCompletionRoutine) { - async_private *ovp; + async_fileio *ovp; int fd; TRACE("file %d to buf %p num %ld %p func %p\n", @@ -1369,10 +1401,11 @@ if(fd<0) { TRACE("Couldn't get FD\n"); + SetLastError( ERROR_INVALID_HANDLE ); return FALSE; } - ovp = (async_private *) HeapAlloc(GetProcessHeap(), 0, sizeof (async_private)); + ovp = (async_fileio *) HeapAlloc(GetProcessHeap(), 0, sizeof (async_fileio)); if(!ovp) { TRACE("HeapAlloc Failed\n"); @@ -1380,30 +1413,23 @@ close(fd); return FALSE; } + + ovp->async.ops = &fileio_async_ops; + ovp->async.handle = hFile; + ovp->async.fd = fd; + ovp->async.type = ASYNC_TYPE_READ; + ovp->async.func = FILE_AsyncReadService; + ovp->lpOverlapped = overlapped; - ovp->count = bytesToRead; ovp->completion_func = lpCompletionRoutine; - ovp->func = FILE_AsyncReadService; ovp->buffer = buffer; - ovp->fd = fd; - ovp->type = ASYNC_TYPE_READ; - ovp->handle = hFile; - - /* hook this overlap into the pending async operation list */ - ovp->next = NtCurrentTeb()->pending_list; - ovp->prev = NULL; - if(ovp->next) - ovp->next->prev = ovp; - NtCurrentTeb()->pending_list = ovp; + ovp->count = bytesToRead; - if ( !FILE_StartAsync(hFile, overlapped, ASYNC_TYPE_READ, bytesToRead, STATUS_PENDING) ) - { - /* FIXME: remove async_private and release memory */ - ERR("FILE_StartAsync failed\n"); - return FALSE; - } + /* Tell register_async that this is a new request */ + overlapped->Internal = STATUS_USER_APC; - return TRUE; + /* The request is hooked into the pending list by register_async() */ + return register_async (&ovp->async); } /*********************************************************************** @@ -1413,7 +1439,6 @@ LPOVERLAPPED overlapped, LPOVERLAPPED_COMPLETION_ROUTINE lpCompletionRoutine) { - overlapped->Internal = STATUS_PENDING; overlapped->InternalHigh = 0; return FILE_ReadFileEx(hFile,buffer,bytesToRead,overlapped,lpCompletionRoutine); } @@ -1470,41 +1495,21 @@ return FALSE; } - /* see if we can read some data already (this shouldn't block) */ - result = pread( unix_handle, buffer, bytesToRead, OVERLAPPED_OFFSET(overlapped) ); - if ((result < 0) && (errno == ESPIPE)) - result = read( unix_handle, buffer, bytesToRead ); close(unix_handle); - - if(result<0) - { - if( (errno!=EAGAIN) && (errno!=EINTR) && - ((errno != EFAULT) || IsBadWritePtr( buffer, bytesToRead )) ) - { - FILE_SetDosError(); - return FALSE; - } - else - result = 0; - } + overlapped->InternalHigh = 0; - /* if we read enough to keep the app happy, then return now */ - if(result>=bytesToRead) - { - *bytesRead = result; - return TRUE; - } + if(!FILE_ReadFileEx(hFile, buffer, bytesToRead, overlapped, FILE_OverlappedComplete)) + return FALSE; - /* at last resort, do an overlapped read */ - overlapped->Internal = STATUS_PENDING; - overlapped->InternalHigh = result; - if(!FILE_ReadFileEx(hFile, buffer, bytesToRead, overlapped, FILE_OverlappedComplete)) + if ( !GetOverlappedResult (hFile, overlapped, bytesRead, FALSE) ) + { + if ( GetLastError() == ERROR_IO_INCOMPLETE ) + SetLastError ( ERROR_IO_PENDING ); return FALSE; + } - /* fail on return, with ERROR_IO_PENDING */ - SetLastError(ERROR_IO_PENDING); - return FALSE; + return TRUE; } else if ( FD_TYPE (type) == FD_TYPE_CONSOLE ) @@ -1554,18 +1559,19 @@ */ static void FILE_AsyncWriteService(struct async_private *ovp) { - LPOVERLAPPED lpOverlapped = ovp->lpOverlapped; + async_fileio *fileio = (async_fileio *) ovp; + LPOVERLAPPED lpOverlapped = fileio->lpOverlapped; int result, r; int already = lpOverlapped->InternalHigh; - TRACE("(%p %p)\n",lpOverlapped,ovp->buffer); + TRACE("(%p %p)\n",lpOverlapped,fileio->buffer); /* write some data (non-blocking) */ - result = pwrite(ovp->fd, &ovp->buffer[already], ovp->count - already, + result = pwrite(ovp->fd, &fileio->buffer[already], fileio->count - already, OVERLAPPED_OFFSET (lpOverlapped) + already); if ((result < 0) && (errno == ESPIPE)) - result = write(ovp->fd, &ovp->buffer[already], ovp->count - already); + result = write(ovp->fd, &fileio->buffer[already], fileio->count - already); if ( (result<0) && ((errno == EAGAIN) || (errno == EINTR))) { @@ -1582,9 +1588,9 @@ lpOverlapped->InternalHigh += result; - TRACE("wrote %d more bytes %ld/%d so far\n",result,lpOverlapped->InternalHigh,ovp->count); + TRACE("wrote %d more bytes %ld/%d so far\n",result,lpOverlapped->InternalHigh,fileio->count); - if(lpOverlapped->InternalHigh < ovp->count) + if(lpOverlapped->InternalHigh < fileio->count) r = STATUS_PENDING; else r = STATUS_SUCCESS; @@ -1600,7 +1606,8 @@ LPOVERLAPPED overlapped, LPOVERLAPPED_COMPLETION_ROUTINE lpCompletionRoutine) { - async_private *ovp; + async_fileio *ovp; + int fd; TRACE("file %d to buf %p num %ld %p func %p stub\n", hFile, buffer, bytesToWrite, overlapped, lpCompletionRoutine); @@ -1611,45 +1618,39 @@ return FALSE; } - overlapped->Internal = STATUS_PENDING; - overlapped->InternalHigh = 0; - - if (!FILE_StartAsync(hFile, overlapped, ASYNC_TYPE_WRITE, bytesToWrite, STATUS_PENDING )) + fd = FILE_GetUnixHandle( hFile, GENERIC_WRITE ); + if(fd<0) { - TRACE("FILE_StartAsync failed\n"); + TRACE("Couldn't get FD\n"); + SetLastError( ERROR_INVALID_HANDLE ); return FALSE; } - ovp = (async_private*) HeapAlloc(GetProcessHeap(), 0, sizeof (async_private)); + ovp = (async_fileio*) HeapAlloc(GetProcessHeap(), 0, sizeof (async_fileio)); if(!ovp) { TRACE("HeapAlloc Failed\n"); SetLastError(ERROR_NOT_ENOUGH_MEMORY); + close(fd); return FALSE; } + + ovp->async.ops = &fileio_async_ops; + ovp->async.handle = hFile; + ovp->async.fd = FILE_GetUnixHandle( hFile, GENERIC_WRITE ); + ovp->async.type = ASYNC_TYPE_WRITE; + ovp->async.func = FILE_AsyncWriteService; + ovp->lpOverlapped = overlapped; - ovp->func = FILE_AsyncWriteService; ovp->buffer = (LPVOID) buffer; - ovp->count = bytesToWrite; ovp->completion_func = lpCompletionRoutine; - ovp->fd = FILE_GetUnixHandle( hFile, GENERIC_WRITE ); - ovp->type = ASYNC_TYPE_WRITE; - ovp->handle = hFile; - - if(ovp->fd <0) - { - HeapFree(GetProcessHeap(), 0, ovp); - return FALSE; - } + ovp->count = bytesToWrite; - /* hook this overlap into the pending async operation list */ - ovp->next = NtCurrentTeb()->pending_list; - ovp->prev = NULL; - if(ovp->next) - ovp->next->prev = ovp; - NtCurrentTeb()->pending_list = ovp; + /* Tell register_async that this is a new request */ + overlapped->Internal = STATUS_USER_APC; - return TRUE; + /* The request is hooked into the pending list by register_async() */ + return register_async (&ovp->async); } /*********************************************************************** @@ -1659,7 +1660,6 @@ LPOVERLAPPED overlapped, LPOVERLAPPED_COMPLETION_ROUTINE lpCompletionRoutine) { - overlapped->Internal = STATUS_PENDING; overlapped->InternalHigh = 0; return FILE_WriteFileEx(hFile, buffer, bytesToWrite, overlapped, lpCompletionRoutine); @@ -1693,44 +1693,20 @@ return FALSE; } - /* see if we can write some data already (this shouldn't block) */ - - result = pwrite( unix_handle, buffer, bytesToWrite, OVERLAPPED_OFFSET (overlapped) ); - if ((result < 0) && (errno == ESPIPE)) - result = write( unix_handle, buffer, bytesToWrite ); - close(unix_handle); - - if(result<0) - { - if( (errno!=EAGAIN) && (errno!=EINTR) && - ((errno != EFAULT) || IsBadReadPtr( buffer, bytesToWrite )) ) - { - FILE_SetDosError(); - return FALSE; - } - else - result = 0; - } - - /* if we wrote enough to keep the app happy, then return now */ - if(result>=bytesToWrite) - { - *bytesWritten = result; - return TRUE; - } - - /* at last resort, do an overlapped read */ - overlapped->Internal = STATUS_PENDING; - overlapped->InternalHigh = result; + overlapped->InternalHigh = 0; - if(!FILE_WriteFileEx(hFile, buffer, bytesToWrite, overlapped, FILE_OverlappedComplete)) + if (!FILE_WriteFileEx(hFile, buffer, bytesToWrite, overlapped, FILE_OverlappedComplete)) return FALSE; + + if ( !GetOverlappedResult (hFile, overlapped, bytesWritten, FALSE) ) + { + if ( GetLastError() == ERROR_IO_INCOMPLETE ) + SetLastError ( ERROR_IO_PENDING ); + return FALSE; + } - /* fail on return, with ERROR_IO_PENDING */ - SetLastError(ERROR_IO_PENDING); - return FALSE; - + return TRUE; } else if ( FD_TYPE (type) == FD_TYPE_CONSOLE ) diff -ruX diffignore CVS/wine/include/file.h MW/wine/include/file.h --- CVS/wine/include/file.h Wed Jan 9 21:27:50 2002 +++ MW/wine/include/file.h Wed Jan 9 20:25:05 2002 @@ -33,23 +33,46 @@ /* overlapped private structure */ struct async_private; -typedef void (*async_handler)(struct async_private *ovp); + +typedef void (*async_handler) (struct async_private *ovp); +typedef void CALLBACK (*async_call_completion_func) (ULONG_PTR data); +typedef DWORD (*async_get_status) (struct async_private *ovp); +typedef DWORD (*async_get_count) (struct async_private *ovp); +typedef void (*async_set_status) (struct async_private *ovp, DWORD status); + +typedef struct async_ops +{ + async_get_status get_status; + async_set_status set_status; + async_get_count get_count; + async_call_completion_func call_completion; +} async_ops; + +extern async_ops fileio_async_ops; + typedef struct async_private { - LPOVERLAPPED lpOverlapped; - HANDLE handle; - int fd; - char *buffer; - async_handler func; - int count; - int type; - LPOVERLAPPED_COMPLETION_ROUTINE completion_func; - struct async_private *next; - struct async_private *prev; + async_ops *ops; + HANDLE handle; + int fd; + int type; + async_handler func; + struct async_private *next; + struct async_private *prev; } async_private; -extern void WINAPI check_async_list(LPOVERLAPPED ov, DWORD status); -extern void finish_async(struct async_private *ovp, DWORD status); +typedef struct async_fileio +{ + async_private async; + LPOVERLAPPED lpOverlapped; + LPOVERLAPPED_COMPLETION_ROUTINE completion_func; + char *buffer; + int count; +} async_fileio; + +extern BOOL register_async (async_private *ovp); +extern void WINAPI check_async_list(async_private *ovp, DWORD status); +extern void finish_async(async_private *ovp); /* locale-independent case conversion */ inline static char FILE_tolower( char c ) @@ -83,7 +106,6 @@ DWORD attributes, HANDLE template, BOOL fail_read_only, UINT drive_type ); extern HANDLE FILE_CreateDevice( int client_id, DWORD access, LPSECURITY_ATTRIBUTES sa ); -extern BOOL FILE_StartAsync(HANDLE handle, LPOVERLAPPED lpOverlapped, DWORD type, DWORD count, DWORD status); extern LONG WINAPI WIN16_hread(HFILE16,SEGPTR,LONG); diff -ruX diffignore CVS/wine/scheduler/synchro.c MW/wine/scheduler/synchro.c --- CVS/wine/scheduler/synchro.c Tue Jan 8 18:53:41 2002 +++ MW/wine/scheduler/synchro.c Wed Jan 9 20:27:16 2002 @@ -36,27 +36,60 @@ } } -static void CALLBACK call_completion_routine(ULONG_PTR data) +/*********************************************************************** + * register_async (INTERNAL) + * + * Manipulate async request queues. + */ +BOOL register_async (async_private *ovp) { - async_private* ovp = (async_private*)data; + BOOL ret; + DWORD status = ovp->ops->get_status (ovp); - ovp->completion_func(ovp->lpOverlapped->Internal, - ovp->lpOverlapped->InternalHigh, - ovp->lpOverlapped); - ovp->completion_func=NULL; - HeapFree(GetProcessHeap(), 0, ovp); -} + /* STATUS_USER_APC means this is a new request */ + /* STATUS_PENDING means this request is rescheduled (IO incomplete) */ -void finish_async(async_private *ovp, DWORD status) -{ - ovp->lpOverlapped->Internal=status; + if (status == STATUS_USER_APC) + { + status = STATUS_PENDING; + ovp->ops->set_status (ovp, status); + + /* hook this overlap into the pending async operation list */ + ovp->next = NtCurrentTeb()->pending_list; + ovp->prev = NULL; + if (ovp->next) ovp->next->prev = ovp; + NtCurrentTeb()->pending_list = ovp; + } - /* call ReadFileEx/WriteFileEx's overlapped completion function */ - if(ovp->completion_func) + /* The server call will destroy all except PENDING requests */ + SERVER_START_REQ(register_async) { - QueueUserAPC(call_completion_routine,GetCurrentThread(),(ULONG_PTR)ovp); + req->handle = ovp->handle; + req->overlapped = ovp; + req->type = ovp->type; + req->count = ovp->ops->get_count (ovp); + req->func = check_async_list; + req->status = status; + ret = wine_server_call( req ); } + SERVER_END_REQ; + + if (ret) ovp->ops->set_status ( ovp, GetLastError() ); + + if ( ovp->ops->get_status (ovp) != STATUS_PENDING ) + finish_async (ovp); + return !ret; +} + +/*********************************************************************** + * finish_async (INTERNAL) + * + * Called after completion or cancellation of an async request. + * Note: The completion status must be set already. + */ +void finish_async(async_private *ovp) +{ /* remove it from the active list */ if(ovp->prev) ovp->prev->next = ovp->next; @@ -70,7 +103,9 @@ ovp->prev=NULL; close(ovp->fd); - if(!ovp->completion_func) HeapFree(GetProcessHeap(), 0, ovp); + + /* Queue completion function unconditionally */ + QueueUserAPC(ovp->ops->call_completion, GetCurrentThread(), (ULONG_PTR)ovp); } /*********************************************************************** @@ -78,30 +113,33 @@ * * Process a status event from the server. */ -void WINAPI check_async_list(LPOVERLAPPED overlapped, DWORD status) +void WINAPI check_async_list(async_private *asp, DWORD status) { - async_private *ovp; + async_private *ovp = NULL; + DWORD ovp_status; /* fprintf(stderr,"overlapped %p status %x\n",overlapped,status); */ - for(ovp = NtCurrentTeb()->pending_list; ovp; ovp = ovp->next) - if(ovp->lpOverlapped == overlapped) - break; + for( ovp = NtCurrentTeb()->pending_list; ovp && ovp != asp; ovp = ovp->next ); - if(!ovp) - return; + if(!ovp) return; if(status != STATUS_ALERTED) - ovp->lpOverlapped->Internal = status; + { + ovp_status = status; + ovp->ops->set_status (ovp, status); + } + else ovp_status = ovp->ops->get_status (ovp); - if(ovp->lpOverlapped->Internal==STATUS_PENDING) - { + if( ovp_status == STATUS_PENDING) + { ovp->func(ovp); - FILE_StartAsync(ovp->handle, ovp->lpOverlapped, ovp->type, 0, ovp->lpOverlapped->Internal); - } + register_async (ovp); + ovp_status = ovp->ops->get_status (ovp); + } - if(ovp->lpOverlapped->Internal!=STATUS_PENDING) - finish_async(ovp,ovp->lpOverlapped->Internal); + if( ovp_status != STATUS_PENDING ) + finish_async(ovp); } diff -ruX diffignore CVS/wine/server/async.c MW/wine/server/async.c --- CVS/wine/server/async.c Wed Jan 2 12:50:04 2002 +++ MW/wine/server/async.c Wed Jan 9 20:04:53 2002 @@ -177,3 +177,28 @@ release_object(obj); } +DECL_HANDLER(cancel_async) +{ + struct object *obj; + + if (!(obj = get_handle_obj( current->process, req->handle, 0, NULL)) ) + return; + + if(obj->ops->queue_async) + { + struct async_queue *q = obj->ops->queue_async(obj, NULL, req->type, 0); + struct async *async; + + async = find_async(q, current, req->overlapped); + if (async) + { + destroy_async(async); + reply->cancelled = 1; + set_select_events(obj,obj->ops->get_poll_events(obj)); + } + else reply->cancelled = 0; + } + + release_object(obj); +} + diff -ruX diffignore CVS/wine/server/protocol.def MW/wine/server/protocol.def --- CVS/wine/server/protocol.def Tue Jan 8 18:58:28 2002 +++ MW/wine/server/protocol.def Wed Jan 9 19:52:33 2002 @@ -1535,7 +1535,7 @@ #define SERIALINFO_SET_ERROR 0x04 -/* Create/Destroy an async I/O */ +/* Create / reschedule an async I/O */ @REQ(register_async) handle_t handle; /* handle to comm port, socket or file */ void* func; @@ -1549,6 +1549,14 @@ #define ASYNC_TYPE_WRITE 0x02 #define ASYNC_TYPE_WAIT 0x03 +/* Cancel an async I/O */ +(a)REQ(cancel_async) + handle_t handle; + int type; + void* overlapped; +(a)REPLY + int cancelled; +(a)END /* Create a named pipe */ @REQ(create_named_pipe)

Jan. 9, 2002

Re: Is W really UTF-16?

by Ove Kaaven

On Wed, 9 Jan 2002, Bill Medland wrote: > While I was working on the DrawText functions over the past many months I > started wondering about when it would fail. (I'm pedantic and such things > fascinate me!). The main concern I have is how to walk a W string > correctly. For example while "ellipsifying" text we will need to "move the > pointer to the previous character" which is currently done by decrementing > the pointer by 1. But from what I currently understand that won't work if > there are surrogate pairs. If you're concerned about that, surrogate pairs are the least of your worries. You should also be concerned about Unicode combining (or composite) characters. I think they might be identified with ctypes C3_NONSPACING and C3_DIACRITIC and that kind of stuff... > 1. Does anyone know under what circumstances CharNextW isn't +1 (apart from > when pointing at the terminating 0) Have you tried low surrogate followed by high surrogate, on a Microsoft OS recent enough that Microsoft *might* have thought about preparing it for dealing with surrogates? > 2. Is e.g. XP really using UTF-16 or is it actually still UCS2? I don't know. But it probably ought to be UTF16. > 3 Have we thought about how we should handle walking along a W string (in a > fashion that doesn't reduced the speed to a crawl). I guess that in the > short term I am expecting some sort of macro or inline. With p++, perhaps? There aren't very many circumstances where that is going to be a problem (where unicode composite characters are not also), is there?

Jan. 9, 2002

Re: Babystep: Testing framework

by Alexandre Julliard

Francois Gouget <fgouget(a)free.fr> writes: > I think we should definitely be able to check the test output against > a reference version. The idea is that there is no test output (except for debugging); everything is checked right inside the test. > Why can't you have one executable per test? Or is there some > test/check confusion here? One check per executable, or perl file for > that matter would be crazy. Because a Winelib executable is too heavyweight; you need a spec file, multiple object files, a lot of disk space, etc. You cannot realistically have hundreds of them without creating a horrible mess. -- Alexandre Julliard julliard(a)winehq.com

Jan. 9, 2002

Re: We *really* need a development model change !

by Andriy Palamarchuk

--- Alexandre Julliard <julliard(a)winehq.com> wrote: > Andriy Palamarchuk <apa3a(a)yahoo.com> writes: > > But I think implementing it now would actually make > your life easier. > This way you wouldn't have to require some magic > call to detect you > are running on Wine, you could simply rely on the > Wine makefiles to > pass you the right option. I don't see any value in information about the first TODO check in a test script. I, personally, prefer using "magic call". One more plus of autodetection - I can run the same Windows test binary on Windows and on Wine without any troubles. But I'd like to use your idea in other place. I feel uncomfortable when test does not show anything. It is not clear whether it runs any checks, whether my last changes were executed. I suggest to have switch which turns off output. By default the test will print some statistics on successful run - number of tests running, number of TODO tests, time of execution - anything else? Make script can run test with output turned off. Andriy Palamarchuk __________________________________________________ Do You Yahoo!? Send FREE video emails in Yahoo! Mail! http://promo.yahoo.com/videomail/

Jan. 9, 2002

Re: Babystep: Testing framework

by Francois Gouget

On Wed, 9 Jan 2002, Jeremy White wrote: [...] > > Also it seems the consensus is that > > we should simply check the exit status of the test, not compare the > > test output. > > Sorry, this wasn't clear. That is the default behavior. > I preserved the pattern and reference checking as an optional > addition. I think we should definitely be able to check the test output against a reference version. [...] > >> 2. I really don't like that a C/Winelib > >> test requires its own directory > >> (see samples/sample3). However, AFAICT, > >> that was the only way to create a simple > >> and clean build environment for the Winelib app. You don't need to. You can do things like winemaker does and then you can put as many executables and libraries as you like in a single directory. But you may not be able to reuse the standard @MAKE_PROG_RULES(a). > > That's the problem with C tests. You cannot have one executable for > > each test, at least not with Winelib, so we need a way to build > > multiple tests inside a single executable. Not very hard, but there's > > a bit of work involved. Why can't you have one executable per test? Or is there some test/check confusion here? One check per executable, or perl file for that matter would be crazy. -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ La terre est une bêta...

Jan. 9, 2002