- Wine-Devel - list.winehq.org

Problems with wizard code in COMCTL32
by Filip Navara July 21, 2004

July 21, 2004

Hi, I took a quick look at the property sheet code in COMCTL32 in order to get the Mozilla Firefox setup working. The attached patch fixes some of the bugs, but the code is too messy for me to understand and fix the other problems: - Line ~1532. When PSH_USEHBMWATERMARK or PSH_USEHBMHEADER is set, the property sheet header is modified and the watermark/header bitmap is loaded using CreateMappedBitmap and injected into the hbmWatermark/hbmHeader field of the global property sheet info structure. This is done for every page in the property sheet. Since the names and bitmaps are stored as union at the same place in memory, on the second page the bitmap is already loaded and the original name is already lost, but the PSH_USEHBM* flag is still set and so the code tries to load the bitmap again passing the bitmap handle instead of the resource ID. I think the code for loading the watermark and header bitmaps should be moved to some other initialization routine, because AFAIK these bitmaps are identical for the whole wizard and not different on per page basis. - The watermark bitmap is displayed under the "dialog" with the controls on the property sheet. The "dialog" itself isn't transparent and so most of the watermark bitmap isn't visible. I thought that setting the window class background color of the "dialog" with controls would solve that, but it seems that the native version of COMCTL32 (from Windows XP) subclasses it. - Caption of the wizard window is never set. If anyone wants to look at these problems it would be great. Regards, Filip Changelog: - Properly calculate caption length when converting multi-byte string to wide character one. - Wizard contents shouldn't be moved by the size of watermark bitmap. - Fix PSH_WIZARD97_[NEW/OLD] checks. - Blit the watermark even if PSH_USEHBMWATERMARK flag isn't set and blit only the part that fits into the dialog. - Fix CreateMappedBitmap for palette-less bitmaps. Index: comctl32/propsheet.c =================================================================== --- comctl32/propsheet.c (revision 1) +++ comctl32/propsheet.c (working copy) @@ -173,7 +173,7 @@ static BOOL PROPSHEET_RemovePage(HWND hwndDlg, int index, HPROPSHEETPAGE hpage); -static void PROPSHEET_CleanUp(); +static void PROPSHEET_CleanUp(HWND hwndDlg); static int PROPSHEET_GetPageIndex(HPROPSHEETPAGE hpage, PropSheetInfo* psInfo); static void PROPSHEET_SetWizButtons(HWND hwndDlg, DWORD dwFlags); static PADDING_INFO PROPSHEET_GetPaddingInfoWizard(HWND hwndDlg, const PropSheetInfo* psInfo); @@ -308,10 +308,9 @@ { if (HIWORD(lppsh->pszCaption)) { - int len = strlen(lppsh->pszCaption); + int len = MultiByteToWideChar(CP_ACP, 0, lppsh->pszCaption, -1, NULL, 0); psInfo->ppshheader.pszCaption = HeapAlloc( GetProcessHeap(), 0, (len+1)*sizeof (WCHAR) ); MultiByteToWideChar(CP_ACP, 0, lppsh->pszCaption, -1, (LPWSTR) psInfo->ppshheader.pszCaption, len+1); - /* strcpy( (char *)psInfo->ppshheader.pszCaption, lppsh->pszCaption ); */ } } psInfo->nPages = lppsh->nPages; @@ -361,7 +360,7 @@ psInfo->ppshheader.pszCaption = NULL; else { - if (!(lppsh->dwFlags & INTRNL_ANY_WIZARD) && HIWORD(lppsh->pszCaption)) + if (HIWORD(lppsh->pszCaption)) { int len = strlenW(lppsh->pszCaption); psInfo->ppshheader.pszCaption = HeapAlloc( GetProcessHeap(), 0, (len+1)*sizeof(WCHAR) ); @@ -1511,7 +1510,6 @@ psInfo->proppage[index].hwndPage = hwndPage; if (psInfo->ppshheader.dwFlags & INTRNL_ANY_WIZARD) { - int offsetx = 0; int offsety = 0; HWND hwndChild; RECT r; @@ -1532,11 +1530,10 @@ pageWidth, pageHeight, padding.x, padding.y); /* If there is a watermark, offset the dialog items */ - if ( (psInfo->ppshheader.dwFlags & (PSH_WATERMARK | PSH_WIZARD97_NEW | PSH_WIZARD97_OLD)) && + if ( (psInfo->ppshheader.dwFlags & (PSH_WIZARD97_NEW | PSH_WIZARD97_OLD)) && + (psInfo->ppshheader.dwFlags & PSH_WATERMARK) && ((index == 0) || (index == psInfo->nPages - 1)) ) { - BITMAP bm; - /* if PSH_USEHBMWATERMARK is not set, load the resource from pszbmWatermark and put the HBITMAP in hbmWatermark. Thus all the rest of the code always considers hbmWatermark as valid. */ @@ -1545,13 +1542,10 @@ ((PropSheetInfo *)psInfo)->ppshheader.u4.hbmWatermark = CreateMappedBitmap(ppshpage->hInstance, (INT)psInfo->ppshheader.u4.pszbmWatermark, 0, NULL, 0); } - - /* Compute the offset x */ - GetObjectA(psInfo->ppshheader.u4.hbmWatermark, sizeof(BITMAP), (LPVOID)&bm); - offsetx = bm.bmWidth; } - if (psInfo->ppshheader.dwFlags & (PSH_HEADER | PSH_WIZARD97_NEW | PSH_WIZARD97_OLD)) + if (psInfo->ppshheader.dwFlags & (PSH_WIZARD97_NEW | PSH_WIZARD97_OLD) && + psInfo->ppshheader.dwFlags & PSH_HEADER) { /* Same behavior as for watermarks */ if (!(psInfo->ppshheader.dwFlags & PSH_USEHBMHEADER)) @@ -1567,21 +1561,10 @@ offsety = (ppshpage->dwFlags & PSP_HIDEHEADER)?0:r.bottom + 1; } - hwndChild = GetWindow(hwndPage, GW_CHILD); - while((offsetx!=0) && (hwndChild)) { - GetWindowRect(hwndChild, &r); - MapWindowPoints(0, hwndPage, (LPPOINT)&r, 2); - OffsetRect(&r, -offsetx, 0); - SetWindowPos(hwndChild, 0, r.left, r.top, - 0, 0, SWP_NOSIZE | SWP_NOZORDER); - - hwndChild = GetWindow(hwndChild, GW_HWNDNEXT); - } - SetWindowPos(hwndPage, HWND_TOP, - rc.left + padding.x/2 + offsetx, + rc.left + padding.x/2, rc.top + padding.y/2 + offsety, - pageWidth - offsetx, pageHeight - offsety, 0); + pageWidth, pageHeight - offsety, 0); } else { /* @@ -2926,7 +2909,8 @@ hOldPal = SelectPalette(hdc, psInfo->ppshheader.hplWatermark, FALSE); if ( (!(ppshpage->dwFlags & PSP_HIDEHEADER)) && - (psInfo->ppshheader.dwFlags & (PSH_WIZARD97_NEW | PSH_HEADER)) ) + (psInfo->ppshheader.dwFlags & (PSH_WIZARD97_OLD | PSH_WIZARD97_NEW)) && + (psInfo->ppshheader.dwFlags & PSH_HEADER) ) { RECT rzone; HWND hwndLineHeader = GetDlgItem(hwnd, IDC_SUNKEN_LINEHEADER); @@ -2982,35 +2966,33 @@ } if ( ((psInfo->active_page == 0) || (psInfo->active_page == psInfo->nPages - 1)) && - (psInfo->ppshheader.dwFlags & (PSH_WIZARD97_NEW | PSH_WATERMARK)) ) + (psInfo->ppshheader.dwFlags & (PSH_WIZARD97_OLD | PSH_WIZARD97_NEW)) && + (psInfo->ppshheader.dwFlags & PSH_WATERMARK) ) { - if (psInfo->ppshheader.dwFlags & PSH_USEHBMWATERMARK ) - { - HWND hwndLine = GetDlgItem(hwnd, IDC_SUNKEN_LINE); + HWND hwndLine = GetDlgItem(hwnd, IDC_SUNKEN_LINE); - GetClientRect(hwndLine, &r); - MapWindowPoints(hwndLine, hwnd, (LPPOINT) &r, 2); + GetClientRect(hwndLine, &r); + MapWindowPoints(hwndLine, hwnd, (LPPOINT) &r, 2); - GetObjectA(psInfo->ppshheader.u4.hbmWatermark, sizeof(BITMAP), (LPVOID)&bm); - hbmp = SelectObject(hdcSrc, psInfo->ppshheader.u4.hbmWatermark); + GetObjectA(psInfo->ppshheader.u4.hbmWatermark, sizeof(BITMAP), (LPVOID)&bm); + hbmp = SelectObject(hdcSrc, psInfo->ppshheader.u4.hbmWatermark); - BitBlt(hdc, 0, offsety, bm.bmWidth, bm.bmHeight, - hdcSrc, 0, 0, SRCCOPY); + BitBlt(hdc, 0, offsety, min(bm.bmWidth, r.right), + min(bm.bmHeight, r.bottom), hdcSrc, 0, 0, SRCCOPY); - /* If the bitmap is not big enough, fill the remaining area - with the color of pixel (0,0) of bitmap - see MSDN */ - if (r.top > bm.bmHeight) { - r.bottom = r.top - 1; - r.top = bm.bmHeight; - r.left = 0; - r.right = bm.bmWidth; - hbr = CreateSolidBrush(GetPixel(hdcSrc, 0, 0)); - FillRect(hdc, &r, hbr); - DeleteObject(hbr); - } - - SelectObject(hdcSrc, hbmp); + /* If the bitmap is not big enough, fill the remaining area + with the color of pixel (0,0) of bitmap - see MSDN */ + if (r.top > bm.bmHeight) { + r.bottom = r.top - 1; + r.top = bm.bmHeight; + r.left = 0; + r.right = bm.bmWidth; + hbr = CreateSolidBrush(GetPixel(hdcSrc, 0, 0)); + FillRect(hdc, &r, hbr); + DeleteObject(hbr); } + + SelectObject(hdcSrc, hbmp); } if (psInfo->ppshheader.dwFlags & PSH_USEHPLWATERMARK) Index: comctl32/commctrl.c =================================================================== --- comctl32/commctrl.c (revision 1) +++ comctl32/commctrl.c (working copy) @@ -862,7 +862,12 @@ if (lpBitmap == NULL) return 0; - nColorTableSize = (1 << lpBitmap->biBitCount); + if (lpBitmap->biSize >= sizeof(BITMAPINFOHEADER) && lpBitmap->biClrUsed) + nColorTableSize = lpBitmap->biClrUsed; + else if (lpBitmap->biBitCount <= 8) + nColorTableSize = (1 << lpBitmap->biBitCount); + else + nColorTableSize = 0; nSize = lpBitmap->biSize + nColorTableSize * sizeof(RGBQUAD); lpBitmapInfo = (LPBITMAPINFOHEADER)GlobalAlloc (GMEM_FIXED, nSize); if (lpBitmapInfo == NULL) @@ -899,7 +904,7 @@ HDC hdcDst = CreateCompatibleDC (hdcScreen); HBITMAP hbmOld = SelectObject (hdcDst, hbm); LPBYTE lpBits = (LPBYTE)(lpBitmap + 1); - lpBits += (1 << (lpBitmapInfo->biBitCount)) * sizeof(RGBQUAD); + lpBits += nColorTableSize * sizeof(RGBQUAD); StretchDIBits (hdcDst, 0, 0, nWidth, nHeight, 0, 0, nWidth, nHeight, lpBits, (LPBITMAPINFO)lpBitmapInfo, DIB_RGB_COLORS, SRCCOPY);

1 0

RFC: API db
by David Miller July 21, 2004

July 21, 2004

Hello, This is a request for comments relating to the API database at http://winedb.pavely.net. The backend database has been redesigned to improve database performance and produce more accurate results. It currently contains import/export and ordinal data for various windows versions as well as WINE. What I would like is some feedback from the WINE development community about what you would like to see in a front-end to the database. This could range from suggestions on presenting data in an easy-to-understand manner to adding additional data to the database. Current suggestions include the following: 1. Cross reference and compare windows versions and dll versions 2. List exports and imports for a given dll 3. create dependency tree for dlls 4. Determine which APIs are used by specific applications 5. Determine most used / least used APIs and dlls (pinpoint what functionality is important for many apps) 6. Compare ordinals across windows versions (helps determine when WINE should / should not export by ordinal)

2 1

Valgrind/Wine
by Zach Gorman July 21, 2004

July 21, 2004

The latest build of valgrind that supports wine is now 8 months stale, and it doesn't seem to work on the wine head. Adam Gundy did some great work on this, and it would be a shame to see it go to waste. Is anyone out there able to pick up the torch on this? Zach

1 0

MFC and RtlpWaitForCriticalSection
by jake＠mail.nnov.ru July 21, 2004

July 21, 2004

Hello, I.m trying to compile MFC with Winelib. I switched off some code in MFC ( richedit, oledb etc ) and have it a draft version compiled/linked but when trying to use it in a simple console or GUI app I have a .err: ntdll:RtlpWaitForCriticalSection section 0x78034120 "?" wait timed out in thread... message on start and the app hangs. Wine ( 04-2004 ) displays this message on LoadLibrary(MFC) in a generated wrapper code. I also compiled the similar .truncated. MFC version on win32 and all works ok there, at least for my simple apps. I.d really appreciate any hints how to debug it and why it may happen. Thanks -- Eugene

1 0

Fix DirectSound's primary buffer Volume and Pan handling
by Francois Gouget July 21, 2004

July 21, 2004

Currently the way DirectSound handles Volume and Pan is wrong: * whenever a primary buffer is created (e.g. in DirectSoundCreate()) we initialize an internal variable that sets the volume to the max and pan to the center. * we also call IDsDriverBuffer_SetVolumePan() or waveOutSetVolume() to set the system volume to that value. On my machine this has the same effect as amixer set PCM 100% * we ignore changes in the volume or balance made outside of DirectSound. So if the user does: amixer set PCM 50% And then the application calls GetVolume() it will still get 0 (i.e. max volume) This is not what happens on Windows. On Windows primary buffers don't have their internal volume. Instead SetVolume()/SetPan() and GetVolume()/GetPan() query directly the underlying volume. I hacked the dsound.c test to verify this. Basically, now it does the following: * it calls GetVolume() and GetPan() and prints the results * it sleeps for 20 seconds * during these 20 seconds, start Windows' mixer control, and change the 'Wave Sound' (translated from French) volume and/or balance. * the test then calls GetVolume() and GetPan() and prints the results again Running the test we can see that the second time DierctSound returns the new volume: dsound.c:771:Primary volume=-566 pan=0 dsound.c:776:Primary volume=-566 pan=0 --- the test sleeps here while we change the volume --- dsound.c:782:Primary volume=-1438 pan=0 I attached the dsound.c test patch so this can be cross-checked. You will note that I also played with the secondary buffers to check the effect changing the volume on a secondary buffer would have on the system volume: it has none but Wine already does the right thing (a bit surprisingly). (for secondary buffers SetVolume() has a call to IDsDriverBuffer_SetVolumePan() but that branch is not taken so we do a DSOUND_ForceRemix() instead. I'm not sure if we should ever go through this IDsDriverBuffer_SetVolumePan() branch) I also patched DirectSound to fix the problem. The one thing that bothers me is that I did not find the equivalent of an IDsDriverBuffer_GetVolumePan(). I.e. we can use the driver to set the volume but not to query it. That's pretty strange. I wonder how DirectSound gets the system volume on Windows. In Wine what I did is simply use waveOutGetVolume() on dsound->hwo. Then I pass this through DSOUND_AmpFactorToVolPan() to convert these 'AmpFactors' into the regular Vol/Pan used by DirectSound. Comment and improvement suggestions welcome. If nobody objects I'll submit this to wine-patches tomorrow. -- Francois Gouget fgouget(a)codeweavers.com Index: dlls/dsound/tests/dsound.c =================================================================== RCS file: /var/cvs/wine/dlls/dsound/tests/dsound.c,v retrieving revision 1.30 diff -u -r1.30 dsound.c --- dlls/dsound/tests/dsound.c 21 Jul 2004 03:23:13 -0000 1.30 +++ dlls/dsound/tests/dsound.c 21 Jul 2004 12:27:48 -0000 @@ -745,11 +745,11 @@ primary=NULL; ZeroMemory(&bufdesc, sizeof(bufdesc)); bufdesc.dwSize=sizeof(bufdesc); - bufdesc.dwFlags=DSBCAPS_PRIMARYBUFFER|DSBCAPS_CTRLVOLUME; + bufdesc.dwFlags=DSBCAPS_PRIMARYBUFFER|DSBCAPS_CTRLVOLUME|DSBCAPS_CTRLPAN; rc=IDirectSound_CreateSoundBuffer(dso,&bufdesc,&primary,NULL); ok(rc==DS_OK && primary!=NULL,"CreateSoundBuffer failed to create a primary buffer: 0x%lx\n",rc); if (rc==DS_OK && primary!=NULL) { - LONG vol; + LONG vol,pan; /* Try to create a second primary buffer */ /* DSOUND: Error: The primary buffer already exists. Any changes made to the buffer description will be ignored. */ @@ -765,7 +765,21 @@ ok(rc!=DS_OK,"IDirectSound_DuplicateSoundBuffer primary buffer should have failed 0x%lx\n",rc); rc=IDirectSoundBuffer_GetVolume(primary,&vol); - ok(rc==DS_OK,"GetVolume failed: 0x%lx\n",rc); + ok(rc==DS_OK,"Primary GetVolume failed: %s\n",DXGetErrorString9(rc)); + rc=IDirectSoundBuffer_GetPan(primary,&pan); + ok(rc==DS_OK,"Primary GetPan failed: %s\n",DXGetErrorString9(rc)); + trace("Primary volume=%ld pan=%ld\n",vol,pan); + rc=IDirectSoundBuffer_SetVolume(primary,vol); + ok(rc==DS_OK,"Primary SetVolume failed: %s\n",DXGetErrorString9(rc)); + rc=IDirectSoundBuffer_SetPan(primary,pan); + ok(rc==DS_OK,"Primary SetPan failed: %s\n",DXGetErrorString9(rc)); + trace("Primary volume=%ld pan=%ld\n",vol,pan); + Sleep(20000); + rc=IDirectSoundBuffer_GetVolume(primary,&vol); + ok(rc==DS_OK,"Primary GetVolume failed: %s\n",DXGetErrorString9(rc)); + rc=IDirectSoundBuffer_GetPan(primary,&pan); + ok(rc==DS_OK,"Primary GetPan failed: %s\n",DXGetErrorString9(rc)); + trace("Primary volume=%ld pan=%ld\n",vol,pan); if (winetest_interactive) { @@ -828,7 +842,7 @@ ZeroMemory(&bufdesc, sizeof(bufdesc)); bufdesc.dwSize=sizeof(bufdesc); - bufdesc.dwFlags=DSBCAPS_PRIMARYBUFFER; + bufdesc.dwFlags=DSBCAPS_PRIMARYBUFFER|DSBCAPS_CTRLVOLUME|DSBCAPS_CTRLPAN; rc=IDirectSound_CreateSoundBuffer(dso,&bufdesc,&primary,NULL); ok(rc==DS_OK && primary!=NULL,"CreateSoundBuffer failed to create a primary buffer 0x%lx\n", rc); @@ -838,7 +852,7 @@ secondary=NULL; ZeroMemory(&bufdesc, sizeof(bufdesc)); bufdesc.dwSize=sizeof(bufdesc); - bufdesc.dwFlags=DSBCAPS_GETCURRENTPOSITION2; + bufdesc.dwFlags=DSBCAPS_GETCURRENTPOSITION2|DSBCAPS_CTRLVOLUME|DSBCAPS_CTRLPAN; bufdesc.dwBufferBytes=wfx.nAvgBytesPerSec*BUFFER_LEN/1000; bufdesc.lpwfxFormat=&wfx; trace(" Testing a secondary buffer at %ldx%dx%d\n", @@ -847,6 +861,27 @@ ok(rc==DS_OK && secondary!=NULL,"CreateSoundBuffer failed to create a secondary buffer 0x%lx\n",rc); if (rc==DS_OK && secondary!=NULL) { + LONG vol,pan; + + rc=IDirectSoundBuffer_GetVolume(secondary,&vol); + ok(rc==DS_OK,"GetVolume(secondary) failed: %s\n",DXGetErrorString9(rc)); + rc=IDirectSoundBuffer_GetPan(secondary,&pan); + ok(rc==DS_OK,"GetPan(secondary) failed: %s\n",DXGetErrorString9(rc)); + trace("Secondary volume=%ld pan=%ld\n",vol,pan); + vol=-1800; + rc=IDirectSoundBuffer_SetVolume(secondary,vol); + ok(rc==DS_OK,"SetVolume(secondary) failed: %s\n",DXGetErrorString9(rc)); + rc=IDirectSoundBuffer_GetVolume(secondary,&vol); + ok(rc==DS_OK,"GetVolume(secondary) failed: %s\n",DXGetErrorString9(rc)); + rc=IDirectSoundBuffer_GetPan(secondary,&pan); + ok(rc==DS_OK,"GetPan(secondary) failed: %s\n",DXGetErrorString9(rc)); + trace("Secondary volume=%ld pan=%ld\n",vol,pan); + rc=IDirectSoundBuffer_GetVolume(primary,&vol); + ok(rc==DS_OK,"Primary GetVolume failed: %s\n",DXGetErrorString9(rc)); + rc=IDirectSoundBuffer_GetPan(primary,&pan); + ok(rc==DS_OK,"Primary GetPan failed: %s\n",DXGetErrorString9(rc)); + trace("Primary volume=%ld pan=%ld\n",vol,pan); + test_buffer(dso,secondary,0,FALSE,0,FALSE,0,winetest_interactive,1.0,0,NULL,0,0); ref=IDirectSoundBuffer_Release(secondary); Index: dlls/dsound/dsound.c =================================================================== RCS file: /var/cvs/wine/dlls/dsound/dsound.c,v retrieving revision 1.6 diff -u -r1.6 dsound.c --- dlls/dsound/dsound.c 19 Jul 2004 20:06:22 -0000 1.6 +++ dlls/dsound/dsound.c 21 Jul 2004 09:34:19 -0000 @@ -892,10 +892,6 @@ pDS->drvcaps.dwPrimaryBuffers = 1; } - pDS->volpan.lVolume = 0; - pDS->volpan.lPan = 0; - DSOUND_RecalcVolPan(&(pDS->volpan)); - InitializeCriticalSection(&(pDS->mixlock)); RtlInitializeResource(&(pDS->lock)); Index: dlls/dsound/dsound_private.h =================================================================== RCS file: /var/cvs/wine/dlls/dsound/dsound_private.h,v retrieving revision 1.19 diff -u -r1.19 dsound_private.h --- dlls/dsound/dsound_private.h 21 Jul 2004 03:23:13 -0000 1.19 +++ dlls/dsound/dsound_private.h 21 Jul 2004 12:27:47 -0000 @@ -91,7 +91,6 @@ IDirectSoundBufferImpl** buffers; RTL_RWLOCK lock; CRITICAL_SECTION mixlock; - DSVOLUMEPAN volpan; PrimaryBufferImpl* primary; DSBUFFERDESC dsbd; DWORD speaker_config; @@ -450,6 +449,7 @@ extern IClassFactoryImpl DSOUND_FULLDUPLEX_CF; void DSOUND_RecalcVolPan(PDSVOLUMEPAN volpan); +void DSOUND_AmpFactorToVolPan(PDSVOLUMEPAN volpan); void DSOUND_RecalcFormat(IDirectSoundBufferImpl *dsb); /* primary.c */ Index: dlls/dsound/mixer.c =================================================================== RCS file: /var/cvs/wine/dlls/dsound/mixer.c,v retrieving revision 1.20 diff -u -r1.20 mixer.c --- dlls/dsound/mixer.c 12 Jan 2004 21:02:22 -0000 1.20 +++ dlls/dsound/mixer.c 21 Jul 2004 11:21:55 -0000 @@ -20,6 +20,7 @@ */ #include "config.h" +#define _GNU_SOURCE /* for round() in math.h */ #include <assert.h> #include <stdarg.h> #include <stdio.h> @@ -54,6 +55,7 @@ double temp; TRACE("(%p)\n",volpan); + TRACE("Vol=%ld Pan=%ld\n", volpan->lVolume, volpan->lPan); /* the AmpFactors are expressed in 16.16 fixed point */ volpan->dwVolAmpFactor = (ULONG) (pow(2.0, volpan->lVolume / 600.0) * 0xffff); /* FIXME: dwPan{Left|Right}AmpFactor */ @@ -67,6 +69,39 @@ TRACE("left = %lx, right = %lx\n", volpan->dwTotalLeftAmpFactor, volpan->dwTotalRightAmpFactor); } +void DSOUND_AmpFactorToVolPan(PDSVOLUMEPAN volpan) +{ + double left,right; + TRACE("(%p)\n",volpan); + + TRACE("left=%lx, right=%lx\n",volpan->dwTotalLeftAmpFactor,volpan->dwTotalRightAmpFactor); + if (volpan->dwTotalLeftAmpFactor==0) + left=-10000; + else + left=600 * log(((double)volpan->dwTotalLeftAmpFactor) / 0xffff) / log(2); + if (volpan->dwTotalRightAmpFactor==0) + right=-10000; + else + right=600 * log(((double)volpan->dwTotalRightAmpFactor) / 0xffff) / log(2); + if (left<right) + { + volpan->lVolume=round(right); + volpan->dwVolAmpFactor=volpan->dwTotalRightAmpFactor; + } + else + { + volpan->lVolume=round(left); + volpan->dwVolAmpFactor=volpan->dwTotalLeftAmpFactor; + } + if (volpan->lVolume < -10000) + volpan->lVolume=-10000; + volpan->lPan=round(right-left); + if (volpan->lPan < -10000) + volpan->lPan=-10000; + + TRACE("Vol=%ld Pan=%ld\n", volpan->lVolume, volpan->lPan); +} + void DSOUND_RecalcFormat(IDirectSoundBufferImpl *dsb) { DWORD sw; Index: dlls/dsound/primary.c =================================================================== RCS file: /var/cvs/wine/dlls/dsound/primary.c,v retrieving revision 1.25 diff -u -r1.25 primary.c --- dlls/dsound/primary.c 13 Jul 2004 23:35:09 -0000 1.25 +++ dlls/dsound/primary.c 21 Jul 2004 11:06:06 -0000 @@ -74,8 +74,6 @@ HRESULT err = DS_OK; TRACE("(%p)\n",This); - DSOUND_RecalcVolPan(&(This->volpan)); - /* are we using waveOut stuff? */ if (!This->driver) { LPBYTE newbuf; @@ -135,11 +133,6 @@ } if ((err == DS_OK) && (merr != DS_OK)) err = merr; - - if (!err) { - DWORD vol = (This->volpan.dwTotalLeftAmpFactor & 0xffff) | (This->volpan.dwTotalRightAmpFactor << 16); - err = mmErr(waveOutSetVolume(This->hwo, vol)); - } } else { if (!This->hwbuf) { err = IDsDriver_CreateSoundBuffer(This->driver,&(This->wfx), @@ -154,7 +147,6 @@ if (dsound->state == STATE_PLAYING) dsound->state = STATE_STARTING; else if (dsound->state == STATE_STOPPING) dsound->state = STATE_STOPPED; } - err = IDsDriverBuffer_SetVolumePan(This->hwbuf, &(This->volpan)); } return err; @@ -458,7 +450,8 @@ ) { ICOM_THIS(PrimaryBufferImpl,iface); IDirectSoundImpl* dsound = This->dsound; - LONG oldVol; + DWORD ampfactors; + DSVOLUMEPAN volpan; TRACE("(%p,%ld)\n",This,vol); @@ -475,24 +468,26 @@ /* **** */ EnterCriticalSection(&(dsound->mixlock)); - oldVol = dsound->volpan.lVolume; - dsound->volpan.lVolume = vol; - DSOUND_RecalcVolPan(&dsound->volpan); - - if (vol != oldVol) { - if (dsound->hwbuf) { - HRESULT hres; - hres = IDsDriverBuffer_SetVolumePan(dsound->hwbuf, &(dsound->volpan)); - if (hres != DS_OK) { - LeaveCriticalSection(&(dsound->mixlock)); - WARN("IDsDriverBuffer_SetVolumePan failed\n"); - return hres; - } - } else { - DWORD vol = (dsound->volpan.dwTotalLeftAmpFactor & 0xffff) | (dsound->volpan.dwTotalRightAmpFactor << 16); - waveOutSetVolume(dsound->hwo, vol); - } - } + waveOutGetVolume(dsound->hwo, &ampfactors); + volpan.dwTotalLeftAmpFactor=ampfactors & 0xffff; + volpan.dwTotalRightAmpFactor=ampfactors >> 16; + DSOUND_AmpFactorToVolPan(&volpan); + if (vol != volpan.lVolume) { + volpan.lVolume=vol; + DSOUND_RecalcVolPan(&volpan); + if (dsound->hwbuf) { + HRESULT hres; + hres = IDsDriverBuffer_SetVolumePan(dsound->hwbuf, &volpan); + if (hres != DS_OK) { + LeaveCriticalSection(&(dsound->mixlock)); + WARN("IDsDriverBuffer_SetVolumePan failed\n"); + return hres; + } + } else { + ampfactors = (volpan.dwTotalLeftAmpFactor & 0xffff) | (volpan.dwTotalRightAmpFactor << 16); + waveOutSetVolume(dsound->hwo, ampfactors); + } + } LeaveCriticalSection(&(dsound->mixlock)); /* **** */ @@ -504,6 +499,8 @@ LPDIRECTSOUNDBUFFER8 iface,LPLONG vol ) { ICOM_THIS(PrimaryBufferImpl,iface); + DWORD ampfactors; + DSVOLUMEPAN volpan; TRACE("(%p,%p)\n",This,vol); if (!(This->dsound->dsbd.dwFlags & DSBCAPS_CTRLVOLUME)) { @@ -516,7 +513,11 @@ return DSERR_INVALIDPARAM; } - *vol = This->dsound->volpan.lVolume; + waveOutGetVolume(dsound->hwo, &ampfactors); + volpan.dwTotalLeftAmpFactor=ampfactors & 0xffff; + volpan.dwTotalRightAmpFactor=ampfactors >> 16; + DSOUND_AmpFactorToVolPan(&volpan); + *vol = volpan.lVolume; return DS_OK; } @@ -769,7 +770,8 @@ ) { ICOM_THIS(PrimaryBufferImpl,iface); IDirectSoundImpl* dsound = This->dsound; - LONG oldPan; + DWORD ampfactors; + DSVOLUMEPAN volpan; TRACE("(%p,%ld)\n",This,pan); @@ -786,25 +788,27 @@ /* **** */ EnterCriticalSection(&(dsound->mixlock)); - oldPan = dsound->volpan.lPan; - dsound->volpan.lPan = pan; - DSOUND_RecalcVolPan(&dsound->volpan); - - if (pan != oldPan) { - if (dsound->hwbuf) { - HRESULT hres; - hres = IDsDriverBuffer_SetVolumePan(dsound->hwbuf, &(dsound->volpan)); - if (hres != DS_OK) { - LeaveCriticalSection(&(dsound->mixlock)); - WARN("IDsDriverBuffer_SetVolumePan failed\n"); - return hres; - } - } - else { - DWORD vol = (dsound->volpan.dwTotalLeftAmpFactor & 0xffff) | (dsound->volpan.dwTotalRightAmpFactor << 16); - waveOutSetVolume(dsound->hwo, vol); - } - } + waveOutGetVolume(dsound->hwo, &ampfactors); + volpan.dwTotalLeftAmpFactor=ampfactors & 0xffff; + volpan.dwTotalRightAmpFactor=ampfactors >> 16; + DSOUND_AmpFactorToVolPan(&volpan); + if (pan != volpan.lPan) { + volpan.lPan=pan; + DSOUND_RecalcVolPan(&volpan); + if (dsound->hwbuf) { + HRESULT hres; + hres = IDsDriverBuffer_SetVolumePan(dsound->hwbuf, &volpan); + if (hres != DS_OK) { + LeaveCriticalSection(&(dsound->mixlock)); + WARN("IDsDriverBuffer_SetVolumePan failed\n"); + return hres; + } + } + else { + ampfactors = (volpan.dwTotalLeftAmpFactor & 0xffff) | (volpan.dwTotalRightAmpFactor << 16); + waveOutSetVolume(dsound->hwo, ampfactors); + } + } LeaveCriticalSection(&(dsound->mixlock)); /* **** */ @@ -816,6 +820,8 @@ LPDIRECTSOUNDBUFFER8 iface,LPLONG pan ) { ICOM_THIS(PrimaryBufferImpl,iface); + DWORD ampfactors; + DSVOLUMEPAN volpan; TRACE("(%p,%p)\n",This,pan); if (!(This->dsound->dsbd.dwFlags & DSBCAPS_CTRLPAN)) { @@ -828,8 +834,11 @@ return DSERR_INVALIDPARAM; } - *pan = This->dsound->volpan.lPan; - + waveOutGetVolume(dsound->hwo, &ampfactors); + volpan.dwTotalLeftAmpFactor=ampfactors & 0xffff; + volpan.dwTotalRightAmpFactor=ampfactors >> 16; + DSOUND_AmpFactorToVolPan(&volpan); + *pan = volpan.lPan; return DS_OK; }

1 0

Serial port problem in 200407016
by jeremy＠fantasticfamily.net July 21, 2004

July 21, 2004

I seem to be having trouble communicating with my serial device. I have a Garming Rino GPS connected on /dev/ttyS0 using Mandrake 10.0, wine build 200407016. I read on 6/14/04 that some serial port issues were solved. permissons on /dev/ttyS0 are set correct and I've used setserial to make the baud rate 9600. In MapSource v2.01, I've set it to com1, 9600 as well. The program runs great under wine, but the is the error I get when I try to transfer maps to the unit (from WINEDEBUG=+comm,+file, +ntdll): <clip> trace:file:GetOverlappedResult (0x6c 0x406cee6c 0x406cdfe0 0) trace:comm:ClearCommError handle 0x6c cbInQue = 0 cbOutQue = 0 trace:file:CreateFileW L"C:\\windows\\win.ini" GENERIC_READ FILE_SHARE_READ attributes 0x80 trace:file:RtlDosPathNameToNtPathName_U (L"C:\\windows\\win.ini",0x406cade8,(nil),(nil)) trace:file:RtlGetFullPathName_U (L"C:\\windows\\win.ini" 520 0x406cab68 (nil)) trace:ntdll:NtCreateFile handle=0x406caddc access=80000000 name=L"\\??\\C:\\windows\\win.ini" objattr=00000040 root=(nil) sec=(nil) io=0x406cade0 alloc_size=(nil) attr=00000080 sharing=00000001 disp=1 options=00000050 ea=(nil).0x00000000 trace:file:wine_nt_to_unix_file_name L"\\??\\C:\\windows\\win.ini" -> "/root/.wine/dosdevices/c:/windows/win.ini" trace:file:CreateFileW returning 0x64 trace:ntdll:NtQueryInformationFile (0x64,0x406cadd8,0x406cade0,0x00000024,0x00000004) trace:file:CreateFileW L"C:\\windows\\win.ini" GENERIC_READ FILE_SHARE_READ OPEN_EXISTING attributes 0x80 trace:file:RtlDosPathNameToNtPathName_U (L"C:\\windows\\win.ini",0x406cad7c,(nil),(nil)) trace:file:RtlGetFullPathName_U (L"C:\\windows\\win.ini" 520 0x406caafc (nil)) trace:ntdll:NtCreateFile handle=0x406cad70 access=80000000 name=L"\\??\\C:\\windows\\win.ini" objattr=00000040 root=(nil) sec=(nil) io=0x406cad74 alloc_size=(nil) </clip> Any Ideas? If you need more output from debug, I'd be happy to get it. Thanks in advance! Jeremy

2 1

Re: Fix winmm's check_position() tests
by Francois Gouget July 21, 2004

July 21, 2004

Robert Reif wrote: > I would just revert the patch and recalculate duration based on the > fragment size. You need to keep the double because of integer > overflow in the MS calculation Well, it won't overflow for the tones we play because they just last a few seconds. But it will start overflowing when we reach 11 seconds so you're right, we should try to avoid that. Rewriting the way we compute samples will fix the first place where we could get an integer overflow: samples=bytes/(pwfx->wBitsPerSample/8*pwfx->nChannels); And since both TIME_MS and TIME_SMPTE essentially use the duration, I added a calculation of the duration, as a double obviously, near the top of check_position(). This fixes the second place where we could potentially get an overflow. > and integer underflow in the SMPTE calculation. The SMPTE calculation is float-based so I don't see how we can get an 'integer underflow'. > We need to add a ceil to the frame calculation to round > up partial frames. That's done already. Changelog: * dlls/winmm/tests/wave.c Francois Gouget <fgouget(a)codeweavers.com> Make sure we won't get an overflow in check_position() even if we play longer test tones. -- Francois Gouget fgouget(a)codeweavers.com Index: dlls/winmm/tests/wave.c =================================================================== RCS file: /var/cvs/wine/dlls/winmm/tests/wave.c,v retrieving revision 1.34 diff -u -r1.34 wave.c --- dlls/winmm/tests/wave.c 20 Jul 2004 22:09:53 -0000 1.34 +++ dlls/winmm/tests/wave.c 20 Jul 2004 22:53:55 -0000 @@ -215,9 +230,11 @@ { MMTIME mmtime; DWORD samples; + double duration; MMRESULT rc; - samples=bytes*8/pwfx->wBitsPerSample/pwfx->nChannels; + samples=bytes/(pwfx->wBitsPerSample/8*pwfx->nChannels); + duration=((double)samples)/pwfx->nSamplesPerSec; mmtime.wType = TIME_BYTES; rc=waveOutGetPosition(wout, &mmtime, sizeof(mmtime)); @@ -246,9 +263,9 @@ ok(rc==MMSYSERR_NOERROR, "waveOutGetPosition: device=%s rc=%s\n",dev_name(device),wave_out_error(rc)); if (mmtime.wType == TIME_MS) - ok(mmtime.u.ms==samples*1000/pwfx->nSamplesPerSec, + ok(mmtime.u.ms==floor(duration*1000.0), "waveOutGetPosition returned %ld ms, should be %ld\n", - mmtime.u.ms, samples*1000/pwfx->nSamplesPerSec); + mmtime.u.ms, floor(duration*1000.0)); else trace("TIME_MS not supported, returned %s\n",wave_time_format(mmtime.wType)); @@ -258,7 +275,6 @@ "waveOutGetPosition: device=%s rc=%s\n",dev_name(device),wave_out_error(rc)); if (mmtime.wType == TIME_SMPTE) { - double duration=((double)samples)/pwfx->nSamplesPerSec; BYTE frames=ceil(fmod(duration*mmtime.u.smpte.fps, mmtime.u.smpte.fps)); ok(mmtime.u.smpte.hour==(BYTE)(floor(duration/(60*60))) && mmtime.u.smpte.min==(BYTE)(fmod(floor(duration/60), 60)) &&

1 0

help with building with winelib on Mac os x
by Jeff Cooper July 20, 2004

July 20, 2004

I am trying to build a "Hello World" app with winelib on Mac os x I have done so by mimicking how winemine is built or modifying the winemine make file to work with the hello world app. This uses GCC to build correct?.. I then try winemaker to create a make file, and when trying to build with that makefile I get lots of errors. This uses winegcc correct?. I think that is where the problem is... Any Ideas? help? What am I missing? Error: winebuild -o main.exe.dbg.c --debug -C. main.c winegcc -c -mno-cygwin -o main.exe.dbg.o main.exe.dbg.c winegcc -c -mno-cygwin -o main.o main.c In file included from /usr/local/include/wine/windows/winnt.h:24, from /usr/local/include/wine/windows/windef.h:203, from /usr/local/include/wine/windows/windows.h:38, from main.c:1: /usr/local/include/wine/windows/basetsd.h:154:3: #error Unknown CPU architecture! In file included from /usr/local/include/wine/windows/winnt.h:29, from /usr/local/include/wine/windows/windef.h:203, from /usr/local/include/wine/windows/windows.h:38, from main.c:1: /usr/local/include/wine/msvcrt/string.h:58: warning: conflicting types for built-in function `memcmp' /usr/local/include/wine/msvcrt/string.h:59: warning: conflicting types for built-in function `memcpy' /usr/local/include/wine/msvcrt/string.h:61: warning: conflicting types for built-in function `memset' /usr/local/include/wine/msvcrt/string.h:70: warning: conflicting types for built-in function `strncat' /usr/local/include/wine/msvcrt/string.h:71: warning: conflicting types for built-in function `strncmp' /usr/local/include/wine/msvcrt/string.h:72: warning: conflicting types for built-in function `strncpy' In file included from /usr/local/include/wine/windows/windef.h:203, from /usr/local/include/wine/windows/windows.h:38, from main.c:1: /usr/local/include/wine/windows/winnt.h:1170:2: #error You need to define a CONTEXT for your CPU In file included from /usr/local/include/wine/windows/windef.h:203, from /usr/local/include/wine/windows/windows.h:38, from main.c:1: /usr/local/include/wine/windows/winnt.h:1173: error: parse error before '*' token /usr/local/include/wine/windows/winnt.h:1173: warning: data definition has no type or storage class /usr/local/include/wine/windows/winnt.h:1489: error: parse error before "PCONTEXT" /usr/local/include/wine/windows/winnt.h:1489: warning: no semicolon at end of struct or union /usr/local/include/wine/windows/winnt.h:1490: warning: data definition has no type or storage class /usr/local/include/wine/windows/winnt.h:1502: error: parse error before "PCONTEXT" /usr/local/include/wine/windows/winnt.h:1514: error: parse error before "ExceptionInfo" /usr/local/include/wine/windows/winnt.h:1514: warning: `__stdcall__' attribute directive ignored /usr/local/include/wine/windows/winnt.h:1515: error: parse error before "ExceptionInfo" /usr/local/include/wine/windows/winnt.h:1515: warning: `__stdcall__' attribute directive ignored /usr/local/include/wine/windows/winnt.h:1518: error: parse error before "epointers" /usr/local/include/wine/windows/winnt.h:1518: warning: `__stdcall__' attribute directive ignored ...{similar errors with more header files}.. /usr/local/include/wine/windows/imm.h:543: warning: `__stdcall__' attribute directive ignored main.c:7: warning: `__stdcall__' attribute directive ignored winegcc -mwindows -mno-cygwin -o main.exe.so main.o main.exe.dbg.o -lodbc32 -lole32 -loleaut32 -lwinspool -luuid ld: table of contents for archive: /usr/local/lib/wine/libuuid.a is out of date; rerun ranlib(1) (can't load from it) winebuild: ld -r failed with status 256 winegcc: winebuild failed. make: *** [main.exe.so] Error 2

2 2

bind() test fails
by James Hawkins July 20, 2004

July 20, 2004

Hi, I ran the winsock tests and the only failure i received was from bind: sock.c:660: Test failed: bind() failed error: 10048 I checked msdn and error 10048 is WSAEADDRINUSE. The code for the test of this part is: rc = bind(s2, (struct sockaddr*)&saddr, sizeof(saddr)); ok(rc==SOCKET_ERROR, "bind() succeeded\n"); reuse = 1; rc = setsockopt(s2, SOL_SOCKET, SO_REUSEADDR, (char*)&reuse, sizeof(reuse)); ok(rc==0, "setsockopt() failed error: %d\n", WSAGetLastError()); todo_wine { rc = bind(s2, (struct sockaddr*)&saddr, sizeof(saddr)); ok(rc==0, "bind() failed error: %d\n", WSAGetLastError()); } For this part of the test, we bind an address, saddr, with a socket, s2, and then set the SO_REUSEADDR option to 1 so that we should be able to bind again, but we're getting the error WSAEADDRINUSE which means that the fact that we are allowing multiple binds with the SO_REUSEADDR option is being ignored somewhere. I looked at the function declarations for the win32 and linux setsockopt. win32: WS_setsockopt(SOCKET s, int level, int optname, const char *optval, int optlen) linux: setsockopt (int __fd, int __level, int __optname, __const void *__optval, socklen_t __optlen) At one point in WS_setsockopt we make a call to the linux setsockopt and im wondering if the problem is when we call this version. Is it not a problem that WS_setsockopt's optval parameter is of type const char * and setsockopt's optval parameter is __const void *? I've searched for a while about this problem and I can't seem to spot the fix. If i'm totally off the mark on this one, let me know as well. Thanks for the help. For further reference see: http://msdn.microsoft.com/library/default.asp?url=/library/en-us/winsock/wi… http://msdn.microsoft.com/library/default.asp?url=/library/en-us/winsock/wi… -- James Hawkins

1 0

Re: Don't crash on NULL data in RegSetValueEx
by Andreas Mohr July 20, 2004

July 20, 2004

Hi, On Tue, Jul 20, 2004 at 03:25:57PM +0100, Mike Hearn wrote: > I tested this on Win98. The platform SDK installer does this. Ah, right, my uncommitted patch from April 14. > + if (!data) return ERROR_INVALID_PARAMETER; > + > if (!is_version_nt()) /* win95 */ > { Are you sure this is correct? free-av.de called RegSetValueEx with a NULL value on Win9x setting only, so maybe NT handles this differently? (I thus placed the check into the win9x branch only) Andreas Mohr

1 0