July 2010 - Wine-devel - list.winehq.org

Re: user32: RegisterClassEx should check for invalid cbSize field.
by Dmitry Timoshkov 18 Jul '10

18 Jul '10

Dylan Smith <dylan.ah.smith(a)gmail.com> wrote: > + wcexA.cbSize = 0; > + wcexA.style = cls.style; > + wcexA.lpfnWndProc = cls.lpfnWndProc; > + wcexA.cbClsExtra = cls.cbClsExtra; > + wcexA.cbWndExtra = cls.cbWndExtra; > + wcexA.hInstance = cls.hInstance; > + wcexA.hIcon = cls.hIcon; > + wcexA.hCursor = cls.hCursor; > + wcexA.hbrBackground = cls.hbrBackground; > + wcexA.lpszMenuName = cls.lpszMenuName; > + wcexA.lpszClassName = cls.lpszClassName; > + wcexA.hIconSm = 0; > + ok( ((RegisterClassExA( &wcexA ) == 0) && (GetLastError() == ERROR_INVALID_PARAMETER)), > + "Failed with invalid number of cbSize bytes\n"); Testing other values besides 0 would be helpful, in particular values like sizeof(class) - 1 and sizeof(class) + 1. Did you inspect all the places in Wine which call RegisterClassEx()? Also, GetClassInfoEx() in Wine is an obvious offender, and this needs a test (and a fix) too. -- Dmitry.

2 1

Re: dlls/user.exec16/user.c: Use FIXME_ONCE for quieter reports
by James McKenzie 18 Jul '10

18 Jul '10

Max TenEyck Woodbury wrote: > --- > dlls/user.exe16/user.c | 48 ++++++++++++++++++++++++------------------------ > 1 files changed, 24 insertions(+), 24 deletions(-) > > This whole line of thought was rejected, at least once. What is your justification to add this to the project. This should be in the changelog section. Also, if this is a resubmission due to the code freeze, this should have a (resend) at the end of the title. James McKenzie

1 0

Re: [Resend] Add '_ONCE' variants of DPRINTF, FIXME, WARN.
by Mike Kaplinskiy 18 Jul '10

18 Jul '10

On Sat, Jul 17, 2010 at 1:29 AM, Max TenEyck Woodbury <max(a)mtew.isa-geek.net> wrote: > +#define WINE_FIXME_ONCE(args...) do { } while(0) > +#define WINE_FIXME_ONCE_ONCE_(ch) WINE_FIXME_ONCE I think you want WINE_FIXME_ONCE_ not WINE_FIXME_ONCE_ONCE_. > +#define WINE_FIXME_ONCE(args...) do { } while(0) > +#define WINE_FIXME_ONCE_ONCE_(ch) WINE_FIXME_ONCE Ditto > +#define __WINE_DPRINTF_ONCE(dbcl,dbch) /* FIXME */ \ > + (!__WINE_GET_DEBUGGING(dbcl,(dbch)) || \ > + (wine_dbg_log(__WINE_DBCL##dbcl,(dbch),__FILE__,"%d: ",__LINE__) == -1)) ? \ > + (void)0 : (void)wine_dbg_printf > + Ask if you don't know how to do it. I don't think we bite too much :). I have no idea what compilers need this instead of the versions for GNU and SunPro. In fact, I don't actually know what compilers those would be (clang may be one of them?). But in the worst case I would suggest at least #error 'ing if someone uses __WINE_DPRINTF_ONCE on these instead of falling back to non-once behavior. > +#define WARN_ONCE_(ch) WINE_WARE_ONCE_(ch) typo > +#define ERR_ONCE WINE_ERR_ONCEE typo

1 0

Re: ws2_32: socket.c: wrapped WSASendTo/WSARecvFrom with a different function name
by Mike Kaplinskiy 17 Jul '10

17 Jul '10

On Sat, Jul 17, 2010 at 7:02 AM, Pigeon <pigeon(a)pigeond.net> wrote: > > WSASendTo/WSARecvFrom, wrapped in a different function name, so that > internal calls from ws2_32 itself will not trigger programs like Garena, > which catches WSASendTo/WSARecvFrom calls. > > Reference: Wine Bug Id #16047 > > --- > dlls/ws2_32/socket.c | 73 ++++++++++++++++++++++++++++++++++++++++--------- > 1 files changed, 59 insertions(+), 14 deletions(-) > > > > > The helpers shouldn't need a WINAPI (those are for exported functions only). And don't use INT in helpers (int is fine). You might also want to give them better names. The names tend to be something like WS2_(lower_case_with_underscore_for_space) in that file. You may want to align the arguments as they were before, but now I'm just nitpicking :). If this does get in, we may want to nominate it for stable - it's a pretty simple change. Mike.

1 0

[PATCH] winealsa.drv: Always scan devices, even if a default device exists.
by Jeff Cook 17 Jul '10

17 Jul '10

Let me know what you guys think of this one, I think it will be somewhat controversial. There is a registry key to enable this currently, but it should really occur by default and maybe be unchangeable. If you have multiple cards, like a USB mic for capture, scan_devices to 0, as would occur on many configurations currently, is undesirable and most users won't be able to work around it. I'm not sure why scan_devices would be 0, maybe to prevent bugs while probing or something, but I don't see much reason to disable scanning all reported audio devices. Thoughts? Please let me know if I'm missing anything. From Jeff --- dlls/winealsa.drv/waveinit.c | 6 +++++- 1 files changed, 5 insertions(+), 1 deletions(-)

1 0

Re: ws2_32: socket.c: wrapped WSASendTo/WSARecvFrom with a different function name
by John Klehm 17 Jul '10

17 Jul '10

On Sat, Jul 17, 2010 at 6:02 AM, Pigeon <pigeon(a)pigeond.net> wrote: > > WSASendTo/WSARecvFrom, wrapped in a different function name, so that > internal calls from ws2_32 itself will not trigger programs like Garena, > which catches WSASendTo/WSARecvFrom calls. > > Reference: Wine Bug Id #16047 > > --- Need your real name when submitting a patch. See http://wiki.winehq.org/SubmittingPatches item 1. Regards, --John

1 0

[RFC PATCH 2/2] wined3d: Let GetRasterStatus return D3DERR_INVALIDCALL.
by Florian Köberle 17 Jul '10

17 Jul '10

--- dlls/d3d9/tests/device.c | 1 - dlls/wined3d/swapchain_base.c | 2 +- 2 files changed, 1 insertions(+), 2 deletions(-) diff --git a/dlls/d3d9/tests/device.c b/dlls/d3d9/tests/device.c index 0acedcb..2345c34 100644 --- a/dlls/d3d9/tests/device.c +++ b/dlls/d3d9/tests/device.c @@ -2771,7 +2771,6 @@ static void test_get_raster_status(void) if ((caps.Caps & D3DCAPS_READ_SCANLINE) == 0) { ZeroMemory( &rasterStatus, sizeof(rasterStatus) ); hr = IDirect3DDevice9_GetRasterStatus(pDevice, 0, &rasterStatus); - todo_wine ok(hr == D3DERR_INVALIDCALL, "GetRasterStatus should return D3DERR_INVALIDCALL\n"); } else { skip("Device supports D3DCAPS_READ_SCANLINE\n"); diff --git a/dlls/wined3d/swapchain_base.c b/dlls/wined3d/swapchain_base.c index 1817de2..3f2a4e5 100644 --- a/dlls/wined3d/swapchain_base.c +++ b/dlls/wined3d/swapchain_base.c @@ -131,7 +131,7 @@ HRESULT WINAPI IWineD3DBaseSwapChainImpl_GetRasterStatus(IWineD3DSwapChain *ifac FIXME("iface %p, raster_status %p stub!\n", iface, pRasterStatus); warned = TRUE; } - return WINED3D_OK; + return WINED3DERR_INVALIDCALL; } HRESULT WINAPI IWineD3DBaseSwapChainImpl_GetDisplayMode(IWineD3DSwapChain *iface, WINED3DDISPLAYMODE*pMode) { -- 1.7.0.4

2 1

[RFC PATCH 1/2] d3d9/tests: Test that GetRasterStatus returns D3DERR_INVALIDCALL if D3DCAPS_READ_SCANLINE isn't set.
by Florian Köberle 17 Jul '10

17 Jul '10

--- dlls/d3d9/tests/device.c | 53 ++++++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 53 insertions(+), 0 deletions(-) diff --git a/dlls/d3d9/tests/device.c b/dlls/d3d9/tests/device.c index ba25760..0acedcb 100644 --- a/dlls/d3d9/tests/device.c +++ b/dlls/d3d9/tests/device.c @@ -2735,6 +2735,58 @@ done: if (d3d9) IDirect3D9_Release(d3d9); } +static void test_get_raster_status(void) +{ + HRESULT hr; + HWND hwnd = NULL; + IDirect3D9 *pD3d = NULL; + IDirect3DDevice9 *pDevice = NULL; + D3DPRESENT_PARAMETERS d3dpp; + D3DDISPLAYMODE d3ddm; + D3DCAPS9 caps; + D3DRASTER_STATUS rasterStatus; + + pD3d = pDirect3DCreate9( D3D_SDK_VERSION ); + ok(pD3d != NULL, "Failed to create IDirect3D9 object\n"); + hwnd = CreateWindow( "static", "d3d9_test", WS_OVERLAPPEDWINDOW, 100, 100, 160, 160, NULL, NULL, NULL, NULL ); + ok(hwnd != NULL, "Failed to create window\n"); + if (!pD3d || !hwnd) goto cleanup; + + IDirect3D9_GetAdapterDisplayMode( pD3d, D3DADAPTER_DEFAULT, &d3ddm ); + ZeroMemory( &d3dpp, sizeof(d3dpp) ); + d3dpp.Windowed = TRUE; + d3dpp.SwapEffect = D3DSWAPEFFECT_DISCARD; + d3dpp.BackBufferFormat = d3ddm.Format; + d3dpp.BackBufferCount = 0; + + hr = IDirect3D9_CreateDevice( pD3d, D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, hwnd, + D3DCREATE_SOFTWARE_VERTEXPROCESSING, &d3dpp, &pDevice ); + ok(hr == S_OK || hr == D3DERR_NOTAVAILABLE, + "Failed to create IDirect3D9Device (%08x)\n", hr); + if (FAILED(hr)) goto cleanup; + + hr = IDirect3DDevice9_GetDeviceCaps(pDevice, &caps); + ok(hr == D3D_OK, "IDirect3DDevice9_GetCaps failed with %08x\n", hr); + + if ((caps.Caps & D3DCAPS_READ_SCANLINE) == 0) { + ZeroMemory( &rasterStatus, sizeof(rasterStatus) ); + hr = IDirect3DDevice9_GetRasterStatus(pDevice, 0, &rasterStatus); + todo_wine + ok(hr == D3DERR_INVALIDCALL, "GetRasterStatus should return D3DERR_INVALIDCALL\n"); + } else { + skip("Device supports D3DCAPS_READ_SCANLINE\n"); + } + +cleanup: + if (pDevice) + { + UINT refcount = IDirect3DDevice9_Release(pDevice); + ok(!refcount, "Device has %u references left.\n", refcount); + } + if (pD3d) IDirect3D9_Release(pD3d); + DestroyWindow( hwnd ); +} + START_TEST(device) { HMODULE d3d9_handle = LoadLibraryA( "d3d9.dll" ); @@ -2778,5 +2830,6 @@ START_TEST(device) test_scissor_size(); test_wndproc(); test_wndproc_windowed(); + test_get_raster_status(); } } -- 1.7.0.4

2 1

Re: [PATCH 1/5] jscript: Fixed sort_cmp for non-string arguments.
by Henri Verbeet 17 Jul '10

17 Jul '10

On 17 July 2010 13:28, Jacek Caban <jacek(a)codeweavers.com> wrote: > + }else if(V_VT(v1) == VT_EMPTY) { > + *cmp = V_VT(v1) == VT_EMPTY ? 0 : 1; That looks wrong, should that read "*cmp = V_VT(v2) == VT_EMPTY ? 0 : 1;" instead?

1 0

re: [2/2] imagehlp: Implement ImageGetDigestStream
by Dan Kegel 17 Jul '10

17 Jul '10

I was worried that mapping the file will fail if you're low on address space and are running this function on a large PE file. It ought to be possible to write a test to see how the function behaves when you've used up oodles of address space, though it'd be tricky to make the test reliable. I see that Windows has a similar problem in MapFileAndCheckSum (which is said to be a caller of this function) http://www.daniweb.com/forums/thread122484.html so who knows, maybe Windows also maps the file...

2 1