November 2014 - Wine-Devel - list.winehq.org

Re: [PATCH 4/7] reg: Add wchar/type conversion functions
by Stefan Dösinger Nov. 4, 2014

Nov. 4, 2014

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Am 2014-11-04 21:15, schrieb Jonathan Vollebregt: > +#define ERROR_INVALID_TYPE 20001 What does RegSetValueEx return if you pass an invalid type? The API makes this error condition possible, so there's likely a return value for it. (Yeah, I know, it's not your fault advapi32/tests has no test for this. If I understand server/registry.c correctly such a call succeeds and creates a blob with an undefined type. Still needs tests) ERROR_INVALID_DATATYPE may be a candidate. > +typedef struct > +{ > + DWORD type; > + const WCHAR *name; > +} type_rel; You can make this anonymous too. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQIcBAEBAgAGBQJUWU1cAAoJEN0/YqbEcdMwBWcQAID2xh321Jci/xtp6ZJMY7cV 6bzXGGBfyBiDaCzyWiUdcZrLaGmPm6ehuV655FylMEAQvY8XGOoifdF1F6rdlqok fwGLfdwWns6EUpdINipGbBrPFn2bKdJyP3ZECY0owEYM8H76CoJDwcgCiIrIM1T1 0uejd152KmVs/4p4kg/XYvWbJqSQc4VceHs2HPW2QG+LZ7prLEK/Cwj7PVu/sIq7 KZKrDEx6UGEkvYCr24ldWcd4qTzicnzipjnbqR8rH0F3Y0ZAL2EaQqrCfSdxj14O Yy8YIn+QQTBa732yahs9hJrVeFYhgLboVrLavTb+AtoLeq5/TnBLwB5Gl1+EqIah HRVK9VBNE1hlIjj84+GdUx4hicNwBwUe0a/cODC6b0LwsBh9aalawWuhOAs8R38/ xt4RNHc/1VUxC3XWNz8s0NuPjIooPLu5z4rnbyl2LM99w5mteac9wqH8mhmJE1qa INB1yBbCGv9dccSZwzRtkyw1Lpj4762uKM6Vxk01ii2lN2uP/mWN91M9o7blga3h +6QTL+TiiUzwVMirkjX2TxLSTDZxHXkYRadKMgerN9V//ztGmie2cIara4J4kLzm PWDpXNguQC4WrcLzAbi7zDY3UOcxvXY4XQSCOBcIAQj254ZfTH+oBgtosWt5x9km dsLZ0QNRy5GmHGm52ine =10ha -----END PGP SIGNATURE-----

1 0

Re: [PATCH 2/7] reg: Check key paths in main
by Stefan Dösinger Nov. 4, 2014

Nov. 4, 2014

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Am 2014-11-04 21:15, schrieb Jonathan Vollebregt: > +static BOOL sane_path(const WCHAR *key) > +{ > + int i = strlenW(key); > + > + if (i < 3 || (key[i - 1] == '\\' && key[i - 2] == '\\')) > + { > + reg_print_error(ERROR_BADKEY); > + return FALSE; > + } I think I have asked this a few times, please forgive me if I forgot the answer: Did you test if RegAddKey and friends handle this for you? dlls/advapi32/tests/registry.c would profit from some tests for the error cases you have in programs/reg/tests/reg.c . > + if (!sane_path(key_name)) > + return 1; > + > for (i = 1; i < argc; i++) > { > if (!lstrcmpiW(argvW[i], slashVW)) > @@ -471,6 +485,9 @@ int wmain(int argc, WCHAR *argvW[]) > } > > key_name = argvW[2]; > + if (!sane_path(key_name)) > + return 1; > + > for (i = 1; i < argc; i++) > { > if (!lstrcmpiW(argvW[i], slashVW)) > @@ -502,6 +519,9 @@ int wmain(int argc, WCHAR *argvW[]) > } > > key_name = argvW[2]; > + if (!sane_path(key_name)) > + return 1; > + The helper function is a good idea if the answer to the above question is "no", but what's the advantage of calling it from main 3 times instead of reg_add, reg_delete and reg_query? All the other validation is done in reg_* . For consistency with the other functions it would be a good idea to return ERROR_BADKEY / ERROR_NO_REMOTE here and call reg_print_error in the caller. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQIcBAEBAgAGBQJUWUE0AAoJEN0/YqbEcdMwNIUP/0bcqgtH0ZPZ3DXmSVYRGq2I mUZmHxZHuGYdfWuiFKt6TjbcVUw/WNNuWn/1hTiSQC4A1wZvnvvT5w6dpKpyd1mP x48OXBPWg1hBALFvqxpK+x9CLUB745ms62wblfsltmJDei/EBnWTXy0jZn4rEshc jqTWesiieev/iaet+QbChMd+28RJmr3MoOaJkYOihi4SnLo6nXD0pcPL+tc3vtpW R0JA0GxmWNUVlqKuvqYvSIEMHoI6XwqL/Zlbr5PP11NtAmEzIEFfMUZ4uXQVY3vy C4qWcQctvQ8Akg+chmpoyZMYJ6+s3udpedzA2RhcKSymFgDRTXwdYnVFTK6BA1Pt jODRZ1RAOG9ShZMzIRNaIK7sBxYbK+wYxQpUvw5Xd5AuDnd2PUNJXnnvNMqW0c3x ZP5NxOpHucmgUgdSxPc0kwTLyC3Dg7KU1npUsxDnpZqgS9nSWZo/ZRp3kuNac6Z3 MSb3gLpvKDK/0mK09dGwJwCNHsIFfg67hjRVcswM4h/9O4gxWQ9tUzW/dwgT343+ iLaso6+Icvq4OFtXnZbZX8yPpK7j7oJgC/YFn94ib1Wz8PwMGb97OEPYF2IEsKWi xBLLDA976QeNKhb7Uxyg3Mr1ER3K4gaCHe0DCG5p4gfdGv5XhlaSOCGAdw/1UFFE IbSp8XZrbY+ZX0Psx/QW =FiSK -----END PGP SIGNATURE-----

1 0

Re: [PATCH 1/7] reg: Add system error printing function
by Stefan Dösinger Nov. 4, 2014

Nov. 4, 2014

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, I conceptually like this function. Am 2014-11-04 21:15, schrieb Jonathan Vollebregt: > +static int reg_print_error(LONG error_code) You are mixing LSTATUS and LONG in your code. It seems like this is ultimately Microsoft's fault since it defines LSTATUS in reg.h but puts error values from winerror.h into it. Since LSTATUS is defined as LONG it is not a big deal, but please pick one and try to be consistent. My choice would be LSTATUS since it is returned from the registry functions, but if you have a reason to choose LONG that's OK with me. > + case ERROR_FILE_NOT_FOUND: > + return reg_message(STRING_CANNOT_FIND); > + case ERROR_BADKEY: > + return reg_message(STRING_INVALID_KEY); These are dead code, please add them in the patches that call them. > + STRING_ERROR, "Error: " Wouldn't "Unexpected error" be a better fit? -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQIcBAEBAgAGBQJUWT4tAAoJEN0/YqbEcdMw3bwP/1//Ah6I2lrHnd0SnJ8m83IM yyrD6D1hL3njOFPCLFykdEbrjf+yDkz1Qi596L/qUbvkK9rbVEo4EtUgtHeJ/D+/ UF7Jr0oB8RWDQZh5ePt+Ns0x9BYUwqbi/fyOda+Xtz0P+Vl3FUH5K3U3OO5cYtlB S52bruK/Bv69FsRQwFwN9vYRX+JnJJq3NJ9Q2xoywBN26iypgTa+SDdKxA9gcDK1 xIfvSDqYsqR5Rq6UnRzroknG0j0aeuDWgHwY5pF1/DxGTbir5wCs6UOVPstIQeqk VxjrInXjGXo4LmJHz+6KPVc3+WuKr5QsmQr9Z/NgZBLN7xJXJ4DY19keFhUQfsLB AjAvf+mJy8GRwoaCihboze9ltQywoM0jZe2zLRxYiiXmD79EDN2huKi9D8g+ur58 0OE37veRYh6IUhjzQvfTxdNZvkNnPxvvP+S51qVEQu6GWPjITe375mQf8HIV3AZO z+NV3TWgsiz/tMqQK5YCPq9oR8vuXzu7H4RBjdUUpKy5yTzEcXSw6ehqj2Qehz5I MTC7QBQVNxHD3MFHGCoPA7Oeb8J2IEDGgZlj9cq9TSLTfy9XOlPQ8Abth29I/fjl GW/9Z10VGQ9F+KQ7E5E0V+Sci93l486Dv2gf579ZYmCa8sfPI7BVj1SQmkOY3bQP i3Ekbq0hR2Yk1FQcciYG =iN7/ -----END PGP SIGNATURE-----

1 0

Re: [PATCH 2/4] d3d9/tests: The device restores the mode settings from the registry (v3).
by Alexandre Julliard Nov. 4, 2014

Nov. 4, 2014

Stefan Dösinger <stefan(a)codeweavers.com> writes: > Version 3: Simplify the mode selection code and hardcode > D3DFMT_X8R8G8B8. Retain the test for mode reset when the device was > originally created with the registry size. > > Version 2: Make sure the mode set with ChangeDisplaySettings is bigger > than the device's mode. Otherwise Windows shrinks the window to the new > mode + 12 pixels. Testing this behavior is not the purpose of this test. > > Implementing this behavior is complicated because ddraw destroys and > recreates swapchains in SetCooperativeLevel. > --- > dlls/d3d9/tests/device.c | 153 ++++++++++++++++++++++++++++++++++++----------- > 1 file changed, 117 insertions(+), 36 deletions(-) It doesn't work here: ../../../tools/runtest -q -P wine -T ../../.. -M d3d9.dll -p d3d9_test.exe.so device && touch device.ok device.c:4034: Test succeeded inside todo block: Got unexpect width 1920, expected 1920. device.c:5516: Tests skipped: This GPU supports SM3, skipping unsupported shader test. device.c:5928: Tests skipped: D3DFMT_A32B32G32R32F supports filtering, skipping tests. device.c:7229: Tests skipped: Dynamic D3DFMT_YUY2 textures not supported, skipping mipmap test. device.c:7229: Tests skipped: Dynamic D3DFMT_UYVY textures not supported, skipping mipmap test. make: *** [device.ok] Error 1 -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: [1/2] d3dx9_36: Implement D3DXCreatePolygon.
by Matteo Bruni Nov. 4, 2014

Nov. 4, 2014

2014-11-03 6:35 GMT+01:00 Sebastian Lackner <sebastian(a)fds-team.de>: > Based on a patch by David Adam. > > For bug 13632. > > --- > dlls/d3dx9_36/d3dx9_36.spec | 2 +- > dlls/d3dx9_36/mesh.c | 89 +++++++++++++++++++++++++++++++++++++++++++++ > include/d3dx9shape.h | 2 + > 3 files changed, 92 insertions(+), 1 deletion(-) > Hi Sebastian, I have a few nitpicks: + TRACE("(%p, %f, %u, %p, %p)\n", device, length, sides, mesh, adjacency); Generally you want that to be something like: TRACE("device %p, length %f, sides %u, mesh %p, adjacency %p.\n", device, length, sides, mesh, adjacency); + if (!device || length < 0.0f || sides == 0 || !mesh) Probably "!sides" is more in the usual Wine d3d* style. Also, no need to use brackets for single-line branches. + if (FAILED(hr = D3DXCreateMeshFVF(sides, sides + 1, D3DXMESH_MANAGED, + D3DFVF_XYZ | D3DFVF_NORMAL,device, &polygon))) Line continuation should be indented by 8 blanks. Also, missing whitespace after a comma. + scale = 0.5f * length / sin(D3DX_PI / sides); You probably want to use sinf() here. Same below. + for (i = 0; i < sides; i++) Not important, but the current wined3d style wants "++i" here. + for (i = 0; i < sides; i++) + { + adjacency_buf[i][0] = i - 1; + adjacency_buf[i][1] = -1; + adjacency_buf[i][2] = i + 1; + } Really a detail, but adjacency_buf and i are both defined as unsigned int, so either defining them as plain, signed int or using ~0 instead of -1 would be nicer. I haven't seen the test yet but in some other D3DX9 functions the "out" pointers are set to NULL on error. If this function has the same behavior you should probably add a "*mesh = NULL;" in the first few lines of the function.

3 2

Re: [PATCH 3/4] d3d8/tests: Check if the current display settings match the registry settings.
by Marvin Nov. 4, 2014

Nov. 4, 2014

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at https://testbot.winehq.org/JobDetails.pl?Key=9978 Your paranoid android. === w8 (32 bit device) === device.c:1173: Test failed: TestCooperativeLevel failed, hr 0x88760868. device.c:1218: Test failed: Reset failed, hr 0x88760868. device.c:1220: Test failed: TestCooperativeLevel failed, hr 0x88760869. device.c:1224: Test failed: Got unexpected value 0 for D3DRS_LIGHTING. device.c:1231: Test failed: D3DVIEWPORT->X = 10, expected 0. device.c:1232: Test failed: D3DVIEWPORT->Y = 20, expected 0. device.c:1233: Test failed: D3DVIEWPORT->Width = 400, expected 800. device.c:1234: Test failed: D3DVIEWPORT->Height = 300, expected 600. device.c:1235: Test failed: D3DVIEWPORT->MinZ = 2.00000003e-001, expected 0. device.c:1236: Test failed: D3DVIEWPORT->MaxZ = 3.00000012e-001, expected 1. device.c:1241: Test failed: Screen width is 640, expected 800. device.c:1242: Test failed: Screen height is 480, expected 600. device.c:1248: Test failed: Back buffer width is 640, expected 800. device.c:1250: Test failed: Back buffer height is 480, expected 600. device.c:1280: Test failed: Screen width is 640, expected 1024. device.c:1281: Test failed: Screen height is 480, expected 768. device.c:2441: Test failed: Expected foreground window 000B00CA, got 0002004C. device.c:2813: Test failed: Expected device window style 0x14cf0000, got 0x4cf0000. device.c:2817: Test failed: Expected device window extended style 0x108, got 0x100. device.c:2839: Test failed: Failed to reset device, hr 0x88760868. device.c:2843: Test failed: Expected device window style 0x14cf0000, got 0x4cf0000. device.c:2847: Test failed: Expected device window extended style 0x108, got 0x100. device.c:3056: Test failed: Got unexpect width 640. device.c:3057: Test failed: Got unexpect height 480. device.c:3068: Test failed: Got unexpect width 640. device.c:3069: Test failed: Got unexpect height 480. device.c:3131: Test failed: Failed to reset device. device.c:3138: Test failed: Expected {0, 0, 1024, 768}, got {0, 0, 512, 384}. device.c:3146: Test failed: Expected wndproc 0x412ed0, got 0x412ed0.

2 1

Re: advapi32/tests: Simplify empty string check (PVS-Studio).
by Marvin Nov. 4, 2014

Nov. 4, 2014

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at https://testbot.winehq.org/JobDetails.pl?Key=9984 Your paranoid android. === w7u (32 bit service) === The test timed out

2 1

Re: [2/2] kernel32/tests: Add tests for FindFirstFileExW level FindExInfoBasic.
by Marvin Nov. 4, 2014

Nov. 4, 2014

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at https://testbot.winehq.org/JobDetails.pl?Key=9994 Your paranoid android. === wvistau64 (32 bit file) === file.c:367: Test failed: attributes of file "testfile.xxx" are 0x0820 === wvistau64 (64 bit file) === file.c:367: Test failed: attributes of file "testfile.xxx" are 0x0820

1 0

[PATCH] iphlpapi: call SetLastError in NotifyAddrChange/NotifyRouteChange(try 2)
by Hao Peng Nov. 4, 2014

Nov. 4, 2014

call SetLastError in NotifyAddrChange/NotifyRouteChange, to make apps can get WSA_IO_PENDING error code after async operation through GetLastError. Thanks to Sebastian Lackner for comments. superseded patch 107390. --- dlls/iphlpapi/iphlpapi_main.c | 18 ++++++++++++++++-- dlls/iphlpapi/tests/iphlpapi.c | 39 +++++++++++++++++++++++---------------- 2 files changed, 39 insertions(+), 18 deletions(-)

1 0

Re: [PATCH 6/6] d3d8/tests: Add an ATI1N entry to the test_volume_blocks test.
by Marvin Nov. 4, 2014

Nov. 4, 2014

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at https://testbot.winehq.org/JobDetails.pl?Key=9983 Your paranoid android. === w8 (32 bit device) === device.c:1174: Test failed: TestCooperativeLevel failed, hr 0x88760868. device.c:1219: Test failed: Reset failed, hr 0x88760868. device.c:1221: Test failed: TestCooperativeLevel failed, hr 0x88760869. device.c:1225: Test failed: Got unexpected value 0 for D3DRS_LIGHTING. device.c:1232: Test failed: D3DVIEWPORT->X = 10, expected 0. device.c:1233: Test failed: D3DVIEWPORT->Y = 20, expected 0. device.c:1234: Test failed: D3DVIEWPORT->Width = 400, expected 800. device.c:1235: Test failed: D3DVIEWPORT->Height = 300, expected 600. device.c:1236: Test failed: D3DVIEWPORT->MinZ = 2.00000003e-001, expected 0. device.c:1237: Test failed: D3DVIEWPORT->MaxZ = 3.00000012e-001, expected 1. device.c:1242: Test failed: Screen width is 640, expected 800. device.c:1243: Test failed: Screen height is 480, expected 600. device.c:1249: Test failed: Back buffer width is 640, expected 800. device.c:1251: Test failed: Back buffer height is 480, expected 600. device.c:1281: Test failed: Screen width is 640, expected 1024. device.c:1282: Test failed: Screen height is 480, expected 768. device.c:2437: Test failed: Expected foreground window 000500F6, got 0002004C. device.c:2808: Test failed: Expected device window style 0x14cf0000, got 0x4cf0000. device.c:2812: Test failed: Expected device window extended style 0x108, got 0x100. device.c:2834: Test failed: Failed to reset device, hr 0x88760868. device.c:2838: Test failed: Expected device window style 0x14cf0000, got 0x4cf0000. device.c:2842: Test failed: Expected device window extended style 0x108, got 0x100. device.c:3051: Test failed: Got unexpect width 640. device.c:3052: Test failed: Got unexpect height 480. device.c:3063: Test failed: Got unexpect width 640. device.c:3064: Test failed: Got unexpect height 480. device.c:3126: Test failed: Failed to reset device. device.c:3133: Test failed: Expected {0, 0, 1024, 768}, got {0, 0, 512, 384}. device.c:3141: Test failed: Expected wndproc 0x412ba0, got 0x412ba0.

1 0