July 2012 - Wine-Devel - list.winehq.org

[RFC] kernel32: Try to get the drive type from the mountmgr first
by GOUJON Alexandre July 28, 2012

July 28, 2012

On 07/22/2012 03:36 PM, Alexandre Goujon wrote: > Should fix bug #17037 > --- > dlls/kernel32/tests/volume.c | 2 ++ > dlls/kernel32/volume.c | 3 ++- > 2 files changed, 4 insertions(+), 1 deletion(-) > > diff --git a/dlls/kernel32/tests/volume.c b/dlls/kernel32/tests/volume.c > index 8152ae1..175095f 100644 > --- a/dlls/kernel32/tests/volume.c > +++ b/dlls/kernel32/tests/volume.c > @@ -908,6 +908,8 @@ static void test_cdrom_ioctl(void) > handle = CreateFileA(drive_full_path, GENERIC_READ, FILE_SHARE_READ|FILE_SHARE_WRITE, NULL, OPEN_EXISTING, 0, 0); > if(handle == INVALID_HANDLE_VALUE) > { > + /* GetDriveType also reports DRIVE_CDROM even if there is no disc in is the drive. > + In this case, the last error is ERROR_ACCESS_DENIED */ > trace("Failed to open the device : %u\n", GetLastError()); > continue; > } > diff --git a/dlls/kernel32/volume.c b/dlls/kernel32/volume.c > index 8377e41..cdd4e46 100644 > --- a/dlls/kernel32/volume.c > +++ b/dlls/kernel32/volume.c > @@ -1538,6 +1538,7 @@ UINT WINAPI GetDriveTypeW(LPCWSTR root) /* [in] String describing drive */ > HANDLE handle; > UINT ret; > > + if ((ret = get_mountmgr_drive_type( root )) != DRIVE_UNKNOWN) return ret; > if (!open_device_root( root, &handle )) return DRIVE_NO_ROOT_DIR; > > status = NtQueryVolumeInformationFile( handle, &io, &info, sizeof(info), FileFsDeviceInformation ); > @@ -1557,7 +1558,7 @@ UINT WINAPI GetDriveTypeW(LPCWSTR root) /* [in] String describing drive */ > case FILE_DEVICE_DISK_FILE_SYSTEM: > if (info.Characteristics & FILE_REMOTE_DEVICE) ret = DRIVE_REMOTE; > else if (info.Characteristics & FILE_REMOVABLE_MEDIA) ret = DRIVE_REMOVABLE; > - else if ((ret = get_mountmgr_drive_type( root )) == DRIVE_UNKNOWN) ret = DRIVE_FIXED; > + else ret = DRIVE_FIXED; > break; > default: > ret = DRIVE_UNKNOWN; Hi, This one is marked as Pending meaning it's unclear at first glance why this change is needed. So here's a detailed explanation. The app in bug #17037 calls GetDriveType on each drive from A to Z looking for a CD drive. It appears genuine GetDriveType returns DRIVE_CDROM even if no disc is in the drive. Let's suppose D is the letter mapped to your CD drive. Calling GetDriveTypeW( L"D:\" ) will first try to get a handle from open_device_root which fails. The last error is STATUS_OBJECT_PATH_NOT_FOUND, hence the bug. In fact, open_device_root calls NtOpenFile with \??\D:\. Following the call chain, the culprit is nt_to_unix_file_name_attr which calls stat() on "/home/<yourlogin>/.wine/dosdevices/d:/". This file doesn't exist because the drive has no mount point but "/home/alex/.wine/dosdevices/d::/" does exist. Either I had to change nt_to_unix_file_name_attr to call stat(D::) in case where stat(D:) fails or use the mountmgr. And as get_mountmgr_drive_type was already called in GetDriveType, I though moving the call before open_device_root was simpler and less invasive than modifying nt_to_unix_file_name_attr. Is it the proper way to fix the bug ? Any idea on how to get that one in ?

1 0

Re: kernel32/tests: Make initonce_callback() static.
by Marvin July 28, 2012

July 28, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=20386 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH] msi/tests: Use always static binding for GetTokenInformation
by Marvin July 28, 2012

July 28, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=20375 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: mshtml: Properly handle NULL VT_DISPATCH refChild in IHTMLDOMNode::insertBefore
by Marvin July 28, 2012

July 28, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=20372 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [2/5] windowscodecs: Add stub IWICMetadataBlockReader to PNG decoder.
by Marvin July 28, 2012

July 28, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=20370 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [3/3] kernel32: Some tests for LOCALE_SNAME and neutral locales
by Marvin July 28, 2012

July 28, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=20368 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [1/3] kernel32: Some test for handling locales specified by language only
by Marvin July 28, 2012

July 28, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=20366 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 2/2] d3drm: Implement IDirect3DRMMeshImpl_SetGroupMaterial and forward material from mesh builder to created mesh
by Marvin July 28, 2012

July 28, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=20365 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [1/5] windowscodecs: Add test for IWICMetadataBlockReader interface.
by Marvin July 28, 2012

July 28, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=20362 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [2/3] kernel32: Fix base for value returned with LOCALE_RETURN_NUMBER
by Marvin July 28, 2012

July 28, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=20359 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0