August 2016 - Wine-devel - list.winehq.org

RE: TestBot job 25201 results: [PATCH] ntdll: Set Rip in for longjmp in RtlRestoreContext (try2)
by Daniel Lehman 18 Aug '16

18 Aug '16

=== wvistau64 (32 bit exception) === ntdll: exception.c:551: Test failed: 42: Unexpected eip 0x350000/0x350001 exception: unhandled exception 80000004 at 00350000 exception.c:548: Test failed: 42: Wrong exception code c0000027/80000004 exception.c:551: Test failed: 42: Unexpected eip 0x77baed97/0x350001 === w7u (32 bit exception) === ntdll: exception.c:551: Test failed: 42: Unexpected eip 0x3f0000/0x3f0001 exception: unhandled exception 80000004 at 003F0000 exception.c:548: Test failed: 42: Wrong exception code c0000027/80000004 exception.c:551: Test failed: 42: Unexpected eip 0x77cde34f/0x3f0001 === w8 (32 bit exception) === ntdll: exception.c:1261: Test failed: Got exception offset 0x1b, expected 0x19 exception.c:1270: Test failed: Got exception offset 0x19, expected 0x17 these are existing failures

1 0

Re: [PATCH] ntdll: Set Rip in for longjmp in RtlRestoreContext (try2)
by Marvin 18 Aug '16

18 Aug '16

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=25201 Your paranoid android. === wvistau64 (32 bit exception) === exception.c:551: Test failed: 42: Unexpected eip 0x350000/0x350001 exception: unhandled exception 80000004 at 00350000 exception.c:551: Test failed: 42: Unexpected eip 0x77baed97/0x350001 === w7u (32 bit exception) === exception.c:551: Test failed: 42: Unexpected eip 0x3f0000/0x3f0001 exception: unhandled exception 80000004 at 003F0000 exception.c:551: Test failed: 42: Unexpected eip 0x77cde34f/0x3f0001 === w1064 (32 bit exception) === An error occurred while waiting for the test to complete: network read timed out (wait2/connect:AgentVersion.h:0/9) The test VM has crashed, rebooted or lost connectivity (or the TestAgent server died) === w1064 (64 bit exception) === An error occurred while waiting for the test to complete: network read timed out (wait2/connect:AgentVersion.h:0/9) The test VM has crashed, rebooted or lost connectivity (or the TestAgent server died)

1 0

Re: [PATCH v7 1/3] shell32/tests: Add tests for FolderItems_Item and FolderItems_get_Count.
by Sebastian Lackner 18 Aug '16

18 Aug '16

On 18.08.2016 06:37, Alex Henrie wrote: > Cc: Sebastian Lackner <sebastian(a)fds-team.de> > > v7 fixes the use-after-free of V_BSTR(&var). > > Signed-off-by: Alex Henrie <alexhenrie24(a)gmail.com> > --- > dlls/shell32/tests/Makefile.in | 2 +- > dlls/shell32/tests/shelldispatch.c | 220 +++++++++++++++++++++++++++++++++++-- > 2 files changed, 210 insertions(+), 12 deletions(-) > > diff --git a/dlls/shell32/tests/Makefile.in b/dlls/shell32/tests/Makefile.in > index cfe0c50..ef95b52 100644 > --- a/dlls/shell32/tests/Makefile.in > +++ b/dlls/shell32/tests/Makefile.in > @@ -1,5 +1,5 @@ > TESTDLL = shell32.dll > -IMPORTS = shell32 ole32 oleaut32 user32 advapi32 > +IMPORTS = shell32 ole32 oleaut32 user32 advapi32 shlwapi > > C_SRCS = \ > appbar.c \ > diff --git a/dlls/shell32/tests/shelldispatch.c b/dlls/shell32/tests/shelldispatch.c > index 9a47c67..02be8fd 100644 > --- a/dlls/shell32/tests/shelldispatch.c > +++ b/dlls/shell32/tests/shelldispatch.c > @@ -25,13 +25,14 @@ > #include "shldisp.h" > #include "shlobj.h" > #include "shlwapi.h" > +#include "shellapi.h" > #include "winsvc.h" > #include "wine/test.h" > > #define EXPECT_HR(hr,hr_exp) \ > ok(hr == hr_exp, "got 0x%08x, expected 0x%08x\n", hr, hr_exp) > > -static const WCHAR winetestW[] = {'w','i','n','e','t','e','s','t',0}; > +static const WCHAR winetestW[] = {'w','i','n','e','t','e','s','t',0,0}; > > static HRESULT (WINAPI *pSHGetFolderPathW)(HWND, int, HANDLE, DWORD, LPWSTR); > static HRESULT (WINAPI *pSHGetNameFromIDList)(PCIDLIST_ABSOLUTE,SIGDN,PWSTR*); > @@ -310,19 +311,46 @@ static void test_namespace(void) > > static void test_items(void) > { > - WCHAR wstr[MAX_PATH], orig_dir[MAX_PATH]; > + static const struct > + { > + char name[32]; > + enum > + { > + DIRECTORY, > + SHORTCUT, > + EMPTY, > + RANDOM, > + } > + type; > + } > + file_defs[] = > + { > + { "00-Myfolder", DIRECTORY }, > + { "01-empty.lnk", EMPTY }, Not really an issue, but is it intentional that the empty file has a .lnk extension? > + { "02-shortcut.lnk", SHORTCUT }, > + { "03-empty.bin", EMPTY }, > + { "04-random.bin", RANDOM }, > + }; > + WCHAR wstr[MAX_PATH], wstr2[MAX_PATH], orig_dir[MAX_PATH]; > HRESULT r; > IShellDispatch *sd = NULL; > Folder *folder = NULL; > FolderItems *items = NULL; > FolderItems2 *items2 = NULL; > FolderItems3 *items3 = NULL; > - FolderItem *item = (FolderItem*)0xdeadbeef; > + FolderItem *item = (FolderItem*)0xdeadbeef, *item2 = (FolderItem*)0xdeadbeef; > IDispatch *disp = NULL; > IUnknown *unk = NULL; > FolderItemVerbs *verbs = (FolderItemVerbs*)0xdeadbeef; > VARIANT var; > LONG lcount = -1; > + HANDLE file; > + DWORD dwcount; > + IPersistFile *shortcut = NULL; > + BSTR bstr; > + SHFILEOPSTRUCTW del; > + char buf[512]; > + int i, j; > > r = CoCreateInstance(&CLSID_Shell, NULL, CLSCTX_INPROC_SERVER, &IID_IShellDispatch, (void**)&sd); > ok(SUCCEEDED(r), "CoCreateInstance failed: %08x\n", r); > @@ -345,13 +373,6 @@ static void test_items(void) > r = Folder_Items(folder, &items); > ok(r == S_OK, "Folder::Items failed: %08x\n", r); > ok(!!items, "items is null\n"); > - r = FolderItems_QueryInterface(items, &IID_FolderItems2, (void**)&items2); > - ok(r == S_OK || broken(E_NOINTERFACE) /* xp and later */, "FolderItems::QueryInterface failed: %08x\n", r); > - ok(!!items2 || broken(!items2) /* xp and later */, "items2 is null\n"); > - r = FolderItems_QueryInterface(items, &IID_FolderItems3, (void**)&items3); > - ok(r == S_OK, "FolderItems::QueryInterface failed: %08x\n", r); > - ok(!!items3, "items3 is null\n"); > - Folder_Release(folder); > > if (0) /* crashes on all versions of Windows */ > r = FolderItems_get_Count(items, NULL); > @@ -373,6 +394,179 @@ todo_wine > ok(r == S_FALSE, "expected S_FALSE, got %08x\n", r); > ok(!item, "item is not null\n"); > > + for (i = 0; i < sizeof(file_defs)/sizeof(file_defs[0]); i++) > + { > + switch (file_defs[i].type) > + { > + case DIRECTORY: > + CreateDirectoryA(file_defs[i].name, NULL); > + PathCombineA(buf, file_defs[i].name, "foo.txt"); > + CloseHandle(CreateFileA(buf, 0, 0, NULL, CREATE_ALWAYS, FILE_ATTRIBUTE_NORMAL, NULL)); > + break; > + > + case SHORTCUT: > + CoCreateInstance(&CLSID_ShellLink, NULL, CLSCTX_INPROC_SERVER, &IID_IPersistFile, (void**)&shortcut); > + MultiByteToWideChar(CP_ACP, 0, file_defs[i].name, -1, wstr, sizeof(wstr)/sizeof(WCHAR)); > + IPersistFile_Save(shortcut, wstr, FALSE); > + IPersistFile_Release(shortcut); > + break; > + > + case EMPTY: > + CloseHandle(CreateFileA(file_defs[i].name, 0, 0, NULL, CREATE_ALWAYS, FILE_ATTRIBUTE_NORMAL, NULL)); > + break; > + > + case RANDOM: > + file = CreateFileA(file_defs[i].name, GENERIC_WRITE, 0, NULL, > + CREATE_ALWAYS, FILE_ATTRIBUTE_NORMAL, NULL); > + for (j = 0; j < sizeof(buf); j++) > + buf[j] = rand() % 256; > + WriteFile(file, buf, sizeof(buf), &dwcount, NULL); > + CloseHandle(file); > + break; I think a few checks would not hurt in the code above. > + } > + } > + > + lcount = -1; > + r = FolderItems_get_Count(items, &lcount); > +todo_wine > + ok(r == S_OK, "FolderItems::get_Count failed: %08x\n", r); > +todo_wine > + ok(!lcount, "expected 0 files, got %d\n", lcount); > + FolderItems_Release(items); > + > + items = NULL; > + r = Folder_Items(folder, &items); > + ok(r == S_OK, "Folder::Items failed: %08x\n", r); > + ok(!!items, "items is null\n"); > + r = FolderItems_QueryInterface(items, &IID_FolderItems2, (void**)&items2); > + ok(r == S_OK || broken(E_NOINTERFACE) /* xp and later */, "FolderItems::QueryInterface failed: %08x\n", r); > + if (r == S_OK) ok(!!items2, "items2 is null\n"); > + r = FolderItems_QueryInterface(items, &IID_FolderItems3, (void**)&items3); > + ok(r == S_OK, "FolderItems::QueryInterface failed: %08x\n", r); > + ok(!!items3, "items3 is null\n"); Isn't it necessary to release those items{2,3} interfaces? > + Folder_Release(folder); > + > + lcount = -1; > + r = FolderItems_get_Count(items, &lcount); > +todo_wine > + ok(r == S_OK, "FolderItems::get_Count failed: %08x\n", r); > +todo_wine > + ok(lcount == sizeof(file_defs)/sizeof(file_defs[0]), > + "expected %d files, got %d\n", sizeof(file_defs)/sizeof(file_defs[0]), lcount); > + > + VariantInit(&var); > + for (i = -10; i < 0x10000; i++) > + { > + item = NULL; > + V_VT(&var) = i; Is such a bruteforce approach also used in other places of Wine? I personally don't think it makes much sense. It causes a huge amount of FIXMEs although there are just a few supported types. For debugging purposes something like this is probably fine, but it might be preferred to test only "reasonable" stuff. Please also note that that VARTYPE is unsigned short, so negative numbers do not make sense. Very high numbers correspond to vector/array/... which also don't make sense, especially when they are not properly initialized. > + r = FolderItems_Item(items, var, &item); > + if (i == VT_I2 || i == VT_I4 || i == VT_ERROR) > + { > +todo_wine > + ok(r == S_OK, "type %d: expected S_OK, got %08x\n", i, r); > +todo_wine > + ok(!!item, "item is null\n"); > + } > + else if (i == VT_BSTR) > + { > +todo_wine > + ok(r == S_FALSE, "type %d: expected S_FALSE, got %08x\n", i, r); > + ok(!item, "item is not null\n"); > + } > + else > + { > + ok(r == E_NOTIMPL, "type %d: expected E_NOTIMPL, got %08x\n", i, r); > + ok(!item, "item is not null\n"); > + } > + if (item) FolderItem_Release(item); > + } > + > + for (i = -10; i < 10; i++) > + { Could you explain why this test is repeated 20 times? It does not look like you actually use the counter variable anywhere. > + item = NULL; > + V_VT(&var) = VT_ERROR; > + V_ERROR(&var) = 0; > + r = FolderItems_Item(items, var, &item); > +todo_wine > + ok(r == S_OK, "expected S_OK, got %08x\n", r); > +todo_wine > + ok(!!item, "item is null\n"); > + if (!item) continue; > + > + bstr = NULL; > + r = FolderItem_get_Path(item, &bstr); > + ok(r == S_OK, "FolderItem::get_Path failed: %08x\n", r); > + ok(!!bstr, "bstr is null\n"); > + GetCurrentDirectoryW(MAX_PATH, wstr); > + GetLongPathNameW(wstr, wstr2, MAX_PATH); > + ok(!lstrcmpW(bstr, wstr2), > + "expected %s, got %s\n", wine_dbgstr_w(wstr2), wine_dbgstr_w(bstr)); > + > + SysFreeString(bstr); > + FolderItem_Release(item); > + } > + > + item = NULL; > + V_VT(&var) = VT_I4; > + V_I4(&var) = -1; > + r = FolderItems_Item(items, var, &item); > +todo_wine > + ok(r == S_FALSE, "expected S_FALSE, got %08x\n", r); > + ok(!item, "item is not null\n"); Not sure if its critical, but the test isn't really convincing when you already initialize item to the correct value. It might be better to set it to something else. The same also applies to similar code below. > + > + for (i = 0; i < sizeof(file_defs)/sizeof(file_defs[0]); i++) > + { > + for (j = 0; j < 2; j++) > + { > + item = NULL; It might be better to do the initialization directly in front of the call. > + MultiByteToWideChar(CP_ACP, 0, file_defs[i].name, -1, wstr, MAX_PATH); > + > + if (j == 0) > + { > + V_VT(&var) = VT_I4; > + V_I4(&var) = i; > + } > + else > + { > + V_VT(&var) = VT_BSTR; > + V_BSTR(&var) = SysAllocString(wstr); > + } > + > + r = FolderItems_Item(items, var, &item); > +todo_wine > + ok(r == S_OK, "file_defs[%d], %d: FolderItems::Item failed: %08x\n", i, j, r); > +todo_wine > + ok(!!item, "file_defs[%d], %d: item is null\n", i, j); > + if (!item) continue; > + You probably should also initialize item2 here. > + r = FolderItems_Item(items, var, &item2); > + ok(r == S_OK, "file_defs[%d], %d: FolderItems::Item failed: %08x\n", i, j, r); > + ok(item2 != item, "file_defs[%d], %d: item and item2 are the same\n", i, j); > + FolderItem_Release(item2); > + if (V_VT(&var) == VT_BSTR) SysFreeString(V_BSTR(&var)); > + > + bstr = NULL; > + r = FolderItem_get_Path(item, &bstr); > + ok(r == S_OK, "file_defs[%d], %d: FolderItem::get_Path failed: %08x\n", i, j, r); > + ok(!!bstr, "file_defs[%d], %d: bstr is null\n", i, j); > + GetFullPathNameW(wstr, MAX_PATH, wstr2, NULL); > + GetLongPathNameW(wstr2, wstr, MAX_PATH); > + ok(!lstrcmpW(bstr, wstr), > + "file_defs[%d], %d: expected %s, got %s\n", i, j, wine_dbgstr_w(wstr), wine_dbgstr_w(bstr)); > + SysFreeString(bstr); > + > + FolderItem_Release(item); > + } > + } > + > + V_VT(&var) = VT_I4; > + V_I4(&var) = sizeof(file_defs)/sizeof(file_defs[0]); > + item = NULL; > + r = FolderItems_Item(items, var, &item); > +todo_wine > + ok(r == S_FALSE, "expected S_FALSE, got %08x\n", r); > + ok(!item, "item is not null\n"); > + > if (0) /* crashes on xp */ > { > r = FolderItems_get_Application(items, NULL); > @@ -430,7 +624,11 @@ todo_wine > > GetTempPathW(MAX_PATH, wstr); > SetCurrentDirectoryW(wstr); > - RemoveDirectoryW(winetestW); > + memset(&del, 0, sizeof(del)); > + del.wFunc = FO_DELETE; > + del.pFrom = winetestW; > + del.fFlags = FOF_NO_UI; > + SHFileOperationW(&del); > SetCurrentDirectoryW(orig_dir); > > FolderItems_Release(items); >

2 1

Re: [PATCH 02/13] wineplugplay.sys: Initialize device driver store
by Sebastian Lackner 18 Aug '16

18 Aug '16

On 17.08.2016 15:36, Aric Stewart wrote: > We are not handling installing drivers just yet, so we just make use of the > critical device store. These are drivers that, on Windows, are automatically > loaded on boot every time. This database is maintained in > [CurrentControlSet/Control/CriticalDeviceDatabase] The keys in this registry key > represent a HardwareID that match a device. We query the IDs from the bus > device’s BusQueryHardwareIDs and if we find a match (from most specific to most > general) we look at that registry key. The CriticalDeviceDatabase entry > specify a [Service] value representing the driver. > > Signed-off-by: Aric Stewart <aric(a)codeweavers.com> > --- > dlls/wineplugplay.sys/Makefile.in | 5 +- > dlls/wineplugplay.sys/main.c | 14 ++++ > dlls/wineplugplay.sys/pnp.h | 20 +++++ > dlls/wineplugplay.sys/pnp_manager.c | 162 ++++++++++++++++++++++++++++++++++++ > 4 files changed, 199 insertions(+), 2 deletions(-) > create mode 100644 dlls/wineplugplay.sys/pnp.h > create mode 100644 dlls/wineplugplay.sys/pnp_manager.c > > > > 0002-wineplugplay.sys-Initialize-device-driver-store.txt > > > diff --git a/dlls/wineplugplay.sys/Makefile.in b/dlls/wineplugplay.sys/Makefile.in > index 4039a8d..618dac7 100644 > --- a/dlls/wineplugplay.sys/Makefile.in > +++ b/dlls/wineplugplay.sys/Makefile.in > @@ -1,6 +1,7 @@ > MODULE = wineplugplay.sys > -IMPORTS = ntoskrnl > +IMPORTS = ntoskrnl advapi32 > EXTRADLLFLAGS = -Wb,--subsystem,native > > C_SRCS = \ > - main.c > + main.c \ > + pnp_manager.c > diff --git a/dlls/wineplugplay.sys/main.c b/dlls/wineplugplay.sys/main.c > index 51b5d97..4ce641b 100644 > --- a/dlls/wineplugplay.sys/main.c > +++ b/dlls/wineplugplay.sys/main.c > @@ -30,11 +30,25 @@ > #include "wine/unicode.h" > #include "wine/debug.h" > > +#include "pnp.h" > + > WINE_DEFAULT_DEBUG_CHANNEL(plugplay); > > +static DWORD CALLBACK Initialize_thread(void *args) > +{ > + Initialize_DriverStore(); > + return 0; > +} > + > /* main entry point for the driver */ > NTSTATUS WINAPI DriverEntry( DRIVER_OBJECT *driver, UNICODE_STRING *path ) > { > TRACE("\n"); > + /* We are doing this in a new thread because we want this DriverEntry > + to return as immediently as possible. The loader lock will prevent > + the thread from starting immediently so this service can finish > + loading before the Initialization start trying to load more > + services */ > + CreateThread(NULL, 0, Initialize_thread, NULL, 0, NULL); > return STATUS_SUCCESS; > } > diff --git a/dlls/wineplugplay.sys/pnp.h b/dlls/wineplugplay.sys/pnp.h > new file mode 100644 > index 0000000..db18202 > --- /dev/null > +++ b/dlls/wineplugplay.sys/pnp.h > @@ -0,0 +1,20 @@ > +/* > + * Copyright 2016 Aric Stewart > + * > + * This library is free software; you can redistribute it and/or > + * modify it under the terms of the GNU Lesser General Public > + * License as published by the Free Software Foundation; either > + * version 2.1 of the License, or (at your option) any later version. > + * > + * This library is distributed in the hope that it will be useful, > + * but WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > + * Lesser General Public License for more details. > + * > + * You should have received a copy of the GNU Lesser General Public > + * License along with this library; if not, write to the Free Software > + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA > + */ > + > +/* Plug and Play Manager */ > +void Initialize_DriverStore(void) DECLSPEC_HIDDEN; > diff --git a/dlls/wineplugplay.sys/pnp_manager.c b/dlls/wineplugplay.sys/pnp_manager.c > new file mode 100644 > index 0000000..0e10240 > --- /dev/null > +++ b/dlls/wineplugplay.sys/pnp_manager.c > @@ -0,0 +1,162 @@ > +/* HID plug and play Manager like functionality > + * > + * Copyright 2016 CodeWeavers, Aric Stewart > + * > + * This library is free software; you can redistribute it and/or > + * modify it under the terms of the GNU Lesser General Public > + * License as published by the Free Software Foundation; either > + * version 2.1 of the License, or (at your option) any later version. > + * > + * This library is distributed in the hope that it will be useful, > + * but WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > + * Lesser General Public License for more details. > + * > + * You should have received a copy of the GNU Lesser General Public > + * License along with this library; if not, write to the Free Software > + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA > + */ > + > +#include <stdarg.h> > +#include "ntstatus.h" > +#define WIN32_NO_STATUS > +#include "windef.h" > +#include "winbase.h" > +#include "winternl.h" > +#include "ddk/wdm.h" > +#include "cfgmgr32.h" > +#include "winsvc.h" > +#include "winreg.h" > +#include "wine/debug.h" > +#include "wine/list.h" > + > +#include "pnp.h" > + > +WINE_DEFAULT_DEBUG_CHANNEL(plugplay); > + > +#define MAX_SERVICE_NAME 260 > + > +typedef struct _device_driver { > + struct list entry; > + > + WCHAR **matching_ids; > + INT id_count; > + > + WCHAR *driver_name; > +} device_driver; > + > +struct list device_drivers = LIST_INIT(device_drivers); > + > +static device_driver* load_device_driver(const WCHAR *drivername, const WCHAR* match) > +{ > + device_driver *ptr; > + > + LIST_FOR_EACH_ENTRY(ptr, &device_drivers, device_driver, entry) > + { > + if (lstrcmpW(drivername, ptr->driver_name) == 0) > + { > + int i; > + /* Check matching ids */ > + for (i = 0; i < ptr->id_count; i++) > + if (lstrcmpW(match, ptr->matching_ids[i]) == 0) > + return ptr; > + /* Add a new matching id, this should be reletivly rare, > + no exponental growth needed */ > + ptr->matching_ids = HeapReAlloc(GetProcessHeap(), 0, ptr->matching_ids, sizeof(WCHAR*) * (ptr->id_count+1)); > + ptr->matching_ids[ptr->id_count] = HeapAlloc(GetProcessHeap(), 0, sizeof(WCHAR) * (lstrlenW(match)+1)); > + lstrcpyW(ptr->matching_ids[ptr->id_count], match); Such code would be easier to read with a strdupW() helper function, like in many other places of Wine. The same also applies to some places below. > + ptr->id_count++; > + > + return ptr; > + } > + } > + > + ptr = HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY, sizeof(*ptr)); > + ptr->driver_name = HeapAlloc(GetProcessHeap(), 0, sizeof(WCHAR) * (lstrlenW(drivername)+1)); > + lstrcpyW(ptr->driver_name, drivername); > + ptr->matching_ids = HeapAlloc(GetProcessHeap(), 0, sizeof(WCHAR*)); > + ptr->matching_ids[0] = HeapAlloc(GetProcessHeap(), 0, sizeof(WCHAR) * (lstrlenW(match)+1)); > + lstrcpyW(ptr->matching_ids[0], match); > + ptr->id_count = 1; > + list_add_tail(&device_drivers, &ptr->entry); > + > + return ptr; > +} > + > +void Initialize_DriverStore(void) > +{ > + static const WCHAR criticalW[] = > + { 'S','y','s','t','e','m','\\', > + 'C','u','r','r','e','n','t','C','o','n','t','r','o','l','S','e','t','\\', > + 'C','o','n','t','r','o','l', '\\', > + 'C','r','i','t','i','c','a','l','D','e','v','i','c','e','D','a','t','a','b','a','s','e', 0}; > + static const WCHAR serviceW[] = {'S','e','r','v','i','c','e',0}; > + HKEY hkey_root; > + WCHAR szMatch[MAX_DEVICE_ID_LEN]; > + SC_HANDLE service_manager; > + > + int i; > + > + if (RegOpenKeyW(HKEY_LOCAL_MACHINE, criticalW, &hkey_root)) > + { > + TRACE("No drivers found\n"); > + return; > + } > + > + service_manager = OpenSCManagerW(NULL, NULL, GENERIC_READ|GENERIC_EXECUTE); > + > + for (i = 0; TRUE; i++) > + { > + HKEY hkey_service; > + DWORD rc; > + device_driver *driver; > + DWORD len = 0; > + WCHAR szService[MAX_SERVICE_NAME]; > + SC_HANDLE service; > + > + rc = RegEnumKeyW(hkey_root, i, szMatch, MAX_DEVICE_ID_LEN); > + if (rc == ERROR_NO_MORE_ITEMS) > + break; > + > + if (rc != 0) > + { > + ERR("Error %d reading key %d - skipping\n", rc, i); > + continue; > + } > + > + if (RegOpenKeyW(hkey_root, szMatch, &hkey_service)) > + { > + ERR("Error opening key %s - skipping\n",debugstr_w(szMatch)); > + continue; > + } > + > + len = sizeof(szService); > + rc = RegQueryValueExW(hkey_service, serviceW, NULL, NULL, (BYTE*)szService, &len); > + if (rc != ERROR_SUCCESS) > + { > + ERR("Error querying service from key %s - skipping\n",debugstr_w(szMatch)); > + RegCloseKey(hkey_service); > + continue; > + } > + RegCloseKey(hkey_service); > + > + service = OpenServiceW(service_manager, szService, GENERIC_READ|GENERIC_EXECUTE); > + if (!service) > + { > + ERR("Unable to open service %s - skipping\n",debugstr_w(szService)); > + continue; > + } > + > + TRACE("Loading service '%s' to match '%s'\n", debugstr_w(szService), debugstr_w(szMatch)); Why are there so many spaces in the middle of the line? > + > + driver = load_device_driver(szService, szMatch); > + if (!driver) > + ERR("failed to load driver %s\n", debugstr_w(szService)); > + > + StartServiceW(service, 0, NULL); > + CloseServiceHandle(service); The registry parsing code is probably fine, but wouldn't it be easier to handle this similar to other autostart services in services.exe? Then there would also be no need to spawn a separate thread for initialization. > + } > + > + CloseServiceHandle(service_manager); > + RegCloseKey(hkey_root); > +} > > >

2 2

Re: [PATCH] api-ms-win-core-console-l2-1-0: Add stub dll.
by Hugh McMaster 18 Aug '16

18 Aug '16

On Wednesday, 17 August 2016 8:39 PM, Hadrien Boizard wrote: > +@ stub AttachConsole > +@ stub CreateConsoleScreenBuffer > +@ stub FillConsoleOutputAttribute > +@ stub FillConsoleOutputCharacterA > +@ stub FillConsoleOutputCharacterW > +@ stub FlushConsoleInputBuffer > +@ stub FreeConsole > +@ stub GenerateConsoleCtrlEvent > +@ stub GetConsoleCursorInfo > +@ stub GetConsoleScreenBufferInfo > +@ stub GetConsoleScreenBufferInfoEx > +@ stub GetConsoleTitleW > +@ stub GetLargestConsoleWindowSize > +@ stub PeekConsoleInputW > +@ stub ReadConsoleOutputA > +@ stub ReadConsoleOutputAttribute > +@ stub ReadConsoleOutputCharacterA > +@ stub ReadConsoleOutputCharacterW > +@ stub ReadConsoleOutputW > +@ stub ScrollConsoleScreenBufferA > +@ stub ScrollConsoleScreenBufferW > +@ stub SetConsoleActiveScreenBuffer > +@ stub SetConsoleCP > +@ stub SetConsoleCursorInfo > +@ stub SetConsoleCursorPosition > +@ stub SetConsoleOutputCP > +@ stub SetConsoleScreenBufferInfoEx > +@ stub SetConsoleScreenBufferSize > +@ stub SetConsoleTextAttribute > +@ stub SetConsoleTitleW > +@ stub SetConsoleWindowInfo > +@ stub WriteConsoleInputA > +@ stub WriteConsoleInputW > +@ stub WriteConsoleOutputA > +@ stub WriteConsoleOutputAttribute > +@ stub WriteConsoleOutputCharacterA > +@ stub WriteConsoleOutputCharacterW > +@ stub WriteConsoleOutputW Hi Hadrien, Most, if not all, of these functions are implemented in dlls/kernel32. You should forward these calls to kernel32, making sure to change 'stub' to 'stdcall' as necessary. -- Hugh McMaster

2 1

Re: msvcrt: Fix strtok_s and _mbstok_s for empty strings. (v2)
by Piotr Caban 18 Aug '16

18 Aug '16

Hi, On 08/18/16 11:06, Bernhard Übelacker wrote: > + if(!p_strtok_s) { > + win_skip("strtok_s not found.\n"); > + return; > + } There's one more thing that I missed earlier - strtok_s/mbstok_s is always available in msvcr90. There's no point in checking it. Sorry for not spotting it earlier. Thanks you, Piotr

1 0

Re: msvcrt: Fix strtok_s for an empty string.
by Piotr Caban 18 Aug '16

18 Aug '16

Hi Bernhard, On 08/18/16 03:48, Bernhard Übelacker wrote: > + old_handler = p_set_invalid_parameter_handler(test_strtok_s_handler); I'm not sure if it makes sense to set invalid parameter handler taking in account it's never called. If you don't set it and it's called the tests will crash. > + context = (char*)0xdeadbeef; > + strret = p__mbstok_s( test, "/", &context); This will cause compilation warnings (signedness mismatch). Thanks, Piotr

2 1

Re: ntdll/tests: Dump the handle table on failure.
by Huw Davies 18 Aug '16

18 Aug '16

> On 17 Aug 2016, at 14:54, Sebastian Lackner <sebastian(a)fds-team.de> wrote: > > On 17.08.2016 14:45, Huw Davies wrote: >> Signed-off-by: Huw Davies <huw(a)codeweavers.com> >> --- >> dlls/ntdll/tests/info.c | 6 +++++- >> 1 file changed, 5 insertions(+), 1 deletion(-) >> > > I assume this is for bug 40106, right? I also investigated it back then, > and it looked like a kernel bug to me. If I remember correctly it was > somehow triggered by a large number of allocated handles. Yes, I was wondering whether it was a simple as needing to compare the LOWORD of the handle. Huw.

2 3

Re: [v4 PATCH] kernel32: Fix RegisterWaitForSingleObject for console handles
by Marvin 18 Aug '16

18 Aug '16

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=25182 Your paranoid android. === w8 (32 bit console) === console.c:1497: Test failed: [13] Expected output count to be 0, got 1354796060 console.c:1497: Test failed: [15] Expected output count to be 0, got 1354796060 console.c:1497: Test failed: [18] Expected output count to be 3735928559, got 1354796060 console.c:1741: Test failed: [13] Expected output count to be 0, got 1354796060 console.c:1741: Test failed: [15] Expected output count to be 0, got 1354796060 console.c:1741: Test failed: [18] Expected output count to be 3735928559, got 1354796060 === w864 (32 bit console) === console.c:2619: Test failed: expected ERROR_INVALID_HANDLE, got 1 console.c:2625: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:2632: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:2639: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:1322: Test failed: expected VerifyConsoleIoHandle to succeed console.c:1497: Test failed: [3] Expected output count to be 3735928559, got 0 console.c:1497: Test failed: [11] Expected output count to be 3735928559, got 0 console.c:1497: Test failed: [13] Expected output count to be 0, got 1429009712 console.c:1497: Test failed: [15] Expected output count to be 0, got 1429009712 console.c:1497: Test failed: [18] Expected output count to be 3735928559, got 1429009712 console.c:1741: Test failed: [3] Expected output count to be 3735928559, got 0 console.c:1741: Test failed: [11] Expected output count to be 3735928559, got 0 console.c:1741: Test failed: [13] Expected output count to be 0, got 1429009712 console.c:1741: Test failed: [15] Expected output count to be 0, got 1429009712 console.c:1741: Test failed: [18] Expected output count to be 3735928559, got 1429009712 console.c:2200: Test failed: [7] Expected count to be 0, got 1 console.c:2266: Test failed: [7] Expected count to be 0, got 1 console.c:2332: Test failed: [7] Expected count to be 0, got 1 === w1064 (32 bit console) === console.c:2619: Test failed: expected ERROR_INVALID_HANDLE, got 1 console.c:2625: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:2632: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:2639: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:1322: Test failed: expected VerifyConsoleIoHandle to succeed console.c:1497: Test failed: [3] Expected output count to be 3735928559, got 0 console.c:1497: Test failed: [11] Expected output count to be 3735928559, got 0 console.c:1497: Test failed: [13] Expected output count to be 0, got 4210212470 console.c:1497: Test failed: [15] Expected output count to be 0, got 4210212470 console.c:1497: Test failed: [18] Expected output count to be 3735928559, got 4210212470 console.c:1741: Test failed: [3] Expected output count to be 3735928559, got 0 console.c:1741: Test failed: [11] Expected output count to be 3735928559, got 0 console.c:1741: Test failed: [13] Expected output count to be 0, got 4210212470 console.c:1741: Test failed: [15] Expected output count to be 0, got 4210212470 console.c:1741: Test failed: [18] Expected output count to be 3735928559, got 4210212470 console.c:2200: Test failed: [7] Expected count to be 0, got 1 console.c:2266: Test failed: [7] Expected count to be 0, got 1 console.c:2332: Test failed: [7] Expected count to be 0, got 1 === w864 (64 bit console) === console.c:2619: Test failed: expected ERROR_INVALID_HANDLE, got 1 console.c:2625: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:2632: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:2639: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:1322: Test failed: expected VerifyConsoleIoHandle to succeed console.c:1497: Test failed: [3] Expected output count to be 3735928559, got 0 console.c:1497: Test failed: [11] Expected output count to be 3735928559, got 0 console.c:1497: Test failed: [18] Expected output count to be 3735928559, got 0 console.c:1741: Test failed: [3] Expected output count to be 3735928559, got 0 console.c:1741: Test failed: [11] Expected output count to be 3735928559, got 0 console.c:1741: Test failed: [18] Expected output count to be 3735928559, got 0 console.c:2200: Test failed: [7] Expected count to be 0, got 1 console.c:2266: Test failed: [7] Expected count to be 0, got 1 console.c:2332: Test failed: [7] Expected count to be 0, got 1 === w1064 (64 bit console) === console.c:2619: Test failed: expected ERROR_INVALID_HANDLE, got 1 console.c:2625: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:2632: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:2639: Test failed: expected ERROR_NOT_ENOUGH_MEMORY, got 998 console.c:1322: Test failed: expected VerifyConsoleIoHandle to succeed console.c:1497: Test failed: [3] Expected output count to be 3735928559, got 0 console.c:1497: Test failed: [11] Expected output count to be 3735928559, got 0 console.c:1497: Test failed: [18] Expected output count to be 3735928559, got 0 console.c:1741: Test failed: [3] Expected output count to be 3735928559, got 0 console.c:1741: Test failed: [11] Expected output count to be 3735928559, got 0 console.c:1741: Test failed: [18] Expected output count to be 3735928559, got 0 console.c:2200: Test failed: [7] Expected count to be 0, got 1 console.c:2266: Test failed: [7] Expected count to be 0, got 1 console.c:2332: Test failed: [7] Expected count to be 0, got 1

1 0

wine-1.8.4-rc2 Available
by Michael Stefaniuc 17 Aug '16

17 Aug '16

Hello guys, wine-1.8.4-rc2 is now available, including the tag: https://github.com/mstefani/wine-stable https://static-winehq.rhcloud.com/stable/ Three more fixes added on Nikolay's request: - t2embed: Fix TTIsEmbeddingEnabled() to properly use otmpFaceName metrics field. - dwrite: Sanitize specified stretch value, valid range is [1, 9]. - dwrite: Fix font stretch name conversion table. As the release commit changed this will be a forced update. bye michael

1 1