February 2017 - Wine-devel - list.winehq.org

Re: [PATCH] iphlpapi: Implement AllocateAndGetTcpExTableFromStack (try 2).
by Hans Leidekker 07 Feb '17

07 Feb '17

On Mon, 2017-02-06 at 21:38 -0700, Erich E. Hoover wrote: > A variety of applications would like to see this function (Bug > #34372). This time without breaking the tests on XP/Win2k3... This export has been removed from Vista. Do these apps break if Wine is set to emulate Vista?

1 0

Re: [PATCH] iphlpapi: Implement AllocateAndGetTcpExTableFromStack.
by Marvin 07 Feb '17

07 Feb '17

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=28372 Your paranoid android. === wxppro (32 bit iphlpapi) === iphlpapi: unhandled exception c0000005 at 76D6F01F === w2003std (32 bit iphlpapi) === iphlpapi: unhandled exception c0000005 at 76CFFAD7

1 0

Re: [1/2] iphlpapi: Add GetUnicastIpAddressEntry implementation (try 3)
by Bruno Jesus 07 Feb '17

07 Feb '17

On Sun, Feb 5, 2017 at 7:25 PM, André Hentschel <nerv(a)dawncrow.de> wrote: > Signed-off-by: André Hentschel <nerv(a)dawncrow.de> > --- > > > For https://bugs.winehq.org/show_bug.cgi?id=41753 > try 2: Many thanks to Bruno and Hans for the feedback! > try 3: Proper patch numbering Thanks for the new patch ;-) I'm testing it now and I just have some doubts. > dlls/iphlpapi/iphlpapi.spec | 2 +- > dlls/iphlpapi/iphlpapi_main.c | 69 +++++++++++++++++++++++++++++++++++++++++++ > include/netioapi.h | 17 +++++++++++ > include/ws2def.h | 13 ++++++++ > include/ws2ipdef.h | 7 +++++ > 5 files changed, 107 insertions(+), 1 deletion(-) > > diff --git a/dlls/iphlpapi/iphlpapi.spec b/dlls/iphlpapi/iphlpapi.spec > index 500267c..1a1f8f7 100644 > --- a/dlls/iphlpapi/iphlpapi.spec > +++ b/dlls/iphlpapi/iphlpapi.spec > @@ -147,7 +147,7 @@ > @ stub GetUdpStatsFromStack > @ stdcall GetUdpTable( ptr ptr long ) > @ stub GetUdpTableFromStack > -#@ stub GetUnicastIpAddressEntry > +@ stdcall GetUnicastIpAddressEntry( ptr ) > #@ stub GetUnicastIpAddressTable > @ stdcall GetUniDirectionalAdapterInfo( ptr ptr ) > @ stdcall Icmp6CreateFile() > diff --git a/dlls/iphlpapi/iphlpapi_main.c b/dlls/iphlpapi/iphlpapi_main.c > index 1daf54d..cfa5d9d 100644 > --- a/dlls/iphlpapi/iphlpapi_main.c > +++ b/dlls/iphlpapi/iphlpapi_main.c > @@ -2467,6 +2467,75 @@ DWORD WINAPI GetExtendedUdpTable(PVOID pUdpTable, PDWORD pdwSize, BOOL bOrder, > return ret; > } > > +DWORD WINAPI GetUnicastIpAddressEntry(MIB_UNICASTIPADDRESS_ROW *row) > +{ > + char buf[MAX_ADAPTER_NAME], *name; > + DWORD ret = ERROR_FILE_NOT_FOUND; > + IP_ADAPTER_ADDRESSES *aa, *ptr; > + ULONG size = 0; > + > + TRACE("%p\n", row); > + > + if (!row) > + return ERROR_INVALID_PARAMETER; > + > + if ((!(name = getInterfaceNameByIndex( row->InterfaceIndex, buf )) && > + !(name = getInterfaceNameByIndex( row->InterfaceLuid.Info.NetLuidIndex, buf )))) > + return ERROR_FILE_NOT_FOUND; The second test for row->InterfaceLuid.Info.NetLuidIndex is really necessary? Because if I remove it the tests don't fail. Actually this is the main point, if I remove the whole if and add a ret = ERROR_FILE_NOT_FOUND before the for (like it was in v1) the tests also pass. > + ret = GetAdaptersAddresses(row->Address.si_family, 0, NULL, NULL, &size); > + if (ret != ERROR_BUFFER_OVERFLOW) > + return ret; > + if (!(ptr = HeapAlloc(GetProcessHeap(), 0, size))) > + return ERROR_OUTOFMEMORY; > + if ((ret = GetAdaptersAddresses(row->Address.si_family, 0, NULL, ptr, &size))) > + { > + HeapFree(GetProcessHeap(), 0, ptr); > + return ret; > + } > + > + for (aa = ptr; aa; aa = aa->Next) > + { > + IP_ADAPTER_UNICAST_ADDRESS *ua; > + > + if (aa->u.s.IfIndex != row->InterfaceIndex) > + continue; > + > + ret = ERROR_NOT_FOUND; > + > + ua = aa->FirstUnicastAddress; > + while (ua) > + { > + SOCKADDR_INET *uaaddr = (SOCKADDR_INET *)ua->Address.lpSockaddr; > + > + if ((row->Address.si_family == WS_AF_INET6 && !memcmp(&row->Address.Ipv6.sin6_addr, &uaaddr->Ipv6.sin6_addr, sizeof(row->Address.Ipv6.sin6_addr))) || > + (row->Address.si_family == WS_AF_INET && !memcmp(&row->Address.Ipv4.sin_addr, &uaaddr->Ipv4.sin_addr, sizeof(row->Address.Ipv4.sin_addr)))) I forgot this in v1 but this line is also too long, sorry. By the way, what about comparing the ipv4 directly (Ipv4.sin_addr.S_un.S_addr ==)?

1 0

Re: [PATCH] Some tests for lsa
by Bruno Jesus 06 Feb '17

06 Feb '17

On Sun, Feb 5, 2017 at 5:57 PM, Julius Schwartzenberg <julius.schwartzenberg(a)gmail.com> wrote: > --- > dlls/secur32/tests/Makefile.in | 1 + > dlls/secur32/tests/lsa.c | 215 +++++++++++++++++++++++++++++++++++++++++ > include/ntsecapi.h | 38 ++++++++ > 3 files changed, 254 insertions(+) > create mode 100644 dlls/secur32/tests/lsa.c Hi, thanks for the patch. Unfortunately as is it cannot be used. Somethings are stated below but mostly you have to check the compiler warnings and guarantee there is none. Also, although very minor, git apply complains about 7 lines adding trailing whitespaces. > diff --git a/dlls/secur32/tests/Makefile.in b/dlls/secur32/tests/Makefile.in > index 089d1c3..d9f140a 100644 > --- a/dlls/secur32/tests/Makefile.in > +++ b/dlls/secur32/tests/Makefile.in > @@ -2,6 +2,7 @@ TESTDLL = secur32.dll > IMPORTS = secur32 crypt32 advapi32 ws2_32 > > C_SRCS = \ > + lsa.c \ > main.c \ > negotiate.c \ > ntlm.c \ > diff --git a/dlls/secur32/tests/lsa.c b/dlls/secur32/tests/lsa.c > new file mode 100644 > index 0000000..6fa1564 > --- /dev/null > +++ b/dlls/secur32/tests/lsa.c > @@ -0,0 +1,215 @@ > +/* > + * Lsa tests The copyright line is missing, example: Copyright 2017 Julius Schwartzenberg > + * 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 <stdio.h> > +#include <windows.h> > +#include <ntsecapi.h> > +#define SECURITY_WIN32 > +#include <ntstatus.h> > +#include <winternl.h> Compiling the patch generates dozens of warnings due to duplicated definitions, please review the headers in use. > +#include "wine/test.h" > + > +static HMODULE secdll; This variable is basically not used anywhere, you don't load the DLL so there is no need to free it. > +char *package_name_kerberos = MICROSOFT_KERBEROS_NAME_A; Is this really necessary? Can't you use the define directly? > +static void test_LsaConnectUntrusted_null_handle(void) > +{ > + NTSTATUS status; > + > + status = LsaConnectUntrusted(NULL); > + > + todo_wine { > + ok(status == STATUS_ACCESS_VIOLATION || status == RPC_NT_NULL_REF_POINTER, > + "LsaConnectUntrusted(NULL) returned 0x%x\n", status); > + } > +} > + > +static void test_LsaConnectUntrsuted_correct_handle(void) > +{ > + NTSTATUS status; > + HANDLE lsa_handle; > + > + status = LsaConnectUntrusted(&lsa_handle); > + > + ok(status == STATUS_SUCCESS, "LsaConnectUntrusted(valid_handle) returned 0x%x\n", status); > +} You are leaking the handle, according to MSDN you need to call LsaDeregisterLogonProcess to close it. This applies to every function that calls this function. > +static void test_LsaLookupAuthenticationPackage_null_handle(void) > +{ > + NTSTATUS status; > + LSA_STRING package_name; > + ULONG auth_package; > + package_name.Buffer = package_name_kerberos; > + package_name.Length = (USHORT)strlen(package_name.Buffer); If you use the #define you can use sizeof(MICROSOFT_KERBEROS_NAME_A) - 1 instead of strlen since the buffer is constant. > + status = LsaLookupAuthenticationPackage(NULL, &package_name, &auth_package); > + > + todo_wine { > + ok(status == STATUS_INVALID_HANDLE || status == STATUS_INVALID_CONNECTION, > + "LsaLookupAuthenticationPackage(NULL, \"Kerberos\", ...) returned 0x%x\n", status); > + } No need for a block in the todo_wine since it is only one ok call. > +} > + > +static void test_LsaLookupAuthenticationPackage_correct_handle(void) > +{ > + NTSTATUS status; > + HANDLE lsa_handle; > + LSA_STRING package_name; > + ULONG auth_package; > + package_name.Buffer = package_name_kerberos; > + package_name.Length = (USHORT)strlen(package_name.Buffer); > + > + LsaConnectUntrusted(&lsa_handle); > + status = LsaLookupAuthenticationPackage(lsa_handle, &package_name, &auth_package); > + > + ok(status == STATUS_SUCCESS, > + "LsaLookupAuthenticationPackage(valid_handle, \"Kerberos\", ...) returned 0x%x\n", status); Is it possible to test that auth_package somehow? > +} > + > +static void test_LsaLookupAuthenticationPackage_correct_handle_wrong_name(void) > +{ > + NTSTATUS status; > + HANDLE lsa_handle; > + LSA_STRING package_name; > + ULONG auth_package; > + package_name.Buffer = "this is probably not a valid authentication package name"; > + package_name.Length = (USHORT)strlen(package_name.Buffer); > + > + LsaConnectUntrusted(&lsa_handle); > + status = LsaLookupAuthenticationPackage(lsa_handle, &package_name, &auth_package); > + > + todo_wine { > + ok(status == STATUS_NO_SUCH_PACKAGE, > + "LsaLookupAuthenticationPackage(valid_handle, invalid, ...) returned 0x%x\n", status); > + } > +} > + > +static void test_LsaCallAuthenticationPackage_invalid_handle(void) > +{ > + NTSTATUS status; > + ULONG submit_buffer_length = 0; > + > + VOID *protocol_return_buffer; > + ULONG *return_buffer_length; > + PNTSTATUS protocol_status; > + > + status = LsaCallAuthenticationPackage( > + NULL, 0xdeadbeef, NULL, submit_buffer_length, > + protocol_return_buffer, return_buffer_length, protocol_status > + ); > + > + todo_wine { > + ok(status == STATUS_INVALID_HANDLE || status == STATUS_INVALID_CONNECTION, > + "LsaCallAuthenticationPackage(NULL, ...) returned 0x%x\n", status); > + } > +} > + > +static void test_LsaCallAuthenticationPackage_invalid_package(void) > +{ > + NTSTATUS status; > + HANDLE lsa_handle; > + ULONG submit_buffer_length = 0; > + > + VOID *protocol_return_buffer; > + ULONG *return_buffer_length; > + PNTSTATUS protocol_status; > + > + LsaConnectUntrusted(&lsa_handle); > + status = LsaCallAuthenticationPackage( > + lsa_handle, 0xdeadbeef, NULL, submit_buffer_length, > + protocol_return_buffer, return_buffer_length, protocol_status > + ); Bad parameters passed to the function, see compiler warnings. > + > + todo_wine { > + ok(status == STATUS_NO_SUCH_PACKAGE, > + "LsaCallAuthenticationPackage(valid_handle, invalid, ...) returned 0x%x\n", status); > + } > +} > + > +static void test_LsaCallAuthenticationPackage_null_submit_buffer(void) > +{ > + NTSTATUS status; > + HANDLE lsa_handle; > + LSA_STRING package_name; > + ULONG auth_package; > + package_name.Buffer = package_name_kerberos; > + package_name.Length = (USHORT)strlen(package_name.Buffer); > + ULONG submit_buffer_length = 0; Very specific compiler warning that you can't declare variables after code. > + > + VOID *protocol_return_buffer; > + ULONG *return_buffer_length; > + PNTSTATUS protocol_status; > + > + LsaConnectUntrusted(&lsa_handle); > + LsaLookupAuthenticationPackage(lsa_handle, &package_name, &auth_package); > + status = LsaCallAuthenticationPackage( > + lsa_handle, auth_package, NULL, submit_buffer_length, > + protocol_return_buffer, return_buffer_length, protocol_status > + ); Bad parameters passed to the function, see compiler warnings. > + todo_wine { > + ok(status == STATUS_INVALID_PARAMETER, > + "LsaCallAuthenticationPackage(...) returned 0x%x\n", status); > + } > +} > + > +static void test_LsaCallAuthenticationPackage_correct_submit_buffer(void) > +{ > + NTSTATUS status; > + HANDLE lsa_handle; > + LSA_STRING package_name; > + ULONG auth_package; > + package_name.Buffer = package_name_kerberos; > + package_name.Length = (USHORT)strlen(package_name.Buffer); > + PKERB_RETRIEVE_TKT_REQUEST protocol_submit_buffer; > + > + VOID *protocol_return_buffer; > + ULONG *return_buffer_length; > + PNTSTATUS protocol_status; > + > + LsaConnectUntrusted(&lsa_handle); > + LsaLookupAuthenticationPackage(lsa_handle, &package_name, &auth_package); > + > + protocol_submit_buffer = LocalAlloc(LMEM_ZEROINIT, sizeof(KERB_RETRIEVE_TKT_REQUEST)); > + protocol_submit_buffer->MessageType = KerbRetrieveEncodedTicketMessage; > + > + status = LsaCallAuthenticationPackage( > + lsa_handle, auth_package, protocol_submit_buffer, sizeof(KERB_RETRIEVE_TKT_REQUEST), > + protocol_return_buffer, return_buffer_length, protocol_status > + ); > + > + ok(status == STATUS_SUCCESS, "LsaCallAuthenticationPackage(...) returned 0x%x\n", status); > +} > + > +START_TEST(lsa) > +{ > + test_LsaConnectUntrusted_null_handle(); > + test_LsaConnectUntrsuted_correct_handle(); Untrusted misspelling. > + test_LsaLookupAuthenticationPackage_null_handle(); > + test_LsaLookupAuthenticationPackage_correct_handle(); > + test_LsaLookupAuthenticationPackage_correct_handle_wrong_name(); > + test_LsaCallAuthenticationPackage_invalid_handle(); > + test_LsaCallAuthenticationPackage_invalid_package(); > + test_LsaCallAuthenticationPackage_null_submit_buffer(); > + test_LsaCallAuthenticationPackage_correct_submit_buffer(); > + > + if(secdll) > + FreeLibrary(secdll); As I said earlier this is not required. > +} > diff --git a/include/ntsecapi.h b/include/ntsecapi.h > index 2bb3d31..439f67e 100644 > --- a/include/ntsecapi.h > +++ b/include/ntsecapi.h > @@ -387,6 +387,44 @@ NTSTATUS WINAPI LsaSetTrustedDomainInformation(LSA_HANDLE,PSID,TRUSTED_INFORMATI > NTSTATUS WINAPI LsaStorePrivateData(LSA_HANDLE,PLSA_UNICODE_STRING,PLSA_UNICODE_STRING); > NTSTATUS WINAPI LsaUnregisterPolicyChangeNotification(POLICY_NOTIFICATION_INFORMATION_CLASS,HANDLE); > > +#ifndef MICROSOFT_KERBEROS_NAME_A > + > +#define MICROSOFT_KERBEROS_NAME_A "Kerberos" > +#define MICROSOFT_KERBEROS_NAME_W L"Kerberos" You can't use L to define WCHAR data. > +#ifdef WIN32_CHICAGO Where is this from? Not a single Wine header has that name. > +#define MICROSOFT_KERBEROS_NAME MICROSOFT_KERBEROS_NAME_A > +#else > +#define MICROSOFT_KERBEROS_NAME MICROSOFT_KERBEROS_NAME_W > +#endif > +#endif > + > +typedef enum _KERB_PROTOCOL_MESSAGE_TYPE { > + KerbDebugRequestMessage = 0,KerbQueryTicketCacheMessage,KerbChangeMachinePasswordMessage,KerbVerifyPacMessage,KerbRetrieveTicketMessage, > + KerbUpdateAddressesMessage,KerbPurgeTicketCacheMessage,KerbChangePasswordMessage,KerbRetrieveEncodedTicketMessage,KerbDecryptDataMessage, > + KerbAddBindingCacheEntryMessage,KerbSetPasswordMessage,KerbSetPasswordExMessage,KerbVerifyCredentialsMessage,KerbQueryTicketCacheExMessage, > + KerbPurgeTicketCacheExMessage,KerbRefreshSmartcardCredentialsMessage,KerbAddExtraCredentialsMessage,KerbQuerySupplementalCredentialsMessage, > + KerbTransferCredentialsMessage,KerbQueryTicketCacheEx2Message These lines are too long, try to keep around 100 columns. > +} KERB_PROTOCOL_MESSAGE_TYPE,*PKERB_PROTOCOL_MESSAGE_TYPE; > + > +#ifndef __SECHANDLE_DEFINED__ > + typedef struct _SecHandle { > + ULONG_PTR dwLower; > + ULONG_PTR dwUpper; > + } SecHandle,*PSecHandle; > + > +#define __SECHANDLE_DEFINED__ > +#endif > + > +typedef struct _KERB_RETRIEVE_TKT_REQUEST { > + KERB_PROTOCOL_MESSAGE_TYPE MessageType; > + LUID LogonId; > + UNICODE_STRING TargetName; > + ULONG TicketFlags; > + ULONG CacheOptions; > + LONG EncryptionType; > + SecHandle CredentialsHandle; > +} KERB_RETRIEVE_TKT_REQUEST,*PKERB_RETRIEVE_TKT_REQUEST; > + > #ifdef __cplusplus > } /* extern "C" */ > #endif /* defined(__cplusplus) */ > -- > 1.9.1 > > >

2 1

Re: [v3 2/3] ntdll: Implement FileIdInformation class support in NtQueryInformationFile.
by Marvin 06 Feb '17

06 Feb '17

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=28366 Your paranoid android. === wxppro (32 bit file) === file.c:1466: Test failed: can't get fid1, res c0000003 file.c:1478: Test failed: can't get fid2, res c0000003 file.c:1491: Test failed: can't get new fid2, res c0000003 file.c:1494: Test failed: file id is not unique === w2003std (32 bit file) === file.c:1466: Test failed: can't get fid1, res c0000003 file.c:1478: Test failed: can't get fid2, res c0000003 file.c:1491: Test failed: can't get new fid2, res c0000003 file.c:1494: Test failed: file id is not unique === wvistau64 (32 bit file) === file.c:1466: Test failed: can't get fid1, res c0000002 file.c:1478: Test failed: can't get fid2, res c0000002 file.c:1491: Test failed: can't get new fid2, res c0000002 file.c:1494: Test failed: file id is not unique === w2008s64 (32 bit file) === file.c:1466: Test failed: can't get fid1, res c0000002 file.c:1478: Test failed: can't get fid2, res c0000002 file.c:1491: Test failed: can't get new fid2, res c0000002 file.c:1494: Test failed: file id is not unique === w7u (32 bit file) === file.c:1466: Test failed: can't get fid1, res c0000003 file.c:1478: Test failed: can't get fid2, res c0000003 file.c:1491: Test failed: can't get new fid2, res c0000003 file.c:1494: Test failed: file id is not unique === w7pro64 (32 bit file) === file.c:1466: Test failed: can't get fid1, res c0000002 file.c:1478: Test failed: can't get fid2, res c0000002 file.c:1491: Test failed: can't get new fid2, res c0000002 file.c:1494: Test failed: file id is not unique === wvistau64 (64 bit file) === file.c:1466: Test failed: can't get fid1, res c0000003 file.c:1478: Test failed: can't get fid2, res c0000003 file.c:1491: Test failed: can't get new fid2, res c0000003 file.c:1494: Test failed: file id is not unique === w2008s64 (64 bit file) === file.c:1466: Test failed: can't get fid1, res c0000003 file.c:1478: Test failed: can't get fid2, res c0000003 file.c:1491: Test failed: can't get new fid2, res c0000003 file.c:1494: Test failed: file id is not unique === w7pro64 (64 bit file) === file.c:1466: Test failed: can't get fid1, res c0000003 file.c:1478: Test failed: can't get fid2, res c0000003 file.c:1491: Test failed: can't get new fid2, res c0000003 file.c:1494: Test failed: file id is not unique

1 0

Re: [v3 1/3] ntdll: Define more file information classes.
by Marvin 06 Feb '17

06 Feb '17

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=28365 Your paranoid android. === wxppro (32 bit directory) === directory.c:369: Test failed: 60: wrong status c0000003 directory.c:369: Test failed: 63: wrong status c0000003 === w2003std (32 bit directory) === directory.c:369: Test failed: 60: wrong status c0000003 directory.c:369: Test failed: 63: wrong status c0000003 === wvistau64 (32 bit directory) === directory.c:369: Test failed: 60: wrong status c0000002 directory.c:369: Test failed: 63: wrong status c0000002 === w2008s64 (32 bit directory) === directory.c:369: Test failed: 60: wrong status c0000002 directory.c:369: Test failed: 63: wrong status c0000002 === w7u (32 bit directory) === directory.c:369: Test failed: 60: wrong status c0000003 directory.c:369: Test failed: 63: wrong status c0000003 === w7pro64 (32 bit directory) === directory.c:369: Test failed: 60: wrong status c0000002 directory.c:369: Test failed: 63: wrong status c0000002 === w864 (32 bit directory) === directory.c:369: Test failed: 63: wrong status c0000002 === wvistau64 (64 bit directory) === directory.c:369: Test failed: 60: wrong status c0000003 directory.c:369: Test failed: 63: wrong status c0000003 === w2008s64 (64 bit directory) === directory.c:369: Test failed: 60: wrong status c0000003 directory.c:369: Test failed: 63: wrong status c0000003 === w7pro64 (64 bit directory) === directory.c:369: Test failed: 60: wrong status c0000003 directory.c:369: Test failed: 63: wrong status c0000003

1 0

Re: [PATCH 1/2] msvcrt/tests: Add tests for _creat.
by Lauri Kenttä 06 Feb '17

06 Feb '17

On 2017-02-06 19:54, Piotr Caban wrote: > On 02/06/17 18:27, Lauri Kenttä wrote: >> + int fd, pos, count, readonly, old_fmode = _fmode; > This will not compile in 64-bit wine. Please use e.g. > __p__fmode instead of _fmode. > > Thanks, > Piotr Thanks. __p__fmode doesn't seem to work on 64-bit either. What a mess. Resubmitted with _get_fmode/_set_fmode. I'm not sure if those should be used without GetProcAddress, but at least MinGW has unresolved symbols. Whatever, at least it's working now on both 32-bit and 64-bit. :) -- Lauri Kenttä

1 0

Re: [PATCH 2/2] msvcrt: Fix _creat and _wcreat.
by Marvin 06 Feb '17

06 Feb '17

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=28359 Your paranoid android. === wvistau64 (64 bit file) === file.c:2416: Test failed: expected pos 4 (binary mode), got 6 === w2008s64 (64 bit file) === file.c:2416: Test failed: expected pos 4 (binary mode), got 6 === w7pro64 (64 bit file) === file.c:2416: Test failed: expected pos 4 (binary mode), got 6 === w864 (64 bit file) === file.c:2416: Test failed: expected pos 4 (binary mode), got 6 === w1064 (64 bit file) === file.c:2416: Test failed: expected pos 4 (binary mode), got 6

1 0

Re: [PATCH] msvcrt: Pass a missing flag to MSVCRT__open.
by Piotr Caban 06 Feb '17

06 Feb '17

On 02/04/17 14:15, Lauri Kenttä wrote: > int CDECL MSVCRT__creat(const char *path, int flags) > { > int usedFlags = (flags & MSVCRT__O_TEXT)| MSVCRT__O_CREAT| MSVCRT__O_WRONLY| MSVCRT__O_TRUNC; > - return MSVCRT__open(path, usedFlags); > + return MSVCRT__open(path, usedFlags, MSVCRT__S_IWRITE); > } According to MSDN the function takes pmode as second argument. Could you please add some tests for it? It looks like wine's implementation is handling second argument incorrectly. Thanks, Piotr

2 1

Re: [PATCH 1/2] msvcrt/tests: Add tests for _creat.
by Marvin 06 Feb '17

06 Feb '17

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=28358 Your paranoid android. === wvistau64 (64 bit file) === file.c:2417: Test failed: expected pos 4 (binary mode), got 6 === w2008s64 (64 bit file) === file.c:2417: Test failed: expected pos 4 (binary mode), got 6 === w7pro64 (64 bit file) === file.c:2417: Test failed: expected pos 4 (binary mode), got 6 === w864 (64 bit file) === file.c:2417: Test failed: expected pos 4 (binary mode), got 6 === w1064 (64 bit file) === file.c:2417: Test failed: expected pos 4 (binary mode), got 6

1 0