Re: Regression in patch wined3d: Implement D3DSIO_MOVA in ARB backend.
Mirek wrote:
Ok, i just found why is it not working with this patch, and why is it
working without this patch. It is based on position of definition MOVA
in vertexshader.c, if MOVA is on after patch position it is not
working, but if i move line
{WINED3DSIO_MOVA, mova, NULL, 1, 2, vshader_mova, vshader_hw_map2gl,
shader_glsl_mov, WINED3DVS_VERSION(2,0), -1}, to position of definition
before patch, evrything is working.
...
that's because Jason seems to have introduced order requirements on the
table that weren't previously there. The mnxn implementation will
attempt to access the table as an array, which will break once MOVA is
moved before DP3:
glsl_shader.c:tmpArg.opcode =
IWineD3DVertexShaderImpl_shader_ins[WINED3DSIO_DP4];
glsl_shader.c:tmpArg.opcode =
IWineD3DVertexShaderImpl_shader_ins[WINED3DSIO_DP3];
The instruction table currently does not support this, and it only
happens to work for the first few instructions [ which the MOVA change
breaks ].
Re: Regression in patch wined3d: Implement D3DSIO_MOVAin ARB backend.
Ivan Gyurdiev ivg231 at gmail.com writes: that's because Jason seems to have introduced order requirements on the table that weren't previously there. The mnxn implementation will attempt to access the table as an array, which will break once MOVA is moved before DP3: glsl_shader.c:tmpArg.opcode = IWineD3DVertexShaderImpl_shader_ins[WINED3DSIO_DP4]; glsl_shader.c:tmpArg.opcode = IWineD3DVertexShaderImpl_shader_ins[WINED3DSIO_DP3]; Yup, seems like the temptation to dereference the array directly was too great. These should be replaced to use baseshader.c:get_shader_opcode(...), I think, instead of the array access. Exact same thing also in arb_program_shader.c too. To prevent such things in the future, is there any way to remove the enum WINED3DSHADER_INSTRUCTION_OPCODE_TYPE and place it somewhere less visible that the included-everywhere wined3d_private_types.h? A quick glance at the code seems to suggest it is not feasible right now, but maybe someone has an idea? Or better yet, since device.c is the only legitimate user of IWineD3DVertexShaderImpl_shader_ins[], don't extern that array in wined3d_private.h, but only in device.c (i.e. move the extern declaration from wined3d_private.h to device.c). Then glsl_shader.c and arb_program_shader.c and future source would properly fail to compile on this type of mistake. In the worst case a comment to the effect of /* don't dereference this array directly */ certainly seems in order... - Aric
Re: [PATCH 5/5] quartzdrv: Implement MsgWaitForMultipleObjectsEx and initial Carbon event handling.
On 26 oct. 06, at 12:47, Alexandre Julliard wrote: I don't think polling is an acceptable way to wait for events. You really need to get a file descriptor or some other handle that you can wait on. I am sure it isn't. But I am not sure of the proper solution. I have two ideas: 1) Change wineserver to relay on Carbon events to send (some of its) message, instead of interrupting the wine thread while waiting for a file descriptor. The wine thread could be waiting via the regular carbon run loop for an event, and the wineserver could send a carbon event as soon as the fd triggers select() end. 2) Launch a 'carbon' thread which will run the carbon run loop (which does not poll), and writes to a fd as soon as it receive events. I think 2) may be preferred but to have a separate thread that watch event and an other that act on the UI tends to be tricky. Pierre.
Checking, if a program is running under WINE
Hi Is there a short way for my windows application to check, if it is running under WINE or under Windows? This should also run in future; no hack. Thanks Gero -- Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen! Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer
Re: Checking, if a program is running under WINE
On Mon, Oct 30, 2006 at 03:33:08PM +0100, Gero Zink wrote: Hi Is there a short way for my windows application to check, if it is running under WINE or under Windows? This should also run in future; no hack. Thanks In general we try to fix any kind of difference between Wine and Windows, so such a check should not be necessary. Can you tell us why you need it? Ciao, Marcus
uploading files to the wikie
Hey, Do we have the ability to upload files to our Wiki? -- James Hawkins
Re: REGRESSION: 0.9.24 crashes on regsvr32 msvbvm60.dll, temporary workaround attached
On 29.10.2006 02:54, [EMAIL PROTECTED] wrote: Apparently, 0x2000 as a flag in FKCCIC indicates that pFuncRec-OptAttr[2] is a pointer to some string. If what little understanding I have of typelib loading is correct, these typelibs are read from DLL resources on disk. Therefore, I fail to grasp how they can possibly refer to valid memory locations. Hmm, perhaps check if interpreting the value as an offset from the start of the typelib data is sensible? -f.r.
Re: uploading files to the wikie
On Mon, October 30, 2006 2:25 pm, James Hawkins wrote: Do we have the ability to upload files to our Wiki? Yes, you do. Avoid uploading GBs of data though :) -- Dimi Paun [EMAIL PROTECTED] Lattica, Inc.
Re: [secur32] Remove test that crashes on Win2k
On 10/30/06, Paul Vriens [EMAIL PROTECTED] wrote: Hi, makes sure the other tests are run. Changelog Remove test that crashes on Win2k Cheers, Paul. --- dlls/secur32/tests/schannel.c |2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/dlls/secur32/tests/schannel.c b/dlls/secur32/tests/schannel.c index 23e2a0a..95669fe 100644 --- a/dlls/secur32/tests/schannel.c +++ b/dlls/secur32/tests/schannel.c @@ -140,10 +140,12 @@ static void testAcquireSecurityContext(v NULL); ok(st == SEC_E_SECPKG_NOT_FOUND, Expected SEC_E_SECPKG_NOT_FOUND, got %08x\n, st); +/* Crashes on Win2K st = pAcquireCredentialsHandleA(NULL, unisp_name_a, 0, NULL, NULL, NULL, NULL, NULL, NULL); ok(st == SEC_E_NO_CREDENTIALS, Expected SEC_E_NO_CREDENTIALS, got %08x\n, st); + */ Please use #if 0/#endif to comment out pieces of code. -- James Hawkins
Re: [secur32] Remove test that crashes on Win2k
On Mon, 2006-10-30 at 23:44 -0800, James Hawkins wrote: On 10/30/06, Paul Vriens [EMAIL PROTECTED] wrote: Hi, makes sure the other tests are run. Changelog Remove test that crashes on Win2k Cheers, Paul. --- dlls/secur32/tests/schannel.c |2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/dlls/secur32/tests/schannel.c b/dlls/secur32/tests/schannel.c index 23e2a0a..95669fe 100644 --- a/dlls/secur32/tests/schannel.c +++ b/dlls/secur32/tests/schannel.c @@ -140,10 +140,12 @@ static void testAcquireSecurityContext(v NULL); ok(st == SEC_E_SECPKG_NOT_FOUND, Expected SEC_E_SECPKG_NOT_FOUND, got %08x\n, st); +/* Crashes on Win2K st = pAcquireCredentialsHandleA(NULL, unisp_name_a, 0, NULL, NULL, NULL, NULL, NULL, NULL); ok(st == SEC_E_NO_CREDENTIALS, Expected SEC_E_NO_CREDENTIALS, got %08x\n, st); + */ Please use #if 0/#endif to comment out pieces of code. Hi James, there is a commented out test a few lines below. I thought I'd stick to how it's done in this file. Cheers, Paul.
Re: [secur32] Remove test that crashes on Win2k
James Hawkins wrote:
+/* Crashes on Win2K
st = pAcquireCredentialsHandleA(NULL, unisp_name_a, 0, NULL,
NULL, NULL,
NULL, NULL, NULL);
ok(st == SEC_E_NO_CREDENTIALS, Expected SEC_E_NO_CREDENTIALS,
got %08x\n,
st);
+ */
Please use #if 0/#endif to comment out pieces of code.
Or even better use if (0) { } so compile problems don't creep in...
Mike
Re: [secur32] Remove test that crashes on Win2k
On 10/30/06, Paul Vriens [EMAIL PROTECTED] wrote: On Mon, 2006-10-30 at 23:44 -0800, James Hawkins wrote: On 10/30/06, Paul Vriens [EMAIL PROTECTED] wrote: Hi, makes sure the other tests are run. Changelog Remove test that crashes on Win2k Cheers, Paul. --- dlls/secur32/tests/schannel.c |2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/dlls/secur32/tests/schannel.c b/dlls/secur32/tests/schannel.c index 23e2a0a..95669fe 100644 --- a/dlls/secur32/tests/schannel.c +++ b/dlls/secur32/tests/schannel.c @@ -140,10 +140,12 @@ static void testAcquireSecurityContext(v NULL); ok(st == SEC_E_SECPKG_NOT_FOUND, Expected SEC_E_SECPKG_NOT_FOUND, got %08x\n, st); +/* Crashes on Win2K st = pAcquireCredentialsHandleA(NULL, unisp_name_a, 0, NULL, NULL, NULL, NULL, NULL, NULL); ok(st == SEC_E_NO_CREDENTIALS, Expected SEC_E_NO_CREDENTIALS, got %08x\n, st); + */ Please use #if 0/#endif to comment out pieces of code. Hi James, there is a commented out test a few lines below. I thought I'd stick to how it's done in this file. In this case, it's not really a 'style' issue, where we keep the style the same in the file. We shouldn't be using actual comments to disable code. If you want to change the other case to #if 0, I wouldn't have a problem with that, but we shouldn't be adding more comments -- James Hawkins
