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

Re: [resend PATCH 2/4] oledb32: IErrorRecords uses 0-based index to access record info
by Alistair Leslie-Hughes 15 Nov '16

15 Nov '16

Hi Nikolay, On 15/11/16 03:56, Nikolay Sivov wrote: > Signed-off-by: Nikolay Sivov <nsivov(a)codeweavers.com> > > + hr = IErrorRecords_GetRecordCount(errrecs, &cnt); > + ok(hr == S_OK, "got %08x\n", hr); > + ok(cnt == 0, "expected 2 got %d\n", cnt); The comment here is incorrect. Best Regards Alistair.

2 1

Re: [PATCH] cabinet: Properly initialize our internal fci structure v2(bug 36115)(valgrind)
by Huw Davies 14 Nov '16

14 Nov '16

On 14 Nov 2016, at 16:46, Detlef Riekenberg <wine.dev(a)web.de> wrote: > > We did not initialize all items for our internal fci structure. > This should fix bug 36115 > > v2: > Remove comments, as suggested by Huw > > -- > bye bye ... Detlef > > Signed-off-by: Detlef Riekenberg <wine.dev(a)web.de> > --- > dlls/cabinet/fci.c | 19 +------------------ > 1 file changed, 1 insertion(+), 18 deletions(-) > > diff --git a/dlls/cabinet/fci.c b/dlls/cabinet/fci.c > index 70d4f3f..03fbf46 100644 > --- a/dlls/cabinet/fci.c > +++ b/dlls/cabinet/fci.c > @@ -1036,6 +1036,7 @@ HFCI __cdecl FCICreate( > return NULL; > } > > + memset(p_fci_internal, 0, sizeof(FCI_Int)); I also suggested changing the sizeof(FCI_Int) -> sizeof(*p_fci_internal), but maybe that wasn’t very clear. Huw.

1 0

WineConf decisions
by Alexandre Julliard 14 Nov '16

14 Nov '16

At this year's WineConf we again had good constructive discussions about the development process, and decided a number of things: - Version numbering will change to a sligthly shorter scheme. The next stable release will be called 2.0. Stable release updates will follow the existing scheme: 2.0.1, 2.0.2, etc. For the developemnt branch we will drop the intermediate number: releases will be numbered 2.1, 2.2, 2.3, etc. and the next stable release in late 2017 will be called 3.0. - Code freeze for the 2.0 release will start after 1.9.24, in 3-4 weeks, with a final release targeted around the end of the year. - Michael Stefaniuc has agreed to continue maintaining the stable branch, and will be in charge of the 2.0.x releases. - Since it sometimes takes months for the stable release to get into distributions, we'll look into providing binary packages of the stable branch to make it easier for users to stay up to date. - The default reported Windows version will be upgraded to Windows 7. This will be done on the development branch soon after 2.0 is released, so that we have the time to catch regressions and address them before the next stable release. - To further the integration of WineHQ and Wine-Staging, we'll look into merging the Staging website into the Wine Wiki. We'll also merge the IRC channels, and encourage people to discuss Staging issues on #winehq and #winehackers, with #wine-staging being retired at some point. - To make sure that no patch falls off the list unreviewed, we are going to implement a mechanism to automatically assign a reviewer from the MAINTAINERS pool when a patch has been in New state for a week without action. - Since reviewers assigned at random may not always be comfortable with fully signing off on a patch, we discussed adding an Acked-by header as a weaker version of Signed-off-by. We finally decided that for now, an informal mail to wine-devel would be good enough. We may revisit that decision if a formal header makes automation easier. I'm hoping that these improvements will make things better for everybody. As always, feedback is encouraged, especially from people who couldn't make it to WineConf. As we implement these changes, please let us know how we are doing, and where we can do better. -- Alexandre Julliard julliard(a)winehq.org

1 0

AppDB Maintainership and Updates
by Jeremy Newman 14 Nov '16

14 Nov '16

Hey all! Just wanted to let you all know that at WineConf we decided that I should once again take up the mantle of AppDB codebase maintainer. Woo! I'll be in charge of reviewing and approving AppDB patches from now on. If you want to submit patches, please send them to the wine-patches mailing list with the subject prefix of [AppDB]. Bugs and feature requests should be filed on Bugzilla as usual. Discussion can happen here on wine-devel, IRC #winehackers, or preferably on the Forums in the new "Website Issues" forum. I've started the process of cleaning up some of the code and moving things around. Please bear with me as I'll be doing this in whatever spare time I can find. I created a ToDo page in the Wiki that will track the various things that I've agreed to work on. If it is not listed there, then it is open for someone else to jump on. https://wiki.winehq.org/AppDB_ToDo_List -N

1 0

Re: [17/23] wined3d: Implement return instruction.
by Matteo Bruni 14 Nov '16

14 Nov '16

2016-11-13 12:35 GMT-06:00 Andrew Wesie <awesie(a)gmail.com>: > @@ -6079,8 +6101,26 @@ static GLuint shader_glsl_generate_vshader(const struct wined3d_context *context > /* Base Shader Body */ > shader_generate_main(shader, buffer, reg_maps, function, &priv_ctx); > > + shader_addline(buffer, "}\n"); > + > + TRACE("Compiling shader object %u.\n", shader_id); > + shader_glsl_compile(gl_info, shader_id, buffer->buffer); > + > + return shader_id; > +} This breaks < SM4 shaders, those don't have an explicit return instruction at the end. In general, each patch should pass the wine tests. Please run the d3d tests and make sure that none are broken by your patchset. Specifically, this patch breaks the d3d9 visual test pretty badly. It looks like there are more failures introduced by other patches in the series but I haven't tracked them down. > +static void shader_glsl_prologue_vs(const struct wined3d_shader_context *ctx) I think this should actually be "epilogue".

1 0

Re: [11/23] wined3d: Implement primitive restart index.
by Matteo Bruni 14 Nov '16

14 Nov '16

2016-11-13 12:35 GMT-06:00 Andrew Wesie <awesie(a)gmail.com>: > Signed-off-by: Andrew Wesie <awesie(a)gmail.com> > --- > dlls/wined3d/directx.c | 3 +++ > dlls/wined3d/glsl_shader.c | 14 ++++++++++++++ > dlls/wined3d/wined3d_gl.h | 2 ++ > 3 files changed, 19 insertions(+) > > diff --git a/dlls/wined3d/directx.c b/dlls/wined3d/directx.c > index a8c8c9f..c930cf6 100644 > --- a/dlls/wined3d/directx.c > +++ b/dlls/wined3d/directx.c > @@ -121,6 +121,7 @@ static const struct wined3d_extension_map gl_extension_map[] = > {"GL_ARB_draw_elements_base_vertex", ARB_DRAW_ELEMENTS_BASE_VERTEX }, > {"GL_ARB_draw_instanced", ARB_DRAW_INSTANCED }, > {"GL_ARB_ES2_compatibility", ARB_ES2_COMPATIBILITY }, > + {"GL_ARB_ES3_compatibility", ARB_ES3_COMPATIBILITY }, > {"GL_ARB_explicit_attrib_location", ARB_EXPLICIT_ATTRIB_LOCATION }, > {"GL_ARB_fragment_coord_conventions", ARB_FRAGMENT_COORD_CONVENTIONS}, > {"GL_ARB_fragment_program", ARB_FRAGMENT_PROGRAM }, > @@ -228,6 +229,7 @@ static const struct wined3d_extension_map gl_extension_map[] = > {"GL_NV_half_float", NV_HALF_FLOAT }, > {"GL_NV_light_max_exponent", NV_LIGHT_MAX_EXPONENT }, > {"GL_NV_point_sprite", NV_POINT_SPRITE }, > + {"GL_NV_primitive_restart", NV_PRIMITIVE_RESTART }, > {"GL_NV_register_combiners", NV_REGISTER_COMBINERS }, > {"GL_NV_register_combiners2", NV_REGISTER_COMBINERS2 }, > {"GL_NV_texgen_reflection", NV_TEXGEN_REFLECTION }, > @@ -3125,6 +3127,7 @@ static void load_gl_funcs(struct wined3d_gl_info *gl_info) > USE_GL_FUNC(glMapBuffer) /* OpenGL 1.5 */ > USE_GL_FUNC(glPointParameteri) /* OpenGL 1.4 */ > USE_GL_FUNC(glPointParameteriv) /* OpenGL 1.4 */ > + USE_GL_FUNC(glPrimitiveRestartIndex) > USE_GL_FUNC(glShaderSource) /* OpenGL 2.0 */ > USE_GL_FUNC(glStencilFuncSeparate) /* OpenGL 2.0 */ > USE_GL_FUNC(glStencilOpSeparate) /* OpenGL 2.0 */ AFAICS this function was added in OpenGL 3.1 and there is no specific ARB extension for it. Maybe adding a new WINED3D_GL_ entry to flag the availability of the function is a good idea. Also you probably want to MAP_GL_FUNCTION() the NV variant to the core one and add an entry for GL_ARB_ES3_compatibility to core_extensions. > diff --git a/dlls/wined3d/glsl_shader.c b/dlls/wined3d/glsl_shader.c > index f136536..27a5533 100644 > --- a/dlls/wined3d/glsl_shader.c > +++ b/dlls/wined3d/glsl_shader.c > @@ -8613,6 +8613,20 @@ static void shader_glsl_init_context_state(struct wined3d_context *context) > > gl_info->gl_ops.gl.p_glEnable(GL_PROGRAM_POINT_SIZE); > checkGLcall("GL_PROGRAM_POINT_SIZE"); > + if (gl_info->supported[ARB_ES3_COMPATIBILITY]) > + { > + /* We prefer this method because it correctly handles 16-bit and 32-bit indices. */ > + gl_info->gl_ops.gl.p_glEnable(GL_PRIMITIVE_RESTART_FIXED_INDEX); > + checkGLcall("GL_PRIMITIVE_RESTART_FIXED_INDEX"); > + } > + else if (gl_info->supported[NV_PRIMITIVE_RESTART] && GL_EXTCALL(glPrimitiveRestartIndex)) > + { > + /* FIXME: Does this handle 16-bit indices correctly? */ > + GL_EXTCALL(glPrimitiveRestartIndex(0xFFFFFFFF)); > + checkGLcall("glPrimitiveRestartIndex"); > + gl_info->gl_ops.gl.p_glEnable(GL_PRIMITIVE_RESTART); > + checkGLcall("GL_PRIMITIVE_RESTART"); Probably not, but nothing prevents you to use 0xffff for 16-bit indices and force re-setting the restart index when changing between 16-bit and 32-bit indices.

2 2

Re: [08/23] wined3d: Allow viewport to be out-of-bounds if using offscreen target.
by Matteo Bruni 14 Nov '16

14 Nov '16

2016-11-13 12:35 GMT-06:00 Andrew Wesie <awesie(a)gmail.com>: > Signed-off-by: Andrew Wesie <awesie(a)gmail.com> > --- > dlls/wined3d/state.c | 42 ++++++++++++++++++++++-------------------- > 1 file changed, 22 insertions(+), 20 deletions(-) > > diff --git a/dlls/wined3d/state.c b/dlls/wined3d/state.c > index cd3763c..ed20e325 100644 > --- a/dlls/wined3d/state.c > +++ b/dlls/wined3d/state.c > @@ -4617,26 +4617,6 @@ static void viewport_miscpart(struct wined3d_context *context, const struct wine > struct wined3d_viewport vp = state->viewport; > unsigned int width, height; > > - if (target) > - { > - if (vp.width > target->width) > - vp.width = target->width; > - if (vp.height > target->height) > - vp.height = target->height; > - > - wined3d_rendertarget_view_get_drawable_size(target, context, &width, &height); > - } > - else if (depth_stencil) > - { > - width = depth_stencil->width; > - height = depth_stencil->height; > - } > - else > - { > - FIXME("No attachments draw calls not supported.\n"); > - return; > - } > - > gl_info->gl_ops.gl.p_glDepthRange(vp.min_z, vp.max_z); > checkGLcall("glDepthRange"); > /* Note: GL requires lower left, DirectX supplies upper left. This is > @@ -4644,7 +4624,29 @@ static void viewport_miscpart(struct wined3d_context *context, const struct wine > if (context->render_offscreen) > gl_info->gl_ops.gl.p_glViewport(vp.x, vp.y, vp.width, vp.height); > else > + { > + if (target) > + { > + if (vp.width > target->width) > + vp.width = target->width; > + if (vp.height > target->height) > + vp.height = target->height; > + > + wined3d_rendertarget_view_get_drawable_size(target, context, &width, &height); > + } > + else if (depth_stencil) > + { > + width = depth_stencil->width; > + height = depth_stencil->height; > + } > + else > + { > + FIXME("No attachments draw calls not supported.\n"); > + return; > + } > + > gl_info->gl_ops.gl.p_glViewport(vp.x, (height - (vp.y + vp.height)), vp.width, vp.height); > + } > checkGLcall("glViewport"); > } You need to make the same change to viewport_miscpart_cc() (you can blame me for that). Again, it would be very nice to have a testcase for this. Extending viewport_test() in d3d9/tests/visual.c and adding similar tests to the other d3d versions shouldn't be too hard.

2 2

Re: [PATCH] hidclass: All reports read or written to user space lead with a reportId
by Sebastian Lackner 14 Nov '16

14 Nov '16

On 14.11.2016 14:35, Aric Stewart wrote: > This is a little confusing because reports to and from the device > itself only have a leading reportId if the reportid is not 0. > Otherwise there is no leading reportId. So we need to add a 0 reportId > to reports coming from such a device and strip the leading 0 from > reports written to the device. > > Signed-off-by: Aric Stewart <aric(a)codeweavers.com> > --- > dlls/hidclass.sys/descriptor.c | 31 ++++----------- > dlls/hidclass.sys/device.c | 90 ++++++++++++++++++++++++++++-------------- > 2 files changed, 67 insertions(+), 54 deletions(-) > > > > 0001-hidclass-All-reports-read-or-written-to-user-space-lea.txt > > > diff --git a/dlls/hidclass.sys/descriptor.c b/dlls/hidclass.sys/descriptor.c > index 9a26f0f4..8bcda43 100644 > --- a/dlls/hidclass.sys/descriptor.c > +++ b/dlls/hidclass.sys/descriptor.c > @@ -887,10 +887,8 @@ static WINE_HIDP_PREPARSED_DATA* build_PreparseData( > new_report(wine_report, input_features[0]); > data->dwInputReportCount++; > > - if (input_features[0]->caps.ReportID != 0) > - bitOffset = 8; > - else > - bitOffset = 0; > + /* Room for the reportID */ > + bitOffset = 8; > > for (i = 0; i < i_count; i++) > { > @@ -901,10 +899,7 @@ static WINE_HIDP_PREPARSED_DATA* build_PreparseData( > new_report(wine_report, input_features[i]); > data->dwInputReportCount++; > bitLength = max(bitOffset, bitLength); > - if (input_features[i]->caps.ReportID != 0) > - bitOffset = 8; > - else > - bitOffset = 0; > + bitOffset = 8; > } > build_elements(wine_report, input_features[i], &bitOffset); > count_elements(input_features[i], &data->caps.NumberInputButtonCaps, > @@ -922,10 +917,7 @@ static WINE_HIDP_PREPARSED_DATA* build_PreparseData( > data->dwOutputReportOffset = (BYTE*)wine_report - (BYTE*)data->InputReports; > new_report(wine_report, output_features[0]); > data->dwOutputReportCount++; > - if (output_features[0]->caps.ReportID != 0) > - bitOffset = 8; > - else > - bitOffset = 0; > + bitOffset = 8; > > for (i = 0; i < o_count; i++) > { > @@ -936,10 +928,7 @@ static WINE_HIDP_PREPARSED_DATA* build_PreparseData( > new_report(wine_report, output_features[i]); > data->dwOutputReportCount++; > bitLength = max(bitOffset, bitLength); > - if (output_features[0]->caps.ReportID != 0) > - bitOffset = 8; > - else > - bitOffset = 0; > + bitOffset = 8; > } > build_elements(wine_report, output_features[i], &bitOffset); > count_elements(output_features[i], &data->caps.NumberOutputButtonCaps, > @@ -957,10 +946,7 @@ static WINE_HIDP_PREPARSED_DATA* build_PreparseData( > data->dwFeatureReportOffset = (BYTE*)wine_report - (BYTE*)data->InputReports; > new_report(wine_report, feature_features[0]); > data->dwFeatureReportCount++; > - if (feature_features[0]->caps.ReportID != 0) > - bitOffset = 8; > - else > - bitOffset = 0; > + bitOffset = 8; > > for (i = 0; i < f_count; i++) > { > @@ -971,10 +957,7 @@ static WINE_HIDP_PREPARSED_DATA* build_PreparseData( > new_report(wine_report, feature_features[i]); > data->dwFeatureReportCount++; > bitLength = max(bitOffset, bitLength); > - if (feature_features[0]->caps.ReportID != 0) > - bitOffset = 8; > - else > - bitOffset = 0; > + bitOffset = 8; > } > build_elements(wine_report, feature_features[i], &bitOffset); > count_elements(feature_features[i], &data->caps.NumberFeatureButtonCaps, > diff --git a/dlls/hidclass.sys/device.c b/dlls/hidclass.sys/device.c > index 5dd5ebb..6454b67 100644 > --- a/dlls/hidclass.sys/device.c > +++ b/dlls/hidclass.sys/device.c > @@ -197,6 +197,28 @@ void HID_DeleteDevice(HID_MINIDRIVER_REGISTRATION *driver, DEVICE_OBJECT *device > IoDeleteDevice(device); > } > > +static NTSTATUS copy_packet_into_buffer(HID_XFER_PACKET *packet, BYTE* buffer, ULONG buffer_length, ULONG *out_length) > +{ > + *out_length = 0; > + if (buffer_length > packet->reportBufferLen) > + { > + if (packet->reportId != 0) > + { > + memcpy(buffer, packet->reportBuffer, packet->reportBufferLen); > + *out_length = packet->reportBufferLen; The length check is a bit too strict in this case. You are returning STATUS_BUFFER_OVERFLOW even when the buffer is sufficient (buffer_length == packet->reportBufferLen). > + } > + else > + { > + buffer[0] = 0; > + memcpy(&buffer[1], packet->reportBuffer, packet->reportBufferLen); > + *out_length = packet->reportBufferLen + 1; > + } > + return STATUS_SUCCESS; > + } > + else > + return STATUS_BUFFER_OVERFLOW; > +} > + > static void HID_Device_processQueue(DEVICE_OBJECT *device) > { > LIST_ENTRY *entry; > @@ -217,20 +239,14 @@ static void HID_Device_processQueue(DEVICE_OBJECT *device) > RingBuffer_Read(ext->ring_buffer, ptr, packet, &buffer_size); > if (buffer_size) > { > + NTSTATUS hr; Maybe its just me, but I think "hr" is really a very bad name for NTSTATUS codes. ;) > + ULONG out_length; > IO_STACK_LOCATION *irpsp = IoGetCurrentIrpStackLocation(irp); > packet->reportBuffer = (BYTE *)packet + sizeof(*packet); > TRACE_(hid_report)("Processing Request (%i)\n",ptr); > - if (irpsp->Parameters.Read.Length >= packet->reportBufferLen) > - { > - memcpy(irp->AssociatedIrp.SystemBuffer, packet->reportBuffer, packet->reportBufferLen); > - irp->IoStatus.Information = packet->reportBufferLen; > - irp->IoStatus.u.Status = STATUS_SUCCESS; > - } > - else > - { > - irp->IoStatus.Information = 0; > - irp->IoStatus.u.Status = STATUS_BUFFER_OVERFLOW; > - } > + hr = copy_packet_into_buffer(packet, irp->AssociatedIrp.SystemBuffer, irpsp->Parameters.Read.Length, &out_length); > + irp->IoStatus.u.Status = hr; > + irp->IoStatus.Information = out_length; > } > else > { > @@ -331,7 +347,10 @@ static DWORD CALLBACK hid_device_thread(void *args) > if (!exit_now && irp->IoStatus.u.Status == STATUS_SUCCESS) > { > packet->reportBufferLen = irp->IoStatus.Information; > - packet->reportId = packet->reportBuffer[0]; > + if (ext->preparseData->InputReports[0].reportID) > + packet->reportId = packet->reportBuffer[0]; > + else > + packet->reportId = 0; > RingBuffer_Write(ext->ring_buffer, packet); > HID_Device_processQueue(device); > } > @@ -461,9 +480,17 @@ static NTSTATUS HID_set_to_device(DEVICE_OBJECT *device, IRP *irp) > NTSTATUS rc; > > TRACE_(hid_report)("Device %p Buffer length %i Buffer %p\n", device, irpsp->Parameters.DeviceIoControl.InputBufferLength, irp->AssociatedIrp.SystemBuffer); > - packet.reportBuffer = irp->AssociatedIrp.SystemBuffer; > - packet.reportId = ((char*)irp->AssociatedIrp.SystemBuffer)[0]; > - packet.reportBufferLen = irpsp->Parameters.DeviceIoControl.InputBufferLength; > + packet.reportId = ((BYTE*)irp->AssociatedIrp.SystemBuffer)[0]; > + if (packet.reportId == 0) > + { > + packet.reportBuffer = &((BYTE*)irp->AssociatedIrp.SystemBuffer)[1]; > + packet.reportBufferLen = irpsp->Parameters.DeviceIoControl.InputBufferLength - 1; > + } > + else > + { > + packet.reportBuffer = irp->AssociatedIrp.SystemBuffer; > + packet.reportBufferLen = irpsp->Parameters.DeviceIoControl.InputBufferLength; > + } > TRACE_(hid_report)("(id %i, len %i buffer %p)\n", packet.reportId, packet.reportBufferLen, packet.reportBuffer); > > rc = call_minidriver(irpsp->Parameters.DeviceIoControl.IoControlCode, > @@ -636,20 +663,15 @@ NTSTATUS WINAPI HID_Device_read(DEVICE_OBJECT *device, IRP *irp) > if (buffer_size) > { > IO_STACK_LOCATION *irpsp = IoGetCurrentIrpStackLocation( irp ); > + NTSTATUS hr; Same here. > + ULONG out_length; > packet->reportBuffer = (BYTE *)packet + sizeof(*packet); > TRACE_(hid_report)("Got Packet %p %i\n", packet->reportBuffer, packet->reportBufferLen); > - if (irpsp->Parameters.Read.Length >= packet->reportBufferLen) > - { > - memcpy(irp->AssociatedIrp.SystemBuffer, packet->reportBuffer, packet->reportBufferLen); > - irp->IoStatus.Information = packet->reportBufferLen; > - irp->IoStatus.u.Status = STATUS_SUCCESS; > - } > - else > - { > - irp->IoStatus.Information = 0; > - irp->IoStatus.u.Status = STATUS_BUFFER_OVERFLOW; > - } > - IoCompleteRequest( irp, IO_NO_INCREMENT ); > + > + hr = copy_packet_into_buffer(packet, irp->AssociatedIrp.SystemBuffer, irpsp->Parameters.Read.Length, &out_length); > + irp->IoStatus.Information = out_length; > + irp->IoStatus.u.Status = hr; > + IoCompleteRequest(irp, IO_NO_INCREMENT); > } > else > { > @@ -671,9 +693,17 @@ NTSTATUS WINAPI HID_Device_write(DEVICE_OBJECT *device, IRP *irp) > irp->IoStatus.Information = 0; > > TRACE_(hid_report)("Device %p Buffer length %i Buffer %p\n", device, irpsp->Parameters.Write.Length, irp->AssociatedIrp.SystemBuffer); > - packet.reportBuffer = irp->AssociatedIrp.SystemBuffer; > - packet.reportId = ((char*)irp->AssociatedIrp.SystemBuffer)[0]; > - packet.reportBufferLen = irpsp->Parameters.Write.Length; > + packet.reportId = ((BYTE*)irp->AssociatedIrp.SystemBuffer)[0]; > + if (packet.reportId == 0) > + { > + packet.reportBuffer = &((BYTE*)irp->AssociatedIrp.SystemBuffer)[1]; > + packet.reportBufferLen = irpsp->Parameters.Write.Length - 1; Besides the fact that the code looks suspicious (couldn't a 0 byte also have a different meaning?) this will probably break the udev backend. As documented in https://www.kernel.org/doc/Documentation/hid/hidraw.txt, the write() command expects that the first byte is 0 if the device does not use numbered reports. > + } > + else > + { > + packet.reportBuffer = irp->AssociatedIrp.SystemBuffer; > + packet.reportBufferLen = irpsp->Parameters.Write.Length; > + } > TRACE_(hid_report)("(id %i, len %i buffer %p)\n", packet.reportId, packet.reportBufferLen, packet.reportBuffer); > > rc = call_minidriver(IOCTL_HID_WRITE_REPORT, device, NULL, 0, &packet, sizeof(packet)); > > >

2 4

Re: [12/23] wined3d: Clear WINED3D_BUFFER_DISCARD flag in unmap.
by Henri Verbeet 14 Nov '16

14 Nov '16

On 13 November 2016 at 19:35, Andrew Wesie <awesie(a)gmail.com> wrote: > @@ -1248,6 +1248,7 @@ static void wined3d_buffer_unmap(struct wined3d_buffer *buffer) > context_release(context); > > buffer_clear_dirty_areas(buffer); > + buffer->flags &= ~WINED3D_BUFFER_DISCARD; Does filtering redundant WINED3D_MAP_DISCARD maps still work after this? Why is this change needed?

1 0

Re: [10/23] d3d11: Reset base vertex index to zero when not using indexed draw.
by Henri Verbeet 14 Nov '16

14 Nov '16

On 13 November 2016 at 19:35, Andrew Wesie <awesie(a)gmail.com> wrote: > Signed-off-by: Andrew Wesie <awesie(a)gmail.com> > --- > dlls/d3d11/device.c | 4 ++++ > 1 file changed, 4 insertions(+) > I think this shouldn't be needed when you do the state invalidation in wined3d_cs_exec_draw(). Ideally we'd probably just get rid of wined3d_device_set_base_vertex_index(), but d3d8 stateblocks complicate that.

1 0