October 2019 - Wine-Devel - list.winehq.org

Re: widl: Set name return value (passed by reference) in the alias case alias case in get_user_type.
by Alexandre Julliard Nov. 22, 2020

Nov. 22, 2020

Gerald Pfeifer <gerald(a)pfeifer.com> writes: > So, I admit I don't really know this code, but looking at it (triggered > by a warning issued by GCC development versions), I noticed that this > variable passed by reference is not initialized here. It's initialized when we return a type, and it doesn't need to be initialized on NULL return. The code is correct, but you could probably set the variable to NULL in the caller to silence the warning. -- Alexandre Julliard julliard(a)winehq.org

2 1

Error handling in RpcBindingVectorFree (dlls/rpcrt4/rpc_binding.c)
by Gerald Pfeifer Nov. 22, 2020

Nov. 22, 2020

Looking at RPC_STATUS WINAPI RpcBindingVectorFree( RPC_BINDING_VECTOR** BindingVector ) { RPC_STATUS status; ULONG c; TRACE("(%p)\n", BindingVector); for (c=0; c<(*BindingVector)->Count; c++) { status = RpcBindingFree(&(*BindingVector)->BindingH[c]); } HeapFree(GetProcessHeap(), 0, *BindingVector); *BindingVector = NULL; return RPC_S_OK; } we currently always ignore the outcome of RpcBindingFree and return RPC_S_OK. However, there is one case where RpcBindingFree returns something different (which is if *Binding is null when RPC_S_INVALID_BINDING is returned). What is the proper way of handling this? Just keeping the code as is and removing the unused status variable? Breaking the loop once RpcBindingFree returns something different from RPC_S_OK? Continuing and returning the first / the last status different from RPC_S_OK? Gerald

1 1

TestBot and graphics tests
by Francois Gouget Aug. 24, 2020

Aug. 24, 2020

Here are some things I've learned about PCI-passthrough recently, which would be one way (probably the best) to add "real hardware" to the TestBot. I don't want to give anyone false hopes though: this just went from "this is a mysterious thing I need to learn about" to "I think I know how to do it but have not tried it yet". So graphics card PCI-passthrough is now relatively well documented on the Internet and seems to have seen some use-cases that would indicate it may even be reasonably usable. * There are two machines intended to run real GPU tests for Wine: cw1-hd6800 and cw2-gtx560. For now they are only used to run WineTest daily on Windows 8.1, Windows 10 1507, 1709, 1809 and Linux. That's quite a bunch but it would be much better if they were integrated with the TestBot as that would allow developers to submit their own tests. So I had a look at what it would imply to convert them to VM hosts using QEmu + PCI-passthrough. * First one needs a processor with hardware virtualisation support. For Intel that's VT-d. Both machines have an Intel Core 2600 which supports VT-d. Good. * Second the motherboard too needs to support VT-d. Both machines have an ASRock P67 Extreme4 motherboard. Unfortunately UEFI says "unsupported" next to the "VT-d" setting for the motherboard :-( It looks like there was some confusion as to whether the P67 chipset supported VT-d initially. From what I gathered it's only Q67 that does but this caused some manufacturers, among which ASRock, to initially claim support and later retract it. * Then one needs to add the intel_iommu=on option to the kernel command line (resp. amd_iommu). This is should make all the PCI devices appear in /sys/kernel/iommu_groups. But that folder remains empty which confirms that full VT-d support is missing. * Another important aspect is to have a graphics card which is hot-restartable. In some cases when a VM's graphics card is crashed the only way to reset it is to reboot the host. The TestBot is likely to crash the graphics card, particularly if we do a hard-power off on the VMs like we currently do, and it would relaly be annoying to have to reboot the host everytime the graphics card goes belly up. I don't know if the AMD HD6800 and Nvidia GTX560 are suitable but it's quite possible they are not. All I know for now is that we should avoid AMD's R9 line of graphics cards. I still need to find a couple of suitable reasonably lower power graphics cards: one AMD and one Nvidia. * Then one needs to prevent the host from using the graphics card. Usually that's done by having the host use the processor's IGP and dedicating the discrete GPU to the VMs. Unfortunately the 2600's IGP cannot be active when there's a discrete card so that route is denied to us. Fortunately there's quite a bit of documentation on how to shut down not just X but also the Linux virtual consoles to free the GPU and hand it over to the VMs after boot. Doing so means losing KVM access to the host which is a bit annoying in case something goes wrong. So ideally we'd make sure this does not happen in grub's "safe mode" boot option. * Although I have not done any test yet I'm reasonably certain that PCI-passthrough rules out live snapshots: QEmu would have no way to restore the graphics card's internal state. - For Windows VMs that's not an issue: if we provide a power off snapshot the TestBot already knows how to power on the VM and wait for it to boot (as long as the boot is shorter than the connection timeout but it works out usually). - For Linux VM's that's more of an issue: the TestBot will power on the VM as usual. The problem is when it updates Wine: after recompiling everything it deletes the old snapshot and creates a new one from the current state of the VM, which means a live snapshot. So the TestBot will need to be modified so it knows when and how to power off the VM and take a powered off snapshot. * Since the VM has full control of the graphics card QEmu has no access to the content of the screen. That's not an issue for the normal TestBot operation, just for the initial VM setup. Fortunately the graphics card is connected to a KVM so the screen can be accessed through that means. It does mean assigning the mouse and keyboard to the VM too. Should that prove impractical there are a bunch of other options too: VNC, LookingGlass, Synergy, etc. But the less needs to be installed in the VMs the better. * Also the TestBot uses QEmu to take the screenshots. But QEmu does not have access to the content of the screen. The fix is to use a tool to take the screenshots from within the VM and use TestAgent to retrieve them. On Linux there are standard tools we can use. On Windows there's code floating around we can use. So the next steps would be: * Maybe test on my box using the builtin IGP. But that likely won't be very conclusive beyond confirming the snapshot issues, screen access, etc. * Find a suitable AMD or Nvidia graphics card and test that on my box. That would allow me to fully test integration with the TestBot, check for stability issues, etc. * Then see what can be done with the existing cw1 and cw2 boxes. -- Francois Gouget <fgouget(a)codeweavers.com>

4 10

About GSoC
by Alex March 3, 2020

March 3, 2020

So, I have this idea of replacing Winetricks with something more beautiful, responsive and functional. I have been developing Wineglass independently as an elementaryOS app (https://github.com/aggalex/Wineglass) and I want to expand it and integrate it with wine in a much better way. If possible, could you inform me about wine's GSoC program? Do you believe such a project is suitable for wine and GSoC?

3 3

[PATCH v2 1/4] d3dcompiler: Get rid of the "node" field from struct hlsl_ir_function_decl.
by Zebediah Figura Feb. 5, 2020

Feb. 5, 2020

Signed-off-by: Zebediah Figura <zfigura(a)codeweavers.com> --- dlls/d3dcompiler_43/d3dcompiler_private.h | 4 ++-- dlls/d3dcompiler_43/hlsl.y | 20 ++++++++++---------- dlls/d3dcompiler_43/utils.c | 4 +--- 3 files changed, 13 insertions(+), 15 deletions(-) diff --git a/dlls/d3dcompiler_43/d3dcompiler_private.h b/dlls/d3dcompiler_43/d3dcompiler_private.h index 1ab343de2d8..01f6ce0e60a 100644 --- a/dlls/d3dcompiler_43/d3dcompiler_private.h +++ b/dlls/d3dcompiler_43/d3dcompiler_private.h @@ -712,7 +712,6 @@ enum hlsl_ir_node_type HLSL_IR_CONSTRUCTOR, HLSL_IR_DEREF, HLSL_IR_EXPR, - HLSL_IR_FUNCTION_DECL, HLSL_IR_IF, HLSL_IR_LOOP, HLSL_IR_JUMP, @@ -776,7 +775,8 @@ struct hlsl_ir_function struct hlsl_ir_function_decl { - struct hlsl_ir_node node; + struct hlsl_type *return_type; + struct source_location loc; struct wine_rb_entry entry; struct hlsl_ir_function *func; const char *semantic; diff --git a/dlls/d3dcompiler_43/hlsl.y b/dlls/d3dcompiler_43/hlsl.y index 5585b899463..7c14ad0091c 100644 --- a/dlls/d3dcompiler_43/hlsl.y +++ b/dlls/d3dcompiler_43/hlsl.y @@ -1084,28 +1084,28 @@ hlsl_prog: /* empty */ { if (decl->body && $2.decl->body) { - hlsl_report_message($2.decl->node.loc.file, $2.decl->node.loc.line, - $2.decl->node.loc.col, HLSL_LEVEL_ERROR, + hlsl_report_message($2.decl->loc.file, $2.decl->loc.line, + $2.decl->loc.col, HLSL_LEVEL_ERROR, "redefinition of function %s", debugstr_a($2.name)); YYABORT; } - else if (!compare_hlsl_types(decl->node.data_type, $2.decl->node.data_type)) + else if (!compare_hlsl_types(decl->return_type, $2.decl->return_type)) { - hlsl_report_message($2.decl->node.loc.file, $2.decl->node.loc.line, - $2.decl->node.loc.col, HLSL_LEVEL_ERROR, + hlsl_report_message($2.decl->loc.file, $2.decl->loc.line, + $2.decl->loc.col, HLSL_LEVEL_ERROR, "redefining function %s with a different return type", debugstr_a($2.name)); - hlsl_report_message(decl->node.loc.file, decl->node.loc.line, decl->node.loc.col, HLSL_LEVEL_NOTE, + hlsl_report_message(decl->loc.file, decl->loc.line, decl->loc.col, HLSL_LEVEL_NOTE, "%s previously declared here", debugstr_a($2.name)); YYABORT; } } - if ($2.decl->node.data_type->base_type == HLSL_TYPE_VOID && $2.decl->semantic) + if ($2.decl->return_type->base_type == HLSL_TYPE_VOID && $2.decl->semantic) { - hlsl_report_message($2.decl->node.loc.file, $2.decl->node.loc.line, - $2.decl->node.loc.col, HLSL_LEVEL_ERROR, + hlsl_report_message($2.decl->loc.file, $2.decl->loc.line, + $2.decl->loc.col, HLSL_LEVEL_ERROR, "void function with a semantic"); } @@ -1280,7 +1280,7 @@ func_prototype: var_modifiers type var_identifier '(' parameters ')' c } $$.name = $3; $$.decl->semantic = $7.semantic; - set_location(&$$.decl->node.loc, &@3); + set_location(&$$.decl->loc, &@3); } compound_statement: '{' '}' diff --git a/dlls/d3dcompiler_43/utils.c b/dlls/d3dcompiler_43/utils.c index 5ecbbc9dc1d..ad2d02e1452 100644 --- a/dlls/d3dcompiler_43/utils.c +++ b/dlls/d3dcompiler_43/utils.c @@ -1577,8 +1577,7 @@ struct hlsl_ir_function_decl *new_func_decl(struct hlsl_type *return_type, struc ERR("Out of memory.\n"); return NULL; } - decl->node.type = HLSL_IR_FUNCTION_DECL; - decl->node.data_type = return_type; + decl->return_type = return_type; decl->parameters = parameters; return decl; @@ -1775,7 +1774,6 @@ static const char *debug_node_type(enum hlsl_ir_node_type type) "HLSL_IR_CONSTRUCTOR", "HLSL_IR_DEREF", "HLSL_IR_EXPR", - "HLSL_IR_FUNCTION_DECL", "HLSL_IR_IF", "HLSL_IR_JUMP", "HLSL_IR_SWIZZLE", -- 2.20.1

2 5

[PATCH 1/2] dlls/d3d12: Fix D3D12SerializeVersionedRootSignature.
by Hans-Kristian Arntzen Nov. 25, 2019

Nov. 25, 2019

Use the proper versioned serializer from vkd3d. Signed-off-by: Hans-Kristian Arntzen <post(a)arntzen-software.no> --- dlls/d3d12/d3d12_main.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/dlls/d3d12/d3d12_main.c b/dlls/d3d12/d3d12_main.c index bd680d8254..9f0fa12ee8 100644 --- a/dlls/d3d12/d3d12_main.c +++ b/dlls/d3d12/d3d12_main.c @@ -420,9 +420,5 @@ HRESULT WINAPI D3D12SerializeVersionedRootSignature(const D3D12_VERSIONED_ROOT_S { TRACE("desc %p, blob %p, error_blob %p.\n", desc, blob, error_blob); - if (desc->Version == D3D_ROOT_SIGNATURE_VERSION_1_0) - return vkd3d_serialize_root_signature(&desc->Desc_1_0, desc->Version, blob, error_blob); - - FIXME("Unsupported version %#x.\n", desc->Version); - return E_NOTIMPL; + return vkd3d_serialize_versioned_root_signature(desc, blob, error_blob); } -- 2.23.0

5 21

[PATCH vkd3d 2/2] vkd3d: Optimise the GPU VA allocator.
by Henri Verbeet Nov. 20, 2019

Nov. 20, 2019

From: Hans-Kristian Arntzen <post(a)arntzen-software.no> The GPU VA allocator was allocating memory in a way where dereferencing GPU VA required a lock + bsearch() to find the right VA range. Rather than going this route, we turn the common case into O(1) and lock-free by creating a slab allocator which allows us to lookup a pointer directly from a GPU VA with (VA - Base) / PageSize. The number of allocations in the fast path must be limited since we cannot trivially grow the allocator while remaining lock-free for dereferences. Signed-off-by: Hans-Kristian Arntzen <post(a)arntzen-software.no> Signed-off-by: Henri Verbeet <hverbeet(a)codeweavers.com> --- This supersedes patch 171219. libs/vkd3d/device.c | 226 +++++++++++++++++++++++++++++++++++++-------- libs/vkd3d/vkd3d_private.h | 28 ++++-- 2 files changed, 204 insertions(+), 50 deletions(-) diff --git a/libs/vkd3d/device.c b/libs/vkd3d/device.c index 2dee5e3..4d6f7c9 100644 --- a/libs/vkd3d/device.c +++ b/libs/vkd3d/device.c @@ -1822,11 +1822,70 @@ static void d3d12_device_destroy_pipeline_cache(struct d3d12_device *device) pthread_mutex_destroy(&device->mutex); } -D3D12_GPU_VIRTUAL_ADDRESS vkd3d_gpu_va_allocator_allocate(struct vkd3d_gpu_va_allocator *allocator, - size_t alignment, size_t size, void *ptr) +#define VKD3D_VA_FALLBACK_BASE 0x8000000000000000ull +#define VKD3D_VA_SLAB_BASE 0x0000001000000000ull +#define VKD3D_VA_SLAB_SIZE_SHIFT 32 +#define VKD3D_VA_SLAB_SIZE (1ull << VKD3D_VA_SLAB_SIZE_SHIFT) +#define VKD3D_VA_SLAB_COUNT (64 * 1024) + +static D3D12_GPU_VIRTUAL_ADDRESS vkd3d_gpu_va_allocator_allocate_slab(struct vkd3d_gpu_va_allocator *allocator, + size_t aligned_size, void *ptr) +{ + struct vkd3d_gpu_va_slab *slab; + D3D12_GPU_VIRTUAL_ADDRESS address; + unsigned slab_idx; + + slab = allocator->free_slab; + allocator->free_slab = slab->ptr; + slab->size = aligned_size; + slab->ptr = ptr; + + /* It is critical that the multiplication happens in 64-bit to not + * overflow. */ + slab_idx = slab - allocator->slabs; + address = VKD3D_VA_SLAB_BASE + slab_idx * VKD3D_VA_SLAB_SIZE; + + TRACE("Allocated address %#"PRIx64", slab %u, size %zu.\n", address, slab_idx, aligned_size); + + return address; +} + +static D3D12_GPU_VIRTUAL_ADDRESS vkd3d_gpu_va_allocator_allocate_fallback(struct vkd3d_gpu_va_allocator *allocator, + size_t alignment, size_t aligned_size, void *ptr) { struct vkd3d_gpu_va_allocation *allocation; D3D12_GPU_VIRTUAL_ADDRESS base, ceiling; + + base = allocator->fallback_floor; + ceiling = ~(D3D12_GPU_VIRTUAL_ADDRESS)0; + ceiling -= alignment - 1; + if (aligned_size > ceiling || ceiling - aligned_size < base) + return 0; + + base = (base + (alignment - 1)) & ~((D3D12_GPU_VIRTUAL_ADDRESS)alignment - 1); + + if (!vkd3d_array_reserve((void **)&allocator->fallback_allocations, &allocator->fallback_allocations_size, + allocator->fallback_allocation_count + 1, sizeof(*allocator->fallback_allocations))) + return 0; + + allocation = &allocator->fallback_allocations[allocator->fallback_allocation_count++]; + allocation->base = base; + allocation->size = aligned_size; + allocation->ptr = ptr; + + /* This pointer is bumped and never lowered on a free. However, this will + * only fail once we have exhausted 63 bits of address space. */ + allocator->fallback_floor = base + aligned_size; + + TRACE("Allocated address %#"PRIx64", size %zu.\n", base, aligned_size); + + return base; +} + +D3D12_GPU_VIRTUAL_ADDRESS vkd3d_gpu_va_allocator_allocate(struct vkd3d_gpu_va_allocator *allocator, + size_t alignment, size_t size, void *ptr) +{ + D3D12_GPU_VIRTUAL_ADDRESS address; int rc; if (size > ~(size_t)0 - (alignment - 1)) @@ -1839,34 +1898,41 @@ D3D12_GPU_VIRTUAL_ADDRESS vkd3d_gpu_va_allocator_allocate(struct vkd3d_gpu_va_al return 0; } - base = allocator->floor; - ceiling = ~(D3D12_GPU_VIRTUAL_ADDRESS)0; - ceiling -= alignment - 1; - if (size > ceiling || ceiling - size < base) - { - pthread_mutex_unlock(&allocator->mutex); - return 0; - } + if (size <= VKD3D_VA_SLAB_SIZE && allocator->free_slab) + address = vkd3d_gpu_va_allocator_allocate_slab(allocator, size, ptr); + else + address = vkd3d_gpu_va_allocator_allocate_fallback(allocator, alignment, size, ptr); - base = (base + (alignment - 1)) & ~((D3D12_GPU_VIRTUAL_ADDRESS)alignment - 1); + pthread_mutex_unlock(&allocator->mutex); - if (!vkd3d_array_reserve((void **)&allocator->allocations, &allocator->allocations_size, - allocator->allocation_count + 1, sizeof(*allocator->allocations))) - { - pthread_mutex_unlock(&allocator->mutex); - return 0; - } + return address; +} - allocation = &allocator->allocations[allocator->allocation_count++]; - allocation->base = base; - allocation->size = size; - allocation->ptr = ptr; +static void *vkd3d_gpu_va_allocator_dereference_slab(struct vkd3d_gpu_va_allocator *allocator, + D3D12_GPU_VIRTUAL_ADDRESS address) +{ + const struct vkd3d_gpu_va_slab *slab; + D3D12_GPU_VIRTUAL_ADDRESS base_offset; + unsigned int slab_idx; - allocator->floor = base + size; + base_offset = address - VKD3D_VA_SLAB_BASE; + slab_idx = base_offset >> VKD3D_VA_SLAB_SIZE_SHIFT; - pthread_mutex_unlock(&allocator->mutex); + if (slab_idx >= VKD3D_VA_SLAB_COUNT) + { + ERR("Invalid slab index %u for address %#"PRIx64".\n", slab_idx, address); + return NULL; + } - return allocation->base; + slab = &allocator->slabs[slab_idx]; + base_offset -= slab_idx * VKD3D_VA_SLAB_SIZE; + if (base_offset >= slab->size) + { + ERR("Address %#"PRIx64" is %#"PRIx64" bytes into slab %u of size %zu.\n", + address, base_offset, slab_idx, slab->size); + return NULL; + } + return slab->ptr; } static int vkd3d_gpu_va_allocation_compare(const void *k, const void *e) @@ -1881,30 +1947,94 @@ static int vkd3d_gpu_va_allocation_compare(const void *k, const void *e) return 0; } -void *vkd3d_gpu_va_allocator_dereference(struct vkd3d_gpu_va_allocator *allocator, +static void *vkd3d_gpu_va_allocator_dereference_fallback(struct vkd3d_gpu_va_allocator *allocator, D3D12_GPU_VIRTUAL_ADDRESS address) { struct vkd3d_gpu_va_allocation *allocation; + + allocation = bsearch(&address, allocator->fallback_allocations, allocator->fallback_allocation_count, + sizeof(*allocation), vkd3d_gpu_va_allocation_compare); + + return allocation ? allocation->ptr : NULL; +} + +void *vkd3d_gpu_va_allocator_dereference(struct vkd3d_gpu_va_allocator *allocator, + D3D12_GPU_VIRTUAL_ADDRESS address) +{ + void *ret; int rc; + /* If we land in the non-fallback region, dereferencing VA is lock-less. + * The base pointer is immutable, and the only way we can have a data race + * is if some other thread is poking into the + * slab_mem_allocation[base_index] block. This can only happen if someone + * is trying to free the entry while we're dereferencing it, which would + * be a serious application bug. */ + if (address < VKD3D_VA_FALLBACK_BASE) + return vkd3d_gpu_va_allocator_dereference_slab(allocator, address); + + /* Slow fallback. */ if ((rc = pthread_mutex_lock(&allocator->mutex))) { ERR("Failed to lock mutex, error %d.\n", rc); return NULL; } - allocation = bsearch(&address, allocator->allocations, allocator->allocation_count, - sizeof(*allocation), vkd3d_gpu_va_allocation_compare); + ret = vkd3d_gpu_va_allocator_dereference_fallback(allocator, address); pthread_mutex_unlock(&allocator->mutex); - return allocation ? allocation->ptr : NULL; + return ret; } -void vkd3d_gpu_va_allocator_free(struct vkd3d_gpu_va_allocator *allocator, D3D12_GPU_VIRTUAL_ADDRESS address) +static void vkd3d_gpu_va_allocator_free_slab(struct vkd3d_gpu_va_allocator *allocator, + D3D12_GPU_VIRTUAL_ADDRESS address) +{ + D3D12_GPU_VIRTUAL_ADDRESS base_offset; + struct vkd3d_gpu_va_slab *slab; + unsigned int slab_idx; + + base_offset = address - VKD3D_VA_SLAB_BASE; + slab_idx = base_offset >> VKD3D_VA_SLAB_SIZE_SHIFT; + + if (slab_idx >= VKD3D_VA_SLAB_COUNT) + { + ERR("Invalid slab index %u for address %#"PRIx64".\n", slab_idx, address); + return; + } + + TRACE("Freeing address %#"PRIx64", slab %u.\n", address, slab_idx); + + slab = &allocator->slabs[slab_idx]; + slab->size = 0; + slab->ptr = allocator->free_slab; + allocator->free_slab = slab; +} + +static void vkd3d_gpu_va_allocator_free_fallback(struct vkd3d_gpu_va_allocator *allocator, + D3D12_GPU_VIRTUAL_ADDRESS address) { struct vkd3d_gpu_va_allocation *allocation; unsigned int index; + + allocation = bsearch(&address, allocator->fallback_allocations, allocator->fallback_allocation_count, + sizeof(*allocation), vkd3d_gpu_va_allocation_compare); + + if (!allocation || allocation->base != address) + { + ERR("Address %#"PRIx64" does not match any allocation.\n", address); + return; + } + + index = allocation - allocator->fallback_allocations; + --allocator->fallback_allocation_count; + if (index != allocator->fallback_allocation_count) + memmove(&allocator->fallback_allocations[index], &allocator->fallback_allocations[index + 1], + (allocator->fallback_allocation_count - index) * sizeof(*allocation)); +} + +void vkd3d_gpu_va_allocator_free(struct vkd3d_gpu_va_allocator *allocator, D3D12_GPU_VIRTUAL_ADDRESS address) +{ int rc; if ((rc = pthread_mutex_lock(&allocator->mutex))) @@ -1913,32 +2043,45 @@ void vkd3d_gpu_va_allocator_free(struct vkd3d_gpu_va_allocator *allocator, D3D12 return; } - allocation = bsearch(&address, allocator->allocations, allocator->allocation_count, - sizeof(*allocation), vkd3d_gpu_va_allocation_compare); - if (allocation && allocation->base == address) + if (address < VKD3D_VA_FALLBACK_BASE) { - index = allocation - allocator->allocations; - --allocator->allocation_count; - if (index != allocator->allocation_count) - { - memmove(&allocator->allocations[index], &allocator->allocations[index + 1], - (allocator->allocation_count - index) * sizeof(*allocation)); - } + vkd3d_gpu_va_allocator_free_slab(allocator, address); + pthread_mutex_unlock(&allocator->mutex); + return; } + vkd3d_gpu_va_allocator_free_fallback(allocator, address); + pthread_mutex_unlock(&allocator->mutex); } static bool vkd3d_gpu_va_allocator_init(struct vkd3d_gpu_va_allocator *allocator) { + unsigned int i; int rc; memset(allocator, 0, sizeof(*allocator)); - allocator->floor = 0x1000; + allocator->fallback_floor = VKD3D_VA_FALLBACK_BASE; + + /* To remain lock-less, we cannot grow the slabs array after the fact. If + * we commit to a maximum number of allocations here, we can dereference + * without taking a lock as the base pointer never changes. We would be + * able to grow more seamlessly using an array of pointers, but that would + * make dereferencing slightly less efficient. */ + if (!(allocator->slabs = vkd3d_calloc(VKD3D_VA_SLAB_COUNT, sizeof(*allocator->slabs)))) + return false; + + /* Mark all slabs as free. */ + allocator->free_slab = &allocator->slabs[0]; + for (i = 0; i < VKD3D_VA_SLAB_COUNT - 1; ++i) + { + allocator->slabs[i].ptr = &allocator->slabs[i + 1]; + } if ((rc = pthread_mutex_init(&allocator->mutex, NULL))) { ERR("Failed to initialize mutex, error %d.\n", rc); + vkd3d_free(allocator->slabs); return false; } @@ -1954,7 +2097,8 @@ static void vkd3d_gpu_va_allocator_cleanup(struct vkd3d_gpu_va_allocator *alloca ERR("Failed to lock mutex, error %d.\n", rc); return; } - vkd3d_free(allocator->allocations); + vkd3d_free(allocator->slabs); + vkd3d_free(allocator->fallback_allocations); pthread_mutex_unlock(&allocator->mutex); pthread_mutex_destroy(&allocator->mutex); } diff --git a/libs/vkd3d/vkd3d_private.h b/libs/vkd3d/vkd3d_private.h index 520dab1..1cb40f2 100644 --- a/libs/vkd3d/vkd3d_private.h +++ b/libs/vkd3d/vkd3d_private.h @@ -202,20 +202,30 @@ HRESULT vkd3d_fence_worker_start(struct vkd3d_fence_worker *worker, HRESULT vkd3d_fence_worker_stop(struct vkd3d_fence_worker *worker, struct d3d12_device *device) DECLSPEC_HIDDEN; +struct vkd3d_gpu_va_allocation +{ + D3D12_GPU_VIRTUAL_ADDRESS base; + SIZE_T size; + void *ptr; +}; + +struct vkd3d_gpu_va_slab +{ + SIZE_T size; + void *ptr; +}; + struct vkd3d_gpu_va_allocator { pthread_mutex_t mutex; - D3D12_GPU_VIRTUAL_ADDRESS floor; + D3D12_GPU_VIRTUAL_ADDRESS fallback_floor; + struct vkd3d_gpu_va_allocation *fallback_allocations; + size_t fallback_allocations_size; + size_t fallback_allocation_count; - struct vkd3d_gpu_va_allocation - { - D3D12_GPU_VIRTUAL_ADDRESS base; - SIZE_T size; - void *ptr; - } *allocations; - size_t allocations_size; - size_t allocation_count; + struct vkd3d_gpu_va_slab *slabs; + struct vkd3d_gpu_va_slab *free_slab; }; D3D12_GPU_VIRTUAL_ADDRESS vkd3d_gpu_va_allocator_allocate(struct vkd3d_gpu_va_allocator *allocator, -- 2.11.0

4 5

[PATCH] wininet/tests: Skip some FTP tests when no connection can be made.
by Sven Baars Nov. 12, 2019

Nov. 12, 2019

Signed-off-by: Sven Baars <sven.wine(a)gmail.com> --- This happens quite often, see https://test.winehq.org/data/tests/wininet:ftp.html Also during bbf74ea22a5310978c9aecbdc648c562c26fa0fb the behavior of this test was changed, so I modified the comments accordingly. dlls/wininet/tests/ftp.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/dlls/wininet/tests/ftp.c b/dlls/wininet/tests/ftp.c index 317b291495..c6d972a352 100644 --- a/dlls/wininet/tests/ftp.c +++ b/dlls/wininet/tests/ftp.c @@ -63,7 +63,7 @@ static void test_connect(HINTERNET hInternet) HINTERNET hFtp; /* Try a few username/password combinations: - * anonymous : NULL + * anonymous : IEUser@ * NULL : IEUser@ * NULL : NULL * "" : IEUser@ @@ -72,14 +72,14 @@ static void test_connect(HINTERNET hInternet) SetLastError(0xdeadbeef); hFtp = InternetConnectA(hInternet, "ftp.winehq.org", INTERNET_DEFAULT_FTP_PORT, "anonymous", "IEUser@", INTERNET_SERVICE_FTP, INTERNET_FLAG_PASSIVE, 0); - if (hFtp) /* some servers accept an empty password */ + if (!hFtp) { - ok ( GetLastError() == ERROR_SUCCESS, "ERROR_SUCCESS, got %d\n", GetLastError()); - InternetCloseHandle(hFtp); + skip("No ftp connection could be made to ftp.winehq.org %u\n", GetLastError()); + return; } - else - ok ( GetLastError() == ERROR_INTERNET_LOGIN_FAILURE, - "Expected ERROR_INTERNET_LOGIN_FAILURE, got %d\n", GetLastError()); + ok(GetLastError() == ERROR_SUCCESS, + "Expected ERROR_SUCCESS, got %d\n", GetLastError()); + InternetCloseHandle(hFtp); SetLastError(0xdeadbeef); hFtp = InternetConnectA(hInternet, "ftp.winehq.org", INTERNET_DEFAULT_FTP_PORT, NULL, "IEUser@", INTERNET_SERVICE_FTP, INTERNET_FLAG_PASSIVE, 0); -- 2.17.1

3 7

Wineconf 2019
by paulo lesgaz Nov. 9, 2019

Nov. 9, 2019

Hello, will videos of wineconf 2019 be available? Thanks in advance

4 4

Wine Gecko status
by Michael Cronenworth Nov. 7, 2019

Nov. 7, 2019

Hi, What is the status of development regarding Wine Gecko? The last general release was June 2016. As the maintainer in Fedora I have a few concerns that I would like to address. 1. The build scripts require Python 2. Fedora is dropping Python 2 in Fedora 31/32. 31 is only a few months away from GA. 2. I have been unable to build from source with the latest MinGW toolset since November 2016. I've last reported this directly to Jacek in 2018, but I have never heard from him. 3. If no support arrives, I don't have the time to patch wine-gecko. It will be dropped from Fedora and users will be downloading the binaries from winehq.org. This provides a reduced user experience and I would like to prevent it. Thanks, Michael

4 15