Files
mesa/src/vulkan/util/vk_util.h
T
Paulo Zanoni b0653370d0 vulkan: don't zero-initialize STACK_ARRAY()'s stack array
STACK_ARRAY() is used in a lot of places. When games are running we
see STACK_ARRAY() arrays being used all the time: each queue
submission uses 6, WaitSemaphores and syncobj waiting also uses them:
they're constantly present in Vulkan runtime.

There's no need for STACK_ARRAY()'s stack array to be initialized,
callers cannot not depend on it. If the number of elements is greater
than STACK_ARRAY_SIZE, then STACK_ARRAY() will just malloc() the array
and return it not initialized: anybody depending of
zero-initialization is going to break when the array is big.

The reason why we're zero-intializing STACK_ARRAY()'s stack array is
to silence -Wmaybe-uninitialized warnings: see commit d7957df318
("vulkan: fix uninitialized variables"). I don't think that commit is
the ideal way to deal with the problem, so this patch proposes a
better solution.

The problem here is that zero-initializing it adds code we don't need
for every single caller. STACK_ARRAY() already has 63 callers and only
3 of them are affected by the -Wmaybe-uninitialized warining. So here
we undo what commit d7957df318 did and instead we fix the 3 cases
that actually generate the -Wmaybe-uninitialized warnings.

Gcc is only emitting those warinings because it knows that the number
of elements in the array may be zero, so the loops we have that set
elements to the array may end up do nothing, and then we pass the
array uninitialized to other functions.

For the cases related to vk_sync this is just returning VK_SUCCESS
earlier, instead of relying on the check that eventually happens at
__vk_sync_wait_many(). For the vkCmdWaitEvents() function, the Vulkan
spec says that "eventCount must be greater than 0", so the early
return doesn't hurt anybody either. In both cases we make the zero
case faster by not defining an 8-sized array, zero-initializing it,
then returning success without using it.

Reference: d7957df318 ("vulkan: fix uninitialized variables")
Acked-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28288>
2024-04-08 17:23:25 +00:00

397 lines
12 KiB
C

/*
* Copyright © 2017 Intel Corporation
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice (including the next
* paragraph) shall be included in all copies or substantial portions of the
* Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
* IN THE SOFTWARE.
*/
#ifndef VK_UTIL_H
#define VK_UTIL_H
#include "util/bitscan.h"
#include "util/macros.h"
#include "compiler/shader_enums.h"
#include <stdlib.h>
#include <string.h>
#include "vk_struct_type_cast.h"
#ifdef __cplusplus
extern "C" {
#endif
/* common inlines and macros for vulkan drivers */
#include <vulkan/vulkan_core.h>
struct vk_pnext_iterator {
VkBaseOutStructure *pos;
#ifndef NDEBUG
VkBaseOutStructure *half_pos;
unsigned idx;
#endif
bool done;
};
static inline struct vk_pnext_iterator
vk_pnext_iterator_init(void *start)
{
struct vk_pnext_iterator iter;
iter.pos = (VkBaseOutStructure *)start;
#ifndef NDEBUG
iter.half_pos = (VkBaseOutStructure *)start;
iter.idx = 0;
#endif
iter.done = false;
return iter;
}
static inline struct vk_pnext_iterator
vk_pnext_iterator_init_const(const void *start)
{
return vk_pnext_iterator_init((void *)start);
}
static inline VkBaseOutStructure *
vk_pnext_iterator_next(struct vk_pnext_iterator *iter)
{
iter->pos = iter->pos->pNext;
#ifndef NDEBUG
if (iter->idx++ & 1) {
/** This the "tortoise and the hare" algorithm. We increment
* chaser->pNext every other time *iter gets incremented. Because *iter
* is incrementing twice as fast as chaser->pNext, the distance between
* them in the list increases by one for each time we get here. If we
* have a loop, eventually, both iterators will be inside the loop and
* this distance will be an integer multiple of the loop length, at
* which point the two pointers will be equal.
*/
iter->half_pos = iter->half_pos->pNext;
if (iter->half_pos == iter->pos)
assert(!"Vulkan input pNext chain has a loop!");
}
#endif
return iter->pos;
}
/* Because the outer loop only executes once, independently of what happens in
* the inner loop, breaks and continues should work exactly the same as if
* there were only one for loop.
*/
#define vk_foreach_struct(__e, __start) \
for (struct vk_pnext_iterator __iter = vk_pnext_iterator_init(__start); \
!__iter.done; __iter.done = true) \
for (VkBaseOutStructure *__e = __iter.pos; \
__e; __e = vk_pnext_iterator_next(&__iter))
#define vk_foreach_struct_const(__e, __start) \
for (struct vk_pnext_iterator __iter = \
vk_pnext_iterator_init_const(__start); \
!__iter.done; __iter.done = true) \
for (const VkBaseInStructure *__e = (VkBaseInStructure *)__iter.pos; \
__e; __e = (VkBaseInStructure *)vk_pnext_iterator_next(&__iter))
static inline void
vk_copy_struct_guts(VkBaseOutStructure *dst, VkBaseInStructure *src, size_t struct_size)
{
STATIC_ASSERT(sizeof(*dst) == sizeof(*src));
memcpy(dst + 1, src + 1, struct_size - sizeof(VkBaseOutStructure));
}
/**
* A wrapper for a Vulkan output array. A Vulkan output array is one that
* follows the convention of the parameters to
* vkGetPhysicalDeviceQueueFamilyProperties().
*
* Example Usage:
*
* VkResult
* vkGetPhysicalDeviceQueueFamilyProperties(
* VkPhysicalDevice physicalDevice,
* uint32_t* pQueueFamilyPropertyCount,
* VkQueueFamilyProperties* pQueueFamilyProperties)
* {
* VK_OUTARRAY_MAKE_TYPED(VkQueueFamilyProperties, props,
* pQueueFamilyProperties,
* pQueueFamilyPropertyCount);
*
* vk_outarray_append_typed(VkQueueFamilyProperties, &props, p) {
* p->queueFlags = ...;
* p->queueCount = ...;
* }
*
* vk_outarray_append_typed(VkQueueFamilyProperties, &props, p) {
* p->queueFlags = ...;
* p->queueCount = ...;
* }
*
* return vk_outarray_status(&props);
* }
*/
struct __vk_outarray {
/** May be null. */
void *data;
/**
* Capacity, in number of elements. Capacity is unlimited (UINT32_MAX) if
* data is null.
*/
uint32_t cap;
/**
* Count of elements successfully written to the array. Every write is
* considered successful if data is null.
*/
uint32_t *filled_len;
/**
* Count of elements that would have been written to the array if its
* capacity were sufficient. Vulkan functions often return VK_INCOMPLETE
* when `*filled_len < wanted_len`.
*/
uint32_t wanted_len;
};
static inline void
__vk_outarray_init(struct __vk_outarray *a,
void *data, uint32_t *restrict len)
{
a->data = data;
a->cap = *len;
a->filled_len = len;
*a->filled_len = 0;
a->wanted_len = 0;
if (a->data == NULL)
a->cap = UINT32_MAX;
}
static inline VkResult
__vk_outarray_status(const struct __vk_outarray *a)
{
if (*a->filled_len < a->wanted_len)
return VK_INCOMPLETE;
else
return VK_SUCCESS;
}
static inline void *
__vk_outarray_next(struct __vk_outarray *a, size_t elem_size)
{
void *p = NULL;
a->wanted_len += 1;
if (*a->filled_len >= a->cap)
return NULL;
if (a->data != NULL)
p = (uint8_t *)a->data + (*a->filled_len) * elem_size;
*a->filled_len += 1;
return p;
}
#define vk_outarray(elem_t) \
struct { \
struct __vk_outarray base; \
elem_t meta[]; \
}
#define vk_outarray_typeof_elem(a) __typeof__((a)->meta[0])
#define vk_outarray_sizeof_elem(a) sizeof((a)->meta[0])
#define vk_outarray_init(a, data, len) \
__vk_outarray_init(&(a)->base, (data), (len))
#define VK_OUTARRAY_MAKE_TYPED(type, name, data, len) \
vk_outarray(type) name; \
vk_outarray_init(&name, (data), (len))
#define vk_outarray_status(a) \
__vk_outarray_status(&(a)->base)
#define vk_outarray_next(a) \
vk_outarray_next_typed(vk_outarray_typeof_elem(a), a)
#define vk_outarray_next_typed(type, a) \
((type *) \
__vk_outarray_next(&(a)->base, vk_outarray_sizeof_elem(a)))
/**
* Append to a Vulkan output array.
*
* This is a block-based macro. For example:
*
* vk_outarray_append_typed(T, &a, elem) {
* elem->foo = ...;
* elem->bar = ...;
* }
*
* The array `a` has type `vk_outarray(elem_t) *`. It is usually declared with
* VK_OUTARRAY_MAKE_TYPED(). The variable `elem` is block-scoped and has type
* `elem_t *`.
*
* The macro unconditionally increments the array's `wanted_len`. If the array
* is not full, then the macro also increment its `filled_len` and then
* executes the block. When the block is executed, `elem` is non-null and
* points to the newly appended element.
*/
#define vk_outarray_append_typed(type, a, elem) \
for (type *elem = vk_outarray_next_typed(type, a); \
elem != NULL; elem = NULL)
static inline void *
__vk_find_struct(void *start, VkStructureType sType)
{
vk_foreach_struct(s, start) {
if (s->sType == sType)
return s;
}
return NULL;
}
#define vk_find_struct(__start, __sType) \
(VK_STRUCTURE_TYPE_##__sType##_cast *)__vk_find_struct( \
(__start), VK_STRUCTURE_TYPE_##__sType)
#define vk_find_struct_const(__start, __sType) \
(const VK_STRUCTURE_TYPE_##__sType##_cast *)__vk_find_struct( \
(void *)(__start), VK_STRUCTURE_TYPE_##__sType)
static inline void
__vk_append_struct(void *start, void *element)
{
vk_foreach_struct(s, start) {
if (s->pNext)
continue;
s->pNext = (struct VkBaseOutStructure *) element;
break;
}
}
uint32_t vk_get_driver_version(void);
uint32_t vk_get_version_override(void);
void vk_warn_non_conformant_implementation(const char *driver_name);
struct vk_pipeline_cache_header {
uint32_t header_size;
uint32_t header_version;
uint32_t vendor_id;
uint32_t device_id;
uint8_t uuid[VK_UUID_SIZE];
};
#define VK_EXT_OFFSET (1000000000UL)
#define VK_ENUM_EXTENSION(__enum) \
((__enum) >= VK_EXT_OFFSET ? ((((__enum) - VK_EXT_OFFSET) / 1000UL) + 1) : 0)
#define VK_ENUM_OFFSET(__enum) \
((__enum) >= VK_EXT_OFFSET ? ((__enum) % 1000) : (__enum))
#define typed_memcpy(dest, src, count) do { \
STATIC_ASSERT(sizeof(*(src)) == sizeof(*(dest))); \
memcpy((dest), (src), (count) * sizeof(*(src))); \
} while (0)
static inline gl_shader_stage
vk_to_mesa_shader_stage(VkShaderStageFlagBits vk_stage)
{
assert(util_bitcount((uint32_t) vk_stage) == 1);
return (gl_shader_stage) (ffs((uint32_t) vk_stage) - 1);
}
static inline VkShaderStageFlagBits
mesa_to_vk_shader_stage(gl_shader_stage mesa_stage)
{
return (VkShaderStageFlagBits) (1 << ((uint32_t) mesa_stage));
}
/* iterate over a sequence of indexed multidraws for VK_EXT_multi_draw extension */
/* 'i' must be explicitly declared */
#define vk_foreach_multi_draw_indexed(_draw, _i, _pDrawInfo, _num_draws, _stride) \
for (const VkMultiDrawIndexedInfoEXT *_draw = (const VkMultiDrawIndexedInfoEXT*)(_pDrawInfo); \
(_i) < (_num_draws); \
(_i)++, (_draw) = (const VkMultiDrawIndexedInfoEXT*)((const uint8_t*)(_draw) + (_stride)))
/* iterate over a sequence of multidraws for VK_EXT_multi_draw extension */
/* 'i' must be explicitly declared */
#define vk_foreach_multi_draw(_draw, _i, _pDrawInfo, _num_draws, _stride) \
for (const VkMultiDrawInfoEXT *_draw = (const VkMultiDrawInfoEXT*)(_pDrawInfo); \
(_i) < (_num_draws); \
(_i)++, (_draw) = (const VkMultiDrawInfoEXT*)((const uint8_t*)(_draw) + (_stride)))
struct nir_spirv_specialization;
struct nir_spirv_specialization*
vk_spec_info_to_nir_spirv(const VkSpecializationInfo *spec_info,
uint32_t *out_num_spec_entries);
#define STACK_ARRAY_SIZE 8
/* Sometimes gcc may claim -Wmaybe-uninitialized for the stack array in some
* places it can't verify that when size is 0 nobody down the call chain reads
* the array. Please don't try to fix it by zero-initializing the array here
* since it's used in a lot of different places. An "if (size == 0) return;"
* may work for you.
*/
#define STACK_ARRAY(type, name, size) \
type _stack_##name[STACK_ARRAY_SIZE]; \
type *const name = \
((size) <= STACK_ARRAY_SIZE ? _stack_##name : (type *)malloc((size) * sizeof(type)))
#define STACK_ARRAY_FINISH(name) \
if (name != _stack_##name) free(name)
static inline uint8_t
vk_index_type_to_bytes(enum VkIndexType type)
{
switch (type) {
case VK_INDEX_TYPE_NONE_KHR: return 0;
case VK_INDEX_TYPE_UINT8_KHR: return 1;
case VK_INDEX_TYPE_UINT16: return 2;
case VK_INDEX_TYPE_UINT32: return 4;
default: unreachable("Invalid index type");
}
}
static inline uint32_t
vk_index_to_restart(enum VkIndexType type)
{
switch (type) {
case VK_INDEX_TYPE_UINT8_KHR: return 0xff;
case VK_INDEX_TYPE_UINT16: return 0xffff;
case VK_INDEX_TYPE_UINT32: return 0xffffffff;
default: unreachable("unexpected index type");
}
}
#ifdef __cplusplus
}
#endif
#endif /* VK_UTIL_H */