diff options
author | Eric Anholt <[email protected]> | 2017-02-02 16:24:13 -0800 |
---|---|---|
committer | Eric Anholt <[email protected]> | 2017-10-10 11:42:04 -0700 |
commit | 45bb8f2957108056de10721d50259e2f8c5a4ddc (patch) | |
tree | 37abe7d19c4e9d20990e1e42b2eeb61f00eab09e /src/gallium/drivers/vc5/vc5_cl.h | |
parent | ade416d02369cc0942d53ad3cce601d66344f9c3 (diff) |
broadcom: Add V3D 3.3 gallium driver called "vc5", for BCM7268.
V3D 3.3 is a continuation of the 3D implementation in VC4 (v2.1 and v2.6).
V3D 3.3 introduces an MMU (no more CMA allocations) and support for
GLES3.1. This driver is not currently conformant, though that will be a
target as soon as possible.
V3D 3.x parts use a new texture tiling layout common across many Broadcom
graphics parts including and the HVS scanout engine. It also massively
changes the QPU instructions, introducing a common physical register file
(no more A/B split) and half-float instructions, while removing the 4x8
unorm instructions in favor of half-float for talking to fixed function
interfaces. Because so much has changed, vc5 is implemented in a separate
gallium driver, using only the XML code-generation support from vc4.
v2: Fix tile layout for 64bpp textures. Fix texture swizzling for 32-bit
returns. Fix up a bit of MRT setup. Sync the simulator to kernel
behavior a bit more. Improve uniform debugging code. Rebase on
QIR->VIR rename. Move texture state mostly to the CSOs. Improve
cache flushing on the simulator. Fix program deletion
use-after-frees.
Acked-by: Dave Airlie <[email protected]> (uabi plan)
Acked-by: Daniel Vetter <[email protected]> (uabi plan)
Diffstat (limited to 'src/gallium/drivers/vc5/vc5_cl.h')
-rw-r--r-- | src/gallium/drivers/vc5/vc5_cl.h | 246 |
1 files changed, 246 insertions, 0 deletions
diff --git a/src/gallium/drivers/vc5/vc5_cl.h b/src/gallium/drivers/vc5/vc5_cl.h new file mode 100644 index 00000000000..e935eeff536 --- /dev/null +++ b/src/gallium/drivers/vc5/vc5_cl.h @@ -0,0 +1,246 @@ +/* + * Copyright © 2014-2017 Broadcom + * + * 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 VC5_CL_H +#define VC5_CL_H + +#include <stdint.h> + +#include "util/u_math.h" +#include "util/macros.h" + +struct vc5_bo; +struct vc5_job; +struct vc5_cl; + +/** + * Undefined structure, used for typechecking that you're passing the pointers + * to these functions correctly. + */ +struct vc5_cl_out; + +/** A reference to a BO used in the CL packing functions */ +struct vc5_cl_reloc { + struct vc5_bo *bo; + uint32_t offset; +}; + +static inline void cl_pack_emit_reloc(struct vc5_cl *cl, const struct vc5_cl_reloc *); + +#define __gen_user_data struct vc5_cl +#define __gen_address_type struct vc5_cl_reloc +#define __gen_address_offset(reloc) (((reloc)->bo ? (reloc)->bo->offset : 0) + \ + (reloc)->offset) +#define __gen_emit_reloc cl_pack_emit_reloc + +struct vc5_cl { + void *base; + struct vc5_job *job; + struct vc5_cl_out *next; + struct vc5_bo *bo; + uint32_t size; +}; + +void vc5_init_cl(struct vc5_job *job, struct vc5_cl *cl); +void vc5_destroy_cl(struct vc5_cl *cl); +void vc5_dump_cl(void *cl, uint32_t size, bool is_render); +uint32_t vc5_gem_hindex(struct vc5_job *job, struct vc5_bo *bo); + +struct PACKED unaligned_16 { uint16_t x; }; +struct PACKED unaligned_32 { uint32_t x; }; + +static inline uint32_t cl_offset(struct vc5_cl *cl) +{ + return (char *)cl->next - (char *)cl->base; +} + +static inline void +cl_advance(struct vc5_cl_out **cl, uint32_t n) +{ + (*cl) = (struct vc5_cl_out *)((char *)(*cl) + n); +} + +static inline struct vc5_cl_out * +cl_start(struct vc5_cl *cl) +{ + return cl->next; +} + +static inline void +cl_end(struct vc5_cl *cl, struct vc5_cl_out *next) +{ + cl->next = next; + assert(cl_offset(cl) <= cl->size); +} + + +static inline void +put_unaligned_32(struct vc5_cl_out *ptr, uint32_t val) +{ + struct unaligned_32 *p = (void *)ptr; + p->x = val; +} + +static inline void +put_unaligned_16(struct vc5_cl_out *ptr, uint16_t val) +{ + struct unaligned_16 *p = (void *)ptr; + p->x = val; +} + +static inline void +cl_u8(struct vc5_cl_out **cl, uint8_t n) +{ + *(uint8_t *)(*cl) = n; + cl_advance(cl, 1); +} + +static inline void +cl_u16(struct vc5_cl_out **cl, uint16_t n) +{ + put_unaligned_16(*cl, n); + cl_advance(cl, 2); +} + +static inline void +cl_u32(struct vc5_cl_out **cl, uint32_t n) +{ + put_unaligned_32(*cl, n); + cl_advance(cl, 4); +} + +static inline void +cl_aligned_u32(struct vc5_cl_out **cl, uint32_t n) +{ + *(uint32_t *)(*cl) = n; + cl_advance(cl, 4); +} + +static inline void +cl_aligned_reloc(struct vc5_cl *cl, + struct vc5_cl_out **cl_out, + struct vc5_bo *bo, uint32_t offset) +{ + cl_aligned_u32(cl_out, bo->offset + offset); + vc5_job_add_bo(cl->job, bo); +} + +static inline void +cl_ptr(struct vc5_cl_out **cl, void *ptr) +{ + *(struct vc5_cl_out **)(*cl) = ptr; + cl_advance(cl, sizeof(void *)); +} + +static inline void +cl_f(struct vc5_cl_out **cl, float f) +{ + cl_u32(cl, fui(f)); +} + +static inline void +cl_aligned_f(struct vc5_cl_out **cl, float f) +{ + cl_aligned_u32(cl, fui(f)); +} + +/** + * Reference to a BO with its associated offset, used in the pack process. + */ +static inline struct vc5_cl_reloc +cl_address(struct vc5_bo *bo, uint32_t offset) +{ + struct vc5_cl_reloc reloc = { + .bo = bo, + .offset = offset, + }; + return reloc; +} + +uint32_t vc5_cl_ensure_space(struct vc5_cl *cl, uint32_t size, uint32_t align); +void vc5_cl_ensure_space_with_branch(struct vc5_cl *cl, uint32_t size); + +#define cl_packet_header(packet) V3D33_ ## packet ## _header +#define cl_packet_length(packet) V3D33_ ## packet ## _length +#define cl_packet_pack(packet) V3D33_ ## packet ## _pack +#define cl_packet_struct(packet) V3D33_ ## packet + +static inline void * +cl_get_emit_space(struct vc5_cl_out **cl, size_t size) +{ + void *addr = *cl; + cl_advance(cl, size); + return addr; +} + +/* Macro for setting up an emit of a CL struct. A temporary unpacked struct + * is created, which you get to set fields in of the form: + * + * cl_emit(bcl, FLAT_SHADE_FLAGS, flags) { + * .flags.flat_shade_flags = 1 << 2, + * } + * + * or default values only can be emitted with just: + * + * cl_emit(bcl, FLAT_SHADE_FLAGS, flags); + * + * The trick here is that we make a for loop that will execute the body + * (either the block or the ';' after the macro invocation) exactly once. + */ +#define cl_emit(cl, packet, name) \ + for (struct cl_packet_struct(packet) name = { \ + cl_packet_header(packet) \ + }, \ + *_loop_terminate = &name; \ + __builtin_expect(_loop_terminate != NULL, 1); \ + ({ \ + struct vc5_cl_out *cl_out = cl_start(cl); \ + cl_packet_pack(packet)(cl, (uint8_t *)cl_out, &name); \ + VG(VALGRIND_CHECK_MEM_IS_DEFINED(cl_out, \ + cl_packet_length(packet))); \ + cl_advance(&cl_out, cl_packet_length(packet)); \ + cl_end(cl, cl_out); \ + _loop_terminate = NULL; \ + })) \ + +#define cl_emit_prepacked(cl, packet) do { \ + memcpy((cl)->next, packet, sizeof(*packet)); \ + cl_advance(&(cl)->next, sizeof(*packet)); \ +} while (0) + +/** + * Helper function called by the XML-generated pack functions for filling in + * an address field in shader records. + * + * Since we have a private address space as of VC5, our BOs can have lifelong + * offsets, and all the kernel needs to know is which BOs need to be paged in + * for this exec. + */ +static inline void +cl_pack_emit_reloc(struct vc5_cl *cl, const struct vc5_cl_reloc *reloc) +{ + if (reloc->bo) + vc5_job_add_bo(cl->job, reloc->bo); +} + +#endif /* VC5_CL_H */ |