diff options
author | Juan A. Suarez Romero <jasuarez@igalia.com> | 2016-07-18 07:17:39 +0000 |
---|---|---|
committer | Francisco Jerez <currojerez@riseup.net> | 2017-04-14 14:56:07 -0700 |
commit | 79af2563889098550de5d4a0955efbeb87a24565 (patch) | |
tree | 6481bb265d63dfdd9194be00663bdfb32b3a5752 | |
parent | fd349d29e43fa9c2227cbf649282810782ecf555 (diff) |
i965/fs: add helper to retrieve instruction execution type
The execution data size is the biggest type size of any instruction
operand.
We will use it to know if the instruction deals with DF, because in Ivy
we need to double the execution size and regioning parameters.
v2:
- Fix typo in commit log (Matt)
- Use static inline function instead of fs_inst's method (Curro).
- Define the result as a constant (Curro).
- Fix indentation (Matt).
- Add braces to nested control flow (Matt).
v3 (Curro):
- Add get_exec_type() and other auxiliary functions and use them to
calculate its size.
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
[ Francisco Jerez: Fix bogus 'type != BAD_FILE' check. Fix deduced
execution type for integer vector types. Take destination type as
execution type where there is no valid source. Assert-fail if the
deduced execution type is byte. Move into brw_ir_fs.h header for
consistency with the VEC4 back-end. ]
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
-rw-r--r-- | src/intel/compiler/brw_fs.cpp | 6 | ||||
-rw-r--r-- | src/intel/compiler/brw_ir_fs.h | 33 | ||||
-rw-r--r-- | src/intel/compiler/brw_reg.h | 30 |
3 files changed, 64 insertions, 5 deletions
diff --git a/src/intel/compiler/brw_fs.cpp b/src/intel/compiler/brw_fs.cpp index 9dc21ac5e3..3fc7ae4894 100644 --- a/src/intel/compiler/brw_fs.cpp +++ b/src/intel/compiler/brw_fs.cpp @@ -4585,11 +4585,7 @@ get_fpu_lowered_simd_width(const struct gen_device_info *devinfo, !inst->force_writemask_all) { const unsigned channels_per_grf = inst->exec_size / DIV_ROUND_UP(inst->size_written, REG_SIZE); - unsigned exec_type_size = 0; - for (int i = 0; i < inst->sources; i++) { - if (inst->src[i].file != BAD_FILE) - exec_type_size = MAX2(exec_type_size, type_sz(inst->src[i].type)); - } + const unsigned exec_type_size = get_exec_type_size(inst); assert(exec_type_size); /* The hardware shifts exactly 8 channels per compressed half of the diff --git a/src/intel/compiler/brw_ir_fs.h b/src/intel/compiler/brw_ir_fs.h index cad371248c..58beae0d1f 100644 --- a/src/intel/compiler/brw_ir_fs.h +++ b/src/intel/compiler/brw_ir_fs.h @@ -448,4 +448,37 @@ regs_read(const fs_inst *inst, unsigned i) reg_size); } +static inline enum brw_reg_type +get_exec_type(const fs_inst *inst) +{ + brw_reg_type exec_type = BRW_REGISTER_TYPE_B; + + for (int i = 0; i < inst->sources; i++) { + if (inst->src[i].file != BAD_FILE) { + const brw_reg_type t = get_exec_type(inst->src[i].type); + if (type_sz(t) > type_sz(exec_type)) + exec_type = t; + else if (type_sz(t) == type_sz(exec_type) && + brw_reg_type_is_floating_point(t)) + exec_type = t; + } + } + + if (exec_type == BRW_REGISTER_TYPE_B) + exec_type = inst->dst.type; + + /* TODO: We need to handle half-float conversions. */ + assert(exec_type != BRW_REGISTER_TYPE_HF || + inst->dst.type == BRW_REGISTER_TYPE_HF); + assert(exec_type != BRW_REGISTER_TYPE_B); + + return exec_type; +} + +static inline unsigned +get_exec_type_size(const fs_inst *inst) +{ + return type_sz(get_exec_type(inst)); +} + #endif diff --git a/src/intel/compiler/brw_reg.h b/src/intel/compiler/brw_reg.h index f8c3340e45..17a51fbd65 100644 --- a/src/intel/compiler/brw_reg.h +++ b/src/intel/compiler/brw_reg.h @@ -325,6 +325,36 @@ type_sz(unsigned type) } } +static inline bool +brw_reg_type_is_floating_point(enum brw_reg_type type) +{ + switch (type) { + case BRW_REGISTER_TYPE_F: + case BRW_REGISTER_TYPE_HF: + case BRW_REGISTER_TYPE_DF: + return true; + default: + return false; + } +} + +static inline enum brw_reg_type +get_exec_type(const enum brw_reg_type type) +{ + switch (type) { + case BRW_REGISTER_TYPE_B: + case BRW_REGISTER_TYPE_V: + return BRW_REGISTER_TYPE_W; + case BRW_REGISTER_TYPE_UB: + case BRW_REGISTER_TYPE_UV: + return BRW_REGISTER_TYPE_UW; + case BRW_REGISTER_TYPE_VF: + return BRW_REGISTER_TYPE_F; + default: + return type; + } +} + /** * Return an integer type of the requested size and signedness. */ |