OpenE2K/gcc
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[nvptx] Rename worker_bcast variables to oacc_bcast

Browse Source 
Rename worker_bcast variables to oacc_bcast, avoiding worker terminology.

Build and reg-tested on x86_64 with nvptx accelerator.

2018-12-19  Tom de Vries  <tdevries@suse.de>

	* config/nvptx/nvptx.c (worker_bcast_size): Rename as
	oacc_bcast_size.
	(worker_bcast_align): Rename as oacc_bcast_align.
	(worker_bcast_sym): Rename as oacc_bcast_sym.
	(nvptx_option_override): Update usage of oacc_bcast_*.
	(struct wcast_data_t): Rename as broadcast_data_t.
	(nvptx_gen_wcast): Update type of data argument and usage of
	oacc_bcast_align.
	(wprop_gen): Update type of data_ and usage of oacc_bcast_align.
	(nvptx_wpropagate): Update type of data and usage of
	oacc_bcast_{sym,size}.
	(nvptx_single): Update type of data and usage of oacc_bcast_size.
	(nvptx_file_end): Update usage of oacc_bcast_{sym,align,size}.

From-SVN: r267259
This commit is contained in:
Tom de Vries 2018-12-19 10:17:11 +00:00 committed by Tom de Vries
parent 1dcf26882b
commit 1ed57fb8b6
2 changed files with 46 additions and 29 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
gcc/ChangeLog
View File

@ -1,3 +1,19 @@
2018-12-19 Tom de Vries <tdevries@suse.de>
* config/nvptx/nvptx.c (worker_bcast_size): Rename as
oacc_bcast_size.
(worker_bcast_align): Rename as oacc_bcast_align.
(worker_bcast_sym): Rename as oacc_bcast_sym.
(nvptx_option_override): Update usage of oacc_bcast_*.
(struct wcast_data_t): Rename as broadcast_data_t.
(nvptx_gen_wcast): Update type of data argument and usage of
oacc_bcast_align.
(wprop_gen): Update type of data_ and usage of oacc_bcast_align.
(nvptx_wpropagate): Update type of data and usage of
oacc_bcast_{sym,size}.
(nvptx_single): Update type of data and usage of oacc_bcast_size.
(nvptx_file_end): Update usage of oacc_bcast_{sym,align,size}.
2018-12-19 Tom de Vries <tdevries@suse.de> 2018-12-19 Tom de Vries <tdevries@suse.de>
* config/nvptx/nvptx.md (nvptx_barsync): Add and handle operand. * config/nvptx/nvptx.md (nvptx_barsync): Add and handle operand.

59
gcc/config/nvptx/nvptx.c
View File

@ -127,14 +127,15 @@ struct tree_hasher : ggc_cache_ptr_hash<tree_node>
static GTY((cache)) hash_table<tree_hasher> *declared_fndecls_htab; static GTY((cache)) hash_table<tree_hasher> *declared_fndecls_htab;
static GTY((cache)) hash_table<tree_hasher> *needed_fndecls_htab; static GTY((cache)) hash_table<tree_hasher> *needed_fndecls_htab;
/* Buffer needed to broadcast across workers. This is used for both /* Buffer needed to broadcast across workers and vectors. This is
worker-neutering and worker broadcasting. It is shared by all used for both worker-neutering and worker broadcasting, and
functions emitted. The buffer is placed in shared memory. It'd be vector-neutering and boardcasting when vector_length > 32. It is
nice if PTX supported common blocks, because then this could be shared by all functions emitted. The buffer is placed in shared
shared across TUs (taking the largest size). */ memory. It'd be nice if PTX supported common blocks, because then
static unsigned worker_bcast_size; this could be shared across TUs (taking the largest size). */
static unsigned worker_bcast_align; static unsigned oacc_bcast_size;
static GTY(()) rtx worker_bcast_sym; static unsigned oacc_bcast_align;
static GTY(()) rtx oacc_bcast_sym;
/* Buffer needed for worker reductions. This has to be distinct from /* Buffer needed for worker reductions. This has to be distinct from
the worker broadcast array, as both may be live concurrently. */ the worker broadcast array, as both may be live concurrently. */
@ -207,9 +208,9 @@ nvptx_option_override (void)
declared_libfuncs_htab declared_libfuncs_htab
= hash_table<declared_libfunc_hasher>::create_ggc (17); = hash_table<declared_libfunc_hasher>::create_ggc (17);
worker_bcast_sym = gen_rtx_SYMBOL_REF (Pmode, "__worker_bcast"); oacc_bcast_sym = gen_rtx_SYMBOL_REF (Pmode, "__oacc_bcast");
SET_SYMBOL_DATA_AREA (worker_bcast_sym, DATA_AREA_SHARED); SET_SYMBOL_DATA_AREA (oacc_bcast_sym, DATA_AREA_SHARED);
worker_bcast_align = GET_MODE_ALIGNMENT (SImode) / BITS_PER_UNIT; oacc_bcast_align = GET_MODE_ALIGNMENT (SImode) / BITS_PER_UNIT;
worker_red_sym = gen_rtx_SYMBOL_REF (Pmode, "__worker_red"); worker_red_sym = gen_rtx_SYMBOL_REF (Pmode, "__worker_red");
SET_SYMBOL_DATA_AREA (worker_red_sym, DATA_AREA_SHARED); SET_SYMBOL_DATA_AREA (worker_red_sym, DATA_AREA_SHARED);
@ -1754,7 +1755,7 @@ nvptx_gen_vcast (rtx reg)
/* Structure used when generating a worker-level spill or fill. */ /* Structure used when generating a worker-level spill or fill. */
struct wcast_data_t struct broadcast_data_t
{ {
rtx base; /* Register holding base addr of buffer. */ rtx base; /* Register holding base addr of buffer. */
rtx ptr; /* Iteration var, if needed. */ rtx ptr; /* Iteration var, if needed. */
@ -1778,7 +1779,7 @@ enum propagate_mask
how many loop iterations will be executed (0 for not a loop). */ how many loop iterations will be executed (0 for not a loop). */
static rtx static rtx
nvptx_gen_wcast (rtx reg, propagate_mask pm, unsigned rep, wcast_data_t *data) nvptx_gen_wcast (rtx reg, propagate_mask pm, unsigned rep, broadcast_data_t *data)
{ {
rtx res; rtx res;
machine_mode mode = GET_MODE (reg); machine_mode mode = GET_MODE (reg);
@ -1808,8 +1809,8 @@ nvptx_gen_wcast (rtx reg, propagate_mask pm, unsigned rep, wcast_data_t *data)
{ {
unsigned align = GET_MODE_ALIGNMENT (mode) / BITS_PER_UNIT; unsigned align = GET_MODE_ALIGNMENT (mode) / BITS_PER_UNIT;
if (align > worker_bcast_align) if (align > oacc_bcast_align)
worker_bcast_align = align; oacc_bcast_align = align;
data->offset = (data->offset + align - 1) & ~(align - 1); data->offset = (data->offset + align - 1) & ~(align - 1);
addr = data->base; addr = data->base;
if (data->offset) if (data->offset)
@ -3914,15 +3915,15 @@ nvptx_vpropagate (bool is_call, basic_block block, rtx_insn *insn)
static rtx static rtx
wprop_gen (rtx reg, propagate_mask pm, unsigned rep, void *data_) wprop_gen (rtx reg, propagate_mask pm, unsigned rep, void *data_)
{ {
wcast_data_t *data = (wcast_data_t *)data_; broadcast_data_t *data = (broadcast_data_t *)data_;
if (pm & PM_loop_begin) if (pm & PM_loop_begin)
{ {
/* Starting a loop, initialize pointer. */ /* Starting a loop, initialize pointer. */
unsigned align = GET_MODE_ALIGNMENT (GET_MODE (reg)) / BITS_PER_UNIT; unsigned align = GET_MODE_ALIGNMENT (GET_MODE (reg)) / BITS_PER_UNIT;
if (align > worker_bcast_align) if (align > oacc_bcast_align)
worker_bcast_align = align; oacc_bcast_align = align;
data->offset = (data->offset + align - 1) & ~(align - 1); data->offset = (data->offset + align - 1) & ~(align - 1);
data->ptr = gen_reg_rtx (Pmode); data->ptr = gen_reg_rtx (Pmode);
@ -3947,7 +3948,7 @@ wprop_gen (rtx reg, propagate_mask pm, unsigned rep, void *data_)
static bool static bool
nvptx_wpropagate (bool pre_p, bool is_call, basic_block block, rtx_insn *insn) nvptx_wpropagate (bool pre_p, bool is_call, basic_block block, rtx_insn *insn)
{ {
wcast_data_t data; broadcast_data_t data;
data.base = gen_reg_rtx (Pmode); data.base = gen_reg_rtx (Pmode);
data.offset = 0; data.offset = 0;
@ -3959,11 +3960,11 @@ nvptx_wpropagate (bool pre_p, bool is_call, basic_block block, rtx_insn *insn)
if (data.offset) if (data.offset)
{ {
/* Stuff was emitted, initialize the base pointer now. */ /* Stuff was emitted, initialize the base pointer now. */
rtx init = gen_rtx_SET (data.base, worker_bcast_sym); rtx init = gen_rtx_SET (data.base, oacc_bcast_sym);
emit_insn_after (init, insn); emit_insn_after (init, insn);
if (worker_bcast_size < data.offset) if (oacc_bcast_size < data.offset)
worker_bcast_size = data.offset; oacc_bcast_size = data.offset;
} }
return empty; return empty;
} }
@ -4333,13 +4334,13 @@ nvptx_single (unsigned mask, basic_block from, basic_block to)
{ {
/* Includes worker mode, do spill & fill. By construction /* Includes worker mode, do spill & fill. By construction
we should never have worker mode only. */ we should never have worker mode only. */
wcast_data_t data; broadcast_data_t data;
data.base = worker_bcast_sym; data.base = oacc_bcast_sym;
data.ptr = 0; data.ptr = 0;
if (worker_bcast_size < GET_MODE_SIZE (SImode)) if (oacc_bcast_size < GET_MODE_SIZE (SImode))
worker_bcast_size = GET_MODE_SIZE (SImode); oacc_bcast_size = GET_MODE_SIZE (SImode);
data.offset = 0; data.offset = 0;
emit_insn_before (nvptx_gen_wcast (pvar, PM_read, 0, &data), emit_insn_before (nvptx_gen_wcast (pvar, PM_read, 0, &data),
@ -4968,9 +4969,9 @@ nvptx_file_end (void)
nvptx_record_fndecl (decl); nvptx_record_fndecl (decl);
fputs (func_decls.str().c_str(), asm_out_file); fputs (func_decls.str().c_str(), asm_out_file);
if (worker_bcast_size) if (oacc_bcast_size)
write_worker_buffer (asm_out_file, worker_bcast_sym, write_worker_buffer (asm_out_file, oacc_bcast_sym,
worker_bcast_align, worker_bcast_size); oacc_bcast_align, oacc_bcast_size);
if (worker_red_size) if (worker_red_size)
write_worker_buffer (asm_out_file, worker_red_sym, write_worker_buffer (asm_out_file, worker_red_sym,