mirrors/qemu
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
bd983f68ac
qemu/target/hexagon/translate.h

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

245 lines
7.4 KiB
C
Raw Normal View History

Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
/*
Hexagon (target/hexagon) Analyze reads before writes We divide gen_analyze_funcs.py into 3 phases Declare the operands Analyze the register reads Analyze the register writes We also create special versions of ctx_log_*_read for new operands Check that the operand is written before the read This is a precursor to improving the analysis for short-circuiting the packet semantics in a subsequent commit Signed-off-by: Taylor Simpson <ltaylorsimpson@gmail.com> Reviewed-by: Brian Cain <bcain@quicinc.com> Message-Id: <20240201103340.119081-2-ltaylorsimpson@gmail.com> Signed-off-by: Brian Cain <bcain@quicinc.com>
2024-02-01 13:33:38 +03:00
* Copyright(c) 2019-2024 Qualcomm Innovation Center, Inc. All Rights Reserved.
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, see <http://www.gnu.org/licenses/>.
*/
#ifndef HEXAGON_TRANSLATE_H
#define HEXAGON_TRANSLATE_H
#include "qemu/bitmap.h"
exec/exec-all: Move 'qemu/log.h' include in units requiring it Many files use "qemu/log.h" declarations but neglect to include it (they inherit it via "exec/exec-all.h"). "exec/exec-all.h" is a core component and shouldn't be used that way. Move the "qemu/log.h" inclusion locally to each unit requiring it. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: Thomas Huth <thuth@redhat.com> Message-Id: <20220207082756.82600-10-f4bug@amsat.org> Signed-off-by: Thomas Huth <thuth@redhat.com>
2022-02-07 11:27:56 +03:00
#include "qemu/log.h"
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
#include "cpu.h"
#include "exec/translator.h"
#include "tcg/tcg-op.h"
Hexagon (target/hexagon) Add pkt and insn to DisasContext This enables us to reduce the number of parameters to many functions In particular, the generated functions previously took all 3 as arguments Not only does this simplify the code, it improves the translation time Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <20221108162906.3166-2-tsimpson@quicinc.com>
2022-11-08 19:28:56 +03:00
#include "insn.h"
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
#include "internal.h"
typedef struct DisasContext {
DisasContextBase base;
Hexagon (target/hexagon) Add pkt and insn to DisasContext This enables us to reduce the number of parameters to many functions In particular, the generated functions previously took all 3 as arguments Not only does this simplify the code, it improves the translation time Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <20221108162906.3166-2-tsimpson@quicinc.com>
2022-11-08 19:28:56 +03:00
Packet *pkt;
Insn *insn;
Hexagon (target/hexagon) Remove next_PC from runtime state The imported files don't properly mark all CONDEXEC instructions, so we add some logic to hex_common.py to add the attribute. Acked-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <20221108162906.3166-7-tsimpson@quicinc.com>
2022-11-08 19:29:01 +03:00
uint32_t next_PC;
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
uint32_t mem_idx;
uint32_t num_packets;
uint32_t num_insns;
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
uint32_t num_hvx_insns;
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
int reg_log[REG_WRITES_MAX];
int reg_log_idx;
DECLARE_BITMAP(regs_written, TOTAL_PER_THREAD_REGS);
Hexagon (target/hexagon) Analyze packet before generating TCG We create a new generator that creates an analyze_<tag> function for each instruction. Currently, these functions record the writes to R, P, and C registers by calling ctx_log_reg_write[_pair] or ctx_log_pred_write. During gen_start_packet, we invoke the analyze_<tag> function for each instruction in the packet, and we mark the implicit register and predicate writes. Doing the analysis up front has several advantages - We remove calls to ctx_log_* from gen_tcg_funcs.py and genptr.c - After the analysis is performed, we can initialize hex_new_value for each of the predicated assignments rather than during TCG generation for the instructions - This is a stepping stone for future work where the analysis will include the set of registers that are read. In cases where the packet doesn't have an overlap between the registers that are written and registers that are read, we can avoid the intermediate step of writing to hex_new_value. Note that other checks will also be needed (e.g., no instructions can raise an exception). Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-6-tsimpson@quicinc.com>
2023-03-07 05:58:19 +03:00
DECLARE_BITMAP(predicated_regs, TOTAL_PER_THREAD_REGS);
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
int preg_log[PRED_WRITES_MAX];
int preg_log_idx;
Hexagon (target/hexagon) decide if pred has been written at TCG gen time Multiple writes to the same preg are and'ed together. Rather than generating a runtime check, we can determine at TCG generation time if the predicate has previously been written in the packet. Test added to tests/tcg/hexagon/misc.c Suggested-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <1617930474-31979-7-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-04-09 04:07:34 +03:00
DECLARE_BITMAP(pregs_written, NUM_PREGS);
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
uint8_t store_width[STORES_MAX];
Hexagon (target/hexagon) change variables from int to bool when appropriate Suggested-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <1617930474-31979-8-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-04-09 04:07:35 +03:00
bool s1_store_processed;
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
int future_vregs_idx;
int future_vregs_num[VECTOR_TEMPS_MAX];
int tmp_vregs_idx;
int tmp_vregs_num[VECTOR_TEMPS_MAX];
int vreg_log[NUM_VREGS];
int vreg_log_idx;
DECLARE_BITMAP(vregs_updated_tmp, NUM_VREGS);
DECLARE_BITMAP(vregs_updated, NUM_VREGS);
DECLARE_BITMAP(vregs_select, NUM_VREGS);
Hexagon (target/hexagon) Analyze packet for HVX Extend the analyze_<tag> functions for HVX vector and predicate writes Remove calls to ctx_log_vreg_write[_pair] from gen_tcg_funcs.py During gen_start_packet, reload the predicated HVX registers into fugure_VRegs and tmp_VRegs Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-8-tsimpson@quicinc.com>
2023-03-07 05:58:21 +03:00
DECLARE_BITMAP(predicated_future_vregs, NUM_VREGS);
DECLARE_BITMAP(predicated_tmp_vregs, NUM_VREGS);
Hexagon (target/hexagon) Mark registers as read during packet analysis Have gen_analyze_funcs mark the registers that are read by the instruction. We also mark the implicit reads using instruction attributes. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-11-tsimpson@quicinc.com>
2023-04-28 02:00:01 +03:00
DECLARE_BITMAP(vregs_read, NUM_VREGS);
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
int qreg_log[NUM_QREGS];
int qreg_log_idx;
Hexagon (target/hexagon) Mark registers as read during packet analysis Have gen_analyze_funcs mark the registers that are read by the instruction. We also mark the implicit reads using instruction attributes. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-11-tsimpson@quicinc.com>
2023-04-28 02:00:01 +03:00
DECLARE_BITMAP(qregs_read, NUM_QREGS);
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
bool pre_commit;
Hexagon (target/hexagon) Short-circuit packet register writes In certain cases, we can avoid the overhead of writing to hex_new_value and write directly to hex_gpr. We add need_commit field to DisasContext indicating if the end-of-packet commit is needed. If it is not needed, get_result_gpr() and get_result_gpr_pair() can return hex_gpr. We pass the ctx->need_commit to helpers when needed. Finally, we can early-exit from gen_reg_writes during packet commit. There are a few instructions whose semantics write to the result before reading all the inputs. Therefore, the idef-parser generated code is incompatible with short-circuit. We tell idef-parser to skip them. For debugging purposes, we add a cpu property to turn off short-circuit. When the short-circuit property is false, we skip the analysis and force the end-of-packet commit. Here's a simple example of the TCG generated for 0x004000b4: 0x7800c020 { R0 = #0x1 } BEFORE: ---- 004000b4 movi_i32 new_r0,$0x1 mov_i32 r0,new_r0 AFTER: ---- 004000b4 movi_i32 r0,$0x1 This patch reintroduces a use of check_for_attrib, so we remove the G_GNUC_UNUSED added earlier in this series. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Brian Cain <bcain@quicinc.com> Message-Id: <20230427230012.3800327-12-tsimpson@quicinc.com>
2023-04-28 02:00:02 +03:00
bool need_commit;
Hexagon (target/hexagon) Use direct block chaining for direct jump/branch Direct block chaining is documented here https://qemu.readthedocs.io/en/latest/devel/tcg.html#direct-block-chaining Recall that Hexagon allows packets with multiple jumps where only the first one with a true predicate will actually jump. We can use tcg_gen_goto_tb/tcg_gen_exit_tb when the packet contains a single PC-relative branch or jump. If not, we use tcg_gen_lookup_and_goto_ptr. We add the following to DisasContext in order to delay the branching until the end of packet commit (in gen_end_tb) branch_cond The TCGCond condition under which the branch is taken When branch_cond == TCG_COND_NEVER, there isn't a single direct branch in this packet. When branch_cond != TCG_COND_ALWAYS, the value is in hex_branch_taken branch_dest The destination of the branch Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <20221108162906.3166-11-tsimpson@quicinc.com>
2022-11-08 19:29:05 +03:00
TCGCond branch_cond;
target_ulong branch_dest;
Hexagon (target/hexagon) Use direct block chaining for tight loops Direct block chaining is documented here https://qemu.readthedocs.io/en/latest/devel/tcg.html#direct-block-chaining Hexagon inner loops end with the endloop0 instruction To go back to the beginning of the loop, this instructions writes to PC from register SA0 (start address 0). To use direct block chaining, we have to assign PC with a constant value. So, we specialize the code generation when the start of the translation block is equal to SA0. When this is the case, we defer the compare/branch from endloop0 to gen_end_tb. When this is done, we can assign the start address of the TB to PC. Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <20221108162906.3166-12-tsimpson@quicinc.com>
2022-11-10 20:49:35 +03:00
bool is_tight_loop;
Hexagon (target/hexagon) Short-circuit packet register writes In certain cases, we can avoid the overhead of writing to hex_new_value and write directly to hex_gpr. We add need_commit field to DisasContext indicating if the end-of-packet commit is needed. If it is not needed, get_result_gpr() and get_result_gpr_pair() can return hex_gpr. We pass the ctx->need_commit to helpers when needed. Finally, we can early-exit from gen_reg_writes during packet commit. There are a few instructions whose semantics write to the result before reading all the inputs. Therefore, the idef-parser generated code is incompatible with short-circuit. We tell idef-parser to skip them. For debugging purposes, we add a cpu property to turn off short-circuit. When the short-circuit property is false, we skip the analysis and force the end-of-packet commit. Here's a simple example of the TCG generated for 0x004000b4: 0x7800c020 { R0 = #0x1 } BEFORE: ---- 004000b4 movi_i32 new_r0,$0x1 mov_i32 r0,new_r0 AFTER: ---- 004000b4 movi_i32 r0,$0x1 This patch reintroduces a use of check_for_attrib, so we remove the G_GNUC_UNUSED added earlier in this series. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Brian Cain <bcain@quicinc.com> Message-Id: <20230427230012.3800327-12-tsimpson@quicinc.com>
2023-04-28 02:00:02 +03:00
bool short_circuit;
Hexagon (target/hexagon) Short-circuit more HVX single instruction packets The generated helpers for HVX use pass-by-reference, so they can't short-circuit when the reads/writes overlap. The instructions with overrides are OK because they use tcg_gen_gvec_*. We add a flag has_hvx_helper to DisasContext and extend gen_analyze_funcs to set the flag when the instruction is an HVX instruction with a generated helper. We add an override for V6_vcombine so that it can be short-circuited along with a test case in tests/tcg/hexagon/hvx_misc.c Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-15-tsimpson@quicinc.com>
2023-04-28 02:00:05 +03:00
bool has_hvx_helper;
Hexagon (target/hexagon) Enable more short-circuit packets (scalar core) Look for read-after-write instead of overlap of reads and writes Here is an example with overalp but no read-after-write: 0x000200fc: 0x38103876 { R0 = add(R0,R1); R6 = add(R6,R7) } BEFORE: ---- 00000000000200fc mov_i32 loc2,$0x0 mov_i32 loc2,r0 add_i32 loc3,loc2,r1 mov_i32 loc2,loc3 mov_i32 loc4,$0x0 mov_i32 loc4,r6 add_i32 loc5,loc4,r7 mov_i32 loc4,loc5 mov_i32 r0,loc2 mov_i32 r6,loc4 AFTER: ---- 00000000000200fc add_i32 loc2,r0,r1 mov_i32 r0,loc2 add_i32 loc3,r6,r7 mov_i32 r6,loc3 We can also short-circuit packets with .new values by reading from the real destination instead of the temporary. 0x00020100: 0x78005ff3 { R19 = #0xff 0x00020104: 0x2002e204 if (cmp.eq(N19.new,R2)) jump:t PC+8 } BEFORE: ---- 0000000000020100 mov_i32 pc,$0x20108 mov_i32 loc8,$0x0 mov_i32 loc8,$0xff setcond_i32 loc10,loc8,r2,eq mov_i32 loc6,loc10 mov_i32 r19,loc8 add_i32 pkt_cnt,pkt_cnt,$0x2 add_i32 insn_cnt,insn_cnt,$0x4 brcond_i32 loc6,$0x0,eq,$L1 goto_tb $0x0 mov_i32 pc,$0x20108 exit_tb $0x7fbb54000040 set_label $L1 goto_tb $0x1 exit_tb $0x7fbb54000041 set_label $L0 exit_tb $0x7fbb54000043 AFTER: ---- 0000000000020100 mov_i32 pc,$0x20108 mov_i32 r19,$0xff setcond_i32 loc7,r19,r2,eq mov_i32 loc4,loc7 add_i32 pkt_cnt,pkt_cnt,$0x2 add_i32 insn_cnt,insn_cnt,$0x4 brcond_i32 loc4,$0x0,eq,$L1 goto_tb $0x0 mov_i32 pc,$0x20108 exit_tb $0x7f9764000040 set_label $L1 goto_tb $0x1 exit_tb $0x7f9764000041 set_label $L0 exit_tb $0x7f9764000043 Signed-off-by: Taylor Simpson <ltaylorsimpson@gmail.com> Reviewed-by: Brian Cain <bcain@quicinc.com> Message-Id: <20240201103340.119081-3-ltaylorsimpson@gmail.com> Signed-off-by: Brian Cain <bcain@quicinc.com>
2024-02-01 13:33:39 +03:00
bool read_after_write;
Hexagon (target/hexagon) Move new_value to DisasContext The new_value array in the CPUHexagonState is only used for bookkeeping within the translation of a packet. With recent changes that eliminate the need to free TCGv variables, these make more sense to be transient and kept in DisasContext. Suggested-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-18-tsimpson@quicinc.com>
2023-04-28 02:00:08 +03:00
TCGv new_value[TOTAL_PER_THREAD_REGS];
Hexagon (target/hexagon) Move new_pred_value to DisasContext The new_pred_value array in the CPUHexagonState is only used for bookkeeping within the translation of a packet. With recent changes that eliminate the need to free TCGv variables, these make more sense to be transient and kept in DisasContext. Suggested-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-19-tsimpson@quicinc.com>
2023-04-28 02:00:09 +03:00
TCGv new_pred_value[NUM_PREGS];
Hexagon (target/hexagon) Move pred_written to DisasContext The pred_written variable in the CPUHexagonState is only used for bookkeeping within the translation of a packet. With recent changes that eliminate the need to free TCGv variables, these make more sense to be transient and kept in DisasContext. Suggested-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-20-tsimpson@quicinc.com>
2023-04-28 02:00:10 +03:00
TCGv pred_written;
Hexagon (target/hexagon) Move items to DisasContext The following items in the CPUHexagonState are only used for bookkeeping within the translation of a packet. With recent changes that eliminate the need to free TCGv variables, these make more sense to be transient and kept in DisasContext. The following items are moved dczero_addr branch_taken this_PC Suggested-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-22-tsimpson@quicinc.com>
2023-04-28 02:00:12 +03:00
TCGv branch_taken;
TCGv dczero_addr;
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
} DisasContext;
Hexagon (target/hexagon) Analyze reads before writes We divide gen_analyze_funcs.py into 3 phases Declare the operands Analyze the register reads Analyze the register writes We also create special versions of ctx_log_*_read for new operands Check that the operand is written before the read This is a precursor to improving the analysis for short-circuiting the packet semantics in a subsequent commit Signed-off-by: Taylor Simpson <ltaylorsimpson@gmail.com> Reviewed-by: Brian Cain <bcain@quicinc.com> Message-Id: <20240201103340.119081-2-ltaylorsimpson@gmail.com> Signed-off-by: Brian Cain <bcain@quicinc.com>
2024-02-01 13:33:38 +03:00
bool is_gather_store_insn(DisasContext *ctx);
Hexagon (target/hexagon) Analyze packet before generating TCG We create a new generator that creates an analyze_<tag> function for each instruction. Currently, these functions record the writes to R, P, and C registers by calling ctx_log_reg_write[_pair] or ctx_log_pred_write. During gen_start_packet, we invoke the analyze_<tag> function for each instruction in the packet, and we mark the implicit register and predicate writes. Doing the analysis up front has several advantages - We remove calls to ctx_log_* from gen_tcg_funcs.py and genptr.c - After the analysis is performed, we can initialize hex_new_value for each of the predicated assignments rather than during TCG generation for the instructions - This is a stepping stone for future work where the analysis will include the set of registers that are read. In cases where the packet doesn't have an overlap between the registers that are written and registers that are read, we can avoid the intermediate step of writing to hex_new_value. Note that other checks will also be needed (e.g., no instructions can raise an exception). Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-6-tsimpson@quicinc.com>
2023-03-07 05:58:19 +03:00
static inline void ctx_log_pred_write(DisasContext *ctx, int pnum)
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
{
Hexagon (target/hexagon) Analyze packet before generating TCG We create a new generator that creates an analyze_<tag> function for each instruction. Currently, these functions record the writes to R, P, and C registers by calling ctx_log_reg_write[_pair] or ctx_log_pred_write. During gen_start_packet, we invoke the analyze_<tag> function for each instruction in the packet, and we mark the implicit register and predicate writes. Doing the analysis up front has several advantages - We remove calls to ctx_log_* from gen_tcg_funcs.py and genptr.c - After the analysis is performed, we can initialize hex_new_value for each of the predicated assignments rather than during TCG generation for the instructions - This is a stepping stone for future work where the analysis will include the set of registers that are read. In cases where the packet doesn't have an overlap between the registers that are written and registers that are read, we can avoid the intermediate step of writing to hex_new_value. Note that other checks will also be needed (e.g., no instructions can raise an exception). Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-6-tsimpson@quicinc.com>
2023-03-07 05:58:19 +03:00
if (!test_bit(pnum, ctx->pregs_written)) {
ctx->preg_log[ctx->preg_log_idx] = pnum;
ctx->preg_log_idx++;
set_bit(pnum, ctx->pregs_written);
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
}
}
Hexagon (target/hexagon) Mark registers as read during packet analysis Have gen_analyze_funcs mark the registers that are read by the instruction. We also mark the implicit reads using instruction attributes. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-11-tsimpson@quicinc.com>
2023-04-28 02:00:01 +03:00
static inline void ctx_log_pred_read(DisasContext *ctx, int pnum)
{
Hexagon (target/hexagon) Enable more short-circuit packets (scalar core) Look for read-after-write instead of overlap of reads and writes Here is an example with overalp but no read-after-write: 0x000200fc: 0x38103876 { R0 = add(R0,R1); R6 = add(R6,R7) } BEFORE: ---- 00000000000200fc mov_i32 loc2,$0x0 mov_i32 loc2,r0 add_i32 loc3,loc2,r1 mov_i32 loc2,loc3 mov_i32 loc4,$0x0 mov_i32 loc4,r6 add_i32 loc5,loc4,r7 mov_i32 loc4,loc5 mov_i32 r0,loc2 mov_i32 r6,loc4 AFTER: ---- 00000000000200fc add_i32 loc2,r0,r1 mov_i32 r0,loc2 add_i32 loc3,r6,r7 mov_i32 r6,loc3 We can also short-circuit packets with .new values by reading from the real destination instead of the temporary. 0x00020100: 0x78005ff3 { R19 = #0xff 0x00020104: 0x2002e204 if (cmp.eq(N19.new,R2)) jump:t PC+8 } BEFORE: ---- 0000000000020100 mov_i32 pc,$0x20108 mov_i32 loc8,$0x0 mov_i32 loc8,$0xff setcond_i32 loc10,loc8,r2,eq mov_i32 loc6,loc10 mov_i32 r19,loc8 add_i32 pkt_cnt,pkt_cnt,$0x2 add_i32 insn_cnt,insn_cnt,$0x4 brcond_i32 loc6,$0x0,eq,$L1 goto_tb $0x0 mov_i32 pc,$0x20108 exit_tb $0x7fbb54000040 set_label $L1 goto_tb $0x1 exit_tb $0x7fbb54000041 set_label $L0 exit_tb $0x7fbb54000043 AFTER: ---- 0000000000020100 mov_i32 pc,$0x20108 mov_i32 r19,$0xff setcond_i32 loc7,r19,r2,eq mov_i32 loc4,loc7 add_i32 pkt_cnt,pkt_cnt,$0x2 add_i32 insn_cnt,insn_cnt,$0x4 brcond_i32 loc4,$0x0,eq,$L1 goto_tb $0x0 mov_i32 pc,$0x20108 exit_tb $0x7f9764000040 set_label $L1 goto_tb $0x1 exit_tb $0x7f9764000041 set_label $L0 exit_tb $0x7f9764000043 Signed-off-by: Taylor Simpson <ltaylorsimpson@gmail.com> Reviewed-by: Brian Cain <bcain@quicinc.com> Message-Id: <20240201103340.119081-3-ltaylorsimpson@gmail.com> Signed-off-by: Brian Cain <bcain@quicinc.com>
2024-02-01 13:33:39 +03:00
if (test_bit(pnum, ctx->pregs_written)) {
ctx->read_after_write = true;
}
Hexagon (target/hexagon) Mark registers as read during packet analysis Have gen_analyze_funcs mark the registers that are read by the instruction. We also mark the implicit reads using instruction attributes. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-11-tsimpson@quicinc.com>
2023-04-28 02:00:01 +03:00
}
Hexagon (target/hexagon) Analyze reads before writes We divide gen_analyze_funcs.py into 3 phases Declare the operands Analyze the register reads Analyze the register writes We also create special versions of ctx_log_*_read for new operands Check that the operand is written before the read This is a precursor to improving the analysis for short-circuiting the packet semantics in a subsequent commit Signed-off-by: Taylor Simpson <ltaylorsimpson@gmail.com> Reviewed-by: Brian Cain <bcain@quicinc.com> Message-Id: <20240201103340.119081-2-ltaylorsimpson@gmail.com> Signed-off-by: Brian Cain <bcain@quicinc.com>
2024-02-01 13:33:38 +03:00
static inline void ctx_log_pred_read_new(DisasContext *ctx, int pnum)
{
g_assert(test_bit(pnum, ctx->pregs_written));
}
Hexagon (target/hexagon) Analyze packet before generating TCG We create a new generator that creates an analyze_<tag> function for each instruction. Currently, these functions record the writes to R, P, and C registers by calling ctx_log_reg_write[_pair] or ctx_log_pred_write. During gen_start_packet, we invoke the analyze_<tag> function for each instruction in the packet, and we mark the implicit register and predicate writes. Doing the analysis up front has several advantages - We remove calls to ctx_log_* from gen_tcg_funcs.py and genptr.c - After the analysis is performed, we can initialize hex_new_value for each of the predicated assignments rather than during TCG generation for the instructions - This is a stepping stone for future work where the analysis will include the set of registers that are read. In cases where the packet doesn't have an overlap between the registers that are written and registers that are read, we can avoid the intermediate step of writing to hex_new_value. Note that other checks will also be needed (e.g., no instructions can raise an exception). Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-6-tsimpson@quicinc.com>
2023-03-07 05:58:19 +03:00
static inline void ctx_log_reg_write(DisasContext *ctx, int rnum,
bool is_predicated)
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
{
Hexagon (target/hexagon) Analyze packet before generating TCG We create a new generator that creates an analyze_<tag> function for each instruction. Currently, these functions record the writes to R, P, and C registers by calling ctx_log_reg_write[_pair] or ctx_log_pred_write. During gen_start_packet, we invoke the analyze_<tag> function for each instruction in the packet, and we mark the implicit register and predicate writes. Doing the analysis up front has several advantages - We remove calls to ctx_log_* from gen_tcg_funcs.py and genptr.c - After the analysis is performed, we can initialize hex_new_value for each of the predicated assignments rather than during TCG generation for the instructions - This is a stepping stone for future work where the analysis will include the set of registers that are read. In cases where the packet doesn't have an overlap between the registers that are written and registers that are read, we can avoid the intermediate step of writing to hex_new_value. Note that other checks will also be needed (e.g., no instructions can raise an exception). Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-6-tsimpson@quicinc.com>
2023-03-07 05:58:19 +03:00
if (rnum == HEX_REG_P3_0_ALIASED) {
for (int i = 0; i < NUM_PREGS; i++) {
ctx_log_pred_write(ctx, i);
}
} else {
if (!test_bit(rnum, ctx->regs_written)) {
ctx->reg_log[ctx->reg_log_idx] = rnum;
ctx->reg_log_idx++;
set_bit(rnum, ctx->regs_written);
}
if (is_predicated) {
set_bit(rnum, ctx->predicated_regs);
}
}
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
}
Hexagon (target/hexagon) Analyze packet before generating TCG We create a new generator that creates an analyze_<tag> function for each instruction. Currently, these functions record the writes to R, P, and C registers by calling ctx_log_reg_write[_pair] or ctx_log_pred_write. During gen_start_packet, we invoke the analyze_<tag> function for each instruction in the packet, and we mark the implicit register and predicate writes. Doing the analysis up front has several advantages - We remove calls to ctx_log_* from gen_tcg_funcs.py and genptr.c - After the analysis is performed, we can initialize hex_new_value for each of the predicated assignments rather than during TCG generation for the instructions - This is a stepping stone for future work where the analysis will include the set of registers that are read. In cases where the packet doesn't have an overlap between the registers that are written and registers that are read, we can avoid the intermediate step of writing to hex_new_value. Note that other checks will also be needed (e.g., no instructions can raise an exception). Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-6-tsimpson@quicinc.com>
2023-03-07 05:58:19 +03:00
static inline void ctx_log_reg_write_pair(DisasContext *ctx, int rnum,
bool is_predicated)
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
{
Hexagon (target/hexagon) Analyze packet before generating TCG We create a new generator that creates an analyze_<tag> function for each instruction. Currently, these functions record the writes to R, P, and C registers by calling ctx_log_reg_write[_pair] or ctx_log_pred_write. During gen_start_packet, we invoke the analyze_<tag> function for each instruction in the packet, and we mark the implicit register and predicate writes. Doing the analysis up front has several advantages - We remove calls to ctx_log_* from gen_tcg_funcs.py and genptr.c - After the analysis is performed, we can initialize hex_new_value for each of the predicated assignments rather than during TCG generation for the instructions - This is a stepping stone for future work where the analysis will include the set of registers that are read. In cases where the packet doesn't have an overlap between the registers that are written and registers that are read, we can avoid the intermediate step of writing to hex_new_value. Note that other checks will also be needed (e.g., no instructions can raise an exception). Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-6-tsimpson@quicinc.com>
2023-03-07 05:58:19 +03:00
ctx_log_reg_write(ctx, rnum, is_predicated);
ctx_log_reg_write(ctx, rnum + 1, is_predicated);
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
}
Hexagon (target/hexagon) Mark registers as read during packet analysis Have gen_analyze_funcs mark the registers that are read by the instruction. We also mark the implicit reads using instruction attributes. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-11-tsimpson@quicinc.com>
2023-04-28 02:00:01 +03:00
static inline void ctx_log_reg_read(DisasContext *ctx, int rnum)
{
Hexagon (target/hexagon) Enable more short-circuit packets (scalar core) Look for read-after-write instead of overlap of reads and writes Here is an example with overalp but no read-after-write: 0x000200fc: 0x38103876 { R0 = add(R0,R1); R6 = add(R6,R7) } BEFORE: ---- 00000000000200fc mov_i32 loc2,$0x0 mov_i32 loc2,r0 add_i32 loc3,loc2,r1 mov_i32 loc2,loc3 mov_i32 loc4,$0x0 mov_i32 loc4,r6 add_i32 loc5,loc4,r7 mov_i32 loc4,loc5 mov_i32 r0,loc2 mov_i32 r6,loc4 AFTER: ---- 00000000000200fc add_i32 loc2,r0,r1 mov_i32 r0,loc2 add_i32 loc3,r6,r7 mov_i32 r6,loc3 We can also short-circuit packets with .new values by reading from the real destination instead of the temporary. 0x00020100: 0x78005ff3 { R19 = #0xff 0x00020104: 0x2002e204 if (cmp.eq(N19.new,R2)) jump:t PC+8 } BEFORE: ---- 0000000000020100 mov_i32 pc,$0x20108 mov_i32 loc8,$0x0 mov_i32 loc8,$0xff setcond_i32 loc10,loc8,r2,eq mov_i32 loc6,loc10 mov_i32 r19,loc8 add_i32 pkt_cnt,pkt_cnt,$0x2 add_i32 insn_cnt,insn_cnt,$0x4 brcond_i32 loc6,$0x0,eq,$L1 goto_tb $0x0 mov_i32 pc,$0x20108 exit_tb $0x7fbb54000040 set_label $L1 goto_tb $0x1 exit_tb $0x7fbb54000041 set_label $L0 exit_tb $0x7fbb54000043 AFTER: ---- 0000000000020100 mov_i32 pc,$0x20108 mov_i32 r19,$0xff setcond_i32 loc7,r19,r2,eq mov_i32 loc4,loc7 add_i32 pkt_cnt,pkt_cnt,$0x2 add_i32 insn_cnt,insn_cnt,$0x4 brcond_i32 loc4,$0x0,eq,$L1 goto_tb $0x0 mov_i32 pc,$0x20108 exit_tb $0x7f9764000040 set_label $L1 goto_tb $0x1 exit_tb $0x7f9764000041 set_label $L0 exit_tb $0x7f9764000043 Signed-off-by: Taylor Simpson <ltaylorsimpson@gmail.com> Reviewed-by: Brian Cain <bcain@quicinc.com> Message-Id: <20240201103340.119081-3-ltaylorsimpson@gmail.com> Signed-off-by: Brian Cain <bcain@quicinc.com>
2024-02-01 13:33:39 +03:00
if (test_bit(rnum, ctx->regs_written)) {
ctx->read_after_write = true;
}
Hexagon (target/hexagon) Mark registers as read during packet analysis Have gen_analyze_funcs mark the registers that are read by the instruction. We also mark the implicit reads using instruction attributes. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-11-tsimpson@quicinc.com>
2023-04-28 02:00:01 +03:00
}
Hexagon (target/hexagon) Analyze reads before writes We divide gen_analyze_funcs.py into 3 phases Declare the operands Analyze the register reads Analyze the register writes We also create special versions of ctx_log_*_read for new operands Check that the operand is written before the read This is a precursor to improving the analysis for short-circuiting the packet semantics in a subsequent commit Signed-off-by: Taylor Simpson <ltaylorsimpson@gmail.com> Reviewed-by: Brian Cain <bcain@quicinc.com> Message-Id: <20240201103340.119081-2-ltaylorsimpson@gmail.com> Signed-off-by: Brian Cain <bcain@quicinc.com>
2024-02-01 13:33:38 +03:00
static inline void ctx_log_reg_read_new(DisasContext *ctx, int rnum)
{
g_assert(test_bit(rnum, ctx->regs_written));
}
Hexagon (target/hexagon) Mark registers as read during packet analysis Have gen_analyze_funcs mark the registers that are read by the instruction. We also mark the implicit reads using instruction attributes. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-11-tsimpson@quicinc.com>
2023-04-28 02:00:01 +03:00
static inline void ctx_log_reg_read_pair(DisasContext *ctx, int rnum)
{
ctx_log_reg_read(ctx, rnum);
ctx_log_reg_read(ctx, rnum + 1);
}
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
intptr_t ctx_future_vreg_off(DisasContext *ctx, int regnum,
int num, bool alloc_ok);
intptr_t ctx_tmp_vreg_off(DisasContext *ctx, int regnum,
int num, bool alloc_ok);
static inline void ctx_log_vreg_write(DisasContext *ctx,
int rnum, VRegWriteType type,
bool is_predicated)
{
if (type != EXT_TMP) {
Hexagon (target/hexagon) Improve code gen for predicated HVX instructions The following improvements are made for predicated HVX instructions During gen_commit_hvx, unconditionally move the "new" value into the dest Don't set slot_cancelled Remove runtime bookkeeping of which registers were updated Reduce the cases where gen_log_vreg_write[_pair] is called It's only needed for special operands VxxV and VyV Remove gen_log_qreg_write Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-15-tsimpson@quicinc.com>
2023-03-07 05:58:28 +03:00
if (!test_bit(rnum, ctx->vregs_updated)) {
ctx->vreg_log[ctx->vreg_log_idx] = rnum;
ctx->vreg_log_idx++;
set_bit(rnum, ctx->vregs_updated);
}
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
set_bit(rnum, ctx->vregs_updated);
Hexagon (target/hexagon) Analyze packet for HVX Extend the analyze_<tag> functions for HVX vector and predicate writes Remove calls to ctx_log_vreg_write[_pair] from gen_tcg_funcs.py During gen_start_packet, reload the predicated HVX registers into fugure_VRegs and tmp_VRegs Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-8-tsimpson@quicinc.com>
2023-03-07 05:58:21 +03:00
if (is_predicated) {
set_bit(rnum, ctx->predicated_future_vregs);
}
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
}
if (type == EXT_NEW) {
set_bit(rnum, ctx->vregs_select);
}
if (type == EXT_TMP) {
set_bit(rnum, ctx->vregs_updated_tmp);
Hexagon (target/hexagon) Analyze packet for HVX Extend the analyze_<tag> functions for HVX vector and predicate writes Remove calls to ctx_log_vreg_write[_pair] from gen_tcg_funcs.py During gen_start_packet, reload the predicated HVX registers into fugure_VRegs and tmp_VRegs Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-8-tsimpson@quicinc.com>
2023-03-07 05:58:21 +03:00
if (is_predicated) {
set_bit(rnum, ctx->predicated_tmp_vregs);
}
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
}
}
static inline void ctx_log_vreg_write_pair(DisasContext *ctx,
int rnum, VRegWriteType type,
bool is_predicated)
{
ctx_log_vreg_write(ctx, rnum ^ 0, type, is_predicated);
ctx_log_vreg_write(ctx, rnum ^ 1, type, is_predicated);
}
Hexagon (target/hexagon) Mark registers as read during packet analysis Have gen_analyze_funcs mark the registers that are read by the instruction. We also mark the implicit reads using instruction attributes. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-11-tsimpson@quicinc.com>
2023-04-28 02:00:01 +03:00
static inline void ctx_log_vreg_read(DisasContext *ctx, int rnum)
{
set_bit(rnum, ctx->vregs_read);
}
Hexagon (target/hexagon) Analyze reads before writes We divide gen_analyze_funcs.py into 3 phases Declare the operands Analyze the register reads Analyze the register writes We also create special versions of ctx_log_*_read for new operands Check that the operand is written before the read This is a precursor to improving the analysis for short-circuiting the packet semantics in a subsequent commit Signed-off-by: Taylor Simpson <ltaylorsimpson@gmail.com> Reviewed-by: Brian Cain <bcain@quicinc.com> Message-Id: <20240201103340.119081-2-ltaylorsimpson@gmail.com> Signed-off-by: Brian Cain <bcain@quicinc.com>
2024-02-01 13:33:38 +03:00
static inline void ctx_log_vreg_read_new(DisasContext *ctx, int rnum)
{
g_assert(is_gather_store_insn(ctx) ||
test_bit(rnum, ctx->vregs_updated) ||
test_bit(rnum, ctx->vregs_select) ||
test_bit(rnum, ctx->vregs_updated_tmp));
set_bit(rnum, ctx->vregs_read);
}
Hexagon (target/hexagon) Mark registers as read during packet analysis Have gen_analyze_funcs mark the registers that are read by the instruction. We also mark the implicit reads using instruction attributes. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-11-tsimpson@quicinc.com>
2023-04-28 02:00:01 +03:00
static inline void ctx_log_vreg_read_pair(DisasContext *ctx, int rnum)
{
ctx_log_vreg_read(ctx, rnum ^ 0);
ctx_log_vreg_read(ctx, rnum ^ 1);
}
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
static inline void ctx_log_qreg_write(DisasContext *ctx,
Hexagon (target/hexagon) Improve code gen for predicated HVX instructions The following improvements are made for predicated HVX instructions During gen_commit_hvx, unconditionally move the "new" value into the dest Don't set slot_cancelled Remove runtime bookkeeping of which registers were updated Reduce the cases where gen_log_vreg_write[_pair] is called It's only needed for special operands VxxV and VyV Remove gen_log_qreg_write Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-15-tsimpson@quicinc.com>
2023-03-07 05:58:28 +03:00
int rnum)
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
{
ctx->qreg_log[ctx->qreg_log_idx] = rnum;
ctx->qreg_log_idx++;
}
Hexagon (target/hexagon) Mark registers as read during packet analysis Have gen_analyze_funcs mark the registers that are read by the instruction. We also mark the implicit reads using instruction attributes. Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-11-tsimpson@quicinc.com>
2023-04-28 02:00:01 +03:00
static inline void ctx_log_qreg_read(DisasContext *ctx, int qnum)
{
set_bit(qnum, ctx->qregs_read);
}
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
extern TCGv hex_gpr[TOTAL_PER_THREAD_REGS];
extern TCGv hex_pred[NUM_PREGS];
extern TCGv hex_slot_cancelled;
Hexagon (target/hexagon) Make special new_value for USR Precursor to moving new_value from the global state to DisasContext USR will need to stay in the global state because some helpers will set it's value Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20230427230012.3800327-17-tsimpson@quicinc.com>
2023-04-28 02:00:07 +03:00
extern TCGv hex_new_value_usr;
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
extern TCGv hex_reg_written[TOTAL_PER_THREAD_REGS];
extern TCGv hex_store_addr[STORES_MAX];
extern TCGv hex_store_width[STORES_MAX];
extern TCGv hex_store_val32[STORES_MAX];
extern TCGv_i64 hex_store_val64[STORES_MAX];
extern TCGv hex_llsc_addr;
extern TCGv hex_llsc_val;
extern TCGv_i64 hex_llsc_val_i64;
Hexagon HVX (target/hexagon) TCG generation Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
2021-09-30 22:29:00 +03:00
extern TCGv hex_vstore_addr[VSTORES_MAX];
extern TCGv hex_vstore_size[VSTORES_MAX];
extern TCGv hex_vstore_pending[VSTORES_MAX];
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
Hexagon (target/hexagon) Add pkt and insn to DisasContext This enables us to reduce the number of parameters to many functions In particular, the generated functions previously took all 3 as arguments Not only does this simplify the code, it improves the translation time Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <20221108162906.3166-2-tsimpson@quicinc.com>
2022-11-08 19:28:56 +03:00
void process_store(DisasContext *ctx, int slot_num);
Hexagon (target/hexagon) Reduce manipulation of slot_cancelled We only need to track slot for predicated stores and predicated HVX instructions. Add arguments to the probe helper functions to indicate if the slot is predicated. Here is a simple example of the differences in the TCG code generated: IN: 0x00400094: 0xf900c102 { if (P0) R2 = and(R0,R1) } BEFORE ---- 00400094 mov_i32 slot_cancelled,$0x0 mov_i32 new_r2,r2 and_i32 tmp0,p0,$0x1 brcond_i32 tmp0,$0x0,eq,$L1 and_i32 tmp0,r0,r1 mov_i32 new_r2,tmp0 br $L2 set_label $L1 or_i32 slot_cancelled,slot_cancelled,$0x8 set_label $L2 mov_i32 r2,new_r2 AFTER ---- 00400094 mov_i32 new_r2,r2 and_i32 tmp0,p0,$0x1 brcond_i32 tmp0,$0x0,eq,$L1 and_i32 tmp0,r0,r1 mov_i32 new_r2,tmp0 br $L2 set_label $L1 set_label $L2 mov_i32 r2,new_r2 Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-14-tsimpson@quicinc.com>
2023-03-07 05:58:27 +03:00
FIELD(PROBE_PKT_SCALAR_STORE_S0, MMU_IDX, 0, 2)
FIELD(PROBE_PKT_SCALAR_STORE_S0, IS_PREDICATED, 2, 1)
FIELD(PROBE_PKT_SCALAR_HVX_STORES, HAS_ST0, 0, 1)
FIELD(PROBE_PKT_SCALAR_HVX_STORES, HAS_ST1, 1, 1)
FIELD(PROBE_PKT_SCALAR_HVX_STORES, HAS_HVX_STORES, 2, 1)
FIELD(PROBE_PKT_SCALAR_HVX_STORES, S0_IS_PRED, 3, 1)
FIELD(PROBE_PKT_SCALAR_HVX_STORES, S1_IS_PRED, 4, 1)
Hexagon (target/hexagon) Merge arguments to probe_pkt_scalar_hvx_stores Reducing the number of arguments reduces the overhead of the helper call Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-Id: <20230405164211.30015-2-tsimpson@quicinc.com>
2023-04-05 19:42:10 +03:00
FIELD(PROBE_PKT_SCALAR_HVX_STORES, MMU_IDX, 5, 2)
Hexagon (target/hexagon) Reduce manipulation of slot_cancelled We only need to track slot for predicated stores and predicated HVX instructions. Add arguments to the probe helper functions to indicate if the slot is predicated. Here is a simple example of the differences in the TCG code generated: IN: 0x00400094: 0xf900c102 { if (P0) R2 = and(R0,R1) } BEFORE ---- 00400094 mov_i32 slot_cancelled,$0x0 mov_i32 new_r2,r2 and_i32 tmp0,p0,$0x1 brcond_i32 tmp0,$0x0,eq,$L1 and_i32 tmp0,r0,r1 mov_i32 new_r2,tmp0 br $L2 set_label $L1 or_i32 slot_cancelled,slot_cancelled,$0x8 set_label $L2 mov_i32 r2,new_r2 AFTER ---- 00400094 mov_i32 new_r2,r2 and_i32 tmp0,p0,$0x1 brcond_i32 tmp0,$0x0,eq,$L1 and_i32 tmp0,r0,r1 mov_i32 new_r2,tmp0 br $L2 set_label $L1 set_label $L2 mov_i32 r2,new_r2 Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230307025828.1612809-14-tsimpson@quicinc.com>
2023-03-07 05:58:27 +03:00
Hexagon (target/hexagon) translation Read the instruction memory Create a packet data structure Generate TCG code for the start of the packet Invoke the generate function for each instruction Generate TCG code for the end of the packet Signed-off-by: Taylor Simpson <tsimpson@quicinc.com> Message-Id: <1612763186-18161-30-git-send-email-tsimpson@quicinc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2021-02-08 08:46:19 +03:00
#endif
Reference in New Issue Copy Permalink