haraldwolff/qemu-patch-raspberry4
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity

target-openrisc: Speed up move instruction

Browse source 
The OpenRISC architecture does not have its own move register
instruction. Instead it uses either "l.addi rd, r0, x" or
"l.ori rd, rs, 0" or "l.or rd, rx, r0"

The l.ori instruction is automatically optimized but not the l.addi instruction.
This patch optimizes for this special case.

Signed-off-by: Sebastian Macke <sebastian@macke.de>
Reviewed-by: Jia Liu <proljc@gmail.com>
Signed-off-by: Jia Liu <proljc@gmail.com>
This commit is contained in:
Sebastian Macke 2013-10-22 02:12:37 +02:00 committed by Jia Liu
parent 394cfa39ba
commit 352367e8bb
1 changed files with 27 additions and 23 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

50
target-openrisc/translate.c
View file

@ -904,29 +904,33 @@ static void dec_misc(DisasContext *dc, uint32_t insn)
case 0x27: /* l.addi */ case 0x27: /* l.addi */
LOG_DIS("l.addi r%d, r%d, %d\n", rd, ra, I16); LOG_DIS("l.addi r%d, r%d, %d\n", rd, ra, I16);
{ {
int lab = gen_new_label(); if (I16 == 0) {
TCGv_i64 ta = tcg_temp_new_i64(); tcg_gen_mov_tl(cpu_R[rd], cpu_R[ra]);
TCGv_i64 td = tcg_temp_local_new_i64(); } else {
TCGv_i32 res = tcg_temp_local_new_i32(); int lab = gen_new_label();
TCGv_i32 sr_ove = tcg_temp_local_new_i32(); TCGv_i64 ta = tcg_temp_new_i64();
tcg_gen_extu_i32_i64(ta, cpu_R[ra]); TCGv_i64 td = tcg_temp_local_new_i64();
tcg_gen_addi_i64(td, ta, sign_extend(I16, 16)); TCGv_i32 res = tcg_temp_local_new_i32();
tcg_gen_trunc_i64_i32(res, td); TCGv_i32 sr_ove = tcg_temp_local_new_i32();
tcg_gen_shri_i64(td, td, 32); tcg_gen_extu_i32_i64(ta, cpu_R[ra]);
tcg_gen_andi_i64(td, td, 0x3); tcg_gen_addi_i64(td, ta, sign_extend(I16, 16));
/* Jump to lab when no overflow. */ tcg_gen_trunc_i64_i32(res, td);
tcg_gen_brcondi_i64(TCG_COND_EQ, td, 0x0, lab); tcg_gen_shri_i64(td, td, 32);
tcg_gen_brcondi_i64(TCG_COND_EQ, td, 0x3, lab); tcg_gen_andi_i64(td, td, 0x3);
tcg_gen_ori_i32(cpu_sr, cpu_sr, (SR_OV | SR_CY)); /* Jump to lab when no overflow. */
tcg_gen_andi_i32(sr_ove, cpu_sr, SR_OVE); tcg_gen_brcondi_i64(TCG_COND_EQ, td, 0x0, lab);
tcg_gen_brcondi_i32(TCG_COND_NE, sr_ove, SR_OVE, lab); tcg_gen_brcondi_i64(TCG_COND_EQ, td, 0x3, lab);
gen_exception(dc, EXCP_RANGE); tcg_gen_ori_i32(cpu_sr, cpu_sr, (SR_OV | SR_CY));
gen_set_label(lab); tcg_gen_andi_i32(sr_ove, cpu_sr, SR_OVE);
tcg_gen_mov_i32(cpu_R[rd], res); tcg_gen_brcondi_i32(TCG_COND_NE, sr_ove, SR_OVE, lab);
tcg_temp_free_i64(ta); gen_exception(dc, EXCP_RANGE);
tcg_temp_free_i64(td); gen_set_label(lab);
tcg_temp_free_i32(res); tcg_gen_mov_i32(cpu_R[rd], res);
tcg_temp_free_i32(sr_ove); tcg_temp_free_i64(ta);
tcg_temp_free_i64(td);
tcg_temp_free_i32(res);
tcg_temp_free_i32(sr_ove);
}
} }
break; break;