[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PATCH 23/65] tcg: Allow an operand to be matching or a con
From: |
Richard Henderson |
Subject: |
[Qemu-devel] [PATCH 23/65] tcg: Allow an operand to be matching or a constant |
Date: |
Fri, 23 Dec 2016 20:00:00 -0800 |
This allows an output operand to match an input operand
only when the input operand needs a register.
Reviewed-by: Alex Bennée <address@hidden>
Signed-off-by: Richard Henderson <address@hidden>
---
tcg/README | 13 +++++++++----
tcg/tcg.c | 63 +++++++++++++++++++++++++++++++-------------------------------
2 files changed, 41 insertions(+), 35 deletions(-)
diff --git a/tcg/README b/tcg/README
index 065d9c2..6946b5b 100644
--- a/tcg/README
+++ b/tcg/README
@@ -539,24 +539,29 @@ version. Aliases are specified in the input operands as
for GCC.
The same register may be used for both an input and an output, even when
they are not explicitly aliased. If an op expands to multiple target
instructions then care must be taken to avoid clobbering input values.
-GCC style "early clobber" outputs are not currently supported.
+GCC style "early clobber" outputs are supported, with '&'.
A target can define specific register or constant constraints. If an
operation uses a constant input constraint which does not allow all
constants, it must also accept registers in order to have a fallback.
+The constraint 'i' is defined generically to accept any constant.
+The constraint 'r' is not defined generically, but is consistently
+used by each backend to indicate all registers.
The movi_i32 and movi_i64 operations must accept any constants.
The mov_i32 and mov_i64 operations must accept any registers of the
same type.
-The ld/st instructions must accept signed 32 bit constant offsets. It
-can be implemented by reserving a specific register to compute the
-address if the offset is too big.
+The ld/st/sti instructions must accept signed 32 bit constant offsets.
+This can be implemented by reserving a specific register in which to
+compute the address if the offset is too big.
The ld/st instructions must accept any destination (ld) or source (st)
register.
+The sti instruction may fail if it cannot store the given constant.
+
4.3) Function call assumptions
- The only supported types for parameters and return value are: 32 and
diff --git a/tcg/tcg.c b/tcg/tcg.c
index 8b4dce7..cb898f1 100644
--- a/tcg/tcg.c
+++ b/tcg/tcg.c
@@ -1256,37 +1256,37 @@ static void process_op_defs(TCGContext *s)
tcg_regset_clear(def->args_ct[i].u.regs);
def->args_ct[i].ct = 0;
- if (ct_str[0] >= '0' && ct_str[0] <= '9') {
- int oarg;
- oarg = ct_str[0] - '0';
- tcg_debug_assert(oarg < def->nb_oargs);
- tcg_debug_assert(def->args_ct[oarg].ct & TCG_CT_REG);
- /* TCG_CT_ALIAS is for the output arguments. The input
- argument is tagged with TCG_CT_IALIAS. */
- def->args_ct[i] = def->args_ct[oarg];
- def->args_ct[oarg].ct = TCG_CT_ALIAS;
- def->args_ct[oarg].alias_index = i;
- def->args_ct[i].ct |= TCG_CT_IALIAS;
- def->args_ct[i].alias_index = oarg;
- } else {
- for(;;) {
- if (*ct_str == '\0')
- break;
- switch(*ct_str) {
- case '&':
- def->args_ct[i].ct |= TCG_CT_NEWREG;
- ct_str++;
- break;
- case 'i':
- def->args_ct[i].ct |= TCG_CT_CONST;
- ct_str++;
- break;
- default:
- ct_str = target_parse_constraint(&def->args_ct[i],
- ct_str, type);
- /* Typo in TCGTargetOpDef constraint. */
- tcg_debug_assert(ct_str != NULL);
+ while (*ct_str != '\0') {
+ switch(*ct_str) {
+ case '0' ... '9':
+ {
+ int oarg = *ct_str - '0';
+ tcg_debug_assert(ct_str == tdefs->args_ct_str[i]);
+ tcg_debug_assert(oarg < def->nb_oargs);
+ tcg_debug_assert(def->args_ct[oarg].ct & TCG_CT_REG);
+ /* TCG_CT_ALIAS is for the output arguments.
+ The input is tagged with TCG_CT_IALIAS. */
+ def->args_ct[i] = def->args_ct[oarg];
+ def->args_ct[oarg].ct |= TCG_CT_ALIAS;
+ def->args_ct[oarg].alias_index = i;
+ def->args_ct[i].ct |= TCG_CT_IALIAS;
+ def->args_ct[i].alias_index = oarg;
}
+ ct_str++;
+ break;
+ case '&':
+ def->args_ct[i].ct |= TCG_CT_NEWREG;
+ ct_str++;
+ break;
+ case 'i':
+ def->args_ct[i].ct |= TCG_CT_CONST;
+ ct_str++;
+ break;
+ default:
+ ct_str = target_parse_constraint(&def->args_ct[i],
+ ct_str, type);
+ /* Typo in TCGTargetOpDef constraint. */
+ tcg_debug_assert(ct_str != NULL);
}
}
}
@@ -2296,7 +2296,8 @@ static void tcg_reg_alloc_op(TCGContext *s,
arg = args[i];
arg_ct = &def->args_ct[i];
ts = &s->temps[arg];
- if (arg_ct->ct & TCG_CT_ALIAS) {
+ if ((arg_ct->ct & TCG_CT_ALIAS)
+ && !const_args[arg_ct->alias_index]) {
reg = new_args[arg_ct->alias_index];
} else if (arg_ct->ct & TCG_CT_NEWREG) {
reg = tcg_reg_alloc(s, arg_ct->u.regs,
--
2.9.3
- [Qemu-devel] [PATCH 06/65] tcg/arm: Implement field extraction opcodes, (continued)
- [Qemu-devel] [PATCH 06/65] tcg/arm: Implement field extraction opcodes, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 10/65] tcg/s390: Expose host facilities to tcg-target.h, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 13/65] target-alpha: Use deposit and extract ops, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 15/65] target-i386: Use new deposit and extract ops, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 14/65] target-arm: Use new deposit and extract ops, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 17/65] target-ppc: Use the new deposit and extract ops, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 18/65] target-s390x: Use the new deposit and extract ops, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 19/65] tcg/optimize: Fold movcond 0/1 into setcond, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 21/65] tcg: Transition flat op_defs array to a target callback, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 20/65] tcg: Add markup for output requires new register, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 23/65] tcg: Allow an operand to be matching or a constant,
Richard Henderson <=
- [Qemu-devel] [PATCH 16/65] target-mips: Use the new extract op, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 25/65] disas/i386.c: Handle tzcnt, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 26/65] disas/ppc: Handle popcnt and cnttz, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 22/65] tcg: Pass the opcode width to target_parse_constraint, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 27/65] target-alpha: Use the ctz and clz opcodes, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 29/65] target-microblaze: Use clz opcode, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 30/65] target-mips: Use clz opcode, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 24/65] tcg: Add clz and ctz opcodes, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 28/65] target-cris: Use clz opcode, Richard Henderson, 2016/12/23
- [Qemu-devel] [PATCH 36/65] target-unicore32: Use clz opcode, Richard Henderson, 2016/12/23