From: Alistair Francis <alistair23@gmail.com>
To: LIU Zhiwei <zhiwei_liu@c-sky.com>
Cc: "open list:RISC-V" <qemu-riscv@nongnu.org>,
Bin Meng <bin.meng@windriver.com>,
Richard Henderson <richard.henderson@linaro.org>,
"qemu-devel@nongnu.org Developers" <qemu-devel@nongnu.org>,
Palmer Dabbelt <palmer@dabbelt.com>,
Alistair Francis <Alistair.Francis@wdc.com>
Subject: Re: [PATCH v4 11/20] target/riscv: Split pm_enabled into mask and base
Date: Fri, 19 Nov 2021 14:51:10 +1000 [thread overview]
Message-ID: <CAKmqyKN9MoJtfxTay2bo2n=EsQ5ZHX1MuM3NA0UHn1fzGxQ8VQ@mail.gmail.com> (raw)
In-Reply-To: <20211111155149.58172-12-zhiwei_liu@c-sky.com>
On Fri, Nov 12, 2021 at 2:05 AM LIU Zhiwei <zhiwei_liu@c-sky.com> wrote:
>
> Use cached cur_pmmask and cur_pmbase to infer the
> current PM mode.
>
> This may decrease the TCG IR by one when pm_enabled
> is true and pm_base_enabled is false.
>
> Signed-off-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Alistair
> ---
> target/riscv/cpu.h | 3 ++-
> target/riscv/cpu_helper.c | 25 +++++++------------------
> target/riscv/translate.c | 12 ++++++++----
> 3 files changed, 17 insertions(+), 23 deletions(-)
>
> diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
> index fa5a6ba1c8..9fba876e08 100644
> --- a/target/riscv/cpu.h
> +++ b/target/riscv/cpu.h
> @@ -411,7 +411,8 @@ FIELD(TB_FLAGS, MSTATUS_HS_FS, 11, 2)
> /* The combination of MXL/SXL/UXL that applies to the current cpu mode. */
> FIELD(TB_FLAGS, XL, 13, 2)
> /* If PointerMasking should be applied */
> -FIELD(TB_FLAGS, PM_ENABLED, 15, 1)
> +FIELD(TB_FLAGS, PM_MASK_ENABLED, 15, 1)
> +FIELD(TB_FLAGS, PM_BASE_ENABLED, 16, 1)
>
> #ifdef TARGET_RISCV32
> #define riscv_cpu_mxl(env) ((void)(env), MXL_RV32)
> diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c
> index 8320f56d9f..a40ed6d748 100644
> --- a/target/riscv/cpu_helper.c
> +++ b/target/riscv/cpu_helper.c
> @@ -108,26 +108,15 @@ void cpu_get_tb_cpu_state(CPURISCVState *env, target_ulong *pc,
> flags = FIELD_DP32(flags, TB_FLAGS, MSTATUS_HS_FS,
> get_field(env->mstatus_hs, MSTATUS_FS));
> }
> - if (riscv_has_ext(env, RVJ)) {
> - int priv = flags & TB_FLAGS_PRIV_MMU_MASK;
> - bool pm_enabled = false;
> - switch (priv) {
> - case PRV_U:
> - pm_enabled = env->mmte & U_PM_ENABLE;
> - break;
> - case PRV_S:
> - pm_enabled = env->mmte & S_PM_ENABLE;
> - break;
> - case PRV_M:
> - pm_enabled = env->mmte & M_PM_ENABLE;
> - break;
> - default:
> - g_assert_not_reached();
> - }
> - flags = FIELD_DP32(flags, TB_FLAGS, PM_ENABLED, pm_enabled);
> - }
> #endif
>
> + if (env->cur_pmmask < (xl == MXL_RV32 ? UINT32_MAX : UINT64_MAX)) {
> + flags = FIELD_DP32(flags, TB_FLAGS, PM_MASK_ENABLED, 1);
> + }
> + if (env->cur_pmbase != 0) {
> + flags = FIELD_DP32(flags, TB_FLAGS, PM_BASE_ENABLED, 1);
> + }
> +
> flags = FIELD_DP32(flags, TB_FLAGS, XL, xl);
>
> *pflags = flags;
> diff --git a/target/riscv/translate.c b/target/riscv/translate.c
> index fd75f7c4bc..10c16e759d 100644
> --- a/target/riscv/translate.c
> +++ b/target/riscv/translate.c
> @@ -87,7 +87,8 @@ typedef struct DisasContext {
> /* Space for 3 operands plus 1 extra for address computation. */
> TCGv temp[4];
> /* PointerMasking extension */
> - bool pm_enabled;
> + bool pm_mask_enabled;
> + bool pm_base_enabled;
> } DisasContext;
>
> static inline bool has_ext(DisasContext *ctx, uint32_t ext)
> @@ -291,12 +292,14 @@ static TCGv get_address(DisasContext *ctx, int rs1, int imm)
> TCGv src1 = get_gpr(ctx, rs1, EXT_NONE);
>
> tcg_gen_addi_tl(addr, src1, imm);
> - if (ctx->pm_enabled) {
> + if (ctx->pm_mask_enabled) {
> tcg_gen_and_tl(addr, addr, pm_mask);
> - tcg_gen_or_tl(addr, addr, pm_base);
> } else if (get_xl(ctx) == MXL_RV32) {
> tcg_gen_ext32u_tl(addr, addr);
> }
> + if (ctx->pm_base_enabled) {
> + tcg_gen_or_tl(addr, addr, pm_base);
> + }
> return addr;
> }
>
> @@ -643,7 +646,8 @@ static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs)
> ctx->cs = cs;
> ctx->ntemp = 0;
> memset(ctx->temp, 0, sizeof(ctx->temp));
> - ctx->pm_enabled = FIELD_EX32(tb_flags, TB_FLAGS, PM_ENABLED);
> + ctx->pm_mask_enabled = FIELD_EX32(tb_flags, TB_FLAGS, PM_MASK_ENABLED);
> + ctx->pm_base_enabled = FIELD_EX32(tb_flags, TB_FLAGS, PM_BASE_ENABLED);
> ctx->zero = tcg_constant_tl(0);
> }
>
> --
> 2.25.1
>
>
next prev parent reply other threads:[~2021-11-19 4:52 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-11-11 15:51 [PATCH v4 00/20] Support UXL filed in xstatus LIU Zhiwei
2021-11-11 15:51 ` [PATCH v4 01/20] target/riscv: Don't save pc when exception return LIU Zhiwei
2021-11-15 4:25 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 02/20] target/riscv: Sign extend pc for different XLEN LIU Zhiwei
2021-11-15 4:26 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 03/20] target/riscv: Ignore the pc bits above XLEN LIU Zhiwei
2021-11-15 4:27 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 04/20] target/riscv: Extend pc for runtime pc write LIU Zhiwei
2021-11-16 0:08 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 05/20] target/riscv: Use gdb xml according to max mxlen LIU Zhiwei
2021-11-16 3:12 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 06/20] target/riscv: Relax debug check for pm write LIU Zhiwei
2021-11-16 3:13 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 07/20] target/riscv: Adjust csr write mask with XLEN LIU Zhiwei
2021-11-16 3:14 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 08/20] target/riscv: Create current pm fields in env LIU Zhiwei
2021-11-19 4:22 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 09/20] target/riscv: Alloc tcg global for cur_pm[mask|base] LIU Zhiwei
2021-11-19 4:29 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 10/20] target/riscv: Calculate address according to XLEN LIU Zhiwei
2021-11-19 4:32 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 11/20] target/riscv: Split pm_enabled into mask and base LIU Zhiwei
2021-11-19 4:51 ` Alistair Francis [this message]
2021-11-11 15:51 ` [PATCH v4 12/20] target/riscv: Split out the vill from vtype LIU Zhiwei
2021-11-19 4:55 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 13/20] target/riscv: Fix RESERVED field length in VTYPE LIU Zhiwei
2021-11-19 4:56 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 14/20] target/riscv: Adjust vsetvl according to XLEN LIU Zhiwei
2021-11-19 12:40 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 15/20] target/riscv: Remove VILL field in VTYPE LIU Zhiwei
2021-11-19 12:33 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 16/20] target/riscv: Ajdust vector atomic check with XLEN LIU Zhiwei
2021-11-19 12:34 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 17/20] target/riscv: Fix check range for first fault only LIU Zhiwei
2021-11-19 12:42 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 18/20] target/riscv: Adjust vector address with mask LIU Zhiwei
2021-11-19 12:46 ` Alistair Francis
2021-11-11 15:51 ` [PATCH v4 19/20] target/riscv: Adjust scalar reg in vector with XLEN LIU Zhiwei
2021-11-11 15:51 ` [PATCH v4 20/20] target/riscv: Enable uxl field write LIU Zhiwei
2021-11-11 18:23 ` Richard Henderson
2021-11-19 12:55 ` Alistair Francis
2021-11-19 12:57 ` [PATCH v4 00/20] Support UXL filed in xstatus Alistair Francis
2021-11-19 13:44 ` LIU Zhiwei
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAKmqyKN9MoJtfxTay2bo2n=EsQ5ZHX1MuM3NA0UHn1fzGxQ8VQ@mail.gmail.com' \
--to=alistair23@gmail.com \
--cc=Alistair.Francis@wdc.com \
--cc=bin.meng@windriver.com \
--cc=palmer@dabbelt.com \
--cc=qemu-devel@nongnu.org \
--cc=qemu-riscv@nongnu.org \
--cc=richard.henderson@linaro.org \
--cc=zhiwei_liu@c-sky.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).