From: Deepak Gupta <debug@rivosinc.com>
To: qemu-riscv@nongnu.org, qemu-devel@nongnu.org
Cc: palmer@dabbelt.com, Alistair.Francis@wdc.com, bmeng.cn@gmail.com,
liwei1518@gmail.com, dbarboza@ventanamicro.com,
zhiwei_liu@linux.alibaba.com, jim.shu@sifive.com,
kito.cheng@sifive.com, Deepak Gupta <debug@rivosinc.com>,
Richard Henderson <richard.henderson@linaro.org>,
Alistair Francis <alistair.francis@wdc.com>
Subject: [PATCH v15 15/21] target/riscv: AMO operations always raise store/AMO fault
Date: Thu, 3 Oct 2024 11:33:36 -0700 [thread overview]
Message-ID: <20241003183342.679249-16-debug@rivosinc.com> (raw)
In-Reply-To: <20241003183342.679249-1-debug@rivosinc.com>
This patch adds one more word for tcg compile which can be obtained during
unwind time to determine fault type for original operation (example AMO).
Depending on that, fault can be promoted to store/AMO fault.
Signed-off-by: Deepak Gupta <debug@rivosinc.com>
Suggested-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
---
target/riscv/cpu.h | 9 ++++++++-
target/riscv/cpu_helper.c | 20 ++++++++++++++++++++
target/riscv/tcg/tcg-cpu.c | 1 +
target/riscv/translate.c | 2 +-
4 files changed, 30 insertions(+), 2 deletions(-)
diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
index d39650636c..fb93b0c859 100644
--- a/target/riscv/cpu.h
+++ b/target/riscv/cpu.h
@@ -46,8 +46,13 @@ typedef struct CPUArchState CPURISCVState;
/*
* RISC-V-specific extra insn start words:
* 1: Original instruction opcode
+ * 2: more information about instruction
*/
-#define TARGET_INSN_START_EXTRA_WORDS 1
+#define TARGET_INSN_START_EXTRA_WORDS 2
+/*
+ * b0: Whether a instruction always raise a store AMO or not.
+ */
+#define RISCV_UW2_ALWAYS_STORE_AMO 1
#define RV(x) ((target_ulong)1 << (x - 'A'))
@@ -234,6 +239,8 @@ struct CPUArchState {
bool elp;
/* shadow stack register for zicfiss extension */
target_ulong ssp;
+ /* env place holder for extra word 2 during unwind */
+ target_ulong excp_uw2;
/* sw check code for sw check exception */
target_ulong sw_check_code;
#ifdef CONFIG_USER_ONLY
diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c
index ab46f694b5..3aae6343c8 100644
--- a/target/riscv/cpu_helper.c
+++ b/target/riscv/cpu_helper.c
@@ -1757,6 +1757,22 @@ static target_ulong riscv_transformed_insn(CPURISCVState *env,
return xinsn;
}
+static target_ulong promote_load_fault(target_ulong orig_cause)
+{
+ switch (orig_cause) {
+ case RISCV_EXCP_LOAD_GUEST_ACCESS_FAULT:
+ return RISCV_EXCP_STORE_GUEST_AMO_ACCESS_FAULT;
+
+ case RISCV_EXCP_LOAD_ACCESS_FAULT:
+ return RISCV_EXCP_STORE_AMO_ACCESS_FAULT;
+
+ case RISCV_EXCP_LOAD_PAGE_FAULT:
+ return RISCV_EXCP_STORE_PAGE_FAULT;
+ }
+
+ /* if no promotion, return original cause */
+ return orig_cause;
+}
/*
* Handle Traps
*
@@ -1769,6 +1785,7 @@ void riscv_cpu_do_interrupt(CPUState *cs)
CPURISCVState *env = &cpu->env;
bool virt = env->virt_enabled;
bool write_gva = false;
+ bool always_storeamo = (env->excp_uw2 & RISCV_UW2_ALWAYS_STORE_AMO);
uint64_t s;
/*
@@ -1804,6 +1821,9 @@ void riscv_cpu_do_interrupt(CPUState *cs)
case RISCV_EXCP_STORE_AMO_ACCESS_FAULT:
case RISCV_EXCP_LOAD_PAGE_FAULT:
case RISCV_EXCP_STORE_PAGE_FAULT:
+ if (always_storeamo) {
+ cause = promote_load_fault(cause);
+ }
write_gva = env->two_stage_lookup;
tval = env->badaddr;
if (env->two_stage_indirect_lookup) {
diff --git a/target/riscv/tcg/tcg-cpu.c b/target/riscv/tcg/tcg-cpu.c
index 6c0c319499..c62c221696 100644
--- a/target/riscv/tcg/tcg-cpu.c
+++ b/target/riscv/tcg/tcg-cpu.c
@@ -129,6 +129,7 @@ static void riscv_restore_state_to_opc(CPUState *cs,
env->pc = pc;
}
env->bins = data[1];
+ env->excp_uw2 = data[2];
}
static const TCGCPUOps riscv_tcg_ops = {
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index afa2ed4e3a..0322597bf6 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1264,7 +1264,7 @@ static void riscv_tr_insn_start(DisasContextBase *dcbase, CPUState *cpu)
pc_next &= ~TARGET_PAGE_MASK;
}
- tcg_gen_insn_start(pc_next, 0);
+ tcg_gen_insn_start(pc_next, 0, 0);
ctx->insn_start_updated = false;
}
--
2.45.0
next prev parent reply other threads:[~2024-10-03 18:37 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-03 18:33 [PATCH v15 00/21] riscv support for control flow integrity extensions Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 01/21] target/riscv: expose *envcfg csr and priv to qemu-user as well Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 02/21] target/riscv: Add zicfilp extension Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 03/21] target/riscv: Introduce elp state and enabling controls for zicfilp Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 04/21] target/riscv: save and restore elp state on priv transitions Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 05/21] target/riscv: additional code information for sw check Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 06/21] target/riscv: tracking indirect branches (fcfi) for zicfilp Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 07/21] target/riscv: zicfilp `lpad` impl and branch tracking Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 08/21] disas/riscv: enable `lpad` disassembly Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 09/21] target/riscv: Expose zicfilp extension as a cpu property Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 10/21] target/riscv: Add zicfiss extension Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 11/21] target/riscv: introduce ssp and enabling controls for zicfiss Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 12/21] target/riscv: tb flag for shadow stack instructions Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 13/21] target/riscv: mmu changes for zicfiss shadow stack protection Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 14/21] target/riscv: disallow probe accesses to shadow stack Deepak Gupta
2024-10-04 0:36 ` Deepak Gupta
2024-10-03 18:33 ` Deepak Gupta [this message]
2024-10-03 18:33 ` [PATCH v15 16/21] target/riscv: update `decode_save_opc` to store extra word2 Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 17/21] target/riscv: implement zicfiss instructions Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 18/21] target/riscv: compressed encodings for sspush and sspopchk Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 19/21] disas/riscv: enable disassembly for zicfiss instructions Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 20/21] disas/riscv: enable disassembly for compressed sspush/sspopchk Deepak Gupta
2024-10-03 18:33 ` [PATCH v15 21/21] target/riscv: Expose zicfiss extension as a cpu property Deepak Gupta
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=20241003183342.679249-16-debug@rivosinc.com \
--to=debug@rivosinc.com \
--cc=Alistair.Francis@wdc.com \
--cc=bmeng.cn@gmail.com \
--cc=dbarboza@ventanamicro.com \
--cc=jim.shu@sifive.com \
--cc=kito.cheng@sifive.com \
--cc=liwei1518@gmail.com \
--cc=palmer@dabbelt.com \
--cc=qemu-devel@nongnu.org \
--cc=qemu-riscv@nongnu.org \
--cc=richard.henderson@linaro.org \
--cc=zhiwei_liu@linux.alibaba.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).