* [PATCH V4 0/4] Basic ASID2 Support
@ 2025-12-02 12:00 Jim MacArthur
2025-12-02 12:00 ` [PATCH V4 1/4] target/arm: Enable ID_AA64MMFR4_EL1 register Jim MacArthur
` (3 more replies)
0 siblings, 4 replies; 10+ messages in thread
From: Jim MacArthur @ 2025-12-02 12:00 UTC (permalink / raw)
To: qemu-devel; +Cc: Jim MacArthur
Added a line to the documentation mentioning that we can now support ASID2.
Changes in v4:
- Patch 3:
- Added mention of ASID2 to docs/system/arm/emulation.rst.
Jim MacArthur (4):
target/arm: Enable ID_AA64MMFR4_EL1 register.
target/arm: Allow writes to FNG1, FNG0, A2
target/arm/tcg/cpu64.c: Enable ASID2 for cpu_max
tests: Add test for ASID2 and write/read of feature bits
docs/system/arm/emulation.rst | 1 +
target/arm/cpu-features.h | 7 +++
target/arm/cpu-sysregs.h.inc | 1 +
target/arm/helper.c | 10 ++++-
target/arm/tcg/cpu64.c | 4 ++
tests/tcg/aarch64/system/asid2.c | 75 ++++++++++++++++++++++++++++++++
6 files changed, 96 insertions(+), 2 deletions(-)
create mode 100644 tests/tcg/aarch64/system/asid2.c
--
2.43.0
^ permalink raw reply [flat|nested] 10+ messages in thread
* [PATCH V4 1/4] target/arm: Enable ID_AA64MMFR4_EL1 register.
2025-12-02 12:00 [PATCH V4 0/4] Basic ASID2 Support Jim MacArthur
@ 2025-12-02 12:00 ` Jim MacArthur
2025-12-02 14:28 ` Gustavo Romero
2025-12-02 12:00 ` [PATCH V4 2/4] target/arm: Allow writes to FNG1, FNG0, A2 Jim MacArthur
` (2 subsequent siblings)
3 siblings, 1 reply; 10+ messages in thread
From: Jim MacArthur @ 2025-12-02 12:00 UTC (permalink / raw)
To: qemu-devel; +Cc: Jim MacArthur
Signed-off-by: Jim MacArthur <jim.macarthur@linaro.org>
---
target/arm/cpu-sysregs.h.inc | 1 +
target/arm/helper.c | 4 ++--
2 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/target/arm/cpu-sysregs.h.inc b/target/arm/cpu-sysregs.h.inc
index 2bb2861c62..2ba49d8478 100644
--- a/target/arm/cpu-sysregs.h.inc
+++ b/target/arm/cpu-sysregs.h.inc
@@ -14,6 +14,7 @@ DEF(ID_AA64MMFR0_EL1, 3, 0, 0, 7, 0)
DEF(ID_AA64MMFR1_EL1, 3, 0, 0, 7, 1)
DEF(ID_AA64MMFR2_EL1, 3, 0, 0, 7, 2)
DEF(ID_AA64MMFR3_EL1, 3, 0, 0, 7, 3)
+DEF(ID_AA64MMFR4_EL1, 3, 0, 0, 7, 4)
DEF(ID_PFR0_EL1, 3, 0, 0, 1, 0)
DEF(ID_PFR1_EL1, 3, 0, 0, 1, 1)
DEF(ID_DFR0_EL1, 3, 0, 0, 1, 2)
diff --git a/target/arm/helper.c b/target/arm/helper.c
index 27ebc6f29b..c20334fa65 100644
--- a/target/arm/helper.c
+++ b/target/arm/helper.c
@@ -6566,11 +6566,11 @@ void register_cp_regs_for_features(ARMCPU *cpu)
.access = PL1_R, .type = ARM_CP_CONST,
.accessfn = access_aa64_tid3,
.resetvalue = GET_IDREG(isar, ID_AA64MMFR3) },
- { .name = "ID_AA64MMFR4_EL1_RESERVED", .state = ARM_CP_STATE_AA64,
+ { .name = "ID_AA64MMFR4_EL1", .state = ARM_CP_STATE_AA64,
.opc0 = 3, .opc1 = 0, .crn = 0, .crm = 7, .opc2 = 4,
.access = PL1_R, .type = ARM_CP_CONST,
.accessfn = access_aa64_tid3,
- .resetvalue = 0 },
+ .resetvalue = GET_IDREG(isar, ID_AA64MMFR4) },
{ .name = "ID_AA64MMFR5_EL1_RESERVED", .state = ARM_CP_STATE_AA64,
.opc0 = 3, .opc1 = 0, .crn = 0, .crm = 7, .opc2 = 5,
.access = PL1_R, .type = ARM_CP_CONST,
--
2.43.0
^ permalink raw reply related [flat|nested] 10+ messages in thread
* [PATCH V4 2/4] target/arm: Allow writes to FNG1, FNG0, A2
2025-12-02 12:00 [PATCH V4 0/4] Basic ASID2 Support Jim MacArthur
2025-12-02 12:00 ` [PATCH V4 1/4] target/arm: Enable ID_AA64MMFR4_EL1 register Jim MacArthur
@ 2025-12-02 12:00 ` Jim MacArthur
2025-12-02 14:29 ` Gustavo Romero
2025-12-02 12:00 ` [PATCH V4 3/4] target/arm/tcg/cpu64.c: Enable ASID2 for cpu_max Jim MacArthur
2025-12-02 12:00 ` [PATCH V4 4/4] tests: Add test for ASID2 and write/read of feature bits Jim MacArthur
3 siblings, 1 reply; 10+ messages in thread
From: Jim MacArthur @ 2025-12-02 12:00 UTC (permalink / raw)
To: qemu-devel; +Cc: Jim MacArthur
This just allows read/write of three feature bits. ASID is still
ignored. Any writes to TTBR0_EL0 and TTBR1_EL0, including changing
the ASID, will still cause a complete flush of the TLB.
Signed-off-by: Jim MacArthur <jim.macarthur@linaro.org>
---
target/arm/cpu-features.h | 7 +++++++
target/arm/helper.c | 6 ++++++
2 files changed, 13 insertions(+)
diff --git a/target/arm/cpu-features.h b/target/arm/cpu-features.h
index 579fa8f8f4..d56bda9ce0 100644
--- a/target/arm/cpu-features.h
+++ b/target/arm/cpu-features.h
@@ -346,6 +346,8 @@ FIELD(ID_AA64MMFR3, SDERR, 52, 4)
FIELD(ID_AA64MMFR3, ADERR, 56, 4)
FIELD(ID_AA64MMFR3, SPEC_FPACC, 60, 4)
+FIELD(ID_AA64MMFR4, ASID2, 8, 4)
+
FIELD(ID_AA64DFR0, DEBUGVER, 0, 4)
FIELD(ID_AA64DFR0, TRACEVER, 4, 4)
FIELD(ID_AA64DFR0, PMUVER, 8, 4)
@@ -1369,6 +1371,11 @@ static inline bool isar_feature_aa64_aie(const ARMISARegisters *id)
return FIELD_EX64_IDREG(id, ID_AA64MMFR3, AIE) != 0;
}
+static inline bool isar_feature_aa64_asid2(const ARMISARegisters *id)
+{
+ return FIELD_EX64_IDREG(id, ID_AA64MMFR4, ASID2) != 0;
+}
+
static inline bool isar_feature_aa64_mec(const ARMISARegisters *id)
{
return FIELD_EX64_IDREG(id, ID_AA64MMFR3, MEC) != 0;
diff --git a/target/arm/helper.c b/target/arm/helper.c
index c20334fa65..7812a82bab 100644
--- a/target/arm/helper.c
+++ b/target/arm/helper.c
@@ -6102,6 +6102,9 @@ static void tcr2_el1_write(CPUARMState *env, const ARMCPRegInfo *ri,
if (cpu_isar_feature(aa64_aie, cpu)) {
valid_mask |= TCR2_AIE;
}
+ if (cpu_isar_feature(aa64_asid2, cpu)) {
+ valid_mask |= TCR2_FNG1 | TCR2_FNG0 | TCR2_A2;
+ }
value &= valid_mask;
raw_write(env, ri, value);
}
@@ -6121,6 +6124,9 @@ static void tcr2_el2_write(CPUARMState *env, const ARMCPRegInfo *ri,
if (cpu_isar_feature(aa64_mec, cpu)) {
valid_mask |= TCR2_AMEC0 | TCR2_AMEC1;
}
+ if (cpu_isar_feature(aa64_asid2, cpu)) {
+ valid_mask |= TCR2_FNG1 | TCR2_FNG0 | TCR2_A2;
+ }
value &= valid_mask;
raw_write(env, ri, value);
}
--
2.43.0
^ permalink raw reply related [flat|nested] 10+ messages in thread
* [PATCH V4 3/4] target/arm/tcg/cpu64.c: Enable ASID2 for cpu_max
2025-12-02 12:00 [PATCH V4 0/4] Basic ASID2 Support Jim MacArthur
2025-12-02 12:00 ` [PATCH V4 1/4] target/arm: Enable ID_AA64MMFR4_EL1 register Jim MacArthur
2025-12-02 12:00 ` [PATCH V4 2/4] target/arm: Allow writes to FNG1, FNG0, A2 Jim MacArthur
@ 2025-12-02 12:00 ` Jim MacArthur
2025-12-02 14:29 ` Gustavo Romero
2025-12-02 12:00 ` [PATCH V4 4/4] tests: Add test for ASID2 and write/read of feature bits Jim MacArthur
3 siblings, 1 reply; 10+ messages in thread
From: Jim MacArthur @ 2025-12-02 12:00 UTC (permalink / raw)
To: qemu-devel; +Cc: Jim MacArthur
docs/system/arm/emulation.rst: Add ASID2
Signed-off-by: Jim MacArthur <jim.macarthur@linaro.org>
---
docs/system/arm/emulation.rst | 1 +
target/arm/tcg/cpu64.c | 4 ++++
2 files changed, 5 insertions(+)
diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst
index 31a5878a8f..3f30ea5a30 100644
--- a/docs/system/arm/emulation.rst
+++ b/docs/system/arm/emulation.rst
@@ -24,6 +24,7 @@ the following architecture extensions:
- FEAT_AIE (Memory Attribute Index Enhancement)
- FEAT_Armv9_Crypto (Armv9 Cryptographic Extension)
- FEAT_ASID16 (16 bit ASID)
+- FEAT_ASID2 (Concurrent use of two ASIDs)
- FEAT_ATS1A (Address Translation operations that ignore stage 1 permissions)
- FEAT_BBM at level 2 (Translation table break-before-make levels)
- FEAT_BF16 (AArch64 BFloat16 instructions)
diff --git a/target/arm/tcg/cpu64.c b/target/arm/tcg/cpu64.c
index 6871956382..ef4c0c8d73 100644
--- a/target/arm/tcg/cpu64.c
+++ b/target/arm/tcg/cpu64.c
@@ -1334,6 +1334,10 @@ void aarch64_max_tcg_initfn(Object *obj)
t = FIELD_DP64(t, ID_AA64MMFR3, AIE, 1); /* FEAT_AIE */
SET_IDREG(isar, ID_AA64MMFR3, t);
+ t = GET_IDREG(isar, ID_AA64MMFR4);
+ t = FIELD_DP64(t, ID_AA64MMFR4, ASID2, 1); /* FEAT_ASID2 */
+ SET_IDREG(isar, ID_AA64MMFR4, t);
+
t = GET_IDREG(isar, ID_AA64ZFR0);
t = FIELD_DP64(t, ID_AA64ZFR0, SVEVER, 2); /* FEAT_SVE2p1 */
t = FIELD_DP64(t, ID_AA64ZFR0, AES, 2); /* FEAT_SVE_PMULL128 */
--
2.43.0
^ permalink raw reply related [flat|nested] 10+ messages in thread
* [PATCH V4 4/4] tests: Add test for ASID2 and write/read of feature bits
2025-12-02 12:00 [PATCH V4 0/4] Basic ASID2 Support Jim MacArthur
` (2 preceding siblings ...)
2025-12-02 12:00 ` [PATCH V4 3/4] target/arm/tcg/cpu64.c: Enable ASID2 for cpu_max Jim MacArthur
@ 2025-12-02 12:00 ` Jim MacArthur
2025-12-02 14:30 ` Gustavo Romero
3 siblings, 1 reply; 10+ messages in thread
From: Jim MacArthur @ 2025-12-02 12:00 UTC (permalink / raw)
To: qemu-devel; +Cc: Jim MacArthur
Test for presence of ASID2; if it is, check FNG1, FNG0, and A2 are
writable, and read value shows the update. If not present, check these
read as RES0.
Signed-off-by: Jim MacArthur <jim.macarthur@linaro.org>
---
tests/tcg/aarch64/system/asid2.c | 75 ++++++++++++++++++++++++++++++++
1 file changed, 75 insertions(+)
create mode 100644 tests/tcg/aarch64/system/asid2.c
diff --git a/tests/tcg/aarch64/system/asid2.c b/tests/tcg/aarch64/system/asid2.c
new file mode 100644
index 0000000000..a4887e4ce2
--- /dev/null
+++ b/tests/tcg/aarch64/system/asid2.c
@@ -0,0 +1,75 @@
+/*
+ * ASID2 Feature presence and enabled TCR2_EL1 bits test
+ *
+ * Copyright (c) 2025 Linaro Ltd
+ *
+ * SPDX-License-Identifier: GPL-2.0-or-later
+ */
+
+#include <stdint.h>
+#include <minilib.h>
+
+#define ID_AA64MMFR3_EL1 "S3_0_C0_C7_3"
+#define ID_AA64MMFR4_EL1 "S3_0_C0_C7_4"
+#define TCR2_EL1 "S3_0_C2_C0_3"
+
+int main()
+{
+ /*
+ * Test for presence of ASID2 and three feature bits enabled by it:
+ * https://developer.arm.com/documentation/109697/2025_09/Feature-descriptions/The-Armv9-5-architecture-extension
+ * Bits added are FNG1, FNG0, and A2. These should be RES0 if A2 is
+ * not enabled and read as the written value if A2 is enabled.
+ */
+
+ uint64_t out;
+ uint64_t idreg3;
+ uint64_t idreg4;
+ int tcr2_present;
+ int asid2_present;
+
+ /* Mask is FNG1, FNG0, and A2 */
+ const uint64_t feature_mask = (1ULL << 18 | 1ULL << 17 | 1ULL << 16);
+ const uint64_t in = feature_mask;
+
+ asm("mrs %[x1], " ID_AA64MMFR3_EL1 "\n\t"
+ : [x1] "=r" (idreg3));
+
+ tcr2_present = ((idreg3 & 0xF) != 0);
+
+ if (!tcr2_present) {
+ ml_printf("TCR2 is not present, cannot perform test");
+ return 0;
+ }
+
+ asm("mrs %[x1], " ID_AA64MMFR4_EL1 "\n\t"
+ : [x1] "=r" (idreg4));
+
+ asid2_present = ((idreg4 & 0xF00) != 0);
+
+ asm("msr " TCR2_EL1 ", %[x0]\n\t"
+ "mrs %[x1], " TCR2_EL1 "\n\t"
+ : [x1] "=r" (out)
+ : [x0] "r" (in));
+
+ if (asid2_present) {
+ if ((out & feature_mask) == in) {
+ ml_printf("OK\n");
+ return 0;
+ } else {
+ ml_printf("FAIL: ASID2 present, but read value %lx != "
+ "written value %lx\n",
+ out & feature_mask, in);
+ return 1;
+ }
+ } else {
+ if (out == 0) {
+ ml_printf("TCR2_EL1 reads as RES0 as expected\n");
+ return 0;
+ } else {
+ ml_printf("FAIL: ASID2, missing but read value %lx != 0\n",
+ out & feature_mask, in);
+ return 1;
+ }
+ }
+}
--
2.43.0
^ permalink raw reply related [flat|nested] 10+ messages in thread
* Re: [PATCH V4 1/4] target/arm: Enable ID_AA64MMFR4_EL1 register.
2025-12-02 12:00 ` [PATCH V4 1/4] target/arm: Enable ID_AA64MMFR4_EL1 register Jim MacArthur
@ 2025-12-02 14:28 ` Gustavo Romero
0 siblings, 0 replies; 10+ messages in thread
From: Gustavo Romero @ 2025-12-02 14:28 UTC (permalink / raw)
To: Jim MacArthur, qemu-devel
Hi Jim,
nit: usually we don't use period in the commit title.
On 12/2/25 09:00, Jim MacArthur wrote:
> Signed-off-by: Jim MacArthur <jim.macarthur@linaro.org>
> ---
> target/arm/cpu-sysregs.h.inc | 1 +
> target/arm/helper.c | 4 ++--
> 2 files changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/target/arm/cpu-sysregs.h.inc b/target/arm/cpu-sysregs.h.inc
> index 2bb2861c62..2ba49d8478 100644
> --- a/target/arm/cpu-sysregs.h.inc
> +++ b/target/arm/cpu-sysregs.h.inc
> @@ -14,6 +14,7 @@ DEF(ID_AA64MMFR0_EL1, 3, 0, 0, 7, 0)
> DEF(ID_AA64MMFR1_EL1, 3, 0, 0, 7, 1)
> DEF(ID_AA64MMFR2_EL1, 3, 0, 0, 7, 2)
> DEF(ID_AA64MMFR3_EL1, 3, 0, 0, 7, 3)
> +DEF(ID_AA64MMFR4_EL1, 3, 0, 0, 7, 4)
> DEF(ID_PFR0_EL1, 3, 0, 0, 1, 0)
> DEF(ID_PFR1_EL1, 3, 0, 0, 1, 1)
> DEF(ID_DFR0_EL1, 3, 0, 0, 1, 2)
> diff --git a/target/arm/helper.c b/target/arm/helper.c
> index 27ebc6f29b..c20334fa65 100644
> --- a/target/arm/helper.c
> +++ b/target/arm/helper.c
> @@ -6566,11 +6566,11 @@ void register_cp_regs_for_features(ARMCPU *cpu)
> .access = PL1_R, .type = ARM_CP_CONST,
> .accessfn = access_aa64_tid3,
> .resetvalue = GET_IDREG(isar, ID_AA64MMFR3) },
> - { .name = "ID_AA64MMFR4_EL1_RESERVED", .state = ARM_CP_STATE_AA64,
> + { .name = "ID_AA64MMFR4_EL1", .state = ARM_CP_STATE_AA64,
> .opc0 = 3, .opc1 = 0, .crn = 0, .crm = 7, .opc2 = 4,
> .access = PL1_R, .type = ARM_CP_CONST,
> .accessfn = access_aa64_tid3,
> - .resetvalue = 0 },
> + .resetvalue = GET_IDREG(isar, ID_AA64MMFR4) },
> { .name = "ID_AA64MMFR5_EL1_RESERVED", .state = ARM_CP_STATE_AA64,
> .opc0 = 3, .opc1 = 0, .crn = 0, .crm = 7, .opc2 = 5,
> .access = PL1_R, .type = ARM_CP_CONST,
Reviewed-by: Gustavo Romero <gustavo.romero@linaro.org>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH V4 2/4] target/arm: Allow writes to FNG1, FNG0, A2
2025-12-02 12:00 ` [PATCH V4 2/4] target/arm: Allow writes to FNG1, FNG0, A2 Jim MacArthur
@ 2025-12-02 14:29 ` Gustavo Romero
2025-12-04 16:25 ` Jim MacArthur
0 siblings, 1 reply; 10+ messages in thread
From: Gustavo Romero @ 2025-12-02 14:29 UTC (permalink / raw)
To: Jim MacArthur, qemu-devel, Richard Henderson
Hi Jim,
On 12/2/25 09:00, Jim MacArthur wrote:
> This just allows read/write of three feature bits. ASID is still
> ignored. Any writes to TTBR0_EL0 and TTBR1_EL0, including changing
> the ASID, will still cause a complete flush of the TLB.
>
> Signed-off-by: Jim MacArthur <jim.macarthur@linaro.org>
> ---
> target/arm/cpu-features.h | 7 +++++++
> target/arm/helper.c | 6 ++++++
> 2 files changed, 13 insertions(+)
>
> diff --git a/target/arm/cpu-features.h b/target/arm/cpu-features.h
> index 579fa8f8f4..d56bda9ce0 100644
> --- a/target/arm/cpu-features.h
> +++ b/target/arm/cpu-features.h
> @@ -346,6 +346,8 @@ FIELD(ID_AA64MMFR3, SDERR, 52, 4)
> FIELD(ID_AA64MMFR3, ADERR, 56, 4)
> FIELD(ID_AA64MMFR3, SPEC_FPACC, 60, 4)
>
> +FIELD(ID_AA64MMFR4, ASID2, 8, 4)
> +
> FIELD(ID_AA64DFR0, DEBUGVER, 0, 4)
> FIELD(ID_AA64DFR0, TRACEVER, 4, 4)
> FIELD(ID_AA64DFR0, PMUVER, 8, 4)
> @@ -1369,6 +1371,11 @@ static inline bool isar_feature_aa64_aie(const ARMISARegisters *id)
> return FIELD_EX64_IDREG(id, ID_AA64MMFR3, AIE) != 0;
> }
>
> +static inline bool isar_feature_aa64_asid2(const ARMISARegisters *id)
> +{
> + return FIELD_EX64_IDREG(id, ID_AA64MMFR4, ASID2) != 0;
> +}
> +
> static inline bool isar_feature_aa64_mec(const ARMISARegisters *id)
> {
> return FIELD_EX64_IDREG(id, ID_AA64MMFR3, MEC) != 0;
> diff --git a/target/arm/helper.c b/target/arm/helper.c
> index c20334fa65..7812a82bab 100644
> --- a/target/arm/helper.c
> +++ b/target/arm/helper.c
> @@ -6102,6 +6102,9 @@ static void tcr2_el1_write(CPUARMState *env, const ARMCPRegInfo *ri,
> if (cpu_isar_feature(aa64_aie, cpu)) {
> valid_mask |= TCR2_AIE;
> }
> + if (cpu_isar_feature(aa64_asid2, cpu)) {
> + valid_mask |= TCR2_FNG1 | TCR2_FNG0 | TCR2_A2;
> + }
> value &= valid_mask;
> raw_write(env, ri, value);
> }
> @@ -6121,6 +6124,9 @@ static void tcr2_el2_write(CPUARMState *env, const ARMCPRegInfo *ri,
> if (cpu_isar_feature(aa64_mec, cpu)) {
> valid_mask |= TCR2_AMEC0 | TCR2_AMEC1;
> }
> + if (cpu_isar_feature(aa64_asid2, cpu)) {
> + valid_mask |= TCR2_FNG1 | TCR2_FNG0 | TCR2_A2;
> + }
> value &= valid_mask;
> raw_write(env, ri, value);
> }
Afaics, we are not flushing the TLB here like we do for TCR_ELx (in vmsa_tcr_el12_write) before
we call raw_write(). Since here we could be changing the A2 & friends bits, which can change
the value of the ASID being using (like the TCR_ELx.A1 bit), I believe we should flush the TLB
explicitly here like we do in vmsa_tcr_el12_write().
@rth wdyt?
Cheers,
Gustavo
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH V4 3/4] target/arm/tcg/cpu64.c: Enable ASID2 for cpu_max
2025-12-02 12:00 ` [PATCH V4 3/4] target/arm/tcg/cpu64.c: Enable ASID2 for cpu_max Jim MacArthur
@ 2025-12-02 14:29 ` Gustavo Romero
0 siblings, 0 replies; 10+ messages in thread
From: Gustavo Romero @ 2025-12-02 14:29 UTC (permalink / raw)
To: Jim MacArthur, qemu-devel
Hi Jim,
On 12/2/25 09:00, Jim MacArthur wrote:
> docs/system/arm/emulation.rst: Add ASID2
>
> Signed-off-by: Jim MacArthur <jim.macarthur@linaro.org>
> ---
> docs/system/arm/emulation.rst | 1 +
> target/arm/tcg/cpu64.c | 4 ++++
> 2 files changed, 5 insertions(+)
>
> diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst
> index 31a5878a8f..3f30ea5a30 100644
> --- a/docs/system/arm/emulation.rst
> +++ b/docs/system/arm/emulation.rst
> @@ -24,6 +24,7 @@ the following architecture extensions:
> - FEAT_AIE (Memory Attribute Index Enhancement)
> - FEAT_Armv9_Crypto (Armv9 Cryptographic Extension)
> - FEAT_ASID16 (16 bit ASID)
> +- FEAT_ASID2 (Concurrent use of two ASIDs)
> - FEAT_ATS1A (Address Translation operations that ignore stage 1 permissions)
> - FEAT_BBM at level 2 (Translation table break-before-make levels)
> - FEAT_BF16 (AArch64 BFloat16 instructions)
> diff --git a/target/arm/tcg/cpu64.c b/target/arm/tcg/cpu64.c
> index 6871956382..ef4c0c8d73 100644
> --- a/target/arm/tcg/cpu64.c
> +++ b/target/arm/tcg/cpu64.c
> @@ -1334,6 +1334,10 @@ void aarch64_max_tcg_initfn(Object *obj)
> t = FIELD_DP64(t, ID_AA64MMFR3, AIE, 1); /* FEAT_AIE */
> SET_IDREG(isar, ID_AA64MMFR3, t);
>
> + t = GET_IDREG(isar, ID_AA64MMFR4);
> + t = FIELD_DP64(t, ID_AA64MMFR4, ASID2, 1); /* FEAT_ASID2 */
> + SET_IDREG(isar, ID_AA64MMFR4, t);
> +
> t = GET_IDREG(isar, ID_AA64ZFR0);
> t = FIELD_DP64(t, ID_AA64ZFR0, SVEVER, 2); /* FEAT_SVE2p1 */
> t = FIELD_DP64(t, ID_AA64ZFR0, AES, 2); /* FEAT_SVE_PMULL128 */
Reviewed-by: Gustavo Romero <gustavo.romero@linaro.org>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH V4 4/4] tests: Add test for ASID2 and write/read of feature bits
2025-12-02 12:00 ` [PATCH V4 4/4] tests: Add test for ASID2 and write/read of feature bits Jim MacArthur
@ 2025-12-02 14:30 ` Gustavo Romero
0 siblings, 0 replies; 10+ messages in thread
From: Gustavo Romero @ 2025-12-02 14:30 UTC (permalink / raw)
To: Jim MacArthur, qemu-devel
Hi Jim,
On 12/2/25 09:00, Jim MacArthur wrote:
> Test for presence of ASID2; if it is, check FNG1, FNG0, and A2 are
> writable, and read value shows the update. If not present, check these
> read as RES0.
>
> Signed-off-by: Jim MacArthur <jim.macarthur@linaro.org>
> ---
> tests/tcg/aarch64/system/asid2.c | 75 ++++++++++++++++++++++++++++++++
> 1 file changed, 75 insertions(+)
> create mode 100644 tests/tcg/aarch64/system/asid2.c
>
> diff --git a/tests/tcg/aarch64/system/asid2.c b/tests/tcg/aarch64/system/asid2.c
> new file mode 100644
> index 0000000000..a4887e4ce2
> --- /dev/null
> +++ b/tests/tcg/aarch64/system/asid2.c
> @@ -0,0 +1,75 @@
> +/*
> + * ASID2 Feature presence and enabled TCR2_EL1 bits test
> + *
> + * Copyright (c) 2025 Linaro Ltd
> + *
> + * SPDX-License-Identifier: GPL-2.0-or-later
The SPDX tag must go in the first line, except in script sources, where it goes
after the shebang line.
+ */
> +
> +#include <stdint.h>
> +#include <minilib.h>
> +
> +#define ID_AA64MMFR3_EL1 "S3_0_C0_C7_3"
> +#define ID_AA64MMFR4_EL1 "S3_0_C0_C7_4"
> +#define TCR2_EL1 "S3_0_C2_C0_3"
> +
> +int main()
> +{
> + /*
> + * Test for presence of ASID2 and three feature bits enabled by it:
> + * https://developer.arm.com/documentation/109697/2025_09/Feature-descriptions/The-Armv9-5-architecture-extension
> + * Bits added are FNG1, FNG0, and A2. These should be RES0 if A2 is
> + * not enabled and read as the written value if A2 is enabled.
> + */
> +
> + uint64_t out;
> + uint64_t idreg3;
> + uint64_t idreg4;
> + int tcr2_present;
> + int asid2_present;
> +
> + /* Mask is FNG1, FNG0, and A2 */
> + const uint64_t feature_mask = (1ULL << 18 | 1ULL << 17 | 1ULL << 16);
> + const uint64_t in = feature_mask;
> +
> + asm("mrs %[x1], " ID_AA64MMFR3_EL1 "\n\t"
> + : [x1] "=r" (idreg3));
You can use the same name for the asm label as the one used for the C variable to
ease even further reading the code, so it's ok to do:
asm("mrs %[idreg3], " ID_AA64MMFR3_EL1 "\n\t"
: [idreg3] "=r" (idreg3));
Likewise for all the other uses of asm() below.
Otherwise the test LGTM.
Cheers,
Gustavo
> + tcr2_present = ((idreg3 & 0xF) != 0);
> +
> + if (!tcr2_present) {
> + ml_printf("TCR2 is not present, cannot perform test");
> + return 0;
> + }
> +
> + asm("mrs %[x1], " ID_AA64MMFR4_EL1 "\n\t"
> + : [x1] "=r" (idreg4));
> +
> + asid2_present = ((idreg4 & 0xF00) != 0);
> +
> + asm("msr " TCR2_EL1 ", %[x0]\n\t"
> + "mrs %[x1], " TCR2_EL1 "\n\t"
> + : [x1] "=r" (out)
> + : [x0] "r" (in));
> +
> + if (asid2_present) {
> + if ((out & feature_mask) == in) {
> + ml_printf("OK\n");
> + return 0;
> + } else {
> + ml_printf("FAIL: ASID2 present, but read value %lx != "
> + "written value %lx\n",
> + out & feature_mask, in);
> + return 1;
> + }
> + } else {
> + if (out == 0) {
> + ml_printf("TCR2_EL1 reads as RES0 as expected\n");
> + return 0;
> + } else {
> + ml_printf("FAIL: ASID2, missing but read value %lx != 0\n",
> + out & feature_mask, in);
> + return 1;
> + }
> + }
> +}
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH V4 2/4] target/arm: Allow writes to FNG1, FNG0, A2
2025-12-02 14:29 ` Gustavo Romero
@ 2025-12-04 16:25 ` Jim MacArthur
0 siblings, 0 replies; 10+ messages in thread
From: Jim MacArthur @ 2025-12-04 16:25 UTC (permalink / raw)
To: Gustavo Romero; +Cc: qemu-devel, Richard Henderson
On Tue, 2 Dec 2025 at 14:29, Gustavo Romero <gustavo.romero@linaro.org> wrote:
> Afaics, we are not flushing the TLB here like we do for TCR_ELx (in vmsa_tcr_el12_write) before
> we call raw_write(). Since here we could be changing the A2 & friends bits, which can change
> the value of the ASID being using (like the TCR_ELx.A1 bit), I believe we should flush the TLB
> explicitly here like we do in vmsa_tcr_el12_write().
>
> @rth wdyt?
I agree; the guest could reasonably expect TLB entries previously in
use to stop being matched. In any case, I don't imagine flushing the
TLB here will be a significant performance problem. Good spot!
Jim
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2025-12-04 16:25 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-12-02 12:00 [PATCH V4 0/4] Basic ASID2 Support Jim MacArthur
2025-12-02 12:00 ` [PATCH V4 1/4] target/arm: Enable ID_AA64MMFR4_EL1 register Jim MacArthur
2025-12-02 14:28 ` Gustavo Romero
2025-12-02 12:00 ` [PATCH V4 2/4] target/arm: Allow writes to FNG1, FNG0, A2 Jim MacArthur
2025-12-02 14:29 ` Gustavo Romero
2025-12-04 16:25 ` Jim MacArthur
2025-12-02 12:00 ` [PATCH V4 3/4] target/arm/tcg/cpu64.c: Enable ASID2 for cpu_max Jim MacArthur
2025-12-02 14:29 ` Gustavo Romero
2025-12-02 12:00 ` [PATCH V4 4/4] tests: Add test for ASID2 and write/read of feature bits Jim MacArthur
2025-12-02 14:30 ` Gustavo Romero
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).