diff mbox series

[11/11] RISC-V: Add initial support for T-Head C906 and C910 CPUs

Message ID 20220906122243.1243354-12-christoph.muellner@vrull.eu (mailing list archive)
State New, archived
Headers show
Series Add support for the T-Head vendor extensions | expand

Commit Message

Christoph Müllner Sept. 6, 2022, 12:22 p.m. UTC 
From: Christoph Müllner <christoph.muellner@vrull.eu>

This patch adds the following T-Head CPUs to the list of known CPUs:
* C906
* C910

Selecting those CPUs will automatically enable the available
ISA extensions of the CPUs (incl. vendor extensions).

Signed-off-by: Christoph Müllner <christoph.muellner@vrull.eu>
---
 target/riscv/cpu.c          | 32 ++++++++++++++++++++++++++++++++
 target/riscv/cpu.h          |  3 +++
 target/riscv/cpu_vendorid.h |  6 ++++++
 3 files changed, 41 insertions(+)
 create mode 100644 target/riscv/cpu_vendorid.h

Comments

Richard Henderson Sept. 8, 2022, 7:46 a.m. UTC | #1 
On 9/6/22 13:22, Christoph Muellner wrote:
> +    DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C906,       rv64_thead_c906_cpu_init),
> +    DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C910,       rv64_thead_c906_cpu_init),

Why model both if they're identical?


r~
Christoph Müllner Sept. 8, 2022, 8:23 a.m. UTC | #2 
On Thu, Sep 8, 2022 at 9:46 AM Richard Henderson <
richard.henderson@linaro.org> wrote:

> On 9/6/22 13:22, Christoph Muellner wrote:
> > +    DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C906,
>  rv64_thead_c906_cpu_init),
> > +    DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C910,
>  rv64_thead_c906_cpu_init),
>
> Why model both if they're identical?
>

I figured that users might expect that (existence of "thead-c906" and
"thead-c910").
And using "thead-c9xx" feels like it would be regretted in the future.

Should I drop "thead-c910"?



>
>
> r~
>
Richard Henderson Sept. 8, 2022, 8:56 a.m. UTC | #3 
On 9/8/22 09:23, Christoph Müllner wrote:
> 
> 
> On Thu, Sep 8, 2022 at 9:46 AM Richard Henderson <richard.henderson@linaro.org 
> <mailto:richard.henderson@linaro.org>> wrote:
> 
>     On 9/6/22 13:22, Christoph Muellner wrote:
>      > +    DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C906,       rv64_thead_c906_cpu_init),
>      > +    DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C910,       rv64_thead_c906_cpu_init),
> 
>     Why model both if they're identical?
> 
> 
> I figured that users might expect that (existence of "thead-c906" and "thead-c910").
> And using "thead-c9xx" feels like it would be regretted in the future.
> 
> Should I drop "thead-c910"?

Quite possibly.  For Arm, we don't try to supply every cpu model, only those that differ 
in some substantial way.


r~
Christoph Müllner Sept. 8, 2022, 9:01 a.m. UTC | #4 
On Thu, Sep 8, 2022 at 10:56 AM Richard Henderson <
richard.henderson@linaro.org> wrote:

> On 9/8/22 09:23, Christoph Müllner wrote:
> >
> >
> > On Thu, Sep 8, 2022 at 9:46 AM Richard Henderson <
> richard.henderson@linaro.org
> > <mailto:richard.henderson@linaro.org>> wrote:
> >
> >     On 9/6/22 13:22, Christoph Muellner wrote:
> >      > +    DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C906,
>  rv64_thead_c906_cpu_init),
> >      > +    DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C910,
>  rv64_thead_c906_cpu_init),
> >
> >     Why model both if they're identical?
> >
> >
> > I figured that users might expect that (existence of "thead-c906" and
> "thead-c910").
> > And using "thead-c9xx" feels like it would be regretted in the future.
> >
> > Should I drop "thead-c910"?
>
> Quite possibly.  For Arm, we don't try to supply every cpu model, only
> those that differ
> in some substantial way.
>

Ok, will do.

Thanks!


>
>
> r~
>
diff mbox series

Patch

diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index 01d85f0f96..1db440e21f 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -194,6 +194,36 @@  static void rv64_sifive_e_cpu_init(Object *obj)
     cpu->cfg.mmu = false;
 }
 
+static void rv64_thead_c906_cpu_init(Object *obj)
+{
+    CPURISCVState *env = &RISCV_CPU(obj)->env;
+    RISCVCPU *cpu = RISCV_CPU(obj);
+
+    set_misa(env, MXL_RV64, RVI | RVM | RVA | RVF | RVD | RVC | RVS | RVU);
+    set_priv_version(env, PRIV_VERSION_1_10_0);
+
+    cpu->cfg.ext_g = true;
+    cpu->cfg.ext_c = true;
+    cpu->cfg.ext_u = true;
+    cpu->cfg.ext_s = true;
+    cpu->cfg.ext_icsr = true;
+    cpu->cfg.ext_zfh = true;
+    cpu->cfg.mmu = true;
+    cpu->cfg.ext_xtheadba = true;
+    cpu->cfg.ext_xtheadbb = true;
+    cpu->cfg.ext_xtheadbs = true;
+    cpu->cfg.ext_xtheadcmo = true;
+    cpu->cfg.ext_xtheadcondmov = true;
+    cpu->cfg.ext_xtheadfmemidx = true;
+    cpu->cfg.ext_xtheadmac = true;
+    cpu->cfg.ext_xtheadmemidx = true;
+    cpu->cfg.ext_xtheadmempair = true;
+    cpu->cfg.ext_xtheadsync = true;
+    cpu->cfg.ext_xtheadxmae = true;
+
+    cpu->cfg.mvendorid = THEAD_VENDOR_ID;
+}
+
 static void rv128_base_cpu_init(Object *obj)
 {
     if (qemu_tcg_mttcg_enabled()) {
@@ -1205,6 +1235,8 @@  static const TypeInfo riscv_cpu_type_infos[] = {
     DEFINE_CPU(TYPE_RISCV_CPU_SIFIVE_E51,       rv64_sifive_e_cpu_init),
     DEFINE_CPU(TYPE_RISCV_CPU_SIFIVE_U54,       rv64_sifive_u_cpu_init),
     DEFINE_CPU(TYPE_RISCV_CPU_SHAKTI_C,         rv64_sifive_u_cpu_init),
+    DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C906,       rv64_thead_c906_cpu_init),
+    DEFINE_CPU(TYPE_RISCV_CPU_THEAD_C910,       rv64_thead_c906_cpu_init),
     DEFINE_CPU(TYPE_RISCV_CPU_BASE128,          rv128_base_cpu_init),
 #endif
 };
diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
index 8b02f530a6..74b291b4e4 100644
--- a/target/riscv/cpu.h
+++ b/target/riscv/cpu.h
@@ -27,6 +27,7 @@ 
 #include "qom/object.h"
 #include "qemu/int128.h"
 #include "cpu_bits.h"
+#include "cpu_vendorid.h"
 
 #define TCG_GUEST_DEFAULT_MO 0
 
@@ -53,6 +54,8 @@ 
 #define TYPE_RISCV_CPU_SIFIVE_E51       RISCV_CPU_TYPE_NAME("sifive-e51")
 #define TYPE_RISCV_CPU_SIFIVE_U34       RISCV_CPU_TYPE_NAME("sifive-u34")
 #define TYPE_RISCV_CPU_SIFIVE_U54       RISCV_CPU_TYPE_NAME("sifive-u54")
+#define TYPE_RISCV_CPU_THEAD_C906       RISCV_CPU_TYPE_NAME("thead-c906")
+#define TYPE_RISCV_CPU_THEAD_C910       RISCV_CPU_TYPE_NAME("thead-c910")
 #define TYPE_RISCV_CPU_HOST             RISCV_CPU_TYPE_NAME("host")
 
 #if defined(TARGET_RISCV32)
diff --git a/target/riscv/cpu_vendorid.h b/target/riscv/cpu_vendorid.h
new file mode 100644
index 0000000000..a5aa249bc9
--- /dev/null
+++ b/target/riscv/cpu_vendorid.h
@@ -0,0 +1,6 @@ 
+#ifndef TARGET_RISCV_CPU_VENDORID_H
+#define TARGET_RISCV_CPU_VENDORID_H
+
+#define THEAD_VENDOR_ID         0x5b7
+
+#endif /*  TARGET_RISCV_CPU_VENDORID_H */