target/riscv: Use TARGET_FMT_lx for env->mhartid

[PATCH] target/riscv: Use TARGET_FMT_lx for env->mhartid

Posted by Bin Meng 1 year, 3 months ago

env->mhartid is currently casted to long before printed, which drops
the high 32-bit for rv64 on 32-bit host. Use TARGET_FMT_lx instead.

Signed-off-by: Bin Meng <bmeng@tinylab.org>
---

 target/riscv/cpu.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index cc75ca7667..a5ed6d3f63 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -660,9 +660,9 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
             (env->priv_ver < isa_edata_arr[i].min_version)) {
             isa_ext_update_enabled(cpu, &isa_edata_arr[i], false);
 #ifndef CONFIG_USER_ONLY
-            warn_report("disabling %s extension for hart 0x%lx because "
-                        "privilege spec version does not match",
-                        isa_edata_arr[i].name, (unsigned long)env->mhartid);
+            warn_report("disabling %s extension for hart 0x" TARGET_FMT_lx
+                        " because privilege spec version does not match",
+                        isa_edata_arr[i].name, env->mhartid);
 #else
             warn_report("disabling %s extension because "
                         "privilege spec version does not match",
-- 
2.34.1

Re: [PATCH] target/riscv: Use TARGET_FMT_lx for env->mhartid

Posted by Alistair Francis 1 year, 3 months ago

On Tue, Jan 10, 2023 at 1:28 AM Bin Meng <bmeng@tinylab.org> wrote:
>
> env->mhartid is currently casted to long before printed, which drops
> the high 32-bit for rv64 on 32-bit host. Use TARGET_FMT_lx instead.
>
> Signed-off-by: Bin Meng <bmeng@tinylab.org>

Thanks!

Applied to riscv-to-apply.next

Alistair

> ---
>
>  target/riscv/cpu.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
> index cc75ca7667..a5ed6d3f63 100644
> --- a/target/riscv/cpu.c
> +++ b/target/riscv/cpu.c
> @@ -660,9 +660,9 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
>              (env->priv_ver < isa_edata_arr[i].min_version)) {
>              isa_ext_update_enabled(cpu, &isa_edata_arr[i], false);
>  #ifndef CONFIG_USER_ONLY
> -            warn_report("disabling %s extension for hart 0x%lx because "
> -                        "privilege spec version does not match",
> -                        isa_edata_arr[i].name, (unsigned long)env->mhartid);
> +            warn_report("disabling %s extension for hart 0x" TARGET_FMT_lx
> +                        " because privilege spec version does not match",
> +                        isa_edata_arr[i].name, env->mhartid);
>  #else
>              warn_report("disabling %s extension because "
>                          "privilege spec version does not match",
> --
> 2.34.1
>
>

Re: [PATCH] target/riscv: Use TARGET_FMT_lx for env->mhartid

Posted by Alistair Francis 1 year, 3 months ago

On Tue, Jan 10, 2023 at 1:28 AM Bin Meng <bmeng@tinylab.org> wrote:
>
> env->mhartid is currently casted to long before printed, which drops
> the high 32-bit for rv64 on 32-bit host. Use TARGET_FMT_lx instead.
>
> Signed-off-by: Bin Meng <bmeng@tinylab.org>

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>

Alistair

> ---
>
>  target/riscv/cpu.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
> index cc75ca7667..a5ed6d3f63 100644
> --- a/target/riscv/cpu.c
> +++ b/target/riscv/cpu.c
> @@ -660,9 +660,9 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
>              (env->priv_ver < isa_edata_arr[i].min_version)) {
>              isa_ext_update_enabled(cpu, &isa_edata_arr[i], false);
>  #ifndef CONFIG_USER_ONLY
> -            warn_report("disabling %s extension for hart 0x%lx because "
> -                        "privilege spec version does not match",
> -                        isa_edata_arr[i].name, (unsigned long)env->mhartid);
> +            warn_report("disabling %s extension for hart 0x" TARGET_FMT_lx
> +                        " because privilege spec version does not match",
> +                        isa_edata_arr[i].name, env->mhartid);
>  #else
>              warn_report("disabling %s extension because "
>                          "privilege spec version does not match",
> --
> 2.34.1
>
>

Re: [PATCH] target/riscv: Use TARGET_FMT_lx for env->mhartid

Posted by Philippe Mathieu-Daudé 1 year, 3 months ago

On 9/1/23 16:26, Bin Meng wrote:
> env->mhartid is currently casted to long before printed, which drops
> the high 32-bit for rv64 on 32-bit host. Use TARGET_FMT_lx instead.

Oh, a 32-bit host user!

> Signed-off-by: Bin Meng <bmeng@tinylab.org>
> ---
> 
>   target/riscv/cpu.c | 6 +++---
>   1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
> index cc75ca7667..a5ed6d3f63 100644
> --- a/target/riscv/cpu.c
> +++ b/target/riscv/cpu.c
> @@ -660,9 +660,9 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
>               (env->priv_ver < isa_edata_arr[i].min_version)) {
>               isa_ext_update_enabled(cpu, &isa_edata_arr[i], false);
>   #ifndef CONFIG_USER_ONLY
> -            warn_report("disabling %s extension for hart 0x%lx because "
> -                        "privilege spec version does not match",
> -                        isa_edata_arr[i].name, (unsigned long)env->mhartid);
> +            warn_report("disabling %s extension for hart 0x" TARGET_FMT_lx
> +                        " because privilege spec version does not match",
> +                        isa_edata_arr[i].name, env->mhartid);

Could we cast it to vaddr instead? I'm trying to remove target_[u]long
from hw/ and restrict it to the target/ directory. Per "exec/cpu-common.h":

  /**
   * vaddr:
   * Type wide enough to contain any #target_ulong virtual address.
   */
  typedef uint64_t vaddr;

Alternatively, since this value has to be accessed out of target/,
can we change its type to vaddr in CPURISCVState?

Thanks,

Phil.

Re: [PATCH] target/riscv: Use TARGET_FMT_lx for env->mhartid

Posted by Richard Henderson 1 year, 3 months ago

On 1/9/23 07:47, Philippe Mathieu-Daudé wrote:
> On 9/1/23 16:26, Bin Meng wrote:
>> env->mhartid is currently casted to long before printed, which drops
>> the high 32-bit for rv64 on 32-bit host. Use TARGET_FMT_lx instead.
> 
> Oh, a 32-bit host user!
> 
>> Signed-off-by: Bin Meng <bmeng@tinylab.org>
>> ---
>>
>>   target/riscv/cpu.c | 6 +++---
>>   1 file changed, 3 insertions(+), 3 deletions(-)
>>
>> diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
>> index cc75ca7667..a5ed6d3f63 100644
>> --- a/target/riscv/cpu.c
>> +++ b/target/riscv/cpu.c
>> @@ -660,9 +660,9 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
>>               (env->priv_ver < isa_edata_arr[i].min_version)) {
>>               isa_ext_update_enabled(cpu, &isa_edata_arr[i], false);
>>   #ifndef CONFIG_USER_ONLY
>> -            warn_report("disabling %s extension for hart 0x%lx because "
>> -                        "privilege spec version does not match",
>> -                        isa_edata_arr[i].name, (unsigned long)env->mhartid);
>> +            warn_report("disabling %s extension for hart 0x" TARGET_FMT_lx
>> +                        " because privilege spec version does not match",
>> +                        isa_edata_arr[i].name, env->mhartid);
> 
> Could we cast it to vaddr instead? I'm trying to remove target_[u]long
> from hw/ and restrict it to the target/ directory. Per "exec/cpu-common.h":

This isn't in hw/, it's in target/.  And no, it's a target-width register.


r~

Re: [PATCH] target/riscv: Use TARGET_FMT_lx for env->mhartid

Posted by Bin Meng 1 year, 3 months ago

On Mon, Jan 9, 2023 at 11:48 PM Philippe Mathieu-Daudé
<philmd@linaro.org> wrote:
>
> On 9/1/23 16:26, Bin Meng wrote:
> > env->mhartid is currently casted to long before printed, which drops
> > the high 32-bit for rv64 on 32-bit host. Use TARGET_FMT_lx instead.
>
> Oh, a 32-bit host user!
>
> > Signed-off-by: Bin Meng <bmeng@tinylab.org>
> > ---
> >
> >   target/riscv/cpu.c | 6 +++---
> >   1 file changed, 3 insertions(+), 3 deletions(-)
> >
> > diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
> > index cc75ca7667..a5ed6d3f63 100644
> > --- a/target/riscv/cpu.c
> > +++ b/target/riscv/cpu.c
> > @@ -660,9 +660,9 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
> >               (env->priv_ver < isa_edata_arr[i].min_version)) {
> >               isa_ext_update_enabled(cpu, &isa_edata_arr[i], false);
> >   #ifndef CONFIG_USER_ONLY
> > -            warn_report("disabling %s extension for hart 0x%lx because "
> > -                        "privilege spec version does not match",
> > -                        isa_edata_arr[i].name, (unsigned long)env->mhartid);
> > +            warn_report("disabling %s extension for hart 0x" TARGET_FMT_lx
> > +                        " because privilege spec version does not match",
> > +                        isa_edata_arr[i].name, env->mhartid);
>
> Could we cast it to vaddr instead? I'm trying to remove target_[u]long
> from hw/ and restrict it to the target/ directory. Per "exec/cpu-common.h":
>
>   /**
>    * vaddr:
>    * Type wide enough to contain any #target_ulong virtual address.
>    */
>   typedef uint64_t vaddr;
>
> Alternatively, since this value has to be accessed out of target/,
> can we change its type to vaddr in CPURISCVState?
>

Technically it does not represent a virtual address but a target
dependent register that can be 32-bit or 64-bit. Change env->mhartid
to vaddr looks weird to me.

Regards,
Bin