[v1] Add ETE devarch condition in etm4_init_iomem_access

[PATCH] [Patch v1]Add ETE devarch condition in etm4_init_iomem_access

Posted by yiru zhang 1 month ago

Due to ETE supported, so add ETE devarch condition in etm4_init_iomem_access.
Signed-off-by: yiru zhang <yiru.zhang@mediatek.com>
---
 drivers/hwtracing/coresight/coresight-etm4x-core.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c b/drivers/hwtracing/coresight/coresight-etm4x-core.c
index 2b8f10463840..971b9f0fe5e4 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x-core.c
+++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c
@@ -1135,8 +1135,9 @@ static bool etm4_init_iomem_access(struct etmv4_drvdata *drvdata,
 	 * with MMIO. But we cannot touch the OSLK until we are
 	 * sure this is an ETM. So rely only on the TRCDEVARCH.
 	 */
-	if ((devarch & ETM_DEVARCH_ID_MASK) != ETM_DEVARCH_ETMv4x_ARCH) {
-		pr_warn_once("TRCDEVARCH doesn't match ETMv4 architecture\n");
+	if ((devarch & ETM_DEVARCH_ID_MASK) != ETM_DEVARCH_ETMv4x_ARCH \
+	&& (devarch & ETM_DEVARCH_ID_MASK) != ETM_DEVARCH_ETE_ARCH) {
+		pr_warn_once("TRCDEVARCH doesn't match ETMv4&ETE architecture\n");
 		return false;
 	}
 
-- 
2.46.0

Re: [PATCH] [Patch v1]Add ETE devarch condition in etm4_init_iomem_access

Posted by Suzuki K Poulose 1 month ago

Hi Yiru



On 08/04/2025 08:36, yiru zhang wrote:
> Due to ETE supported, so add ETE devarch condition in etm4_init_iomem_access.

Is there a reason why you cannot use the "system instructions to access 
the ETE" ?

The patch as such is fine by me (withe some minor styling nits). But, we
do not recommend using the MMIO for ETE, when it can be accessed
directly by sysreg.

FWIW, if you remove the "mmio base" address from the DT/ACPI, it
should automatically use the system instructions for ETE.




> Signed-off-by: yiru zhang <yiru.zhang@mediatek.com>
> ---
>   drivers/hwtracing/coresight/coresight-etm4x-core.c | 5 +++--
>   1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c b/drivers/hwtracing/coresight/coresight-etm4x-core.c
> index 2b8f10463840..971b9f0fe5e4 100644
> --- a/drivers/hwtracing/coresight/coresight-etm4x-core.c
> +++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c
> @@ -1135,8 +1135,9 @@ static bool etm4_init_iomem_access(struct etmv4_drvdata *drvdata,
>   	 * with MMIO. But we cannot touch the OSLK until we are
>   	 * sure this is an ETM. So rely only on the TRCDEVARCH.
>   	 */
> -	if ((devarch & ETM_DEVARCH_ID_MASK) != ETM_DEVARCH_ETMv4x_ARCH) {
> -		pr_warn_once("TRCDEVARCH doesn't match ETMv4 architecture\n");
> +	if ((devarch & ETM_DEVARCH_ID_MASK) != ETM_DEVARCH_ETMv4x_ARCH \
> +	&& (devarch & ETM_DEVARCH_ID_MASK) != ETM_DEVARCH_ETE_ARCH) {

We could use a switch case ?

	switch (devarch & ETM_DEVARCH_ID_MASK) {
	case ETM_DEVARCH_ETMv4x_ARCH:
	case ETM_DEVARCH_ETE_ARCH:
		break;
	default:
		pr_warn_once("Unknown ETM architecture: %x\n",
				devarch & ETM_DEVARCH_ID_MASK);
		return false;
	}

Suzuki

> +		pr_warn_once("TRCDEVARCH doesn't match ETMv4&ETE architecture\n");
>   		return false;
>   	}
>

Re: [PATCH] [Patch v1]Add ETE devarch condition in etm4_init_iomem_access

Posted by Yiru Zhang (张熠儒) 1 month ago

hi Suzuki, thanks for your reply.
If no ETE condition in etm4_init_iomem_access, the function will return
false and influence following process.
I will send v2 patch using a switch case.

Thanks,
Yiru

On Tue, 2025-04-08 at 10:05 +0100, Suzuki K Poulose wrote:
> External email : Please do not click links or open attachments until
> you have verified the sender or the content.
> 
> 
> Hi Yiru
> 
> 
> 
> On 08/04/2025 08:36, yiru zhang wrote:
> > Due to ETE supported, so add ETE devarch condition in
> > etm4_init_iomem_access.
> 
> Is there a reason why you cannot use the "system instructions to
> access
> the ETE" ?
> 
> The patch as such is fine by me (withe some minor styling nits). But,
> we
> do not recommend using the MMIO for ETE, when it can be accessed
> directly by sysreg.
> 
> FWIW, if you remove the "mmio base" address from the DT/ACPI, it
> should automatically use the system instructions for ETE.
> 
> 
> 
> 
> > Signed-off-by: yiru zhang <yiru.zhang@mediatek.com>
> > ---
> >   drivers/hwtracing/coresight/coresight-etm4x-core.c | 5 +++--
> >   1 file changed, 3 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c
> > b/drivers/hwtracing/coresight/coresight-etm4x-core.c
> > index 2b8f10463840..971b9f0fe5e4 100644
> > --- a/drivers/hwtracing/coresight/coresight-etm4x-core.c
> > +++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c
> > @@ -1135,8 +1135,9 @@ static bool etm4_init_iomem_access(struct
> > etmv4_drvdata *drvdata,
> >        * with MMIO. But we cannot touch the OSLK until we are
> >        * sure this is an ETM. So rely only on the TRCDEVARCH.
> >        */
> > -     if ((devarch & ETM_DEVARCH_ID_MASK) !=
> > ETM_DEVARCH_ETMv4x_ARCH) {
> > -             pr_warn_once("TRCDEVARCH doesn't match ETMv4
> > architecture\n");
> > +     if ((devarch & ETM_DEVARCH_ID_MASK) !=
> > ETM_DEVARCH_ETMv4x_ARCH \
> > +     && (devarch & ETM_DEVARCH_ID_MASK) != ETM_DEVARCH_ETE_ARCH) {
> 
> We could use a switch case ?
> 
>         switch (devarch & ETM_DEVARCH_ID_MASK) {
>         case ETM_DEVARCH_ETMv4x_ARCH:
>         case ETM_DEVARCH_ETE_ARCH:
>                 break;
>         default:
>                 pr_warn_once("Unknown ETM architecture: %x\n",
>                                 devarch & ETM_DEVARCH_ID_MASK);
>                 return false;
>         }
> 
> Suzuki
> 
> > +             pr_warn_once("TRCDEVARCH doesn't match ETMv4&ETE
> > architecture\n");
> >               return false;
> >       }
> > 
> 
>