diff mbox series

arm64/sysreg: Update ID_AA64DFR0_EL1 and ID_DFR0_EL1

Message ID 20240215055159.2440898-1-anshuman.khandual@arm.com (mailing list archive)
State New, archived
Headers show
Series arm64/sysreg: Update ID_AA64DFR0_EL1 and ID_DFR0_EL1 | expand

Commit Message

Anshuman Khandual Feb. 15, 2024, 5:51 a.m. UTC
This just updates ID_DFR0_EL1.CopDbg and ID_AA64DFR0_EL1.DebugVer register
fields as per the definitions based on DDI0601 2023-12.

Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
 arch/arm64/tools/sysreg | 2 ++
 1 file changed, 2 insertions(+)

Comments

Mark Brown Feb. 15, 2024, 1 p.m. UTC | #1
On Thu, Feb 15, 2024 at 11:21:59AM +0530, Anshuman Khandual wrote:

> This just updates ID_DFR0_EL1.CopDbg and ID_AA64DFR0_EL1.DebugVer register
> fields as per the definitions based on DDI0601 2023-12.

Just as a general note for non-trival registers it's easier to review if
we change one register per patch, it avoids having to re-review the same
registers repeatedly and for incremental changes like this it helps with
the fact that diff does badly at offering context.

> --- a/arch/arm64/tools/sysreg
> +++ b/arch/arm64/tools/sysreg
> @@ -231,6 +231,7 @@ Enum	3:0	CopDbg
>  	0b1000	Debugv8p2
>  	0b1001	Debugv8p4
>  	0b1010	Debugv8p8
> +	0b1011	Debugv8p9
>  EndEnum
>  EndSysreg

This is ID_AA64DFR0_EL1 and is missing at least an update to PMSVer.

>  
> @@ -1247,6 +1248,7 @@ UnsignedEnum	3:0	DebugVer
>  	0b1000	V8P2
>  	0b1001	V8P4
>  	0b1010	V8P8
> +	0b1011	V8P9
>  EndEnum
>  EndSysreg

This is ID_DFR0_EL1 and is missing an update to at least PerfMon.
Anshuman Khandual Feb. 18, 2024, 12:54 a.m. UTC | #2
On 2/15/24 18:30, Mark Brown wrote:
> On Thu, Feb 15, 2024 at 11:21:59AM +0530, Anshuman Khandual wrote:
> 
>> This just updates ID_DFR0_EL1.CopDbg and ID_AA64DFR0_EL1.DebugVer register
>> fields as per the definitions based on DDI0601 2023-12.
> 
> Just as a general note for non-trival registers it's easier to review if
> we change one register per patch, it avoids having to re-review the same
> registers repeatedly and for incremental changes like this it helps with
> the fact that diff does badly at offering context.

Agreed, and noted.

> 
>> --- a/arch/arm64/tools/sysreg
>> +++ b/arch/arm64/tools/sysreg
>> @@ -231,6 +231,7 @@ Enum	3:0	CopDbg
>>  	0b1000	Debugv8p2
>>  	0b1001	Debugv8p4
>>  	0b1010	Debugv8p8
>> +	0b1011	Debugv8p9
>>  EndEnum
>>  EndSysreg
> 
> This is ID_AA64DFR0_EL1 and is missing at least an update to PMSVer.
> 
>>  
>> @@ -1247,6 +1248,7 @@ UnsignedEnum	3:0	DebugVer
>>  	0b1000	V8P2
>>  	0b1001	V8P4
>>  	0b1010	V8P8
>> +	0b1011	V8P9
>>  EndEnum
>>  EndSysreg
> 
> This is ID_DFR0_EL1 and is missing an update to at least PerfMon.

Will fold in the following changes i.e for ID_DFR0_EL1 and for
ID_AA64DFR0_EL1 respectively and then split the patch into two
- one for each register.

diff --git a/arch/arm64/tools/sysreg b/arch/arm64/tools/sysreg
index cc2914d6ebdc..508224a0e078 100644
--- a/arch/arm64/tools/sysreg
+++ b/arch/arm64/tools/sysreg
@@ -200,6 +200,7 @@ UnsignedEnum        27:24   PerfMon
        0b0110  PMUv3p5
        0b0111  PMUv3p7
        0b1000  PMUv3p8
+       0b1001  PMUv3p9
        0b1111  IMPDEF
 EndEnum
 Enum   23:20   MProfDbg
@@ -1222,6 +1223,7 @@ UnsignedEnum      35:32   PMSVer
        0b0010  V1P1
        0b0011  V1P2
        0b0100  V1P3
+       0b0101  V1P4
 EndEnum
 Field  31:28   CTX_CMPs
 Res0   27:24
diff mbox series

Patch

diff --git a/arch/arm64/tools/sysreg b/arch/arm64/tools/sysreg
index 4c9b67934367..94692abfeeb9 100644
--- a/arch/arm64/tools/sysreg
+++ b/arch/arm64/tools/sysreg
@@ -231,6 +231,7 @@  Enum	3:0	CopDbg
 	0b1000	Debugv8p2
 	0b1001	Debugv8p4
 	0b1010	Debugv8p8
+	0b1011	Debugv8p9
 EndEnum
 EndSysreg
 
@@ -1247,6 +1248,7 @@  UnsignedEnum	3:0	DebugVer
 	0b1000	V8P2
 	0b1001	V8P4
 	0b1010	V8P8
+	0b1011	V8P9
 EndEnum
 EndSysreg