Re: [PATCH v14 05/44] arm64: RMI: Add wrappers for RMI calls

[Steven Price <steven.price@arm.com>]

On 21/05/2026 01:21, Gavin Shan wrote:
> Hi Steven,
> 
> On 5/13/26 11:17 PM, Steven Price wrote:
>> The wrappers make the call sites easier to read and deal with the
>> boiler plate of handling the error codes from the RMM.
>>
>> Signed-off-by: Steven Price <steven.price@arm.com>
>> ---
>> Changes from v13:
>>   * Update to RMM v2.0-bet1 spec including some SRO support (there still
>>     some FIXMEs where SRO support is incomplete).
>> Changes from v12:
>>   * Update to RMM v2.0 specification
>> Changes from v8:
>>   * Switch from arm_smccc_1_2_smc() to arm_smccc_1_2_invoke() in
>>     rmi_rtt_read_entry() for consistency.
>> Changes from v7:
>>   * Minor renaming of parameters and updated comments
>> Changes from v5:
>>   * Further improve comments
>> Changes from v4:
>>   * Improve comments
>> Changes from v2:
>>   * Make output arguments optional.
>>   * Mask RIPAS value rmi_rtt_read_entry()
>>   * Drop unused rmi_rtt_get_phys()
>> ---
>>   arch/arm64/include/asm/rmi_cmds.h | 661 ++++++++++++++++++++++++++++++
>>   1 file changed, 661 insertions(+)
>>   create mode 100644 arch/arm64/include/asm/rmi_cmds.h
>>
>> diff --git a/arch/arm64/include/asm/rmi_cmds.h b/arch/arm64/include/
>> asm/rmi_cmds.h
>> new file mode 100644
>> index 000000000000..04f7066894e9
>> --- /dev/null
>> +++ b/arch/arm64/include/asm/rmi_cmds.h
>> @@ -0,0 +1,661 @@
>> +/* SPDX-License-Identifier: GPL-2.0 */
>> +/*
>> + * Copyright (C) 2023 ARM Ltd.
>> + */
>> +
>> +#ifndef __ASM_RMI_CMDS_H
>> +#define __ASM_RMI_CMDS_H
>> +
>> +#include <linux/arm-smccc.h>
>> +
> 
> [...]
> 
>> +
>> +/**
>> + * rmi_rtt_destroy() - Destroy an RTT
>> + * @rd: PA of the RD
>> + * @ipa: Base of the IPA range described by the RTT
>> + * @level: Depth of the RTT within the tree
>> + * @out_rtt: Pointer to write the PA of the RTT which was destroyed
>> + * @out_top: Pointer to write the top IPA of non-live RTT entries
>> + *
> 
> In most cases, the parameters are well explained in RMM-v2.0-bet1 spec,
> I think
> it's nice to keep the code and the spec synchronized. For those specific
> parameters
> of this function, they're well explained in RMM-v2.0-bet1 spec as below.
> 
>    @rd: PA of the RD for the target realm
>    @ipa: Base of the IPA range described by the RTT
>    @level: RTT level
>    @out_rtt: PA of the RTT which was destroyed
>    @out_top: Top IPA of non-live RTT entries, from entry at which the
> RTT walk terminated

I have attempted to keep the descriptions consistent with the spec - I'm
not quite sure what you think the issue is here. The @rd parameter gains
a "for the target realm" - which isn't really very informative (clearly
rmi_rtt_destroy() is targetting the realm which is being passed into the
function). @level is less informative. @out_xxx are prefixed with
"Pointer to write the" because the C function does indeed take a pointer
for the output parameter to be written.

But fair enough I can align them more precisely. In some cases I've
written the code before the final spec wording has been available which
might explain some differences.

Thanks,
Steve

>> + * Destroys an RTT. The RTT must be non-live, i.e. none of the
>> entries in the
>> + * table are in ASSIGNED or TABLE state.
>> + *
>> + * Return: RMI return code.
>> + */
>> +static inline int rmi_rtt_destroy(unsigned long rd,
>> +                  unsigned long ipa,
>> +                  long level,
>> +                  unsigned long *out_rtt,
>> +                  unsigned long *out_top)
>> +{
>> +    struct arm_smccc_res res;
>> +
>> +    arm_smccc_1_1_invoke(SMC_RMI_RTT_DESTROY, rd, ipa, level, &res);
>> +
>> +    if (out_rtt)
>> +        *out_rtt = res.a1;
>> +    if (out_top)
>> +        *out_top = res.a2;
>> +
>> +    return res.a0;
>> +}
>> +
> 
> [...]
> 
> Thanks,
> Gavin
>