Re: [PATCH v11 03/26] gunyah: Common types and error codes for Gunyah hypercalls

[Elliot Berman <quic_eberman@quicinc.com>]

On 3/31/2023 7:24 AM, Alex Elder wrote:
> On 3/3/23 7:06 PM, Elliot Berman wrote:
>> Add architecture-independent standard error codes, types, and macros for
>> Gunyah hypercalls.
>>
>> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
>> Signed-off-by: Elliot Berman <quic_eberman@quicinc.com>
> 
> See a few comments below.    -Alex
> 
>> ---
>>   include/linux/gunyah.h | 83 ++++++++++++++++++++++++++++++++++++++++++
>>   1 file changed, 83 insertions(+)
>>   create mode 100644 include/linux/gunyah.h
>>
>> diff --git a/include/linux/gunyah.h b/include/linux/gunyah.h
>> new file mode 100644
>> index 000000000000..54b4be71caf7
>> --- /dev/null
>> +++ b/include/linux/gunyah.h
>> @@ -0,0 +1,83 @@
>> +/* SPDX-License-Identifier: GPL-2.0-only */
>> +/*
>> + * Copyright (c) 2022-2023 Qualcomm Innovation Center, Inc. All 
>> rights reserved.
>> + */
>> +
>> +#ifndef _LINUX_GUNYAH_H
>> +#define _LINUX_GUNYAH_H
>> +
>> +#include <linux/errno.h>
>> +#include <linux/limits.h>
>> +
>> +/******************************************************************************/
>> +/* Common arch-independent definitions for Gunyah 
>> hypercalls                  */
>> +#define GH_CAPID_INVAL    U64_MAX
>> +#define GH_VMID_ROOT_VM    0xff
> 
> The above definition doesn't seem to be used anywhere, but seeing
> it begs the question to me of what type it is expected to have.
> If it were used, where would it be used in an 8 bit field?
> 

VMIDs are u16, the root VM (Resource Manager) VMID is 0xff. I think this 
definition snuck in from the downstream code and is indeed not being 
needed/used anywhere. I'll remove it.

>> +
>> +enum gh_error {
>> +    GH_ERROR_OK            = 0,
>> +    GH_ERROR_UNIMPLEMENTED        = -1,
>> +    GH_ERROR_RETRY            = -2,
> 
> There might be nothing fundamentally wrong with this, but I
> dislike seeing negative values assigned to enums.
> 
> These error values are returned from the hypervisor, and it
> looks like they'll likely truncated from a 64-bit unsigned
> value.  Are they *sent* from the hypervisor as 64-bit signed
> values?  Or 32-bit signed values?  (In that case, the
> 
> I just wonder if you can use 0xffffffff or 0xffff for example
> rather than -1, depending on the actual value that gets passed.
> 

They are sent from the hypervisor as 64-bit signed values (it's filling 
a register). I think truncating should be OK because Gunyah wants to 
maintain capability with 32-bit architectures and we would not see an 
error number that truly requires more than 32 bits to represent.

>> +
>> +    GH_ERROR_ARG_INVAL        = 1,
>> +    GH_ERROR_ARG_SIZE        = 2,
>> +    GH_ERROR_ARG_ALIGN        = 3,
>> +
>> +    GH_ERROR_NOMEM            = 10,
>> +
>> +    GH_ERROR_ADDR_OVFL        = 20,
>> +    GH_ERROR_ADDR_UNFL        = 21,
>> +    GH_ERROR_ADDR_INVAL        = 22,
>> +
>> +    GH_ERROR_DENIED            = 30,
>> +    GH_ERROR_BUSY            = 31,
>> +    GH_ERROR_IDLE            = 32,
>> +
>> +    GH_ERROR_IRQ_BOUND        = 40,
>> +    GH_ERROR_IRQ_UNBOUND        = 41,
>> +
>> +    GH_ERROR_CSPACE_CAP_NULL    = 50,
>> +    GH_ERROR_CSPACE_CAP_REVOKED    = 51,
>> +    GH_ERROR_CSPACE_WRONG_OBJ_TYPE    = 52,
>> +    GH_ERROR_CSPACE_INSUF_RIGHTS    = 53,
>> +    GH_ERROR_CSPACE_FULL        = 54,
>> +
>> +    GH_ERROR_MSGQUEUE_EMPTY        = 60,
>> +    GH_ERROR_MSGQUEUE_FULL        = 61,
>> +};
>> +
>> +/**
>> + * gh_remap_error() - Remap Gunyah hypervisor errors into a Linux 
>> error code
>> + * @gh_error: Gunyah hypercall return value
>> + */
>> +static inline int gh_remap_error(enum gh_error gh_error)
> 
> Since you're remapping a gh_error, I would have named this
> gh_error_remap().
> 

Done.

>> +{
>> +    switch (gh_error) {
>> +    case GH_ERROR_OK:
>> +        return 0;
>> +    case GH_ERROR_NOMEM:
>> +        return -ENOMEM;
>> +    case GH_ERROR_DENIED:
>> +    case GH_ERROR_CSPACE_CAP_NULL:
>> +    case GH_ERROR_CSPACE_CAP_REVOKED:
>> +    case GH_ERROR_CSPACE_WRONG_OBJ_TYPE:
>> +    case GH_ERROR_CSPACE_INSUF_RIGHTS:
>> +    case GH_ERROR_CSPACE_FULL:
>> +        return -EACCES;
>> +    case GH_ERROR_BUSY:
>> +    case GH_ERROR_IDLE:
>> +        return -EBUSY;
>> +    case GH_ERROR_IRQ_BOUND:
>> +    case GH_ERROR_IRQ_UNBOUND:
>> +    case GH_ERROR_MSGQUEUE_FULL:
>> +    case GH_ERROR_MSGQUEUE_EMPTY:
>> +        return -EIO;
>> +    case GH_ERROR_UNIMPLEMENTED:
>> +    case GH_ERROR_RETRY:
>> +        return -EOPNOTSUPP;
>> +    default:
>> +        return -EINVAL;
>> +    }
>> +}
>> +
>> +#endif
>