linux-kernel - Re: [PATCH v4 2/8] scpi: Add alternative legacy structures, functions and macros

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <8f47801c-558d-a270-b66e-a7da1b4d7a60@arm.com>
Date:   Wed, 19 Oct 2016 11:42:00 +0100
From:   Sudeep Holla <sudeep.holla@....com>
To:     Neil Armstrong <narmstrong@...libre.com>,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Cc:     Sudeep Holla <sudeep.holla@....com>,
        linux-amlogic@...ts.infradead.org, khilman@...libre.com,
        heiko@...ech.de, wxt@...k-chips.com, frank.wang@...k-chips.com
Subject: Re: [PATCH v4 2/8] scpi: Add alternative legacy structures, functions
 and macros



On 19/10/16 11:37, Sudeep Holla wrote:
>
>
> On 19/10/16 11:28, Neil Armstrong wrote:
>> On 10/17/2016 01:16 PM, Sudeep Holla wrote:
>
> [...]
>
>>>
>>> and the above 2 can be moved out of the conditions, no ?
>>>
>>> if (scpi_info->is_legacy) {
>>>     struct legacy_scpi_shared_mem *mem = ch->rx_payload;
>>>     len = match->rx_len;
>>> } else {
>>>     struct scpi_shared_mem *mem = ch->rx_payload;
>>>     len = min(match->rx_len, CMD_SIZE(cmd));
>>> }
>>> match->status = le32_to_cpu(mem->status);
>>> memcpy_fromio(match->rx_buf, mem->payload, len);
>>>
>>> should work.
>>
>> Well, we will have "error: ‘mem’ undeclared (first use in this
>> function)" since mem is not declared outside the if/else.
>>
>> I don't see good solutions even with an union.
>
> Right, I missed to see that. You can leave it as you had before then.
>
>
> [...]
>>>>>>      if (t->rx_buf) {
>>>>>>          if (!(++ch->token))
>>>>>>              ++ch->token;
>>>>>>          ADD_SCPI_TOKEN(t->cmd, ch->token);
>>>>>> +        if (scpi_info->is_legacy)
>>>>>> +            t->slot = t->cmd;
>>>>>
>>>>> I thought passing token was not an issue from your previous response,
>>>>> but you are overriding it here, why ?
>>>>
>>>> Indeed, I can leave it, but it's useless since it won't serve to
>>>> distinguish multiple similar commands.
>>>>
>>>
>>> OK, I don't see any point in such micro optimization, so please
>>> retain it.
>>>
>>
>> I misread my code, I leaved the token passing, but I copy back the
>> cmd  to the slot which is used by the MHU.
>> If I remove the "t->slot = t->cmd;", the token won't be passed to the
>> FW.
>>
>
> Right, sorry I think I misled you :)
>

On the other hand, if we are unable to use that in the driver as you
just have FIFO in the list for this legacy mode, I think we can drop it
completely. I know I had argued otherwise before, but that was before I
realized that we need to keep FIFO in the list ;). Let me know if you
are OK to drop it.

-- 
Regards,
Sudeep