linux-kernel - Re: [Patch v5 2/4] memory: tegra: Add MC error logging on tegra186 onward

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 for Android: free password hash cracker in your pocket

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

Message-ID: <44235c65-160c-04c7-294d-16b13d25605c@nvidia.com>
Date:   Wed, 30 Mar 2022 15:46:59 +0530
From:   Ashish Mhetre <amhetre@...dia.com>
To:     Dmitry Osipenko <dmitry.osipenko@...labora.com>,
        Dmitry Osipenko <digetx@...il.com>,
        krzysztof.kozlowski@...onical.com, robh+dt@...nel.org,
        thierry.reding@...il.com, jonathanh@...dia.com,
        linux-kernel@...r.kernel.org, devicetree@...r.kernel.org,
        linux-tegra@...r.kernel.org
Cc:     vdumpa@...dia.com, Snikam@...dia.com
Subject: Re: [Patch v5 2/4] memory: tegra: Add MC error logging on tegra186
 onward



On 3/30/2022 5:31 AM, Dmitry Osipenko wrote:
> External email: Use caution opening links or attachments
> 
> 
> On 3/22/22 20:34, Ashish Mhetre wrote:
>>>> +     switch (status & mc->soc->int_channel_mask) {
>>>> +     case BIT(0):
>>>> +             *mc_channel = 0;
>>>> +             break;
>>>> +
>>>> +     case BIT(1):
>>>> +             *mc_channel = 1;
>>>> +             break;
>>>> +
>>>> +     case BIT(2):
>>>> +             *mc_channel = 2;
>>>> +             break;
>>>> +
>>>> +     case BIT(3):
>>>> +             *mc_channel = 3;
>>>> +             break;
>>>> +
>>>> +     case BIT(24):
>>>> +             *mc_channel = MC_BROADCAST_CHANNEL;
>>>> +             break;
>>>> +
>>>> +     default:
>>>> +             pr_err("Unknown interrupt source\n");
>>>
>>> dev_err_ratelimited("unknown interrupt channel 0x%08x\n", status) and
>>> should be moved to the common interrupt handler.
>>>
>> So return just error from default case and handle error in common
>> interrupt handler with this print, right? I'll update this in next
>> version.
> 
> Yes, just move out the common print.
> 
> Although, you could parameterize the shift per SoC and then have a
> common helper that does "status >> intmask_chan_shift", couldn't you?

Do you mean shift to get the channel, like
"channel = status >> intmask_chan_shift"?
So to get rid of this callback completely and adding a variable in
tegra_mc_soc for intmask_chan_shift, right? Or compute shift in this
callback and use it in common handler?
If we are to remove this callback then how to handle unknown interrupt
channel error?
Also we want to handle interrupts on one channel at a time and then
clear it from status register. There can be interrupts on multiple
channel. So multiple bits from status will be set. Hence it will be
hard to parameterize shift such that it gives appropriate channel.
So I think current approach is fine. Please correct me if I am wrong
somewhere.