linux-hardening - Re: [PATCH v4 1/2] dt-bindings: ramoops: Inherit reserve memory property

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] [day] [month] [year] [list]

Message-ID: <7ef34bc2-f8ef-72e1-b97e-7c274b671550@quicinc.com>
Date:   Mon, 30 Jan 2023 14:02:05 +0530
From:   Mukesh Ojha <quic_mojha@...cinc.com>
To:     Rob Herring <robh+dt@...nel.org>
CC:     <linux-hardening@...r.kernel.org>, <linux-doc@...r.kernel.org>,
        <linux-kernel@...r.kernel.org>, <devicetree@...r.kernel.org>,
        Kees Cook <keescook@...omium.org>,
        Tony Luck <tony.luck@...el.com>,
        "Guilherme G . Piccoli" <gpiccoli@...lia.com>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>
Subject: Re: [PATCH v4 1/2] dt-bindings: ramoops: Inherit reserve memory
 property



On 1/28/2023 1:35 AM, Rob Herring wrote:
> On Fri, Jan 27, 2023 at 10:01 AM Mukesh Ojha <quic_mojha@...cinc.com> wrote:
>>
>> The reserved memory region for ramoops is assumed to be at a
>> fixed and known location when read from the devicetree. This
>> is not desirable in an environment where it is preferred the
>> region to be dynamically allocated at runtime, as opposed to
>> being fixed at compile time.
> 
> I don't see how this can work. How do you find the region after
> rebooting? Or this is only used for the current boot? If so, provide a
> userspace interface to configure it (or the existing module params).
> 

Client can do of_reserved_mem_lookup() on this ramoops node to get 
region addresses and it is for regular boots and it is not for
current boot.

Basically, here we are trying to support system which does not 
completely support warm reset(content could be inconsistent across 
boots) and has their own mechanism to dump ramoops region on crash.

Also, If it is required, to override the existing ramoops region of 
current boot(from static to dynamic), would provide user space interface 
as well.

> The addition of 'size' type regions was primarily for large carveouts
> that needed to be allocated before anything else. ramoops is not that.
> It's 10s or 100s of KB at most.

I agree..

Is this the only issue ?

I thought if it is inheriting reserve-memory binding, it can be used if 
it is not breaking anything what ramoops already supports..
you can have a look at 2/2 patch.


-Mukesh

> 
> 
>> So, update the ramoops binding by inheriting some reserve memory
>> property to allocate the ramoops region dynamically.
>>
>> Cc: Kees Cook <keescook@...omium.org>
>> Cc: Tony Luck <tony.luck@...el.com>
>> Cc: Guilherme G. Piccoli <gpiccoli@...lia.com>
>> Cc: Rob Herring <robh+dt@...nel.org>
>> Cc: Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>
>> Signed-off-by: Mukesh Ojha <quic_mojha@...cinc.com>
>> ---
>> Changes in v4:
>>   - Addressed comment made by Krzysztof on ramoops node name.
>>
>> Changes in v3:
>>   - Fixed yaml error and updated commit text as per comment.
>>
>> Change in v2:
>>    - Added this patch as per changes going to be done in patch 3/3
>>
>>   .../bindings/reserved-memory/ramoops.yaml          | 34 ++++++++++++++++++++--
>>   1 file changed, 32 insertions(+), 2 deletions(-)
>>
>> diff --git a/Documentation/devicetree/bindings/reserved-memory/ramoops.yaml b/Documentation/devicetree/bindings/reserved-memory/ramoops.yaml
>> index 0391871..8741626 100644
>> --- a/Documentation/devicetree/bindings/reserved-memory/ramoops.yaml
>> +++ b/Documentation/devicetree/bindings/reserved-memory/ramoops.yaml
>> @@ -10,7 +10,8 @@ description: |
>>     ramoops provides persistent RAM storage for oops and panics, so they can be
>>     recovered after a reboot. This is a child-node of "/reserved-memory", and
>>     is named "ramoops" after the backend, rather than "pstore" which is the
>> -  subsystem.
>> +  subsystem. This region can be reserved both statically or dynamically by
>> +  using appropriate property in device tree.
>>
>>     Parts of this storage may be set aside for other persistent log buffers, such
>>     as kernel log messages, or for optional ECC error-correction data.  The total
>> @@ -112,7 +113,13 @@ unevaluatedProperties: false
>>
>>   required:
>>     - compatible
>> -  - reg
>> +
>> +oneOf:
>> +  - required:
>> +      - reg
>> +
>> +  - required:
>> +      - size
>>
>>   anyOf:
>>     - required: [record-size]
>> @@ -142,3 +149,26 @@ examples:
>>               };
>>           };
>>       };
>> +
>> +  - |
>> +    / {
> 
> You can't have multiple root node examples. Check the example dts, the
> examples will be merged together.

Sure, will remove this..

> 
>> +        compatible = "foo";
>> +        model = "foo";
>> +        #address-cells = <1>;
>> +        #size-cells = <1>;
>> +
>> +        reserved-memory {
>> +            #address-cells = <1>;
>> +            #size-cells = <1>;
>> +            ranges;
>> +
>> +            ramoops_region: ramoops {
>> +                compatible = "ramoops";
>> +                alloc-ranges = <0x00000000 0xffffffff>;
>> +                size = <0x0 0x10000>;       /* 64kB */
>> +                console-size = <0x8000>;    /* 32kB */
>> +                record-size = <0x400>;      /*  1kB */
>> +                ecc-size = <16>;
>> +            };
>> +        };
>> +    };
>> --
>> 2.7.4
>>