linux-kernel - Re: [RFC] Make dynamic debug infrastructure more flexible

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: <11115196-54b0-5948-a6b1-87994239f2a7@akamai.com>
Date:   Thu, 21 May 2020 17:10:33 -0400
From:   Jason Baron <jbaron@...mai.com>
To:     Joe Perches <joe@...ches.com>
Cc:     Vikash Garodia <vgarodia@...eaurora.org>,
        Stanimir Varbanov <stanimir.varbanov@...aro.org>,
        linux-media@...r.kernel.org, linux-arm-msm@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [RFC] Make dynamic debug infrastructure more flexible



On 5/21/20 4:06 PM, Joe Perches wrote:
> On Thu, 2020-05-21 at 09:08 -0700, Joe Perches wrote:
>> On Thu, 2020-05-21 at 16:28 +0300, Stanimir Varbanov wrote:
>>> Here we introduce few debug macros with levels (low, medium and
>>> high) and debug macro for firmware. Enabling the particular level
>>> will be done by dynamic debug.
>>
>> I'd rather make the logging level facility generic in
>> dynamic debug than prefix all formats with what could
>> be non-specific content.
>>
>> From a long time ago:
>>
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__groups.google.com_forum_-23-21msg_linux.kernel_VlWbno-2DZAFw_k-5FfFadhNHXcJ&d=DwICAw&c=96ZbZZcaMF4w0F4jpN6LZg&r=1fLh1mlLqbfetnnGsbwXfpwmGlG4m83mXgtV4vZ1B1A&m=frs_z9MmQ_bzjMWZc5gnzdbuAHVi2EWsCO7ikUkszv4&s=Cbij9ptIrxbwB3XJNp8lxXrdTDi8T2s5XlB1llRbmjU&e= 
> 
> Hey Jason.
> 
> I believe there are 6 bits left in the unsigned int
> use for the line number and flags in struct _ddebug
> 
> Assuming the use of a mechanism like
> 
> 	pr_debug_level(level, fmt, ...)
> 
> would you be OK with something like this to enable a
> level or bitmask test of dynamic debug logging output?
> 
> where the output is controlled by something like
> 
> echo 'file <filename> level <n> +p' > <debugfs>/dynamic_debug/control
> 
> to enable dynamic debug output only at level <n> or higher
> or maybe match a bitmap of <n>
> 
> (modulo all the rest of the code necessary to use it?)

Hi Joe,

Yes, I think its good idea. I think in the past I felt that since
we could enable/disable statements individually it wasn't needed.
However, the ability to group in the kernel seems like a better
layer. I think we were also missing a specific use-case. Seems
like we have that here. So makes sense to me.

Thanks,

-Jason



> ---
>  include/linux/dynamic_debug.h | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/include/linux/dynamic_debug.h b/include/linux/dynamic_debug.h
> index abcd5fde30eb..616dbb2b5921 100644
> --- a/include/linux/dynamic_debug.h
> +++ b/include/linux/dynamic_debug.h
> @@ -38,6 +38,8 @@ struct _ddebug {
>  #define _DPRINTK_FLAGS_DEFAULT 0
>  #endif
>  	unsigned int flags:8;
> +	unsigned int level:5;
> +	unsigned int level_is_bitmask:1;
>  #ifdef CONFIG_JUMP_LABEL
>  	union {
>  		struct static_key_true dd_key_true;
> 
>