Re: [PATCH, v2] [AArch64] MTE corefile support

[Luis Machado <luis.machado@arm.com>]

On 4/22/22 09:43, Eli Zaretskii wrote:
>> Authentication-Results-Original: dkim=none (message not signed)
>>   header.d=none;dmarc=none action=none header.from=arm.com;
>> Date: Fri, 22 Apr 2022 09:37:47 +0100
>> Cc: gdb-patches@sourceware.org
>> From: Luis Machado <luis.machado@arm.com>
>> NoDisclaimer: true
>> Original-Authentication-Results: dkim=none (message not signed)
>>   header.d=none;dmarc=none action=none header.from=arm.com;
>>
>> On 4/22/22 09:30, Eli Zaretskii wrote:
>>>> Date: Fri, 22 Apr 2022 09:12:44 +0100
>>>> Cc: gdb-patches@sourceware.org
>>>> From: Luis Machado <luis.machado@arm.com>
>>>>
>>>>>>     *** Changes since GDB 12
>>>>>>     
>>>>>> +* GDB now supports dumping memory tag data for AArch64 MTE.  It also supports
>>>>>> +  reading memory tag data for AArch64 MTE from core files generated by
>>>>>> +  the gcore command or the Linux kernel.
>>>>>> +
>>>>>
>>>>> I think this should say something about the significance of this
>>>>> feature.  Otherwise it is completely unclear why would GDB want to
>>>>> support that.
>>>>>
>>>>> Same comment for the addition to the manual.
>>>>>
>>>>> Thanks.
>>>>
>>>> It is a bit short. How about the following additional paragraph?
>>>>
>>>> "When a process uses memory-mapped pages protected by memory tags (for
>>>> example, AArch64 MTE), this additional information will be recorded in
>>>> the core file in the event of a crash or if GDB generates a core file
>>>> from the current process state.
>>>>
>>>> The memory tag data will be used so developers can display the memory
>>>> tags from a particular memory region, and will also be used to show a
>>>> detailed message about a crash that happened due to a memory tag violation."
>>>
>>> This is fine, but can what's described in the last paragraph be done
>>> by GDB commands, or does it require external tools?  In the former
>>> case, we should mention those GDB commands; in the latter case I'd
>>> settle for mentioning the fact that external tools are required.
>>
>> No external tools are required. GDB already supports displaying memory
>> tags automatically if it detects the process is using those. We can
>> explicitly mention the 'm' modifier for the x command and the memory-tag
>> sub-commands.
>>
>> Would this be more appropriate for the manual entry? Or do you think it
>> makes sense to mention it in the NEWS as well?
> 
> It's okay to only mention that in the manual.  NEWS should just say
> that GDB can display those tags automatically.
> 
> Thanks.

Thanks. I'll update these entries and will submit a v3.