Re: [PATCH v5] arm pl011 serial: support multi-irq request

From: Robin Murphy <robin.murphy@arm.com>
To: Qian Cai <quic_qiancai@quicinc.com>, Bing Fan <hptsfb@gmail.com>,
	gregkh@linuxfoundation.org, Bing Fan <tombinfan@tencent.com>
Cc: linux-serial@vger.kernel.org, linux-kernel@vger.kernel.org,
	Linux ARM <linux-arm-kernel@lists.infradead.org>
Subject: Re: [PATCH v5] arm pl011 serial: support multi-irq request
Date: Mon, 26 Jul 2021 22:14:04 +0100	[thread overview]
Message-ID: <e98962f3-9232-4abf-ec27-a7524a9e786d@arm.com> (raw)
In-Reply-To: <290c01ec-173f-755f-788e-2a33a69586e8@quicinc.com>

On 2021-07-26 21:56, Qian Cai wrote:
> 
> 
> On 7/26/2021 4:36 PM, Robin Murphy wrote:
>> The important point you're missing, but which the KASAN dump does hint at, is
>> that that is a machine with SBSA generic UARTs booting via ACPI - I know it
>> doesn't do DT at all because I have one too. What matters there is that pl011
>> binds as a platform driver, *not* an amba driver.
> 
> Thanks for pointing out, Robin. I just yet to see an ARM server booting from DT
> those days.

Unlikely in production datacentre/cloud environments, indeed, although 
some of the mid-range kit like LX2160 does start to blur the line of 
what might be considered "server", and that's one example which *does* 
have full-featured DT support (even if it also aspires to ACPI...)

What I thought was worth clarifying for the general audience is that the 
relevant aspects of "server" here should in fact still be possible to 
reproduce on something like a Raspberry Pi or a tiny QEMU VM, if one can 
figure out the ACPI runes :)

Thanks,
Robin.