Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   John Ogness <john.ogness@linutronix.de>
To:     "Hongzhi\, Song" <hongzhi.song@windriver.com>
Cc:     <linux-kernel@vger.kernel.org>,
        linux-rt-users <linux-rt-users@vger.kernel.org>,
        <bigeasy@linutronix.de>
Subject: Re: Bug report: A commit about serial8250 cause the output disorderly at the phase of startup
References: <5f1398dd-d505-dd8a-0ea1-8186618d56d3@windriver.com>
        <82793501-9ddf-11ec-3f91-73872bf3bd54@windriver.com>
Date:   Tue, 23 Apr 2019 08:48:15 +0200
In-Reply-To: <82793501-9ddf-11ec-3f91-73872bf3bd54@windriver.com> (Song
        Hongzhi's message of "Mon, 22 Apr 2019 09:12:38 +0800")
Message-ID: <871s1tusi8.fsf@linutronix.de>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/23.4 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8BIT
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 2019-04-22, Hongzhi, Song <hongzhi.song@windriver.com> wrote:
> Anyone notice this issue?

Yes, I am aware of the issue. It is actually a feature, not a bug. ;-)
Individual LOG_CONT messages, when classified as emergency messages, are
printed immediately to the console.  This makes them appear
"disorderly". It is not yet clear how best to deal with LOG_CONT
emergency messages. But I suppose we should introduce a config option to
disable atomic consoles altogether if users preferred buffered/delayed
printk output.

John Ogness


> On 4/19/19 10:24 AM, Hongzhi, Song wrote:
>> 1. Issue description:
>>
>> Boot kernel( >= linux-rt-devel-v5.0.3 ) with qemu.
>> Then qemu will print following disorderly messages.
>>
>> At the beginning, the messages are disorderly. But then it becomes
>> normally from printing "[    0.000000] 000: Linux version..."
>>
>> --------------------------------------------------
>> [    0.019000] 000: tsc: Unable to calibrate against PIT
>> [    0.002583] 000: 6199.83 BogoMIPS (lpj=3099918)
>> [    0.521247] 000: Intel(R) Core(TM)2 Duo CPU     T7700  @ 2.40GHz
>> [    0.521247] 000:  (family: 0x6, model: 0xf
>> [    0.521247] 000: , stepping: 0xb)
>> [    0.533126] 000: unsupported p6 CPU model 15
>> [    0.533318] 000: no PMU driver, software events only.
>> [    0.765082] 000: 1 ACPI AML tables successfully acquired and loaded
>> [    0.765274] 000:
>> [    0.785903] 000: Enabled 2 GPEs in block 00 to 0F
>> [    0.786128] 000:
>> [    0.835675] 000: acpi PNP0A03:00: fail to add MMCONFIG
>> information, can't access extended PCI configuration space under
>> this bridge.
>> [    0.892056] 000:  5
>> [    0.892289] 000:  *10
>> [    0.892416] 000:  11
>> [    0.892527] 000: )
>> [    0.892661] 000:
>>
>> /* skip some repeated contents */
>>
>> [    5.052149] 000: , 512kB Cache
>> [    0.000000] 000: Linux version 5.0.3-yocto-preempt-rt+
>> (hsong@pek-lpggp1) (gcc version 8.3.0 (GCC)) #24 SMP PREEMPT Thu Apr
>> 18 03:29:58 EDT 2019
>> [    0.000000] 000: Command line: root=/dev/vda rw highres=off
>> console=ttyS0 mem=256M ip=192.168.7.4::192.168.7.3:255.255.255.0
>> vga=0 uvesafb.mode_opti0
>> [    0.000000] 000: x86/fpu: x87 FPU will use FXSAVE
>> [    0.000000] 000: BIOS-provided physical RAM map:
>> [    0.000000] 000: BIOS-e820: [mem
>> 0x0000000000000000-0x000000000009fbff]
>> [    0.000000] 000: BIOS-e820: [mem
>> 0x000000000009fc00-0x000000000009ffff]
>> [    0.000000] 000: BIOS-e820: [mem
>> 0x00000000000f0000-0x00000000000fffff]
>> [    0.000000] 000: BIOS-e820: [mem
>> 0x0000000000100000-0x000000000ffdbfff]
>>
>> --------------------------------------------------
>>
>> 2. Reproduce:
>> (1)build kernel: (Attachment is my .config)
>> make ARCH=x86_64
>> CROSS_COMPILE=[path-to-my-cross-toolchain]/x86_64-wrs-linux-
>>
>> (2)boot kernel with qemu:
>>
>> qemu-system-x86_64 \
>> -drive file=qemux86-64.rootfs.ext4,if=virtio,format=raw \
>> -nographic \
>> -kernel arch/x86/boot/bzImage \
>> -append 'root=/dev/vda rw highres=off  console=ttyS0 mem=256M ip=dhcp'
>>
>> 3. Analysis:
>> I find the following commit from >=linux-rt-devel-v5.0.3. cause the
>> issue.
>>
>> b9d460e serial: 8250: implement write_atomic