linux-kernel - Re: [PATCH 0/2] Fix v4.11 malta

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: <f15b4357-0e26-7c31-a279-d6ecad536d4d@imgtec.com>
Date:   Fri, 31 Mar 2017 13:55:51 +0100
From:   Matt Redfearn <matt.redfearn@...tec.com>
To:     Marc Zyngier <marc.zyngier@....com>,
        Ralf Baechle <ralf@...ux-mips.org>,
        James Hogan <james.hogan@...tec.com>
CC:     <linux-mips@...ux-mips.org>, <linux-kernel@...r.kernel.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Jason Cooper <jason@...edaemon.net>,
        Paul Burton <paul.burton@...tec.com>
Subject: Re: [PATCH 0/2] Fix v4.11 malta_defconfig regressions

Hi Marc,


On 31/03/17 13:04, Marc Zyngier wrote:
> Hi Matt,
>
> On 31/03/17 12:05, Matt Redfearn wrote:
>> Since v4.11-rc1, 3 regressions have been observed on the Malta platform,
>> using malta_defconfig. which prevent it booting. These patches fix 2 of
>> them. The third one is that malta_defconfig, which uses SMP-MT, no
>> longer sets up its IPIs correctly resulting is a string of messages
>> like:
>>
>> irq 23: nobody cared (try booting with the "irqpoll" option)
>> CPU: 1 PID: 0 Comm: swapper/1 Tainted: G        W       4.11.0-rc4 #421
>> Stack : 00000000 00000000 00000000 00000000 807cdff2 00000047 00000000 0000003d
>>          80741327 8f093194 806c191c 00000000 00000001 807c9acc 80756078 807d0000
>>          807cdbe4 80177c78 00000003 0000003c 00000006 80177a04 806c70a8 8f02be8c
>>          00000006 801b4c8c 00000000 00000000 ffffffff 00000000 8f02be8c 80740000
>>          00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
>>          ...
>> Call Trace:
>> [<8010c6c0>] show_stack+0x88/0xa4
>> [<80380fb8>] dump_stack+0x88/0xd0
>> [<8017cf64>] __report_bad_irq+0x48/0x108
>> [<8017d2d4>] note_interrupt+0x1c0/0x2fc
>> [<80179ed4>] handle_irq_event_percpu+0x4c/0x64
>> [<8017eafc>] handle_percpu_irq+0x88/0xb8
>> [<801791c0>] generic_handle_irq+0x40/0x58
>> [<80108664>] do_IRQ+0x18/0x24
>> [<803b83fc>] plat_irq_dispatch+0x54/0xa8
>> handlers:
>> Disabling IRQ #23
>>
>> This regression is fixed by Paul Burtons series "MIPS/irqchip: Use IPI
>> IRQ domains for CPU interrupt controller IPIs", but it is a large change
>> for this stage in the cycle so I don't know how best to proceed with
>> that one.
>>
>>
>>
>> Matt Redfearn (2):
>>    MIPS: Malta: Fix i8259 irqchip setup
>>    irqchip/mips-gic: Fix Local compare interrupt
>>
>>   arch/mips/mti-malta/malta-int.c | 13 +++++++++++++
>>   drivers/irqchip/irq-mips-gic.c  |  4 ++++
>>   2 files changed, 17 insertions(+)
>>
> I can take the GIC patch through the irq tree if that's convenient (I
> was about to send a PR anyway). Just let me know.
>
> Thanks,
>
> 	M.

Yes that would be great, please. Ralf has just acked it.

Thanks,
Matt