linux-kernel - RE: [PATCH] scsi: pm8001: Fix phys_to_virt() usage on dma_addr

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: <PH0PR11MB5112E2E7D00D95F32C86677AEC6E9@PH0PR11MB5112.namprd11.prod.outlook.com>
Date:   Tue, 7 Dec 2021 10:36:18 +0000
From:   <Ajish.Koshy@...rochip.com>
To:     <john.garry@...wei.com>
CC:     <jinpu.wang@...ud.ionos.com>, <jejb@...ux.ibm.com>,
        <martin.petersen@...cle.com>, <Viswas.G@...rochip.com>,
        <linux-scsi@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
        <damien.lemoal@...nsource.wdc.com>, <Niklas.Cassel@....com>,
        <Ajish.Koshy@...rochip.com>,
        <Vasanthalakshmi.Tharmarajan@...rochip.com>
Subject: RE: [PATCH] scsi: pm8001: Fix phys_to_virt() usage on dma_addr_t

Hi John,

Well I could see this kernel panic [  126.843958] RIP: 0010:pm80xx_chip_smp_req+0x1d6/0x2e0 [pm80xx] with respect to existing driver on my current system x86 64bit after enabling the following 2 kernel boot arguments:
-intel_iommu=on
-iommu.passthrough=0

Post application of this patch,  now not observing the kernel panic and device discovery happens smoothly.


Thanks,
Ajish


 

-----Original Message-----
From: Niklas Cassel <Niklas.Cassel@....com> 
Sent: Monday, December 6, 2021 06:54 PM
To: John Garry <john.garry@...wei.com>
Cc: Ajish Koshy - I30923 <Ajish.Koshy@...rochip.com>; jinpu.wang@...ud.ionos.com; jejb@...ux.ibm.com; martin.petersen@...cle.com; Viswas G - I30667 <Viswas.G@...rochip.com>; linux-scsi@...r.kernel.org; linux-kernel@...r.kernel.org; Damien Le Moal <damien.lemoal@...nsource.wdc.com>
Subject: Re: [PATCH] scsi: pm8001: Fix phys_to_virt() usage on dma_addr_t

[Some people who received this message don't often get email from niklas.cassel@....com. Learn why this is important at http://aka.ms/LearnAboutSenderIdentification.]

EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe

On Mon, Dec 06, 2021 at 09:42:23AM +0000, John Garry wrote:
> On 29/11/2021 10:46, Ajish.Koshy@...rochip.com wrote:
> > Thanks John for the update. Based on the given issue,
> > we never tested on arm server.
> >
> > Further arm testing will depend on the availability of
> > the server.
> >
> > Meanwhile will do further test on x86 and update
> > on the observations.
>
> Have you tested on x86 with the IOMMU enabled? From my limited experience,
> out of the box the IOMMU is disabled in the BIOS on x86 machines - that is a
> very general statement. But this is not just an issue specific to arm64.

My limited experience tells me the opposite, that modern x64 servers
have IOMMU enabled by default:

E.g.:
https://www.dell.com/support/manuals/en-uk/poweredge-r7515/r7515_bios_pub/processor-settings?guid=guid-ec36d324-4f45-4bc1-bb51-de2db7cc5cd9&lang=en-us
https://support.hpe.com/hpesc/public/docDisplay?docId=a00025662en_us&docLocale=en_US


This driver not working properly on a system with an IOMMU just tells
us that the number of mainline users of this driver is very low :/


Kind regards,
Niklas