netdev - Re: [PATCH net-next 2/4] ionic: fix for ppc msix layout

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] [day] [month] [year] [list]

Message-ID: <1f3ea5ee-da18-7695-a477-d65384a480bd@pensando.io>
Date:   Mon, 6 Jan 2020 19:05:59 -0800
From:   Shannon Nelson <snelson@...sando.io>
To:     David Miller <davem@...emloft.net>
Cc:     netdev@...r.kernel.org
Subject: Re: [PATCH net-next 2/4] ionic: fix for ppc msix layout

On 1/6/20 6:42 PM, David Miller wrote:
> From: Shannon Nelson <snelson@...sando.io>
> Date: Mon,  6 Jan 2020 13:05:10 -0800
>
>> The IBM Power9 ppc64 seems to have a problem with not wanting
>> to limit the address space used by a PCI device.  The Naples
>> internal HW can only address up to 52 bits, but the ppc does
>> not play well with that limitation.  This patch tells the
>> system how to work with Naples successfully.
>>
>> Signed-off-by: Shannon Nelson <snelson@...sando.io>
> Then dma_set_mask_and_coherent() should fail, either that or the mask
> being passed in isn't correct.
>
> There is no reason this hack should be necessary in any driver.
>
> Our DMA abstractions are built to handle exactly this kind of
> situation.
>
> Please find out what is really going on and fix this properly.
>
> Thank you.

Unfortunately, we spent a bunch of time with some IBM support folks and 
this is the best they could do for us to get it to work on their 
particular (peculiar?) box.

I'll pull this out of the patchset and let them know its not going to go 
upstream.

Thanks,
sln