linux-kernel - Re: [PATCH 1/4] ARM: footbridge: remove addin mode

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]

Date:   Wed, 24 Aug 2022 12:33:49 +0200
From:   Arnd Bergmann <arnd@...nel.org>
To:     Christoph Hellwig <hch@....de>
Cc:     linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        Arnd Bergmann <arnd@...db.de>,
        Russell King <linux@...linux.org.uk>,
        Linus Walleij <linus.walleij@...aro.org>,
        Marc Zyngier <maz@...nel.org>
Subject: Re: [PATCH 1/4] ARM: footbridge: remove addin mode

On Sun, Aug 21, 2022 at 7:53 AM Christoph Hellwig <hch@....de> wrote:
>
> On Thu, Aug 18, 2022 at 11:15:47PM +0200, Arnd Bergmann wrote:
> > The recently added phys_to_dma() functions are now trivial and
> > could probably be removed again as a follow-up, if anyone knows
> > how.
>
> The normal way to that would be with a call to dma_direct_set_offset in
> the init code.

Ah good, that is what I was looking for. Unfortunately it seems that the
platform does not itself create any 'struct device'. The DMA masters
are either ISA devices, which I think go through a dummy device an
should also use offset 0 (the same offset seems to get added
in dma-mapping code but subtracted again in the ISA DMA API)
and in PCI bus masters.

Is it enough to just call dma_direct_set_offset() on the PCI root
device, or do I have to add a pci_host_bridge callback to get
called from pcibios_bus_add_device()?

>  But for that we'd need to find someone to care enough
> about footbridge to test it.

LinusW, MarcZ and RMK all have one of these, I'm fairly sure
one of them can test the PCI side, though possibly not
the ISA DMA ones (parport, floppy, ne2000, and soundblaster
in case of netwinder).

         Arnd