linux-kernel - Re: [PATCH] iommu/rockchip: Make use of 'struct iommu

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:   Mon, 03 Apr 2017 11:56:59 +0200
From:   Heiko Stübner <heiko@...ech.de>
To:     Joerg Roedel <joro@...tes.org>
Cc:     iommu@...ts.linux-foundation.org,
        linux-arm-kernel@...ts.infradead.org,
        linux-rockchip@...ts.infradead.org, linux-kernel@...r.kernel.org,
        Joerg Roedel <jroedel@...e.de>
Subject: Re: [PATCH] iommu/rockchip: Make use of 'struct iommu_device'

Hi Joerg,

Am Freitag, 31. März 2017, 16:30:24 CEST schrieb Joerg Roedel:
> From: Joerg Roedel <jroedel@...e.de>
> 
> Register hardware IOMMUs seperatly with the iommu-core code
> and add a sysfs representation of the iommu topology.
> 
> Signed-off-by: Joerg Roedel <jroedel@...e.de>

In general works, and I still keep a working iommu-based display :-)
I can also see my two vop iommus under /sys/class/iommu now.

Links in the devices subdirectory do not work though, see below:

> ---
>  drivers/iommu/rockchip-iommu.c | 30 ++++++++++++++++++++++++++++--
>  1 file changed, 28 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/iommu/rockchip-iommu.c b/drivers/iommu/rockchip-iommu.c
> index 9afcbf7..36d0890 100644
> --- a/drivers/iommu/rockchip-iommu.c
> +++ b/drivers/iommu/rockchip-iommu.c
> @@ -1054,6 +1056,10 @@ static int rk_iommu_add_device(struct device *dev)
>  	if (ret)
>  		goto err_remove_device;
> 
> +	iommu = rk_iommu_from_dev(dev);
> +	if (iommu)
> +		iommu_device_link(&iommu->iommu, dev);
> +

It looks like the iommu association can very well run before the iommu
actually probes, via the bus_set_iommu and which happens on my veyron-pinky
device). Thus iommu is NULL, as platform_set_drvdata in probe hasn't run yet
and the link is never created.

[    0.034180] iommu: Adding device ff930000.vop to group 0
[    0.034198] rk_iommu_from_dev: found iommu ff930300.iommu for ff930000.vop with data at   (null)
[    0.034209] rk_iommu_add_device: rk_iommu_from_dev returned   (null) for ff930000.vop
[    0.034232] iommu: Adding device ff940000.vop to group 1
[    0.034247] rk_iommu_from_dev: found iommu ff940300.iommu for ff940000.vop with data at   (null)
[    0.034258] rk_iommu_add_device: rk_iommu_from_dev returned   (null) for ff940000.vop
[    0.034342] rk_iommu_probe: set_drvdata for ff930300.iommu at ee2c7a10
[    0.034411] rk_iommu_probe: set_drvdata for ff940300.iommu at ee2c7c10
[    7.326395] rk_iommu_from_dev: found iommu ff930300.iommu for ff930000.vop with data at ee2c7a10


Not sure what the proper solution is though.


Heiko