linux-kernel - Re: [PATCH v2 3/3] iommufd/device: Change iommufd_hw_pagetable_has

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, 1 Feb 2023 22:36:00 -0800
From:   Nicolin Chen <nicolinc@...dia.com>
To:     "Tian, Kevin" <kevin.tian@...el.com>
CC:     Baolu Lu <baolu.lu@...ux.intel.com>,
        "jgg@...dia.com" <jgg@...dia.com>,
        "Lu, Baolu" <baolu.lu@...el.com>,
        "robin.murphy@....com" <robin.murphy@....com>,
        "Liu, Yi L" <yi.l.liu@...el.com>,
        "iommu@...ts.linux.dev" <iommu@...ts.linux.dev>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2 3/3] iommufd/device: Change
 iommufd_hw_pagetable_has_group to device centric

On Thu, Feb 02, 2023 at 06:32:36AM +0000, Tian, Kevin wrote:
> External email: Use caution opening links or attachments
> 
> 
> > From: Nicolin Chen <nicolinc@...dia.com>
> > Sent: Wednesday, February 1, 2023 3:20 PM
> >
> > So, assuming we drop this series and move the first two patches
> > back to the nesting series or the replace series, one of them
> > would end up doing something ugly:
> >
> >       if (cur_hwpt != hwpt)
> >               mutex_lock(&cur_hwpt->device_lock);
> >       mutex_lock(&hwpt->device_lock);
> >       ...
> >       mutex_unlock(&hwpt->device_lock);
> >       if (cur_hwpt != hwpt)
> >               mutex_unlock(&cur_hwpt->device_lock);
> >
> > So, perhaps we should discuss about which way we want to choose.
> 
> from your discussion with Jason I think this locking open has
> been settled down.

Yes :)

> >
> > Btw, Baolu's version has a similar patch as mine changing the
> > iommu_attach/detach_device(), yet also touches _group(). Could
> > we bisect that series into _device() first and _group() later?
> > Given that we only need a device-centric API at this moment...
> >
> 
> I'll let Baolu to decide after he re-catches up the comments in
> that thread. But overall I think we now agreed that removing the
> device list/lock can be kept out of your replace/nesting series
> and let it cleaned up after Baolu's work completes, correct? 😊

Correct. It's not a blocker. And I am going to post the replace
series today -- running some additional sanity now.

Thanks
Nic