linux-kernel - Re: linux-next: build failure after merge of the drm tree

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, 28 Sep 2020 13:46:55 +0200
From:   Paul Cercueil <paul@...pouillou.net>
To:     Christoph Hellwig <hch@....de>
Cc:     Stephen Rothwell <sfr@...b.auug.org.au>,
        Dave Airlie <airlied@...ux.ie>,
        DRI <dri-devel@...ts.freedesktop.org>,
        Linux Next Mailing List <linux-next@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: linux-next: build failure after merge of the drm tree



Le lun. 28 sept. 2020 à 13:34, Christoph Hellwig <hch@....de> a écrit 
:
> On Mon, Sep 28, 2020 at 12:15:56PM +0200, Paul Cercueil wrote:
>>  Hi Christoph,
>> 
>>  Le lun. 28 sept. 2020 à 8:04, Christoph Hellwig <hch@....de> a 
>> écrit :
>>>  On Mon, Sep 28, 2020 at 01:54:05PM +1000, Stephen Rothwell wrote:
>>>>   Hi all,
>>>> 
>>>>   After merging the drm tree, today's linux-next build (x86_64
>>>>  allmodconfig)
>>>>   failed like this:
>>> 
>>>  The driver needs to switch do dma_alloc_noncoherent + 
>>> dma_sync_single*
>>>  like the other drivers converted in the dma tree.  Paul, let me 
>>> know if
>>>  you have any questions.
>> 
>>  I don't dma_alloc* anything, DRM core does. I use the
>>  DMA_ATTR_NON_CONSISTENT attr with dma_mmap_attrs(). Is there a 
>> replacement
>>  for that?
> 
> dma_mmap_attrs can only be used on allocations from dma_mmap_attrs 
> with
> the same attrs.  As there is no allocation using 
> DMA_ATTR_NON_CONSISTENT
> in the drm core, something looks very fishy here.

Is that a fact? I don't see why you couldn't change the cache settings 
after allocation. In practice it works just fine.

> Where does the allocation you try to mmap come from?  All the 
> allocations
> in drivers/gpu/drm/drm_gem_cma_helper.c seems to use dma_alloc_wc (aka
> dma_allloc_attrs with the DMA_ATTR_WRITE_COMBINE flag).

It's the dma_alloc_wc.

-Paul