linux-kernel - Re: [PATCH] nouveau/hmm: fix nouveau_dmem

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: <a4a405d7-c136-b479-2bbe-5a2304f59c99@nvidia.com>
Date:   Thu, 23 Apr 2020 11:32:31 -0700
From:   Ralph Campbell <rcampbell@...dia.com>
To:     Jason Gunthorpe <jgg@...lanox.com>
CC:     <nouveau@...ts.freedesktop.org>, <linux-rdma@...r.kernel.org>,
        <linux-kernel@...r.kernel.org>, Jerome Glisse <jglisse@...hat.com>,
        "John Hubbard" <jhubbard@...dia.com>,
        Christoph Hellwig <hch@....de>, Ben Skeggs <bskeggs@...hat.com>
Subject: Re: [PATCH] nouveau/hmm: fix nouveau_dmem_chunk allocations


On 4/23/20 5:17 AM, Jason Gunthorpe wrote:
> On Tue, Apr 21, 2020 at 04:11:07PM -0700, Ralph Campbell wrote:
>> In nouveau_dmem_init(), a number of struct nouveau_dmem_chunk are allocated
>> and put on the dmem->chunk_empty list. Then in nouveau_dmem_pages_alloc(),
>> a nouveau_dmem_chunk is removed from the list and GPU memory is allocated.
>> However, the nouveau_dmem_chunk is never removed from the chunk_empty
>> list nor placed on the chunk_free or chunk_full lists. This results
>> in only one chunk ever being actually used (2MB) and quickly leads to
>> migration to device private memory failures.
>>
>> Fix this by having just one list of free device private pages and if no
>> pages are free, allocate a chunk of device private pages and GPU memory.
>>
>> Signed-off-by: Ralph Campbell <rcampbell@...dia.com>
>> ---
>>   drivers/gpu/drm/nouveau/nouveau_dmem.c | 304 +++++++++----------------
>>   1 file changed, 112 insertions(+), 192 deletions(-)
> 
> Does this generate any conflicts with my series to rework
> hmm_range_fault?
> 
> Jason

I based it on top of your series and the patch that Ben already has queued
("nouveau/hmm: map pages after migration") so it shouldn't have any conflicts.

I guess I should have mentioned that in a cover letter.