From: Balbir Singh <balbirs@nvidia.com>
To: "Bert Karwatzki" <spasswolf@web.de>,
"Christian König" <christian.koenig@amd.com>
Cc: Ingo Molnar <mingo@kernel.org>, Kees Cook <kees@kernel.org>,
Bjorn Helgaas <bhelgaas@google.com>,
Linus Torvalds <torvalds@linux-foundation.org>,
Peter Zijlstra <peterz@infradead.org>,
Andy Lutomirski <luto@kernel.org>,
Alex Deucher <alexander.deucher@amd.com>,
linux-kernel@vger.kernel.org, amd-gfx@lists.freedesktop.org
Subject: Re: commit 7ffb791423c7 breaks steam game
Date: Wed, 26 Mar 2025 12:50:11 +1100 [thread overview]
Message-ID: <a9f37e3b-2192-42d2-8d5d-c38c0d3fe509@nvidia.com> (raw)
In-Reply-To: <76672910-423c-4664-a1bd-da5c1d7d6afd@nvidia.com>
On 3/26/25 10:43, Balbir Singh wrote:
> On 3/26/25 10:21, Bert Karwatzki wrote:
>> Am Mittwoch, dem 26.03.2025 um 09:45 +1100 schrieb Balbir Singh:
>>>
>>>
>>> The second region seems to be additional, I suspect that is HMM mapping from kgd2kfd_init_zone_device()
>>>
>>> Balbir Singh
>>>
>> Good guess! I inserted a printk into kgd2kfd_init_zone_device():
>>
>> diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
>> b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
>> index d05d199b5e44..201220e2ac42 100644
>> --- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
>> +++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
>> @@ -1049,6 +1049,8 @@ int kgd2kfd_init_zone_device(struct amdgpu_device *adev)
>> pgmap->range.end = res->end;
>> pgmap->type = MEMORY_DEVICE_PRIVATE;
>> }
>> + dev_info(adev->dev, "%s: range.start = 0x%llx ranges.end = 0x%llx\n",
>> + __func__, pgmap->range.start, pgmap->range.end);
>>
>> pgmap->nr_range = 1;
>> pgmap->ops = &svm_migrate_pgmap_ops;
>>
>>
>> and get this in the case without nokaslr:
>>
>> [ T367] amdgpu 0000:03:00.0: kfd_migrate: kgd2kfd_init_zone_device:
>> range.start = 0xafe00000000 ranges.end = 0xaffffffffff
>>
>> and this in the case with nokaslr:
>>
>> [ T365] amdgpu 0000:03:00.0: kfd_migrate: kgd2kfd_init_zone_device:
>> range.start = 0x3ffe00000000 ranges.end = 0x3fffffffffff
>>
>
> So we should ignore the second region then for the purposes of this issue.
>
> I think this now boils down to
>
> Why is the dma_get_required_mask set to all of addressable memory (46 bits)
> when we have nokaslr
>
I think I know the root cause of the required_mask going up and hence the
use of DMA32
1. HMM calls add_pages()
2. add_pages calls update_end_of_memory_vars()
3. This updates max_pfn and that causes required_mask to go up to 46 bits
Do you have CONFIG_HSA_AMD_SVM enabled? Does turning it off, fix the issue?
The actual issue is the update of max_pfn.
Balbir Singh
next prev parent reply other threads:[~2025-03-26 1:50 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-22 12:23 Bert Karwatzki
2025-03-23 6:51 ` Balbir Singh
2025-03-24 11:23 ` Bert Karwatzki
2025-03-24 12:14 ` Christian König
2025-03-24 22:48 ` Balbir Singh
[not found] ` <938c2cbd-c47f-4925-ba82-94eef54d9ebc@amd.com>
2025-03-25 22:45 ` Balbir Singh
2025-03-25 23:21 ` Bert Karwatzki
2025-03-25 23:43 ` Balbir Singh
2025-03-26 1:50 ` Balbir Singh [this message]
2025-03-26 10:10 ` Bert Karwatzki
2025-03-26 10:36 ` Balbir Singh
2025-03-26 11:14 ` Bert Karwatzki
2025-03-27 10:53 ` Ingo Molnar
2025-03-27 22:03 ` Balbir Singh
2025-03-24 21:43 ` Balbir Singh
-- strict thread matches above, loose matches on Subject: below --
2025-03-25 10:14 Bert Karwatzki
2025-03-25 12:23 ` Christian König
2025-03-26 22:00 ` Bert Karwatzki
2025-03-26 22:58 ` Linus Torvalds
2025-03-27 0:57 ` Balbir Singh
2025-03-27 0:58 ` Bert Karwatzki
2025-03-27 10:40 ` Ingo Molnar
2025-03-24 23:07 Bert Karwatzki
[not found] ` <634e77d7-4f3c-4d1a-8aa3-1978896f9bf2@amd.com>
2025-03-25 10:25 ` Balbir Singh
2025-03-10 11:22 Bert Karwatzki
2025-03-10 21:48 ` Balbir Singh
2025-03-11 7:19 ` Balbir Singh
2025-03-11 7:28 ` Ingo Molnar
2025-03-11 11:15 ` Bert Karwatzki
2025-03-11 18:24 ` Bert Karwatzki
2025-03-11 22:10 ` Balbir Singh
2025-03-11 23:09 ` Bert Karwatzki
2025-03-12 0:26 ` Bert Karwatzki
2025-03-12 2:23 ` Balbir Singh
2025-03-12 1:24 ` Balbir Singh
2025-03-13 9:22 ` Bert Karwatzki
2025-03-13 10:40 ` Balbir Singh
2025-03-13 10:53 ` Bert Karwatzki
2025-03-13 11:47 ` Balbir Singh
2025-03-13 18:12 ` Bert Karwatzki
2025-03-13 21:54 ` Balbir Singh
2025-03-13 22:22 ` Bert Karwatzki
2025-03-14 6:14 ` Balbir Singh
2025-03-14 13:34 ` Balbir Singh
2025-03-14 14:18 ` Bert Karwatzki
2025-03-15 0:16 ` Balbir Singh
2025-03-15 17:40 ` Alex Deucher
2025-03-16 13:09 ` Bert Karwatzki
2025-03-16 20:06 ` Bert Karwatzki
2025-03-17 9:13 ` Balbir Singh
2025-03-20 9:01 ` Ingo Molnar
2025-03-20 23:55 ` Balbir Singh
2025-03-21 10:24 ` Ingo Molnar
2025-03-21 11:05 ` Balbir Singh
2025-03-22 8:04 ` Ingo Molnar
2025-03-22 9:40 ` Balbir Singh
2025-03-20 23:43 ` Bert Karwatzki
2025-03-21 4:55 ` Balbir Singh
2025-03-21 12:26 ` Bert Karwatzki
2025-03-22 2:06 ` Balbir Singh
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=a9f37e3b-2192-42d2-8d5d-c38c0d3fe509@nvidia.com \
--to=balbirs@nvidia.com \
--cc=alexander.deucher@amd.com \
--cc=amd-gfx@lists.freedesktop.org \
--cc=bhelgaas@google.com \
--cc=christian.koenig@amd.com \
--cc=kees@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@kernel.org \
--cc=mingo@kernel.org \
--cc=peterz@infradead.org \
--cc=spasswolf@web.de \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®