Linux IOMMU Development
 help / color / mirror / Atom feed
From: Baolu Lu <baolu.lu@linux.intel.com>
To: Jason Gunthorpe <jgg@nvidia.com>
Cc: baolu.lu@linux.intel.com,
	"iommu@lists.linux.dev" <iommu@lists.linux.dev>,
	Kevin Tian <kevin.tian@intel.com>, Yi Liu <yi.l.liu@intel.com>,
	Nicolin Chen <nicolinc@nvidia.com>
Subject: Re: an iommufd selftest issue with 6.4-rc1
Date: Thu, 18 May 2023 15:22:31 +0800	[thread overview]
Message-ID: <8a211d2e-27ac-90ca-584c-c438cd3b535b@linux.intel.com> (raw)
In-Reply-To: <ZGTedVao04FKpl7A@nvidia.com>

On 2023/5/17 22:02, Jason Gunthorpe wrote:
> On Wed, May 17, 2023 at 09:56:47PM +0800, Baolu Lu wrote:
>> Hi Jason,
>>
>> I compiled a 6.4-rc1 kernel and ran the iommufd selftest. The test hung,
>> and I had to break it with Ctrl+C. Is this a known issue, or there is
>> something wrong with my test setup?
> 
> There is an annoying bug in the test harness and it infinite loops for
> some failure cases. That explains the hang

Get it, thank you!

> 
>> # #  RUN           iommufd_ioas.no_domain.ioas_destroy ...
>> # #            OK  iommufd_ioas.no_domain.ioas_destroy
>> # ok 13 iommufd_ioas.no_domain.ioas_destroy
>> # #  RUN           iommufd_ioas.no_domain.hwpt_attach ...
>> # # hwpt_attach: Test terminated unexpectedly by signal 9
> 
> That seems bad

I don't know yet, need some investigation.

> 
>> # #          FAIL  iommufd_ioas.no_domain.hwpt_attach
>> # not ok 14 iommufd_ioas.no_domain.hwpt_attach
>> # #  RUN           iommufd_ioas.no_domain.ioas_area_destroy ...
>> # #            OK  iommufd_ioas.no_domain.ioas_area_destroy
>> # ok 15 iommufd_ioas.no_domain.ioas_area_destroy
>> # #  RUN           iommufd_ioas.no_domain.ioas_area_auto_destroy ...
>> # #            OK  iommufd_ioas.no_domain.ioas_area_auto_destroy
>> # ok 16 iommufd_ioas.no_domain.ioas_area_auto_destroy
>> # #  RUN           iommufd_ioas.no_domain.area ...
>> # #            OK  iommufd_ioas.no_domain.area
>> # ok 17 iommufd_ioas.no_domain.area
>> # #  RUN           iommufd_ioas.no_domain.unmap_fully_contained_areas ...
>> # #            OK  iommufd_ioas.no_domain.unmap_fully_contained_areas
>> # ok 18 iommufd_ioas.no_domain.unmap_fully_contained_areas
>> # #  RUN           iommufd_ioas.no_domain.area_auto_iova ...
>> # #            OK  iommufd_ioas.no_domain.area_auto_iova
>> # ok 19 iommufd_ioas.no_domain.area_auto_iova
>> # #  RUN           iommufd_ioas.no_domain.area_allowed ...
>> # #            OK  iommufd_ioas.no_domain.area_allowed
>> # ok 20 iommufd_ioas.no_domain.area_allowed
>> # #  RUN           iommufd_ioas.no_domain.copy_area ...
>> # #            OK  iommufd_ioas.no_domain.copy_area
>> # ok 21 iommufd_ioas.no_domain.copy_area
>> # #  RUN           iommufd_ioas.no_domain.iova_ranges ...
>> # #            OK  iommufd_ioas.no_domain.iova_ranges
>> # ok 22 iommufd_ioas.no_domain.iova_ranges
>> # #  RUN           iommufd_ioas.no_domain.access_domain_destory ...
>> # # iommufd.c:611:access_domain_destory:Expected MAP_FAILED
>> (18446744073709551615) != buf (18446744073709551615)
> 
> And you had a failed mmp, so probably missing THPs or something

Seems to look good.

# grep THP .config
CONFIG_ARCH_ENABLE_THP_MIGRATION=y
CONFIG_ARCH_WANTS_THP_SWAP=y
CONFIG_THP_SWAP=y
# CONFIG_READ_ONLY_THP_FOR_FS is not set

# cat /sys/kernel/mm/transparent_hugepage/enabled
[always] madvise never

> I don't see any failures when I run it..

Okay, that's fine.

Best regards,
baolu


      reply	other threads:[~2023-05-18  7:22 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-05-17 13:56 an iommufd selftest issue with 6.4-rc1 Baolu Lu
2023-05-17 14:02 ` Jason Gunthorpe
2023-05-18  7:22   ` Baolu Lu [this message]

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=8a211d2e-27ac-90ca-584c-c438cd3b535b@linux.intel.com \
    --to=baolu.lu@linux.intel.com \
    --cc=iommu@lists.linux.dev \
    --cc=jgg@nvidia.com \
    --cc=kevin.tian@intel.com \
    --cc=nicolinc@nvidia.com \
    --cc=yi.l.liu@intel.com \
    /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