kvm-riscv.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: Shaoqin Huang <shahuang@redhat.com>
To: Alexandru Elisei <alexandru.elisei@arm.com>,
	andrew.jones@linux.dev, eric.auger@redhat.com,
	lvivier@redhat.com, thuth@redhat.com, frankja@linux.ibm.com,
	imbrenda@linux.ibm.com, nrb@linux.ibm.com, david@redhat.com,
	pbonzini@redhat.com
Cc: kvm@vger.kernel.org, kvmarm@lists.linux.dev,
	linuxppc-dev@lists.ozlabs.org, kvm-riscv@lists.infradead.org,
	linux-s390@vger.kernel.org, will@kernel.org,
	julien.thierry.kdev@gmail.com, maz@kernel.org,
	oliver.upton@linux.dev, suzuki.poulose@arm.com,
	yuzenghui@huawei.com, joey.gouly@arm.com, andre.przywara@arm.com
Subject: Re: [kvm-unit-tests PATCH v3 00/16] arm/arm64: Add kvmtool to the runner script
Date: Mon, 19 May 2025 16:56:56 +0800	[thread overview]
Message-ID: <1ecedde8-8d18-43f1-a3b7-e8bcc82a61f0@redhat.com> (raw)
In-Reply-To: <20250507151256.167769-1-alexandru.elisei@arm.com>

Hi Alexandru,

For this series, I've tested it, everything works good.

On 5/7/25 11:12 PM, Alexandru Elisei wrote:
> v2 can be found here [1].
> 
> To goal is to allow the user to do:
> 
> $ ./configure --target=kvmtool
> $ make clean && make
> $ ./run_tests.sh
> 
> to run all the tests automatically with kvmtool.
> 
> Reasons to use kvmtool:
> 
> * kvmtool is smaller and a lot easier to hack than qemu, which means
> developers may prefer it when adding or prototyping new features to KVM.
> Being able to run all the tests reliably and automatically is very useful
> in the development process.
> 
> * kvmtool is faster to run the tests (a couple of times faster on
> my rockpro64), making for a quick turnaround. But do keep in mind that not
> all tests work on kvmtool because of missing features compared to qemu.
> 
> * kvmtool does things differently than qemu: different memory layout,
> different uart, PMU emulation is disabled by default, etc. This makes it a
> good testing vehicule for kvm-unit-tests itself.
> 
> Changes in v3
> -------------
> 
> Lots of changes following the excellent feedback I got. A bird's eye view:
> 
> * Split extra_params into qemu_params and test_args: qemu_params for qemu
> arguments and test_args for the test's main() function.
> 
> Now that I'm putting the cover letter together I'm considering that maybe
> having qemu_params, kvmtool_params and test_params (instead of test_args)
> might be a better naming scheme.
> 
> * TARGET is now exported unconditionally. Unfortunately a side effect of
> this is that checking out these series and running the tests will end up
> with an error because the scripts now expect TARGET to be defined in
> config.mak.
> 
> If it's unacceptable, I can drop this and handle everything in vmm.bash by
> converting direct accesses to vmm_opts with functions defined in vmm.bash
> (vmm_opts[$TARGET:parse_premature_failure] becomes
> vmm_parse_premature_failure(), for example).
> 
> * Introduced scripts/vmm.bash to keep the vmm stuff contained. As a
> consequence there's very little $TARGET stuff in scripts/runtime.bash (only
> for premature_failure(), and no more 'case' statements anywhere) and
> instead scripts/common.bash passes the correct arguments directly to
> runtime.bash::run().
> 
> Unfortunately, because of all the changes, I decided not to keep some of
> the Reviewed-by tags. That's not to say that the effort is not appreciated,
> on the contrary, these changes are a direct result of the review; I dropped
> the tags because I was worried they might not apply to the current content
> of the patches.
> 
> If no major changes are needed following this round of review, for the next
> iteration I'm planning to send the first two patches (extra_params renamed
> to qemu_params and the new test_args) separately, to make sure it gets the
> review it deserves from the rest of the architectures.
> 
> Still haven't managed to get EDK2 to work with kvmtool, so I've decided to
> explicitely disabled UEFI tests in the last patch ("scripts: Enable
> kvmtool") - this is new.
> 
> I would also like to point out that despite Drew's comment I kept the
> 'disabled_if' test definition because I think using 'targets', with the
> default value of 'qemu', will probably lead to most, if not all, of the new
> tests which will be added never being run or tested with kvmtool. More
> details in patch #15 ("scripts: Add 'disabled_if' test definition parameter
> for kvmtool to use").
> 
> [1] https://lore.kernel.org/kvm/20250120164316.31473-1-alexandru.elisei@arm.com/
> 
> Alexandru Elisei (16):
>    scripts: unittests.cfg: Rename 'extra_params' to 'qemu_params'
>    scripts: Add 'test_args' test definition parameter
>    configure: Export TARGET unconditionally
>    run_tests.sh: Document --probe-maxsmp argument
>    scripts: Document environment variables
>    scripts: Refuse to run the tests if not configured for qemu
>    scripts: Use an associative array for qemu argument names
>    scripts: Add 'kvmtool_params' to test definition
>    scripts: Add support for kvmtool
>    scripts: Add default arguments for kvmtool
>    scripts: Add KVMTOOL environment variable for kvmtool binary path
>    scripts: Detect kvmtool failure in premature_failure()
>    scripts: Do not probe for maximum number of VCPUs when using kvmtool
>    scripts/mkstandalone: Export $TARGET
>    scripts: Add 'disabled_if' test definition parameter for kvmtool to
>      use
>    scripts: Enable kvmtool
> 
>   README.md               |  18 ++++-
>   arm/efi/run             |   8 ++
>   arm/run                 | 161 +++++++++++++++++++++++--------------
>   arm/unittests.cfg       | 125 ++++++++++++++++++++---------
>   configure               |  37 ++++++---
>   docs/unittests.txt      |  54 +++++++++++--
>   powerpc/run             |   4 +-
>   powerpc/unittests.cfg   |  21 ++---
>   riscv/run               |   4 +-
>   riscv/unittests.cfg     |   2 +-
>   run_tests.sh            |  35 ++++++---
>   s390x/run               |   2 +-
>   s390x/unittests.cfg     |  53 +++++++------
>   scripts/arch-run.bash   | 113 ++++++++++----------------
>   scripts/common.bash     |  71 +++++++++++------
>   scripts/mkstandalone.sh |   4 +
>   scripts/runtime.bash    |  51 +++++-------
>   scripts/vmm.bash        | 170 ++++++++++++++++++++++++++++++++++++++++
>   x86/run                 |   4 +-
>   x86/unittests.cfg       | 164 +++++++++++++++++++++-----------------
>   20 files changed, 730 insertions(+), 371 deletions(-)
>   create mode 100644 scripts/vmm.bash
> 
> 
> base-commit: 08db0f5cfbca16b36f200b7bc54a78fa4941bcce

-- 
Shaoqin


-- 
kvm-riscv mailing list
kvm-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kvm-riscv

      parent reply	other threads:[~2025-05-19  8:57 UTC|newest]

Thread overview: 50+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-05-07 15:12 [kvm-unit-tests PATCH v3 00/16] arm/arm64: Add kvmtool to the runner script Alexandru Elisei
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 01/16] scripts: unittests.cfg: Rename 'extra_params' to 'qemu_params' Alexandru Elisei
2025-05-07 15:40   ` Andrew Jones
2025-05-14  2:57   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 02/16] scripts: Add 'test_args' test definition parameter Alexandru Elisei
2025-05-07 15:58   ` Andrew Jones
2025-05-14  3:16   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 03/16] configure: Export TARGET unconditionally Alexandru Elisei
2025-05-07 16:02   ` Andrew Jones
2025-05-08  8:52     ` Alexandru Elisei
2025-05-08  9:39       ` Andrew Jones
2025-05-08 10:05         ` Alexandru Elisei
2025-05-08 10:17           ` Andrew Jones
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 04/16] run_tests.sh: Document --probe-maxsmp argument Alexandru Elisei
2025-05-14  3:29   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 05/16] scripts: Document environment variables Alexandru Elisei
2025-05-14  3:36   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 06/16] scripts: Refuse to run the tests if not configured for qemu Alexandru Elisei
2025-05-07 16:10   ` Andrew Jones
2025-05-07 16:14     ` Alexandru Elisei
2025-05-14  7:49   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 07/16] scripts: Use an associative array for qemu argument names Alexandru Elisei
2025-05-07 16:17   ` Andrew Jones
2025-05-14  8:02   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 08/16] scripts: Add 'kvmtool_params' to test definition Alexandru Elisei
2025-05-07 16:28   ` Andrew Jones
2025-05-08 15:54     ` Alexandru Elisei
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 09/16] scripts: Add support for kvmtool Alexandru Elisei
2025-05-07 16:38   ` Andrew Jones
2025-05-19  8:55   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 10/16] scripts: Add default arguments " Alexandru Elisei
2025-05-07 16:43   ` Andrew Jones
2025-05-21  3:21   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 11/16] scripts: Add KVMTOOL environment variable for kvmtool binary path Alexandru Elisei
2025-05-07 16:45   ` Andrew Jones
2025-05-19  8:13   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 12/16] scripts: Detect kvmtool failure in premature_failure() Alexandru Elisei
2025-05-07 16:47   ` Andrew Jones
2025-05-21  5:58   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 13/16] scripts: Do not probe for maximum number of VCPUs when using kvmtool Alexandru Elisei
2025-05-07 16:48   ` Andrew Jones
2025-05-21  6:02   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 14/16] scripts/mkstandalone: Export $TARGET Alexandru Elisei
2025-05-21  6:16   ` Shaoqin Huang
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 15/16] scripts: Add 'disabled_if' test definition parameter for kvmtool to use Alexandru Elisei
2025-05-07 16:56   ` Andrew Jones
2025-05-07 15:12 ` [kvm-unit-tests PATCH v3 16/16] scripts: Enable kvmtool Alexandru Elisei
2025-05-07 16:59   ` Andrew Jones
2025-05-21  6:20   ` Shaoqin Huang
2025-05-19  8:56 ` Shaoqin Huang [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=1ecedde8-8d18-43f1-a3b7-e8bcc82a61f0@redhat.com \
    --to=shahuang@redhat.com \
    --cc=alexandru.elisei@arm.com \
    --cc=andre.przywara@arm.com \
    --cc=andrew.jones@linux.dev \
    --cc=david@redhat.com \
    --cc=eric.auger@redhat.com \
    --cc=frankja@linux.ibm.com \
    --cc=imbrenda@linux.ibm.com \
    --cc=joey.gouly@arm.com \
    --cc=julien.thierry.kdev@gmail.com \
    --cc=kvm-riscv@lists.infradead.org \
    --cc=kvm@vger.kernel.org \
    --cc=kvmarm@lists.linux.dev \
    --cc=linux-s390@vger.kernel.org \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=lvivier@redhat.com \
    --cc=maz@kernel.org \
    --cc=nrb@linux.ibm.com \
    --cc=oliver.upton@linux.dev \
    --cc=pbonzini@redhat.com \
    --cc=suzuki.poulose@arm.com \
    --cc=thuth@redhat.com \
    --cc=will@kernel.org \
    --cc=yuzenghui@huawei.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;
as well as URLs for NNTP newsgroup(s).