| Message ID | 20220311162445.346685-1-vincent.whitchurch@axis.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | roadtest: a driver testing framework | expand |
+Kees Cook - I imagine you have already seen this, but I figured you would be interested because of your recent work on the KUnit UAPI and the mocking discussions. +Dmitry Vyukov - This made me think of the syzkaller/KUnit experiments we did a couple of years back - this would probably work a bit better. On Fri, Mar 11, 2022 at 11:24 AM Vincent Whitchurch <vincent.whitchurch@axis.com> wrote: > > This patchset proposes roadtest, a device-driver testing framework. Drivers > are tested under User Mode Linux (UML) and interact with mocked/modelled > hardware. The tests and hardware models are written in Python, the former > using Python's built-in unittest framework. Wow! This sounds awesome! I was hoping to get some kind of hardware modeling with KUnit eventually. I did some experiments, but this looks way more mature. > Drivers are tested via their userspace interfaces. The hardware models allow > tests to inject values into registers and assert that drivers control the > hardware in the right way and react as expected to stimuli. I already took a look at the documentation patch - I'll comment there more in detail, but I like the hardware modelling and device tree code; it seems very usable. > Roadtest is meant to be used for relatively simple drivers, such as the ones > part of the IIO, regulator and RTC subsystems. Obviously for an initial version going after simple stuff makes sense, but I would hope there is applicability to any driver stack eventually. > Questions and answers: > > = Why do we need this? > > There are a large amount of these kind of drivers in the kernel. Most of the > hardware is not available in current CI systems so most drivers can only, at > best, be build-tested there. Even basic soundness such as a driver > successfully probing and binding to the devices it tries to be support cannot > be tested. Drivers cannot be easily regression-tested to ensure that bugs > fixed once do not get reintroduced. > > Many drivers support multiple related hardware variants, and far from all patch > submitters have access to all the variants which the driver that they are > patching supports, so there is no way for them to easily verify that they > haven't broken something basic on a variant which they do not own. > > Furthermore, hardware can be used in many different configurations with drivers > supporting many different devicetree properties, so even just having access to > all the variants would be insufficient. > > On top of that, some of the chips measure environmental conditions such as > temperature, so testing extreme cases may not be simple even if one has access > to the hardware. > > All this makes development, modification, maintenance, and reviewing of these > drivers harder than it necessarily needs to be. Roadtest hopes to make some of > these things slightly easier by providing a framework to create hardware > models/mocks and to write testcases which exercise drivers using these models. Very much agree. I used to do driver development and these difficulties are what prompted me to do KUnit. > = Do you have some specific examples of the kind of code this could be used to > test? > > Here is an example of a patch which can easily be regression-tested using > roadtest (in fact, this series includes such a regression test) but is much > harder to do so automatically with real hardware since it requires specific > environmental conditions: > > iio: light: opt3001: Fixed timeout error when 0 lux > https://lore.kernel.org/lkml/20210920125351.6569-1-valek@2n.cz/ > > Here is another example. This driver has code which correctly parses a > documented devicetree property (amstaos,proximity-diodes) but which then fails > to actually communicate this setting to the hardware in any way. Such code can > be easily tested with roadtest since the framework integrates devicetree > support and provides functions to assert that drivers writes expected registers > with expected values: > > drivers/iio/light/tsl2772.c tsl2772_read_prox_diodes() > > (Both the above examples happen to be from the same subsystem but that should > in no way be taken to imply that such issues are unique to that subsystem or > that that subsystem has more of them.) > > = How does this relate to kselftests? > > Tests in kselftests also test kernel code using the userspace interfaces, but > that's about what's common between the frameworks. kselftests has other goals > and does not provide any kind of mechanism for hardware mocking. I had a question that after thinking about it; I think I know the answer, so I am going to ask the question anyway and attempt to answer it myself: I agree in regard to mocking, but why not use kselftest for driving tests that check drivers from userspace? I believe there are other kselftest tests implemented in Python, why can't you just run your tests inside of kselftest? Now, I believe the answer to this question is that you need to control spinning up your own kernel to run inside your test harness because you need to control the environment that the kernel runs in - is this correct? > = How does this relate to kunit? > > Kunit is for unit testing of functions in kernel code, and is not meant for > testing kernel code via userspace interfaces. It could in theory be used to > test some of the simple drivers too, but that would require (1) a large amount > of mocking code in various kernel frameworks, and, more importantly, (2) > refactoring of the drivers to be tested. I mostly agree, but I think there is something that is missing here: so roadtest seems to depend on having a user interface to test a driver - for a simple smoke test on a simple driver without a big driver stack on top, that makes sense, but what about testing error paths or a platform driver buried beneath a deep driver stack? I think there is potential for a powerful combination using KUnit to test the low level kernel API and using roadtest to mock the hardware environment and provide configuration. I am imagining that we could have an in-kernel KUnit/roadtest API that we can use to have an in-kernel test request changes to the environment for creating error cases and the like that can be validated by KUnit test cases. Going even further, I wonder if we could run kselftests inside of roadtest since roadtest allows us to change the environment on the fly. > This can be contrasted with roadtest which works with mostly unmodified drivers > and which mocks the hardware at the lowest level without having to change > kernel frameworks. I think that is both potentially an advantage and a disadvantage. The advantage is that your test is very general; roadtests would likely be portable across kernel versions. The disadvantage is that you don't get as much code introspection: I imagine roadtest is not as good as testing error paths for example. I also think that having to change code to make it more testable is often an advantage as much as a disadvantage. Still, I think that is a good set of tradeoffs for roadtest to make when set against KUnit and kselftest since roadtest seems to fit in where kselftest and KUnit are weak. > = How do I use it? > > See Documentation/dev-tools/roadtest.rst added by the documentation patch for > more information about running and writing tests using this framework. > > = What's included in the patchset? > > The current framework allows developing tests for hardware which uses the I2C > bus. Hardware models can also control GPIOs and use them to trigger > interrupts. > > This series includes tests for some IIO, regulator and RTC drivers. The > regulator and RTC tests depend on a few driver patches which are either in > review or in linux-next. These are noted in the commit messages. > > The entire patch set, including the required dependencies, is also available in > a git tree: > > https://github.com/vwax/linux/commits/roadtest/rfc-v1 > > Cc: linux-kernel@vger.kernel.org > Cc: devicetree@vger.kernel.org > Cc: linux-um@lists.infradead.org > > Cc: shuah@kernel.org > Cc: brendanhiggins@google.com > Cc: linux-kselftest@vger.kernel.org > > Cc: jic23@kernel.org > Cc: linux-iio@vger.kernel.org > > Cc: lgirdwood@gmail.com > Cc: broonie@kernel.org > > Cc: a.zummo@towertech.it > Cc: alexandre.belloni@bootlin.com > Cc: linux-rtc@vger.kernel.org > > Cc: corbet@lwn.net > Cc: linux-doc@vger.kernel.org > > Vincent Whitchurch (10): > roadtest: import libvhost-user from QEMU > roadtest: add C backend > roadtest: add framework > roadtest: add base config > roadtest: add build files > roadtest: add documentation > iio: light: opt3001: add roadtest > iio: light: vcnl4000: add roadtest > regulator: tps62864: add roadtest > rtc: pcf8563: add roadtest > > Documentation/dev-tools/index.rst | 1 + > Documentation/dev-tools/roadtest.rst | 669 ++++ > tools/testing/roadtest/.gitignore | 2 + > tools/testing/roadtest/Dockerfile | 25 + > tools/testing/roadtest/Makefile | 84 + > tools/testing/roadtest/init.sh | 19 + > tools/testing/roadtest/pyproject.toml | 10 + > tools/testing/roadtest/requirements.txt | 4 + > tools/testing/roadtest/roadtest/__init__.py | 2 + > .../roadtest/roadtest/backend/__init__.py | 0 > .../roadtest/roadtest/backend/backend.py | 32 + > .../testing/roadtest/roadtest/backend/gpio.py | 111 + > .../testing/roadtest/roadtest/backend/i2c.py | 123 + > .../testing/roadtest/roadtest/backend/main.py | 13 + > .../testing/roadtest/roadtest/backend/mock.py | 20 + > .../roadtest/roadtest/backend/test_gpio.py | 98 + > .../roadtest/roadtest/backend/test_i2c.py | 84 + > .../testing/roadtest/roadtest/cmd/__init__.py | 0 > tools/testing/roadtest/roadtest/cmd/main.py | 146 + > tools/testing/roadtest/roadtest/cmd/remote.py | 48 + > .../roadtest/roadtest/core/__init__.py | 0 > .../testing/roadtest/roadtest/core/control.py | 52 + > .../roadtest/roadtest/core/devicetree.py | 155 + > .../roadtest/roadtest/core/hardware.py | 94 + > tools/testing/roadtest/roadtest/core/log.py | 42 + > .../testing/roadtest/roadtest/core/modules.py | 38 + > .../testing/roadtest/roadtest/core/opslog.py | 35 + > tools/testing/roadtest/roadtest/core/proxy.py | 48 + > tools/testing/roadtest/roadtest/core/suite.py | 286 ++ > tools/testing/roadtest/roadtest/core/sysfs.py | 77 + > .../roadtest/roadtest/core/test_control.py | 35 + > .../roadtest/roadtest/core/test_devicetree.py | 31 + > .../roadtest/roadtest/core/test_hardware.py | 41 + > .../roadtest/roadtest/core/test_log.py | 54 + > .../roadtest/roadtest/core/test_opslog.py | 27 + > .../roadtest/roadtest/tests/__init__.py | 0 > .../roadtest/roadtest/tests/base/config | 84 + > .../roadtest/roadtest/tests/iio/__init__.py | 0 > .../roadtest/roadtest/tests/iio/config | 1 + > .../roadtest/roadtest/tests/iio/iio.py | 112 + > .../roadtest/tests/iio/light/__init__.py | 0 > .../roadtest/roadtest/tests/iio/light/config | 2 + > .../roadtest/tests/iio/light/test_opt3001.py | 95 + > .../roadtest/tests/iio/light/test_vcnl4000.py | 132 + > .../roadtest/tests/iio/light/test_vcnl4010.py | 282 ++ > .../roadtest/tests/iio/light/test_vcnl4040.py | 104 + > .../roadtest/tests/iio/light/test_vcnl4200.py | 96 + > .../roadtest/tests/regulator/__init__.py | 0 > .../roadtest/roadtest/tests/regulator/config | 4 + > .../roadtest/tests/regulator/test_tps62864.py | 187 ++ > .../roadtest/roadtest/tests/rtc/__init__.py | 0 > .../roadtest/roadtest/tests/rtc/config | 1 + > .../roadtest/roadtest/tests/rtc/rtc.py | 73 + > .../roadtest/tests/rtc/test_pcf8563.py | 348 ++ > tools/testing/roadtest/src/.gitignore | 1 + > tools/testing/roadtest/src/backend.c | 884 +++++ > .../src/libvhost-user/include/atomic.h | 310 ++ > .../src/libvhost-user/libvhost-user.c | 2885 +++++++++++++++++ > .../src/libvhost-user/libvhost-user.h | 691 ++++ > 59 files changed, 8798 insertions(+) > create mode 100644 Documentation/dev-tools/roadtest.rst > create mode 100644 tools/testing/roadtest/.gitignore > create mode 100644 tools/testing/roadtest/Dockerfile > create mode 100644 tools/testing/roadtest/Makefile > create mode 100755 tools/testing/roadtest/init.sh > create mode 100644 tools/testing/roadtest/pyproject.toml > create mode 100644 tools/testing/roadtest/requirements.txt > create mode 100644 tools/testing/roadtest/roadtest/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/backend/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/backend/backend.py > create mode 100644 tools/testing/roadtest/roadtest/backend/gpio.py > create mode 100644 tools/testing/roadtest/roadtest/backend/i2c.py > create mode 100644 tools/testing/roadtest/roadtest/backend/main.py > create mode 100644 tools/testing/roadtest/roadtest/backend/mock.py > create mode 100644 tools/testing/roadtest/roadtest/backend/test_gpio.py > create mode 100644 tools/testing/roadtest/roadtest/backend/test_i2c.py > create mode 100644 tools/testing/roadtest/roadtest/cmd/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/cmd/main.py > create mode 100644 tools/testing/roadtest/roadtest/cmd/remote.py > create mode 100644 tools/testing/roadtest/roadtest/core/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/core/control.py > create mode 100644 tools/testing/roadtest/roadtest/core/devicetree.py > create mode 100644 tools/testing/roadtest/roadtest/core/hardware.py > create mode 100644 tools/testing/roadtest/roadtest/core/log.py > create mode 100644 tools/testing/roadtest/roadtest/core/modules.py > create mode 100644 tools/testing/roadtest/roadtest/core/opslog.py > create mode 100644 tools/testing/roadtest/roadtest/core/proxy.py > create mode 100644 tools/testing/roadtest/roadtest/core/suite.py > create mode 100644 tools/testing/roadtest/roadtest/core/sysfs.py > create mode 100644 tools/testing/roadtest/roadtest/core/test_control.py > create mode 100644 tools/testing/roadtest/roadtest/core/test_devicetree.py > create mode 100644 tools/testing/roadtest/roadtest/core/test_hardware.py > create mode 100644 tools/testing/roadtest/roadtest/core/test_log.py > create mode 100644 tools/testing/roadtest/roadtest/core/test_opslog.py > create mode 100644 tools/testing/roadtest/roadtest/tests/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/tests/base/config > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/config > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/iio.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/config > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/test_opt3001.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/test_vcnl4000.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/test_vcnl4010.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/test_vcnl4040.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/test_vcnl4200.py > create mode 100644 tools/testing/roadtest/roadtest/tests/regulator/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/tests/regulator/config > create mode 100644 tools/testing/roadtest/roadtest/tests/regulator/test_tps62864.py > create mode 100644 tools/testing/roadtest/roadtest/tests/rtc/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/tests/rtc/config > create mode 100644 tools/testing/roadtest/roadtest/tests/rtc/rtc.py > create mode 100644 tools/testing/roadtest/roadtest/tests/rtc/test_pcf8563.py > create mode 100644 tools/testing/roadtest/src/.gitignore > create mode 100644 tools/testing/roadtest/src/backend.c > create mode 100644 tools/testing/roadtest/src/libvhost-user/include/atomic.h > create mode 100644 tools/testing/roadtest/src/libvhost-user/libvhost-user.c > create mode 100644 tools/testing/roadtest/src/libvhost-user/libvhost-user.h > > -- > 2.34.1 >
On Mon, Mar 14, 2022 at 11:24:59PM +0100, Brendan Higgins wrote: > +Kees Cook - I imagine you have already seen this, but I figured you > would be interested because of your recent work on the KUnit UAPI and > the mocking discussions. > +Dmitry Vyukov - This made me think of the syzkaller/KUnit experiments > we did a couple of years back - this would probably work a bit better. > > On Fri, Mar 11, 2022 at 11:24 AM Vincent Whitchurch > <vincent.whitchurch@axis.com> wrote: > > > > This patchset proposes roadtest, a device-driver testing framework. Drivers > > are tested under User Mode Linux (UML) and interact with mocked/modelled > > hardware. The tests and hardware models are written in Python, the former > > using Python's built-in unittest framework. > > Wow! This sounds awesome! I was hoping to get some kind of hardware > modeling with KUnit eventually. I did some experiments, but this looks > way more mature. Thank you for the comments! > > Drivers are tested via their userspace interfaces. The hardware models allow > > tests to inject values into registers and assert that drivers control the > > hardware in the right way and react as expected to stimuli. > > I already took a look at the documentation patch - I'll comment there > more in detail, but I like the hardware modelling and device tree > code; it seems very usable. > > > Roadtest is meant to be used for relatively simple drivers, such as the ones > > part of the IIO, regulator and RTC subsystems. > > Obviously for an initial version going after simple stuff makes sense, > but I would hope there is applicability to any driver stack > eventually. Yes, there is no inherent restriction to only simple hardware, but these kinds of subsystem are the ones where it's easier to apply the framework since there's simply less stuff to model/mock in the hardware. Supporting different busses also requires some work in the framework and potentially some new drivers. For I2C we use virtio-i2c but there's no ready-made virtio-spi for example. For MMIO (PCI / platform drivers), I did some basic experiments with UML's virtio-mmio in the early stages of writing this framework. > > = How does this relate to kselftests? > > > > Tests in kselftests also test kernel code using the userspace interfaces, but > > that's about what's common between the frameworks. kselftests has other goals > > and does not provide any kind of mechanism for hardware mocking. > > I had a question that after thinking about it; I think I know the > answer, so I am going to ask the question anyway and attempt to answer > it myself: > > I agree in regard to mocking, but why not use kselftest for driving > tests that check drivers from userspace? I believe there are other > kselftest tests implemented in Python, why can't you just run your > tests inside of kselftest? > > Now, I believe the answer to this question is that you need to control > spinning up your own kernel to run inside your test harness because > you need to control the environment that the kernel runs in - is this > correct? Yes, that is correct. For example, the devicetree stuff requires that the kernel be booted with the devicetree. For the other tests also it's simpler to have a controlled environment without being affected by other stuff going on on the host. And generally it's of course easier if the kernel which is inevitably going to crash and burn due to buggy drivers isn't the one powering your workstation. Also, there is no currently way to use virtio drivers such as virtio-i2c and virtio-gpio (which roadtest uses) with the virtio device side implemented in userspace on the same system, so that would have also required a fair bit of work to get running. (On a side note, I've wondered why kselftest doesn't provide a standard way to run all the tests under kvm or something similar with all the correct configs. For example, the kernels I work with are on embedded systems and I rarely recompile my host kernel, and I assume that there are plenty of others in the same situation.) > > = How does this relate to kunit? > > > > Kunit is for unit testing of functions in kernel code, and is not meant for > > testing kernel code via userspace interfaces. It could in theory be used to > > test some of the simple drivers too, but that would require (1) a large amount > > of mocking code in various kernel frameworks, and, more importantly, (2) > > refactoring of the drivers to be tested. > > I mostly agree, but I think there is something that is missing here: > so roadtest seems to depend on having a user interface to test a > driver - for a simple smoke test on a simple driver without a big > driver stack on top, that makes sense, but what about testing error > paths or a platform driver buried beneath a deep driver stack? I think > there is potential for a powerful combination using KUnit to test the > low level kernel API and using roadtest to mock the hardware > environment and provide configuration. Yes, that could be useful. I have previously written some kunit tests for some experimental memory management code which required different devicetree reserved-memory nodes and arm64 (no hardware mocking) to run, and I ran them by having a shell script which ran QEMU several times with appropriate -append kunit.filter_glob=foo and -dtb options and post-processing the logs with kunit.py. > I am imagining that we could have an in-kernel KUnit/roadtest API that > we can use to have an in-kernel test request changes to the > environment for creating error cases and the like that can be > validated by KUnit test cases. > > Going even further, I wonder if we could run kselftests inside of > roadtest since roadtest allows us to change the environment on the > fly. Sounds interesting, but I would likely need to see concrete examples to understand what kind of environment we'd want to change from within the kernel. > > This can be contrasted with roadtest which works with mostly unmodified drivers > > and which mocks the hardware at the lowest level without having to change > > kernel frameworks. > > I think that is both potentially an advantage and a disadvantage. > > The advantage is that your test is very general; roadtests would > likely be portable across kernel versions. > > The disadvantage is that you don't get as much code introspection: I > imagine roadtest is not as good as testing error paths for example. > > I also think that having to change code to make it more testable is > often an advantage as much as a disadvantage. Yes, that's true, but I highlighted the unmodified drivers bit because (1) the process of refactoring drivers which don't have tests to make them testable in itself carries it with a risk of breaking stuff, and (2) and there are simply so many existing drivers that it's very unlikely that most of them get refactored, but it should be relatively easy to, for example, add a regression test for a specific bug fix with roadtest. > > Still, I think that is a good set of tradeoffs for roadtest to make > when set against KUnit and kselftest since roadtest seems to fit in > where kselftest and KUnit are weak.
On Fri, 2022-03-11 at 17:24 +0100, Vincent Whitchurch wrote: > Import the libvhost-user from QEMU for use in the implementation of the > virtio devices in the roadtest backend. > So hm, I wonder if this is the sensible thing to do? Not that I mind importing qemu code, but: 1) the implementation is rather complex in some places, and has support for a LOT of virtio/vhost-user features that are really not needed in these cases, for performance etc. It's also close to 4k LOC. 2) the implementation doesn't support time-travel mode which might come in handy We have another implementation that might be simpler: https://github.com/linux-test-project/usfstl/blob/main/src/vhost.c but it probably has dependencies on other things in this library, but vhost.c itself is only ~1k LOC. (But I need to update it, I'm sure we have some unpublished bugfixes etc. in this code) johannes
On Thu, Mar 24, 2022 at 02:00:10PM +0100, Johannes Berg wrote: > On Fri, 2022-03-11 at 17:24 +0100, Vincent Whitchurch wrote: > > Import the libvhost-user from QEMU for use in the implementation of the > > virtio devices in the roadtest backend. > > So hm, I wonder if this is the sensible thing to do? > > Not that I mind importing qemu code, but: > > 1) the implementation is rather complex in some places, and has support > for a LOT of virtio/vhost-user features that are really not needed > in these cases, for performance etc. It's also close to 4k LOC. Is this really a problem given that the code is imported as-is? The intention is not to have to make a lot of local modifications to it in the kernel tree. The code is stable and presumably well-tested upstream, and upstream maintains it as a separate library (in the QEMU source tree though) to encourage reuse. > 2) the implementation doesn't support time-travel mode which might come > in handy True, but I don't see the external time-travel controller stuff being too useful for the kinds of tests this framework is targeting.
On Fri, 11 Mar 2022 17:24:35 +0100 Vincent Whitchurch <vincent.whitchurch@axis.com> wrote: > This patchset proposes roadtest, a device-driver testing framework. Drivers > are tested under User Mode Linux (UML) and interact with mocked/modelled > hardware. The tests and hardware models are written in Python, the former > using Python's built-in unittest framework. > > Drivers are tested via their userspace interfaces. The hardware models allow > tests to inject values into registers and assert that drivers control the > hardware in the right way and react as expected to stimuli. > > Roadtest is meant to be used for relatively simple drivers, such as the ones > part of the IIO, regulator and RTC subsystems. Hi All, Just wanted to very briefly report back on my experience of using this framework. Given I wanted a suitable job to try out it's usefulness when doing refactoring /development, I decided to tidy up one of the remaining IIO drivers in staging and see how things went in developing tests to hit the particular code I was modifying. At some point I might extend this to a more comprehensive test suite for that driver, but for now it does basic channel reading and a few other things + verifies some of the register state changes seen on the hardware side of things. Whilst my python could be said to be decidedly rusty (last time I recall writing some was for an intern project 20 years back), it was fairly easy to get something working using the docs in this series and the fine engineering tool of cut and paste. Road test worked very well. Was it easier than my existing hacked up QEMU board emulation that lets me instantiate minimal emulation pretty quickly? Pretty similar on balance but big advantage here is I'm not having to ask people to go fetch a tree and build QEMU just to sanity check the driver changes. Also note I'm doing a lot of QEMU work for the day job at the moment, so it's not really a fair comparison if the question is what would most kernel driver developers find useful. So for now I'll probably mix and match depending on the complexity of the device I'm emulating, but roadtest is definitely a good addition to the toolkit. Note this is putting aside all the advantages of having tests in tree and the much lighter amount of infrastructure needed to run those over QEMU CI. If anyone is curious patch set with tests and the staging graduation of the AD7746 CDC driver. https://lore.kernel.org/all/20220418192907.763933-18-jic23@kernel.org/ Great work Vincent. I'm looking forward to using this more. Thanks, Jonathan > > Questions and answers: > > = Why do we need this? > > There are a large amount of these kind of drivers in the kernel. Most of the > hardware is not available in current CI systems so most drivers can only, at > best, be build-tested there. Even basic soundness such as a driver > successfully probing and binding to the devices it tries to be support cannot > be tested. Drivers cannot be easily regression-tested to ensure that bugs > fixed once do not get reintroduced. > > Many drivers support multiple related hardware variants, and far from all patch > submitters have access to all the variants which the driver that they are > patching supports, so there is no way for them to easily verify that they > haven't broken something basic on a variant which they do not own. > > Furthermore, hardware can be used in many different configurations with drivers > supporting many different devicetree properties, so even just having access to > all the variants would be insufficient. > > On top of that, some of the chips measure environmental conditions such as > temperature, so testing extreme cases may not be simple even if one has access > to the hardware. > > All this makes development, modification, maintenance, and reviewing of these > drivers harder than it necessarily needs to be. Roadtest hopes to make some of > these things slightly easier by providing a framework to create hardware > models/mocks and to write testcases which exercise drivers using these models. > > = Do you have some specific examples of the kind of code this could be used to > test? > > Here is an example of a patch which can easily be regression-tested using > roadtest (in fact, this series includes such a regression test) but is much > harder to do so automatically with real hardware since it requires specific > environmental conditions: > > iio: light: opt3001: Fixed timeout error when 0 lux > https://lore.kernel.org/lkml/20210920125351.6569-1-valek@2n.cz/ > > Here is another example. This driver has code which correctly parses a > documented devicetree property (amstaos,proximity-diodes) but which then fails > to actually communicate this setting to the hardware in any way. Such code can > be easily tested with roadtest since the framework integrates devicetree > support and provides functions to assert that drivers writes expected registers > with expected values: > > drivers/iio/light/tsl2772.c tsl2772_read_prox_diodes() > > (Both the above examples happen to be from the same subsystem but that should > in no way be taken to imply that such issues are unique to that subsystem or > that that subsystem has more of them.) > > = How does this relate to kselftests? > > Tests in kselftests also test kernel code using the userspace interfaces, but > that's about what's common between the frameworks. kselftests has other goals > and does not provide any kind of mechanism for hardware mocking. > > = How does this relate to kunit? > > Kunit is for unit testing of functions in kernel code, and is not meant for > testing kernel code via userspace interfaces. It could in theory be used to > test some of the simple drivers too, but that would require (1) a large amount > of mocking code in various kernel frameworks, and, more importantly, (2) > refactoring of the drivers to be tested. > > This can be contrasted with roadtest which works with mostly unmodified drivers > and which mocks the hardware at the lowest level without having to change > kernel frameworks. > > = How do I use it? > > See Documentation/dev-tools/roadtest.rst added by the documentation patch for > more information about running and writing tests using this framework. > > = What's included in the patchset? > > The current framework allows developing tests for hardware which uses the I2C > bus. Hardware models can also control GPIOs and use them to trigger > interrupts. > > This series includes tests for some IIO, regulator and RTC drivers. The > regulator and RTC tests depend on a few driver patches which are either in > review or in linux-next. These are noted in the commit messages. > > The entire patch set, including the required dependencies, is also available in > a git tree: > > https://github.com/vwax/linux/commits/roadtest/rfc-v1 > > Cc: linux-kernel@vger.kernel.org > Cc: devicetree@vger.kernel.org > Cc: linux-um@lists.infradead.org > > Cc: shuah@kernel.org > Cc: brendanhiggins@google.com > Cc: linux-kselftest@vger.kernel.org > > Cc: jic23@kernel.org > Cc: linux-iio@vger.kernel.org > > Cc: lgirdwood@gmail.com > Cc: broonie@kernel.org > > Cc: a.zummo@towertech.it > Cc: alexandre.belloni@bootlin.com > Cc: linux-rtc@vger.kernel.org > > Cc: corbet@lwn.net > Cc: linux-doc@vger.kernel.org > > Vincent Whitchurch (10): > roadtest: import libvhost-user from QEMU > roadtest: add C backend > roadtest: add framework > roadtest: add base config > roadtest: add build files > roadtest: add documentation > iio: light: opt3001: add roadtest > iio: light: vcnl4000: add roadtest > regulator: tps62864: add roadtest > rtc: pcf8563: add roadtest > > Documentation/dev-tools/index.rst | 1 + > Documentation/dev-tools/roadtest.rst | 669 ++++ > tools/testing/roadtest/.gitignore | 2 + > tools/testing/roadtest/Dockerfile | 25 + > tools/testing/roadtest/Makefile | 84 + > tools/testing/roadtest/init.sh | 19 + > tools/testing/roadtest/pyproject.toml | 10 + > tools/testing/roadtest/requirements.txt | 4 + > tools/testing/roadtest/roadtest/__init__.py | 2 + > .../roadtest/roadtest/backend/__init__.py | 0 > .../roadtest/roadtest/backend/backend.py | 32 + > .../testing/roadtest/roadtest/backend/gpio.py | 111 + > .../testing/roadtest/roadtest/backend/i2c.py | 123 + > .../testing/roadtest/roadtest/backend/main.py | 13 + > .../testing/roadtest/roadtest/backend/mock.py | 20 + > .../roadtest/roadtest/backend/test_gpio.py | 98 + > .../roadtest/roadtest/backend/test_i2c.py | 84 + > .../testing/roadtest/roadtest/cmd/__init__.py | 0 > tools/testing/roadtest/roadtest/cmd/main.py | 146 + > tools/testing/roadtest/roadtest/cmd/remote.py | 48 + > .../roadtest/roadtest/core/__init__.py | 0 > .../testing/roadtest/roadtest/core/control.py | 52 + > .../roadtest/roadtest/core/devicetree.py | 155 + > .../roadtest/roadtest/core/hardware.py | 94 + > tools/testing/roadtest/roadtest/core/log.py | 42 + > .../testing/roadtest/roadtest/core/modules.py | 38 + > .../testing/roadtest/roadtest/core/opslog.py | 35 + > tools/testing/roadtest/roadtest/core/proxy.py | 48 + > tools/testing/roadtest/roadtest/core/suite.py | 286 ++ > tools/testing/roadtest/roadtest/core/sysfs.py | 77 + > .../roadtest/roadtest/core/test_control.py | 35 + > .../roadtest/roadtest/core/test_devicetree.py | 31 + > .../roadtest/roadtest/core/test_hardware.py | 41 + > .../roadtest/roadtest/core/test_log.py | 54 + > .../roadtest/roadtest/core/test_opslog.py | 27 + > .../roadtest/roadtest/tests/__init__.py | 0 > .../roadtest/roadtest/tests/base/config | 84 + > .../roadtest/roadtest/tests/iio/__init__.py | 0 > .../roadtest/roadtest/tests/iio/config | 1 + > .../roadtest/roadtest/tests/iio/iio.py | 112 + > .../roadtest/tests/iio/light/__init__.py | 0 > .../roadtest/roadtest/tests/iio/light/config | 2 + > .../roadtest/tests/iio/light/test_opt3001.py | 95 + > .../roadtest/tests/iio/light/test_vcnl4000.py | 132 + > .../roadtest/tests/iio/light/test_vcnl4010.py | 282 ++ > .../roadtest/tests/iio/light/test_vcnl4040.py | 104 + > .../roadtest/tests/iio/light/test_vcnl4200.py | 96 + > .../roadtest/tests/regulator/__init__.py | 0 > .../roadtest/roadtest/tests/regulator/config | 4 + > .../roadtest/tests/regulator/test_tps62864.py | 187 ++ > .../roadtest/roadtest/tests/rtc/__init__.py | 0 > .../roadtest/roadtest/tests/rtc/config | 1 + > .../roadtest/roadtest/tests/rtc/rtc.py | 73 + > .../roadtest/tests/rtc/test_pcf8563.py | 348 ++ > tools/testing/roadtest/src/.gitignore | 1 + > tools/testing/roadtest/src/backend.c | 884 +++++ > .../src/libvhost-user/include/atomic.h | 310 ++ > .../src/libvhost-user/libvhost-user.c | 2885 +++++++++++++++++ > .../src/libvhost-user/libvhost-user.h | 691 ++++ > 59 files changed, 8798 insertions(+) > create mode 100644 Documentation/dev-tools/roadtest.rst > create mode 100644 tools/testing/roadtest/.gitignore > create mode 100644 tools/testing/roadtest/Dockerfile > create mode 100644 tools/testing/roadtest/Makefile > create mode 100755 tools/testing/roadtest/init.sh > create mode 100644 tools/testing/roadtest/pyproject.toml > create mode 100644 tools/testing/roadtest/requirements.txt > create mode 100644 tools/testing/roadtest/roadtest/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/backend/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/backend/backend.py > create mode 100644 tools/testing/roadtest/roadtest/backend/gpio.py > create mode 100644 tools/testing/roadtest/roadtest/backend/i2c.py > create mode 100644 tools/testing/roadtest/roadtest/backend/main.py > create mode 100644 tools/testing/roadtest/roadtest/backend/mock.py > create mode 100644 tools/testing/roadtest/roadtest/backend/test_gpio.py > create mode 100644 tools/testing/roadtest/roadtest/backend/test_i2c.py > create mode 100644 tools/testing/roadtest/roadtest/cmd/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/cmd/main.py > create mode 100644 tools/testing/roadtest/roadtest/cmd/remote.py > create mode 100644 tools/testing/roadtest/roadtest/core/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/core/control.py > create mode 100644 tools/testing/roadtest/roadtest/core/devicetree.py > create mode 100644 tools/testing/roadtest/roadtest/core/hardware.py > create mode 100644 tools/testing/roadtest/roadtest/core/log.py > create mode 100644 tools/testing/roadtest/roadtest/core/modules.py > create mode 100644 tools/testing/roadtest/roadtest/core/opslog.py > create mode 100644 tools/testing/roadtest/roadtest/core/proxy.py > create mode 100644 tools/testing/roadtest/roadtest/core/suite.py > create mode 100644 tools/testing/roadtest/roadtest/core/sysfs.py > create mode 100644 tools/testing/roadtest/roadtest/core/test_control.py > create mode 100644 tools/testing/roadtest/roadtest/core/test_devicetree.py > create mode 100644 tools/testing/roadtest/roadtest/core/test_hardware.py > create mode 100644 tools/testing/roadtest/roadtest/core/test_log.py > create mode 100644 tools/testing/roadtest/roadtest/core/test_opslog.py > create mode 100644 tools/testing/roadtest/roadtest/tests/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/tests/base/config > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/config > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/iio.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/config > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/test_opt3001.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/test_vcnl4000.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/test_vcnl4010.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/test_vcnl4040.py > create mode 100644 tools/testing/roadtest/roadtest/tests/iio/light/test_vcnl4200.py > create mode 100644 tools/testing/roadtest/roadtest/tests/regulator/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/tests/regulator/config > create mode 100644 tools/testing/roadtest/roadtest/tests/regulator/test_tps62864.py > create mode 100644 tools/testing/roadtest/roadtest/tests/rtc/__init__.py > create mode 100644 tools/testing/roadtest/roadtest/tests/rtc/config > create mode 100644 tools/testing/roadtest/roadtest/tests/rtc/rtc.py > create mode 100644 tools/testing/roadtest/roadtest/tests/rtc/test_pcf8563.py > create mode 100644 tools/testing/roadtest/src/.gitignore > create mode 100644 tools/testing/roadtest/src/backend.c > create mode 100644 tools/testing/roadtest/src/libvhost-user/include/atomic.h > create mode 100644 tools/testing/roadtest/src/libvhost-user/libvhost-user.c > create mode 100644 tools/testing/roadtest/src/libvhost-user/libvhost-user.h >