mmc_test framework: failures at rtsx_usb

mmc_test framework: failures at rtsx_usb

[Matwey V. Kornilov]

Hello,


I've just run mmc_test suite on rtsx_usb mmc host controller module
with RTS5170 hardware attached.
Kernel version is 6.14.0. The following four test has been failed:

May 05 21:06:47 localhost.localdomain kernel: mmc0: Test case 15.
Proper xfer_size at write (start failure)...
May 05 21:06:47 localhost.localdomain kernel: mmc0: Result: FAILED
May 05 21:06:47 localhost.localdomain kernel: mmc0: Test case 16.
Proper xfer_size at read (start failure)...
May 05 21:06:57 localhost.localdomain kernel: mmc0: Result: FAILED
May 05 21:06:57 localhost.localdomain kernel: mmc0: Test case 17.
Proper xfer_size at write (midway failure)...
May 05 21:06:57 localhost.localdomain kernel: mmc0: Result: FAILED
May 05 21:06:57 localhost.localdomain kernel: mmc0: Test case 18.
Proper xfer_size at read (midway failure)...
May 05 21:07:08 localhost.localdomain kernel: mmc0: Result: FAILED

Should I normally report it or it is known not to work?


-- 
With best regards,
Matwey V. Kornilov

Re: mmc_test framework: failures at rtsx_usb

[Ulf Hansson]

+ Ricky

On Mon, 5 May 2025 at 20:56, Matwey V. Kornilov
<matwey.kornilov@gmail.com> wrote:
>
> Hello,
>
>
> I've just run mmc_test suite on rtsx_usb mmc host controller module
> with RTS5170 hardware attached.
> Kernel version is 6.14.0. The following four test has been failed:
>
> May 05 21:06:47 localhost.localdomain kernel: mmc0: Test case 15.
> Proper xfer_size at write (start failure)...
> May 05 21:06:47 localhost.localdomain kernel: mmc0: Result: FAILED
> May 05 21:06:47 localhost.localdomain kernel: mmc0: Test case 16.
> Proper xfer_size at read (start failure)...
> May 05 21:06:57 localhost.localdomain kernel: mmc0: Result: FAILED
> May 05 21:06:57 localhost.localdomain kernel: mmc0: Test case 17.
> Proper xfer_size at write (midway failure)...
> May 05 21:06:57 localhost.localdomain kernel: mmc0: Result: FAILED
> May 05 21:06:57 localhost.localdomain kernel: mmc0: Test case 18.
> Proper xfer_size at read (midway failure)...
> May 05 21:07:08 localhost.localdomain kernel: mmc0: Result: FAILED
>
> Should I normally report it or it is known not to work?
>

I have looped in Ricky Wu from Realtek, let's see if he has some
comments to this.

I personally haven't used mmc_test for rtsx_usb, so I can't tell if
these are real problems.

Kind regards
Uffe

RE: mmc_test framework: failures at rtsx_usb

[Ricky WU]

> On Mon, 5 May 2025 at 20:56, Matwey V. Kornilov
> <matwey.kornilov@gmail.com> wrote:
> >
> > Hello,
> >
> >
> > I've just run mmc_test suite on rtsx_usb mmc host controller module
> > with RTS5170 hardware attached.
> > Kernel version is 6.14.0. The following four test has been failed:
> >
> > May 05 21:06:47 localhost.localdomain kernel: mmc0: Test case 15.
> > Proper xfer_size at write (start failure)...
> > May 05 21:06:47 localhost.localdomain kernel: mmc0: Result: FAILED May
> > 05 21:06:47 localhost.localdomain kernel: mmc0: Test case 16.
> > Proper xfer_size at read (start failure)...
> > May 05 21:06:57 localhost.localdomain kernel: mmc0: Result: FAILED May
> > 05 21:06:57 localhost.localdomain kernel: mmc0: Test case 17.
> > Proper xfer_size at write (midway failure)...
> > May 05 21:06:57 localhost.localdomain kernel: mmc0: Result: FAILED May
> > 05 21:06:57 localhost.localdomain kernel: mmc0: Test case 18.
> > Proper xfer_size at read (midway failure)...
> > May 05 21:07:08 localhost.localdomain kernel: mmc0: Result: FAILED
> >
> > Should I normally report it or it is known not to work?
> >

Hi Matwey,

I have not used the mmc_test suite before. 
Could you provide me with your experimental procedure and steps?

Ricky

> 
> I have looped in Ricky Wu from Realtek, let's see if he has some comments to
> this.
> 
> I personally haven't used mmc_test for rtsx_usb, so I can't tell if these are real
> problems.
> 
> Kind regards
> Uffe

Re: mmc_test framework: failures at rtsx_usb

[Matwey V. Kornilov]

ср, 14 мая 2025 г. в 06:25, Ricky WU <ricky_wu@realtek.com>:
>
> > On Mon, 5 May 2025 at 20:56, Matwey V. Kornilov
> > <matwey.kornilov@gmail.com> wrote:
> > >
> > > Hello,
> > >
> > >
> > > I've just run mmc_test suite on rtsx_usb mmc host controller module
> > > with RTS5170 hardware attached.
> > > Kernel version is 6.14.0. The following four test has been failed:
> > >
> > > May 05 21:06:47 localhost.localdomain kernel: mmc0: Test case 15.
> > > Proper xfer_size at write (start failure)...
> > > May 05 21:06:47 localhost.localdomain kernel: mmc0: Result: FAILED May
> > > 05 21:06:47 localhost.localdomain kernel: mmc0: Test case 16.
> > > Proper xfer_size at read (start failure)...
> > > May 05 21:06:57 localhost.localdomain kernel: mmc0: Result: FAILED May
> > > 05 21:06:57 localhost.localdomain kernel: mmc0: Test case 17.
> > > Proper xfer_size at write (midway failure)...
> > > May 05 21:06:57 localhost.localdomain kernel: mmc0: Result: FAILED May
> > > 05 21:06:57 localhost.localdomain kernel: mmc0: Test case 18.
> > > Proper xfer_size at read (midway failure)...
> > > May 05 21:07:08 localhost.localdomain kernel: mmc0: Result: FAILED
> > >
> > > Should I normally report it or it is known not to work?
> > >
>
> Hi Matwey,
>
> I have not used the mmc_test suite before.
> Could you provide me with your experimental procedure and steps?
>
> Ricky
>

Hi,

I did as described in the documentation. First, the kernel was rebuilt
with CONFIG_MMC_TEST=m. I run this kernel in a KVM based virtual
machine with the USB device passed through inside. Then I attached an
microSD card (attention: the data will be lost, so this should be card
for testing) and did some magic:

modprobe mmc_test
echo 'mmc0:1234' > /sys/bus/mmc/drivers/mmcblk/unbind
echo 'mmc0:1234' > /sys/bus/mmc/drivers/mmc_test/bind
echo 0 > /sys/kernel/debug/mmc0/mmc0:1234/test

The latter means to run all the tests. The results are located in dmesg.

> >
> > I have looped in Ricky Wu from Realtek, let's see if he has some comments to
> > this.
> >
> > I personally haven't used mmc_test for rtsx_usb, so I can't tell if these are real
> > problems.
> >
> > Kind regards
> > Uffe



-- 
With best regards,
Matwey V. Kornilov