mbox series

[0/2] mmc: sdhci-acpi: Introduce device specific quirks, fix issues on 2 device models

Message ID 20200108093903.57620-1-hdegoede@redhat.com (mailing list archive)
Headers show
Series mmc: sdhci-acpi: Introduce device specific quirks, fix issues on 2 device models | expand

Message

Hans de Goede Jan. 8, 2020, 9:39 a.m. UTC
Hi Adrian,

I know you have sofar mostly resisted adding device specific (DMI based)
quirks to the sdhci-acpi.c driver and I agree with you that whenever
possible, those should be avoided.

But yesterday I was debugging an issue where using the microSD slot
causes the LCD panel of a tablet to go black. This turns out to be a
bug in the DSDT which gets triggered when using 1.8V modes, see the first
patch for details. In this case I really so no other option then disabling
1.8V modes and only doing so only on the affected device model.

Another issue which I had on my TODO list of things to fix is the Acer
SW5-012 version of the Acer Switch 10 models always reporting the microSD
as being write-protected. Here too I see no other option then a model
specific quirk, since some BYT devices may use a normal SD slot with
actual write-protect capabilities and we do not want to disable
write-protect checking everywhere just because it is broken on one model.
The workaround for this is the second patch in this series.

Regards,

Hans

Comments

Adrian Hunter Jan. 15, 2020, 12:51 p.m. UTC | #1
On 8/01/20 11:39 am, Hans de Goede wrote:
> Hi Adrian,
> 
> I know you have sofar mostly resisted adding device specific (DMI based)
> quirks to the sdhci-acpi.c driver and I agree with you that whenever
> possible, those should be avoided.

Not exactly.  I objected to matching against the string "UNKNOWN" because it
is insufficiently specific.

Matching specific DMI strings to identify board issues is fine.

> 
> But yesterday I was debugging an issue where using the microSD slot
> causes the LCD panel of a tablet to go black. This turns out to be a
> bug in the DSDT which gets triggered when using 1.8V modes, see the first
> patch for details. In this case I really so no other option then disabling
> 1.8V modes and only doing so only on the affected device model.
> 
> Another issue which I had on my TODO list of things to fix is the Acer
> SW5-012 version of the Acer Switch 10 models always reporting the microSD
> as being write-protected. Here too I see no other option then a model
> specific quirk, since some BYT devices may use a normal SD slot with
> actual write-protect capabilities and we do not want to disable
> write-protect checking everywhere just because it is broken on one model.
> The workaround for this is the second patch in this series.
> 
> Regards,
> 
> Hans
>