On Thu, Aug 17, 2017 at 10:16:55PM +0200, Mark Kettenis wrote:
> If we attach a device to dwiic(4), we should mark it as attached such
> that we don't try to attach it again. Before this diff, I had:
>
> ihidev0 at iic6 addr 0x5c gpio 77, vendor 0x457 product 0x1133, SIS0457
> ...
> "SIS0457" at acpi0 not configured
>
> with this diff, only the first line shows up.
>
> ok?
Yes, thanks.
It looks like the King Jim Portabook shows the same bug:
ihidev0 at iic1 addr 0x34 gpio 5, vendor 0x8566 product 0x101, MSFT1234
...
"MSFT1234" at acpi0 not configured
> Index: dev/acpi/dwiic.c
> ===================================================================
> RCS file: /cvs/src/sys/dev/acpi/dwiic.c,v
> retrieving revision 1.23
> diff -u -p -r1.23 dwiic.c
> --- dev/acpi/dwiic.c 17 Aug 2017 05:16:27 -0000 1.23
> +++ dev/acpi/dwiic.c 17 Aug 2017 20:11:24 -0000
> @@ -647,8 +647,10 @@ dwiic_acpi_found_ihidev(struct dwiic_sof
> }
> ia.ia_intr = &crs;
>
> - if (config_found(sc->sc_iic, &ia, dwiic_i2c_print))
> + if (config_found(sc->sc_iic, &ia, dwiic_i2c_print)) {
> + node->parent->attached = 1;
> return 0;
> + }
>
> return 1;
> }
> @@ -678,8 +680,10 @@ dwiic_acpi_found_iatp(struct dwiic_softc
> }
> ia.ia_intr = &crs;
>
> - if (config_found(sc->sc_iic, &ia, dwiic_i2c_print))
> + if (config_found(sc->sc_iic, &ia, dwiic_i2c_print)) {
> + node->parent->attached = 1;
> return 0;
> + }
>
> return 1;
> }
