[PATCH v1 1/2] arm64: dts: freescale: imx95-toradex-osm-dev: Remove ETH_A PHY interrupt

[Frank Li]

On Thu, Oct 15, 2026 at 09:37:19AM +0200, Francesco Dolcini wrote:
> Hello Frank,
>
> On Wed, Oct 14, 2026 at 09:20:23AM -0500, Frank Li wrote:
> > On Wed, Oct 14, 2026 at 09:58:09AM +0200, Francesco Dolcini wrote:
> > > From: Francesco Dolcini <francesco.dolcini at toradex.com>
> > >
> > > GPIO_A7 is a boot strap on the OSM iMX95 module. The pull-up on the OSM
> > > Development Board Ethernet PHY interrupt line causes the module to enter
> > > USB loader mode instead of booting normally when the reset button is
> > > held for more than 0.5 seconds.
> > >
> > > The PHY interrupt connection has been removed from the carrier board,
> >
> > Look like it is hardware modification, can declear the version number here
> >
> > Since V_x.y, the PHY interrupt connection has been removed...
>
> Valid point. It deserves a better explanation.
>
>  1. There is no change in the actual board revision.
>  2. Moving from interrupt to polling works always, so it does not
>     require any change on the board.
>  3. On our board is possible to disconnect the interrupt just removing a
>     jumper. The actual board revision stay the same. This is a new
>     product, that is going to be generally available in a few days. We
>     are going to ship boards with such jumper removed to
>     solve the i.MX95 strap option issue on reset.

Please add this information at commit message, otherwise, it will be strange,
why wire trace suddently route to HEADER from irq line.

Frank

>  4. Future board revisions might completely disconnect the ethernet
>     interrupt, removing the actual connection from the schematics. Given
>     we work in polling (after this patch is merged), this will be a
>     transparent change that will not create any kind of software
>     incompatibility.
>
>
> Should I incorporate this detailed information in the commit message and
> send a v2 or the current commit message is fine ?
>
> Francesco
>