Patch Set 2:

Patch Set 2:

Yes, the question is why it is fixed. Is this because there's an inversion being done somewhere where it shouldn't?

We should dump the the gpio config for the one in question and compare w/ the coreboot settings vs what the kerne's pinmux might be doing.

every cyan variant has the jack detect GPIO defined as 95, PCIE_CLKREQ3B/AUDIO_CODEC_IRQ,
and every one sets it as a GPI, edge triggered both. The kernel cherrytrail pinmux driver actually skips over 95 since it's only listed as having a single/primary function (PCIE_CLKREQ3B)

Given that, 'ActiveBoth' would seem to be the correct setting for both the codec and LPE device. I'll test and adjust the patch if that works here

Nice analysis. Can you add that to the commit message too?

View Change

2 comments:

• Commit Message:

  • Patch Set #4, Line 16: Cherrytrail

    Cherry Trail

  • Patch Set #4, Line 21: Linux

    What version?

To view, visit change 24989. To unsubscribe, or for help writing mail filters, visit settings.