Re: git: c2e9c5bbf022 - main - tpm: Refactor TIS and add a SPI attachment
- In reply to: Warner Losh : "Re: git: c2e9c5bbf022 - main - tpm: Refactor TIS and add a SPI attachment"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Tue, 25 Jun 2024 13:15:59 UTC
On Mon, 24 Jun 2024 16:07:56 -0600 Warner Losh <imp@bsdimp.com> wrote: > On Mon, Jun 24, 2024, 3:16 PM Gleb Smirnoff <glebius@freebsd.org> > wrote: > > > On Mon, Jun 24, 2024 at 04:11:28PM -0400, Justin Hibbits wrote: > > J> > sorry for late notice, but we just discovered that after this > > J> > commit tpm(4) on ACPI no longer attaches. From what I see > > J> > ther probe method isn't even called. > > J> > > J> Hi Gleb, > > J> > > J> That's very odd. The probe method hasn't changed, nor has the > > J> tpm_acpi attachment (still acpi), so I'm really confused why it > > J> wouldn't be called now. Is there any debugging you can do to > > J> further narrow it down? I haven't seen any issues reported > > J> internally at Juniper on x86 hardware. I just verified > > J> everything should get built properly. > > > > I started adding printfs to the probe method and as said - it is not > > even called. So I puzzled on where to start the debugging. I can > > test patches/ideas if you have any. > > > > Warner is also looking into this problem. > > > > I'm looking at name mismatched and/or moved parents to see if that's > in play. I thought I'd found something like this earlier, but it was > a false alarm I think. > > Warner > > > I just saw bbecd3148abf. Sorry for the breakage, thanks for fixing it! - Justin