On 18/04/2021, Jan Kiszka <[email protected]> wrote: > On 23.03.21 17:26, Angelo Ruocco wrote: >> On 23/03/2021, Jan Kiszka <[email protected]> wrote: >>> On 23.03.21 16:36, Angelo Ruocco wrote: >>>> On 23/03/2021, Jan Kiszka <[email protected]> wrote: >>>>> On 23.03.21 12:33, Angelo Ruocco wrote: >>>>>> On 23/03/2021, Jan Kiszka <[email protected]> wrote: >>>>>>> On 23.03.21 11:17, Angelo Ruocco wrote: >>>>>>>> Hi Jan, all, >>>>>>>> >>>>>>>> I have seen that arm-zero-exits has been merged into next. >>>>>>>> >>>>>>>> My understanding is that right now the only officially supported >>>>>>>> boards, meaning supporting both jailhouse and SDEI, are NXP's ones, >>>>>>>> with only a experimental patch for the xilinx zynqmp firmware to >>>>>>>> enable SDEI. >>>>>>>> >>>>>>>> Regarding that patch, I have looked at it, tested it and added the >>>>>>>> missing non-secure-boundary check. >>>>>>>> >>>>>>>> Are you thinking of submitting it upstream? If so, how do you want >>>>>>>> to >>>>>>>> proceed? >>>>>>> >>>>>>> If you have everything ready with the patch, I would not mind if you >>>>>>> pushed that upstream. Just CC me. >>>>>> >>>>>> The only thing missing is a proper commit message, I could also write >>>>>> that, but I don't really know whether it's appropriate for me to >>>>>> submit a patch of which I don't have the authorship. What do you >>>>>> think? >>>>> >>>>> But you added something to it as you wrote above, no? >>>> >>>> Yes, but that's hardly worth adding my signature to the patch, not to >>>> mention authorship. >>>> >>>>> Otherwise, share what you have, and I will kick off the gerrit mess. >>>> >>>> That is ok by me. How do you want me to send the patch? Here in the >>>> mailing list? >>>> >>> >>> Yes, just drop it here, attached or inline. >>> >>> Jan >>> >>> -- >>> Siemens AG, T RDA IOT >>> Corporate Competence Center Embedded Linux >>> >> >> >> You'll find the patch attached. >> > > Thanks for that, I've finally found the time to finish it and create > > https://review.trustedfirmware.org/c/TF-A/trusted-firmware-a/+/9673 > > I'm not getting you on CC there, likely because gerrit hates me (and > vice versa).
Hi Jan, Thank you for the update. > BTW, how did you address > https://lists.trustedfirmware.org/pipermail/tf-a/2021-April/001093.html > in your setup? Took me a while to understand this, at least with 5.10 > (wasn't happening with 5.4? didn't try). With DEBUG/SPDs enabled I am not having any issue on booting Linux (5.4), I had some random errors using psci calls (cores weren't waking up/suspending) under jailhouse, probably jailhouse or Linux at some point were overriding those ATF functions, (as you found out, in DEBUG mode that region isn't protected in any way). I Just removed the region from the root-cell configuration, and the problem went away. You're right though, even though I never experienced issues, with jailhouse a reservation in the dtb is the better way of handling this. I don't know if it's possible to do it "dinamically", i.e. to have a universal dtb and let Linux decide whether it can or cannot use that region. To be honest, sizing less than a megabyte, I never really cared much and just reserved it for ATF even when not in DEBUG mode or with SPDs enabled... -- Angelo > Jan > -- You received this message because you are subscribed to the Google Groups "Jailhouse" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/jailhouse-dev/CADiTV-2Sqqce9me0Myp1T9FP0KRg0cOoUDQXekrereqXiVpF1w%40mail.gmail.com.
