On 27.10.20 10:22, Jan Kiszka wrote: > On 27.10.20 02:25, Peng Fan wrote: >> Jan, >> >>> Subject: Re: [PATCH v2 00/46] arm64: Rework SMMUv2 support >>> >>> On 14.10.20 10:28, Jan Kiszka wrote: >>>> Changes in v2: >>>> - map 52-bit parange to 48 >>>> >>>> That wasn't the plan when I started, but the more I dug into the >>>> details and started to understand the hardware, the more issues I >>>> found and the more dead code fragments from the Linux usage became >>> visible. >>>> >>>> Highlights of the outcome: >>>> - Fix stall of SMMU due to unhandled stalled contexts (took me a while >>>> to understand that...) >>>> - Fix programming of CBn_TCR and TTBR >>>> - Fix TLB flush on cell exit >>>> - Fix bogus handling of Extended StreamID support >>>> - Do not pass-through unknown streams >>>> - Disable SMMU on shutdown >>>> - Reassign StreamIDs to the root cell >>>> - 225 insertions(+), 666 deletions(-) >>>> >>>> The code works as expected on the Ultra96-v2 here, but due to all the >>>> time that went into the rework, I had no chance to bring up my MX8QM >>>> so far. I'm fairly optimistic that things are not broken there as >>>> well, but if they are, bisecting should be rather simple with this >>>> series. So please test and review. >>>> >>> >>> Alice, Peng, already had a chance to review or test (ie. next)? >> >> I gave a test, sometimes I met SDHC ADMA error when >> `jailhouse enable imx8qm.cell`, sometimes it work well. >> >> I suspect when during jailhouse enable phase, there might be >> ongoing sdhc transactions not finished, not sure. >> >> I have not find time to look into details. >> >> Anyway, you could check in to master I think, we could address >> the issue later when I have time. >> > > Hmm, I would still like to understand this first... Do you have the > chance to bisect this effect to a commit? Otherwise, I guess I finally > need to get my board running. >
It's running now (quite some effort due to the incomplete upstream state - e.g. upstream u-boot runs but cannot boot all downstream kernels...), but I wasn't able to reproduce startup issues. Shutting down Jailhouse often hangs, though, at least restarting does all the time. And that even with next. Seems we still do not properly turn off/on something here. Interestingly, this issue was not present on the zynqmp. Jan -- Siemens AG, T RDA IOT Corporate Competence Center Embedded Linux -- 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/fd8fe9cc-9e69-4406-21e2-979b282b6d16%40siemens.com.
