On 17.02.21 08:14, Sainz Markel wrote: > On 12.02.21 10:05, Sainz Markel wrote: >>> >>> -----Mensaje original----- >>> De: Jan Kiszka <[email protected]> Enviado el: miércoles, 10 de >>> febrero de 2021 14:33 >>> Para: Sainz Markel <[email protected]>; [email protected] >>> Asunto: Re: Overflow in MMIO registration >>> >>> CAUTION: This email originated from outside of the organization. Do not >>> click links or open attachments unless you recognize the sender and know >>> the content is safe. >>> >>> >>> On 10.02.21 14:31, Jan Kiszka wrote: >>>> On 10.02.21 12:15, Sainz Markel wrote: >>>>> On 08.02.21 12:24, Kizska Jan wrote: >>>>> >>>>>> Something may have caused a subpage MMIO region to be created >>>>>> (page-unaligned memory region). If that region split-up didn't exist in >>>>>> the root cell already, things will fail. But I suspect that subpage is >>>>>> rather a symptom, not the >cause. >>>>> >>>>>> Where there any compiler warnings when building the config? I would also >>>>>> recommend checking our configs via "jailhouse config check root.cell >>>>>> cell1.cell cell2.cell ...". Or share what you created. >>>>> >>>>>> Jan >>>>> >>>>>> -- >>>>>> Siemens AG, T RDA IOT >>>>>> Corporate Competence Center Embedded Linux >>>>> >>>>> There are no warning during compilation and jailhouse config check >>>>> command says it is everything correct. >>>>> >>>>> #jailhouse config check sysconfig.cell core1-cell.cell core2-cell.cell >>>>> core3-cell.cell core4-cell.cell core5-cell.cell >>>>> Reading configuration set: >>>>> Root cell: RootCell (sysconfig.cell) >>>>> Non-root cell: core1-cell (core1-cell.cell) >>>>> Non-root cell: core2-cell (core2-cell.cell) >>>>> Non-root cell: core3-cell (core3-cell.cell) >>>>> Non-root cell: core4-cell (core4-cell.cell) >>>>> Non-root cell: core5-cell (core5-cell.cell) Overlapping memory >>>>> regions inside cell: None Overlapping memory regions with hypervisor: >>>>> None Missing PCI MMCONFIG interceptions: None Missing IOMMU >>>>> interceptions: None Missing resource interceptions for architecture >>>>> x86: None >>>>> >>>>> Attached to the mail the cell files I'm using. The .bin file is the same >>>>> as for apic-demo. >>>>> >>>> >>>> Just looking at core1-cell.c reveals that you are neither aligning on >>>> physical page boundaries for the base nor using page-aligned sizes. >>>> So it is like I predicted. And it is NOT a simple copy of apic-demo.c. >>>> >>> >>>> ...and we should likely enhance jailhouse-config-check to catch also this >>>> case. >>> >>> God, don't know what was thinking at... That's working now, cell starts >>> correctly. I can't see any output in the console, what did you mean in a >>> previous main with " Hope you do not let all cells use the same UART >>> console in this duplication (PIO 0x3f8 etc.)." I'm now just creating a >>> single cell and no output is being shown... >>> > >> I meant that you cannot hand out the single UART to multiple cells at the >> same time. There is no virtualization of this resource. You can >use it, >> obviously, to bring up one instance of a cell and then disable it for that >> cell again. > >> Connection to the cell should go via ivshmem: ivshmem-net or >> virtio-over-ivshmem or something custom. > > After having modified the .c files to align sizes and boundaries, apparently > cell starts with no issues. I understood what you said about using the same > UART interface among all cell files and that it won't work. ATM I'm just > starting a single cell and no console output is being displayed. What am I > doing wrong? Attached the updates .c files. > >
Systematic debugging is needed then: You can get the apic-demo in its original form work and printing output? Then compare to all your changes, revert them as long as it's not working. You will likely end up at the relocated comm region, maybe more. 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/41084eda-f26d-5538-f2fe-568cba2285fd%40siemens.com.
