July 2024 - Devel - lists.sel4.systems

Stuck during invalidate local TLB in Tx2
by Muneeswaran Rajendran 16 Oct '24

16 Oct '24

Hi seL4 Dev team, Can someone explain the reason (share the reference) to invalidate the local TLB during kernel vspace activation as part CPU initialisation. The Tx2 board stuck while performing this operation. As per my understanding both Tx1 and Tx2 are using same A57 core. so the TLB initialisation and invalidation remain same for Tx2 and should not stuck during invalidation.Please correct me if anything wrong here. I understand from that instruction (tlbi vmalle1) which is invalidating stage 1 entries associated to the current VMID. Also from arm mmu code the seL4 setting up (TCR_EL1) TCR register addressing space as 48bit, ASID as 16bit and page table setup 4k granular size. I am using MRS macro to dump the spsr_el1 and dspsr_el0 register to see the state of processor but its throwing synchronous exception. Do we have any other option to dump all the register content to see the state of processor like we have one in 'sel4debug_dump_registers' which is used in sel4test code. Regards, Muneeswaran R

3 4

Gpu passthrough
by Hugo V.C. 15 Aug '24

15 Aug '24

Hi all, here a simple question: is there any example/tutorial off gpu passthrough to a Linux guest? Best,

4 6

Build failure for sel4 on raspberry pi 5
by l4＠photiczone.com 14 Aug '24

14 Aug '24

Hi everyone, I am getting a link failure during the build of sel4test-tests: /usr/bin/ld: apps/sel4test-driver/musllibc/build-temp/stage/lib/libc.a(__stdio_exit.o): copy relocation against non-copyable protected symbol `__stdin_used' collect2: error: ld returned 1 exit status Searching t'interwebs seems to imply this is a gnarly problem caused by the nuances of various build tools - but one that should not apply to Arm. You can summarise everything I did before the build like this: sudo apt install ninja-build libxml2-utils repo pip3 install pyyaml typing hardware pyfdt jinja2 jsonschema six future ply python3-protobuf grpcio-tools cd && mkdir xxx && cd xxx Then to follow your build repo init -u https://github.com/seL4/sel4test-manifest.git repo sync mkdir build && cd build ../init-build.sh -DPLATFORM=bcm2711 ninja And yes I know it is not supported and the BCM2711 is for the Pi 4. I am evaluating ;) Any help greatly appreciated. Thanks, Smiley

3 4

I am interested in porting LionOS to X64 architecture
by Peter Relich 14 Aug '24

14 Aug '24

Hello, I am a researcher at Intellisense Systems, Inc. We are a small business that performs R&D for the U.S. Military. We have an active contractor to build a hypervisor for a microkernel. We recently selected Sel4 as the microkernel and I learned that the Army (our customer) used a camKes implementation of a VMM to host a fork of Ubuntu, but the limitations prompted them to try another approach. I think since that time, microkit has developed quite a bit and I saw the trustworthy systems has even been developing an operating system on the Sel4 microkernel. However, LionOS appears to be developed for ARM systems as the use case. For the Army use, we are using a Nuvo7000LP: https://www.neousys-tech.com/Resource/Product_Document/Nuvo-7000LP/Nuvo-700… Which is an X64 architecture. I just started learning microkit through your available tutorials and I am aware that building a comprehensive VMM is no easy task. However, I was wondering if you had some intuition or if you could provide me an estimate of the libraries in the LionOS that I would need to adjust/redesign to make an X64 architecture fork? Is this even a feasible approach? It looks like I just need to swap out the aarch64 libraries with compatible x64 libraries and do some testing. Before I commit, I just wanted to get some advice on the endeavor. Is forking LionOS to an X64 architecture feasible, or should I start at a lower level (i.e., redesigning the vmm?) Thanks so much for your time. Respectfully, Peter Relich Research Scientist Intellisense Systems, Inc. ________________________________ The information contained in this e-mail and any attachments from Intellisense Systems may contain confidential and/or proprietary information, and is intended only for the named recipient to whom it was originally addressed. If you are not the intended recipient, any disclosure, distribution, or copying of this e-mail or its attachments is strictly prohibited. If you have received this e-mail in error, please notify the sender immediately by return e-mail and permanently delete the e-mail and any attachments.

5 5

MCS round-robin scheduling not behaving as expected
by liam.vervecken＠gmail.com 12 Aug '24

12 Aug '24

Hello, I'm trying to understand MCS and how the scheduling works but I'm running into some issues. My setup: RISCV on BeagleVFire in sel4test I have configured 2 threads: seL4_SchedControl_ConfigureFlags(info->schedcontrol.start, sched_context_thread2, 1000000, 1000000, 0, 0, 0); seL4_SchedContext_Bind(sched_context_thread2, thread2_TCB); seL4_SchedControl_ConfigureFlags(info->schedcontrol.start, sched_context_thread3, 1000000, 1000000, 0, 0, 0); seL4_SchedContext_Bind(sched_context_thread3, thread3_TCB); The threads have the same priority. From the documentation I was expecting them to behave like this: - Thread1 does something - Thread2 does something - Thread1 does something - ... But it seems instead only the first thread gets scheduled, and it never allows the second thread to get any CPU time. Am I correct in assuming that 2 threads with period=budget and equal priority should be scheduled round-robin? I found that if I call seL4_SchedContext_Consumed(sched_context_thread) periodically in the threads they will switch if the budget has expired. It doesn't make much sense to me why this is the case, I thought they would switch automatically once their budget is expired and that seL4_SchedContext_Consumed() was simply used to keep track of this. I'm hoping it can offer some clue as to what I'm doing wrong. Any help or guidance would be greatly appreciated. Kind Regards, Liam

6 6

How do I reduce the size of the seL4 executable?
by liam.vervecken＠gmail.com 26 Jul '24

26 Jul '24

Hello, I'm trying to compile a simple 'hello world' sel4 program for the polarfire (riscv) and I want the executable to be as small as possible. According to the FAQ (https://docs.sel4.systems/projects/sel4/frequently-asked-questions.html) the single core kernel can compile to a <200kb executable but in my testing, I always end up with an image of approx 2MB. I worked with this (https://github.com/mskordal/SeL4-hello-world) example project and changed main.c to work with seL4_DebugPutChar instead of printf. Besides that, I added opensbi to the tools folder and slightly modified line 32 of projects/hello/CMakeList.txt to target_link_libraries(hello sel4 sel4_autoconf sel4runtime sel4platsupport). After running ../init-build.sh -DPLATFORM=polarfire and ninja, I end up with an executable of about 2MB. I'm guessing musllibc, util_libs and sel4_libs contribute to the size, but if I change my cmakelist file to try to get rid of them the build will fail. I was hoping since my hello world only uses sel4 APIs I could compile it without musllibc, util_libs and sel4_libs but I haven't been able to figure out how. Any help or tips would be greatly appreciated. Kind Regards, Liam

1 0

next seL4 TSC meeting on Thu 25 Jul
by Gerwin Klein 24 Jul '24

24 Jul '24

The Technical Steering Committee (TSC) of the seL4 Foundation will hold its next public meting on: - Sydney: Thu, 25 Jul, 4pm - Central Europe: Thu, 25 Jul, 8am - US West Coast (LA/SF): Thu, 25 Jul, 0:00h See also the seL4 calendar feed available from here: https://sel4.systems/contact/home.pml The meeting is by Zoom, and you can join from here: https://unsw.zoom.us/j/89949317341?pwd=JSC976ciVYmbDJByjSFH0chh78sCsg.1 Anyone interested can join the meeting and listen to the discussion and request to speak, which will be granted if time permits. Only TSC members can vote. The seL4 code of conduct applies to the meeting [1]. The agenda for the meeting so far is: - action items from last meeting - discuss new FPU RFC: https://github.com/seL4/rfcs/pull/26 - hear about progress on PMU, budget limit, device driver framework - make decision on old RFC-4: https://github.com/seL4/rfcs/pull/12 - any feedback on GitHub RFC implementation - bump version of cmake-format (will break current formatting) to reduce python dependency hell If there are other items that I missed or you think should be on the agenda, please let me know. If we can't get to an item this time, we can still schedule another meeting. Cheers, Gerwin (TSC chair) [1]: https://docs.sel4.systems/processes/conduct.html

3 6

seL4 developer hangout reminder
by Birgit Brecknell 22 Jul '24

22 Jul '24

Hi all A friendly reminder that the seL4 developer hangout is on again this week. Wed 24 Jul, 8:00am UTC. For your local date and time please see https://sel4.systems/contact/. Zoom link: https://unsw.zoom.us/j/82640784431 Birgit Brecknell seL4 Foundation Project Coordinator Sydney, Australia Mon 9-5 Wed 2-5 Fri 9-5 birgit(a)sel4.systems <mailto:birgit@sel4.systems> bbrcknl(a)gmail.com <mailto:bbrcknl@gmail.com>

1 0

Announcing beta.sel4.systems running on LionsOS
by James Archer 11 Jul '24

11 Jul '24

Trustworthy Systems is now serving a copy of the seL4.systems website from a beta LionsOS-based[1] webserver. It can be accessed at https://beta.sel4.systems/. The source code can be found in the LionsOS repository[2]. More information about the implementation can be found at https://beta.sel4.systems/Info/Website/. Any feedback or attempts to break it are welcome. [1] https://lionsos.org/ [2] https://github.com/au-ts/lionsos James

3 3

seL4_X86_Page_Map on seL4_X86_LargePageObject != 2MiB
by Ben McCart 10 Jul '24

10 Jul '24

I have created my own project on the PC99 architecture (x64) and have been working through the tutorial source (copy & pasted into my project) to learn and then prototype with seL4. I have done the mapping tutorial (pc99) and utilized that code within my project and everything seems to work just fine, but if I change out the seL4_X86_4K to seL4_X86_LargePageObject I don't end up with 2MiB of mappable virtual space. All the seL4 system calls through the client API return seL4_NoError. The seL4_LargePageBits value is 21 (2^21 == 2MB, and intel documentation for IA-32e shows that 2 MiB should be the large page size. However, if I address memory outside of the first 256 KiB it halts. The following works: seL4_Word *x = (seL4_Word *) TEST_VADDR; printf("Read x: %lu\n", *x); And this works: seL4_Word *x = (seL4_Word *) TEST_VADDR + 0x0003FFF0; printf("Read x: %lu\n", *x); But this halts when trying to dereference x to print it: seL4_Word *x = (seL4_Word *) TEST_VADDR + 0x00040000; printf("Read x: %lu\n", *x); I am running seL4 12.1 using QEMU with the following command line: qemu-system-x86_64 -cpu Nehalem,-vme,+pdpe1gb,-fsgsbase,-invpcid,+syscall,+lm,enforce -nographic -serial mon:stdio -m size=512M -s -S -kernel ../sel4-project/build/images/kernel-x86_64-pc99 -initrd build/cxxkit Does anybody have any idea why I might be having this issue and what I may do about it? -Ben McCart

2 2