January 2017 - Devel - lists.sel4.systems

Questions about Memory Management in seL4
by Daniel Wang 19 May '23

19 May '23

Hi all, Sorry to bother, I have some basic questions about seL4 memory management. Can someone help me? 1. where is the kernel’s address space? I know there are basically three methods: a) separate virtual address space, e.g. through changing page tables on entry into privileged mode to access b) physical space, e.g. through disable automatic hardware translation of virtual addresses on entry into privilege mode c) privilege region in each process’s virtual address space, e.g. through page table to map kernel address into use-mode process. My question is which of this is being used by seL4 microkernel? 2. Objects like CSpace, VSpace, CNode, TCB, Endpoint, etc. are called kernel objects, but it seems all dynamically allocated in thread running/creating time. Are those objects (e.g. Figure 3.3 in seL4 manual) in kernel space or in user space? 3. I read that after the kernel boot up, it passes all resources, untyped memory to the first thread. Are those untyped memory physical memory or virtual memory in a single page? The reason I ask is I’m learning the code of the UNSW AOS project (two of the assignment projects are implementing frame and papers). I’m confused that before the implementation of frames and pages what kind of memory is passed to the the first thread? I saw int AOS project, it uses ut_table_init(), ut_steal_mem(), I wonder what kind of memory it operates on? Thanks -Dan

5 4

seL4 on Zedboard
by Brandon, Jeffrey - 0553 - MITLL 01 Feb '17

01 Feb '17

Hi, I've been trying to get seL4 running on a Zedboard and I was wondering if there was any documentation on how to boot sle4 from an SD card. I've read http://sel4.systems/pipermail/devel/2016-December/001160.html and https://sel4.systems/Info/Hardware/General/ which mentions using ${loadaddr} but that doesn't seem to be defined on the Zedboard's u-boot. So far I've compiled the image targeting the zynq7000 and used the mkimage command from the general hardware page to try u-boot's fastboot, but fastboot doesn't seem to be available on the zynq either. I am a bit of a novice here. I worry that the solution may be simple but I am currently taking shots in the dark. I wanted to ask this list for any resources you may be aware of for getting sel4 test cases running on the Zed. Thanks for the help, Jeff Brandon 05-53 Secure Resilient Systems and Technology

2 3

arm_data_abort_exception
by Wladislav Wiebe 31 Jan '17

31 Jan '17

Hello, has somebody experience with running into arm_data_abort_exception after enabling the MMU/paging and jumping to the kernel image: .. ELF-loader started on CPU: ARM Ltd. Cortex-A15 r2p4 paddr=[90000000..904b041f] ELF-loading image 'kernel' paddr=[80000000..80033fff] vaddr=[f0000000..f0033fff] virt_entry=f0000000 ELF-loading image 'sel4test-driver' paddr=[80034000..8042dfff] vaddr=[10000..409fff] virt_entry=1e924 Enabling MMU and paging Jumping to kernel-image entry point... check_data_abort_exception() DFAR = 0xfff01014 DFSR = 0x7 ADFSR = 0x0 abort() called. I've dumped the DFAR/DFSR register -- Relevant config parts about the setup: CONFIG_ARCH_ARM_V7A=y CONFIG_ARCH_ARM=y CONFIG_ARCH_AARCH32=y CONFIG_ARM_CORTEX_A15=y CONFIG_PLAT_KEYSTONE=y Thanks in advance! -- WBR, Wladislav WIebe

2 12

Exposing ticks to the rt API
by Corey Richardson 30 Jan '17

30 Jan '17

Hi Anna, While analyzing some time related APIs, I've noticed that the x86 rt kernel assumes that the tsc frequency is constant and well represented by the measurement made at init by comparing against the PIT. It also exposes only microseconds and uses the internal calibration to convert that to/from ticks. I'm concerned about this behavior. The tsc can have up to 200ppm error if uncalibrated and not temperature controlled, from what I'm able to determine. I'd rather have all the public APIs take ticks and allow user space to determine how their timing requirements best correspond to ticks. This came up when wanting to measure how much running time a process/thread has used. Ideally, it's kept as a running count of ticks. In the current API, one can only observe microseconds consumed, converted by a potentially lossy ticksToUs (lossy when the rate isn't exactly known or tick->us conversion isn't an exact integer multiplication). It'd be more useful for me to have tick count directly. I'm also somewhat concerned with using the PIT to estimate the TSC rate, as I can't find anything about expected error of the PIT, and any slop there will influence all future kernel behaviors until reboot. One potential alternative is the ACPI PM timer or the HPET. Those run at higher frequencies, which would help reduce the error bounds on the rate estimation. Let me know what you think! If it's useful at all, I've written up some stuff about the work that inspired these thoughts: https://gitlab.com/robigalia/meta/blob/master/blog/_drafts/2016-12-23-what-… Merry Christmas, -- cmr http://octayn.net/

4 4

strange code in thread.c
by Berkus 30 Jan '17

30 Jan '17

Hi! Details here https://github.com/seL4/seL4/issues/57 do you guys keep an eye on github activity actually or it’s a bad medium for communcation?

2 1

I2C Driver for BeagleBone (AM335x) for CAmkES
by Richard Habeeb 25 Jan '17

25 Jan '17

Hey all, Has anyone setup I2C on a Beaglebone CAmkES application? Is the code in libplatsupport all that's needed? The i2d_id enum doesn't seem to include all the i2c modules ( https://github.com/seL4/libplatsupport/blob/master/plat_include/am335x/plat…) I didn't see an init function for the i2c_bus for am335x; for example, exynos uses the exynos_i2c_init. Any help is appreciated. Thanks, Richard -- Richard Habeeb Research Assistant, Computer Science, USF PGP: https://goo.gl/VhmGGu https://habeebr.bitbucket.io/

1 0

ARM LPAE?
by Wladislav Wiebe 21 Jan '17

21 Jan '17

Hello, I am trying to integrate TI Keystone 2 platform to seL4. What I have done so far is basically an initial port of the Kernel/libs/elfloader/Platform components and introducing a new keystone defconfig. Relevant config parts about the setup: CONFIG_ARCH_ARM_V7A=y CONFIG_ARCH_ARM=y CONFIG_ARCH_AARCH32=y CONFIG_ARM_CORTEX_A15=y CONFIG_PLAT_KEYSTONE=y At the end I am able to generate an image which looks like: $ arm-cortexa15-linux-gnueabihf-readelf -e images/sel4test-driver-image-arm-keystone ELF Header: Class: ELF32 .. Entry point address: 0xc0008000 .. (I am using the same entry point as Linux use it) So far, it stops booting in the elfloader -- output log: .. [ 29.705] ## Starting application at 0xc0008000 ... ELF-loader started on CPU: ARM Ltd. Cortex-A15 r2p4 paddr=[c0008000..c032841f] -> Stop here <- Seems it stops loading @ elf_getMemoryBounds(). I suppose it is related to fact that the physical start address starts @ 0x840000000 which is beyond 32 Bit - so, LPAE support is required. The physical RAM setup looks like: bank 0: addr: 0x840000000 size: 0x17ec5000 (382 MB) bank 1: addr: 0x880000000 size: 0x80000000 (2048 MB) Do you have an idea if this is supposed to be supported on seL4? What should be the "physBase" and the "kernelBase" for such a setup? How should the "static const p_region_t BOOT_RODATA avail_p_regs" looks like? Thanks in advance! -- WBR, Wladislav WIebe

3 10

Re: [seL4] SDHC drivers (cont.)
by Siwei.Zhuang＠data61.csiro.au 20 Jan '17

20 Jan '17

Hi Oak, I tried your code. It works for me except that it seems that you are using old versions of seL4 kernel and libraries. Attached is the code I'm using, I merely add some initialization code to your code. It should work on the latest version of seL4 kernel and related libraries. If the code doesn't work for you. I would suggest, 1. If you are using u-boot, try its mmc command, see if you can access the card. 2. Have a look at the imx6 reference manual, chapter 67.2 and 67.3, check if the host controller is enabled properly. - Siwei ________________________________ From: Norrathep Rattanavipanon <nrattana(a)uci.edu> Sent: Friday, 20 January 2017 6:12 AM To: Zhuang, Siwei (Data61, Kensington NSW) Subject: Re: [seL4] SDHC drivers (cont.) Hi Siwei, Yes, I believe mmc_card_capacity returns the correct value (4026531840 - my microsd is 4GB). And my code does not do any interrupt handling. In fact, that's all the code in the root userspace process (besides initializing vka, vspace and simple objects) and there is no other process. I also tried using a callback as mmc_block_read's argument, instead of using NULL, but it also does not work. On Wed, Jan 18, 2017 at 8:41 PM, <Siwei.Zhuang(a)data61.csiro.au<mailto:Siwei.Zhuang@data61.csiro.au>> wrote: Hi Oak, Do you get the correct card capacity by calling mmc_card_capacity? If you don't, you'd need to check if the hardware is enabled in your bootloader. Are you doing any interrupt handling? As you don't supply a callback function to mmc_block_read, the read function would block until the hardware responses. In this case, mmc_block_read handles the interrupt itself. It would trigger a dead lock if you also handles the interrupt. - Siwei ________________________________ From: Devel <devel-bounces(a)sel4.systems> on behalf of Norrathep Rattanavipanon <nrattana(a)uci.edu<mailto:nrattana@uci.edu>> Sent: Thursday, 19 January 2017 9:54 AM To: Danis, Adrian (Data61, Kensington NSW) Cc: devel(a)sel4.systems; GTS Subject: Re: [seL4] SDHC drivers (cont.) Hi Adrian, Thank you for the reply. I fix it according to what you suggest but it still does not solve the problem (still hang at the same place. Do you have any other suggestions on what I should fix or how I should look into debugging it? Best, Oak On Tue, Jan 17, 2017 at 8:58 PM, <Adrian.Danis(a)data61.csiro.au<mailto:Adrian.Danis@data61.csiro.au>> wrote: Hi Oak, I don't know much about SD cards or this driver but it seems to me like passing it a physical address of 0 isn't what you want, and whilst I know nothing about the mmc hardware it seems reasonable to me that it might hang trying to DMA to memory that doesn't exist. There is also the additional problem that even if paddr was correct for the first page, there is no guarantee that the 5 pages you have allocated are contiguously physically even though they will be mapped contiguous virtually. My recommendation is to use the page dma allocator in seL4_libs/libsel4utils/include/sel4utils/page_dma.h, this will provide you an easy way to allocate/map pages for the purposes of DMA. So instead of void *vaddr = vspace_new_pages(&vspace, seL4_AllRights, 5, seL4_PageBits); uintptr_t paddr = 0; You would have ps_dma_man_t dma_man; error = sel4utils_new_page_dma_alloc(&vka, &vspace, &dma_man); void *vaddr = ps_dma_alloc(&dma_man, 5 * PAGE_SIZE_4K, PAGE_SIZE_4K, 0, PS_MEM_NORMAL); uintptr_t paddr = ps_dma_pin(&dma_man, vaddr, 5 * PAGE_SIZE_4K); As I said at the start, I don't know if this will actually fix your current problem, but it's still something you will need to fix. Adrian On Wed 18-Jan-2017 6:31 AM, Norrathep Rattanavipanon wrote: My apology that I have to continue the discussion from the old thread: (http://sel4.systems/pipermail/devel/2016-October/001056.html) My goal is to read and write data from/to micro sd in SabreLite platform. I tried both Adrain and Anna's suggestions in that thread and they dont seem to fix the problem. Nothing returns error and I also changed cacheable parameter in sel4utils_new_pages_at_vaddr to 0 (I believe this is what Anna meant instead of setting 3rd argument of vspace_new_pages). The code's still stuck inside mmc_block_read function. I tracked it down and it seems like BRR and BWR interrupt statuses are never set to 1 I add my code below, please let me know if anything is wrong with my code. ps_io_mapper_t io_mapper = {0}; error = sel4platsupport_new_io_mapper(simple, vspace, vka, &io_mapper); assert(error == 0); ps_io_ops_t io_ops = { .io_mapper = io_mapper }; sdio_host_dev_t* dev = (sdio_host_dev_t*) malloc(sizeof(*dev)); assert(dev != NULL); memset(dev, 0, sizeof(*dev)); enum sdio_id id = sdio_default_id(); // return id=3 error = sdio_init(id, &io_ops, dev); assert(error == 0); mmc_card_t* mmc_card = (mmc_card_t*) malloc(sizeof(*mmc_card)); error = mmc_init(dev, &io_ops, mmc_card); assert(error == 0 && mmc_card != NULL); void *vaddr = vspace_new_pages(&vspace, seL4_AllRights, 5, seL4_PageBits); assert(vaddr != NULL); uintptr_t paddr = 0; printf("mmc card capacity %llu bytes\n", mmc_card_capacity(*mmc_card)); long read_len = mmc_block_read(*mmc_card, 0x50000, 1, vaddr, paddr, NULL, NULL); // Stuck here printf("read %lu bytes\n", read_len); Thanks Oak -- Norrathep (Oak) Rattanavipanon M.S. in Computer Science University of California - Irvine _______________________________________________ Devel mailing list Devel(a)sel4.systems<mailto:Devel@sel4.systems> https://sel4.systems/lists/listinfo/devel -- Norrathep (Oak) Rattanavipanon M.S. in Computer Science University of California - Irvine -- Norrathep (Oak) Rattanavipanon M.S. in Computer Science University of California - Irvine

1 0

seL4 System Utilization / Benchmark
by Andrew Gacek 20 Jan '17

20 Jan '17

Hi, I'm interested in seeing how much of the CPU is being used by the various threads in my (camkes) application. I'd like a simple "Task Manager" style report of the form: 70% idle 25% can_obj_7_0_control_9_tcb 5% can_obj_7_Int_3_0000_tcb 0% can_obj_7_0_fault_handler_15_0000_tcb etc I've tried poking around libsel4bench, but I don't see any obvious way of benchmarking the whole system. From what I can tell, it's mostly designed to get cycle counts from specific code segments. Is that right? What's the best way to go about benchmarking the whole system? It's only for debugging, so I'm willing to make some quick-and-dirty changes to the kernel just to see the results. Thanks, Andrew

4 11

SDHC drivers (cont.)
by Norrathep Rattanavipanon 19 Jan '17

19 Jan '17

My apology that I have to continue the discussion from the old thread: ( http://sel4.systems/pipermail/devel/2016-October/001056.html) My goal is to read and write data from/to micro sd in SabreLite platform. I tried both Adrain and Anna's suggestions in that thread and they dont seem to fix the problem. Nothing returns error and I also changed cacheable parameter in sel4utils_new_pages_at_vaddr to 0 (I believe this is what Anna meant instead of setting 3rd argument of vspace_new_pages). The code's still stuck inside mmc_block_read function. I tracked it down and it seems like BRR and BWR interrupt statuses are never set to 1 I add my code below, please let me know if anything is wrong with my code. ps_io_mapper_t io_mapper = {0}; error = sel4platsupport_new_io_mapper(simple, vspace, vka, &io_mapper); assert(error == 0); ps_io_ops_t io_ops = { .io_mapper = io_mapper }; sdio_host_dev_t* dev = (sdio_host_dev_t*) malloc(sizeof(*dev)); assert(dev != NULL); memset(dev, 0, sizeof(*dev)); enum sdio_id id = sdio_default_id(); // return id=3 error = sdio_init(id, &io_ops, dev); assert(error == 0); mmc_card_t* mmc_card = (mmc_card_t*) malloc(sizeof(*mmc_card)); error = mmc_init(dev, &io_ops, mmc_card); assert(error == 0 && mmc_card != NULL); void *vaddr = vspace_new_pages(&vspace, seL4_AllRights, 5, seL4_PageBits); assert(vaddr != NULL); uintptr_t paddr = 0; printf("mmc card capacity %llu bytes\n", mmc_card_capacity(*mmc_card)); long read_len = mmc_block_read(*mmc_card, 0x50000, 1, vaddr, paddr, NULL, NULL); // Stuck here printf("read %lu bytes\n", read_len); Thanks Oak -- Norrathep (Oak) Rattanavipanon M.S. in Computer Science University of California - Irvine

3 3